Merge drm/drm-next into drm-intel-next-queued

We need a backmerge to get DP_DPCD_REV_14 before we push other
i915 changes to dinq that could break compilation.

Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

1340 files changed, 76955 insertions, 33017 deletions

diff --git a/drivers/gpu/drm/Kconfig b/drivers/gpu/drm/Kconfig
index 757825ac60df..cb88528e7b10 100644
--- a/drivers/gpu/drm/Kconfig
+++ b/drivers/gpu/drm/Kconfig
@@ -49,16 +49,17 @@ config DRM_DEBUG_MM
 
 	  If in doubt, say "N".
 
-config DRM_DEBUG_MM_SELFTEST
-	tristate "kselftests for DRM range manager (struct drm_mm)"
+config DRM_DEBUG_SELFTEST
+	tristate "kselftests for DRM"
 	depends on DRM
 	depends on DEBUG_KERNEL
 	select PRIME_NUMBERS
 	select DRM_LIB_RANDOM
+	select DRM_KMS_HELPER
 	default n
 	help
-	  This option provides a kernel module that can be used to test
-	  the DRM range manager (drm_mm) and its API. This option is not
+	  This option provides kernel modules that can be used to run
+	  various selftests on parts of the DRM api. This option is not
 	  useful for distributions or general kernels, but only for kernel
 	  developers working on DRM and associated drivers.
 
@@ -121,6 +122,16 @@ config DRM_LOAD_EDID_FIRMWARE
 	  default case is N. Details and instructions how to build your own
 	  EDID data are given in Documentation/EDID/HOWTO.txt.
 
+config DRM_DP_CEC
+	bool "Enable DisplayPort CEC-Tunneling-over-AUX HDMI support"
+	select CEC_CORE
+	help
+	  Choose this option if you want to enable HDMI CEC support for
+	  DisplayPort/USB-C to HDMI adapters.
+
+	  Note: not all adapters support this feature, and even for those
+	  that do support this they often do not hook up the CEC pin.
+
 config DRM_TTM
 	tristate
 	depends on DRM && MMU
@@ -212,6 +223,17 @@ config DRM_VGEM
 	  as used by Mesa's software renderer for enhanced performance.
 	  If M is selected the module will be called vgem.
 
+config DRM_VKMS
+	tristate "Virtual KMS (EXPERIMENTAL)"
+	depends on DRM
+	select DRM_KMS_HELPER
+	default n
+	help
+	  Virtual Kernel Mode-Setting (VKMS) is used for testing or for
+	  running GPU in a headless machines. Choose this option to get
+	  a VKMS.
+
+	  If M is selected the module will be called vkms.
 
 source "drivers/gpu/drm/exynos/Kconfig"
 
@@ -267,6 +289,8 @@ source "drivers/gpu/drm/amd/amdkfd/Kconfig"
 
 source "drivers/gpu/drm/imx/Kconfig"
 
+source "drivers/gpu/drm/v3d/Kconfig"
+
 source "drivers/gpu/drm/vc4/Kconfig"
 
 source "drivers/gpu/drm/etnaviv/Kconfig"
diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile
index 9d66657ea117..a6771cef85e2 100644
--- a/drivers/gpu/drm/Makefile
+++ b/drivers/gpu/drm/Makefile
@@ -18,7 +18,7 @@ drm-y       :=	drm_auth.o drm_bufs.o drm_cache.o \
 		drm_encoder.o drm_mode_object.o drm_property.o \
 		drm_plane.o drm_color_mgmt.o drm_print.o \
 		drm_dumb_buffers.o drm_mode_config.o drm_vblank.o \
-		drm_syncobj.o drm_lease.o
+		drm_syncobj.o drm_lease.o drm_writeback.o drm_client.o
 
 drm-$(CONFIG_DRM_LIB_RANDOM) += lib/drm_random.o
 drm-$(CONFIG_DRM_VM) += drm_vm.o
@@ -41,9 +41,10 @@ drm_kms_helper-$(CONFIG_DRM_PANEL_BRIDGE) += bridge/panel.o
 drm_kms_helper-$(CONFIG_DRM_FBDEV_EMULATION) += drm_fb_helper.o
 drm_kms_helper-$(CONFIG_DRM_KMS_CMA_HELPER) += drm_fb_cma_helper.o
 drm_kms_helper-$(CONFIG_DRM_DP_AUX_CHARDEV) += drm_dp_aux_dev.o
+drm_kms_helper-$(CONFIG_DRM_DP_CEC) += drm_dp_cec.o
 
 obj-$(CONFIG_DRM_KMS_HELPER) += drm_kms_helper.o
-obj-$(CONFIG_DRM_DEBUG_MM_SELFTEST) += selftests/
+obj-$(CONFIG_DRM_DEBUG_SELFTEST) += selftests/
 
 obj-$(CONFIG_DRM)	+= drm.o
 obj-$(CONFIG_DRM_MIPI_DSI) += drm_mipi_dsi.o
@@ -61,6 +62,7 @@ obj-$(CONFIG_DRM_MGA)	+= mga/
 obj-$(CONFIG_DRM_I810)	+= i810/
 obj-$(CONFIG_DRM_I915)	+= i915/
 obj-$(CONFIG_DRM_MGAG200) += mgag200/
+obj-$(CONFIG_DRM_V3D)  += v3d/
 obj-$(CONFIG_DRM_VC4)  += vc4/
 obj-$(CONFIG_DRM_CIRRUS_QEMU) += cirrus/
 obj-$(CONFIG_DRM_SIS)   += sis/
@@ -68,6 +70,7 @@ obj-$(CONFIG_DRM_SAVAGE)+= savage/
 obj-$(CONFIG_DRM_VMWGFX)+= vmwgfx/
 obj-$(CONFIG_DRM_VIA)	+=via/
 obj-$(CONFIG_DRM_VGEM)	+= vgem/
+obj-$(CONFIG_DRM_VKMS)	+= vkms/
 obj-$(CONFIG_DRM_NOUVEAU) +=nouveau/
 obj-$(CONFIG_DRM_EXYNOS) +=exynos/
 obj-$(CONFIG_DRM_ROCKCHIP) +=rockchip/
diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile
index 2ca2b5154d52..bfd332c95b61 100644
--- a/drivers/gpu/drm/amd/amdgpu/Makefile
+++ b/drivers/gpu/drm/amd/amdgpu/Makefile
@@ -56,13 +56,18 @@ amdgpu-y += amdgpu_device.o amdgpu_kms.o \
 
 # add asic specific block
 amdgpu-$(CONFIG_DRM_AMDGPU_CIK)+= cik.o cik_ih.o kv_smc.o kv_dpm.o \
-	ci_smc.o ci_dpm.o dce_v8_0.o gfx_v7_0.o cik_sdma.o uvd_v4_2.o vce_v2_0.o \
-	amdgpu_amdkfd_gfx_v7.o
+	ci_smc.o ci_dpm.o dce_v8_0.o gfx_v7_0.o cik_sdma.o uvd_v4_2.o vce_v2_0.o
 
 amdgpu-$(CONFIG_DRM_AMDGPU_SI)+= si.o gmc_v6_0.o gfx_v6_0.o si_ih.o si_dma.o dce_v6_0.o si_dpm.o si_smc.o
 
 amdgpu-y += \
-	vi.o mxgpu_vi.o nbio_v6_1.o soc15.o emu_soc.o mxgpu_ai.o nbio_v7_0.o vega10_reg_init.o
+	vi.o mxgpu_vi.o nbio_v6_1.o soc15.o emu_soc.o mxgpu_ai.o nbio_v7_0.o vega10_reg_init.o \
+	vega20_reg_init.o
+
+# add DF block
+amdgpu-y += \
+	df_v1_7.o \
+	df_v3_6.o
 
 # add GMC block
 amdgpu-y += \
@@ -126,11 +131,20 @@ amdgpu-y += \
 	vcn_v1_0.o
 
 # add amdkfd interfaces
+amdgpu-y += amdgpu_amdkfd.o
+
+ifneq ($(CONFIG_HSA_AMD),)
 amdgpu-y += \
-	 amdgpu_amdkfd.o \
 	 amdgpu_amdkfd_fence.o \
 	 amdgpu_amdkfd_gpuvm.o \
-	 amdgpu_amdkfd_gfx_v8.o
+	 amdgpu_amdkfd_gfx_v8.o \
+	 amdgpu_amdkfd_gfx_v9.o
+
+ifneq ($(CONFIG_DRM_AMDGPU_CIK),)
+amdgpu-y += amdgpu_amdkfd_gfx_v7.o
+endif
+
+endif
 
 # add cgs
 amdgpu-y += amdgpu_cgs.o
diff --git a/drivers/gpu/drm/amd/amdgpu/ObjectID.h b/drivers/gpu/drm/amd/amdgpu/ObjectID.h
index 06192698bd96..5b393622f592 100644
--- a/drivers/gpu/drm/amd/amdgpu/ObjectID.h
+++ b/drivers/gpu/drm/amd/amdgpu/ObjectID.h
@@ -136,6 +136,7 @@
 #define GENERIC_OBJECT_ID_PX2_NON_DRIVABLE        0x02
 #define GENERIC_OBJECT_ID_MXM_OPM                 0x03
 #define GENERIC_OBJECT_ID_STEREO_PIN              0x04        //This object could show up from Misc Object table, it follows ATOM_OBJECT format, and contains one ATOM_OBJECT_GPIO_CNTL_RECORD for the stereo pin
+#define GENERIC_OBJECT_ID_BRACKET_LAYOUT          0x05
 
 /****************************************************/
 /* Graphics Object ENUM ID Definition               */
@@ -714,6 +715,13 @@
                                                  GRAPH_OBJECT_ENUM_ID1 << ENUM_ID_SHIFT |\
                                                  GENERIC_OBJECT_ID_STEREO_PIN << OBJECT_ID_SHIFT)
 
+#define GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID1    (GRAPH_OBJECT_TYPE_GENERIC << OBJECT_TYPE_SHIFT |\
+                                                 GRAPH_OBJECT_ENUM_ID1 << ENUM_ID_SHIFT |\
+                                                 GENERIC_OBJECT_ID_BRACKET_LAYOUT << OBJECT_ID_SHIFT)
+
+#define GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID2    (GRAPH_OBJECT_TYPE_GENERIC << OBJECT_TYPE_SHIFT |\
+                                                 GRAPH_OBJECT_ENUM_ID2 << ENUM_ID_SHIFT |\
+                                                 GENERIC_OBJECT_ID_BRACKET_LAYOUT << OBJECT_ID_SHIFT)
 /****************************************************/
 /* Object Cap definition - Shared with BIOS         */
 /****************************************************/
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index c8b605f3dc05..44f62fda4022 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -73,6 +73,7 @@
 #include "amdgpu_virt.h"
 #include "amdgpu_gart.h"
 #include "amdgpu_debugfs.h"
+#include "amdgpu_job.h"
 
 /*
  * Modules parameters.
@@ -105,11 +106,8 @@ extern int amdgpu_vm_fault_stop;
 extern int amdgpu_vm_debug;
 extern int amdgpu_vm_update_mode;
 extern int amdgpu_dc;
-extern int amdgpu_dc_log;
 extern int amdgpu_sched_jobs;
 extern int amdgpu_sched_hw_submission;
-extern int amdgpu_no_evict;
-extern int amdgpu_direct_gma_size;
 extern uint amdgpu_pcie_gen_cap;
 extern uint amdgpu_pcie_lane_cap;
 extern uint amdgpu_cg_mask;
@@ -129,6 +127,7 @@ extern int amdgpu_lbpw;
 extern int amdgpu_compute_multipipe;
 extern int amdgpu_gpu_recovery;
 extern int amdgpu_emu_mode;
+extern uint amdgpu_smu_memory_pool_size;
 
 #ifdef CONFIG_DRM_AMDGPU_SI
 extern int amdgpu_si_support;
@@ -137,6 +136,7 @@ extern int amdgpu_si_support;
 extern int amdgpu_cik_support;
 #endif
 
+#define AMDGPU_SG_THRESHOLD			(256*1024*1024)
 #define AMDGPU_DEFAULT_GTT_SIZE_MB		3072ULL /* 3GB by default */
 #define AMDGPU_WAIT_IDLE_TIMEOUT_IN_MS	        3000
 #define AMDGPU_MAX_USEC_TIMEOUT			100000	/* 100 ms */
@@ -188,6 +188,7 @@ struct amdgpu_job;
 struct amdgpu_irq_src;
 struct amdgpu_fpriv;
 struct amdgpu_bo_va_mapping;
+struct amdgpu_atif;
 
 enum amdgpu_cp_irq {
 	AMDGPU_CP_IRQ_GFX_EOP = 0,
@@ -222,10 +223,10 @@ enum amdgpu_kiq_irq {
 	AMDGPU_CP_KIQ_IRQ_LAST
 };
 
-int amdgpu_device_ip_set_clockgating_state(struct amdgpu_device *adev,
+int amdgpu_device_ip_set_clockgating_state(void *dev,
 					   enum amd_ip_block_type block_type,
 					   enum amd_clockgating_state state);
-int amdgpu_device_ip_set_powergating_state(struct amdgpu_device *adev,
+int amdgpu_device_ip_set_powergating_state(void *dev,
 					   enum amd_ip_block_type block_type,
 					   enum amd_powergating_state state);
 void amdgpu_device_ip_get_clockgating_state(struct amdgpu_device *adev,
@@ -597,17 +598,6 @@ struct amdgpu_ib {
 
 extern const struct drm_sched_backend_ops amdgpu_sched_ops;
 
-int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs,
-		     struct amdgpu_job **job, struct amdgpu_vm *vm);
-int amdgpu_job_alloc_with_ib(struct amdgpu_device *adev, unsigned size,
-			     struct amdgpu_job **job);
-
-void amdgpu_job_free_resources(struct amdgpu_job *job);
-void amdgpu_job_free(struct amdgpu_job *job);
-int amdgpu_job_submit(struct amdgpu_job *job, struct amdgpu_ring *ring,
-		      struct drm_sched_entity *entity, void *owner,
-		      struct dma_fence **f);
-
 /*
  * Queue manager
  */
@@ -681,6 +671,8 @@ int amdgpu_ctx_ioctl(struct drm_device *dev, void *data,
 int amdgpu_ctx_wait_prev_fence(struct amdgpu_ctx *ctx, unsigned ring_id);
 
 void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr);
+void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr);
+void amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr);
 void amdgpu_ctx_mgr_fini(struct amdgpu_ctx_mgr *mgr);
 
 
@@ -727,6 +719,14 @@ void amdgpu_bo_list_get_list(struct amdgpu_bo_list *list,
 			     struct list_head *validated);
 void amdgpu_bo_list_put(struct amdgpu_bo_list *list);
 void amdgpu_bo_list_free(struct amdgpu_bo_list *list);
+int amdgpu_bo_create_list_entry_array(struct drm_amdgpu_bo_list_in *in,
+				      struct drm_amdgpu_bo_list_entry **info_param);
+
+int amdgpu_bo_list_create(struct amdgpu_device *adev,
+				 struct drm_file *filp,
+				 struct drm_amdgpu_bo_list_entry *info,
+				 unsigned num_entries,
+				 struct amdgpu_bo_list **list);
 
 /*
  * GFX stuff
@@ -771,9 +771,18 @@ struct amdgpu_rlc {
 	u32 starting_offsets_start;
 	u32 reg_list_format_size_bytes;
 	u32 reg_list_size_bytes;
+	u32 reg_list_format_direct_reg_list_length;
+	u32 save_restore_list_cntl_size_bytes;
+	u32 save_restore_list_gpm_size_bytes;
+	u32 save_restore_list_srm_size_bytes;
 
 	u32 *register_list_format;
 	u32 *register_restore;
+	u8 *save_restore_list_cntl;
+	u8 *save_restore_list_gpm;
+	u8 *save_restore_list_srm;
+
+	bool is_rlc_v2_1;
 };
 
 #define AMDGPU_MAX_COMPUTE_QUEUES KGD_MAX_QUEUES
@@ -867,6 +876,8 @@ struct amdgpu_gfx_config {
 
 	/* gfx configure feature */
 	uint32_t double_offchip_lds_buf;
+	/* cached value of DB_DEBUG2 */
+	uint32_t db_debug2;
 };
 
 struct amdgpu_cu_info {
@@ -915,6 +926,11 @@ struct amdgpu_ngg {
 	bool			init;
 };
 
+struct sq_work {
+	struct work_struct	work;
+	unsigned ih_data;
+};
+
 struct amdgpu_gfx {
 	struct mutex			gpu_clock_mutex;
 	struct amdgpu_gfx_config	config;
@@ -938,6 +954,12 @@ struct amdgpu_gfx {
 	uint32_t			ce_feature_version;
 	uint32_t			pfp_feature_version;
 	uint32_t			rlc_feature_version;
+	uint32_t			rlc_srlc_fw_version;
+	uint32_t			rlc_srlc_feature_version;
+	uint32_t			rlc_srlg_fw_version;
+	uint32_t			rlc_srlg_feature_version;
+	uint32_t			rlc_srls_fw_version;
+	uint32_t			rlc_srls_feature_version;
 	uint32_t			mec_feature_version;
 	uint32_t			mec2_feature_version;
 	struct amdgpu_ring		gfx_ring[AMDGPU_MAX_GFX_RINGS];
@@ -947,6 +969,10 @@ struct amdgpu_gfx {
 	struct amdgpu_irq_src		eop_irq;
 	struct amdgpu_irq_src		priv_reg_irq;
 	struct amdgpu_irq_src		priv_inst_irq;
+	struct amdgpu_irq_src		cp_ecc_error_irq;
+	struct amdgpu_irq_src		sq_irq;
+	struct sq_work			sq_work;
+
 	/* gfx status */
 	uint32_t			gfx_current_status;
 	/* ce ram size*/
@@ -998,6 +1024,7 @@ struct amdgpu_cs_parser {
 
 	/* scheduler job object */
 	struct amdgpu_job	*job;
+	struct amdgpu_ring	*ring;
 
 	/* buffer objects */
 	struct ww_acquire_ctx		ticket;
@@ -1019,40 +1046,6 @@ struct amdgpu_cs_parser {
 	struct drm_syncobj **post_dep_syncobjs;
 };
 
-#define AMDGPU_PREAMBLE_IB_PRESENT          (1 << 0) /* bit set means command submit involves a preamble IB */
-#define AMDGPU_PREAMBLE_IB_PRESENT_FIRST    (1 << 1) /* bit set means preamble IB is first presented in belonging context */
-#define AMDGPU_HAVE_CTX_SWITCH              (1 << 2) /* bit set means context switch occured */
-
-struct amdgpu_job {
-	struct drm_sched_job    base;
-	struct amdgpu_device	*adev;
-	struct amdgpu_vm	*vm;
-	struct amdgpu_ring	*ring;
-	struct amdgpu_sync	sync;
-	struct amdgpu_sync	sched_sync;
-	struct amdgpu_ib	*ibs;
-	struct dma_fence	*fence; /* the hw fence */
-	uint32_t		preamble_status;
-	uint32_t		num_ibs;
-	void			*owner;
-	uint64_t		fence_ctx; /* the fence_context this job uses */
-	bool                    vm_needs_flush;
-	uint64_t		vm_pd_addr;
-	unsigned		vmid;
-	unsigned		pasid;
-	uint32_t		gds_base, gds_size;
-	uint32_t		gws_base, gws_size;
-	uint32_t		oa_base, oa_size;
-	uint32_t		vram_lost_counter;
-
-	/* user fence handling */
-	uint64_t		uf_addr;
-	uint64_t		uf_sequence;
-
-};
-#define to_amdgpu_job(sched_job)		\
-		container_of((sched_job), struct amdgpu_job, base)
-
 static inline u32 amdgpu_get_ib_value(struct amdgpu_cs_parser *p,
 				      uint32_t ib_idx, int idx)
 {
@@ -1204,6 +1197,8 @@ struct amdgpu_asic_funcs {
 	/* invalidate hdp read cache */
 	void (*invalidate_hdp)(struct amdgpu_device *adev,
 			       struct amdgpu_ring *ring);
+	/* check if the asic needs a full reset of if soft reset will work */
+	bool (*need_full_reset)(struct amdgpu_device *adev);
 };
 
 /*
@@ -1246,43 +1241,6 @@ struct amdgpu_vram_scratch {
 /*
  * ACPI
  */
-struct amdgpu_atif_notification_cfg {
-	bool enabled;
-	int command_code;
-};
-
-struct amdgpu_atif_notifications {
-	bool display_switch;
-	bool expansion_mode_change;
-	bool thermal_state;
-	bool forced_power_state;
-	bool system_power_state;
-	bool display_conf_change;
-	bool px_gfx_switch;
-	bool brightness_change;
-	bool dgpu_display_event;
-};
-
-struct amdgpu_atif_functions {
-	bool system_params;
-	bool sbios_requests;
-	bool select_active_disp;
-	bool lid_state;
-	bool get_tv_standard;
-	bool set_tv_standard;
-	bool get_panel_expansion_mode;
-	bool set_panel_expansion_mode;
-	bool temperature_change;
-	bool graphics_device_types;
-};
-
-struct amdgpu_atif {
-	struct amdgpu_atif_notifications notifications;
-	struct amdgpu_atif_functions functions;
-	struct amdgpu_atif_notification_cfg notification_cfg;
-	struct amdgpu_encoder *encoder_for_bl;
-};
-
 struct amdgpu_atcs_functions {
 	bool get_ext_state;
 	bool pcie_perf_req;
@@ -1368,7 +1326,19 @@ struct amdgpu_nbio_funcs {
 	void (*detect_hw_virt)(struct amdgpu_device *adev);
 };
 
-
+struct amdgpu_df_funcs {
+	void (*init)(struct amdgpu_device *adev);
+	void (*enable_broadcast_mode)(struct amdgpu_device *adev,
+				      bool enable);
+	u32 (*get_fb_channel_number)(struct amdgpu_device *adev);
+	u32 (*get_hbm_channel_number)(struct amdgpu_device *adev);
+	void (*update_medium_grain_clock_gating)(struct amdgpu_device *adev,
+						 bool enable);
+	void (*get_clockgating_state)(struct amdgpu_device *adev,
+				      u32 *flags);
+	void (*enable_ecc_force_par_wr_rmw)(struct amdgpu_device *adev,
+					    bool enable);
+};
 /* Define the HW IP blocks will be used in driver , add more if necessary */
 enum amd_hw_ip_block_type {
 	GC_HWIP = 1,
@@ -1390,6 +1360,7 @@ enum amd_hw_ip_block_type {
 	PWR_HWIP,
 	NBIF_HWIP,
 	THM_HWIP,
+	CLK_HWIP,
 	MAX_HWIP
 };
 
@@ -1398,6 +1369,7 @@ enum amd_hw_ip_block_type {
 struct amd_powerplay {
 	void *pp_handle;
 	const struct amd_pm_funcs *pp_funcs;
+	uint32_t pp_feature;
 };
 
 #define AMDGPU_RESET_MAGIC_NUM 64
@@ -1430,7 +1402,7 @@ struct amdgpu_device {
 #if defined(CONFIG_DEBUG_FS)
 	struct dentry			*debugfs_regs[AMDGPU_DEBUGFS_MAX_COMPONENTS];
 #endif
-	struct amdgpu_atif		atif;
+	struct amdgpu_atif		*atif;
 	struct amdgpu_atcs		atcs;
 	struct mutex			srbm_mutex;
 	/* GRBM index mutex. Protects concurrent access to GRBM index */
@@ -1579,9 +1551,9 @@ struct amdgpu_device {
 	DECLARE_HASHTABLE(mn_hash, 7);
 
 	/* tracking pinned memory */
-	u64 vram_pin_size;
-	u64 invisible_pin_size;
-	u64 gart_pin_size;
+	atomic64_t vram_pin_size;
+	atomic64_t visible_pin_size;
+	atomic64_t gart_pin_size;
 
 	/* amdkfd interface */
 	struct kfd_dev          *kfd;
@@ -1590,6 +1562,7 @@ struct amdgpu_device {
 	uint32_t 		*reg_offset[MAX_HWIP][HWIP_MAX_INSTANCE];
 
 	const struct amdgpu_nbio_funcs	*nbio_funcs;
+	const struct amdgpu_df_funcs	*df_funcs;
 
 	/* delayed work_func for deferring clockgating during resume */
 	struct delayed_work     late_init_work;
@@ -1764,6 +1737,7 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring)
 #define amdgpu_asic_get_config_memsize(adev) (adev)->asic_funcs->get_config_memsize((adev))
 #define amdgpu_asic_flush_hdp(adev, r) (adev)->asic_funcs->flush_hdp((adev), (r))
 #define amdgpu_asic_invalidate_hdp(adev, r) (adev)->asic_funcs->invalidate_hdp((adev), (r))
+#define amdgpu_asic_need_full_reset(adev) (adev)->asic_funcs->need_full_reset((adev))
 #define amdgpu_gmc_flush_gpu_tlb(adev, vmid) (adev)->gmc.gmc_funcs->flush_gpu_tlb((adev), (vmid))
 #define amdgpu_gmc_emit_flush_gpu_tlb(r, vmid, addr) (r)->adev->gmc.gmc_funcs->emit_flush_gpu_tlb((r), (vmid), (addr))
 #define amdgpu_gmc_emit_pasid_mapping(r, vmid, pasid) (r)->adev->gmc.gmc_funcs->emit_pasid_mapping((r), (vmid), (pasid))
@@ -1790,6 +1764,7 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring)
 #define amdgpu_ring_emit_rreg(r, d) (r)->funcs->emit_rreg((r), (d))
 #define amdgpu_ring_emit_wreg(r, d, v) (r)->funcs->emit_wreg((r), (d), (v))
 #define amdgpu_ring_emit_reg_wait(r, d, v, m) (r)->funcs->emit_reg_wait((r), (d), (v), (m))
+#define amdgpu_ring_emit_reg_write_reg_wait(r, d0, d1, v, m) (r)->funcs->emit_reg_write_reg_wait((r), (d0), (d1), (v), (m))
 #define amdgpu_ring_emit_tmz(r, b) (r)->funcs->emit_tmz((r), (b))
 #define amdgpu_ring_pad_ib(r, ib) ((r)->funcs->pad_ib((r), (ib)))
 #define amdgpu_ring_init_cond_exec(r) (r)->funcs->init_cond_exec((r))
@@ -1855,6 +1830,12 @@ static inline bool amdgpu_atpx_dgpu_req_power_for_displays(void) { return false;
 static inline bool amdgpu_has_atpx(void) { return false; }
 #endif
 
+#if defined(CONFIG_VGA_SWITCHEROO) && defined(CONFIG_ACPI)
+void *amdgpu_atpx_get_dhandle(void);
+#else
+static inline void *amdgpu_atpx_get_dhandle(void) { return NULL; }
+#endif
+
 /*
  * KMS
  */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c
index a29362f9ef41..f4c474a95875 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c
@@ -290,12 +290,11 @@ static int acp_hw_init(void *handle)
 	else if (r)
 		return r;
 
-	r = cgs_get_pci_resource(adev->acp.cgs_device, CGS_RESOURCE_TYPE_MMIO,
-			0x5289, 0, &acp_base);
-	if (r == -ENODEV)
-		return 0;
-	else if (r)
-		return r;
+	if (adev->rmmio_size == 0 || adev->rmmio_size < 0x5289)
+		return -EINVAL;
+
+	acp_base = adev->rmmio_base;
+
 	if (adev->asic_type != CHIP_STONEY) {
 		adev->acp.acp_genpd = kzalloc(sizeof(struct acp_pm_domain), GFP_KERNEL);
 		if (adev->acp.acp_genpd == NULL)
@@ -311,20 +310,20 @@ static int acp_hw_init(void *handle)
 		pm_genpd_init(&adev->acp.acp_genpd->gpd, NULL, false);
 	}
 
-	adev->acp.acp_cell = kzalloc(sizeof(struct mfd_cell) * ACP_DEVS,
+	adev->acp.acp_cell = kcalloc(ACP_DEVS, sizeof(struct mfd_cell),
 							GFP_KERNEL);
 
 	if (adev->acp.acp_cell == NULL)
 		return -ENOMEM;
 
-	adev->acp.acp_res = kzalloc(sizeof(struct resource) * 4, GFP_KERNEL);
+	adev->acp.acp_res = kcalloc(4, sizeof(struct resource), GFP_KERNEL);
 
 	if (adev->acp.acp_res == NULL) {
 		kfree(adev->acp.acp_cell);
 		return -ENOMEM;
 	}
 
-	i2s_pdata = kzalloc(sizeof(struct i2s_platform_data) * 2, GFP_KERNEL);
+	i2s_pdata = kcalloc(2, sizeof(struct i2s_platform_data), GFP_KERNEL);
 	if (i2s_pdata == NULL) {
 		kfree(adev->acp.acp_res);
 		kfree(adev->acp.acp_cell);
@@ -513,7 +512,7 @@ static int acp_hw_fini(void *handle)
 	if (adev->acp.acp_genpd) {
 		for (i = 0; i < ACP_DEVS ; i++) {
 			dev = get_mfd_cell_dev(adev->acp.acp_cell[i].name, i);
-			ret = pm_genpd_remove_device(&adev->acp.acp_genpd->gpd, dev);
+			ret = pm_genpd_remove_device(dev);
 			/* If removal fails, dont giveup and try rest */
 			if (ret)
 				dev_err(dev, "remove dev from genpd failed\n");
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c
index 8fa850a070e0..0d8c3fc6eace 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c
@@ -34,6 +34,45 @@
 #include "amd_acpi.h"
 #include "atom.h"
 
+struct amdgpu_atif_notification_cfg {
+	bool enabled;
+	int command_code;
+};
+
+struct amdgpu_atif_notifications {
+	bool display_switch;
+	bool expansion_mode_change;
+	bool thermal_state;
+	bool forced_power_state;
+	bool system_power_state;
+	bool display_conf_change;
+	bool px_gfx_switch;
+	bool brightness_change;
+	bool dgpu_display_event;
+};
+
+struct amdgpu_atif_functions {
+	bool system_params;
+	bool sbios_requests;
+	bool select_active_disp;
+	bool lid_state;
+	bool get_tv_standard;
+	bool set_tv_standard;
+	bool get_panel_expansion_mode;
+	bool set_panel_expansion_mode;
+	bool temperature_change;
+	bool graphics_device_types;
+};
+
+struct amdgpu_atif {
+	acpi_handle handle;
+
+	struct amdgpu_atif_notifications notifications;
+	struct amdgpu_atif_functions functions;
+	struct amdgpu_atif_notification_cfg notification_cfg;
+	struct amdgpu_encoder *encoder_for_bl;
+};
+
 /* Call the ATIF method
  */
 /**
@@ -46,8 +85,9 @@
  * Executes the requested ATIF function (all asics).
  * Returns a pointer to the acpi output buffer.
  */
-static union acpi_object *amdgpu_atif_call(acpi_handle handle, int function,
-		struct acpi_buffer *params)
+static union acpi_object *amdgpu_atif_call(struct amdgpu_atif *atif,
+					   int function,
+					   struct acpi_buffer *params)
 {
 	acpi_status status;
 	union acpi_object atif_arg_elements[2];
@@ -70,7 +110,8 @@ static union acpi_object *amdgpu_atif_call(acpi_handle handle, int function,
 		atif_arg_elements[1].integer.value = 0;
 	}
 
-	status = acpi_evaluate_object(handle, "ATIF", &atif_arg, &buffer);
+	status = acpi_evaluate_object(atif->handle, NULL, &atif_arg,
+				      &buffer);
 
 	/* Fail only if calling the method fails and ATIF is supported */
 	if (ACPI_FAILURE(status) && status != AE_NOT_FOUND) {
@@ -141,15 +182,14 @@ static void amdgpu_atif_parse_functions(struct amdgpu_atif_functions *f, u32 mas
  * (all asics).
  * returns 0 on success, error on failure.
  */
-static int amdgpu_atif_verify_interface(acpi_handle handle,
-		struct amdgpu_atif *atif)
+static int amdgpu_atif_verify_interface(struct amdgpu_atif *atif)
 {
 	union acpi_object *info;
 	struct atif_verify_interface output;
 	size_t size;
 	int err = 0;
 
-	info = amdgpu_atif_call(handle, ATIF_FUNCTION_VERIFY_INTERFACE, NULL);
+	info = amdgpu_atif_call(atif, ATIF_FUNCTION_VERIFY_INTERFACE, NULL);
 	if (!info)
 		return -EIO;
 
@@ -176,6 +216,35 @@ out:
 	return err;
 }
 
+static acpi_handle amdgpu_atif_probe_handle(acpi_handle dhandle)
+{
+	acpi_handle handle = NULL;
+	char acpi_method_name[255] = { 0 };
+	struct acpi_buffer buffer = { sizeof(acpi_method_name), acpi_method_name };
+	acpi_status status;
+
+	/* For PX/HG systems, ATIF and ATPX are in the iGPU's namespace, on dGPU only
+	 * systems, ATIF is in the dGPU's namespace.
+	 */
+	status = acpi_get_handle(dhandle, "ATIF", &handle);
+	if (ACPI_SUCCESS(status))
+		goto out;
+
+	if (amdgpu_has_atpx()) {
+		status = acpi_get_handle(amdgpu_atpx_get_dhandle(), "ATIF",
+					 &handle);
+		if (ACPI_SUCCESS(status))
+			goto out;
+	}
+
+	DRM_DEBUG_DRIVER("No ATIF handle found\n");
+	return NULL;
+out:
+	acpi_get_name(handle, ACPI_FULL_PATHNAME, &buffer);
+	DRM_DEBUG_DRIVER("Found ATIF handle %s\n", acpi_method_name);
+	return handle;
+}
+
 /**
  * amdgpu_atif_get_notification_params - determine notify configuration
  *
@@ -188,15 +257,16 @@ out:
  * where n is specified in the result if a notifier is used.
  * Returns 0 on success, error on failure.
  */
-static int amdgpu_atif_get_notification_params(acpi_handle handle,
-		struct amdgpu_atif_notification_cfg *n)
+static int amdgpu_atif_get_notification_params(struct amdgpu_atif *atif)
 {
 	union acpi_object *info;
+	struct amdgpu_atif_notification_cfg *n = &atif->notification_cfg;
 	struct atif_system_params params;
 	size_t size;
 	int err = 0;
 
-	info = amdgpu_atif_call(handle, ATIF_FUNCTION_GET_SYSTEM_PARAMETERS, NULL);
+	info = amdgpu_atif_call(atif, ATIF_FUNCTION_GET_SYSTEM_PARAMETERS,
+				NULL);
 	if (!info) {
 		err = -EIO;
 		goto out;
@@ -250,14 +320,15 @@ out:
  * (all asics).
  * Returns 0 on success, error on failure.
  */
-static int amdgpu_atif_get_sbios_requests(acpi_handle handle,
-		struct atif_sbios_requests *req)
+static int amdgpu_atif_get_sbios_requests(struct amdgpu_atif *atif,
+					  struct atif_sbios_requests *req)
 {
 	union acpi_object *info;
 	size_t size;
 	int count = 0;
 
-	info = amdgpu_atif_call(handle, ATIF_FUNCTION_GET_SYSTEM_BIOS_REQUESTS, NULL);
+	info = amdgpu_atif_call(atif, ATIF_FUNCTION_GET_SYSTEM_BIOS_REQUESTS,
+				NULL);
 	if (!info)
 		return -EIO;
 
@@ -290,11 +361,10 @@ out:
  * Returns NOTIFY code
  */
 static int amdgpu_atif_handler(struct amdgpu_device *adev,
-			struct acpi_bus_event *event)
+			       struct acpi_bus_event *event)
 {
-	struct amdgpu_atif *atif = &adev->atif;
+	struct amdgpu_atif *atif = adev->atif;
 	struct atif_sbios_requests req;
-	acpi_handle handle;
 	int count;
 
 	DRM_DEBUG_DRIVER("event, device_class = %s, type = %#x\n",
@@ -303,14 +373,14 @@ static int amdgpu_atif_handler(struct amdgpu_device *adev,
 	if (strcmp(event->device_class, ACPI_VIDEO_CLASS) != 0)
 		return NOTIFY_DONE;
 
-	if (!atif->notification_cfg.enabled ||
+	if (!atif ||
+	    !atif->notification_cfg.enabled ||
 	    event->type != atif->notification_cfg.command_code)
 		/* Not our event */
 		return NOTIFY_DONE;
 
 	/* Check pending SBIOS requests */
-	handle = ACPI_HANDLE(&adev->pdev->dev);
-	count = amdgpu_atif_get_sbios_requests(handle, &req);
+	count = amdgpu_atif_get_sbios_requests(atif, &req);
 
 	if (count <= 0)
 		return NOTIFY_DONE;
@@ -641,8 +711,8 @@ static int amdgpu_acpi_event(struct notifier_block *nb,
  */
 int amdgpu_acpi_init(struct amdgpu_device *adev)
 {
-	acpi_handle handle;
-	struct amdgpu_atif *atif = &adev->atif;
+	acpi_handle handle, atif_handle;
+	struct amdgpu_atif *atif;
 	struct amdgpu_atcs *atcs = &adev->atcs;
 	int ret;
 
@@ -658,12 +728,26 @@ int amdgpu_acpi_init(struct amdgpu_device *adev)
 		DRM_DEBUG_DRIVER("Call to ATCS verify_interface failed: %d\n", ret);
 	}
 
+	/* Probe for ATIF, and initialize it if found */
+	atif_handle = amdgpu_atif_probe_handle(handle);
+	if (!atif_handle)
+		goto out;
+
+	atif = kzalloc(sizeof(*atif), GFP_KERNEL);
+	if (!atif) {
+		DRM_WARN("Not enough memory to initialize ATIF\n");
+		goto out;
+	}
+	atif->handle = atif_handle;
+
 	/* Call the ATIF method */
-	ret = amdgpu_atif_verify_interface(handle, atif);
+	ret = amdgpu_atif_verify_interface(atif);
 	if (ret) {
 		DRM_DEBUG_DRIVER("Call to ATIF verify_interface failed: %d\n", ret);
+		kfree(atif);
 		goto out;
 	}
+	adev->atif = atif;
 
 	if (atif->notifications.brightness_change) {
 		struct drm_encoder *tmp;
@@ -693,8 +777,7 @@ int amdgpu_acpi_init(struct amdgpu_device *adev)
 	}
 
 	if (atif->functions.system_params) {
-		ret = amdgpu_atif_get_notification_params(handle,
-				&atif->notification_cfg);
+		ret = amdgpu_atif_get_notification_params(atif);
 		if (ret) {
 			DRM_DEBUG_DRIVER("Call to GET_SYSTEM_PARAMS failed: %d\n",
 					ret);
@@ -720,4 +803,6 @@ out:
 void amdgpu_acpi_fini(struct amdgpu_device *adev)
 {
 	unregister_acpi_notifier(&adev->acpi_nb);
+	if (adev->atif)
+		kfree(adev->atif);
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
index 4d36203ffb11..e3ed08dca7b7 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c
@@ -50,15 +50,21 @@ int amdgpu_amdkfd_init(void)
 		kgd2kfd = NULL;
 	}
 
+
 #elif defined(CONFIG_HSA_AMD)
+
 	ret = kgd2kfd_init(KFD_INTERFACE_VERSION, &kgd2kfd);
 	if (ret)
 		kgd2kfd = NULL;
 
 #else
+	kgd2kfd = NULL;
 	ret = -ENOENT;
 #endif
+
+#if defined(CONFIG_HSA_AMD_MODULE) || defined(CONFIG_HSA_AMD)
 	amdgpu_amdkfd_gpuvm_init_mem_limits();
+#endif
 
 	return ret;
 }
@@ -92,8 +98,12 @@ void amdgpu_amdkfd_device_probe(struct amdgpu_device *adev)
 	case CHIP_POLARIS11:
 		kfd2kgd = amdgpu_amdkfd_gfx_8_0_get_functions();
 		break;
+	case CHIP_VEGA10:
+	case CHIP_RAVEN:
+		kfd2kgd = amdgpu_amdkfd_gfx_9_0_get_functions();
+		break;
 	default:
-		dev_dbg(adev->dev, "kfd not supported on this ASIC\n");
+		dev_info(adev->dev, "kfd not supported on this ASIC\n");
 		return;
 	}
 
@@ -175,6 +185,28 @@ void amdgpu_amdkfd_device_init(struct amdgpu_device *adev)
 				&gpu_resources.doorbell_physical_address,
 				&gpu_resources.doorbell_aperture_size,
 				&gpu_resources.doorbell_start_offset);
+		if (adev->asic_type >= CHIP_VEGA10) {
+			/* On SOC15 the BIF is involved in routing
+			 * doorbells using the low 12 bits of the
+			 * address. Communicate the assignments to
+			 * KFD. KFD uses two doorbell pages per
+			 * process in case of 64-bit doorbells so we
+			 * can use each doorbell assignment twice.
+			 */
+			gpu_resources.sdma_doorbell[0][0] =
+				AMDGPU_DOORBELL64_sDMA_ENGINE0;
+			gpu_resources.sdma_doorbell[0][1] =
+				AMDGPU_DOORBELL64_sDMA_ENGINE0 + 0x200;
+			gpu_resources.sdma_doorbell[1][0] =
+				AMDGPU_DOORBELL64_sDMA_ENGINE1;
+			gpu_resources.sdma_doorbell[1][1] =
+				AMDGPU_DOORBELL64_sDMA_ENGINE1 + 0x200;
+			/* Doorbells 0x0f0-0ff and 0x2f0-2ff are reserved for
+			 * SDMA, IH and VCN. So don't use them for the CP.
+			 */
+			gpu_resources.reserved_doorbell_mask = 0x1f0;
+			gpu_resources.reserved_doorbell_val  = 0x0f0;
+		}
 
 		kgd2kfd->device_init(adev->kfd, &gpu_resources);
 	}
@@ -217,13 +249,18 @@ int alloc_gtt_mem(struct kgd_dev *kgd, size_t size,
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
 	struct amdgpu_bo *bo = NULL;
+	struct amdgpu_bo_param bp;
 	int r;
-	uint64_t gpu_addr_tmp = 0;
 	void *cpu_ptr_tmp = NULL;
 
-	r = amdgpu_bo_create(adev, size, PAGE_SIZE, AMDGPU_GEM_DOMAIN_GTT,
-			     AMDGPU_GEM_CREATE_CPU_GTT_USWC, ttm_bo_type_kernel,
-			     NULL, &bo);
+	memset(&bp, 0, sizeof(bp));
+	bp.size = size;
+	bp.byte_align = PAGE_SIZE;
+	bp.domain = AMDGPU_GEM_DOMAIN_GTT;
+	bp.flags = AMDGPU_GEM_CREATE_CPU_GTT_USWC;
+	bp.type = ttm_bo_type_kernel;
+	bp.resv = NULL;
+	r = amdgpu_bo_create(adev, &bp, &bo);
 	if (r) {
 		dev_err(adev->dev,
 			"failed to allocate BO for amdkfd (%d)\n", r);
@@ -237,13 +274,18 @@ int alloc_gtt_mem(struct kgd_dev *kgd, size_t size,
 		goto allocate_mem_reserve_bo_failed;
 	}
 
-	r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT,
-				&gpu_addr_tmp);
+	r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT);
 	if (r) {
 		dev_err(adev->dev, "(%d) failed to pin bo for amdkfd\n", r);
 		goto allocate_mem_pin_bo_failed;
 	}
 
+	r = amdgpu_ttm_alloc_gart(&bo->tbo);
+	if (r) {
+		dev_err(adev->dev, "%p bind failed\n", bo);
+		goto allocate_mem_kmap_bo_failed;
+	}
+
 	r = amdgpu_bo_kmap(bo, &cpu_ptr_tmp);
 	if (r) {
 		dev_err(adev->dev,
@@ -252,7 +294,7 @@ int alloc_gtt_mem(struct kgd_dev *kgd, size_t size,
 	}
 
 	*mem_obj = bo;
-	*gpu_addr = gpu_addr_tmp;
+	*gpu_addr = amdgpu_bo_gpu_offset(bo);
 	*cpu_ptr = cpu_ptr_tmp;
 
 	amdgpu_bo_unreserve(bo);
@@ -304,15 +346,12 @@ void get_local_mem_info(struct kgd_dev *kgd,
 			mem_info->local_mem_size_public,
 			mem_info->local_mem_size_private);
 
-	if (amdgpu_emu_mode == 1) {
-		mem_info->mem_clk_max = 100;
-		return;
-	}
-
 	if (amdgpu_sriov_vf(adev))
 		mem_info->mem_clk_max = adev->clock.default_mclk / 100;
-	else
+	else if (adev->powerplay.pp_funcs)
 		mem_info->mem_clk_max = amdgpu_dpm_get_mclk(adev, false) / 100;
+	else
+		mem_info->mem_clk_max = 100;
 }
 
 uint64_t get_gpu_clock_counter(struct kgd_dev *kgd)
@@ -329,13 +368,12 @@ uint32_t get_max_engine_clock_in_mhz(struct kgd_dev *kgd)
 	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
 
 	/* the sclk is in quantas of 10kHz */
-	if (amdgpu_emu_mode == 1)
-		return 100;
-
 	if (amdgpu_sriov_vf(adev))
 		return adev->clock.default_sclk / 100;
-
-	return amdgpu_dpm_get_sclk(adev, false) / 100;
+	else if (adev->powerplay.pp_funcs)
+		return amdgpu_dpm_get_sclk(adev, false) / 100;
+	else
+		return 100;
 }
 
 void get_cu_info(struct kgd_dev *kgd, struct kfd_cu_info *cu_info)
@@ -432,3 +470,44 @@ bool amdgpu_amdkfd_is_kfd_vmid(struct amdgpu_device *adev, u32 vmid)
 
 	return false;
 }
+
+#if !defined(CONFIG_HSA_AMD_MODULE) && !defined(CONFIG_HSA_AMD)
+bool amdkfd_fence_check_mm(struct dma_fence *f, struct mm_struct *mm)
+{
+	return false;
+}
+
+void amdgpu_amdkfd_unreserve_system_memory_limit(struct amdgpu_bo *bo)
+{
+}
+
+void amdgpu_amdkfd_gpuvm_destroy_cb(struct amdgpu_device *adev,
+					struct amdgpu_vm *vm)
+{
+}
+
+struct amdgpu_amdkfd_fence *to_amdgpu_amdkfd_fence(struct dma_fence *f)
+{
+	return NULL;
+}
+
+int amdgpu_amdkfd_evict_userptr(struct kgd_mem *mem, struct mm_struct *mm)
+{
+	return 0;
+}
+
+struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void)
+{
+	return NULL;
+}
+
+struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void)
+{
+	return NULL;
+}
+
+struct kfd2kgd_calls *amdgpu_amdkfd_gfx_9_0_get_functions(void)
+{
+	return NULL;
+}
+#endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
index c2c2bea731e0..a8418a3f4e9d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
@@ -28,6 +28,7 @@
 #include <linux/types.h>
 #include <linux/mm.h>
 #include <linux/mmu_context.h>
+#include <linux/workqueue.h>
 #include <kgd_kfd_interface.h>
 #include <drm/ttm/ttm_execbuf_util.h>
 #include "amdgpu_sync.h"
@@ -59,7 +60,9 @@ struct kgd_mem {
 
 	uint32_t mapping_flags;
 
+	atomic_t invalid;
 	struct amdkfd_process_info *process_info;
+	struct page **user_pages;
 
 	struct amdgpu_sync sync;
 
@@ -84,6 +87,9 @@ struct amdkfd_process_info {
 	struct list_head vm_list_head;
 	/* List head for all KFD BOs that belong to a KFD process. */
 	struct list_head kfd_bo_list;
+	/* List of userptr BOs that are valid or invalid */
+	struct list_head userptr_valid_list;
+	struct list_head userptr_inval_list;
 	/* Lock to protect kfd_bo_list */
 	struct mutex lock;
 
@@ -91,6 +97,11 @@ struct amdkfd_process_info {
 	unsigned int n_vms;
 	/* Eviction Fence */
 	struct amdgpu_amdkfd_fence *eviction_fence;
+
+	/* MMU-notifier related fields */
+	atomic_t evicted_bos;
+	struct delayed_work restore_userptr_work;
+	struct pid *pid;
 };
 
 int amdgpu_amdkfd_init(void);
@@ -104,12 +115,14 @@ void amdgpu_amdkfd_device_probe(struct amdgpu_device *adev);
 void amdgpu_amdkfd_device_init(struct amdgpu_device *adev);
 void amdgpu_amdkfd_device_fini(struct amdgpu_device *adev);
 
+int amdgpu_amdkfd_evict_userptr(struct kgd_mem *mem, struct mm_struct *mm);
 int amdgpu_amdkfd_submit_ib(struct kgd_dev *kgd, enum kgd_engine_type engine,
 				uint32_t vmid, uint64_t gpu_addr,
 				uint32_t *ib_cmd, uint32_t ib_len);
 
 struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void);
 struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void);
+struct kfd2kgd_calls *amdgpu_amdkfd_gfx_9_0_get_functions(void);
 
 bool amdgpu_amdkfd_is_kfd_vmid(struct amdgpu_device *adev, u32 vmid);
 
@@ -143,14 +156,14 @@ uint64_t amdgpu_amdkfd_get_vram_usage(struct kgd_dev *kgd);
 
 /* GPUVM API */
 int amdgpu_amdkfd_gpuvm_create_process_vm(struct kgd_dev *kgd, void **vm,
-					  void **process_info,
-					  struct dma_fence **ef);
+					void **process_info,
+					struct dma_fence **ef);
 int amdgpu_amdkfd_gpuvm_acquire_process_vm(struct kgd_dev *kgd,
-					   struct file *filp,
-					   void **vm, void **process_info,
-					   struct dma_fence **ef);
+					struct file *filp,
+					void **vm, void **process_info,
+					struct dma_fence **ef);
 void amdgpu_amdkfd_gpuvm_destroy_cb(struct amdgpu_device *adev,
-				    struct amdgpu_vm *vm);
+				struct amdgpu_vm *vm);
 void amdgpu_amdkfd_gpuvm_destroy_process_vm(struct kgd_dev *kgd, void *vm);
 uint32_t amdgpu_amdkfd_gpuvm_get_process_page_dir(void *vm);
 int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c
index 2c14025e5e76..574c1181ae9a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_fence.c
@@ -173,7 +173,5 @@ static const struct dma_fence_ops amdkfd_fence_ops = {
 	.get_driver_name = amdkfd_fence_get_driver_name,
 	.get_timeline_name = amdkfd_fence_get_timeline_name,
 	.enable_signaling = amdkfd_fence_enable_signaling,
-	.signaled = NULL,
-	.wait = dma_fence_default_wait,
 	.release = amdkfd_fence_release,
 };
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c
index ea54e53172b9..ea79908dac4c 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v7.c
@@ -98,8 +98,6 @@ static void kgd_program_sh_mem_settings(struct kgd_dev *kgd, uint32_t vmid,
 static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
 					unsigned int vmid);
 
-static int kgd_init_pipeline(struct kgd_dev *kgd, uint32_t pipe_id,
-				uint32_t hpd_size, uint64_t hpd_gpu_addr);
 static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id);
 static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
 			uint32_t queue_id, uint32_t __user *wptr,
@@ -183,7 +181,6 @@ static const struct kfd2kgd_calls kfd2kgd = {
 	.free_pasid = amdgpu_pasid_free,
 	.program_sh_mem_settings = kgd_program_sh_mem_settings,
 	.set_pasid_vmid_mapping = kgd_set_pasid_vmid_mapping,
-	.init_pipeline = kgd_init_pipeline,
 	.init_interrupts = kgd_init_interrupts,
 	.hqd_load = kgd_hqd_load,
 	.hqd_sdma_load = kgd_hqd_sdma_load,
@@ -309,13 +306,6 @@ static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
 	return 0;
 }
 
-static int kgd_init_pipeline(struct kgd_dev *kgd, uint32_t pipe_id,
-				uint32_t hpd_size, uint64_t hpd_gpu_addr)
-{
-	/* amdgpu owns the per-pipe state */
-	return 0;
-}
-
 static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id)
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
@@ -417,7 +407,7 @@ static int kgd_hqd_dump(struct kgd_dev *kgd,
 		(*dump)[i++][1] = RREG32(addr);		\
 	} while (0)
 
-	*dump = kmalloc(HQD_N_REGS*2*sizeof(uint32_t), GFP_KERNEL);
+	*dump = kmalloc_array(HQD_N_REGS * 2, sizeof(uint32_t), GFP_KERNEL);
 	if (*dump == NULL)
 		return -ENOMEM;
 
@@ -514,7 +504,7 @@ static int kgd_hqd_sdma_dump(struct kgd_dev *kgd,
 #undef HQD_N_REGS
 #define HQD_N_REGS (19+4)
 
-	*dump = kmalloc(HQD_N_REGS*2*sizeof(uint32_t), GFP_KERNEL);
+	*dump = kmalloc_array(HQD_N_REGS * 2, sizeof(uint32_t), GFP_KERNEL);
 	if (*dump == NULL)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c
index 89264c9a5e9f..19dd665e7307 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v8.c
@@ -57,8 +57,6 @@ static void kgd_program_sh_mem_settings(struct kgd_dev *kgd, uint32_t vmid,
 		uint32_t sh_mem_bases);
 static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
 		unsigned int vmid);
-static int kgd_init_pipeline(struct kgd_dev *kgd, uint32_t pipe_id,
-		uint32_t hpd_size, uint64_t hpd_gpu_addr);
 static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id);
 static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
 			uint32_t queue_id, uint32_t __user *wptr,
@@ -141,7 +139,6 @@ static const struct kfd2kgd_calls kfd2kgd = {
 	.free_pasid = amdgpu_pasid_free,
 	.program_sh_mem_settings = kgd_program_sh_mem_settings,
 	.set_pasid_vmid_mapping = kgd_set_pasid_vmid_mapping,
-	.init_pipeline = kgd_init_pipeline,
 	.init_interrupts = kgd_init_interrupts,
 	.hqd_load = kgd_hqd_load,
 	.hqd_sdma_load = kgd_hqd_sdma_load,
@@ -270,13 +267,6 @@ static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
 	return 0;
 }
 
-static int kgd_init_pipeline(struct kgd_dev *kgd, uint32_t pipe_id,
-				uint32_t hpd_size, uint64_t hpd_gpu_addr)
-{
-	/* amdgpu owns the per-pipe state */
-	return 0;
-}
-
 static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id)
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
@@ -405,7 +395,7 @@ static int kgd_hqd_dump(struct kgd_dev *kgd,
 		(*dump)[i++][1] = RREG32(addr);		\
 	} while (0)
 
-	*dump = kmalloc(HQD_N_REGS*2*sizeof(uint32_t), GFP_KERNEL);
+	*dump = kmalloc_array(HQD_N_REGS * 2, sizeof(uint32_t), GFP_KERNEL);
 	if (*dump == NULL)
 		return -ENOMEM;
 
@@ -501,7 +491,7 @@ static int kgd_hqd_sdma_dump(struct kgd_dev *kgd,
 #undef HQD_N_REGS
 #define HQD_N_REGS (19+4+2+3+7)
 
-	*dump = kmalloc(HQD_N_REGS*2*sizeof(uint32_t), GFP_KERNEL);
+	*dump = kmalloc_array(HQD_N_REGS * 2, sizeof(uint32_t), GFP_KERNEL);
 	if (*dump == NULL)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
new file mode 100644
index 000000000000..1db60aa5b7f0
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gfx_v9.c
@@ -0,0 +1,1043 @@
+/*
+ * Copyright 2014-2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#define pr_fmt(fmt) "kfd2kgd: " fmt
+
+#include <linux/module.h>
+#include <linux/fdtable.h>
+#include <linux/uaccess.h>
+#include <linux/firmware.h>
+#include <drm/drmP.h>
+#include "amdgpu.h"
+#include "amdgpu_amdkfd.h"
+#include "amdgpu_ucode.h"
+#include "soc15_hw_ip.h"
+#include "gc/gc_9_0_offset.h"
+#include "gc/gc_9_0_sh_mask.h"
+#include "vega10_enum.h"
+#include "sdma0/sdma0_4_0_offset.h"
+#include "sdma0/sdma0_4_0_sh_mask.h"
+#include "sdma1/sdma1_4_0_offset.h"
+#include "sdma1/sdma1_4_0_sh_mask.h"
+#include "athub/athub_1_0_offset.h"
+#include "athub/athub_1_0_sh_mask.h"
+#include "oss/osssys_4_0_offset.h"
+#include "oss/osssys_4_0_sh_mask.h"
+#include "soc15_common.h"
+#include "v9_structs.h"
+#include "soc15.h"
+#include "soc15d.h"
+
+/* HACK: MMHUB and GC both have VM-related register with the same
+ * names but different offsets. Define the MMHUB register we need here
+ * with a prefix. A proper solution would be to move the functions
+ * programming these registers into gfx_v9_0.c and mmhub_v1_0.c
+ * respectively.
+ */
+#define mmMMHUB_VM_INVALIDATE_ENG16_REQ				0x06f3
+#define mmMMHUB_VM_INVALIDATE_ENG16_REQ_BASE_IDX		0
+
+#define mmMMHUB_VM_INVALIDATE_ENG16_ACK				0x0705
+#define mmMMHUB_VM_INVALIDATE_ENG16_ACK_BASE_IDX		0
+
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32		0x072b
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32_BASE_IDX	0
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32		0x072c
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32_BASE_IDX	0
+
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32		0x074b
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32_BASE_IDX	0
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32		0x074c
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32_BASE_IDX	0
+
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32		0x076b
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32_BASE_IDX	0
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32		0x076c
+#define mmMMHUB_VM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32_BASE_IDX	0
+
+#define mmMMHUB_VM_INVALIDATE_ENG16_ADDR_RANGE_LO32		0x0727
+#define mmMMHUB_VM_INVALIDATE_ENG16_ADDR_RANGE_LO32_BASE_IDX	0
+#define mmMMHUB_VM_INVALIDATE_ENG16_ADDR_RANGE_HI32		0x0728
+#define mmMMHUB_VM_INVALIDATE_ENG16_ADDR_RANGE_HI32_BASE_IDX	0
+
+#define V9_PIPE_PER_MEC		(4)
+#define V9_QUEUES_PER_PIPE_MEC	(8)
+
+enum hqd_dequeue_request_type {
+	NO_ACTION = 0,
+	DRAIN_PIPE,
+	RESET_WAVES
+};
+
+/*
+ * Register access functions
+ */
+
+static void kgd_program_sh_mem_settings(struct kgd_dev *kgd, uint32_t vmid,
+		uint32_t sh_mem_config,
+		uint32_t sh_mem_ape1_base, uint32_t sh_mem_ape1_limit,
+		uint32_t sh_mem_bases);
+static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
+		unsigned int vmid);
+static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id);
+static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
+			uint32_t queue_id, uint32_t __user *wptr,
+			uint32_t wptr_shift, uint32_t wptr_mask,
+			struct mm_struct *mm);
+static int kgd_hqd_dump(struct kgd_dev *kgd,
+			uint32_t pipe_id, uint32_t queue_id,
+			uint32_t (**dump)[2], uint32_t *n_regs);
+static int kgd_hqd_sdma_load(struct kgd_dev *kgd, void *mqd,
+			     uint32_t __user *wptr, struct mm_struct *mm);
+static int kgd_hqd_sdma_dump(struct kgd_dev *kgd,
+			     uint32_t engine_id, uint32_t queue_id,
+			     uint32_t (**dump)[2], uint32_t *n_regs);
+static bool kgd_hqd_is_occupied(struct kgd_dev *kgd, uint64_t queue_address,
+		uint32_t pipe_id, uint32_t queue_id);
+static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd);
+static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd,
+				enum kfd_preempt_type reset_type,
+				unsigned int utimeout, uint32_t pipe_id,
+				uint32_t queue_id);
+static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd,
+				unsigned int utimeout);
+static int kgd_address_watch_disable(struct kgd_dev *kgd);
+static int kgd_address_watch_execute(struct kgd_dev *kgd,
+					unsigned int watch_point_id,
+					uint32_t cntl_val,
+					uint32_t addr_hi,
+					uint32_t addr_lo);
+static int kgd_wave_control_execute(struct kgd_dev *kgd,
+					uint32_t gfx_index_val,
+					uint32_t sq_cmd);
+static uint32_t kgd_address_watch_get_offset(struct kgd_dev *kgd,
+					unsigned int watch_point_id,
+					unsigned int reg_offset);
+
+static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd,
+		uint8_t vmid);
+static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
+		uint8_t vmid);
+static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
+		uint32_t page_table_base);
+static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type);
+static void set_scratch_backing_va(struct kgd_dev *kgd,
+					uint64_t va, uint32_t vmid);
+static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid);
+static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid);
+
+/* Because of REG_GET_FIELD() being used, we put this function in the
+ * asic specific file.
+ */
+static int amdgpu_amdkfd_get_tile_config(struct kgd_dev *kgd,
+		struct tile_config *config)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)kgd;
+
+	config->gb_addr_config = adev->gfx.config.gb_addr_config;
+
+	config->tile_config_ptr = adev->gfx.config.tile_mode_array;
+	config->num_tile_configs =
+			ARRAY_SIZE(adev->gfx.config.tile_mode_array);
+	config->macro_tile_config_ptr =
+			adev->gfx.config.macrotile_mode_array;
+	config->num_macro_tile_configs =
+			ARRAY_SIZE(adev->gfx.config.macrotile_mode_array);
+
+	return 0;
+}
+
+static const struct kfd2kgd_calls kfd2kgd = {
+	.init_gtt_mem_allocation = alloc_gtt_mem,
+	.free_gtt_mem = free_gtt_mem,
+	.get_local_mem_info = get_local_mem_info,
+	.get_gpu_clock_counter = get_gpu_clock_counter,
+	.get_max_engine_clock_in_mhz = get_max_engine_clock_in_mhz,
+	.alloc_pasid = amdgpu_pasid_alloc,
+	.free_pasid = amdgpu_pasid_free,
+	.program_sh_mem_settings = kgd_program_sh_mem_settings,
+	.set_pasid_vmid_mapping = kgd_set_pasid_vmid_mapping,
+	.init_interrupts = kgd_init_interrupts,
+	.hqd_load = kgd_hqd_load,
+	.hqd_sdma_load = kgd_hqd_sdma_load,
+	.hqd_dump = kgd_hqd_dump,
+	.hqd_sdma_dump = kgd_hqd_sdma_dump,
+	.hqd_is_occupied = kgd_hqd_is_occupied,
+	.hqd_sdma_is_occupied = kgd_hqd_sdma_is_occupied,
+	.hqd_destroy = kgd_hqd_destroy,
+	.hqd_sdma_destroy = kgd_hqd_sdma_destroy,
+	.address_watch_disable = kgd_address_watch_disable,
+	.address_watch_execute = kgd_address_watch_execute,
+	.wave_control_execute = kgd_wave_control_execute,
+	.address_watch_get_offset = kgd_address_watch_get_offset,
+	.get_atc_vmid_pasid_mapping_pasid =
+			get_atc_vmid_pasid_mapping_pasid,
+	.get_atc_vmid_pasid_mapping_valid =
+			get_atc_vmid_pasid_mapping_valid,
+	.get_fw_version = get_fw_version,
+	.set_scratch_backing_va = set_scratch_backing_va,
+	.get_tile_config = amdgpu_amdkfd_get_tile_config,
+	.get_cu_info = get_cu_info,
+	.get_vram_usage = amdgpu_amdkfd_get_vram_usage,
+	.create_process_vm = amdgpu_amdkfd_gpuvm_create_process_vm,
+	.acquire_process_vm = amdgpu_amdkfd_gpuvm_acquire_process_vm,
+	.destroy_process_vm = amdgpu_amdkfd_gpuvm_destroy_process_vm,
+	.get_process_page_dir = amdgpu_amdkfd_gpuvm_get_process_page_dir,
+	.set_vm_context_page_table_base = set_vm_context_page_table_base,
+	.alloc_memory_of_gpu = amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu,
+	.free_memory_of_gpu = amdgpu_amdkfd_gpuvm_free_memory_of_gpu,
+	.map_memory_to_gpu = amdgpu_amdkfd_gpuvm_map_memory_to_gpu,
+	.unmap_memory_to_gpu = amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu,
+	.sync_memory = amdgpu_amdkfd_gpuvm_sync_memory,
+	.map_gtt_bo_to_kernel = amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel,
+	.restore_process_bos = amdgpu_amdkfd_gpuvm_restore_process_bos,
+	.invalidate_tlbs = invalidate_tlbs,
+	.invalidate_tlbs_vmid = invalidate_tlbs_vmid,
+	.submit_ib = amdgpu_amdkfd_submit_ib,
+};
+
+struct kfd2kgd_calls *amdgpu_amdkfd_gfx_9_0_get_functions(void)
+{
+	return (struct kfd2kgd_calls *)&kfd2kgd;
+}
+
+static inline struct amdgpu_device *get_amdgpu_device(struct kgd_dev *kgd)
+{
+	return (struct amdgpu_device *)kgd;
+}
+
+static void lock_srbm(struct kgd_dev *kgd, uint32_t mec, uint32_t pipe,
+			uint32_t queue, uint32_t vmid)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+
+	mutex_lock(&adev->srbm_mutex);
+	soc15_grbm_select(adev, mec, pipe, queue, vmid);
+}
+
+static void unlock_srbm(struct kgd_dev *kgd)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+
+	soc15_grbm_select(adev, 0, 0, 0, 0);
+	mutex_unlock(&adev->srbm_mutex);
+}
+
+static void acquire_queue(struct kgd_dev *kgd, uint32_t pipe_id,
+				uint32_t queue_id)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+
+	uint32_t mec = (pipe_id / adev->gfx.mec.num_pipe_per_mec) + 1;
+	uint32_t pipe = (pipe_id % adev->gfx.mec.num_pipe_per_mec);
+
+	lock_srbm(kgd, mec, pipe, queue_id, 0);
+}
+
+static uint32_t get_queue_mask(struct amdgpu_device *adev,
+			       uint32_t pipe_id, uint32_t queue_id)
+{
+	unsigned int bit = (pipe_id * adev->gfx.mec.num_queue_per_pipe +
+			    queue_id) & 31;
+
+	return ((uint32_t)1) << bit;
+}
+
+static void release_queue(struct kgd_dev *kgd)
+{
+	unlock_srbm(kgd);
+}
+
+static void kgd_program_sh_mem_settings(struct kgd_dev *kgd, uint32_t vmid,
+					uint32_t sh_mem_config,
+					uint32_t sh_mem_ape1_base,
+					uint32_t sh_mem_ape1_limit,
+					uint32_t sh_mem_bases)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+
+	lock_srbm(kgd, 0, 0, 0, vmid);
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmSH_MEM_CONFIG), sh_mem_config);
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmSH_MEM_BASES), sh_mem_bases);
+	/* APE1 no longer exists on GFX9 */
+
+	unlock_srbm(kgd);
+}
+
+static int kgd_set_pasid_vmid_mapping(struct kgd_dev *kgd, unsigned int pasid,
+					unsigned int vmid)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+
+	/*
+	 * We have to assume that there is no outstanding mapping.
+	 * The ATC_VMID_PASID_MAPPING_UPDATE_STATUS bit could be 0 because
+	 * a mapping is in progress or because a mapping finished
+	 * and the SW cleared it.
+	 * So the protocol is to always wait & clear.
+	 */
+	uint32_t pasid_mapping = (pasid == 0) ? 0 : (uint32_t)pasid |
+			ATC_VMID0_PASID_MAPPING__VALID_MASK;
+
+	/*
+	 * need to do this twice, once for gfx and once for mmhub
+	 * for ATC add 16 to VMID for mmhub, for IH different registers.
+	 * ATC_VMID0..15 registers are separate from ATC_VMID16..31.
+	 */
+
+	WREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID0_PASID_MAPPING) + vmid,
+	       pasid_mapping);
+
+	while (!(RREG32(SOC15_REG_OFFSET(
+				ATHUB, 0,
+				mmATC_VMID_PASID_MAPPING_UPDATE_STATUS)) &
+		 (1U << vmid)))
+		cpu_relax();
+
+	WREG32(SOC15_REG_OFFSET(ATHUB, 0,
+				mmATC_VMID_PASID_MAPPING_UPDATE_STATUS),
+	       1U << vmid);
+
+	/* Mapping vmid to pasid also for IH block */
+	WREG32(SOC15_REG_OFFSET(OSSSYS, 0, mmIH_VMID_0_LUT) + vmid,
+	       pasid_mapping);
+
+	WREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID16_PASID_MAPPING) + vmid,
+	       pasid_mapping);
+
+	while (!(RREG32(SOC15_REG_OFFSET(
+				ATHUB, 0,
+				mmATC_VMID_PASID_MAPPING_UPDATE_STATUS)) &
+		 (1U << (vmid + 16))))
+		cpu_relax();
+
+	WREG32(SOC15_REG_OFFSET(ATHUB, 0,
+				mmATC_VMID_PASID_MAPPING_UPDATE_STATUS),
+	       1U << (vmid + 16));
+
+	/* Mapping vmid to pasid also for IH block */
+	WREG32(SOC15_REG_OFFSET(OSSSYS, 0, mmIH_VMID_0_LUT_MM) + vmid,
+	       pasid_mapping);
+	return 0;
+}
+
+/* TODO - RING0 form of field is obsolete, seems to date back to SI
+ * but still works
+ */
+
+static int kgd_init_interrupts(struct kgd_dev *kgd, uint32_t pipe_id)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	uint32_t mec;
+	uint32_t pipe;
+
+	mec = (pipe_id / adev->gfx.mec.num_pipe_per_mec) + 1;
+	pipe = (pipe_id % adev->gfx.mec.num_pipe_per_mec);
+
+	lock_srbm(kgd, mec, pipe, 0, 0);
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmCPC_INT_CNTL),
+		CP_INT_CNTL_RING0__TIME_STAMP_INT_ENABLE_MASK |
+		CP_INT_CNTL_RING0__OPCODE_ERROR_INT_ENABLE_MASK);
+
+	unlock_srbm(kgd);
+
+	return 0;
+}
+
+static uint32_t get_sdma_base_addr(struct amdgpu_device *adev,
+				unsigned int engine_id,
+				unsigned int queue_id)
+{
+	uint32_t base[2] = {
+		SOC15_REG_OFFSET(SDMA0, 0,
+				 mmSDMA0_RLC0_RB_CNTL) - mmSDMA0_RLC0_RB_CNTL,
+		SOC15_REG_OFFSET(SDMA1, 0,
+				 mmSDMA1_RLC0_RB_CNTL) - mmSDMA1_RLC0_RB_CNTL
+	};
+	uint32_t retval;
+
+	retval = base[engine_id] + queue_id * (mmSDMA0_RLC1_RB_CNTL -
+					       mmSDMA0_RLC0_RB_CNTL);
+
+	pr_debug("sdma base address: 0x%x\n", retval);
+
+	return retval;
+}
+
+static inline struct v9_mqd *get_mqd(void *mqd)
+{
+	return (struct v9_mqd *)mqd;
+}
+
+static inline struct v9_sdma_mqd *get_sdma_mqd(void *mqd)
+{
+	return (struct v9_sdma_mqd *)mqd;
+}
+
+static int kgd_hqd_load(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
+			uint32_t queue_id, uint32_t __user *wptr,
+			uint32_t wptr_shift, uint32_t wptr_mask,
+			struct mm_struct *mm)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	struct v9_mqd *m;
+	uint32_t *mqd_hqd;
+	uint32_t reg, hqd_base, data;
+
+	m = get_mqd(mqd);
+
+	acquire_queue(kgd, pipe_id, queue_id);
+
+	/* HIQ is set during driver init period with vmid set to 0*/
+	if (m->cp_hqd_vmid == 0) {
+		uint32_t value, mec, pipe;
+
+		mec = (pipe_id / adev->gfx.mec.num_pipe_per_mec) + 1;
+		pipe = (pipe_id % adev->gfx.mec.num_pipe_per_mec);
+
+		pr_debug("kfd: set HIQ, mec:%d, pipe:%d, queue:%d.\n",
+			mec, pipe, queue_id);
+		value = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CP_SCHEDULERS));
+		value = REG_SET_FIELD(value, RLC_CP_SCHEDULERS, scheduler1,
+			((mec << 5) | (pipe << 3) | queue_id | 0x80));
+		WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_CP_SCHEDULERS), value);
+	}
+
+	/* HQD registers extend from CP_MQD_BASE_ADDR to CP_HQD_EOP_WPTR_MEM. */
+	mqd_hqd = &m->cp_mqd_base_addr_lo;
+	hqd_base = SOC15_REG_OFFSET(GC, 0, mmCP_MQD_BASE_ADDR);
+
+	for (reg = hqd_base;
+	     reg <= SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_HI); reg++)
+		WREG32(reg, mqd_hqd[reg - hqd_base]);
+
+
+	/* Activate doorbell logic before triggering WPTR poll. */
+	data = REG_SET_FIELD(m->cp_hqd_pq_doorbell_control,
+			     CP_HQD_PQ_DOORBELL_CONTROL, DOORBELL_EN, 1);
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_DOORBELL_CONTROL), data);
+
+	if (wptr) {
+		/* Don't read wptr with get_user because the user
+		 * context may not be accessible (if this function
+		 * runs in a work queue). Instead trigger a one-shot
+		 * polling read from memory in the CP. This assumes
+		 * that wptr is GPU-accessible in the queue's VMID via
+		 * ATC or SVM. WPTR==RPTR before starting the poll so
+		 * the CP starts fetching new commands from the right
+		 * place.
+		 *
+		 * Guessing a 64-bit WPTR from a 32-bit RPTR is a bit
+		 * tricky. Assume that the queue didn't overflow. The
+		 * number of valid bits in the 32-bit RPTR depends on
+		 * the queue size. The remaining bits are taken from
+		 * the saved 64-bit WPTR. If the WPTR wrapped, add the
+		 * queue size.
+		 */
+		uint32_t queue_size =
+			2 << REG_GET_FIELD(m->cp_hqd_pq_control,
+					   CP_HQD_PQ_CONTROL, QUEUE_SIZE);
+		uint64_t guessed_wptr = m->cp_hqd_pq_rptr & (queue_size - 1);
+
+		if ((m->cp_hqd_pq_wptr_lo & (queue_size - 1)) < guessed_wptr)
+			guessed_wptr += queue_size;
+		guessed_wptr += m->cp_hqd_pq_wptr_lo & ~(queue_size - 1);
+		guessed_wptr += (uint64_t)m->cp_hqd_pq_wptr_hi << 32;
+
+		WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_LO),
+		       lower_32_bits(guessed_wptr));
+		WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_HI),
+		       upper_32_bits(guessed_wptr));
+		WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR),
+		       lower_32_bits((uintptr_t)wptr));
+		WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_POLL_ADDR_HI),
+		       upper_32_bits((uintptr_t)wptr));
+		WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_PQ_WPTR_POLL_CNTL1),
+		       get_queue_mask(adev, pipe_id, queue_id));
+	}
+
+	/* Start the EOP fetcher */
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_EOP_RPTR),
+	       REG_SET_FIELD(m->cp_hqd_eop_rptr,
+			     CP_HQD_EOP_RPTR, INIT_FETCHER, 1));
+
+	data = REG_SET_FIELD(m->cp_hqd_active, CP_HQD_ACTIVE, ACTIVE, 1);
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE), data);
+
+	release_queue(kgd);
+
+	return 0;
+}
+
+static int kgd_hqd_dump(struct kgd_dev *kgd,
+			uint32_t pipe_id, uint32_t queue_id,
+			uint32_t (**dump)[2], uint32_t *n_regs)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	uint32_t i = 0, reg;
+#define HQD_N_REGS 56
+#define DUMP_REG(addr) do {				\
+		if (WARN_ON_ONCE(i >= HQD_N_REGS))	\
+			break;				\
+		(*dump)[i][0] = (addr) << 2;		\
+		(*dump)[i++][1] = RREG32(addr);		\
+	} while (0)
+
+	*dump = kmalloc_array(HQD_N_REGS * 2, sizeof(uint32_t), GFP_KERNEL);
+	if (*dump == NULL)
+		return -ENOMEM;
+
+	acquire_queue(kgd, pipe_id, queue_id);
+
+	for (reg = SOC15_REG_OFFSET(GC, 0, mmCP_MQD_BASE_ADDR);
+	     reg <= SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_WPTR_HI); reg++)
+		DUMP_REG(reg);
+
+	release_queue(kgd);
+
+	WARN_ON_ONCE(i != HQD_N_REGS);
+	*n_regs = i;
+
+	return 0;
+}
+
+static int kgd_hqd_sdma_load(struct kgd_dev *kgd, void *mqd,
+			     uint32_t __user *wptr, struct mm_struct *mm)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	struct v9_sdma_mqd *m;
+	uint32_t sdma_base_addr, sdmax_gfx_context_cntl;
+	unsigned long end_jiffies;
+	uint32_t data;
+	uint64_t data64;
+	uint64_t __user *wptr64 = (uint64_t __user *)wptr;
+
+	m = get_sdma_mqd(mqd);
+	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+					    m->sdma_queue_id);
+	sdmax_gfx_context_cntl = m->sdma_engine_id ?
+		SOC15_REG_OFFSET(SDMA1, 0, mmSDMA1_GFX_CONTEXT_CNTL) :
+		SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_GFX_CONTEXT_CNTL);
+
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL,
+		m->sdmax_rlcx_rb_cntl & (~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK));
+
+	end_jiffies = msecs_to_jiffies(2000) + jiffies;
+	while (true) {
+		data = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS);
+		if (data & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK)
+			break;
+		if (time_after(jiffies, end_jiffies))
+			return -ETIME;
+		usleep_range(500, 1000);
+	}
+	data = RREG32(sdmax_gfx_context_cntl);
+	data = REG_SET_FIELD(data, SDMA0_GFX_CONTEXT_CNTL,
+			     RESUME_CTX, 0);
+	WREG32(sdmax_gfx_context_cntl, data);
+
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL_OFFSET,
+	       m->sdmax_rlcx_doorbell_offset);
+
+	data = REG_SET_FIELD(m->sdmax_rlcx_doorbell, SDMA0_RLC0_DOORBELL,
+			     ENABLE, 1);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, data);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR, m->sdmax_rlcx_rb_rptr);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_HI,
+				m->sdmax_rlcx_rb_rptr_hi);
+
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 1);
+	if (read_user_wptr(mm, wptr64, data64)) {
+		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR,
+		       lower_32_bits(data64));
+		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR_HI,
+		       upper_32_bits(data64));
+	} else {
+		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR,
+		       m->sdmax_rlcx_rb_rptr);
+		WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_WPTR_HI,
+		       m->sdmax_rlcx_rb_rptr_hi);
+	}
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_MINOR_PTR_UPDATE, 0);
+
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE, m->sdmax_rlcx_rb_base);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_BASE_HI,
+			m->sdmax_rlcx_rb_base_hi);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_LO,
+			m->sdmax_rlcx_rb_rptr_addr_lo);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_ADDR_HI,
+			m->sdmax_rlcx_rb_rptr_addr_hi);
+
+	data = REG_SET_FIELD(m->sdmax_rlcx_rb_cntl, SDMA0_RLC0_RB_CNTL,
+			     RB_ENABLE, 1);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, data);
+
+	return 0;
+}
+
+static int kgd_hqd_sdma_dump(struct kgd_dev *kgd,
+			     uint32_t engine_id, uint32_t queue_id,
+			     uint32_t (**dump)[2], uint32_t *n_regs)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	uint32_t sdma_base_addr = get_sdma_base_addr(adev, engine_id, queue_id);
+	uint32_t i = 0, reg;
+#undef HQD_N_REGS
+#define HQD_N_REGS (19+6+7+10)
+
+	*dump = kmalloc_array(HQD_N_REGS * 2, sizeof(uint32_t), GFP_KERNEL);
+	if (*dump == NULL)
+		return -ENOMEM;
+
+	for (reg = mmSDMA0_RLC0_RB_CNTL; reg <= mmSDMA0_RLC0_DOORBELL; reg++)
+		DUMP_REG(sdma_base_addr + reg);
+	for (reg = mmSDMA0_RLC0_STATUS; reg <= mmSDMA0_RLC0_CSA_ADDR_HI; reg++)
+		DUMP_REG(sdma_base_addr + reg);
+	for (reg = mmSDMA0_RLC0_IB_SUB_REMAIN;
+	     reg <= mmSDMA0_RLC0_MINOR_PTR_UPDATE; reg++)
+		DUMP_REG(sdma_base_addr + reg);
+	for (reg = mmSDMA0_RLC0_MIDCMD_DATA0;
+	     reg <= mmSDMA0_RLC0_MIDCMD_CNTL; reg++)
+		DUMP_REG(sdma_base_addr + reg);
+
+	WARN_ON_ONCE(i != HQD_N_REGS);
+	*n_regs = i;
+
+	return 0;
+}
+
+static bool kgd_hqd_is_occupied(struct kgd_dev *kgd, uint64_t queue_address,
+				uint32_t pipe_id, uint32_t queue_id)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	uint32_t act;
+	bool retval = false;
+	uint32_t low, high;
+
+	acquire_queue(kgd, pipe_id, queue_id);
+	act = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE));
+	if (act) {
+		low = lower_32_bits(queue_address >> 8);
+		high = upper_32_bits(queue_address >> 8);
+
+		if (low == RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_BASE)) &&
+		   high == RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_PQ_BASE_HI)))
+			retval = true;
+	}
+	release_queue(kgd);
+	return retval;
+}
+
+static bool kgd_hqd_sdma_is_occupied(struct kgd_dev *kgd, void *mqd)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	struct v9_sdma_mqd *m;
+	uint32_t sdma_base_addr;
+	uint32_t sdma_rlc_rb_cntl;
+
+	m = get_sdma_mqd(mqd);
+	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+					    m->sdma_queue_id);
+
+	sdma_rlc_rb_cntl = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL);
+
+	if (sdma_rlc_rb_cntl & SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK)
+		return true;
+
+	return false;
+}
+
+static int kgd_hqd_destroy(struct kgd_dev *kgd, void *mqd,
+				enum kfd_preempt_type reset_type,
+				unsigned int utimeout, uint32_t pipe_id,
+				uint32_t queue_id)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	enum hqd_dequeue_request_type type;
+	unsigned long end_jiffies;
+	uint32_t temp;
+	struct v9_mqd *m = get_mqd(mqd);
+
+	acquire_queue(kgd, pipe_id, queue_id);
+
+	if (m->cp_hqd_vmid == 0)
+		WREG32_FIELD15(GC, 0, RLC_CP_SCHEDULERS, scheduler1, 0);
+
+	switch (reset_type) {
+	case KFD_PREEMPT_TYPE_WAVEFRONT_DRAIN:
+		type = DRAIN_PIPE;
+		break;
+	case KFD_PREEMPT_TYPE_WAVEFRONT_RESET:
+		type = RESET_WAVES;
+		break;
+	default:
+		type = DRAIN_PIPE;
+		break;
+	}
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_DEQUEUE_REQUEST), type);
+
+	end_jiffies = (utimeout * HZ / 1000) + jiffies;
+	while (true) {
+		temp = RREG32(SOC15_REG_OFFSET(GC, 0, mmCP_HQD_ACTIVE));
+		if (!(temp & CP_HQD_ACTIVE__ACTIVE_MASK))
+			break;
+		if (time_after(jiffies, end_jiffies)) {
+			pr_err("cp queue preemption time out.\n");
+			release_queue(kgd);
+			return -ETIME;
+		}
+		usleep_range(500, 1000);
+	}
+
+	release_queue(kgd);
+	return 0;
+}
+
+static int kgd_hqd_sdma_destroy(struct kgd_dev *kgd, void *mqd,
+				unsigned int utimeout)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	struct v9_sdma_mqd *m;
+	uint32_t sdma_base_addr;
+	uint32_t temp;
+	unsigned long end_jiffies = (utimeout * HZ / 1000) + jiffies;
+
+	m = get_sdma_mqd(mqd);
+	sdma_base_addr = get_sdma_base_addr(adev, m->sdma_engine_id,
+					    m->sdma_queue_id);
+
+	temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL);
+	temp = temp & ~SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK;
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL, temp);
+
+	while (true) {
+		temp = RREG32(sdma_base_addr + mmSDMA0_RLC0_CONTEXT_STATUS);
+		if (temp & SDMA0_RLC0_CONTEXT_STATUS__IDLE_MASK)
+			break;
+		if (time_after(jiffies, end_jiffies))
+			return -ETIME;
+		usleep_range(500, 1000);
+	}
+
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_DOORBELL, 0);
+	WREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL,
+		RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_CNTL) |
+		SDMA0_RLC0_RB_CNTL__RB_ENABLE_MASK);
+
+	m->sdmax_rlcx_rb_rptr = RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR);
+	m->sdmax_rlcx_rb_rptr_hi =
+		RREG32(sdma_base_addr + mmSDMA0_RLC0_RB_RPTR_HI);
+
+	return 0;
+}
+
+static bool get_atc_vmid_pasid_mapping_valid(struct kgd_dev *kgd,
+							uint8_t vmid)
+{
+	uint32_t reg;
+	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+
+	reg = RREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID0_PASID_MAPPING)
+		     + vmid);
+	return reg & ATC_VMID0_PASID_MAPPING__VALID_MASK;
+}
+
+static uint16_t get_atc_vmid_pasid_mapping_pasid(struct kgd_dev *kgd,
+								uint8_t vmid)
+{
+	uint32_t reg;
+	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+
+	reg = RREG32(SOC15_REG_OFFSET(ATHUB, 0, mmATC_VMID0_PASID_MAPPING)
+		     + vmid);
+	return reg & ATC_VMID0_PASID_MAPPING__PASID_MASK;
+}
+
+static void write_vmid_invalidate_request(struct kgd_dev *kgd, uint8_t vmid)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+	uint32_t req = (1 << vmid) |
+		(0 << VM_INVALIDATE_ENG16_REQ__FLUSH_TYPE__SHIFT) | /* legacy */
+		VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PTES_MASK |
+		VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PDE0_MASK |
+		VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PDE1_MASK |
+		VM_INVALIDATE_ENG16_REQ__INVALIDATE_L2_PDE2_MASK |
+		VM_INVALIDATE_ENG16_REQ__INVALIDATE_L1_PTES_MASK;
+
+	mutex_lock(&adev->srbm_mutex);
+
+	/* Use legacy mode tlb invalidation.
+	 *
+	 * Currently on Raven the code below is broken for anything but
+	 * legacy mode due to a MMHUB power gating problem. A workaround
+	 * is for MMHUB to wait until the condition PER_VMID_INVALIDATE_REQ
+	 * == PER_VMID_INVALIDATE_ACK instead of simply waiting for the ack
+	 * bit.
+	 *
+	 * TODO 1: agree on the right set of invalidation registers for
+	 * KFD use. Use the last one for now. Invalidate both GC and
+	 * MMHUB.
+	 *
+	 * TODO 2: support range-based invalidation, requires kfg2kgd
+	 * interface change
+	 */
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_INVALIDATE_ENG16_ADDR_RANGE_LO32),
+				0xffffffff);
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_INVALIDATE_ENG16_ADDR_RANGE_HI32),
+				0x0000001f);
+
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0,
+				mmMMHUB_VM_INVALIDATE_ENG16_ADDR_RANGE_LO32),
+				0xffffffff);
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0,
+				mmMMHUB_VM_INVALIDATE_ENG16_ADDR_RANGE_HI32),
+				0x0000001f);
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_INVALIDATE_ENG16_REQ), req);
+
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmMMHUB_VM_INVALIDATE_ENG16_REQ),
+				req);
+
+	while (!(RREG32(SOC15_REG_OFFSET(GC, 0, mmVM_INVALIDATE_ENG16_ACK)) &
+					(1 << vmid)))
+		cpu_relax();
+
+	while (!(RREG32(SOC15_REG_OFFSET(MMHUB, 0,
+					mmMMHUB_VM_INVALIDATE_ENG16_ACK)) &
+					(1 << vmid)))
+		cpu_relax();
+
+	mutex_unlock(&adev->srbm_mutex);
+
+}
+
+static int invalidate_tlbs_with_kiq(struct amdgpu_device *adev, uint16_t pasid)
+{
+	signed long r;
+	uint32_t seq;
+	struct amdgpu_ring *ring = &adev->gfx.kiq.ring;
+
+	spin_lock(&adev->gfx.kiq.ring_lock);
+	amdgpu_ring_alloc(ring, 12); /* fence + invalidate_tlbs package*/
+	amdgpu_ring_write(ring, PACKET3(PACKET3_INVALIDATE_TLBS, 0));
+	amdgpu_ring_write(ring,
+			PACKET3_INVALIDATE_TLBS_DST_SEL(1) |
+			PACKET3_INVALIDATE_TLBS_ALL_HUB(1) |
+			PACKET3_INVALIDATE_TLBS_PASID(pasid) |
+			PACKET3_INVALIDATE_TLBS_FLUSH_TYPE(0)); /* legacy */
+	amdgpu_fence_emit_polling(ring, &seq);
+	amdgpu_ring_commit(ring);
+	spin_unlock(&adev->gfx.kiq.ring_lock);
+
+	r = amdgpu_fence_wait_polling(ring, seq, adev->usec_timeout);
+	if (r < 1) {
+		DRM_ERROR("wait for kiq fence error: %ld.\n", r);
+		return -ETIME;
+	}
+
+	return 0;
+}
+
+static int invalidate_tlbs(struct kgd_dev *kgd, uint16_t pasid)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+	int vmid;
+	struct amdgpu_ring *ring = &adev->gfx.kiq.ring;
+
+	if (ring->ready)
+		return invalidate_tlbs_with_kiq(adev, pasid);
+
+	for (vmid = 0; vmid < 16; vmid++) {
+		if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid))
+			continue;
+		if (get_atc_vmid_pasid_mapping_valid(kgd, vmid)) {
+			if (get_atc_vmid_pasid_mapping_pasid(kgd, vmid)
+				== pasid) {
+				write_vmid_invalidate_request(kgd, vmid);
+				break;
+			}
+		}
+	}
+
+	return 0;
+}
+
+static int invalidate_tlbs_vmid(struct kgd_dev *kgd, uint16_t vmid)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+
+	if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) {
+		pr_err("non kfd vmid %d\n", vmid);
+		return 0;
+	}
+
+	write_vmid_invalidate_request(kgd, vmid);
+	return 0;
+}
+
+static int kgd_address_watch_disable(struct kgd_dev *kgd)
+{
+	return 0;
+}
+
+static int kgd_address_watch_execute(struct kgd_dev *kgd,
+					unsigned int watch_point_id,
+					uint32_t cntl_val,
+					uint32_t addr_hi,
+					uint32_t addr_lo)
+{
+	return 0;
+}
+
+static int kgd_wave_control_execute(struct kgd_dev *kgd,
+					uint32_t gfx_index_val,
+					uint32_t sq_cmd)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	uint32_t data = 0;
+
+	mutex_lock(&adev->grbm_idx_mutex);
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_GFX_INDEX), gfx_index_val);
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmSQ_CMD), sq_cmd);
+
+	data = REG_SET_FIELD(data, GRBM_GFX_INDEX,
+		INSTANCE_BROADCAST_WRITES, 1);
+	data = REG_SET_FIELD(data, GRBM_GFX_INDEX,
+		SH_BROADCAST_WRITES, 1);
+	data = REG_SET_FIELD(data, GRBM_GFX_INDEX,
+		SE_BROADCAST_WRITES, 1);
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmGRBM_GFX_INDEX), data);
+	mutex_unlock(&adev->grbm_idx_mutex);
+
+	return 0;
+}
+
+static uint32_t kgd_address_watch_get_offset(struct kgd_dev *kgd,
+					unsigned int watch_point_id,
+					unsigned int reg_offset)
+{
+	return 0;
+}
+
+static void set_scratch_backing_va(struct kgd_dev *kgd,
+					uint64_t va, uint32_t vmid)
+{
+	/* No longer needed on GFXv9. The scratch base address is
+	 * passed to the shader by the CP. It's the user mode driver's
+	 * responsibility.
+	 */
+}
+
+/* FIXME: Does this need to be ASIC-specific code? */
+static uint16_t get_fw_version(struct kgd_dev *kgd, enum kgd_engine_type type)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *) kgd;
+	const union amdgpu_firmware_header *hdr;
+
+	switch (type) {
+	case KGD_ENGINE_PFP:
+		hdr = (const union amdgpu_firmware_header *)adev->gfx.pfp_fw->data;
+		break;
+
+	case KGD_ENGINE_ME:
+		hdr = (const union amdgpu_firmware_header *)adev->gfx.me_fw->data;
+		break;
+
+	case KGD_ENGINE_CE:
+		hdr = (const union amdgpu_firmware_header *)adev->gfx.ce_fw->data;
+		break;
+
+	case KGD_ENGINE_MEC1:
+		hdr = (const union amdgpu_firmware_header *)adev->gfx.mec_fw->data;
+		break;
+
+	case KGD_ENGINE_MEC2:
+		hdr = (const union amdgpu_firmware_header *)adev->gfx.mec2_fw->data;
+		break;
+
+	case KGD_ENGINE_RLC:
+		hdr = (const union amdgpu_firmware_header *)adev->gfx.rlc_fw->data;
+		break;
+
+	case KGD_ENGINE_SDMA1:
+		hdr = (const union amdgpu_firmware_header *)adev->sdma.instance[0].fw->data;
+		break;
+
+	case KGD_ENGINE_SDMA2:
+		hdr = (const union amdgpu_firmware_header *)adev->sdma.instance[1].fw->data;
+		break;
+
+	default:
+		return 0;
+	}
+
+	if (hdr == NULL)
+		return 0;
+
+	/* Only 12 bit in use*/
+	return hdr->common.ucode_version;
+}
+
+static void set_vm_context_page_table_base(struct kgd_dev *kgd, uint32_t vmid,
+		uint32_t page_table_base)
+{
+	struct amdgpu_device *adev = get_amdgpu_device(kgd);
+	uint64_t base = (uint64_t)page_table_base << PAGE_SHIFT |
+		AMDGPU_PTE_VALID;
+
+	if (!amdgpu_amdkfd_is_kfd_vmid(adev, vmid)) {
+		pr_err("trying to set page table base for wrong VMID %u\n",
+		       vmid);
+		return;
+	}
+
+	/* TODO: take advantage of per-process address space size. For
+	 * now, all processes share the same address space size, like
+	 * on GFX8 and older.
+	 */
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmMMHUB_VM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32) + (vmid*2), 0);
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmMMHUB_VM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32) + (vmid*2), 0);
+
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmMMHUB_VM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32) + (vmid*2),
+			lower_32_bits(adev->vm_manager.max_pfn - 1));
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmMMHUB_VM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32) + (vmid*2),
+			upper_32_bits(adev->vm_manager.max_pfn - 1));
+
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmMMHUB_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32) + (vmid*2), lower_32_bits(base));
+	WREG32(SOC15_REG_OFFSET(MMHUB, 0, mmMMHUB_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32) + (vmid*2), upper_32_bits(base));
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_START_ADDR_LO32) + (vmid*2), 0);
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_START_ADDR_HI32) + (vmid*2), 0);
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_END_ADDR_LO32) + (vmid*2),
+			lower_32_bits(adev->vm_manager.max_pfn - 1));
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_END_ADDR_HI32) + (vmid*2),
+			upper_32_bits(adev->vm_manager.max_pfn - 1));
+
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32) + (vmid*2), lower_32_bits(base));
+	WREG32(SOC15_REG_OFFSET(GC, 0, mmVM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32) + (vmid*2), upper_32_bits(base));
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
index 1d6e1479da38..079af8ac2636 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
@@ -23,6 +23,8 @@
 #define pr_fmt(fmt) "kfd2kgd: " fmt
 
 #include <linux/list.h>
+#include <linux/pagemap.h>
+#include <linux/sched/mm.h>
 #include <drm/drmP.h>
 #include "amdgpu_object.h"
 #include "amdgpu_vm.h"
@@ -33,10 +35,20 @@
  */
 #define VI_BO_SIZE_ALIGN (0x8000)
 
+/* BO flag to indicate a KFD userptr BO */
+#define AMDGPU_AMDKFD_USERPTR_BO (1ULL << 63)
+
+/* Userptr restore delay, just long enough to allow consecutive VM
+ * changes to accumulate
+ */
+#define AMDGPU_USERPTR_RESTORE_DELAY_MS 1
+
 /* Impose limit on how much memory KFD can use */
 static struct {
 	uint64_t max_system_mem_limit;
+	uint64_t max_userptr_mem_limit;
 	int64_t system_mem_used;
+	int64_t userptr_mem_used;
 	spinlock_t mem_limit_lock;
 } kfd_mem_limit;
 
@@ -57,6 +69,7 @@ static const char * const domain_bit_to_string[] = {
 
 #define domain_string(domain) domain_bit_to_string[ffs(domain)-1]
 
+static void amdgpu_amdkfd_restore_userptr_worker(struct work_struct *work);
 
 
 static inline struct amdgpu_device *get_amdgpu_device(struct kgd_dev *kgd)
@@ -78,6 +91,7 @@ static bool check_if_add_bo_to_vm(struct amdgpu_vm *avm,
 
 /* Set memory usage limits. Current, limits are
  *  System (kernel) memory - 3/8th System RAM
+ *  Userptr memory - 3/4th System RAM
  */
 void amdgpu_amdkfd_gpuvm_init_mem_limits(void)
 {
@@ -90,8 +104,10 @@ void amdgpu_amdkfd_gpuvm_init_mem_limits(void)
 
 	spin_lock_init(&kfd_mem_limit.mem_limit_lock);
 	kfd_mem_limit.max_system_mem_limit = (mem >> 1) - (mem >> 3);
-	pr_debug("Kernel memory limit %lluM\n",
-		(kfd_mem_limit.max_system_mem_limit >> 20));
+	kfd_mem_limit.max_userptr_mem_limit = mem - (mem >> 2);
+	pr_debug("Kernel memory limit %lluM, userptr limit %lluM\n",
+		(kfd_mem_limit.max_system_mem_limit >> 20),
+		(kfd_mem_limit.max_userptr_mem_limit >> 20));
 }
 
 static int amdgpu_amdkfd_reserve_system_mem_limit(struct amdgpu_device *adev,
@@ -111,6 +127,16 @@ static int amdgpu_amdkfd_reserve_system_mem_limit(struct amdgpu_device *adev,
 			goto err_no_mem;
 		}
 		kfd_mem_limit.system_mem_used += (acc_size + size);
+	} else if (domain == AMDGPU_GEM_DOMAIN_CPU) {
+		if ((kfd_mem_limit.system_mem_used + acc_size >
+			kfd_mem_limit.max_system_mem_limit) ||
+			(kfd_mem_limit.userptr_mem_used + (size + acc_size) >
+			kfd_mem_limit.max_userptr_mem_limit)) {
+			ret = -ENOMEM;
+			goto err_no_mem;
+		}
+		kfd_mem_limit.system_mem_used += acc_size;
+		kfd_mem_limit.userptr_mem_used += size;
 	}
 err_no_mem:
 	spin_unlock(&kfd_mem_limit.mem_limit_lock);
@@ -126,10 +152,16 @@ static void unreserve_system_mem_limit(struct amdgpu_device *adev,
 				       sizeof(struct amdgpu_bo));
 
 	spin_lock(&kfd_mem_limit.mem_limit_lock);
-	if (domain == AMDGPU_GEM_DOMAIN_GTT)
+	if (domain == AMDGPU_GEM_DOMAIN_GTT) {
 		kfd_mem_limit.system_mem_used -= (acc_size + size);
+	} else if (domain == AMDGPU_GEM_DOMAIN_CPU) {
+		kfd_mem_limit.system_mem_used -= acc_size;
+		kfd_mem_limit.userptr_mem_used -= size;
+	}
 	WARN_ONCE(kfd_mem_limit.system_mem_used < 0,
 		  "kfd system memory accounting unbalanced");
+	WARN_ONCE(kfd_mem_limit.userptr_mem_used < 0,
+		  "kfd userptr memory accounting unbalanced");
 
 	spin_unlock(&kfd_mem_limit.mem_limit_lock);
 }
@@ -138,12 +170,17 @@ void amdgpu_amdkfd_unreserve_system_memory_limit(struct amdgpu_bo *bo)
 {
 	spin_lock(&kfd_mem_limit.mem_limit_lock);
 
-	if (bo->preferred_domains == AMDGPU_GEM_DOMAIN_GTT) {
+	if (bo->flags & AMDGPU_AMDKFD_USERPTR_BO) {
+		kfd_mem_limit.system_mem_used -= bo->tbo.acc_size;
+		kfd_mem_limit.userptr_mem_used -= amdgpu_bo_size(bo);
+	} else if (bo->preferred_domains == AMDGPU_GEM_DOMAIN_GTT) {
 		kfd_mem_limit.system_mem_used -=
 			(bo->tbo.acc_size + amdgpu_bo_size(bo));
 	}
 	WARN_ONCE(kfd_mem_limit.system_mem_used < 0,
 		  "kfd system memory accounting unbalanced");
+	WARN_ONCE(kfd_mem_limit.userptr_mem_used < 0,
+		  "kfd userptr memory accounting unbalanced");
 
 	spin_unlock(&kfd_mem_limit.mem_limit_lock);
 }
@@ -506,7 +543,8 @@ static void remove_bo_from_vm(struct amdgpu_device *adev,
 }
 
 static void add_kgd_mem_to_kfd_bo_list(struct kgd_mem *mem,
-				struct amdkfd_process_info *process_info)
+				struct amdkfd_process_info *process_info,
+				bool userptr)
 {
 	struct ttm_validate_buffer *entry = &mem->validate_list;
 	struct amdgpu_bo *bo = mem->bo;
@@ -515,10 +553,95 @@ static void add_kgd_mem_to_kfd_bo_list(struct kgd_mem *mem,
 	entry->shared = true;
 	entry->bo = &bo->tbo;
 	mutex_lock(&process_info->lock);
-	list_add_tail(&entry->head, &process_info->kfd_bo_list);
+	if (userptr)
+		list_add_tail(&entry->head, &process_info->userptr_valid_list);
+	else
+		list_add_tail(&entry->head, &process_info->kfd_bo_list);
 	mutex_unlock(&process_info->lock);
 }
 
+/* Initializes user pages. It registers the MMU notifier and validates
+ * the userptr BO in the GTT domain.
+ *
+ * The BO must already be on the userptr_valid_list. Otherwise an
+ * eviction and restore may happen that leaves the new BO unmapped
+ * with the user mode queues running.
+ *
+ * Takes the process_info->lock to protect against concurrent restore
+ * workers.
+ *
+ * Returns 0 for success, negative errno for errors.
+ */
+static int init_user_pages(struct kgd_mem *mem, struct mm_struct *mm,
+			   uint64_t user_addr)
+{
+	struct amdkfd_process_info *process_info = mem->process_info;
+	struct amdgpu_bo *bo = mem->bo;
+	struct ttm_operation_ctx ctx = { true, false };
+	int ret = 0;
+
+	mutex_lock(&process_info->lock);
+
+	ret = amdgpu_ttm_tt_set_userptr(bo->tbo.ttm, user_addr, 0);
+	if (ret) {
+		pr_err("%s: Failed to set userptr: %d\n", __func__, ret);
+		goto out;
+	}
+
+	ret = amdgpu_mn_register(bo, user_addr);
+	if (ret) {
+		pr_err("%s: Failed to register MMU notifier: %d\n",
+		       __func__, ret);
+		goto out;
+	}
+
+	/* If no restore worker is running concurrently, user_pages
+	 * should not be allocated
+	 */
+	WARN(mem->user_pages, "Leaking user_pages array");
+
+	mem->user_pages = kvmalloc_array(bo->tbo.ttm->num_pages,
+					   sizeof(struct page *),
+					   GFP_KERNEL | __GFP_ZERO);
+	if (!mem->user_pages) {
+		pr_err("%s: Failed to allocate pages array\n", __func__);
+		ret = -ENOMEM;
+		goto unregister_out;
+	}
+
+	ret = amdgpu_ttm_tt_get_user_pages(bo->tbo.ttm, mem->user_pages);
+	if (ret) {
+		pr_err("%s: Failed to get user pages: %d\n", __func__, ret);
+		goto free_out;
+	}
+
+	amdgpu_ttm_tt_set_user_pages(bo->tbo.ttm, mem->user_pages);
+
+	ret = amdgpu_bo_reserve(bo, true);
+	if (ret) {
+		pr_err("%s: Failed to reserve BO\n", __func__);
+		goto release_out;
+	}
+	amdgpu_ttm_placement_from_domain(bo, mem->domain);
+	ret = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
+	if (ret)
+		pr_err("%s: failed to validate BO\n", __func__);
+	amdgpu_bo_unreserve(bo);
+
+release_out:
+	if (ret)
+		release_pages(mem->user_pages, bo->tbo.ttm->num_pages);
+free_out:
+	kvfree(mem->user_pages);
+	mem->user_pages = NULL;
+unregister_out:
+	if (ret)
+		amdgpu_mn_unregister(bo);
+out:
+	mutex_unlock(&process_info->lock);
+	return ret;
+}
+
 /* Reserving a BO and its page table BOs must happen atomically to
  * avoid deadlocks. Some operations update multiple VMs at once. Track
  * all the reservation info in a context structure. Optionally a sync
@@ -748,7 +871,8 @@ static int update_gpuvm_pte(struct amdgpu_device *adev,
 }
 
 static int map_bo_to_gpuvm(struct amdgpu_device *adev,
-		struct kfd_bo_va_list *entry, struct amdgpu_sync *sync)
+		struct kfd_bo_va_list *entry, struct amdgpu_sync *sync,
+		bool no_update_pte)
 {
 	int ret;
 
@@ -762,6 +886,9 @@ static int map_bo_to_gpuvm(struct amdgpu_device *adev,
 		return ret;
 	}
 
+	if (no_update_pte)
+		return 0;
+
 	ret = update_gpuvm_pte(adev, entry, sync);
 	if (ret) {
 		pr_err("update_gpuvm_pte() failed\n");
@@ -820,6 +947,8 @@ static int init_kfd_vm(struct amdgpu_vm *vm, void **process_info,
 		mutex_init(&info->lock);
 		INIT_LIST_HEAD(&info->vm_list_head);
 		INIT_LIST_HEAD(&info->kfd_bo_list);
+		INIT_LIST_HEAD(&info->userptr_valid_list);
+		INIT_LIST_HEAD(&info->userptr_inval_list);
 
 		info->eviction_fence =
 			amdgpu_amdkfd_fence_create(dma_fence_context_alloc(1),
@@ -830,6 +959,11 @@ static int init_kfd_vm(struct amdgpu_vm *vm, void **process_info,
 			goto create_evict_fence_fail;
 		}
 
+		info->pid = get_task_pid(current->group_leader, PIDTYPE_PID);
+		atomic_set(&info->evicted_bos, 0);
+		INIT_DELAYED_WORK(&info->restore_userptr_work,
+				  amdgpu_amdkfd_restore_userptr_worker);
+
 		*process_info = info;
 		*ef = dma_fence_get(&info->eviction_fence->base);
 	}
@@ -872,6 +1006,7 @@ reserve_pd_fail:
 		dma_fence_put(*ef);
 		*ef = NULL;
 		*process_info = NULL;
+		put_pid(info->pid);
 create_evict_fence_fail:
 		mutex_destroy(&info->lock);
 		kfree(info);
@@ -967,8 +1102,12 @@ void amdgpu_amdkfd_gpuvm_destroy_cb(struct amdgpu_device *adev,
 	/* Release per-process resources when last compute VM is destroyed */
 	if (!process_info->n_vms) {
 		WARN_ON(!list_empty(&process_info->kfd_bo_list));
+		WARN_ON(!list_empty(&process_info->userptr_valid_list));
+		WARN_ON(!list_empty(&process_info->userptr_inval_list));
 
 		dma_fence_put(&process_info->eviction_fence->base);
+		cancel_delayed_work_sync(&process_info->restore_userptr_work);
+		put_pid(process_info->pid);
 		mutex_destroy(&process_info->lock);
 		kfree(process_info);
 	}
@@ -1003,9 +1142,11 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 {
 	struct amdgpu_device *adev = get_amdgpu_device(kgd);
 	struct amdgpu_vm *avm = (struct amdgpu_vm *)vm;
+	uint64_t user_addr = 0;
 	struct amdgpu_bo *bo;
+	struct amdgpu_bo_param bp;
 	int byte_align;
-	u32 alloc_domain;
+	u32 domain, alloc_domain;
 	u64 alloc_flags;
 	uint32_t mapping_flags;
 	int ret;
@@ -1014,14 +1155,21 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 	 * Check on which domain to allocate BO
 	 */
 	if (flags & ALLOC_MEM_FLAGS_VRAM) {
-		alloc_domain = AMDGPU_GEM_DOMAIN_VRAM;
+		domain = alloc_domain = AMDGPU_GEM_DOMAIN_VRAM;
 		alloc_flags = AMDGPU_GEM_CREATE_VRAM_CLEARED;
 		alloc_flags |= (flags & ALLOC_MEM_FLAGS_PUBLIC) ?
 			AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED :
 			AMDGPU_GEM_CREATE_NO_CPU_ACCESS;
 	} else if (flags & ALLOC_MEM_FLAGS_GTT) {
-		alloc_domain = AMDGPU_GEM_DOMAIN_GTT;
+		domain = alloc_domain = AMDGPU_GEM_DOMAIN_GTT;
+		alloc_flags = 0;
+	} else if (flags & ALLOC_MEM_FLAGS_USERPTR) {
+		domain = AMDGPU_GEM_DOMAIN_GTT;
+		alloc_domain = AMDGPU_GEM_DOMAIN_CPU;
 		alloc_flags = 0;
+		if (!offset || !*offset)
+			return -EINVAL;
+		user_addr = *offset;
 	} else {
 		return -EINVAL;
 	}
@@ -1069,8 +1217,14 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 	pr_debug("\tcreate BO VA 0x%llx size 0x%llx domain %s\n",
 			va, size, domain_string(alloc_domain));
 
-	ret = amdgpu_bo_create(adev, size, byte_align,
-				alloc_domain, alloc_flags, ttm_bo_type_device, NULL, &bo);
+	memset(&bp, 0, sizeof(bp));
+	bp.size = size;
+	bp.byte_align = byte_align;
+	bp.domain = alloc_domain;
+	bp.flags = alloc_flags;
+	bp.type = ttm_bo_type_device;
+	bp.resv = NULL;
+	ret = amdgpu_bo_create(adev, &bp, &bo);
 	if (ret) {
 		pr_debug("Failed to create BO on domain %s. ret %d\n",
 				domain_string(alloc_domain), ret);
@@ -1078,18 +1232,34 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 	}
 	bo->kfd_bo = *mem;
 	(*mem)->bo = bo;
+	if (user_addr)
+		bo->flags |= AMDGPU_AMDKFD_USERPTR_BO;
 
 	(*mem)->va = va;
-	(*mem)->domain = alloc_domain;
+	(*mem)->domain = domain;
 	(*mem)->mapped_to_gpu_memory = 0;
 	(*mem)->process_info = avm->process_info;
-	add_kgd_mem_to_kfd_bo_list(*mem, avm->process_info);
+	add_kgd_mem_to_kfd_bo_list(*mem, avm->process_info, user_addr);
+
+	if (user_addr) {
+		ret = init_user_pages(*mem, current->mm, user_addr);
+		if (ret) {
+			mutex_lock(&avm->process_info->lock);
+			list_del(&(*mem)->validate_list.head);
+			mutex_unlock(&avm->process_info->lock);
+			goto allocate_init_user_pages_failed;
+		}
+	}
 
 	if (offset)
 		*offset = amdgpu_bo_mmap_offset(bo);
 
 	return 0;
 
+allocate_init_user_pages_failed:
+	amdgpu_bo_unref(&bo);
+	/* Don't unreserve system mem limit twice */
+	goto err_reserve_system_mem;
 err_bo_create:
 	unreserve_system_mem_limit(adev, size, alloc_domain);
 err_reserve_system_mem:
@@ -1122,12 +1292,24 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu(
 	 * be freed anyway
 	 */
 
+	/* No more MMU notifiers */
+	amdgpu_mn_unregister(mem->bo);
+
 	/* Make sure restore workers don't access the BO any more */
 	bo_list_entry = &mem->validate_list;
 	mutex_lock(&process_info->lock);
 	list_del(&bo_list_entry->head);
 	mutex_unlock(&process_info->lock);
 
+	/* Free user pages if necessary */
+	if (mem->user_pages) {
+		pr_debug("%s: Freeing user_pages array\n", __func__);
+		if (mem->user_pages[0])
+			release_pages(mem->user_pages,
+					mem->bo->tbo.ttm->num_pages);
+		kvfree(mem->user_pages);
+	}
+
 	ret = reserve_bo_and_cond_vms(mem, NULL, BO_VM_ALL, &ctx);
 	if (unlikely(ret))
 		return ret;
@@ -1173,21 +1355,32 @@ int amdgpu_amdkfd_gpuvm_map_memory_to_gpu(
 	struct kfd_bo_va_list *bo_va_entry = NULL;
 	struct kfd_bo_va_list *bo_va_entry_aql = NULL;
 	unsigned long bo_size;
-
-	/* Make sure restore is not running concurrently.
-	 */
-	mutex_lock(&mem->process_info->lock);
-
-	mutex_lock(&mem->lock);
+	bool is_invalid_userptr = false;
 
 	bo = mem->bo;
-
 	if (!bo) {
 		pr_err("Invalid BO when mapping memory to GPU\n");
-		ret = -EINVAL;
-		goto out;
+		return -EINVAL;
 	}
 
+	/* Make sure restore is not running concurrently. Since we
+	 * don't map invalid userptr BOs, we rely on the next restore
+	 * worker to do the mapping
+	 */
+	mutex_lock(&mem->process_info->lock);
+
+	/* Lock mmap-sem. If we find an invalid userptr BO, we can be
+	 * sure that the MMU notifier is no longer running
+	 * concurrently and the queues are actually stopped
+	 */
+	if (amdgpu_ttm_tt_get_usermm(bo->tbo.ttm)) {
+		down_write(&current->mm->mmap_sem);
+		is_invalid_userptr = atomic_read(&mem->invalid);
+		up_write(&current->mm->mmap_sem);
+	}
+
+	mutex_lock(&mem->lock);
+
 	domain = mem->domain;
 	bo_size = bo->tbo.mem.size;
 
@@ -1200,6 +1393,14 @@ int amdgpu_amdkfd_gpuvm_map_memory_to_gpu(
 	if (unlikely(ret))
 		goto out;
 
+	/* Userptr can be marked as "not invalid", but not actually be
+	 * validated yet (still in the system domain). In that case
+	 * the queues are still stopped and we can leave mapping for
+	 * the next restore worker
+	 */
+	if (bo->tbo.mem.mem_type == TTM_PL_SYSTEM)
+		is_invalid_userptr = true;
+
 	if (check_if_add_bo_to_vm(avm, mem)) {
 		ret = add_bo_to_vm(adev, mem, avm, false,
 				&bo_va_entry);
@@ -1217,7 +1418,8 @@ int amdgpu_amdkfd_gpuvm_map_memory_to_gpu(
 			goto add_bo_to_vm_failed;
 	}
 
-	if (mem->mapped_to_gpu_memory == 0) {
+	if (mem->mapped_to_gpu_memory == 0 &&
+	    !amdgpu_ttm_tt_get_usermm(bo->tbo.ttm)) {
 		/* Validate BO only once. The eviction fence gets added to BO
 		 * the first time it is mapped. Validate will wait for all
 		 * background evictions to complete.
@@ -1235,7 +1437,8 @@ int amdgpu_amdkfd_gpuvm_map_memory_to_gpu(
 					entry->va, entry->va + bo_size,
 					entry);
 
-			ret = map_bo_to_gpuvm(adev, entry, ctx.sync);
+			ret = map_bo_to_gpuvm(adev, entry, ctx.sync,
+					      is_invalid_userptr);
 			if (ret) {
 				pr_err("Failed to map radeon bo to gpuvm\n");
 				goto map_bo_to_gpuvm_failed;
@@ -1384,7 +1587,7 @@ int amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel(struct kgd_dev *kgd,
 		goto bo_reserve_failed;
 	}
 
-	ret = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT, NULL);
+	ret = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT);
 	if (ret) {
 		pr_err("Failed to pin bo. ret %d\n", ret);
 		goto pin_failed;
@@ -1418,6 +1621,337 @@ bo_reserve_failed:
 	return ret;
 }
 
+/* Evict a userptr BO by stopping the queues if necessary
+ *
+ * Runs in MMU notifier, may be in RECLAIM_FS context. This means it
+ * cannot do any memory allocations, and cannot take any locks that
+ * are held elsewhere while allocating memory. Therefore this is as
+ * simple as possible, using atomic counters.
+ *
+ * It doesn't do anything to the BO itself. The real work happens in
+ * restore, where we get updated page addresses. This function only
+ * ensures that GPU access to the BO is stopped.
+ */
+int amdgpu_amdkfd_evict_userptr(struct kgd_mem *mem,
+				struct mm_struct *mm)
+{
+	struct amdkfd_process_info *process_info = mem->process_info;
+	int invalid, evicted_bos;
+	int r = 0;
+
+	invalid = atomic_inc_return(&mem->invalid);
+	evicted_bos = atomic_inc_return(&process_info->evicted_bos);
+	if (evicted_bos == 1) {
+		/* First eviction, stop the queues */
+		r = kgd2kfd->quiesce_mm(mm);
+		if (r)
+			pr_err("Failed to quiesce KFD\n");
+		schedule_delayed_work(&process_info->restore_userptr_work,
+			msecs_to_jiffies(AMDGPU_USERPTR_RESTORE_DELAY_MS));
+	}
+
+	return r;
+}
+
+/* Update invalid userptr BOs
+ *
+ * Moves invalidated (evicted) userptr BOs from userptr_valid_list to
+ * userptr_inval_list and updates user pages for all BOs that have
+ * been invalidated since their last update.
+ */
+static int update_invalid_user_pages(struct amdkfd_process_info *process_info,
+				     struct mm_struct *mm)
+{
+	struct kgd_mem *mem, *tmp_mem;
+	struct amdgpu_bo *bo;
+	struct ttm_operation_ctx ctx = { false, false };
+	int invalid, ret;
+
+	/* Move all invalidated BOs to the userptr_inval_list and
+	 * release their user pages by migration to the CPU domain
+	 */
+	list_for_each_entry_safe(mem, tmp_mem,
+				 &process_info->userptr_valid_list,
+				 validate_list.head) {
+		if (!atomic_read(&mem->invalid))
+			continue; /* BO is still valid */
+
+		bo = mem->bo;
+
+		if (amdgpu_bo_reserve(bo, true))
+			return -EAGAIN;
+		amdgpu_ttm_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_CPU);
+		ret = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
+		amdgpu_bo_unreserve(bo);
+		if (ret) {
+			pr_err("%s: Failed to invalidate userptr BO\n",
+			       __func__);
+			return -EAGAIN;
+		}
+
+		list_move_tail(&mem->validate_list.head,
+			       &process_info->userptr_inval_list);
+	}
+
+	if (list_empty(&process_info->userptr_inval_list))
+		return 0; /* All evicted userptr BOs were freed */
+
+	/* Go through userptr_inval_list and update any invalid user_pages */
+	list_for_each_entry(mem, &process_info->userptr_inval_list,
+			    validate_list.head) {
+		invalid = atomic_read(&mem->invalid);
+		if (!invalid)
+			/* BO hasn't been invalidated since the last
+			 * revalidation attempt. Keep its BO list.
+			 */
+			continue;
+
+		bo = mem->bo;
+
+		if (!mem->user_pages) {
+			mem->user_pages =
+				kvmalloc_array(bo->tbo.ttm->num_pages,
+						 sizeof(struct page *),
+						 GFP_KERNEL | __GFP_ZERO);
+			if (!mem->user_pages) {
+				pr_err("%s: Failed to allocate pages array\n",
+				       __func__);
+				return -ENOMEM;
+			}
+		} else if (mem->user_pages[0]) {
+			release_pages(mem->user_pages, bo->tbo.ttm->num_pages);
+		}
+
+		/* Get updated user pages */
+		ret = amdgpu_ttm_tt_get_user_pages(bo->tbo.ttm,
+						   mem->user_pages);
+		if (ret) {
+			mem->user_pages[0] = NULL;
+			pr_info("%s: Failed to get user pages: %d\n",
+				__func__, ret);
+			/* Pretend it succeeded. It will fail later
+			 * with a VM fault if the GPU tries to access
+			 * it. Better than hanging indefinitely with
+			 * stalled user mode queues.
+			 */
+		}
+
+		/* Mark the BO as valid unless it was invalidated
+		 * again concurrently
+		 */
+		if (atomic_cmpxchg(&mem->invalid, invalid, 0) != invalid)
+			return -EAGAIN;
+	}
+
+	return 0;
+}
+
+/* Validate invalid userptr BOs
+ *
+ * Validates BOs on the userptr_inval_list, and moves them back to the
+ * userptr_valid_list. Also updates GPUVM page tables with new page
+ * addresses and waits for the page table updates to complete.
+ */
+static int validate_invalid_user_pages(struct amdkfd_process_info *process_info)
+{
+	struct amdgpu_bo_list_entry *pd_bo_list_entries;
+	struct list_head resv_list, duplicates;
+	struct ww_acquire_ctx ticket;
+	struct amdgpu_sync sync;
+
+	struct amdgpu_vm *peer_vm;
+	struct kgd_mem *mem, *tmp_mem;
+	struct amdgpu_bo *bo;
+	struct ttm_operation_ctx ctx = { false, false };
+	int i, ret;
+
+	pd_bo_list_entries = kcalloc(process_info->n_vms,
+				     sizeof(struct amdgpu_bo_list_entry),
+				     GFP_KERNEL);
+	if (!pd_bo_list_entries) {
+		pr_err("%s: Failed to allocate PD BO list entries\n", __func__);
+		return -ENOMEM;
+	}
+
+	INIT_LIST_HEAD(&resv_list);
+	INIT_LIST_HEAD(&duplicates);
+
+	/* Get all the page directory BOs that need to be reserved */
+	i = 0;
+	list_for_each_entry(peer_vm, &process_info->vm_list_head,
+			    vm_list_node)
+		amdgpu_vm_get_pd_bo(peer_vm, &resv_list,
+				    &pd_bo_list_entries[i++]);
+	/* Add the userptr_inval_list entries to resv_list */
+	list_for_each_entry(mem, &process_info->userptr_inval_list,
+			    validate_list.head) {
+		list_add_tail(&mem->resv_list.head, &resv_list);
+		mem->resv_list.bo = mem->validate_list.bo;
+		mem->resv_list.shared = mem->validate_list.shared;
+	}
+
+	/* Reserve all BOs and page tables for validation */
+	ret = ttm_eu_reserve_buffers(&ticket, &resv_list, false, &duplicates);
+	WARN(!list_empty(&duplicates), "Duplicates should be empty");
+	if (ret)
+		goto out;
+
+	amdgpu_sync_create(&sync);
+
+	/* Avoid triggering eviction fences when unmapping invalid
+	 * userptr BOs (waits for all fences, doesn't use
+	 * FENCE_OWNER_VM)
+	 */
+	list_for_each_entry(peer_vm, &process_info->vm_list_head,
+			    vm_list_node)
+		amdgpu_amdkfd_remove_eviction_fence(peer_vm->root.base.bo,
+						process_info->eviction_fence,
+						NULL, NULL);
+
+	ret = process_validate_vms(process_info);
+	if (ret)
+		goto unreserve_out;
+
+	/* Validate BOs and update GPUVM page tables */
+	list_for_each_entry_safe(mem, tmp_mem,
+				 &process_info->userptr_inval_list,
+				 validate_list.head) {
+		struct kfd_bo_va_list *bo_va_entry;
+
+		bo = mem->bo;
+
+		/* Copy pages array and validate the BO if we got user pages */
+		if (mem->user_pages[0]) {
+			amdgpu_ttm_tt_set_user_pages(bo->tbo.ttm,
+						     mem->user_pages);
+			amdgpu_ttm_placement_from_domain(bo, mem->domain);
+			ret = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
+			if (ret) {
+				pr_err("%s: failed to validate BO\n", __func__);
+				goto unreserve_out;
+			}
+		}
+
+		/* Validate succeeded, now the BO owns the pages, free
+		 * our copy of the pointer array. Put this BO back on
+		 * the userptr_valid_list. If we need to revalidate
+		 * it, we need to start from scratch.
+		 */
+		kvfree(mem->user_pages);
+		mem->user_pages = NULL;
+		list_move_tail(&mem->validate_list.head,
+			       &process_info->userptr_valid_list);
+
+		/* Update mapping. If the BO was not validated
+		 * (because we couldn't get user pages), this will
+		 * clear the page table entries, which will result in
+		 * VM faults if the GPU tries to access the invalid
+		 * memory.
+		 */
+		list_for_each_entry(bo_va_entry, &mem->bo_va_list, bo_list) {
+			if (!bo_va_entry->is_mapped)
+				continue;
+
+			ret = update_gpuvm_pte((struct amdgpu_device *)
+					       bo_va_entry->kgd_dev,
+					       bo_va_entry, &sync);
+			if (ret) {
+				pr_err("%s: update PTE failed\n", __func__);
+				/* make sure this gets validated again */
+				atomic_inc(&mem->invalid);
+				goto unreserve_out;
+			}
+		}
+	}
+
+	/* Update page directories */
+	ret = process_update_pds(process_info, &sync);
+
+unreserve_out:
+	list_for_each_entry(peer_vm, &process_info->vm_list_head,
+			    vm_list_node)
+		amdgpu_bo_fence(peer_vm->root.base.bo,
+				&process_info->eviction_fence->base, true);
+	ttm_eu_backoff_reservation(&ticket, &resv_list);
+	amdgpu_sync_wait(&sync, false);
+	amdgpu_sync_free(&sync);
+out:
+	kfree(pd_bo_list_entries);
+
+	return ret;
+}
+
+/* Worker callback to restore evicted userptr BOs
+ *
+ * Tries to update and validate all userptr BOs. If successful and no
+ * concurrent evictions happened, the queues are restarted. Otherwise,
+ * reschedule for another attempt later.
+ */
+static void amdgpu_amdkfd_restore_userptr_worker(struct work_struct *work)
+{
+	struct delayed_work *dwork = to_delayed_work(work);
+	struct amdkfd_process_info *process_info =
+		container_of(dwork, struct amdkfd_process_info,
+			     restore_userptr_work);
+	struct task_struct *usertask;
+	struct mm_struct *mm;
+	int evicted_bos;
+
+	evicted_bos = atomic_read(&process_info->evicted_bos);
+	if (!evicted_bos)
+		return;
+
+	/* Reference task and mm in case of concurrent process termination */
+	usertask = get_pid_task(process_info->pid, PIDTYPE_PID);
+	if (!usertask)
+		return;
+	mm = get_task_mm(usertask);
+	if (!mm) {
+		put_task_struct(usertask);
+		return;
+	}
+
+	mutex_lock(&process_info->lock);
+
+	if (update_invalid_user_pages(process_info, mm))
+		goto unlock_out;
+	/* userptr_inval_list can be empty if all evicted userptr BOs
+	 * have been freed. In that case there is nothing to validate
+	 * and we can just restart the queues.
+	 */
+	if (!list_empty(&process_info->userptr_inval_list)) {
+		if (atomic_read(&process_info->evicted_bos) != evicted_bos)
+			goto unlock_out; /* Concurrent eviction, try again */
+
+		if (validate_invalid_user_pages(process_info))
+			goto unlock_out;
+	}
+	/* Final check for concurrent evicton and atomic update. If
+	 * another eviction happens after successful update, it will
+	 * be a first eviction that calls quiesce_mm. The eviction
+	 * reference counting inside KFD will handle this case.
+	 */
+	if (atomic_cmpxchg(&process_info->evicted_bos, evicted_bos, 0) !=
+	    evicted_bos)
+		goto unlock_out;
+	evicted_bos = 0;
+	if (kgd2kfd->resume_mm(mm)) {
+		pr_err("%s: Failed to resume KFD\n", __func__);
+		/* No recovery from this failure. Probably the CP is
+		 * hanging. No point trying again.
+		 */
+	}
+unlock_out:
+	mutex_unlock(&process_info->lock);
+	mmput(mm);
+	put_task_struct(usertask);
+
+	/* If validation failed, reschedule another attempt */
+	if (evicted_bos)
+		schedule_delayed_work(&process_info->restore_userptr_work,
+			msecs_to_jiffies(AMDGPU_USERPTR_RESTORE_DELAY_MS));
+}
+
 /** amdgpu_amdkfd_gpuvm_restore_process_bos - Restore all BOs for the given
  *   KFD process identified by process_info
  *
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
index a0f48cb9b8f0..236915849cfe 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.c
@@ -322,3 +322,47 @@ int amdgpu_atomfirmware_get_clock_info(struct amdgpu_device *adev)
 
 	return ret;
 }
+
+union gfx_info {
+	struct  atom_gfx_info_v2_4 v24;
+};
+
+int amdgpu_atomfirmware_get_gfx_info(struct amdgpu_device *adev)
+{
+	struct amdgpu_mode_info *mode_info = &adev->mode_info;
+	int index;
+	uint8_t frev, crev;
+	uint16_t data_offset;
+
+	index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1,
+					    gfx_info);
+	if (amdgpu_atom_parse_data_header(mode_info->atom_context, index, NULL,
+				   &frev, &crev, &data_offset)) {
+		union gfx_info *gfx_info = (union gfx_info *)
+			(mode_info->atom_context->bios + data_offset);
+		switch (crev) {
+		case 4:
+			adev->gfx.config.max_shader_engines = gfx_info->v24.gc_num_se;
+			adev->gfx.config.max_cu_per_sh = gfx_info->v24.gc_num_cu_per_sh;
+			adev->gfx.config.max_sh_per_se = gfx_info->v24.gc_num_sh_per_se;
+			adev->gfx.config.max_backends_per_se = gfx_info->v24.gc_num_rb_per_se;
+			adev->gfx.config.max_texture_channel_caches = gfx_info->v24.gc_num_tccs;
+			adev->gfx.config.max_gprs = le16_to_cpu(gfx_info->v24.gc_num_gprs);
+			adev->gfx.config.max_gs_threads = gfx_info->v24.gc_num_max_gs_thds;
+			adev->gfx.config.gs_vgt_table_depth = gfx_info->v24.gc_gs_table_depth;
+			adev->gfx.config.gs_prim_buffer_depth =
+				le16_to_cpu(gfx_info->v24.gc_gsprim_buff_depth);
+			adev->gfx.config.double_offchip_lds_buf =
+				gfx_info->v24.gc_double_offchip_lds_buffer;
+			adev->gfx.cu_info.wave_front_size = le16_to_cpu(gfx_info->v24.gc_wave_size);
+			adev->gfx.cu_info.max_waves_per_simd = le16_to_cpu(gfx_info->v24.gc_max_waves_per_simd);
+			adev->gfx.cu_info.max_scratch_slots_per_cu = gfx_info->v24.gc_max_scratch_slots_per_cu;
+			adev->gfx.cu_info.lds_size = le16_to_cpu(gfx_info->v24.gc_lds_size);
+			return 0;
+		default:
+			return -EINVAL;
+		}
+
+	}
+	return -EINVAL;
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h
index 7689c961c4ef..20f158fd3b76 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atomfirmware.h
@@ -30,5 +30,6 @@ int amdgpu_atomfirmware_allocate_fb_scratch(struct amdgpu_device *adev);
 int amdgpu_atomfirmware_get_vram_width(struct amdgpu_device *adev);
 int amdgpu_atomfirmware_get_vram_type(struct amdgpu_device *adev);
 int amdgpu_atomfirmware_get_clock_info(struct amdgpu_device *adev);
+int amdgpu_atomfirmware_get_gfx_info(struct amdgpu_device *adev);
 
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
index 1ae5ae8c45a4..b33f1680c9a3 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_atpx_handler.c
@@ -32,7 +32,7 @@ struct amdgpu_atpx_functions {
 	bool switch_start;
 	bool switch_end;
 	bool disp_connectors_mapping;
-	bool disp_detetion_ports;
+	bool disp_detection_ports;
 };
 
 struct amdgpu_atpx {
@@ -90,6 +90,12 @@ bool amdgpu_atpx_dgpu_req_power_for_displays(void) {
 	return amdgpu_atpx_priv.atpx.dgpu_req_power_for_displays;
 }
 
+#if defined(CONFIG_ACPI)
+void *amdgpu_atpx_get_dhandle(void) {
+	return amdgpu_atpx_priv.dhandle;
+}
+#endif
+
 /**
  * amdgpu_atpx_call - call an ATPX method
  *
@@ -156,7 +162,7 @@ static void amdgpu_atpx_parse_functions(struct amdgpu_atpx_functions *f, u32 mas
 	f->switch_start = mask & ATPX_GRAPHICS_DEVICE_SWITCH_START_NOTIFICATION_SUPPORTED;
 	f->switch_end = mask & ATPX_GRAPHICS_DEVICE_SWITCH_END_NOTIFICATION_SUPPORTED;
 	f->disp_connectors_mapping = mask & ATPX_GET_DISPLAY_CONNECTORS_MAPPING_SUPPORTED;
-	f->disp_detetion_ports = mask & ATPX_GET_DISPLAY_DETECTION_PORTS_SUPPORTED;
+	f->disp_detection_ports = mask & ATPX_GET_DISPLAY_DETECTION_PORTS_SUPPORTED;
 }
 
 /**
@@ -550,7 +556,7 @@ static int amdgpu_atpx_init(void)
  * look up whether we are the integrated or discrete GPU (all asics).
  * Returns the client id.
  */
-static int amdgpu_atpx_get_client_id(struct pci_dev *pdev)
+static enum vga_switcheroo_client_id amdgpu_atpx_get_client_id(struct pci_dev *pdev)
 {
 	if (amdgpu_atpx_priv.dhandle == ACPI_HANDLE(&pdev->dev))
 		return VGA_SWITCHEROO_IGD;
@@ -569,7 +575,6 @@ static const struct amdgpu_px_quirk amdgpu_px_quirk_list[] = {
 	{ 0x1002, 0x6900, 0x1002, 0x0124, AMDGPU_PX_QUIRK_FORCE_ATPX },
 	{ 0x1002, 0x6900, 0x1028, 0x0812, AMDGPU_PX_QUIRK_FORCE_ATPX },
 	{ 0x1002, 0x6900, 0x1028, 0x0813, AMDGPU_PX_QUIRK_FORCE_ATPX },
-	{ 0x1002, 0x67DF, 0x1028, 0x0774, AMDGPU_PX_QUIRK_FORCE_ATPX },
 	{ 0, 0, 0, 0, 0 },
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c
index 02b849be083b..3079ea8523c5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_benchmark.c
@@ -75,37 +75,56 @@ static void amdgpu_benchmark_move(struct amdgpu_device *adev, unsigned size,
 {
 	struct amdgpu_bo *dobj = NULL;
 	struct amdgpu_bo *sobj = NULL;
+	struct amdgpu_bo_param bp;
 	uint64_t saddr, daddr;
 	int r, n;
 	int time;
 
+	memset(&bp, 0, sizeof(bp));
+	bp.size = size;
+	bp.byte_align = PAGE_SIZE;
+	bp.domain = sdomain;
+	bp.flags = 0;
+	bp.type = ttm_bo_type_kernel;
+	bp.resv = NULL;
 	n = AMDGPU_BENCHMARK_ITERATIONS;
-	r = amdgpu_bo_create(adev, size, PAGE_SIZE,sdomain, 0,
-			     ttm_bo_type_kernel, NULL, &sobj);
+	r = amdgpu_bo_create(adev, &bp, &sobj);
 	if (r) {
 		goto out_cleanup;
 	}
 	r = amdgpu_bo_reserve(sobj, false);
 	if (unlikely(r != 0))
 		goto out_cleanup;
-	r = amdgpu_bo_pin(sobj, sdomain, &saddr);
+	r = amdgpu_bo_pin(sobj, sdomain);
+	if (r) {
+		amdgpu_bo_unreserve(sobj);
+		goto out_cleanup;
+	}
+	r = amdgpu_ttm_alloc_gart(&sobj->tbo);
 	amdgpu_bo_unreserve(sobj);
 	if (r) {
 		goto out_cleanup;
 	}
-	r = amdgpu_bo_create(adev, size, PAGE_SIZE, ddomain, 0,
-			     ttm_bo_type_kernel, NULL, &dobj);
+	saddr = amdgpu_bo_gpu_offset(sobj);
+	bp.domain = ddomain;
+	r = amdgpu_bo_create(adev, &bp, &dobj);
 	if (r) {
 		goto out_cleanup;
 	}
 	r = amdgpu_bo_reserve(dobj, false);
 	if (unlikely(r != 0))
 		goto out_cleanup;
-	r = amdgpu_bo_pin(dobj, ddomain, &daddr);
+	r = amdgpu_bo_pin(dobj, ddomain);
+	if (r) {
+		amdgpu_bo_unreserve(sobj);
+		goto out_cleanup;
+	}
+	r = amdgpu_ttm_alloc_gart(&dobj->tbo);
 	amdgpu_bo_unreserve(dobj);
 	if (r) {
 		goto out_cleanup;
 	}
+	daddr = amdgpu_bo_gpu_offset(dobj);
 
 	if (adev->mman.buffer_funcs) {
 		time = amdgpu_benchmark_do_move(adev, size, saddr, daddr, n);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
index 92be7f6de197..7679c068c89a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_bo_list.c
@@ -55,15 +55,15 @@ static void amdgpu_bo_list_release_rcu(struct kref *ref)
 	kfree_rcu(list, rhead);
 }
 
-static int amdgpu_bo_list_create(struct amdgpu_device *adev,
+int amdgpu_bo_list_create(struct amdgpu_device *adev,
 				 struct drm_file *filp,
 				 struct drm_amdgpu_bo_list_entry *info,
 				 unsigned num_entries,
-				 int *id)
+				 struct amdgpu_bo_list **list_out)
 {
-	int r;
-	struct amdgpu_fpriv *fpriv = filp->driver_priv;
 	struct amdgpu_bo_list *list;
+	int r;
+
 
 	list = kzalloc(sizeof(struct amdgpu_bo_list), GFP_KERNEL);
 	if (!list)
@@ -78,16 +78,7 @@ static int amdgpu_bo_list_create(struct amdgpu_device *adev,
 		return r;
 	}
 
-	/* idr alloc should be called only after initialization of bo list. */
-	mutex_lock(&fpriv->bo_list_lock);
-	r = idr_alloc(&fpriv->bo_list_handles, list, 1, 0, GFP_KERNEL);
-	mutex_unlock(&fpriv->bo_list_lock);
-	if (r < 0) {
-		amdgpu_bo_list_free(list);
-		return r;
-	}
-	*id = r;
-
+	*list_out = list;
 	return 0;
 }
 
@@ -263,55 +254,79 @@ void amdgpu_bo_list_free(struct amdgpu_bo_list *list)
 	kfree(list);
 }
 
-int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data,
-				struct drm_file *filp)
+int amdgpu_bo_create_list_entry_array(struct drm_amdgpu_bo_list_in *in,
+				      struct drm_amdgpu_bo_list_entry **info_param)
 {
+	const void __user *uptr = u64_to_user_ptr(in->bo_info_ptr);
 	const uint32_t info_size = sizeof(struct drm_amdgpu_bo_list_entry);
-
-	struct amdgpu_device *adev = dev->dev_private;
-	struct amdgpu_fpriv *fpriv = filp->driver_priv;
-	union drm_amdgpu_bo_list *args = data;
-	uint32_t handle = args->in.list_handle;
-	const void __user *uptr = u64_to_user_ptr(args->in.bo_info_ptr);
-
 	struct drm_amdgpu_bo_list_entry *info;
-	struct amdgpu_bo_list *list;
-
 	int r;
 
-	info = kvmalloc_array(args->in.bo_number,
-			     sizeof(struct drm_amdgpu_bo_list_entry), GFP_KERNEL);
+	info = kvmalloc_array(in->bo_number, info_size, GFP_KERNEL);
 	if (!info)
 		return -ENOMEM;
 
 	/* copy the handle array from userspace to a kernel buffer */
 	r = -EFAULT;
-	if (likely(info_size == args->in.bo_info_size)) {
-		unsigned long bytes = args->in.bo_number *
-			args->in.bo_info_size;
+	if (likely(info_size == in->bo_info_size)) {
+		unsigned long bytes = in->bo_number *
+			in->bo_info_size;
 
 		if (copy_from_user(info, uptr, bytes))
 			goto error_free;
 
 	} else {
-		unsigned long bytes = min(args->in.bo_info_size, info_size);
+		unsigned long bytes = min(in->bo_info_size, info_size);
 		unsigned i;
 
-		memset(info, 0, args->in.bo_number * info_size);
-		for (i = 0; i < args->in.bo_number; ++i) {
+		memset(info, 0, in->bo_number * info_size);
+		for (i = 0; i < in->bo_number; ++i) {
 			if (copy_from_user(&info[i], uptr, bytes))
 				goto error_free;
 
-			uptr += args->in.bo_info_size;
+			uptr += in->bo_info_size;
 		}
 	}
 
+	*info_param = info;
+	return 0;
+
+error_free:
+	kvfree(info);
+	return r;
+}
+
+int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data,
+				struct drm_file *filp)
+{
+	struct amdgpu_device *adev = dev->dev_private;
+	struct amdgpu_fpriv *fpriv = filp->driver_priv;
+	union drm_amdgpu_bo_list *args = data;
+	uint32_t handle = args->in.list_handle;
+	struct drm_amdgpu_bo_list_entry *info = NULL;
+	struct amdgpu_bo_list *list;
+	int r;
+
+	r = amdgpu_bo_create_list_entry_array(&args->in, &info);
+	if (r)
+		goto error_free;
+
 	switch (args->in.operation) {
 	case AMDGPU_BO_LIST_OP_CREATE:
 		r = amdgpu_bo_list_create(adev, filp, info, args->in.bo_number,
-					  &handle);
+					  &list);
 		if (r)
 			goto error_free;
+
+		mutex_lock(&fpriv->bo_list_lock);
+		r = idr_alloc(&fpriv->bo_list_handles, list, 1, 0, GFP_KERNEL);
+		mutex_unlock(&fpriv->bo_list_lock);
+		if (r < 0) {
+			amdgpu_bo_list_free(list);
+			return r;
+		}
+
+		handle = r;
 		break;
 
 	case AMDGPU_BO_LIST_OP_DESTROY:
@@ -345,6 +360,7 @@ int amdgpu_bo_list_ioctl(struct drm_device *dev, void *data,
 	return 0;
 
 error_free:
-	kvfree(info);
+	if (info)
+		kvfree(info);
 	return r;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
index 71a57b2f7f04..693ec5ea4950 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c
@@ -23,7 +23,6 @@
  */
 #include <linux/list.h>
 #include <linux/slab.h>
-#include <linux/pci.h>
 #include <drm/drmP.h>
 #include <linux/firmware.h>
 #include <drm/amdgpu_drm.h>
@@ -109,121 +108,6 @@ static void amdgpu_cgs_write_ind_register(struct cgs_device *cgs_device,
 	WARN(1, "Invalid indirect register space");
 }
 
-static int amdgpu_cgs_get_pci_resource(struct cgs_device *cgs_device,
-				       enum cgs_resource_type resource_type,
-				       uint64_t size,
-				       uint64_t offset,
-				       uint64_t *resource_base)
-{
-	CGS_FUNC_ADEV;
-
-	if (resource_base == NULL)
-		return -EINVAL;
-
-	switch (resource_type) {
-	case CGS_RESOURCE_TYPE_MMIO:
-		if (adev->rmmio_size == 0)
-			return -ENOENT;
-		if ((offset + size) > adev->rmmio_size)
-			return -EINVAL;
-		*resource_base = adev->rmmio_base;
-		return 0;
-	case CGS_RESOURCE_TYPE_DOORBELL:
-		if (adev->doorbell.size == 0)
-			return -ENOENT;
-		if ((offset + size) > adev->doorbell.size)
-			return -EINVAL;
-		*resource_base = adev->doorbell.base;
-		return 0;
-	case CGS_RESOURCE_TYPE_FB:
-	case CGS_RESOURCE_TYPE_IO:
-	case CGS_RESOURCE_TYPE_ROM:
-	default:
-		return -EINVAL;
-	}
-}
-
-static const void *amdgpu_cgs_atom_get_data_table(struct cgs_device *cgs_device,
-						  unsigned table, uint16_t *size,
-						  uint8_t *frev, uint8_t *crev)
-{
-	CGS_FUNC_ADEV;
-	uint16_t data_start;
-
-	if (amdgpu_atom_parse_data_header(
-		    adev->mode_info.atom_context, table, size,
-		    frev, crev, &data_start))
-		return (uint8_t*)adev->mode_info.atom_context->bios +
-			data_start;
-
-	return NULL;
-}
-
-static int amdgpu_cgs_atom_get_cmd_table_revs(struct cgs_device *cgs_device, unsigned table,
-					      uint8_t *frev, uint8_t *crev)
-{
-	CGS_FUNC_ADEV;
-
-	if (amdgpu_atom_parse_cmd_header(
-		    adev->mode_info.atom_context, table,
-		    frev, crev))
-		return 0;
-
-	return -EINVAL;
-}
-
-static int amdgpu_cgs_atom_exec_cmd_table(struct cgs_device *cgs_device, unsigned table,
-					  void *args)
-{
-	CGS_FUNC_ADEV;
-
-	return amdgpu_atom_execute_table(
-		adev->mode_info.atom_context, table, args);
-}
-
-static int amdgpu_cgs_set_clockgating_state(struct cgs_device *cgs_device,
-				  enum amd_ip_block_type block_type,
-				  enum amd_clockgating_state state)
-{
-	CGS_FUNC_ADEV;
-	int i, r = -1;
-
-	for (i = 0; i < adev->num_ip_blocks; i++) {
-		if (!adev->ip_blocks[i].status.valid)
-			continue;
-
-		if (adev->ip_blocks[i].version->type == block_type) {
-			r = adev->ip_blocks[i].version->funcs->set_clockgating_state(
-								(void *)adev,
-									state);
-			break;
-		}
-	}
-	return r;
-}
-
-static int amdgpu_cgs_set_powergating_state(struct cgs_device *cgs_device,
-				  enum amd_ip_block_type block_type,
-				  enum amd_powergating_state state)
-{
-	CGS_FUNC_ADEV;
-	int i, r = -1;
-
-	for (i = 0; i < adev->num_ip_blocks; i++) {
-		if (!adev->ip_blocks[i].status.valid)
-			continue;
-
-		if (adev->ip_blocks[i].version->type == block_type) {
-			r = adev->ip_blocks[i].version->funcs->set_powergating_state(
-								(void *)adev,
-									state);
-			break;
-		}
-	}
-	return r;
-}
-
-
 static uint32_t fw_type_convert(struct cgs_device *cgs_device, uint32_t fw_type)
 {
 	CGS_FUNC_ADEV;
@@ -271,18 +155,6 @@ static uint32_t fw_type_convert(struct cgs_device *cgs_device, uint32_t fw_type)
 	return result;
 }
 
-static int amdgpu_cgs_rel_firmware(struct cgs_device *cgs_device, enum cgs_ucode_id type)
-{
-	CGS_FUNC_ADEV;
-	if ((CGS_UCODE_ID_SMU == type) || (CGS_UCODE_ID_SMU_SK == type)) {
-		release_firmware(adev->pm.fw);
-		adev->pm.fw = NULL;
-		return 0;
-	}
-	/* cannot release other firmware because they are not created by cgs */
-	return -EINVAL;
-}
-
 static uint16_t amdgpu_get_firmware_version(struct cgs_device *cgs_device,
 					enum cgs_ucode_id type)
 {
@@ -326,34 +198,6 @@ static uint16_t amdgpu_get_firmware_version(struct cgs_device *cgs_device,
 	return fw_version;
 }
 
-static int amdgpu_cgs_enter_safe_mode(struct cgs_device *cgs_device,
-					bool en)
-{
-	CGS_FUNC_ADEV;
-
-	if (adev->gfx.rlc.funcs->enter_safe_mode == NULL ||
-		adev->gfx.rlc.funcs->exit_safe_mode == NULL)
-		return 0;
-
-	if (en)
-		adev->gfx.rlc.funcs->enter_safe_mode(adev);
-	else
-		adev->gfx.rlc.funcs->exit_safe_mode(adev);
-
-	return 0;
-}
-
-static void amdgpu_cgs_lock_grbm_idx(struct cgs_device *cgs_device,
-					bool lock)
-{
-	CGS_FUNC_ADEV;
-
-	if (lock)
-		mutex_lock(&adev->grbm_idx_mutex);
-	else
-		mutex_unlock(&adev->grbm_idx_mutex);
-}
-
 static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 					enum cgs_ucode_id type,
 					struct cgs_firmware_info *info)
@@ -470,17 +314,17 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 					(adev->pdev->revision == 0x81) ||
 					(adev->pdev->device == 0x665f)) {
 					info->is_kicker = true;
-					strcpy(fw_name, "radeon/bonaire_k_smc.bin");
+					strcpy(fw_name, "amdgpu/bonaire_k_smc.bin");
 				} else {
-					strcpy(fw_name, "radeon/bonaire_smc.bin");
+					strcpy(fw_name, "amdgpu/bonaire_smc.bin");
 				}
 				break;
 			case CHIP_HAWAII:
 				if (adev->pdev->revision == 0x80) {
 					info->is_kicker = true;
-					strcpy(fw_name, "radeon/hawaii_k_smc.bin");
+					strcpy(fw_name, "amdgpu/hawaii_k_smc.bin");
 				} else {
-					strcpy(fw_name, "radeon/hawaii_smc.bin");
+					strcpy(fw_name, "amdgpu/hawaii_smc.bin");
 				}
 				break;
 			case CHIP_TOPAZ:
@@ -541,6 +385,9 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 			case CHIP_POLARIS12:
 				strcpy(fw_name, "amdgpu/polaris12_smc.bin");
 				break;
+			case CHIP_VEGAM:
+				strcpy(fw_name, "amdgpu/vegam_smc.bin");
+				break;
 			case CHIP_VEGA10:
 				if ((adev->pdev->device == 0x687f) &&
 					((adev->pdev->revision == 0xc0) ||
@@ -553,6 +400,9 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 			case CHIP_VEGA12:
 				strcpy(fw_name, "amdgpu/vega12_smc.bin");
 				break;
+			case CHIP_VEGA20:
+				strcpy(fw_name, "amdgpu/vega20_smc.bin");
+				break;
 			default:
 				DRM_ERROR("SMC firmware not supported\n");
 				return -EINVAL;
@@ -598,97 +448,12 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
 	return 0;
 }
 
-static int amdgpu_cgs_is_virtualization_enabled(void *cgs_device)
-{
-	CGS_FUNC_ADEV;
-	return amdgpu_sriov_vf(adev);
-}
-
-static int amdgpu_cgs_get_active_displays_info(struct cgs_device *cgs_device,
-					  struct cgs_display_info *info)
-{
-	CGS_FUNC_ADEV;
-	struct cgs_mode_info *mode_info;
-
-	if (info == NULL)
-		return -EINVAL;
-
-	mode_info = info->mode_info;
-	if (mode_info)
-		/* if the displays are off, vblank time is max */
-		mode_info->vblank_time_us = 0xffffffff;
-
-	if (!amdgpu_device_has_dc_support(adev)) {
-		struct amdgpu_crtc *amdgpu_crtc;
-		struct drm_device *ddev = adev->ddev;
-		struct drm_crtc *crtc;
-		uint32_t line_time_us, vblank_lines;
-
-		if (adev->mode_info.num_crtc && adev->mode_info.mode_config_initialized) {
-			list_for_each_entry(crtc,
-					&ddev->mode_config.crtc_list, head) {
-				amdgpu_crtc = to_amdgpu_crtc(crtc);
-				if (crtc->enabled) {
-					info->active_display_mask |= (1 << amdgpu_crtc->crtc_id);
-					info->display_count++;
-				}
-				if (mode_info != NULL &&
-					crtc->enabled && amdgpu_crtc->enabled &&
-					amdgpu_crtc->hw_mode.clock) {
-					line_time_us = (amdgpu_crtc->hw_mode.crtc_htotal * 1000) /
-								amdgpu_crtc->hw_mode.clock;
-					vblank_lines = amdgpu_crtc->hw_mode.crtc_vblank_end -
-								amdgpu_crtc->hw_mode.crtc_vdisplay +
-								(amdgpu_crtc->v_border * 2);
-					mode_info->vblank_time_us = vblank_lines * line_time_us;
-					mode_info->refresh_rate = drm_mode_vrefresh(&amdgpu_crtc->hw_mode);
-					/* we have issues with mclk switching with refresh rates
-					 * over 120 hz on the non-DC code.
-					 */
-					if (mode_info->refresh_rate > 120)
-						mode_info->vblank_time_us = 0;
-					mode_info = NULL;
-				}
-			}
-		}
-	} else {
-		info->display_count = adev->pm.pm_display_cfg.num_display;
-		if (mode_info != NULL) {
-			mode_info->vblank_time_us = adev->pm.pm_display_cfg.min_vblank_time;
-			mode_info->refresh_rate = adev->pm.pm_display_cfg.vrefresh;
-		}
-	}
-	return 0;
-}
-
-
-static int amdgpu_cgs_notify_dpm_enabled(struct cgs_device *cgs_device, bool enabled)
-{
-	CGS_FUNC_ADEV;
-
-	adev->pm.dpm_enabled = enabled;
-
-	return 0;
-}
-
 static const struct cgs_ops amdgpu_cgs_ops = {
 	.read_register = amdgpu_cgs_read_register,
 	.write_register = amdgpu_cgs_write_register,
 	.read_ind_register = amdgpu_cgs_read_ind_register,
 	.write_ind_register = amdgpu_cgs_write_ind_register,
-	.get_pci_resource = amdgpu_cgs_get_pci_resource,
-	.atom_get_data_table = amdgpu_cgs_atom_get_data_table,
-	.atom_get_cmd_table_revs = amdgpu_cgs_atom_get_cmd_table_revs,
-	.atom_exec_cmd_table = amdgpu_cgs_atom_exec_cmd_table,
 	.get_firmware_info = amdgpu_cgs_get_firmware_info,
-	.rel_firmware = amdgpu_cgs_rel_firmware,
-	.set_powergating_state = amdgpu_cgs_set_powergating_state,
-	.set_clockgating_state = amdgpu_cgs_set_clockgating_state,
-	.get_active_displays_info = amdgpu_cgs_get_active_displays_info,
-	.notify_dpm_enabled = amdgpu_cgs_notify_dpm_enabled,
-	.is_virtualization_enabled = amdgpu_cgs_is_virtualization_enabled,
-	.enter_safe_mode = amdgpu_cgs_enter_safe_mode,
-	.lock_grbm_idx = amdgpu_cgs_lock_grbm_idx,
 };
 
 struct cgs_device *amdgpu_cgs_create_device(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c
index 96501ff0e55b..c770d73352a7 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c
@@ -212,30 +212,21 @@ static void
 amdgpu_connector_update_scratch_regs(struct drm_connector *connector,
 				      enum drm_connector_status status)
 {
-	struct drm_encoder *best_encoder = NULL;
-	struct drm_encoder *encoder = NULL;
+	struct drm_encoder *best_encoder;
+	struct drm_encoder *encoder;
 	const struct drm_connector_helper_funcs *connector_funcs = connector->helper_private;
 	bool connected;
 	int i;
 
 	best_encoder = connector_funcs->best_encoder(connector);
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-
-		encoder = drm_encoder_find(connector->dev, NULL,
-					connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
-
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		if ((encoder == best_encoder) && (status == connector_status_connected))
 			connected = true;
 		else
 			connected = false;
 
 		amdgpu_atombios_encoder_set_bios_scratch_regs(connector, encoder, connected);
-
 	}
 }
 
@@ -246,17 +237,11 @@ amdgpu_connector_find_encoder(struct drm_connector *connector,
 	struct drm_encoder *encoder;
 	int i;
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-		encoder = drm_encoder_find(connector->dev, NULL,
-					connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
-
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		if (encoder->encoder_type == encoder_type)
 			return encoder;
 	}
+
 	return NULL;
 }
 
@@ -349,22 +334,24 @@ static int amdgpu_connector_ddc_get_modes(struct drm_connector *connector)
 	int ret;
 
 	if (amdgpu_connector->edid) {
-		drm_mode_connector_update_edid_property(connector, amdgpu_connector->edid);
+		drm_connector_update_edid_property(connector, amdgpu_connector->edid);
 		ret = drm_add_edid_modes(connector, amdgpu_connector->edid);
 		return ret;
 	}
-	drm_mode_connector_update_edid_property(connector, NULL);
+	drm_connector_update_edid_property(connector, NULL);
 	return 0;
 }
 
 static struct drm_encoder *
 amdgpu_connector_best_single_encoder(struct drm_connector *connector)
 {
-	int enc_id = connector->encoder_ids[0];
+	struct drm_encoder *encoder;
+	int i;
+
+	/* pick the first one */
+	drm_connector_for_each_possible_encoder(connector, encoder, i)
+		return encoder;
 
-	/* pick the encoder ids */
-	if (enc_id)
-		return drm_encoder_find(connector->dev, NULL, enc_id);
 	return NULL;
 }
 
@@ -691,7 +678,7 @@ static int amdgpu_connector_lvds_get_modes(struct drm_connector *connector)
 	return ret;
 }
 
-static int amdgpu_connector_lvds_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status amdgpu_connector_lvds_mode_valid(struct drm_connector *connector,
 					     struct drm_display_mode *mode)
 {
 	struct drm_encoder *encoder = amdgpu_connector_best_single_encoder(connector);
@@ -843,7 +830,7 @@ static int amdgpu_connector_vga_get_modes(struct drm_connector *connector)
 	return ret;
 }
 
-static int amdgpu_connector_vga_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status amdgpu_connector_vga_mode_valid(struct drm_connector *connector,
 					    struct drm_display_mode *mode)
 {
 	struct drm_device *dev = connector->dev;
@@ -985,9 +972,8 @@ amdgpu_connector_dvi_detect(struct drm_connector *connector, bool force)
 	struct drm_device *dev = connector->dev;
 	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
-	struct drm_encoder *encoder = NULL;
 	const struct drm_encoder_helper_funcs *encoder_funcs;
-	int i, r;
+	int r;
 	enum drm_connector_status ret = connector_status_disconnected;
 	bool dret = false, broken_edid = false;
 
@@ -1077,14 +1063,10 @@ amdgpu_connector_dvi_detect(struct drm_connector *connector, bool force)
 
 	/* find analog encoder */
 	if (amdgpu_connector->dac_load_detect) {
-		for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-			if (connector->encoder_ids[i] == 0)
-				break;
-
-			encoder = drm_encoder_find(connector->dev, NULL, connector->encoder_ids[i]);
-			if (!encoder)
-				continue;
+		struct drm_encoder *encoder;
+		int i;
 
+		drm_connector_for_each_possible_encoder(connector, encoder, i) {
 			if (encoder->encoder_type != DRM_MODE_ENCODER_DAC &&
 			    encoder->encoder_type != DRM_MODE_ENCODER_TVDAC)
 				continue;
@@ -1132,18 +1114,11 @@ exit:
 static struct drm_encoder *
 amdgpu_connector_dvi_encoder(struct drm_connector *connector)
 {
-	int enc_id = connector->encoder_ids[0];
 	struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
 	struct drm_encoder *encoder;
 	int i;
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-
-		encoder = drm_encoder_find(connector->dev, NULL, connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
 
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		if (amdgpu_connector->use_digital == true) {
 			if (encoder->encoder_type == DRM_MODE_ENCODER_TMDS)
 				return encoder;
@@ -1158,8 +1133,9 @@ amdgpu_connector_dvi_encoder(struct drm_connector *connector)
 
 	/* then check use digitial */
 	/* pick the first one */
-	if (enc_id)
-		return drm_encoder_find(connector->dev, NULL, enc_id);
+	drm_connector_for_each_possible_encoder(connector, encoder, i)
+		return encoder;
+
 	return NULL;
 }
 
@@ -1172,7 +1148,7 @@ static void amdgpu_connector_dvi_force(struct drm_connector *connector)
 		amdgpu_connector->use_digital = true;
 }
 
-static int amdgpu_connector_dvi_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status amdgpu_connector_dvi_mode_valid(struct drm_connector *connector,
 					    struct drm_display_mode *mode)
 {
 	struct drm_device *dev = connector->dev;
@@ -1296,15 +1272,7 @@ u16 amdgpu_connector_encoder_get_dp_bridge_encoder_id(struct drm_connector *conn
 	struct amdgpu_encoder *amdgpu_encoder;
 	int i;
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-
-		encoder = drm_encoder_find(connector->dev, NULL,
-					connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
-
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		amdgpu_encoder = to_amdgpu_encoder(encoder);
 
 		switch (amdgpu_encoder->encoder_id) {
@@ -1326,14 +1294,7 @@ static bool amdgpu_connector_encoder_is_hbr2(struct drm_connector *connector)
 	int i;
 	bool found = false;
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-		encoder = drm_encoder_find(connector->dev, NULL,
-					connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
-
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		amdgpu_encoder = to_amdgpu_encoder(encoder);
 		if (amdgpu_encoder->caps & ATOM_ENCODER_CAP_RECORD_HBR2)
 			found = true;
@@ -1448,7 +1409,7 @@ out:
 	return ret;
 }
 
-static int amdgpu_connector_dp_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status amdgpu_connector_dp_mode_valid(struct drm_connector *connector,
 					   struct drm_display_mode *mode)
 {
 	struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index dc34b50e6b29..7c5cc33d0cda 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -31,6 +31,7 @@
 #include <drm/drm_syncobj.h>
 #include "amdgpu.h"
 #include "amdgpu_trace.h"
+#include "amdgpu_gmc.h"
 
 static int amdgpu_cs_user_fence_chunk(struct amdgpu_cs_parser *p,
 				      struct drm_amdgpu_cs_chunk_fence *data,
@@ -65,11 +66,35 @@ static int amdgpu_cs_user_fence_chunk(struct amdgpu_cs_parser *p,
 	return 0;
 }
 
-static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data)
+static int amdgpu_cs_bo_handles_chunk(struct amdgpu_cs_parser *p,
+				      struct drm_amdgpu_bo_list_in *data)
+{
+	int r;
+	struct drm_amdgpu_bo_list_entry *info = NULL;
+
+	r = amdgpu_bo_create_list_entry_array(data, &info);
+	if (r)
+		return r;
+
+	r = amdgpu_bo_list_create(p->adev, p->filp, info, data->bo_number,
+				  &p->bo_list);
+	if (r)
+		goto error_free;
+
+	kvfree(info);
+	return 0;
+
+error_free:
+	if (info)
+		kvfree(info);
+
+	return r;
+}
+
+static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, union drm_amdgpu_cs *cs)
 {
 	struct amdgpu_fpriv *fpriv = p->filp->driver_priv;
 	struct amdgpu_vm *vm = &fpriv->vm;
-	union drm_amdgpu_cs *cs = data;
 	uint64_t *chunk_array_user;
 	uint64_t *chunk_array;
 	unsigned size, num_ibs = 0;
@@ -163,6 +188,19 @@ static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data)
 
 			break;
 
+		case AMDGPU_CHUNK_ID_BO_HANDLES:
+			size = sizeof(struct drm_amdgpu_bo_list_in);
+			if (p->chunks[i].length_dw * sizeof(uint32_t) < size) {
+				ret = -EINVAL;
+				goto free_partial_kdata;
+			}
+
+			ret = amdgpu_cs_bo_handles_chunk(p, p->chunks[i].kdata);
+			if (ret)
+				goto free_partial_kdata;
+
+			break;
+
 		case AMDGPU_CHUNK_ID_DEPENDENCIES:
 		case AMDGPU_CHUNK_ID_SYNCOBJ_IN:
 		case AMDGPU_CHUNK_ID_SYNCOBJ_OUT:
@@ -186,6 +224,10 @@ static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data)
 	if (p->uf_entry.robj)
 		p->job->uf_addr = uf_offset;
 	kfree(chunk_array);
+
+	/* Use this opportunity to fill in task info for the vm */
+	amdgpu_vm_set_task_info(vm);
+
 	return 0;
 
 free_all_kdata:
@@ -257,7 +299,7 @@ static void amdgpu_cs_get_threshold_for_moves(struct amdgpu_device *adev,
 		return;
 	}
 
-	total_vram = adev->gmc.real_vram_size - adev->vram_pin_size;
+	total_vram = adev->gmc.real_vram_size - atomic64_read(&adev->vram_pin_size);
 	used_vram = amdgpu_vram_mgr_usage(&adev->mman.bdev.man[TTM_PL_VRAM]);
 	free_vram = used_vram >= total_vram ? 0 : total_vram - used_vram;
 
@@ -302,7 +344,7 @@ static void amdgpu_cs_get_threshold_for_moves(struct amdgpu_device *adev,
 	*max_bytes = us_to_bytes(adev, adev->mm_stats.accum_us);
 
 	/* Do the same for visible VRAM if half of it is free */
-	if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size) {
+	if (!amdgpu_gmc_vram_full_visible(&adev->gmc)) {
 		u64 total_vis_vram = adev->gmc.visible_vram_size;
 		u64 used_vis_vram =
 			amdgpu_vram_mgr_vis_usage(&adev->mman.bdev.man[TTM_PL_VRAM]);
@@ -359,7 +401,7 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p,
 	 * to move it. Don't move anything if the threshold is zero.
 	 */
 	if (p->bytes_moved < p->bytes_moved_threshold) {
-		if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size &&
+		if (!amdgpu_gmc_vram_full_visible(&adev->gmc) &&
 		    (bo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED)) {
 			/* And don't move a CPU_ACCESS_REQUIRED BO to limited
 			 * visible VRAM if we've depleted our allowance to do
@@ -381,9 +423,8 @@ retry:
 	r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
 
 	p->bytes_moved += ctx.bytes_moved;
-	if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size &&
-	    bo->tbo.mem.mem_type == TTM_PL_VRAM &&
-	    bo->tbo.mem.start < adev->gmc.visible_vram_size >> PAGE_SHIFT)
+	if (!amdgpu_gmc_vram_full_visible(&adev->gmc) &&
+	    amdgpu_bo_in_cpu_visible_vram(bo))
 		p->bytes_moved_vis += ctx.bytes_moved;
 
 	if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) {
@@ -411,7 +452,6 @@ static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p,
 		struct amdgpu_bo_list_entry *candidate = p->evictable;
 		struct amdgpu_bo *bo = candidate->robj;
 		struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
-		u64 initial_bytes_moved, bytes_moved;
 		bool update_bytes_moved_vis;
 		uint32_t other;
 
@@ -435,18 +475,14 @@ static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p,
 			continue;
 
 		/* Good we can try to move this BO somewhere else */
-		amdgpu_ttm_placement_from_domain(bo, other);
 		update_bytes_moved_vis =
-			adev->gmc.visible_vram_size < adev->gmc.real_vram_size &&
-			bo->tbo.mem.mem_type == TTM_PL_VRAM &&
-			bo->tbo.mem.start < adev->gmc.visible_vram_size >> PAGE_SHIFT;
-		initial_bytes_moved = atomic64_read(&adev->num_bytes_moved);
+				!amdgpu_gmc_vram_full_visible(&adev->gmc) &&
+				amdgpu_bo_in_cpu_visible_vram(bo);
+		amdgpu_ttm_placement_from_domain(bo, other);
 		r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
-		bytes_moved = atomic64_read(&adev->num_bytes_moved) -
-			initial_bytes_moved;
-		p->bytes_moved += bytes_moved;
+		p->bytes_moved += ctx.bytes_moved;
 		if (update_bytes_moved_vis)
-			p->bytes_moved_vis += bytes_moved;
+			p->bytes_moved_vis += ctx.bytes_moved;
 
 		if (unlikely(r))
 			break;
@@ -528,15 +564,23 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 	struct amdgpu_bo_list_entry *e;
 	struct list_head duplicates;
 	unsigned i, tries = 10;
+	struct amdgpu_bo *gds;
+	struct amdgpu_bo *gws;
+	struct amdgpu_bo *oa;
 	int r;
 
 	INIT_LIST_HEAD(&p->validated);
 
-	p->bo_list = amdgpu_bo_list_get(fpriv, cs->in.bo_list_handle);
+	/* p->bo_list could already be assigned if AMDGPU_CHUNK_ID_BO_HANDLES is present */
+	if (!p->bo_list)
+		p->bo_list = amdgpu_bo_list_get(fpriv, cs->in.bo_list_handle);
+	else
+		mutex_lock(&p->bo_list->lock);
+
 	if (p->bo_list) {
 		amdgpu_bo_list_get_list(p->bo_list, &p->validated);
 		if (p->bo_list->first_userptr != p->bo_list->num_entries)
-			p->mn = amdgpu_mn_get(p->adev);
+			p->mn = amdgpu_mn_get(p->adev, AMDGPU_MN_TYPE_GFX);
 	}
 
 	INIT_LIST_HEAD(&duplicates);
@@ -658,31 +702,36 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 
 	amdgpu_cs_report_moved_bytes(p->adev, p->bytes_moved,
 				     p->bytes_moved_vis);
+
 	if (p->bo_list) {
-		struct amdgpu_bo *gds = p->bo_list->gds_obj;
-		struct amdgpu_bo *gws = p->bo_list->gws_obj;
-		struct amdgpu_bo *oa = p->bo_list->oa_obj;
 		struct amdgpu_vm *vm = &fpriv->vm;
 		unsigned i;
 
+		gds = p->bo_list->gds_obj;
+		gws = p->bo_list->gws_obj;
+		oa = p->bo_list->oa_obj;
 		for (i = 0; i < p->bo_list->num_entries; i++) {
 			struct amdgpu_bo *bo = p->bo_list->array[i].robj;
 
 			p->bo_list->array[i].bo_va = amdgpu_vm_bo_find(vm, bo);
 		}
+	} else {
+		gds = p->adev->gds.gds_gfx_bo;
+		gws = p->adev->gds.gws_gfx_bo;
+		oa = p->adev->gds.oa_gfx_bo;
+	}
 
-		if (gds) {
-			p->job->gds_base = amdgpu_bo_gpu_offset(gds);
-			p->job->gds_size = amdgpu_bo_size(gds);
-		}
-		if (gws) {
-			p->job->gws_base = amdgpu_bo_gpu_offset(gws);
-			p->job->gws_size = amdgpu_bo_size(gws);
-		}
-		if (oa) {
-			p->job->oa_base = amdgpu_bo_gpu_offset(oa);
-			p->job->oa_size = amdgpu_bo_size(oa);
-		}
+	if (gds) {
+		p->job->gds_base = amdgpu_bo_gpu_offset(gds);
+		p->job->gds_size = amdgpu_bo_size(gds);
+	}
+	if (gws) {
+		p->job->gws_base = amdgpu_bo_gpu_offset(gws);
+		p->job->gws_size = amdgpu_bo_size(gws);
+	}
+	if (oa) {
+		p->job->oa_base = amdgpu_bo_gpu_offset(oa);
+		p->job->oa_size = amdgpu_bo_size(oa);
 	}
 
 	if (!r && p->uf_entry.robj) {
@@ -863,11 +912,11 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev,
 {
 	struct amdgpu_fpriv *fpriv = p->filp->driver_priv;
 	struct amdgpu_vm *vm = &fpriv->vm;
-	struct amdgpu_ring *ring = p->job->ring;
+	struct amdgpu_ring *ring = p->ring;
 	int r;
 
 	/* Only for UVD/VCE VM emulation */
-	if (p->job->ring->funcs->parse_cs) {
+	if (p->ring->funcs->parse_cs) {
 		unsigned i, j;
 
 		for (i = 0, j = 0; i < p->nchunks && j < p->job->num_ibs; i++) {
@@ -925,6 +974,10 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev,
 		r = amdgpu_bo_vm_update_pte(p);
 		if (r)
 			return r;
+
+		r = reservation_object_reserve_shared(vm->root.base.bo->tbo.resv);
+		if (r)
+			return r;
 	}
 
 	return amdgpu_cs_sync_rings(p);
@@ -977,10 +1030,10 @@ static int amdgpu_cs_ib_fill(struct amdgpu_device *adev,
 			}
 		}
 
-		if (parser->job->ring && parser->job->ring != ring)
+		if (parser->ring && parser->ring != ring)
 			return -EINVAL;
 
-		parser->job->ring = ring;
+		parser->ring = ring;
 
 		r =  amdgpu_ib_get(adev, vm,
 					ring->funcs->parse_cs ? chunk_ib->ib_bytes : 0,
@@ -999,11 +1052,11 @@ static int amdgpu_cs_ib_fill(struct amdgpu_device *adev,
 
 	/* UVD & VCE fw doesn't support user fences */
 	if (parser->job->uf_addr && (
-	    parser->job->ring->funcs->type == AMDGPU_RING_TYPE_UVD ||
-	    parser->job->ring->funcs->type == AMDGPU_RING_TYPE_VCE))
+	    parser->ring->funcs->type == AMDGPU_RING_TYPE_UVD ||
+	    parser->ring->funcs->type == AMDGPU_RING_TYPE_VCE))
 		return -EINVAL;
 
-	return amdgpu_ctx_wait_prev_fence(parser->ctx, parser->job->ring->idx);
+	return amdgpu_ctx_wait_prev_fence(parser->ctx, parser->ring->idx);
 }
 
 static int amdgpu_cs_process_fence_dep(struct amdgpu_cs_parser *p,
@@ -1154,8 +1207,9 @@ static void amdgpu_cs_post_dependencies(struct amdgpu_cs_parser *p)
 static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 			    union drm_amdgpu_cs *cs)
 {
-	struct amdgpu_ring *ring = p->job->ring;
+	struct amdgpu_ring *ring = p->ring;
 	struct drm_sched_entity *entity = &p->ctx->rings[ring->idx].entity;
+	enum drm_sched_priority priority;
 	struct amdgpu_job *job;
 	unsigned i;
 	uint64_t seq;
@@ -1186,7 +1240,6 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 	}
 
 	job->owner = p->filp;
-	job->fence_ctx = entity->fence_context;
 	p->fence = dma_fence_get(&job->base.s_fence->finished);
 
 	r = amdgpu_ctx_add_fence(p->ctx, ring, p->fence, &seq);
@@ -1204,11 +1257,14 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
 	job->uf_sequence = seq;
 
 	amdgpu_job_free_resources(job);
-	amdgpu_ring_priority_get(job->ring, job->base.s_priority);
 
 	trace_amdgpu_cs_ioctl(job);
+	priority = job->base.s_priority;
 	drm_sched_entity_push_job(&job->base, entity);
 
+	ring = to_amdgpu_ring(entity->sched);
+	amdgpu_ring_priority_get(ring, priority);
+
 	ttm_eu_fence_buffer_objects(&p->ticket, &p->validated, p->fence);
 	amdgpu_mn_unlock(p->mn);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
index 09d35051fdd6..83e3b320a793 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
@@ -90,8 +90,8 @@ static int amdgpu_ctx_init(struct amdgpu_device *adev,
 		if (ring == &adev->gfx.kiq.ring)
 			continue;
 
-		r = drm_sched_entity_init(&ring->sched, &ctx->rings[i].entity,
-					  rq, amdgpu_sched_jobs, &ctx->guilty);
+		r = drm_sched_entity_init(&ctx->rings[i].entity,
+					  &rq, 1, &ctx->guilty);
 		if (r)
 			goto failed;
 	}
@@ -104,15 +104,16 @@ static int amdgpu_ctx_init(struct amdgpu_device *adev,
 
 failed:
 	for (j = 0; j < i; j++)
-		drm_sched_entity_fini(&adev->rings[j]->sched,
+		drm_sched_entity_destroy(&adev->rings[j]->sched,
 				      &ctx->rings[j].entity);
 	kfree(ctx->fences);
 	ctx->fences = NULL;
 	return r;
 }
 
-static void amdgpu_ctx_fini(struct amdgpu_ctx *ctx)
+static void amdgpu_ctx_fini(struct kref *ref)
 {
+	struct amdgpu_ctx *ctx = container_of(ref, struct amdgpu_ctx, refcount);
 	struct amdgpu_device *adev = ctx->adev;
 	unsigned i, j;
 
@@ -125,13 +126,11 @@ static void amdgpu_ctx_fini(struct amdgpu_ctx *ctx)
 	kfree(ctx->fences);
 	ctx->fences = NULL;
 
-	for (i = 0; i < adev->num_rings; i++)
-		drm_sched_entity_fini(&adev->rings[i]->sched,
-				      &ctx->rings[i].entity);
-
 	amdgpu_queue_mgr_fini(adev, &ctx->queue_mgr);
 
 	mutex_destroy(&ctx->lock);
+
+	kfree(ctx);
 }
 
 static int amdgpu_ctx_alloc(struct amdgpu_device *adev,
@@ -170,12 +169,20 @@ static int amdgpu_ctx_alloc(struct amdgpu_device *adev,
 static void amdgpu_ctx_do_release(struct kref *ref)
 {
 	struct amdgpu_ctx *ctx;
+	u32 i;
 
 	ctx = container_of(ref, struct amdgpu_ctx, refcount);
 
-	amdgpu_ctx_fini(ctx);
+	for (i = 0; i < ctx->adev->num_rings; i++) {
 
-	kfree(ctx);
+		if (ctx->adev->rings[i] == &ctx->adev->gfx.kiq.ring)
+			continue;
+
+		drm_sched_entity_destroy(&ctx->adev->rings[i]->sched,
+			&ctx->rings[i].entity);
+	}
+
+	amdgpu_ctx_fini(ref);
 }
 
 static int amdgpu_ctx_free(struct amdgpu_fpriv *fpriv, uint32_t id)
@@ -419,9 +426,11 @@ int amdgpu_ctx_wait_prev_fence(struct amdgpu_ctx *ctx, unsigned ring_id)
 
 	if (other) {
 		signed long r;
-		r = dma_fence_wait_timeout(other, false, MAX_SCHEDULE_TIMEOUT);
+		r = dma_fence_wait(other, true);
 		if (r < 0) {
-			DRM_ERROR("Error (%ld) waiting for fence!\n", r);
+			if (r != -ERESTARTSYS)
+				DRM_ERROR("Error (%ld) waiting for fence!\n", r);
+
 			return r;
 		}
 	}
@@ -435,16 +444,74 @@ void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr)
 	idr_init(&mgr->ctx_handles);
 }
 
+void amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr)
+{
+	struct amdgpu_ctx *ctx;
+	struct idr *idp;
+	uint32_t id, i;
+	long max_wait = MAX_WAIT_SCHED_ENTITY_Q_EMPTY;
+
+	idp = &mgr->ctx_handles;
+
+	mutex_lock(&mgr->lock);
+	idr_for_each_entry(idp, ctx, id) {
+
+		if (!ctx->adev) {
+			mutex_unlock(&mgr->lock);
+			return;
+		}
+
+		for (i = 0; i < ctx->adev->num_rings; i++) {
+
+			if (ctx->adev->rings[i] == &ctx->adev->gfx.kiq.ring)
+				continue;
+
+			max_wait = drm_sched_entity_flush(&ctx->adev->rings[i]->sched,
+					  &ctx->rings[i].entity, max_wait);
+		}
+	}
+	mutex_unlock(&mgr->lock);
+}
+
+void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr)
+{
+	struct amdgpu_ctx *ctx;
+	struct idr *idp;
+	uint32_t id, i;
+
+	idp = &mgr->ctx_handles;
+
+	idr_for_each_entry(idp, ctx, id) {
+
+		if (!ctx->adev)
+			return;
+
+		for (i = 0; i < ctx->adev->num_rings; i++) {
+
+			if (ctx->adev->rings[i] == &ctx->adev->gfx.kiq.ring)
+				continue;
+
+			if (kref_read(&ctx->refcount) == 1)
+				drm_sched_entity_fini(&ctx->adev->rings[i]->sched,
+					&ctx->rings[i].entity);
+			else
+				DRM_ERROR("ctx %p is still alive\n", ctx);
+		}
+	}
+}
+
 void amdgpu_ctx_mgr_fini(struct amdgpu_ctx_mgr *mgr)
 {
 	struct amdgpu_ctx *ctx;
 	struct idr *idp;
 	uint32_t id;
 
+	amdgpu_ctx_mgr_entity_fini(mgr);
+
 	idp = &mgr->ctx_handles;
 
 	idr_for_each_entry(idp, ctx, id) {
-		if (kref_put(&ctx->refcount, amdgpu_ctx_do_release) != 1)
+		if (kref_put(&ctx->refcount, amdgpu_ctx_fini) != 1)
 			DRM_ERROR("ctx %p is still alive\n", ctx);
 	}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
index 448d69fe3756..f5fb93795a69 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
@@ -28,8 +28,13 @@
 #include <linux/debugfs.h>
 #include "amdgpu.h"
 
-/*
- * Debugfs
+/**
+ * amdgpu_debugfs_add_files - Add simple debugfs entries
+ *
+ * @adev:  Device to attach debugfs entries to
+ * @files:  Array of function callbacks that respond to reads
+ * @nfiles: Number of callbacks to register
+ *
  */
 int amdgpu_debugfs_add_files(struct amdgpu_device *adev,
 			     const struct drm_info_list *files,
@@ -64,7 +69,33 @@ int amdgpu_debugfs_add_files(struct amdgpu_device *adev,
 
 #if defined(CONFIG_DEBUG_FS)
 
-
+/**
+ * amdgpu_debugfs_process_reg_op - Handle MMIO register reads/writes
+ *
+ * @read: True if reading
+ * @f: open file handle
+ * @buf: User buffer to write/read to
+ * @size: Number of bytes to write/read
+ * @pos:  Offset to seek to
+ *
+ * This debugfs entry has special meaning on the offset being sought.
+ * Various bits have different meanings:
+ *
+ * Bit 62:  Indicates a GRBM bank switch is needed
+ * Bit 61:  Indicates a SRBM bank switch is needed (implies bit 62 is
+ * 			zero)
+ * Bits 24..33: The SE or ME selector if needed
+ * Bits 34..43: The SH (or SA) or PIPE selector if needed
+ * Bits 44..53: The INSTANCE (or CU/WGP) or QUEUE selector if needed
+ *
+ * Bit 23:  Indicates that the PM power gating lock should be held
+ * 			This is necessary to read registers that might be
+ * 			unreliable during a power gating transistion.
+ *
+ * The lower bits are the BYTE offset of the register to read.  This
+ * allows reading multiple registers in a single call and having
+ * the returned size reflect that.
+ */
 static int  amdgpu_debugfs_process_reg_op(bool read, struct file *f,
 		char __user *buf, size_t size, loff_t *pos)
 {
@@ -164,19 +195,37 @@ end:
 	return result;
 }
 
-
+/**
+ * amdgpu_debugfs_regs_read - Callback for reading MMIO registers
+ */
 static ssize_t amdgpu_debugfs_regs_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
 	return amdgpu_debugfs_process_reg_op(true, f, buf, size, pos);
 }
 
+/**
+ * amdgpu_debugfs_regs_write - Callback for writing MMIO registers
+ */
 static ssize_t amdgpu_debugfs_regs_write(struct file *f, const char __user *buf,
 					 size_t size, loff_t *pos)
 {
 	return amdgpu_debugfs_process_reg_op(false, f, (char __user *)buf, size, pos);
 }
 
+
+/**
+ * amdgpu_debugfs_regs_pcie_read - Read from a PCIE register
+ *
+ * @f: open file handle
+ * @buf: User buffer to store read data in
+ * @size: Number of bytes to read
+ * @pos:  Offset to seek to
+ *
+ * The lower bits are the BYTE offset of the register to read.  This
+ * allows reading multiple registers in a single call and having
+ * the returned size reflect that.
+ */
 static ssize_t amdgpu_debugfs_regs_pcie_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
@@ -204,6 +253,18 @@ static ssize_t amdgpu_debugfs_regs_pcie_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_regs_pcie_write - Write to a PCIE register
+ *
+ * @f: open file handle
+ * @buf: User buffer to write data from
+ * @size: Number of bytes to write
+ * @pos:  Offset to seek to
+ *
+ * The lower bits are the BYTE offset of the register to write.  This
+ * allows writing multiple registers in a single call and having
+ * the returned size reflect that.
+ */
 static ssize_t amdgpu_debugfs_regs_pcie_write(struct file *f, const char __user *buf,
 					 size_t size, loff_t *pos)
 {
@@ -232,6 +293,18 @@ static ssize_t amdgpu_debugfs_regs_pcie_write(struct file *f, const char __user
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_regs_didt_read - Read from a DIDT register
+ *
+ * @f: open file handle
+ * @buf: User buffer to store read data in
+ * @size: Number of bytes to read
+ * @pos:  Offset to seek to
+ *
+ * The lower bits are the BYTE offset of the register to read.  This
+ * allows reading multiple registers in a single call and having
+ * the returned size reflect that.
+ */
 static ssize_t amdgpu_debugfs_regs_didt_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
@@ -259,6 +332,18 @@ static ssize_t amdgpu_debugfs_regs_didt_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_regs_didt_write - Write to a DIDT register
+ *
+ * @f: open file handle
+ * @buf: User buffer to write data from
+ * @size: Number of bytes to write
+ * @pos:  Offset to seek to
+ *
+ * The lower bits are the BYTE offset of the register to write.  This
+ * allows writing multiple registers in a single call and having
+ * the returned size reflect that.
+ */
 static ssize_t amdgpu_debugfs_regs_didt_write(struct file *f, const char __user *buf,
 					 size_t size, loff_t *pos)
 {
@@ -287,6 +372,18 @@ static ssize_t amdgpu_debugfs_regs_didt_write(struct file *f, const char __user
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_regs_smc_read - Read from a SMC register
+ *
+ * @f: open file handle
+ * @buf: User buffer to store read data in
+ * @size: Number of bytes to read
+ * @pos:  Offset to seek to
+ *
+ * The lower bits are the BYTE offset of the register to read.  This
+ * allows reading multiple registers in a single call and having
+ * the returned size reflect that.
+ */
 static ssize_t amdgpu_debugfs_regs_smc_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
@@ -314,6 +411,18 @@ static ssize_t amdgpu_debugfs_regs_smc_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_regs_smc_write - Write to a SMC register
+ *
+ * @f: open file handle
+ * @buf: User buffer to write data from
+ * @size: Number of bytes to write
+ * @pos:  Offset to seek to
+ *
+ * The lower bits are the BYTE offset of the register to write.  This
+ * allows writing multiple registers in a single call and having
+ * the returned size reflect that.
+ */
 static ssize_t amdgpu_debugfs_regs_smc_write(struct file *f, const char __user *buf,
 					 size_t size, loff_t *pos)
 {
@@ -342,6 +451,20 @@ static ssize_t amdgpu_debugfs_regs_smc_write(struct file *f, const char __user *
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_gca_config_read - Read from gfx config data
+ *
+ * @f: open file handle
+ * @buf: User buffer to store read data in
+ * @size: Number of bytes to read
+ * @pos:  Offset to seek to
+ *
+ * This file is used to access configuration data in a somewhat
+ * stable fashion.  The format is a series of DWORDs with the first
+ * indicating which revision it is.  New content is appended to the
+ * end so that older software can still read the data.
+ */
+
 static ssize_t amdgpu_debugfs_gca_config_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
@@ -418,6 +541,19 @@ static ssize_t amdgpu_debugfs_gca_config_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/**
+ * amdgpu_debugfs_sensor_read - Read from the powerplay sensors
+ *
+ * @f: open file handle
+ * @buf: User buffer to store read data in
+ * @size: Number of bytes to read
+ * @pos:  Offset to seek to
+ *
+ * The offset is treated as the BYTE address of one of the sensors
+ * enumerated in amd/include/kgd_pp_interface.h under the
+ * 'amd_pp_sensors' enumeration.  For instance to read the UVD VCLK
+ * you would use the offset 3 * 4 = 12.
+ */
 static ssize_t amdgpu_debugfs_sensor_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
@@ -428,7 +564,7 @@ static ssize_t amdgpu_debugfs_sensor_read(struct file *f, char __user *buf,
 	if (size & 3 || *pos & 0x3)
 		return -EINVAL;
 
-	if (amdgpu_dpm == 0)
+	if (!adev->pm.dpm_enabled)
 		return -EINVAL;
 
 	/* convert offset to sensor number */
@@ -457,6 +593,27 @@ static ssize_t amdgpu_debugfs_sensor_read(struct file *f, char __user *buf,
 	return !r ? outsize : r;
 }
 
+/** amdgpu_debugfs_wave_read - Read WAVE STATUS data
+ *
+ * @f: open file handle
+ * @buf: User buffer to store read data in
+ * @size: Number of bytes to read
+ * @pos:  Offset to seek to
+ *
+ * The offset being sought changes which wave that the status data
+ * will be returned for.  The bits are used as follows:
+ *
+ * Bits 0..6: 	Byte offset into data
+ * Bits 7..14:	SE selector
+ * Bits 15..22:	SH/SA selector
+ * Bits 23..30: CU/{WGP+SIMD} selector
+ * Bits 31..36: WAVE ID selector
+ * Bits 37..44: SIMD ID selector
+ *
+ * The returned data begins with one DWORD of version information
+ * Followed by WAVE STATUS registers relevant to the GFX IP version
+ * being used.  See gfx_v8_0_read_wave_data() for an example output.
+ */
 static ssize_t amdgpu_debugfs_wave_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
@@ -507,6 +664,28 @@ static ssize_t amdgpu_debugfs_wave_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/** amdgpu_debugfs_gpr_read - Read wave gprs
+ *
+ * @f: open file handle
+ * @buf: User buffer to store read data in
+ * @size: Number of bytes to read
+ * @pos:  Offset to seek to
+ *
+ * The offset being sought changes which wave that the status data
+ * will be returned for.  The bits are used as follows:
+ *
+ * Bits 0..11:	Byte offset into data
+ * Bits 12..19:	SE selector
+ * Bits 20..27:	SH/SA selector
+ * Bits 28..35: CU/{WGP+SIMD} selector
+ * Bits 36..43: WAVE ID selector
+ * Bits 37..44: SIMD ID selector
+ * Bits 52..59: Thread selector
+ * Bits 60..61: Bank selector (VGPR=0,SGPR=1)
+ *
+ * The return data comes from the SGPR or VGPR register bank for
+ * the selected operational unit.
+ */
 static ssize_t amdgpu_debugfs_gpr_read(struct file *f, char __user *buf,
 					size_t size, loff_t *pos)
 {
@@ -637,6 +816,12 @@ static const char *debugfs_regs_names[] = {
 	"amdgpu_gpr",
 };
 
+/**
+ * amdgpu_debugfs_regs_init -	Initialize debugfs entries that provide
+ * 								register access.
+ *
+ * @adev: The device to attach the debugfs entries to
+ */
 int amdgpu_debugfs_regs_init(struct amdgpu_device *adev)
 {
 	struct drm_minor *minor = adev->ddev->primary;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 34af664b9f93..386a7b34d2f4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -25,6 +25,7 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
+#include <linux/power_supply.h>
 #include <linux/kthread.h>
 #include <linux/console.h>
 #include <linux/slab.h>
@@ -83,8 +84,10 @@ static const char *amdgpu_asic_name[] = {
 	"POLARIS10",
 	"POLARIS11",
 	"POLARIS12",
+	"VEGAM",
 	"VEGA10",
 	"VEGA12",
+	"VEGA20",
 	"RAVEN",
 	"LAST",
 };
@@ -673,34 +676,34 @@ void amdgpu_device_vram_location(struct amdgpu_device *adev,
 }
 
 /**
- * amdgpu_device_gart_location - try to find GTT location
+ * amdgpu_device_gart_location - try to find GART location
  *
  * @adev: amdgpu device structure holding all necessary informations
  * @mc: memory controller structure holding memory informations
  *
- * Function will place try to place GTT before or after VRAM.
+ * Function will place try to place GART before or after VRAM.
  *
- * If GTT size is bigger than space left then we ajust GTT size.
+ * If GART size is bigger than space left then we ajust GART size.
  * Thus function will never fails.
- *
- * FIXME: when reducing GTT size align new size on power of 2.
  */
 void amdgpu_device_gart_location(struct amdgpu_device *adev,
 				 struct amdgpu_gmc *mc)
 {
 	u64 size_af, size_bf;
 
+	mc->gart_size += adev->pm.smu_prv_buffer_size;
+
 	size_af = adev->gmc.mc_mask - mc->vram_end;
 	size_bf = mc->vram_start;
 	if (size_bf > size_af) {
 		if (mc->gart_size > size_bf) {
-			dev_warn(adev->dev, "limiting GTT\n");
+			dev_warn(adev->dev, "limiting GART\n");
 			mc->gart_size = size_bf;
 		}
 		mc->gart_start = 0;
 	} else {
 		if (mc->gart_size > size_af) {
-			dev_warn(adev->dev, "limiting GTT\n");
+			dev_warn(adev->dev, "limiting GART\n");
 			mc->gart_size = size_af;
 		}
 		/* VCE doesn't like it when BOs cross a 4GB segment, so align
@@ -709,7 +712,7 @@ void amdgpu_device_gart_location(struct amdgpu_device *adev,
 		mc->gart_start = ALIGN(mc->vram_end + 1, 0x100000000ULL);
 	}
 	mc->gart_end = mc->gart_start + mc->gart_size - 1;
-	dev_info(adev->dev, "GTT: %lluM 0x%016llX - 0x%016llX\n",
+	dev_info(adev->dev, "GART: %lluM 0x%016llX - 0x%016llX\n",
 			mc->gart_size >> 20, mc->gart_start, mc->gart_end);
 }
 
@@ -907,6 +910,46 @@ static void amdgpu_device_check_vm_size(struct amdgpu_device *adev)
 	}
 }
 
+static void amdgpu_device_check_smu_prv_buffer_size(struct amdgpu_device *adev)
+{
+	struct sysinfo si;
+	bool is_os_64 = (sizeof(void *) == 8) ? true : false;
+	uint64_t total_memory;
+	uint64_t dram_size_seven_GB = 0x1B8000000;
+	uint64_t dram_size_three_GB = 0xB8000000;
+
+	if (amdgpu_smu_memory_pool_size == 0)
+		return;
+
+	if (!is_os_64) {
+		DRM_WARN("Not 64-bit OS, feature not supported\n");
+		goto def_value;
+	}
+	si_meminfo(&si);
+	total_memory = (uint64_t)si.totalram * si.mem_unit;
+
+	if ((amdgpu_smu_memory_pool_size == 1) ||
+		(amdgpu_smu_memory_pool_size == 2)) {
+		if (total_memory < dram_size_three_GB)
+			goto def_value1;
+	} else if ((amdgpu_smu_memory_pool_size == 4) ||
+		(amdgpu_smu_memory_pool_size == 8)) {
+		if (total_memory < dram_size_seven_GB)
+			goto def_value1;
+	} else {
+		DRM_WARN("Smu memory pool size not supported\n");
+		goto def_value;
+	}
+	adev->pm.smu_prv_buffer_size = amdgpu_smu_memory_pool_size << 28;
+
+	return;
+
+def_value1:
+	DRM_WARN("No enough system memory\n");
+def_value:
+	adev->pm.smu_prv_buffer_size = 0;
+}
+
 /**
  * amdgpu_device_check_arguments - validate module params
  *
@@ -948,6 +991,8 @@ static void amdgpu_device_check_arguments(struct amdgpu_device *adev)
 		amdgpu_vm_fragment_size = -1;
 	}
 
+	amdgpu_device_check_smu_prv_buffer_size(adev);
+
 	amdgpu_device_check_vm_size(adev);
 
 	amdgpu_device_check_block_size(adev);
@@ -1031,7 +1076,7 @@ static const struct vga_switcheroo_client_ops amdgpu_switcheroo_ops = {
 /**
  * amdgpu_device_ip_set_clockgating_state - set the CG state
  *
- * @adev: amdgpu_device pointer
+ * @dev: amdgpu_device pointer
  * @block_type: Type of hardware IP (SMU, GFX, UVD, etc.)
  * @state: clockgating state (gate or ungate)
  *
@@ -1039,10 +1084,11 @@ static const struct vga_switcheroo_client_ops amdgpu_switcheroo_ops = {
  * the hardware IP specified.
  * Returns the error code from the last instance.
  */
-int amdgpu_device_ip_set_clockgating_state(struct amdgpu_device *adev,
+int amdgpu_device_ip_set_clockgating_state(void *dev,
 					   enum amd_ip_block_type block_type,
 					   enum amd_clockgating_state state)
 {
+	struct amdgpu_device *adev = dev;
 	int i, r = 0;
 
 	for (i = 0; i < adev->num_ip_blocks; i++) {
@@ -1064,7 +1110,7 @@ int amdgpu_device_ip_set_clockgating_state(struct amdgpu_device *adev,
 /**
  * amdgpu_device_ip_set_powergating_state - set the PG state
  *
- * @adev: amdgpu_device pointer
+ * @dev: amdgpu_device pointer
  * @block_type: Type of hardware IP (SMU, GFX, UVD, etc.)
  * @state: powergating state (gate or ungate)
  *
@@ -1072,10 +1118,11 @@ int amdgpu_device_ip_set_clockgating_state(struct amdgpu_device *adev,
  * the hardware IP specified.
  * Returns the error code from the last instance.
  */
-int amdgpu_device_ip_set_powergating_state(struct amdgpu_device *adev,
+int amdgpu_device_ip_set_powergating_state(void *dev,
 					   enum amd_ip_block_type block_type,
 					   enum amd_powergating_state state)
 {
+	struct amdgpu_device *adev = dev;
 	int i, r = 0;
 
 	for (i = 0; i < adev->num_ip_blocks; i++) {
@@ -1174,7 +1221,7 @@ bool amdgpu_device_ip_is_idle(struct amdgpu_device *adev,
  * amdgpu_device_ip_get_ip_block - get a hw IP pointer
  *
  * @adev: amdgpu_device pointer
- * @block_type: Type of hardware IP (SMU, GFX, UVD, etc.)
+ * @type: Type of hardware IP (SMU, GFX, UVD, etc.)
  *
  * Returns a pointer to the hardware IP block structure
  * if it exists for the asic, otherwise NULL.
@@ -1320,9 +1367,10 @@ static int amdgpu_device_parse_gpu_info_fw(struct amdgpu_device *adev)
 	case CHIP_TOPAZ:
 	case CHIP_TONGA:
 	case CHIP_FIJI:
-	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 	case CHIP_CARRIZO:
 	case CHIP_STONEY:
 #ifdef CONFIG_DRM_AMDGPU_SI
@@ -1339,6 +1387,7 @@ static int amdgpu_device_parse_gpu_info_fw(struct amdgpu_device *adev)
 	case CHIP_KABINI:
 	case CHIP_MULLINS:
 #endif
+	case CHIP_VEGA20:
 	default:
 		return 0;
 	case CHIP_VEGA10:
@@ -1428,9 +1477,10 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev)
 	case CHIP_TOPAZ:
 	case CHIP_TONGA:
 	case CHIP_FIJI:
-	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 	case CHIP_CARRIZO:
 	case CHIP_STONEY:
 		if (adev->asic_type == CHIP_CARRIZO || adev->asic_type == CHIP_STONEY)
@@ -1472,6 +1522,7 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev)
 #endif
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 	case CHIP_RAVEN:
 		if (adev->asic_type == CHIP_RAVEN)
 			adev->family = AMDGPU_FAMILY_RV;
@@ -1499,6 +1550,8 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev)
 			return -EAGAIN;
 	}
 
+	adev->powerplay.pp_feature = amdgpu_pp_feature_mask;
+
 	for (i = 0; i < adev->num_ip_blocks; i++) {
 		if ((amdgpu_ip_block_mask & (1 << i)) == 0) {
 			DRM_ERROR("disabled ip block: %d <%s>\n",
@@ -1660,6 +1713,7 @@ static int amdgpu_device_ip_late_set_cg_state(struct amdgpu_device *adev)
 		/* skip CG for VCE/UVD, it's handled specially */
 		if (adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_UVD &&
 		    adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCE &&
+		    adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCN &&
 		    adev->ip_blocks[i].version->funcs->set_clockgating_state) {
 			/* enable clockgating to save power */
 			r = adev->ip_blocks[i].version->funcs->set_clockgating_state((void *)adev,
@@ -1671,6 +1725,35 @@ static int amdgpu_device_ip_late_set_cg_state(struct amdgpu_device *adev)
 			}
 		}
 	}
+
+	return 0;
+}
+
+static int amdgpu_device_ip_late_set_pg_state(struct amdgpu_device *adev)
+{
+	int i = 0, r;
+
+	if (amdgpu_emu_mode == 1)
+		return 0;
+
+	for (i = 0; i < adev->num_ip_blocks; i++) {
+		if (!adev->ip_blocks[i].status.valid)
+			continue;
+		/* skip CG for VCE/UVD, it's handled specially */
+		if (adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_UVD &&
+		    adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCE &&
+		    adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCN &&
+		    adev->ip_blocks[i].version->funcs->set_powergating_state) {
+			/* enable powergating to save power */
+			r = adev->ip_blocks[i].version->funcs->set_powergating_state((void *)adev,
+										     AMD_PG_STATE_GATE);
+			if (r) {
+				DRM_ERROR("set_powergating_state(gate) of IP block <%s> failed %d\n",
+					  adev->ip_blocks[i].version->funcs->name, r);
+				return r;
+			}
+		}
+	}
 	return 0;
 }
 
@@ -1704,8 +1787,11 @@ static int amdgpu_device_ip_late_init(struct amdgpu_device *adev)
 		}
 	}
 
-	mod_delayed_work(system_wq, &adev->late_init_work,
-			msecs_to_jiffies(AMDGPU_RESUME_MS));
+	amdgpu_device_ip_late_set_cg_state(adev);
+	amdgpu_device_ip_late_set_pg_state(adev);
+
+	queue_delayed_work(system_wq, &adev->late_init_work,
+			   msecs_to_jiffies(AMDGPU_RESUME_MS));
 
 	amdgpu_device_fill_reset_magic(adev);
 
@@ -1742,6 +1828,8 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev)
 					  adev->ip_blocks[i].version->funcs->name, r);
 				return r;
 			}
+			if (adev->powerplay.pp_funcs->set_powergating_by_smu)
+				amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_GFX, false);
 			r = adev->ip_blocks[i].version->funcs->hw_fini((void *)adev);
 			/* XXX handle errors */
 			if (r) {
@@ -1759,6 +1847,7 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev)
 
 		if (adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_UVD &&
 			adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCE &&
+			adev->ip_blocks[i].version->type != AMD_IP_BLOCK_TYPE_VCN &&
 			adev->ip_blocks[i].version->funcs->set_clockgating_state) {
 			/* ungate blocks before hw fini so that we can shutdown the blocks safely */
 			r = adev->ip_blocks[i].version->funcs->set_clockgating_state((void *)adev,
@@ -1829,7 +1918,11 @@ static void amdgpu_device_ip_late_init_func_handler(struct work_struct *work)
 {
 	struct amdgpu_device *adev =
 		container_of(work, struct amdgpu_device, late_init_work.work);
-	amdgpu_device_ip_late_set_cg_state(adev);
+	int r;
+
+	r = amdgpu_ib_ring_tests(adev);
+	if (r)
+		DRM_ERROR("ib ring test failed (%d).\n", r);
 }
 
 /**
@@ -1857,6 +1950,10 @@ int amdgpu_device_ip_suspend(struct amdgpu_device *adev)
 		DRM_ERROR("set_clockgating_state(ungate) SMC failed %d\n", r);
 	}
 
+	/* call smu to disable gfx off feature first when suspend */
+	if (adev->powerplay.pp_funcs->set_powergating_by_smu)
+		amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_GFX, false);
+
 	for (i = adev->num_ip_blocks - 1; i >= 0; i--) {
 		if (!adev->ip_blocks[i].status.valid)
 			continue;
@@ -2080,23 +2177,30 @@ bool amdgpu_device_asic_has_dc_support(enum amd_asic_type asic_type)
 	switch (asic_type) {
 #if defined(CONFIG_DRM_AMD_DC)
 	case CHIP_BONAIRE:
-	case CHIP_HAWAII:
 	case CHIP_KAVERI:
 	case CHIP_KABINI:
 	case CHIP_MULLINS:
+		/*
+		 * We have systems in the wild with these ASICs that require
+		 * LVDS and VGA support which is not supported with DC.
+		 *
+		 * Fallback to the non-DC driver here by default so as not to
+		 * cause regressions.
+		 */
+		return amdgpu_dc > 0;
+	case CHIP_HAWAII:
 	case CHIP_CARRIZO:
 	case CHIP_STONEY:
-	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 	case CHIP_TONGA:
 	case CHIP_FIJI:
-#if defined(CONFIG_DRM_AMD_DC_PRE_VEGA)
-		return amdgpu_dc != 0;
-#endif
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+	case CHIP_VEGA20:
+#ifdef CONFIG_X86
 	case CHIP_RAVEN:
 #endif
 		return amdgpu_dc != 0;
@@ -2125,7 +2229,7 @@ bool amdgpu_device_has_dc_support(struct amdgpu_device *adev)
  * amdgpu_device_init - initialize the driver
  *
  * @adev: amdgpu_device pointer
- * @pdev: drm dev pointer
+ * @ddev: drm dev pointer
  * @pdev: pci dev pointer
  * @flags: driver flags
  *
@@ -2216,6 +2320,8 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	INIT_DELAYED_WORK(&adev->late_init_work,
 			  amdgpu_device_ip_late_init_func_handler);
 
+	adev->pm.ac_power = power_supply_is_system_supplied() > 0 ? true : false;
+
 	/* Registers mapping */
 	/* TODO: block userspace mapping of io register */
 	if (adev->asic_type >= CHIP_BONAIRE) {
@@ -2375,10 +2481,6 @@ fence_driver_init:
 		goto failed;
 	}
 
-	r = amdgpu_ib_ring_tests(adev);
-	if (r)
-		DRM_ERROR("ib ring test failed (%d).\n", r);
-
 	if (amdgpu_sriov_vf(adev))
 		amdgpu_virt_init_data_exchange(adev);
 
@@ -2500,8 +2602,9 @@ void amdgpu_device_fini(struct amdgpu_device *adev)
 /**
  * amdgpu_device_suspend - initiate device suspend
  *
- * @pdev: drm dev pointer
- * @state: suspend state
+ * @dev: drm dev pointer
+ * @suspend: suspend state
+ * @fbcon : notify the fbdev of suspend
  *
  * Puts the hw in the suspend state (all asics).
  * Returns 0 for success or an error on failure.
@@ -2539,7 +2642,7 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon)
 	/* unpin the front buffers and cursors */
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
 		struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
-		struct amdgpu_framebuffer *rfb = to_amdgpu_framebuffer(crtc->primary->fb);
+		struct drm_framebuffer *fb = crtc->primary->fb;
 		struct amdgpu_bo *robj;
 
 		if (amdgpu_crtc->cursor_bo) {
@@ -2551,10 +2654,10 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon)
 			}
 		}
 
-		if (rfb == NULL || rfb->obj == NULL) {
+		if (fb == NULL || fb->obj[0] == NULL) {
 			continue;
 		}
-		robj = gem_to_amdgpu_bo(rfb->obj);
+		robj = gem_to_amdgpu_bo(fb->obj[0]);
 		/* don't unpin kernel fb objects */
 		if (!amdgpu_fbdev_robj_is_fb(adev, robj)) {
 			r = amdgpu_bo_reserve(robj, true);
@@ -2599,7 +2702,9 @@ int amdgpu_device_suspend(struct drm_device *dev, bool suspend, bool fbcon)
 /**
  * amdgpu_device_resume - initiate device resume
  *
- * @pdev: drm dev pointer
+ * @dev: drm dev pointer
+ * @resume: resume state
+ * @fbcon : notify the fbdev of resume
  *
  * Bring the hw back to operating state (all asics).
  * Returns 0 for success or an error on failure.
@@ -2640,11 +2745,6 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon)
 	}
 	amdgpu_fence_driver_resume(adev);
 
-	if (resume) {
-		r = amdgpu_ib_ring_tests(adev);
-		if (r)
-			DRM_ERROR("ib ring test failed (%d).\n", r);
-	}
 
 	r = amdgpu_device_ip_late_init(adev);
 	if (r)
@@ -2658,11 +2758,10 @@ int amdgpu_device_resume(struct drm_device *dev, bool resume, bool fbcon)
 			struct amdgpu_bo *aobj = gem_to_amdgpu_bo(amdgpu_crtc->cursor_bo);
 			r = amdgpu_bo_reserve(aobj, true);
 			if (r == 0) {
-				r = amdgpu_bo_pin(aobj,
-						  AMDGPU_GEM_DOMAIN_VRAM,
-						  &amdgpu_crtc->cursor_addr);
+				r = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM);
 				if (r != 0)
 					DRM_ERROR("Failed to pin cursor BO (%d)\n", r);
+				amdgpu_crtc->cursor_addr = amdgpu_bo_gpu_offset(aobj);
 				amdgpu_bo_unreserve(aobj);
 			}
 		}
@@ -2736,6 +2835,9 @@ static bool amdgpu_device_ip_check_soft_reset(struct amdgpu_device *adev)
 	if (amdgpu_sriov_vf(adev))
 		return true;
 
+	if (amdgpu_asic_need_full_reset(adev))
+		return true;
+
 	for (i = 0; i < adev->num_ip_blocks; i++) {
 		if (!adev->ip_blocks[i].status.valid)
 			continue;
@@ -2792,6 +2894,9 @@ static bool amdgpu_device_ip_need_full_reset(struct amdgpu_device *adev)
 {
 	int i;
 
+	if (amdgpu_asic_need_full_reset(adev))
+		return true;
+
 	for (i = 0; i < adev->num_ip_blocks; i++) {
 		if (!adev->ip_blocks[i].status.valid)
 			continue;
@@ -3061,6 +3166,7 @@ out:
  * amdgpu_device_reset_sriov - reset ASIC for SR-IOV vf
  *
  * @adev: amdgpu device pointer
+ * @from_hypervisor: request from hypervisor
  *
  * do VF FLR and reinitialize Asic
  * return 0 means successed otherwise failed
@@ -3087,20 +3193,19 @@ static int amdgpu_device_reset_sriov(struct amdgpu_device *adev,
 
 	/* now we are okay to resume SMC/CP/SDMA */
 	r = amdgpu_device_ip_reinit_late_sriov(adev);
-	amdgpu_virt_release_full_gpu(adev, true);
 	if (r)
 		goto error;
 
 	amdgpu_irq_gpu_reset_resume_helper(adev);
 	r = amdgpu_ib_ring_tests(adev);
 
+error:
+	amdgpu_virt_release_full_gpu(adev, true);
 	if (!r && adev->virt.gim_feature & AMDGIM_FEATURE_GIM_FLR_VRAMLOST) {
 		atomic_inc(&adev->vram_lost_counter);
 		r = amdgpu_device_handle_vram_lost(adev);
 	}
 
-error:
-
 	return r;
 }
 
@@ -3109,7 +3214,7 @@ error:
  *
  * @adev: amdgpu device pointer
  * @job: which job trigger hang
- * @force forces reset regardless of amdgpu_gpu_recovery
+ * @force: forces reset regardless of amdgpu_gpu_recovery
  *
  * Attempt to reset the GPU if it has hung (all asics).
  * Returns 0 for success or an error on failure.
@@ -3117,7 +3222,6 @@ error:
 int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 			      struct amdgpu_job *job, bool force)
 {
-	struct drm_atomic_state *state = NULL;
 	int i, r, resched;
 
 	if (!force && !amdgpu_device_ip_check_soft_reset(adev)) {
@@ -3140,10 +3244,6 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 	/* block TTM */
 	resched = ttm_bo_lock_delayed_workqueue(&adev->mman.bdev);
 
-	/* store modesetting */
-	if (amdgpu_device_has_dc_support(adev))
-		state = drm_atomic_helper_suspend(adev->ddev);
-
 	/* block all schedulers and reset given job's ring */
 	for (i = 0; i < AMDGPU_MAX_RINGS; ++i) {
 		struct amdgpu_ring *ring = adev->rings[i];
@@ -3153,7 +3253,7 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 
 		kthread_park(ring->sched.thread);
 
-		if (job && job->ring->idx != i)
+		if (job && job->base.sched == &ring->sched)
 			continue;
 
 		drm_sched_hw_job_reset(&ring->sched, &job->base);
@@ -3177,16 +3277,13 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
 		 * or all rings (in the case @job is NULL)
 		 * after above amdgpu_reset accomplished
 		 */
-		if ((!job || job->ring->idx == i) && !r)
+		if ((!job || job->base.sched == &ring->sched) && !r)
 			drm_sched_job_recovery(&ring->sched);
 
 		kthread_unpark(ring->sched.thread);
 	}
 
-	if (amdgpu_device_has_dc_support(adev)) {
-		if (drm_atomic_helper_resume(adev->ddev, state))
-			dev_info(adev->dev, "drm resume failed:%d\n", r);
-	} else {
+	if (!amdgpu_device_has_dc_support(adev)) {
 		drm_helper_resume_force_mode(adev->ddev);
 	}
 
@@ -3217,8 +3314,9 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev,
  */
 static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev)
 {
-	u32 mask;
-	int ret;
+	struct pci_dev *pdev;
+	enum pci_bus_speed speed_cap;
+	enum pcie_link_width link_width;
 
 	if (amdgpu_pcie_gen_cap)
 		adev->pm.pcie_gen_mask = amdgpu_pcie_gen_cap;
@@ -3236,27 +3334,61 @@ static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev)
 	}
 
 	if (adev->pm.pcie_gen_mask == 0) {
-		ret = drm_pcie_get_speed_cap_mask(adev->ddev, &mask);
-		if (!ret) {
-			adev->pm.pcie_gen_mask = (CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1 |
+		/* asic caps */
+		pdev = adev->pdev;
+		speed_cap = pcie_get_speed_cap(pdev);
+		if (speed_cap == PCI_SPEED_UNKNOWN) {
+			adev->pm.pcie_gen_mask |= (CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1 |
 						  CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN2 |
 						  CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN3);
-
-			if (mask & DRM_PCIE_SPEED_25)
-				adev->pm.pcie_gen_mask |= CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1;
-			if (mask & DRM_PCIE_SPEED_50)
-				adev->pm.pcie_gen_mask |= CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2;
-			if (mask & DRM_PCIE_SPEED_80)
-				adev->pm.pcie_gen_mask |= CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3;
 		} else {
-			adev->pm.pcie_gen_mask = AMDGPU_DEFAULT_PCIE_GEN_MASK;
+			if (speed_cap == PCIE_SPEED_16_0GT)
+				adev->pm.pcie_gen_mask |= (CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1 |
+							  CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN2 |
+							  CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN3 |
+							  CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN4);
+			else if (speed_cap == PCIE_SPEED_8_0GT)
+				adev->pm.pcie_gen_mask |= (CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1 |
+							  CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN2 |
+							  CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN3);
+			else if (speed_cap == PCIE_SPEED_5_0GT)
+				adev->pm.pcie_gen_mask |= (CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1 |
+							  CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN2);
+			else
+				adev->pm.pcie_gen_mask |= CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1;
+		}
+		/* platform caps */
+		pdev = adev->ddev->pdev->bus->self;
+		speed_cap = pcie_get_speed_cap(pdev);
+		if (speed_cap == PCI_SPEED_UNKNOWN) {
+			adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 |
+						   CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2);
+		} else {
+			if (speed_cap == PCIE_SPEED_16_0GT)
+				adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 |
+							   CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2 |
+							   CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3 |
+							   CAIL_PCIE_LINK_SPEED_SUPPORT_GEN4);
+			else if (speed_cap == PCIE_SPEED_8_0GT)
+				adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 |
+							   CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2 |
+							   CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3);
+			else if (speed_cap == PCIE_SPEED_5_0GT)
+				adev->pm.pcie_gen_mask |= (CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1 |
+							   CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2);
+			else
+				adev->pm.pcie_gen_mask |= CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1;
+
 		}
 	}
 	if (adev->pm.pcie_mlw_mask == 0) {
-		ret = drm_pcie_get_max_link_width(adev->ddev, &mask);
-		if (!ret) {
-			switch (mask) {
-			case 32:
+		pdev = adev->ddev->pdev->bus->self;
+		link_width = pcie_get_width_cap(pdev);
+		if (link_width == PCIE_LNK_WIDTH_UNKNOWN) {
+			adev->pm.pcie_mlw_mask |= AMDGPU_DEFAULT_PCIE_MLW_MASK;
+		} else {
+			switch (link_width) {
+			case PCIE_LNK_X32:
 				adev->pm.pcie_mlw_mask = (CAIL_PCIE_LINK_WIDTH_SUPPORT_X32 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X16 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X12 |
@@ -3265,7 +3397,7 @@ static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev)
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X1);
 				break;
-			case 16:
+			case PCIE_LNK_X16:
 				adev->pm.pcie_mlw_mask = (CAIL_PCIE_LINK_WIDTH_SUPPORT_X16 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X12 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X8 |
@@ -3273,36 +3405,34 @@ static void amdgpu_device_get_pcie_info(struct amdgpu_device *adev)
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X1);
 				break;
-			case 12:
+			case PCIE_LNK_X12:
 				adev->pm.pcie_mlw_mask = (CAIL_PCIE_LINK_WIDTH_SUPPORT_X12 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X8 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X4 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X1);
 				break;
-			case 8:
+			case PCIE_LNK_X8:
 				adev->pm.pcie_mlw_mask = (CAIL_PCIE_LINK_WIDTH_SUPPORT_X8 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X4 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X1);
 				break;
-			case 4:
+			case PCIE_LNK_X4:
 				adev->pm.pcie_mlw_mask = (CAIL_PCIE_LINK_WIDTH_SUPPORT_X4 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X1);
 				break;
-			case 2:
+			case PCIE_LNK_X2:
 				adev->pm.pcie_mlw_mask = (CAIL_PCIE_LINK_WIDTH_SUPPORT_X2 |
 							  CAIL_PCIE_LINK_WIDTH_SUPPORT_X1);
 				break;
-			case 1:
+			case PCIE_LNK_X1:
 				adev->pm.pcie_mlw_mask = CAIL_PCIE_LINK_WIDTH_SUPPORT_X1;
 				break;
 			default:
 				break;
 			}
-		} else {
-			adev->pm.pcie_mlw_mask = AMDGPU_DEFAULT_PCIE_MLW_MASK;
 		}
 	}
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
index 93f700ab1bfb..6748cd7fc129 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.c
@@ -35,6 +35,7 @@
 #include <linux/pm_runtime.h>
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_edid.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 #include <drm/drm_fb_helper.h>
 
 static void amdgpu_display_flip_callback(struct dma_fence *f,
@@ -151,14 +152,11 @@ int amdgpu_display_crtc_page_flip_target(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
-	struct amdgpu_framebuffer *old_amdgpu_fb;
-	struct amdgpu_framebuffer *new_amdgpu_fb;
 	struct drm_gem_object *obj;
 	struct amdgpu_flip_work *work;
 	struct amdgpu_bo *new_abo;
 	unsigned long flags;
 	u64 tiling_flags;
-	u64 base;
 	int i, r;
 
 	work = kzalloc(sizeof *work, GFP_KERNEL);
@@ -174,15 +172,13 @@ int amdgpu_display_crtc_page_flip_target(struct drm_crtc *crtc,
 	work->async = (page_flip_flags & DRM_MODE_PAGE_FLIP_ASYNC) != 0;
 
 	/* schedule unpin of the old buffer */
-	old_amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
-	obj = old_amdgpu_fb->obj;
+	obj = crtc->primary->fb->obj[0];
 
 	/* take a reference to the old object */
 	work->old_abo = gem_to_amdgpu_bo(obj);
 	amdgpu_bo_ref(work->old_abo);
 
-	new_amdgpu_fb = to_amdgpu_framebuffer(fb);
-	obj = new_amdgpu_fb->obj;
+	obj = fb->obj[0];
 	new_abo = gem_to_amdgpu_bo(obj);
 
 	/* pin the new buffer */
@@ -192,12 +188,18 @@ int amdgpu_display_crtc_page_flip_target(struct drm_crtc *crtc,
 		goto cleanup;
 	}
 
-	r = amdgpu_bo_pin(new_abo, amdgpu_display_framebuffer_domains(adev), &base);
+	r = amdgpu_bo_pin(new_abo, amdgpu_display_supported_domains(adev));
 	if (unlikely(r != 0)) {
 		DRM_ERROR("failed to pin new abo buffer before flip\n");
 		goto unreserve;
 	}
 
+	r = amdgpu_ttm_alloc_gart(&new_abo->tbo);
+	if (unlikely(r != 0)) {
+		DRM_ERROR("%p bind failed\n", new_abo);
+		goto unpin;
+	}
+
 	r = reservation_object_get_fences_rcu(new_abo->tbo.resv, &work->excl,
 					      &work->shared_count,
 					      &work->shared);
@@ -209,7 +211,7 @@ int amdgpu_display_crtc_page_flip_target(struct drm_crtc *crtc,
 	amdgpu_bo_get_tiling_flags(new_abo, &tiling_flags);
 	amdgpu_bo_unreserve(new_abo);
 
-	work->base = base;
+	work->base = amdgpu_bo_gpu_offset(new_abo);
 	work->target_vblank = target - (uint32_t)drm_crtc_vblank_count(crtc) +
 		amdgpu_get_vblank_counter_kms(dev, work->crtc_id);
 
@@ -482,31 +484,12 @@ bool amdgpu_display_ddc_probe(struct amdgpu_connector *amdgpu_connector,
 	return true;
 }
 
-static void amdgpu_display_user_framebuffer_destroy(struct drm_framebuffer *fb)
-{
-	struct amdgpu_framebuffer *amdgpu_fb = to_amdgpu_framebuffer(fb);
-
-	drm_gem_object_put_unlocked(amdgpu_fb->obj);
-	drm_framebuffer_cleanup(fb);
-	kfree(amdgpu_fb);
-}
-
-static int amdgpu_display_user_framebuffer_create_handle(
-			struct drm_framebuffer *fb,
-			struct drm_file *file_priv,
-			unsigned int *handle)
-{
-	struct amdgpu_framebuffer *amdgpu_fb = to_amdgpu_framebuffer(fb);
-
-	return drm_gem_handle_create(file_priv, amdgpu_fb->obj, handle);
-}
-
 static const struct drm_framebuffer_funcs amdgpu_fb_funcs = {
-	.destroy = amdgpu_display_user_framebuffer_destroy,
-	.create_handle = amdgpu_display_user_framebuffer_create_handle,
+	.destroy = drm_gem_fb_destroy,
+	.create_handle = drm_gem_fb_create_handle,
 };
 
-uint32_t amdgpu_display_framebuffer_domains(struct amdgpu_device *adev)
+uint32_t amdgpu_display_supported_domains(struct amdgpu_device *adev)
 {
 	uint32_t domain = AMDGPU_GEM_DOMAIN_VRAM;
 
@@ -526,11 +509,11 @@ int amdgpu_display_framebuffer_init(struct drm_device *dev,
 				    struct drm_gem_object *obj)
 {
 	int ret;
-	rfb->obj = obj;
+	rfb->base.obj[0] = obj;
 	drm_helper_mode_fill_fb_struct(dev, &rfb->base, mode_cmd);
 	ret = drm_framebuffer_init(dev, &rfb->base, &amdgpu_fb_funcs);
 	if (ret) {
-		rfb->obj = NULL;
+		rfb->base.obj[0] = NULL;
 		return ret;
 	}
 	return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.h
index 2b11d808f297..f66e3e3fef0a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_display.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_display.h
@@ -23,7 +23,7 @@
 #ifndef __AMDGPU_DISPLAY_H__
 #define __AMDGPU_DISPLAY_H__
 
-uint32_t amdgpu_display_framebuffer_domains(struct amdgpu_device *adev);
+uint32_t amdgpu_display_supported_domains(struct amdgpu_device *adev);
 struct drm_framebuffer *
 amdgpu_display_user_framebuffer_create(struct drm_device *dev,
 				       struct drm_file *file_priv,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
index e997ebbe43ea..1c4595562f8f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.c
@@ -28,6 +28,7 @@
 #include "amdgpu_i2c.h"
 #include "amdgpu_dpm.h"
 #include "atom.h"
+#include "amd_pcie.h"
 
 void amdgpu_dpm_print_class_info(u32 class, u32 class2)
 {
@@ -115,6 +116,26 @@ void amdgpu_dpm_print_ps_status(struct amdgpu_device *adev,
 	pr_cont("\n");
 }
 
+void amdgpu_dpm_get_active_displays(struct amdgpu_device *adev)
+{
+	struct drm_device *ddev = adev->ddev;
+	struct drm_crtc *crtc;
+	struct amdgpu_crtc *amdgpu_crtc;
+
+	adev->pm.dpm.new_active_crtcs = 0;
+	adev->pm.dpm.new_active_crtc_count = 0;
+	if (adev->mode_info.num_crtc && adev->mode_info.mode_config_initialized) {
+		list_for_each_entry(crtc,
+				    &ddev->mode_config.crtc_list, head) {
+			amdgpu_crtc = to_amdgpu_crtc(crtc);
+			if (amdgpu_crtc->enabled) {
+				adev->pm.dpm.new_active_crtcs |= (1 << amdgpu_crtc->crtc_id);
+				adev->pm.dpm.new_active_crtc_count++;
+			}
+		}
+	}
+}
+
 
 u32 amdgpu_dpm_get_vblank_time(struct amdgpu_device *adev)
 {
@@ -432,7 +453,7 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev)
 			ATOM_PPLIB_PhaseSheddingLimits_Record *entry;
 
 			adev->pm.dpm.dyn_state.phase_shedding_limits_table.entries =
-				kzalloc(psl->ucNumEntries *
+				kcalloc(psl->ucNumEntries,
 					sizeof(struct amdgpu_phase_shedding_limits_entry),
 					GFP_KERNEL);
 			if (!adev->pm.dpm.dyn_state.phase_shedding_limits_table.entries) {
@@ -916,9 +937,11 @@ enum amdgpu_pcie_gen amdgpu_get_pcie_gen_support(struct amdgpu_device *adev,
 	case AMDGPU_PCIE_GEN3:
 		return AMDGPU_PCIE_GEN3;
 	default:
-		if ((sys_mask & DRM_PCIE_SPEED_80) && (default_gen == AMDGPU_PCIE_GEN3))
+		if ((sys_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3) &&
+		    (default_gen == AMDGPU_PCIE_GEN3))
 			return AMDGPU_PCIE_GEN3;
-		else if ((sys_mask & DRM_PCIE_SPEED_50) && (default_gen == AMDGPU_PCIE_GEN2))
+		else if ((sys_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2) &&
+			 (default_gen == AMDGPU_PCIE_GEN2))
 			return AMDGPU_PCIE_GEN2;
 		else
 			return AMDGPU_PCIE_GEN1;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
index 643d008410c6..ff24e1cc5b65 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_dpm.h
@@ -52,8 +52,6 @@ enum amdgpu_dpm_event_src {
 	AMDGPU_DPM_EVENT_SRC_DIGIAL_OR_EXTERNAL = 4
 };
 
-#define SCLK_DEEP_SLEEP_MASK 0x8
-
 struct amdgpu_ps {
 	u32 caps; /* vbios flags */
 	u32 class; /* vbios flags */
@@ -289,12 +287,6 @@ enum amdgpu_pcie_gen {
 #define amdgpu_dpm_force_performance_level(adev, l) \
 		((adev)->powerplay.pp_funcs->force_performance_level((adev)->powerplay.pp_handle, (l)))
 
-#define amdgpu_dpm_powergate_uvd(adev, g) \
-		((adev)->powerplay.pp_funcs->powergate_uvd((adev)->powerplay.pp_handle, (g)))
-
-#define amdgpu_dpm_powergate_vce(adev, g) \
-		((adev)->powerplay.pp_funcs->powergate_vce((adev)->powerplay.pp_handle, (g)))
-
 #define amdgpu_dpm_get_current_power_state(adev) \
 		((adev)->powerplay.pp_funcs->get_current_power_state((adev)->powerplay.pp_handle))
 
@@ -349,11 +341,9 @@ enum amdgpu_pcie_gen {
 		((adev)->powerplay.pp_funcs->set_clockgating_by_smu(\
 			(adev)->powerplay.pp_handle, msg_id))
 
-#define amdgpu_dpm_notify_smu_memory_info(adev, virtual_addr_low, \
-			virtual_addr_hi, mc_addr_low, mc_addr_hi, size) \
-		((adev)->powerplay.pp_funcs->notify_smu_memory_info)( \
-			(adev)->powerplay.pp_handle, virtual_addr_low, \
-			virtual_addr_hi, mc_addr_low, mc_addr_hi, size)
+#define amdgpu_dpm_set_powergating_by_smu(adev, block_type, gate) \
+		((adev)->powerplay.pp_funcs->set_powergating_by_smu(\
+			(adev)->powerplay.pp_handle, block_type, gate))
 
 #define amdgpu_dpm_get_power_profile_mode(adev, buf) \
 		((adev)->powerplay.pp_funcs->get_power_profile_mode(\
@@ -367,10 +357,6 @@ enum amdgpu_pcie_gen {
 		((adev)->powerplay.pp_funcs->odn_edit_dpm_table(\
 			(adev)->powerplay.pp_handle, type, parameter, size))
 
-#define amdgpu_dpm_set_mmhub_powergating_by_smu(adev) \
-		((adev)->powerplay.pp_funcs->set_mmhub_powergating_by_smu( \
-		(adev)->powerplay.pp_handle))
-
 struct amdgpu_dpm {
 	struct amdgpu_ps        *ps;
 	/* number of valid power states */
@@ -410,7 +396,6 @@ struct amdgpu_dpm {
 	u32 tdp_adjustment;
 	u16 load_line_slope;
 	bool power_control;
-	bool ac_power;
 	/* special states active */
 	bool                    thermal_active;
 	bool                    uvd_active;
@@ -445,6 +430,9 @@ struct amdgpu_pm {
 	uint32_t                pcie_gen_mask;
 	uint32_t                pcie_mlw_mask;
 	struct amd_pp_display_configuration pm_display_cfg;/* set by dc */
+	uint32_t                smu_prv_buffer_size;
+	struct amdgpu_bo        *smu_prv_buffer;
+	bool ac_power;
 };
 
 #define R600_SSTU_DFLT                               0
@@ -482,6 +470,7 @@ void amdgpu_dpm_print_ps_status(struct amdgpu_device *adev,
 				struct amdgpu_ps *rps);
 u32 amdgpu_dpm_get_vblank_time(struct amdgpu_device *adev);
 u32 amdgpu_dpm_get_vrefresh(struct amdgpu_device *adev);
+void amdgpu_dpm_get_active_displays(struct amdgpu_device *adev);
 bool amdgpu_is_uvd_state(u32 class, u32 class2);
 void amdgpu_calculate_u_and_p(u32 i, u32 r_c, u32 p_b,
 			      u32 *p, u32 *u);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index 0b19482b36b8..8843a06360fa 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -1,10 +1,3 @@
-/**
- * \file amdgpu_drv.c
- * AMD Amdgpu driver
- *
- * \author Gareth Hughes <gareth@valinux.com>
- */
-
 /*
  * Copyright 2000 VA Linux Systems, Inc., Sunnyvale, California.
  * All Rights Reserved.
@@ -75,9 +68,11 @@
  * - 3.23.0 - Add query for VRAM lost counter
  * - 3.24.0 - Add high priority compute support for gfx9
  * - 3.25.0 - Add support for sensor query info (stable pstate sclk/mclk).
+ * - 3.26.0 - GFX9: Process AMDGPU_IB_FLAG_TC_WB_NOT_INVALIDATE.
+ * - 3.27.0 - Add new chunk to to AMDGPU_CS to enable BO_LIST creation.
  */
 #define KMS_DRIVER_MAJOR	3
-#define KMS_DRIVER_MINOR	25
+#define KMS_DRIVER_MINOR	27
 #define KMS_DRIVER_PATCHLEVEL	0
 
 int amdgpu_vram_limit = 0;
@@ -109,11 +104,8 @@ int amdgpu_vram_page_split = 512;
 int amdgpu_vm_update_mode = -1;
 int amdgpu_exp_hw_support = 0;
 int amdgpu_dc = -1;
-int amdgpu_dc_log = 0;
 int amdgpu_sched_jobs = 32;
 int amdgpu_sched_hw_submission = 2;
-int amdgpu_no_evict = 0;
-int amdgpu_direct_gma_size = 0;
 uint amdgpu_pcie_gen_cap = 0;
 uint amdgpu_pcie_lane_cap = 0;
 uint amdgpu_cg_mask = 0xffffffff;
@@ -121,7 +113,8 @@ uint amdgpu_pg_mask = 0xffffffff;
 uint amdgpu_sdma_phase_quantum = 32;
 char *amdgpu_disable_cu = NULL;
 char *amdgpu_virtual_display = NULL;
-uint amdgpu_pp_feature_mask = 0xffffbfff;
+/* OverDrive(bit 14),gfxoff(bit 15),stutter mode(bit 17) disabled by default*/
+uint amdgpu_pp_feature_mask = 0xfffd3fff;
 int amdgpu_ngg = 0;
 int amdgpu_prim_buf_per_se = 0;
 int amdgpu_pos_buf_per_se = 0;
@@ -132,164 +125,370 @@ int amdgpu_lbpw = -1;
 int amdgpu_compute_multipipe = -1;
 int amdgpu_gpu_recovery = -1; /* auto */
 int amdgpu_emu_mode = 0;
+uint amdgpu_smu_memory_pool_size = 0;
 
+/**
+ * DOC: vramlimit (int)
+ * Restrict the total amount of VRAM in MiB for testing.  The default is 0 (Use full VRAM).
+ */
 MODULE_PARM_DESC(vramlimit, "Restrict VRAM for testing, in megabytes");
 module_param_named(vramlimit, amdgpu_vram_limit, int, 0600);
 
+/**
+ * DOC: vis_vramlimit (int)
+ * Restrict the amount of CPU visible VRAM in MiB for testing.  The default is 0 (Use full CPU visible VRAM).
+ */
 MODULE_PARM_DESC(vis_vramlimit, "Restrict visible VRAM for testing, in megabytes");
 module_param_named(vis_vramlimit, amdgpu_vis_vram_limit, int, 0444);
 
+/**
+ * DOC: gartsize (uint)
+ * Restrict the size of GART in Mib (32, 64, etc.) for testing. The default is -1 (The size depends on asic).
+ */
 MODULE_PARM_DESC(gartsize, "Size of GART to setup in megabytes (32, 64, etc., -1=auto)");
 module_param_named(gartsize, amdgpu_gart_size, uint, 0600);
 
+/**
+ * DOC: gttsize (int)
+ * Restrict the size of GTT domain in MiB for testing. The default is -1 (It's VRAM size if 3GB < VRAM < 3/4 RAM,
+ * otherwise 3/4 RAM size).
+ */
 MODULE_PARM_DESC(gttsize, "Size of the GTT domain in megabytes (-1 = auto)");
 module_param_named(gttsize, amdgpu_gtt_size, int, 0600);
 
+/**
+ * DOC: moverate (int)
+ * Set maximum buffer migration rate in MB/s. The default is -1 (8 MB/s).
+ */
 MODULE_PARM_DESC(moverate, "Maximum buffer migration rate in MB/s. (32, 64, etc., -1=auto, 0=1=disabled)");
 module_param_named(moverate, amdgpu_moverate, int, 0600);
 
+/**
+ * DOC: benchmark (int)
+ * Run benchmarks. The default is 0 (Skip benchmarks).
+ */
 MODULE_PARM_DESC(benchmark, "Run benchmark");
 module_param_named(benchmark, amdgpu_benchmarking, int, 0444);
 
+/**
+ * DOC: test (int)
+ * Test BO GTT->VRAM and VRAM->GTT GPU copies. The default is 0 (Skip test, only set 1 to run test).
+ */
 MODULE_PARM_DESC(test, "Run tests");
 module_param_named(test, amdgpu_testing, int, 0444);
 
+/**
+ * DOC: audio (int)
+ * Set HDMI/DPAudio. Only affects non-DC display handling. The default is -1 (Enabled), set 0 to disabled it.
+ */
 MODULE_PARM_DESC(audio, "Audio enable (-1 = auto, 0 = disable, 1 = enable)");
 module_param_named(audio, amdgpu_audio, int, 0444);
 
+/**
+ * DOC: disp_priority (int)
+ * Set display Priority (1 = normal, 2 = high). Only affects non-DC display handling. The default is 0 (auto).
+ */
 MODULE_PARM_DESC(disp_priority, "Display Priority (0 = auto, 1 = normal, 2 = high)");
 module_param_named(disp_priority, amdgpu_disp_priority, int, 0444);
 
+/**
+ * DOC: hw_i2c (int)
+ * To enable hw i2c engine. Only affects non-DC display handling. The default is 0 (Disabled).
+ */
 MODULE_PARM_DESC(hw_i2c, "hw i2c engine enable (0 = disable)");
 module_param_named(hw_i2c, amdgpu_hw_i2c, int, 0444);
 
+/**
+ * DOC: pcie_gen2 (int)
+ * To disable PCIE Gen2/3 mode (0 = disable, 1 = enable). The default is -1 (auto, enabled).
+ */
 MODULE_PARM_DESC(pcie_gen2, "PCIE Gen2 mode (-1 = auto, 0 = disable, 1 = enable)");
 module_param_named(pcie_gen2, amdgpu_pcie_gen2, int, 0444);
 
+/**
+ * DOC: msi (int)
+ * To disable Message Signaled Interrupts (MSI) functionality (1 = enable, 0 = disable). The default is -1 (auto, enabled).
+ */
 MODULE_PARM_DESC(msi, "MSI support (1 = enable, 0 = disable, -1 = auto)");
 module_param_named(msi, amdgpu_msi, int, 0444);
 
+/**
+ * DOC: lockup_timeout (int)
+ * Set GPU scheduler timeout value in ms. Value 0 is invalidated, will be adjusted to 10000.
+ * Negative values mean 'infinite timeout' (MAX_JIFFY_OFFSET). The default is 10000.
+ */
 MODULE_PARM_DESC(lockup_timeout, "GPU lockup timeout in ms > 0 (default 10000)");
 module_param_named(lockup_timeout, amdgpu_lockup_timeout, int, 0444);
 
+/**
+ * DOC: dpm (int)
+ * Override for dynamic power management setting (1 = enable, 0 = disable). The default is -1 (auto).
+ */
 MODULE_PARM_DESC(dpm, "DPM support (1 = enable, 0 = disable, -1 = auto)");
 module_param_named(dpm, amdgpu_dpm, int, 0444);
 
+/**
+ * DOC: fw_load_type (int)
+ * Set different firmware loading type for debugging (0 = direct, 1 = SMU, 2 = PSP). The default is -1 (auto).
+ */
 MODULE_PARM_DESC(fw_load_type, "firmware loading type (0 = direct, 1 = SMU, 2 = PSP, -1 = auto)");
 module_param_named(fw_load_type, amdgpu_fw_load_type, int, 0444);
 
+/**
+ * DOC: aspm (int)
+ * To disable ASPM (1 = enable, 0 = disable). The default is -1 (auto, enabled).
+ */
 MODULE_PARM_DESC(aspm, "ASPM support (1 = enable, 0 = disable, -1 = auto)");
 module_param_named(aspm, amdgpu_aspm, int, 0444);
 
+/**
+ * DOC: runpm (int)
+ * Override for runtime power management control for dGPUs in PX/HG laptops. The amdgpu driver can dynamically power down
+ * the dGPU on PX/HG laptops when it is idle. The default is -1 (auto enable). Setting the value to 0 disables this functionality.
+ */
 MODULE_PARM_DESC(runpm, "PX runtime pm (1 = force enable, 0 = disable, -1 = PX only default)");
 module_param_named(runpm, amdgpu_runtime_pm, int, 0444);
 
+/**
+ * DOC: ip_block_mask (uint)
+ * Override what IP blocks are enabled on the GPU. Each GPU is a collection of IP blocks (gfx, display, video, etc.).
+ * Use this parameter to disable specific blocks. Note that the IP blocks do not have a fixed index. Some asics may not have
+ * some IPs or may include multiple instances of an IP so the ordering various from asic to asic. See the driver output in
+ * the kernel log for the list of IPs on the asic. The default is 0xffffffff (enable all blocks on a device).
+ */
 MODULE_PARM_DESC(ip_block_mask, "IP Block Mask (all blocks enabled (default))");
 module_param_named(ip_block_mask, amdgpu_ip_block_mask, uint, 0444);
 
+/**
+ * DOC: bapm (int)
+ * Bidirectional Application Power Management (BAPM) used to dynamically share TDP between CPU and GPU. Set value 0 to disable it.
+ * The default -1 (auto, enabled)
+ */
 MODULE_PARM_DESC(bapm, "BAPM support (1 = enable, 0 = disable, -1 = auto)");
 module_param_named(bapm, amdgpu_bapm, int, 0444);
 
+/**
+ * DOC: deep_color (int)
+ * Set 1 to enable Deep Color support. Only affects non-DC display handling. The default is 0 (disabled).
+ */
 MODULE_PARM_DESC(deep_color, "Deep Color support (1 = enable, 0 = disable (default))");
 module_param_named(deep_color, amdgpu_deep_color, int, 0444);
 
+/**
+ * DOC: vm_size (int)
+ * Override the size of the GPU's per client virtual address space in GiB.  The default is -1 (automatic for each asic).
+ */
 MODULE_PARM_DESC(vm_size, "VM address space size in gigabytes (default 64GB)");
 module_param_named(vm_size, amdgpu_vm_size, int, 0444);
 
+/**
+ * DOC: vm_fragment_size (int)
+ * Override VM fragment size in bits (4, 5, etc. 4 = 64K, 9 = 2M). The default is -1 (automatic for each asic).
+ */
 MODULE_PARM_DESC(vm_fragment_size, "VM fragment size in bits (4, 5, etc. 4 = 64K (default), Max 9 = 2M)");
 module_param_named(vm_fragment_size, amdgpu_vm_fragment_size, int, 0444);
 
+/**
+ * DOC: vm_block_size (int)
+ * Override VM page table size in bits (default depending on vm_size and hw setup). The default is -1 (automatic for each asic).
+ */
 MODULE_PARM_DESC(vm_block_size, "VM page table size in bits (default depending on vm_size)");
 module_param_named(vm_block_size, amdgpu_vm_block_size, int, 0444);
 
+/**
+ * DOC: vm_fault_stop (int)
+ * Stop on VM fault for debugging (0 = never, 1 = print first, 2 = always). The default is 0 (No stop).
+ */
 MODULE_PARM_DESC(vm_fault_stop, "Stop on VM fault (0 = never (default), 1 = print first, 2 = always)");
 module_param_named(vm_fault_stop, amdgpu_vm_fault_stop, int, 0444);
 
+/**
+ * DOC: vm_debug (int)
+ * Debug VM handling (0 = disabled, 1 = enabled). The default is 0 (Disabled).
+ */
 MODULE_PARM_DESC(vm_debug, "Debug VM handling (0 = disabled (default), 1 = enabled)");
 module_param_named(vm_debug, amdgpu_vm_debug, int, 0644);
 
+/**
+ * DOC: vm_update_mode (int)
+ * Override VM update mode. VM updated by using CPU (0 = never, 1 = Graphics only, 2 = Compute only, 3 = Both). The default
+ * is -1 (Only in large BAR(LB) systems Compute VM tables will be updated by CPU, otherwise 0, never).
+ */
 MODULE_PARM_DESC(vm_update_mode, "VM update using CPU (0 = never (default except for large BAR(LB)), 1 = Graphics only, 2 = Compute only (default for LB), 3 = Both");
 module_param_named(vm_update_mode, amdgpu_vm_update_mode, int, 0444);
 
+/**
+ * DOC: vram_page_split (int)
+ * Override the number of pages after we split VRAM allocations (default 512, -1 = disable). The default is 512.
+ */
 MODULE_PARM_DESC(vram_page_split, "Number of pages after we split VRAM allocations (default 512, -1 = disable)");
 module_param_named(vram_page_split, amdgpu_vram_page_split, int, 0444);
 
+/**
+ * DOC: exp_hw_support (int)
+ * Enable experimental hw support (1 = enable). The default is 0 (disabled).
+ */
 MODULE_PARM_DESC(exp_hw_support, "experimental hw support (1 = enable, 0 = disable (default))");
 module_param_named(exp_hw_support, amdgpu_exp_hw_support, int, 0444);
 
+/**
+ * DOC: dc (int)
+ * Disable/Enable Display Core driver for debugging (1 = enable, 0 = disable). The default is -1 (automatic for each asic).
+ */
 MODULE_PARM_DESC(dc, "Display Core driver (1 = enable, 0 = disable, -1 = auto (default))");
 module_param_named(dc, amdgpu_dc, int, 0444);
 
-MODULE_PARM_DESC(dc_log, "Display Core Log Level (0 = minimal (default), 1 = chatty");
-module_param_named(dc_log, amdgpu_dc_log, int, 0444);
-
+/**
+ * DOC: sched_jobs (int)
+ * Override the max number of jobs supported in the sw queue. The default is 32.
+ */
 MODULE_PARM_DESC(sched_jobs, "the max number of jobs supported in the sw queue (default 32)");
 module_param_named(sched_jobs, amdgpu_sched_jobs, int, 0444);
 
+/**
+ * DOC: sched_hw_submission (int)
+ * Override the max number of HW submissions. The default is 2.
+ */
 MODULE_PARM_DESC(sched_hw_submission, "the max number of HW submissions (default 2)");
 module_param_named(sched_hw_submission, amdgpu_sched_hw_submission, int, 0444);
 
+/**
+ * DOC: ppfeaturemask (uint)
+ * Override power features enabled. See enum PP_FEATURE_MASK in drivers/gpu/drm/amd/include/amd_shared.h.
+ * The default is the current set of stable power features.
+ */
 MODULE_PARM_DESC(ppfeaturemask, "all power features enabled (default))");
 module_param_named(ppfeaturemask, amdgpu_pp_feature_mask, uint, 0444);
 
-MODULE_PARM_DESC(no_evict, "Support pinning request from user space (1 = enable, 0 = disable (default))");
-module_param_named(no_evict, amdgpu_no_evict, int, 0444);
-
-MODULE_PARM_DESC(direct_gma_size, "Direct GMA size in megabytes (max 96MB)");
-module_param_named(direct_gma_size, amdgpu_direct_gma_size, int, 0444);
-
+/**
+ * DOC: pcie_gen_cap (uint)
+ * Override PCIE gen speed capabilities. See the CAIL flags in drivers/gpu/drm/amd/include/amd_pcie.h.
+ * The default is 0 (automatic for each asic).
+ */
 MODULE_PARM_DESC(pcie_gen_cap, "PCIE Gen Caps (0: autodetect (default))");
 module_param_named(pcie_gen_cap, amdgpu_pcie_gen_cap, uint, 0444);
 
+/**
+ * DOC: pcie_lane_cap (uint)
+ * Override PCIE lanes capabilities. See the CAIL flags in drivers/gpu/drm/amd/include/amd_pcie.h.
+ * The default is 0 (automatic for each asic).
+ */
 MODULE_PARM_DESC(pcie_lane_cap, "PCIE Lane Caps (0: autodetect (default))");
 module_param_named(pcie_lane_cap, amdgpu_pcie_lane_cap, uint, 0444);
 
+/**
+ * DOC: cg_mask (uint)
+ * Override Clockgating features enabled on GPU (0 = disable clock gating). See the AMD_CG_SUPPORT flags in
+ * drivers/gpu/drm/amd/include/amd_shared.h. The default is 0xffffffff (all enabled).
+ */
 MODULE_PARM_DESC(cg_mask, "Clockgating flags mask (0 = disable clock gating)");
 module_param_named(cg_mask, amdgpu_cg_mask, uint, 0444);
 
+/**
+ * DOC: pg_mask (uint)
+ * Override Powergating features enabled on GPU (0 = disable power gating). See the AMD_PG_SUPPORT flags in
+ * drivers/gpu/drm/amd/include/amd_shared.h. The default is 0xffffffff (all enabled).
+ */
 MODULE_PARM_DESC(pg_mask, "Powergating flags mask (0 = disable power gating)");
 module_param_named(pg_mask, amdgpu_pg_mask, uint, 0444);
 
+/**
+ * DOC: sdma_phase_quantum (uint)
+ * Override SDMA context switch phase quantum (x 1K GPU clock cycles, 0 = no change). The default is 32.
+ */
 MODULE_PARM_DESC(sdma_phase_quantum, "SDMA context switch phase quantum (x 1K GPU clock cycles, 0 = no change (default 32))");
 module_param_named(sdma_phase_quantum, amdgpu_sdma_phase_quantum, uint, 0444);
 
+/**
+ * DOC: disable_cu (charp)
+ * Set to disable CUs (It's set like se.sh.cu,...). The default is NULL.
+ */
 MODULE_PARM_DESC(disable_cu, "Disable CUs (se.sh.cu,...)");
 module_param_named(disable_cu, amdgpu_disable_cu, charp, 0444);
 
+/**
+ * DOC: virtual_display (charp)
+ * Set to enable virtual display feature. This feature provides a virtual display hardware on headless boards
+ * or in virtualized environments. It will be set like xxxx:xx:xx.x,x;xxxx:xx:xx.x,x. It's the pci address of
+ * the device, plus the number of crtcs to expose. E.g., 0000:26:00.0,4 would enable 4 virtual crtcs on the pci
+ * device at 26:00.0. The default is NULL.
+ */
 MODULE_PARM_DESC(virtual_display,
 		 "Enable virtual display feature (the virtual_display will be set like xxxx:xx:xx.x,x;xxxx:xx:xx.x,x)");
 module_param_named(virtual_display, amdgpu_virtual_display, charp, 0444);
 
+/**
+ * DOC: ngg (int)
+ * Set to enable Next Generation Graphics (1 = enable). The default is 0 (disabled).
+ */
 MODULE_PARM_DESC(ngg, "Next Generation Graphics (1 = enable, 0 = disable(default depending on gfx))");
 module_param_named(ngg, amdgpu_ngg, int, 0444);
 
+/**
+ * DOC: prim_buf_per_se (int)
+ * Override the size of Primitive Buffer per Shader Engine in Byte. The default is 0 (depending on gfx).
+ */
 MODULE_PARM_DESC(prim_buf_per_se, "the size of Primitive Buffer per Shader Engine (default depending on gfx)");
 module_param_named(prim_buf_per_se, amdgpu_prim_buf_per_se, int, 0444);
 
+/**
+ * DOC: pos_buf_per_se (int)
+ * Override the size of Position Buffer per Shader Engine in Byte. The default is 0 (depending on gfx).
+ */
 MODULE_PARM_DESC(pos_buf_per_se, "the size of Position Buffer per Shader Engine (default depending on gfx)");
 module_param_named(pos_buf_per_se, amdgpu_pos_buf_per_se, int, 0444);
 
+/**
+ * DOC: cntl_sb_buf_per_se (int)
+ * Override the size of Control Sideband per Shader Engine in Byte. The default is 0 (depending on gfx).
+ */
 MODULE_PARM_DESC(cntl_sb_buf_per_se, "the size of Control Sideband per Shader Engine (default depending on gfx)");
 module_param_named(cntl_sb_buf_per_se, amdgpu_cntl_sb_buf_per_se, int, 0444);
 
+/**
+ * DOC: param_buf_per_se (int)
+ * Override the size of Off-Chip Pramater Cache per Shader Engine in Byte. The default is 0 (depending on gfx).
+ */
 MODULE_PARM_DESC(param_buf_per_se, "the size of Off-Chip Pramater Cache per Shader Engine (default depending on gfx)");
 module_param_named(param_buf_per_se, amdgpu_param_buf_per_se, int, 0444);
 
+/**
+ * DOC: job_hang_limit (int)
+ * Set how much time allow a job hang and not drop it. The default is 0.
+ */
 MODULE_PARM_DESC(job_hang_limit, "how much time allow a job hang and not drop it (default 0)");
 module_param_named(job_hang_limit, amdgpu_job_hang_limit, int ,0444);
 
+/**
+ * DOC: lbpw (int)
+ * Override Load Balancing Per Watt (LBPW) support (1 = enable, 0 = disable). The default is -1 (auto, enabled).
+ */
 MODULE_PARM_DESC(lbpw, "Load Balancing Per Watt (LBPW) support (1 = enable, 0 = disable, -1 = auto)");
 module_param_named(lbpw, amdgpu_lbpw, int, 0444);
 
 MODULE_PARM_DESC(compute_multipipe, "Force compute queues to be spread across pipes (1 = enable, 0 = disable, -1 = auto)");
 module_param_named(compute_multipipe, amdgpu_compute_multipipe, int, 0444);
 
+/**
+ * DOC: gpu_recovery (int)
+ * Set to enable GPU recovery mechanism (1 = enable, 0 = disable). The default is -1 (auto, disabled except SRIOV).
+ */
 MODULE_PARM_DESC(gpu_recovery, "Enable GPU recovery mechanism, (1 = enable, 0 = disable, -1 = auto)");
 module_param_named(gpu_recovery, amdgpu_gpu_recovery, int, 0444);
 
+/**
+ * DOC: emu_mode (int)
+ * Set value 1 to enable emulation mode. This is only needed when running on an emulator. The default is 0 (disabled).
+ */
 MODULE_PARM_DESC(emu_mode, "Emulation mode, (1 = enable, 0 = disable)");
 module_param_named(emu_mode, amdgpu_emu_mode, int, 0444);
 
+/**
+ * DOC: si_support (int)
+ * Set SI support driver. This parameter works after set config CONFIG_DRM_AMDGPU_SI. For SI asic, when radeon driver is enabled,
+ * set value 0 to use radeon driver, while set value 1 to use amdgpu driver. The default is using radeon driver when it available,
+ * otherwise using amdgpu driver.
+ */
 #ifdef CONFIG_DRM_AMDGPU_SI
 
 #if defined(CONFIG_DRM_RADEON) || defined(CONFIG_DRM_RADEON_MODULE)
@@ -303,6 +502,12 @@ MODULE_PARM_DESC(si_support, "SI support (1 = enabled (default), 0 = disabled)")
 module_param_named(si_support, amdgpu_si_support, int, 0444);
 #endif
 
+/**
+ * DOC: cik_support (int)
+ * Set CIK support driver. This parameter works after set config CONFIG_DRM_AMDGPU_CIK. For CIK asic, when radeon driver is enabled,
+ * set value 0 to use radeon driver, while set value 1 to use amdgpu driver. The default is using radeon driver when it available,
+ * otherwise using amdgpu driver.
+ */
 #ifdef CONFIG_DRM_AMDGPU_CIK
 
 #if defined(CONFIG_DRM_RADEON) || defined(CONFIG_DRM_RADEON_MODULE)
@@ -316,6 +521,16 @@ MODULE_PARM_DESC(cik_support, "CIK support (1 = enabled (default), 0 = disabled)
 module_param_named(cik_support, amdgpu_cik_support, int, 0444);
 #endif
 
+/**
+ * DOC: smu_memory_pool_size (uint)
+ * It is used to reserve gtt for smu debug usage, setting value 0 to disable it. The actual size is value * 256MiB.
+ * E.g. 0x1 = 256Mbyte, 0x2 = 512Mbyte, 0x4 = 1 Gbyte, 0x8 = 2GByte. The default is 0 (disabled).
+ */
+MODULE_PARM_DESC(smu_memory_pool_size,
+	"reserve gtt for smu debug usage, 0 = disable,"
+		"0x1 = 256Mbyte, 0x2 = 512Mbyte, 0x4 = 1 Gbyte, 0x8 = 2GByte");
+module_param_named(smu_memory_pool_size, amdgpu_smu_memory_pool_size, uint, 0444);
+
 static const struct pci_device_id pciidlist[] = {
 #ifdef  CONFIG_DRM_AMDGPU_SI
 	{0x1002, 0x6780, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TAHITI},
@@ -534,6 +749,9 @@ static const struct pci_device_id pciidlist[] = {
 	{0x1002, 0x6995, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS12},
 	{0x1002, 0x6997, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS12},
 	{0x1002, 0x699F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_POLARIS12},
+	/* VEGAM */
+	{0x1002, 0x694C, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGAM},
+	{0x1002, 0x694E, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGAM},
 	/* Vega 10 */
 	{0x1002, 0x6860, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
 	{0x1002, 0x6861, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA10},
@@ -550,6 +768,13 @@ static const struct pci_device_id pciidlist[] = {
 	{0x1002, 0x69A2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA12},
 	{0x1002, 0x69A3, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA12},
 	{0x1002, 0x69AF, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA12},
+	/* Vega 20 */
+	{0x1002, 0x66A0, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20|AMD_EXP_HW_SUPPORT},
+	{0x1002, 0x66A1, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20|AMD_EXP_HW_SUPPORT},
+	{0x1002, 0x66A2, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20|AMD_EXP_HW_SUPPORT},
+	{0x1002, 0x66A3, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20|AMD_EXP_HW_SUPPORT},
+	{0x1002, 0x66A7, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20|AMD_EXP_HW_SUPPORT},
+	{0x1002, 0x66AF, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_VEGA20|AMD_EXP_HW_SUPPORT},
 	/* Raven */
 	{0x1002, 0x15dd, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_RAVEN|AMD_IS_APU},
 
@@ -647,7 +872,7 @@ retry_init:
 err_pci:
 	pci_disable_device(pdev);
 err_free:
-	drm_dev_unref(dev);
+	drm_dev_put(dev);
 	return ret;
 }
 
@@ -657,7 +882,7 @@ amdgpu_pci_remove(struct pci_dev *pdev)
 	struct drm_device *dev = pci_get_drvdata(pdev);
 
 	drm_dev_unregister(dev);
-	drm_dev_unref(dev);
+	drm_dev_put(dev);
 	pci_disable_device(pdev);
 	pci_set_drvdata(pdev, NULL);
 }
@@ -838,9 +1063,21 @@ static const struct dev_pm_ops amdgpu_pm_ops = {
 	.runtime_idle = amdgpu_pmops_runtime_idle,
 };
 
+static int amdgpu_flush(struct file *f, fl_owner_t id)
+{
+	struct drm_file *file_priv = f->private_data;
+	struct amdgpu_fpriv *fpriv = file_priv->driver_priv;
+
+	amdgpu_ctx_mgr_entity_flush(&fpriv->ctx_mgr);
+
+	return 0;
+}
+
+
 static const struct file_operations amdgpu_driver_kms_fops = {
 	.owner = THIS_MODULE,
 	.open = drm_open,
+	.flush = amdgpu_flush,
 	.release = drm_release,
 	.unlocked_ioctl = amdgpu_drm_ioctl,
 	.mmap = amdgpu_mmap,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_encoders.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_encoders.c
index 94138abe093b..ae8fac34f7a5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_encoders.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_encoders.c
@@ -46,7 +46,7 @@ amdgpu_link_encoder_connector(struct drm_device *dev)
 		list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
 			amdgpu_encoder = to_amdgpu_encoder(encoder);
 			if (amdgpu_encoder->devices & amdgpu_connector->devices) {
-				drm_mode_connector_attach_encoder(connector, encoder);
+				drm_connector_attach_encoder(connector, encoder);
 				if (amdgpu_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) {
 					amdgpu_atombios_encoder_init_backlight(amdgpu_encoder, connector);
 					adev->mode_info.bl_encoder = amdgpu_encoder;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
index 12063019751b..d44b76455e89 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
@@ -137,7 +137,7 @@ static int amdgpufb_create_pinned_object(struct amdgpu_fbdev *rfbdev,
 	/* need to align pitch with crtc limits */
 	mode_cmd->pitches[0] = amdgpu_align_pitch(adev, mode_cmd->width, cpp,
 						  fb_tiled);
-	domain = amdgpu_display_framebuffer_domains(adev);
+	domain = amdgpu_display_supported_domains(adev);
 
 	height = ALIGN(mode_cmd->height, 8);
 	size = mode_cmd->pitches[0] * height;
@@ -146,7 +146,7 @@ static int amdgpufb_create_pinned_object(struct amdgpu_fbdev *rfbdev,
 				       AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
 				       AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS |
 				       AMDGPU_GEM_CREATE_VRAM_CLEARED,
-				       true, NULL, &gobj);
+				       ttm_bo_type_kernel, NULL, &gobj);
 	if (ret) {
 		pr_err("failed to allocate framebuffer (%d)\n", aligned_size);
 		return -ENOMEM;
@@ -168,11 +168,19 @@ static int amdgpufb_create_pinned_object(struct amdgpu_fbdev *rfbdev,
 	}
 
 
-	ret = amdgpu_bo_pin(abo, domain, NULL);
+	ret = amdgpu_bo_pin(abo, domain);
 	if (ret) {
 		amdgpu_bo_unreserve(abo);
 		goto out_unref;
 	}
+
+	ret = amdgpu_ttm_alloc_gart(&abo->tbo);
+	if (ret) {
+		amdgpu_bo_unreserve(abo);
+		dev_err(adev->dev, "%p bind failed\n", abo);
+		goto out_unref;
+	}
+
 	ret = amdgpu_bo_kmap(abo, NULL);
 	amdgpu_bo_unreserve(abo);
 	if (ret) {
@@ -292,9 +300,9 @@ static int amdgpu_fbdev_destroy(struct drm_device *dev, struct amdgpu_fbdev *rfb
 
 	drm_fb_helper_unregister_fbi(&rfbdev->helper);
 
-	if (rfb->obj) {
-		amdgpufb_destroy_pinned_object(rfb->obj);
-		rfb->obj = NULL;
+	if (rfb->base.obj[0]) {
+		amdgpufb_destroy_pinned_object(rfb->base.obj[0]);
+		rfb->base.obj[0] = NULL;
 		drm_framebuffer_unregister_private(&rfb->base);
 		drm_framebuffer_cleanup(&rfb->base);
 	}
@@ -377,7 +385,7 @@ int amdgpu_fbdev_total_size(struct amdgpu_device *adev)
 	if (!adev->mode_info.rfbdev)
 		return 0;
 
-	robj = gem_to_amdgpu_bo(adev->mode_info.rfbdev->rfb.obj);
+	robj = gem_to_amdgpu_bo(adev->mode_info.rfbdev->rfb.base.obj[0]);
 	size += amdgpu_bo_size(robj);
 	return size;
 }
@@ -386,7 +394,7 @@ bool amdgpu_fbdev_robj_is_fb(struct amdgpu_device *adev, struct amdgpu_bo *robj)
 {
 	if (!adev->mode_info.rfbdev)
 		return false;
-	if (robj == gem_to_amdgpu_bo(adev->mode_info.rfbdev->rfb.obj))
+	if (robj == gem_to_amdgpu_bo(adev->mode_info.rfbdev->rfb.base.obj[0]))
 		return true;
 	return false;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
index 97449e06a242..7056925eb386 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c
@@ -131,7 +131,8 @@ static u32 amdgpu_fence_read(struct amdgpu_ring *ring)
  * Emits a fence command on the requested ring (all asics).
  * Returns 0 on success, -ENOMEM on failure.
  */
-int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f)
+int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f,
+		      unsigned flags)
 {
 	struct amdgpu_device *adev = ring->adev;
 	struct amdgpu_fence *fence;
@@ -149,7 +150,7 @@ int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **f)
 		       adev->fence_context + ring->idx,
 		       seq);
 	amdgpu_ring_emit_fence(ring, ring->fence_drv.gpu_addr,
-			       seq, AMDGPU_FENCE_FLAG_INT);
+			       seq, flags | AMDGPU_FENCE_FLAG_INT);
 
 	ptr = &ring->fence_drv.fences[seq & ring->fence_drv.num_fences_mask];
 	/* This function can't be called concurrently anyway, otherwise
@@ -375,14 +376,14 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring,
 	struct amdgpu_device *adev = ring->adev;
 	uint64_t index;
 
-	if (ring != &adev->uvd.ring) {
+	if (ring->funcs->type != AMDGPU_RING_TYPE_UVD) {
 		ring->fence_drv.cpu_addr = &adev->wb.wb[ring->fence_offs];
 		ring->fence_drv.gpu_addr = adev->wb.gpu_addr + (ring->fence_offs * 4);
 	} else {
 		/* put fence directly behind firmware */
 		index = ALIGN(adev->uvd.fw->size, 8);
-		ring->fence_drv.cpu_addr = adev->uvd.cpu_addr + index;
-		ring->fence_drv.gpu_addr = adev->uvd.gpu_addr + index;
+		ring->fence_drv.cpu_addr = adev->uvd.inst[ring->me].cpu_addr + index;
+		ring->fence_drv.gpu_addr = adev->uvd.inst[ring->me].gpu_addr + index;
 	}
 	amdgpu_fence_write(ring, atomic_read(&ring->fence_drv.last_seq));
 	amdgpu_irq_get(adev, irq_src, irq_type);
@@ -645,7 +646,6 @@ static const struct dma_fence_ops amdgpu_fence_ops = {
 	.get_driver_name = amdgpu_fence_get_driver_name,
 	.get_timeline_name = amdgpu_fence_get_timeline_name,
 	.enable_signaling = amdgpu_fence_enable_signaling,
-	.wait = dma_fence_default_wait,
 	.release = amdgpu_fence_release,
 };
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
index cf0f186c6092..a54d5655a191 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.c
@@ -113,12 +113,17 @@ int amdgpu_gart_table_vram_alloc(struct amdgpu_device *adev)
 	int r;
 
 	if (adev->gart.robj == NULL) {
-		r = amdgpu_bo_create(adev, adev->gart.table_size, PAGE_SIZE,
-				     AMDGPU_GEM_DOMAIN_VRAM,
-				     AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
-				     AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS,
-				     ttm_bo_type_kernel, NULL,
-				     &adev->gart.robj);
+		struct amdgpu_bo_param bp;
+
+		memset(&bp, 0, sizeof(bp));
+		bp.size = adev->gart.table_size;
+		bp.byte_align = PAGE_SIZE;
+		bp.domain = AMDGPU_GEM_DOMAIN_VRAM;
+		bp.flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
+			AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS;
+		bp.type = ttm_bo_type_kernel;
+		bp.resv = NULL;
+		r = amdgpu_bo_create(adev, &bp, &adev->gart.robj);
 		if (r) {
 			return r;
 		}
@@ -138,14 +143,12 @@ int amdgpu_gart_table_vram_alloc(struct amdgpu_device *adev)
  */
 int amdgpu_gart_table_vram_pin(struct amdgpu_device *adev)
 {
-	uint64_t gpu_addr;
 	int r;
 
 	r = amdgpu_bo_reserve(adev->gart.robj, false);
 	if (unlikely(r != 0))
 		return r;
-	r = amdgpu_bo_pin(adev->gart.robj,
-				AMDGPU_GEM_DOMAIN_VRAM, &gpu_addr);
+	r = amdgpu_bo_pin(adev->gart.robj, AMDGPU_GEM_DOMAIN_VRAM);
 	if (r) {
 		amdgpu_bo_unreserve(adev->gart.robj);
 		return r;
@@ -154,7 +157,7 @@ int amdgpu_gart_table_vram_pin(struct amdgpu_device *adev)
 	if (r)
 		amdgpu_bo_unpin(adev->gart.robj);
 	amdgpu_bo_unreserve(adev->gart.robj);
-	adev->gart.table_addr = gpu_addr;
+	adev->gart.table_addr = amdgpu_bo_gpu_offset(adev->gart.robj);
 	return r;
 }
 
@@ -229,7 +232,7 @@ int amdgpu_gart_unbind(struct amdgpu_device *adev, uint64_t offset,
 	}
 
 	t = offset / AMDGPU_GPU_PAGE_SIZE;
-	p = t / (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE);
+	p = t / AMDGPU_GPU_PAGES_IN_CPU_PAGE;
 	for (i = 0; i < pages; i++, p++) {
 #ifdef CONFIG_DRM_AMDGPU_GART_DEBUGFS
 		adev->gart.pages[p] = NULL;
@@ -238,7 +241,7 @@ int amdgpu_gart_unbind(struct amdgpu_device *adev, uint64_t offset,
 		if (!adev->gart.ptr)
 			continue;
 
-		for (j = 0; j < (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); j++, t++) {
+		for (j = 0; j < AMDGPU_GPU_PAGES_IN_CPU_PAGE; j++, t++) {
 			amdgpu_gmc_set_pte_pde(adev, adev->gart.ptr,
 					       t, page_base, flags);
 			page_base += AMDGPU_GPU_PAGE_SIZE;
@@ -277,7 +280,7 @@ int amdgpu_gart_map(struct amdgpu_device *adev, uint64_t offset,
 
 	for (i = 0; i < pages; i++) {
 		page_base = dma_addr[i];
-		for (j = 0; j < (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE); j++, t++) {
+		for (j = 0; j < AMDGPU_GPU_PAGES_IN_CPU_PAGE; j++, t++) {
 			amdgpu_gmc_set_pte_pde(adev, dst, t, page_base, flags);
 			page_base += AMDGPU_GPU_PAGE_SIZE;
 		}
@@ -314,7 +317,7 @@ int amdgpu_gart_bind(struct amdgpu_device *adev, uint64_t offset,
 
 #ifdef CONFIG_DRM_AMDGPU_GART_DEBUGFS
 	t = offset / AMDGPU_GPU_PAGE_SIZE;
-	p = t / (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE);
+	p = t / AMDGPU_GPU_PAGES_IN_CPU_PAGE;
 	for (i = 0; i < pages; i++, p++)
 		adev->gart.pages[p] = pagelist ? pagelist[i] : NULL;
 #endif
@@ -364,7 +367,8 @@ int amdgpu_gart_init(struct amdgpu_device *adev)
 
 #ifdef CONFIG_DRM_AMDGPU_GART_DEBUGFS
 	/* Allocate pages table */
-	adev->gart.pages = vzalloc(sizeof(void *) * adev->gart.num_cpu_pages);
+	adev->gart.pages = vzalloc(array_size(sizeof(void *),
+					      adev->gart.num_cpu_pages));
 	if (adev->gart.pages == NULL)
 		return -ENOMEM;
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h
index 456295c00291..9f9e9dc87da1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gart.h
@@ -37,6 +37,8 @@ struct amdgpu_bo;
 #define AMDGPU_GPU_PAGE_SHIFT 12
 #define AMDGPU_GPU_PAGE_ALIGN(a) (((a) + AMDGPU_GPU_PAGE_MASK) & ~AMDGPU_GPU_PAGE_MASK)
 
+#define AMDGPU_GPU_PAGES_IN_CPU_PAGE (PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE)
+
 struct amdgpu_gart {
 	u64				table_addr;
 	struct amdgpu_bo		*robj;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
index 46b9ea4e6103..bcbdcf997d20 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
@@ -30,6 +30,7 @@
 #include <drm/drmP.h>
 #include <drm/amdgpu_drm.h>
 #include "amdgpu.h"
+#include "amdgpu_display.h"
 
 void amdgpu_gem_object_free(struct drm_gem_object *gobj)
 {
@@ -48,17 +49,25 @@ int amdgpu_gem_object_create(struct amdgpu_device *adev, unsigned long size,
 			     struct drm_gem_object **obj)
 {
 	struct amdgpu_bo *bo;
+	struct amdgpu_bo_param bp;
 	int r;
 
+	memset(&bp, 0, sizeof(bp));
 	*obj = NULL;
 	/* At least align on page size */
 	if (alignment < PAGE_SIZE) {
 		alignment = PAGE_SIZE;
 	}
 
+	bp.size = size;
+	bp.byte_align = alignment;
+	bp.type = type;
+	bp.resv = resv;
+	bp.preferred_domain = initial_domain;
 retry:
-	r = amdgpu_bo_create(adev, size, alignment, initial_domain,
-			     flags, type, resv, &bo);
+	bp.flags = flags;
+	bp.domain = initial_domain;
+	r = amdgpu_bo_create(adev, &bp, &bo);
 	if (r) {
 		if (r != -ERESTARTSYS) {
 			if (flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) {
@@ -221,17 +230,19 @@ int amdgpu_gem_create_ioctl(struct drm_device *dev, void *data,
 		return -EINVAL;
 
 	/* reject invalid gem domains */
-	if (args->in.domains & ~(AMDGPU_GEM_DOMAIN_CPU |
-				 AMDGPU_GEM_DOMAIN_GTT |
-				 AMDGPU_GEM_DOMAIN_VRAM |
-				 AMDGPU_GEM_DOMAIN_GDS |
-				 AMDGPU_GEM_DOMAIN_GWS |
-				 AMDGPU_GEM_DOMAIN_OA))
+	if (args->in.domains & ~AMDGPU_GEM_DOMAIN_MASK)
 		return -EINVAL;
 
 	/* create a gem object to contain this object in */
 	if (args->in.domains & (AMDGPU_GEM_DOMAIN_GDS |
 	    AMDGPU_GEM_DOMAIN_GWS | AMDGPU_GEM_DOMAIN_OA)) {
+		if (flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID) {
+			/* if gds bo is created from user space, it must be
+			 * passed to bo list
+			 */
+			DRM_ERROR("GDS bo cannot be per-vm-bo\n");
+			return -EINVAL;
+		}
 		flags |= AMDGPU_GEM_CREATE_NO_CPU_ACCESS;
 		if (args->in.domains == AMDGPU_GEM_DOMAIN_GDS)
 			size = size << AMDGPU_GDS_SHIFT;
@@ -254,7 +265,7 @@ int amdgpu_gem_create_ioctl(struct drm_device *dev, void *data,
 
 	r = amdgpu_gem_object_create(adev, size, args->in.alignment,
 				     (u32)(0xffffffff & args->in.domains),
-				     flags, false, resv, &gobj);
+				     flags, ttm_bo_type_device, resv, &gobj);
 	if (flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID) {
 		if (!r) {
 			struct amdgpu_bo *abo = gem_to_amdgpu_bo(gobj);
@@ -306,7 +317,7 @@ int amdgpu_gem_userptr_ioctl(struct drm_device *dev, void *data,
 
 	/* create a gem object to contain this object in */
 	r = amdgpu_gem_object_create(adev, args->size, 0, AMDGPU_GEM_DOMAIN_CPU,
-				     0, 0, NULL, &gobj);
+				     0, ttm_bo_type_device, NULL, &gobj);
 	if (r)
 		return r;
 
@@ -499,7 +510,6 @@ out:
  * @adev: amdgpu_device pointer
  * @vm: vm to update
  * @bo_va: bo_va to update
- * @list: validation list
  * @operation: map, unmap or clear
  *
  * Update the bo_va directly after setting its address. Errors are not
@@ -508,7 +518,6 @@ out:
 static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev,
 				    struct amdgpu_vm *vm,
 				    struct amdgpu_bo_va *bo_va,
-				    struct list_head *list,
 				    uint32_t operation)
 {
 	int r;
@@ -601,7 +610,7 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
 			return -ENOENT;
 		abo = gem_to_amdgpu_bo(gobj);
 		tv.bo = &abo->tbo;
-		tv.shared = false;
+		tv.shared = !!(abo->flags & AMDGPU_GEM_CREATE_VM_ALWAYS_VALID);
 		list_add(&tv.head, &list);
 	} else {
 		gobj = NULL;
@@ -662,7 +671,7 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
 		break;
 	}
 	if (!r && !(args->flags & AMDGPU_VM_DELAY_UPDATE) && !amdgpu_vm_debug)
-		amdgpu_gem_va_update_vm(adev, &fpriv->vm, bo_va, &list,
+		amdgpu_gem_va_update_vm(adev, &fpriv->vm, bo_va,
 					args->operation);
 
 error_backoff:
@@ -746,17 +755,18 @@ int amdgpu_mode_dumb_create(struct drm_file *file_priv,
 	struct amdgpu_device *adev = dev->dev_private;
 	struct drm_gem_object *gobj;
 	uint32_t handle;
+	u32 domain;
 	int r;
 
 	args->pitch = amdgpu_align_pitch(adev, args->width,
 					 DIV_ROUND_UP(args->bpp, 8), 0);
 	args->size = (u64)args->pitch * args->height;
 	args->size = ALIGN(args->size, PAGE_SIZE);
-
-	r = amdgpu_gem_object_create(adev, args->size, 0,
-				     AMDGPU_GEM_DOMAIN_VRAM,
+	domain = amdgpu_bo_get_preferred_pin_domain(adev,
+				amdgpu_display_supported_domains(adev));
+	r = amdgpu_gem_object_create(adev, args->size, 0, domain,
 				     AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
-				     false, NULL, &gobj);
+				     ttm_bo_type_device, NULL, &gobj);
 	if (r)
 		return -ENOMEM;
 
@@ -771,16 +781,23 @@ int amdgpu_mode_dumb_create(struct drm_file *file_priv,
 }
 
 #if defined(CONFIG_DEBUG_FS)
+
+#define amdgpu_debugfs_gem_bo_print_flag(m, bo, flag)	\
+	if (bo->flags & (AMDGPU_GEM_CREATE_ ## flag)) {	\
+		seq_printf((m), " " #flag);		\
+	}
+
 static int amdgpu_debugfs_gem_bo_info(int id, void *ptr, void *data)
 {
 	struct drm_gem_object *gobj = ptr;
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(gobj);
 	struct seq_file *m = data;
 
+	struct dma_buf_attachment *attachment;
+	struct dma_buf *dma_buf;
 	unsigned domain;
 	const char *placement;
 	unsigned pin_count;
-	uint64_t offset;
 
 	domain = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type);
 	switch (domain) {
@@ -798,13 +815,27 @@ static int amdgpu_debugfs_gem_bo_info(int id, void *ptr, void *data)
 	seq_printf(m, "\t0x%08x: %12ld byte %s",
 		   id, amdgpu_bo_size(bo), placement);
 
-	offset = READ_ONCE(bo->tbo.mem.start);
-	if (offset != AMDGPU_BO_INVALID_OFFSET)
-		seq_printf(m, " @ 0x%010Lx", offset);
-
 	pin_count = READ_ONCE(bo->pin_count);
 	if (pin_count)
 		seq_printf(m, " pin count %d", pin_count);
+
+	dma_buf = READ_ONCE(bo->gem_base.dma_buf);
+	attachment = READ_ONCE(bo->gem_base.import_attach);
+
+	if (attachment)
+		seq_printf(m, " imported from %p", dma_buf);
+	else if (dma_buf)
+		seq_printf(m, " exported as %p", dma_buf);
+
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, CPU_ACCESS_REQUIRED);
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, NO_CPU_ACCESS);
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, CPU_GTT_USWC);
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, VRAM_CLEARED);
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, SHADOW);
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, VRAM_CONTIGUOUS);
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, VM_ALWAYS_VALID);
+	amdgpu_debugfs_gem_bo_print_flag(m, bo, EXPLICIT_SYNC);
+
 	seq_printf(m, "\n");
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
index 893c2490b783..6cb4948233cb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gmc.h
@@ -109,4 +109,19 @@ struct amdgpu_gmc {
 	const struct amdgpu_gmc_funcs	*gmc_funcs;
 };
 
+/**
+ * amdgpu_gmc_vram_full_visible - Check if full VRAM is visible through the BAR
+ *
+ * @adev: amdgpu_device pointer
+ *
+ * Returns:
+ * True if full VRAM is visible through the BAR
+ */
+static inline bool amdgpu_gmc_vram_full_visible(struct amdgpu_gmc *gmc)
+{
+	WARN_ON(gmc->real_vram_size < gmc->visible_vram_size);
+
+	return (gmc->real_vram_size == gmc->visible_vram_size);
+}
+
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
index 311589e02d17..5518e623fed2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
@@ -127,6 +127,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
 	struct amdgpu_vm *vm;
 	uint64_t fence_ctx;
 	uint32_t status = 0, alloc_size;
+	unsigned fence_flags = 0;
 
 	unsigned i;
 	int r = 0;
@@ -138,7 +139,7 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
 	/* ring tests don't use a job */
 	if (job) {
 		vm = job->vm;
-		fence_ctx = job->fence_ctx;
+		fence_ctx = job->base.s_fence->scheduled.context;
 	} else {
 		vm = NULL;
 		fence_ctx = 0;
@@ -227,7 +228,16 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
 #endif
 		amdgpu_asic_invalidate_hdp(adev, ring);
 
-	r = amdgpu_fence_emit(ring, f);
+	if (ib->flags & AMDGPU_IB_FLAG_TC_WB_NOT_INVALIDATE)
+		fence_flags |= AMDGPU_FENCE_FLAG_TC_WB_ONLY;
+
+	/* wrap the last IB with fence */
+	if (job && job->uf_addr) {
+		amdgpu_ring_emit_fence(ring, job->uf_addr, job->uf_sequence,
+				       fence_flags | AMDGPU_FENCE_FLAG_64BIT);
+	}
+
+	r = amdgpu_fence_emit(ring, f, fence_flags);
 	if (r) {
 		dev_err(adev->dev, "failed to emit fence (%d)\n", r);
 		if (job && job->vmid)
@@ -239,12 +249,6 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
 	if (ring->funcs->insert_end)
 		ring->funcs->insert_end(ring);
 
-	/* wrap the last IB with fence */
-	if (job && job->uf_addr) {
-		amdgpu_ring_emit_fence(ring, job->uf_addr, job->uf_sequence,
-				       AMDGPU_FENCE_FLAG_64BIT);
-	}
-
 	if (patch_offset != ~0 && ring->funcs->patch_cond_exec)
 		amdgpu_ring_patch_cond_exec(ring, patch_offset);
 
@@ -349,7 +353,8 @@ int amdgpu_ib_ring_tests(struct amdgpu_device *adev)
 			ring->funcs->type == AMDGPU_RING_TYPE_VCE ||
 			ring->funcs->type == AMDGPU_RING_TYPE_UVD_ENC ||
 			ring->funcs->type == AMDGPU_RING_TYPE_VCN_DEC ||
-			ring->funcs->type == AMDGPU_RING_TYPE_VCN_ENC)
+			ring->funcs->type == AMDGPU_RING_TYPE_VCN_ENC ||
+			ring->funcs->type == AMDGPU_RING_TYPE_VCN_JPEG)
 			tmo = tmo_mm;
 		else
 			tmo = tmo_gfx;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c
index a1c78f90eadf..3a072a7a39f0 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ids.c
@@ -578,11 +578,6 @@ void amdgpu_vmid_mgr_init(struct amdgpu_device *adev)
 			list_add_tail(&id_mgr->ids[j].list, &id_mgr->ids_lru);
 		}
 	}
-
-	adev->vm_manager.fence_context =
-		dma_fence_context_alloc(AMDGPU_MAX_RINGS);
-	for (i = 0; i < AMDGPU_MAX_RINGS; ++i)
-		adev->vm_manager.seqno[i] = 0;
 }
 
 /**
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c
index 3a5ca462abf0..1abf5b5bac9e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c
@@ -25,6 +25,23 @@
  *          Alex Deucher
  *          Jerome Glisse
  */
+
+/**
+ * DOC: Interrupt Handling
+ *
+ * Interrupts generated within GPU hardware raise interrupt requests that are
+ * passed to amdgpu IRQ handler which is responsible for detecting source and
+ * type of the interrupt and dispatching matching handlers. If handling an
+ * interrupt requires calling kernel functions that may sleep processing is
+ * dispatched to work handlers.
+ *
+ * If MSI functionality is not disabled by module parameter then MSI
+ * support will be enabled.
+ *
+ * For GPU interrupt sources that may be driven by another driver, IRQ domain
+ * support is used (with mapping between virtual and hardware IRQs).
+ */
+
 #include <linux/irq.h>
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
@@ -43,19 +60,21 @@
 
 #define AMDGPU_WAIT_IDLE_TIMEOUT 200
 
-/*
- * Handle hotplug events outside the interrupt handler proper.
- */
 /**
- * amdgpu_hotplug_work_func - display hotplug work handler
+ * amdgpu_hotplug_work_func - work handler for display hotplug event
  *
- * @work: work struct
+ * @work: work struct pointer
  *
- * This is the hot plug event work handler (all asics).
- * The work gets scheduled from the irq handler if there
- * was a hot plug interrupt.  It walks the connector table
- * and calls the hotplug handler for each one, then sends
- * a drm hotplug event to alert userspace.
+ * This is the hotplug event work handler (all ASICs).
+ * The work gets scheduled from the IRQ handler if there
+ * was a hotplug interrupt.  It walks through the connector table
+ * and calls hotplug handler for each connector. After this, it sends
+ * a DRM hotplug event to alert userspace.
+ *
+ * This design approach is required in order to defer hotplug event handling
+ * from the IRQ handler to a work handler because hotplug handler has to use
+ * mutexes which cannot be locked in an IRQ handler (since &mutex_lock may
+ * sleep).
  */
 static void amdgpu_hotplug_work_func(struct work_struct *work)
 {
@@ -74,13 +93,12 @@ static void amdgpu_hotplug_work_func(struct work_struct *work)
 }
 
 /**
- * amdgpu_irq_reset_work_func - execute gpu reset
+ * amdgpu_irq_reset_work_func - execute GPU reset
  *
- * @work: work struct
+ * @work: work struct pointer
  *
- * Execute scheduled gpu reset (cayman+).
- * This function is called when the irq handler
- * thinks we need a gpu reset.
+ * Execute scheduled GPU reset (Cayman+).
+ * This function is called when the IRQ handler thinks we need a GPU reset.
  */
 static void amdgpu_irq_reset_work_func(struct work_struct *work)
 {
@@ -91,7 +109,13 @@ static void amdgpu_irq_reset_work_func(struct work_struct *work)
 		amdgpu_device_gpu_recover(adev, NULL, false);
 }
 
-/* Disable *all* interrupts */
+/**
+ * amdgpu_irq_disable_all - disable *all* interrupts
+ *
+ * @adev: amdgpu device pointer
+ *
+ * Disable all types of interrupts from all sources.
+ */
 void amdgpu_irq_disable_all(struct amdgpu_device *adev)
 {
 	unsigned long irqflags;
@@ -123,11 +147,15 @@ void amdgpu_irq_disable_all(struct amdgpu_device *adev)
 }
 
 /**
- * amdgpu_irq_handler - irq handler
+ * amdgpu_irq_handler - IRQ handler
+ *
+ * @irq: IRQ number (unused)
+ * @arg: pointer to DRM device
  *
- * @int irq, void *arg: args
+ * IRQ handler for amdgpu driver (all ASICs).
  *
- * This is the irq handler for the amdgpu driver (all asics).
+ * Returns:
+ * result of handling the IRQ, as defined by &irqreturn_t
  */
 irqreturn_t amdgpu_irq_handler(int irq, void *arg)
 {
@@ -142,18 +170,18 @@ irqreturn_t amdgpu_irq_handler(int irq, void *arg)
 }
 
 /**
- * amdgpu_msi_ok - asic specific msi checks
+ * amdgpu_msi_ok - check whether MSI functionality is enabled
  *
- * @adev: amdgpu device pointer
+ * @adev: amdgpu device pointer (unused)
+ *
+ * Checks whether MSI functionality has been disabled via module parameter
+ * (all ASICs).
  *
- * Handles asic specific MSI checks to determine if
- * MSIs should be enabled on a particular chip (all asics).
- * Returns true if MSIs should be enabled, false if MSIs
- * should not be enabled.
+ * Returns:
+ * *true* if MSIs are allowed to be enabled or *false* otherwise
  */
 static bool amdgpu_msi_ok(struct amdgpu_device *adev)
 {
-	/* force MSI on */
 	if (amdgpu_msi == 1)
 		return true;
 	else if (amdgpu_msi == 0)
@@ -163,12 +191,15 @@ static bool amdgpu_msi_ok(struct amdgpu_device *adev)
 }
 
 /**
- * amdgpu_irq_init - init driver interrupt info
+ * amdgpu_irq_init - initialize interrupt handling
  *
  * @adev: amdgpu device pointer
  *
- * Sets up the work irq handlers, vblank init, MSIs, etc. (all asics).
- * Returns 0 for success, error for failure.
+ * Sets up work functions for hotplug and reset interrupts, enables MSI
+ * functionality, initializes vblank, hotplug and reset interrupt handling.
+ *
+ * Returns:
+ * 0 on success or error code on failure
  */
 int amdgpu_irq_init(struct amdgpu_device *adev)
 {
@@ -176,7 +207,7 @@ int amdgpu_irq_init(struct amdgpu_device *adev)
 
 	spin_lock_init(&adev->irq.lock);
 
-	/* enable msi */
+	/* Enable MSI if not disabled by module parameter */
 	adev->irq.msi_enabled = false;
 
 	if (amdgpu_msi_ok(adev)) {
@@ -189,7 +220,7 @@ int amdgpu_irq_init(struct amdgpu_device *adev)
 
 	if (!amdgpu_device_has_dc_support(adev)) {
 		if (!adev->enable_virtual_display)
-			/* Disable vblank irqs aggressively for power-saving */
+			/* Disable vblank IRQs aggressively for power-saving */
 			/* XXX: can this be enabled for DC? */
 			adev->ddev->vblank_disable_immediate = true;
 
@@ -197,7 +228,7 @@ int amdgpu_irq_init(struct amdgpu_device *adev)
 		if (r)
 			return r;
 
-		/* pre DCE11 */
+		/* Pre-DCE11 */
 		INIT_WORK(&adev->hotplug_work,
 				amdgpu_hotplug_work_func);
 	}
@@ -220,11 +251,13 @@ int amdgpu_irq_init(struct amdgpu_device *adev)
 }
 
 /**
- * amdgpu_irq_fini - tear down driver interrupt info
+ * amdgpu_irq_fini - shut down interrupt handling
  *
  * @adev: amdgpu device pointer
  *
- * Tears down the work irq handlers, vblank handlers, MSIs, etc. (all asics).
+ * Tears down work functions for hotplug and reset interrupts, disables MSI
+ * functionality, shuts down vblank, hotplug and reset interrupt handling,
+ * turns off interrupts from all sources (all ASICs).
  */
 void amdgpu_irq_fini(struct amdgpu_device *adev)
 {
@@ -264,12 +297,17 @@ void amdgpu_irq_fini(struct amdgpu_device *adev)
 }
 
 /**
- * amdgpu_irq_add_id - register irq source
+ * amdgpu_irq_add_id - register IRQ source
  *
  * @adev: amdgpu device pointer
- * @src_id: source id for this source
- * @source: irq source
+ * @client_id: client id
+ * @src_id: source id
+ * @source: IRQ source pointer
+ *
+ * Registers IRQ source on a client.
  *
+ * Returns:
+ * 0 on success or error code otherwise
  */
 int amdgpu_irq_add_id(struct amdgpu_device *adev,
 		      unsigned client_id, unsigned src_id,
@@ -312,12 +350,12 @@ int amdgpu_irq_add_id(struct amdgpu_device *adev,
 }
 
 /**
- * amdgpu_irq_dispatch - dispatch irq to IP blocks
+ * amdgpu_irq_dispatch - dispatch IRQ to IP blocks
  *
  * @adev: amdgpu device pointer
- * @entry: interrupt vector
+ * @entry: interrupt vector pointer
  *
- * Dispatches the irq to the different IP blocks
+ * Dispatches IRQ to IP blocks.
  */
 void amdgpu_irq_dispatch(struct amdgpu_device *adev,
 			 struct amdgpu_iv_entry *entry)
@@ -361,13 +399,13 @@ void amdgpu_irq_dispatch(struct amdgpu_device *adev,
 }
 
 /**
- * amdgpu_irq_update - update hw interrupt state
+ * amdgpu_irq_update - update hardware interrupt state
  *
  * @adev: amdgpu device pointer
- * @src: interrupt src you want to enable
- * @type: type of interrupt you want to update
+ * @src: interrupt source pointer
+ * @type: type of interrupt
  *
- * Updates the interrupt state for a specific src (all asics).
+ * Updates interrupt state for the specific source (all ASICs).
  */
 int amdgpu_irq_update(struct amdgpu_device *adev,
 			     struct amdgpu_irq_src *src, unsigned type)
@@ -378,7 +416,7 @@ int amdgpu_irq_update(struct amdgpu_device *adev,
 
 	spin_lock_irqsave(&adev->irq.lock, irqflags);
 
-	/* we need to determine after taking the lock, otherwise
+	/* We need to determine after taking the lock, otherwise
 	   we might disable just enabled interrupts again */
 	if (amdgpu_irq_enabled(adev, src, type))
 		state = AMDGPU_IRQ_STATE_ENABLE;
@@ -390,6 +428,14 @@ int amdgpu_irq_update(struct amdgpu_device *adev,
 	return r;
 }
 
+/**
+ * amdgpu_irq_gpu_reset_resume_helper - update interrupt states on all sources
+ *
+ * @adev: amdgpu device pointer
+ *
+ * Updates state of all types of interrupts on all sources on resume after
+ * reset.
+ */
 void amdgpu_irq_gpu_reset_resume_helper(struct amdgpu_device *adev)
 {
 	int i, j, k;
@@ -413,10 +459,13 @@ void amdgpu_irq_gpu_reset_resume_helper(struct amdgpu_device *adev)
  * amdgpu_irq_get - enable interrupt
  *
  * @adev: amdgpu device pointer
- * @src: interrupt src you want to enable
- * @type: type of interrupt you want to enable
+ * @src: interrupt source pointer
+ * @type: type of interrupt
  *
- * Enables the interrupt type for a specific src (all asics).
+ * Enables specified type of interrupt on the specified source (all ASICs).
+ *
+ * Returns:
+ * 0 on success or error code otherwise
  */
 int amdgpu_irq_get(struct amdgpu_device *adev, struct amdgpu_irq_src *src,
 		   unsigned type)
@@ -440,10 +489,13 @@ int amdgpu_irq_get(struct amdgpu_device *adev, struct amdgpu_irq_src *src,
  * amdgpu_irq_put - disable interrupt
  *
  * @adev: amdgpu device pointer
- * @src: interrupt src you want to disable
- * @type: type of interrupt you want to disable
+ * @src: interrupt source pointer
+ * @type: type of interrupt
+ *
+ * Enables specified type of interrupt on the specified source (all ASICs).
  *
- * Disables the interrupt type for a specific src (all asics).
+ * Returns:
+ * 0 on success or error code otherwise
  */
 int amdgpu_irq_put(struct amdgpu_device *adev, struct amdgpu_irq_src *src,
 		   unsigned type)
@@ -464,12 +516,17 @@ int amdgpu_irq_put(struct amdgpu_device *adev, struct amdgpu_irq_src *src,
 }
 
 /**
- * amdgpu_irq_enabled - test if irq is enabled or not
+ * amdgpu_irq_enabled - check whether interrupt is enabled or not
  *
  * @adev: amdgpu device pointer
- * @idx: interrupt src you want to test
+ * @src: interrupt source pointer
+ * @type: type of interrupt
  *
- * Tests if the given interrupt source is enabled or not
+ * Checks whether the given type of interrupt is enabled on the given source.
+ *
+ * Returns:
+ * *true* if interrupt is enabled, *false* if interrupt is disabled or on
+ * invalid parameters
  */
 bool amdgpu_irq_enabled(struct amdgpu_device *adev, struct amdgpu_irq_src *src,
 			unsigned type)
@@ -486,7 +543,7 @@ bool amdgpu_irq_enabled(struct amdgpu_device *adev, struct amdgpu_irq_src *src,
 	return !!atomic_read(&src->enabled_types[type]);
 }
 
-/* gen irq */
+/* XXX: Generic IRQ handling */
 static void amdgpu_irq_mask(struct irq_data *irqd)
 {
 	/* XXX */
@@ -497,12 +554,26 @@ static void amdgpu_irq_unmask(struct irq_data *irqd)
 	/* XXX */
 }
 
+/* amdgpu hardware interrupt chip descriptor */
 static struct irq_chip amdgpu_irq_chip = {
 	.name = "amdgpu-ih",
 	.irq_mask = amdgpu_irq_mask,
 	.irq_unmask = amdgpu_irq_unmask,
 };
 
+/**
+ * amdgpu_irqdomain_map - create mapping between virtual and hardware IRQ numbers
+ *
+ * @d: amdgpu IRQ domain pointer (unused)
+ * @irq: virtual IRQ number
+ * @hwirq: hardware irq number
+ *
+ * Current implementation assigns simple interrupt handler to the given virtual
+ * IRQ.
+ *
+ * Returns:
+ * 0 on success or error code otherwise
+ */
 static int amdgpu_irqdomain_map(struct irq_domain *d,
 				unsigned int irq, irq_hw_number_t hwirq)
 {
@@ -514,17 +585,21 @@ static int amdgpu_irqdomain_map(struct irq_domain *d,
 	return 0;
 }
 
+/* Implementation of methods for amdgpu IRQ domain */
 static const struct irq_domain_ops amdgpu_hw_irqdomain_ops = {
 	.map = amdgpu_irqdomain_map,
 };
 
 /**
- * amdgpu_irq_add_domain - create a linear irq domain
+ * amdgpu_irq_add_domain - create a linear IRQ domain
  *
  * @adev: amdgpu device pointer
  *
- * Create an irq domain for GPU interrupt sources
+ * Creates an IRQ domain for GPU interrupt sources
  * that may be driven by another driver (e.g., ACP).
+ *
+ * Returns:
+ * 0 on success or error code otherwise
  */
 int amdgpu_irq_add_domain(struct amdgpu_device *adev)
 {
@@ -539,11 +614,11 @@ int amdgpu_irq_add_domain(struct amdgpu_device *adev)
 }
 
 /**
- * amdgpu_irq_remove_domain - remove the irq domain
+ * amdgpu_irq_remove_domain - remove the IRQ domain
  *
  * @adev: amdgpu device pointer
  *
- * Remove the irq domain for GPU interrupt sources
+ * Removes the IRQ domain for GPU interrupt sources
  * that may be driven by another driver (e.g., ACP).
  */
 void amdgpu_irq_remove_domain(struct amdgpu_device *adev)
@@ -555,16 +630,17 @@ void amdgpu_irq_remove_domain(struct amdgpu_device *adev)
 }
 
 /**
- * amdgpu_irq_create_mapping - create a mapping between a domain irq and a
- *                             Linux irq
+ * amdgpu_irq_create_mapping - create mapping between domain Linux IRQs
  *
  * @adev: amdgpu device pointer
  * @src_id: IH source id
  *
- * Create a mapping between a domain irq (GPU IH src id) and a Linux irq
+ * Creates mapping between a domain IRQ (GPU IH src id) and a Linux IRQ
  * Use this for components that generate a GPU interrupt, but are driven
  * by a different driver (e.g., ACP).
- * Returns the Linux irq.
+ *
+ * Returns:
+ * Linux IRQ
  */
 unsigned amdgpu_irq_create_mapping(struct amdgpu_device *adev, unsigned src_id)
 {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
index 2bd56760c744..5a2c26a85984 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.c
@@ -30,14 +30,14 @@
 
 static void amdgpu_job_timedout(struct drm_sched_job *s_job)
 {
-	struct amdgpu_job *job = container_of(s_job, struct amdgpu_job, base);
+	struct amdgpu_ring *ring = to_amdgpu_ring(s_job->sched);
+	struct amdgpu_job *job = to_amdgpu_job(s_job);
 
-	DRM_ERROR("ring %s timeout, last signaled seq=%u, last emitted seq=%u\n",
-		  job->base.sched->name,
-		  atomic_read(&job->ring->fence_drv.last_seq),
-		  job->ring->fence_drv.sync_seq);
+	DRM_ERROR("ring %s timeout, signaled seq=%u, emitted seq=%u\n",
+		  job->base.sched->name, atomic_read(&ring->fence_drv.last_seq),
+		  ring->fence_drv.sync_seq);
 
-	amdgpu_device_gpu_recover(job->adev, job, false);
+	amdgpu_device_gpu_recover(ring->adev, job, false);
 }
 
 int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs,
@@ -54,7 +54,11 @@ int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs,
 	if (!*job)
 		return -ENOMEM;
 
-	(*job)->adev = adev;
+	/*
+	 * Initialize the scheduler to at least some ring so that we always
+	 * have a pointer to adev.
+	 */
+	(*job)->base.sched = &adev->rings[0]->sched;
 	(*job)->vm = vm;
 	(*job)->ibs = (void *)&(*job)[1];
 	(*job)->num_ibs = num_ibs;
@@ -86,6 +90,7 @@ int amdgpu_job_alloc_with_ib(struct amdgpu_device *adev, unsigned size,
 
 void amdgpu_job_free_resources(struct amdgpu_job *job)
 {
+	struct amdgpu_ring *ring = to_amdgpu_ring(job->base.sched);
 	struct dma_fence *f;
 	unsigned i;
 
@@ -93,14 +98,15 @@ void amdgpu_job_free_resources(struct amdgpu_job *job)
 	f = job->base.s_fence ? &job->base.s_fence->finished : job->fence;
 
 	for (i = 0; i < job->num_ibs; ++i)
-		amdgpu_ib_free(job->adev, &job->ibs[i], f);
+		amdgpu_ib_free(ring->adev, &job->ibs[i], f);
 }
 
 static void amdgpu_job_free_cb(struct drm_sched_job *s_job)
 {
-	struct amdgpu_job *job = container_of(s_job, struct amdgpu_job, base);
+	struct amdgpu_ring *ring = to_amdgpu_ring(s_job->sched);
+	struct amdgpu_job *job = to_amdgpu_job(s_job);
 
-	amdgpu_ring_priority_put(job->ring, s_job->s_priority);
+	amdgpu_ring_priority_put(ring, s_job->s_priority);
 	dma_fence_put(job->fence);
 	amdgpu_sync_free(&job->sync);
 	amdgpu_sync_free(&job->sched_sync);
@@ -117,50 +123,68 @@ void amdgpu_job_free(struct amdgpu_job *job)
 	kfree(job);
 }
 
-int amdgpu_job_submit(struct amdgpu_job *job, struct amdgpu_ring *ring,
-		      struct drm_sched_entity *entity, void *owner,
-		      struct dma_fence **f)
+int amdgpu_job_submit(struct amdgpu_job *job, struct drm_sched_entity *entity,
+		      void *owner, struct dma_fence **f)
 {
+	enum drm_sched_priority priority;
+	struct amdgpu_ring *ring;
 	int r;
-	job->ring = ring;
 
 	if (!f)
 		return -EINVAL;
 
-	r = drm_sched_job_init(&job->base, &ring->sched, entity, owner);
+	r = drm_sched_job_init(&job->base, entity->sched, entity, owner);
 	if (r)
 		return r;
 
 	job->owner = owner;
-	job->fence_ctx = entity->fence_context;
 	*f = dma_fence_get(&job->base.s_fence->finished);
 	amdgpu_job_free_resources(job);
-	amdgpu_ring_priority_get(job->ring, job->base.s_priority);
+	priority = job->base.s_priority;
 	drm_sched_entity_push_job(&job->base, entity);
 
+	ring = to_amdgpu_ring(entity->sched);
+	amdgpu_ring_priority_get(ring, priority);
+
+	return 0;
+}
+
+int amdgpu_job_submit_direct(struct amdgpu_job *job, struct amdgpu_ring *ring,
+			     struct dma_fence **fence)
+{
+	int r;
+
+	job->base.sched = &ring->sched;
+	r = amdgpu_ib_schedule(ring, job->num_ibs, job->ibs, NULL, fence);
+	job->fence = dma_fence_get(*fence);
+	if (r)
+		return r;
+
+	amdgpu_job_free(job);
 	return 0;
 }
 
 static struct dma_fence *amdgpu_job_dependency(struct drm_sched_job *sched_job,
 					       struct drm_sched_entity *s_entity)
 {
+	struct amdgpu_ring *ring = to_amdgpu_ring(s_entity->sched);
 	struct amdgpu_job *job = to_amdgpu_job(sched_job);
 	struct amdgpu_vm *vm = job->vm;
+	struct dma_fence *fence;
 	bool explicit = false;
 	int r;
-	struct dma_fence *fence = amdgpu_sync_get_fence(&job->sync, &explicit);
 
+	fence = amdgpu_sync_get_fence(&job->sync, &explicit);
 	if (fence && explicit) {
 		if (drm_sched_dependency_optimized(fence, s_entity)) {
-			r = amdgpu_sync_fence(job->adev, &job->sched_sync, fence, false);
+			r = amdgpu_sync_fence(ring->adev, &job->sched_sync,
+					      fence, false);
 			if (r)
-				DRM_ERROR("Error adding fence to sync (%d)\n", r);
+				DRM_ERROR("Error adding fence (%d)\n", r);
 		}
 	}
 
 	while (fence == NULL && vm && !job->vmid) {
-		struct amdgpu_ring *ring = job->ring;
-
 		r = amdgpu_vmid_grab(vm, ring, &job->sync,
 				     &job->base.s_fence->finished,
 				     job);
@@ -175,30 +199,25 @@ static struct dma_fence *amdgpu_job_dependency(struct drm_sched_job *sched_job,
 
 static struct dma_fence *amdgpu_job_run(struct drm_sched_job *sched_job)
 {
+	struct amdgpu_ring *ring = to_amdgpu_ring(sched_job->sched);
 	struct dma_fence *fence = NULL, *finished;
-	struct amdgpu_device *adev;
 	struct amdgpu_job *job;
 	int r;
 
-	if (!sched_job) {
-		DRM_ERROR("job is null\n");
-		return NULL;
-	}
 	job = to_amdgpu_job(sched_job);
 	finished = &job->base.s_fence->finished;
-	adev = job->adev;
 
 	BUG_ON(amdgpu_sync_peek_fence(&job->sync, NULL));
 
 	trace_amdgpu_sched_run_job(job);
 
-	if (job->vram_lost_counter != atomic_read(&adev->vram_lost_counter))
+	if (job->vram_lost_counter != atomic_read(&ring->adev->vram_lost_counter))
 		dma_fence_set_error(finished, -ECANCELED);/* skip IB as well if VRAM lost */
 
 	if (finished->error < 0) {
 		DRM_INFO("Skip scheduling IBs!\n");
 	} else {
-		r = amdgpu_ib_schedule(job->ring, job->num_ibs, job->ibs, job,
+		r = amdgpu_ib_schedule(ring, job->num_ibs, job->ibs, job,
 				       &fence);
 		if (r)
 			DRM_ERROR("Error scheduling IBs (%d)\n", r);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_job.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.h
new file mode 100644
index 000000000000..57cfe78a262b
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_job.h
@@ -0,0 +1,74 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef __AMDGPU_JOB_H__
+#define __AMDGPU_JOB_H__
+
+/* bit set means command submit involves a preamble IB */
+#define AMDGPU_PREAMBLE_IB_PRESENT          (1 << 0)
+/* bit set means preamble IB is first presented in belonging context */
+#define AMDGPU_PREAMBLE_IB_PRESENT_FIRST    (1 << 1)
+/* bit set means context switch occured */
+#define AMDGPU_HAVE_CTX_SWITCH              (1 << 2)
+
+#define to_amdgpu_job(sched_job)		\
+		container_of((sched_job), struct amdgpu_job, base)
+
+struct amdgpu_fence;
+
+struct amdgpu_job {
+	struct drm_sched_job    base;
+	struct amdgpu_vm	*vm;
+	struct amdgpu_sync	sync;
+	struct amdgpu_sync	sched_sync;
+	struct amdgpu_ib	*ibs;
+	struct dma_fence	*fence; /* the hw fence */
+	uint32_t		preamble_status;
+	uint32_t		num_ibs;
+	void			*owner;
+	bool                    vm_needs_flush;
+	uint64_t		vm_pd_addr;
+	unsigned		vmid;
+	unsigned		pasid;
+	uint32_t		gds_base, gds_size;
+	uint32_t		gws_base, gws_size;
+	uint32_t		oa_base, oa_size;
+	uint32_t		vram_lost_counter;
+
+	/* user fence handling */
+	uint64_t		uf_addr;
+	uint64_t		uf_sequence;
+
+};
+
+int amdgpu_job_alloc(struct amdgpu_device *adev, unsigned num_ibs,
+		     struct amdgpu_job **job, struct amdgpu_vm *vm);
+int amdgpu_job_alloc_with_ib(struct amdgpu_device *adev, unsigned size,
+			     struct amdgpu_job **job);
+
+void amdgpu_job_free_resources(struct amdgpu_job *job);
+void amdgpu_job_free(struct amdgpu_job *job);
+int amdgpu_job_submit(struct amdgpu_job *job, struct drm_sched_entity *entity,
+		      void *owner, struct dma_fence **f);
+int amdgpu_job_submit_direct(struct amdgpu_job *job, struct amdgpu_ring *ring,
+			     struct dma_fence **fence);
+#endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
index 4b7824d30e73..207f238649b4 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
@@ -31,6 +31,7 @@
 #include "amdgpu_sched.h"
 #include "amdgpu_uvd.h"
 #include "amdgpu_vce.h"
+#include "atom.h"
 
 #include <linux/vga_switcheroo.h>
 #include <linux/slab.h>
@@ -214,6 +215,18 @@ static int amdgpu_firmware_info(struct drm_amdgpu_info_firmware *fw_info,
 		fw_info->ver = adev->gfx.rlc_fw_version;
 		fw_info->feature = adev->gfx.rlc_feature_version;
 		break;
+	case AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_CNTL:
+		fw_info->ver = adev->gfx.rlc_srlc_fw_version;
+		fw_info->feature = adev->gfx.rlc_srlc_feature_version;
+		break;
+	case AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_GPM_MEM:
+		fw_info->ver = adev->gfx.rlc_srlg_fw_version;
+		fw_info->feature = adev->gfx.rlc_srlg_feature_version;
+		break;
+	case AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_SRM_MEM:
+		fw_info->ver = adev->gfx.rlc_srls_fw_version;
+		fw_info->feature = adev->gfx.rlc_srls_feature_version;
+		break;
 	case AMDGPU_INFO_FW_GFX_MEC:
 		if (query_fw->index == 0) {
 			fw_info->ver = adev->gfx.mec_fw_version;
@@ -273,12 +286,15 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 	struct drm_crtc *crtc;
 	uint32_t ui32 = 0;
 	uint64_t ui64 = 0;
-	int i, found;
+	int i, j, found;
 	int ui32_size = sizeof(ui32);
 
 	if (!info->return_size || !info->return_pointer)
 		return -EINVAL;
 
+	/* Ensure IB tests are run on ring */
+	flush_delayed_work(&adev->late_init_work);
+
 	switch (info->query) {
 	case AMDGPU_INFO_ACCEL_WORKING:
 		ui32 = adev->accel_working;
@@ -313,56 +329,66 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 			type = AMD_IP_BLOCK_TYPE_GFX;
 			for (i = 0; i < adev->gfx.num_gfx_rings; i++)
 				ring_mask |= ((adev->gfx.gfx_ring[i].ready ? 1 : 0) << i);
-			ib_start_alignment = AMDGPU_GPU_PAGE_SIZE;
-			ib_size_alignment = 8;
+			ib_start_alignment = 32;
+			ib_size_alignment = 32;
 			break;
 		case AMDGPU_HW_IP_COMPUTE:
 			type = AMD_IP_BLOCK_TYPE_GFX;
 			for (i = 0; i < adev->gfx.num_compute_rings; i++)
 				ring_mask |= ((adev->gfx.compute_ring[i].ready ? 1 : 0) << i);
-			ib_start_alignment = AMDGPU_GPU_PAGE_SIZE;
-			ib_size_alignment = 8;
+			ib_start_alignment = 32;
+			ib_size_alignment = 32;
 			break;
 		case AMDGPU_HW_IP_DMA:
 			type = AMD_IP_BLOCK_TYPE_SDMA;
 			for (i = 0; i < adev->sdma.num_instances; i++)
 				ring_mask |= ((adev->sdma.instance[i].ring.ready ? 1 : 0) << i);
-			ib_start_alignment = AMDGPU_GPU_PAGE_SIZE;
-			ib_size_alignment = 1;
+			ib_start_alignment = 256;
+			ib_size_alignment = 4;
 			break;
 		case AMDGPU_HW_IP_UVD:
 			type = AMD_IP_BLOCK_TYPE_UVD;
-			ring_mask = adev->uvd.ring.ready ? 1 : 0;
-			ib_start_alignment = AMDGPU_GPU_PAGE_SIZE;
-			ib_size_alignment = 16;
+			for (i = 0; i < adev->uvd.num_uvd_inst; i++)
+				ring_mask |= ((adev->uvd.inst[i].ring.ready ? 1 : 0) << i);
+			ib_start_alignment = 64;
+			ib_size_alignment = 64;
 			break;
 		case AMDGPU_HW_IP_VCE:
 			type = AMD_IP_BLOCK_TYPE_VCE;
 			for (i = 0; i < adev->vce.num_rings; i++)
 				ring_mask |= ((adev->vce.ring[i].ready ? 1 : 0) << i);
-			ib_start_alignment = AMDGPU_GPU_PAGE_SIZE;
+			ib_start_alignment = 4;
 			ib_size_alignment = 1;
 			break;
 		case AMDGPU_HW_IP_UVD_ENC:
 			type = AMD_IP_BLOCK_TYPE_UVD;
-			for (i = 0; i < adev->uvd.num_enc_rings; i++)
-				ring_mask |= ((adev->uvd.ring_enc[i].ready ? 1 : 0) << i);
-			ib_start_alignment = AMDGPU_GPU_PAGE_SIZE;
-			ib_size_alignment = 1;
+			for (i = 0; i < adev->uvd.num_uvd_inst; i++)
+				for (j = 0; j < adev->uvd.num_enc_rings; j++)
+					ring_mask |=
+					((adev->uvd.inst[i].ring_enc[j].ready ? 1 : 0) <<
+					(j + i * adev->uvd.num_enc_rings));
+			ib_start_alignment = 64;
+			ib_size_alignment = 64;
 			break;
 		case AMDGPU_HW_IP_VCN_DEC:
 			type = AMD_IP_BLOCK_TYPE_VCN;
 			ring_mask = adev->vcn.ring_dec.ready ? 1 : 0;
-			ib_start_alignment = AMDGPU_GPU_PAGE_SIZE;
+			ib_start_alignment = 16;
 			ib_size_alignment = 16;
 			break;
 		case AMDGPU_HW_IP_VCN_ENC:
 			type = AMD_IP_BLOCK_TYPE_VCN;
 			for (i = 0; i < adev->vcn.num_enc_rings; i++)
 				ring_mask |= ((adev->vcn.ring_enc[i].ready ? 1 : 0) << i);
-			ib_start_alignment = AMDGPU_GPU_PAGE_SIZE;
+			ib_start_alignment = 64;
 			ib_size_alignment = 1;
 			break;
+		case AMDGPU_HW_IP_VCN_JPEG:
+			type = AMD_IP_BLOCK_TYPE_VCN;
+			ring_mask = adev->vcn.ring_jpeg.ready ? 1 : 0;
+			ib_start_alignment = 16;
+			ib_size_alignment = 16;
+			break;
 		default:
 			return -EINVAL;
 		}
@@ -407,6 +433,7 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 			break;
 		case AMDGPU_HW_IP_VCN_DEC:
 		case AMDGPU_HW_IP_VCN_ENC:
+		case AMDGPU_HW_IP_VCN_JPEG:
 			type = AMD_IP_BLOCK_TYPE_VCN;
 			break;
 		default:
@@ -474,13 +501,13 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 	case AMDGPU_INFO_VRAM_GTT: {
 		struct drm_amdgpu_info_vram_gtt vram_gtt;
 
-		vram_gtt.vram_size = adev->gmc.real_vram_size;
-		vram_gtt.vram_size -= adev->vram_pin_size;
-		vram_gtt.vram_cpu_accessible_size = adev->gmc.visible_vram_size;
-		vram_gtt.vram_cpu_accessible_size -= (adev->vram_pin_size - adev->invisible_pin_size);
+		vram_gtt.vram_size = adev->gmc.real_vram_size -
+			atomic64_read(&adev->vram_pin_size);
+		vram_gtt.vram_cpu_accessible_size = adev->gmc.visible_vram_size -
+			atomic64_read(&adev->visible_pin_size);
 		vram_gtt.gtt_size = adev->mman.bdev.man[TTM_PL_TT].size;
 		vram_gtt.gtt_size *= PAGE_SIZE;
-		vram_gtt.gtt_size -= adev->gart_pin_size;
+		vram_gtt.gtt_size -= atomic64_read(&adev->gart_pin_size);
 		return copy_to_user(out, &vram_gtt,
 				    min((size_t)size, sizeof(vram_gtt))) ? -EFAULT : 0;
 	}
@@ -489,17 +516,16 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 
 		memset(&mem, 0, sizeof(mem));
 		mem.vram.total_heap_size = adev->gmc.real_vram_size;
-		mem.vram.usable_heap_size =
-			adev->gmc.real_vram_size - adev->vram_pin_size;
+		mem.vram.usable_heap_size = adev->gmc.real_vram_size -
+			atomic64_read(&adev->vram_pin_size);
 		mem.vram.heap_usage =
 			amdgpu_vram_mgr_usage(&adev->mman.bdev.man[TTM_PL_VRAM]);
 		mem.vram.max_allocation = mem.vram.usable_heap_size * 3 / 4;
 
 		mem.cpu_accessible_vram.total_heap_size =
 			adev->gmc.visible_vram_size;
-		mem.cpu_accessible_vram.usable_heap_size =
-			adev->gmc.visible_vram_size -
-			(adev->vram_pin_size - adev->invisible_pin_size);
+		mem.cpu_accessible_vram.usable_heap_size = adev->gmc.visible_vram_size -
+			atomic64_read(&adev->visible_pin_size);
 		mem.cpu_accessible_vram.heap_usage =
 			amdgpu_vram_mgr_vis_usage(&adev->mman.bdev.man[TTM_PL_VRAM]);
 		mem.cpu_accessible_vram.max_allocation =
@@ -507,8 +533,8 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 
 		mem.gtt.total_heap_size = adev->mman.bdev.man[TTM_PL_TT].size;
 		mem.gtt.total_heap_size *= PAGE_SIZE;
-		mem.gtt.usable_heap_size = mem.gtt.total_heap_size
-			- adev->gart_pin_size;
+		mem.gtt.usable_heap_size = mem.gtt.total_heap_size -
+			atomic64_read(&adev->gart_pin_size);
 		mem.gtt.heap_usage =
 			amdgpu_gtt_mgr_usage(&adev->mman.bdev.man[TTM_PL_TT]);
 		mem.gtt.max_allocation = mem.gtt.usable_heap_size * 3 / 4;
@@ -701,10 +727,7 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 		}
 	}
 	case AMDGPU_INFO_SENSOR: {
-		struct pp_gpu_power query = {0};
-		int query_size = sizeof(query);
-
-		if (amdgpu_dpm == 0)
+		if (!adev->pm.dpm_enabled)
 			return -ENOENT;
 
 		switch (info->sensor_info.type) {
@@ -746,10 +769,10 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 			/* get average GPU power */
 			if (amdgpu_dpm_read_sensor(adev,
 						   AMDGPU_PP_SENSOR_GPU_POWER,
-						   (void *)&query, &query_size)) {
+						   (void *)&ui32, &ui32_size)) {
 				return -EINVAL;
 			}
-			ui32 = query.average_gpu_power >> 8;
+			ui32 >>= 8;
 			break;
 		case AMDGPU_INFO_SENSOR_VDDNB:
 			/* get VDDNB in millivolts */
@@ -914,8 +937,6 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev,
 
 	pm_runtime_get_sync(dev->dev);
 
-	amdgpu_ctx_mgr_fini(&fpriv->ctx_mgr);
-
 	if (adev->asic_type != CHIP_RAVEN) {
 		amdgpu_uvd_free_handles(adev, file_priv);
 		amdgpu_vce_free_handles(adev, file_priv);
@@ -935,6 +956,8 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev,
 	pd = amdgpu_bo_ref(fpriv->vm.root.base.bo);
 
 	amdgpu_vm_fini(adev, &fpriv->vm);
+	amdgpu_ctx_mgr_fini(&fpriv->ctx_mgr);
+
 	if (pasid)
 		amdgpu_pasid_free_delayed(pd->tbo.resv, pasid);
 	amdgpu_bo_unref(&pd);
@@ -1088,6 +1111,7 @@ static int amdgpu_debugfs_firmware_info(struct seq_file *m, void *data)
 	struct amdgpu_device *adev = dev->dev_private;
 	struct drm_amdgpu_info_firmware fw_info;
 	struct drm_amdgpu_query_fw query_fw;
+	struct atom_context *ctx = adev->mode_info.atom_context;
 	int ret, i;
 
 	/* VCE */
@@ -1146,6 +1170,30 @@ static int amdgpu_debugfs_firmware_info(struct seq_file *m, void *data)
 	seq_printf(m, "RLC feature version: %u, firmware version: 0x%08x\n",
 		   fw_info.feature, fw_info.ver);
 
+	/* RLC SAVE RESTORE LIST CNTL */
+	query_fw.fw_type = AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_CNTL;
+	ret = amdgpu_firmware_info(&fw_info, &query_fw, adev);
+	if (ret)
+		return ret;
+	seq_printf(m, "RLC SRLC feature version: %u, firmware version: 0x%08x\n",
+		   fw_info.feature, fw_info.ver);
+
+	/* RLC SAVE RESTORE LIST GPM MEM */
+	query_fw.fw_type = AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_GPM_MEM;
+	ret = amdgpu_firmware_info(&fw_info, &query_fw, adev);
+	if (ret)
+		return ret;
+	seq_printf(m, "RLC SRLG feature version: %u, firmware version: 0x%08x\n",
+		   fw_info.feature, fw_info.ver);
+
+	/* RLC SAVE RESTORE LIST SRM MEM */
+	query_fw.fw_type = AMDGPU_INFO_FW_GFX_RLC_RESTORE_LIST_SRM_MEM;
+	ret = amdgpu_firmware_info(&fw_info, &query_fw, adev);
+	if (ret)
+		return ret;
+	seq_printf(m, "RLC SRLS feature version: %u, firmware version: 0x%08x\n",
+		   fw_info.feature, fw_info.ver);
+
 	/* MEC */
 	query_fw.fw_type = AMDGPU_INFO_FW_GFX_MEC;
 	query_fw.index = 0;
@@ -1210,6 +1258,9 @@ static int amdgpu_debugfs_firmware_info(struct seq_file *m, void *data)
 	seq_printf(m, "VCN feature version: %u, firmware version: 0x%08x\n",
 		   fw_info.feature, fw_info.ver);
 
+
+	seq_printf(m, "VBIOS version: %s\n", ctx->vbios_version);
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
index bd67f4cb8e6c..a365ea2383d1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
@@ -28,6 +28,21 @@
  *    Christian König <christian.koenig@amd.com>
  */
 
+/**
+ * DOC: MMU Notifier
+ *
+ * For coherent userptr handling registers an MMU notifier to inform the driver
+ * about updates on the page tables of a process.
+ *
+ * When somebody tries to invalidate the page tables we block the update until
+ * all operations on the pages in question are completed, then those pages are
+ * marked as accessed and also dirty if it wasn't a read only access.
+ *
+ * New command submissions using the userptrs in question are delayed until all
+ * page table invalidation are completed and we once more see a coherent process
+ * address space.
+ */
+
 #include <linux/firmware.h>
 #include <linux/module.h>
 #include <linux/mmu_notifier.h>
@@ -36,12 +51,30 @@
 #include <drm/drm.h>
 
 #include "amdgpu.h"
+#include "amdgpu_amdkfd.h"
 
+/**
+ * struct amdgpu_mn
+ *
+ * @adev: amdgpu device pointer
+ * @mm: process address space
+ * @mn: MMU notifier structure
+ * @type: type of MMU notifier
+ * @work: destruction work item
+ * @node: hash table node to find structure by adev and mn
+ * @lock: rw semaphore protecting the notifier nodes
+ * @objects: interval tree containing amdgpu_mn_nodes
+ * @read_lock: mutex for recursive locking of @lock
+ * @recursion: depth of recursion
+ *
+ * Data for each amdgpu device and process address space.
+ */
 struct amdgpu_mn {
 	/* constant after initialisation */
 	struct amdgpu_device	*adev;
 	struct mm_struct	*mm;
 	struct mmu_notifier	mn;
+	enum amdgpu_mn_type	type;
 
 	/* only used on destruction */
 	struct work_struct	work;
@@ -56,13 +89,21 @@ struct amdgpu_mn {
 	atomic_t		recursion;
 };
 
+/**
+ * struct amdgpu_mn_node
+ *
+ * @it: interval node defining start-last of the affected address range
+ * @bos: list of all BOs in the affected address range
+ *
+ * Manages all BOs which are affected of a certain range of address space.
+ */
 struct amdgpu_mn_node {
 	struct interval_tree_node	it;
 	struct list_head		bos;
 };
 
 /**
- * amdgpu_mn_destroy - destroy the rmn
+ * amdgpu_mn_destroy - destroy the MMU notifier
  *
  * @work: previously sheduled work item
  *
@@ -70,47 +111,50 @@ struct amdgpu_mn_node {
  */
 static void amdgpu_mn_destroy(struct work_struct *work)
 {
-	struct amdgpu_mn *rmn = container_of(work, struct amdgpu_mn, work);
-	struct amdgpu_device *adev = rmn->adev;
+	struct amdgpu_mn *amn = container_of(work, struct amdgpu_mn, work);
+	struct amdgpu_device *adev = amn->adev;
 	struct amdgpu_mn_node *node, *next_node;
 	struct amdgpu_bo *bo, *next_bo;
 
 	mutex_lock(&adev->mn_lock);
-	down_write(&rmn->lock);
-	hash_del(&rmn->node);
+	down_write(&amn->lock);
+	hash_del(&amn->node);
 	rbtree_postorder_for_each_entry_safe(node, next_node,
-					     &rmn->objects.rb_root, it.rb) {
+					     &amn->objects.rb_root, it.rb) {
 		list_for_each_entry_safe(bo, next_bo, &node->bos, mn_list) {
 			bo->mn = NULL;
 			list_del_init(&bo->mn_list);
 		}
 		kfree(node);
 	}
-	up_write(&rmn->lock);
+	up_write(&amn->lock);
 	mutex_unlock(&adev->mn_lock);
-	mmu_notifier_unregister_no_release(&rmn->mn, rmn->mm);
-	kfree(rmn);
+	mmu_notifier_unregister_no_release(&amn->mn, amn->mm);
+	kfree(amn);
 }
 
 /**
  * amdgpu_mn_release - callback to notify about mm destruction
  *
  * @mn: our notifier
- * @mn: the mm this callback is about
+ * @mm: the mm this callback is about
  *
  * Shedule a work item to lazy destroy our notifier.
  */
 static void amdgpu_mn_release(struct mmu_notifier *mn,
 			      struct mm_struct *mm)
 {
-	struct amdgpu_mn *rmn = container_of(mn, struct amdgpu_mn, mn);
-	INIT_WORK(&rmn->work, amdgpu_mn_destroy);
-	schedule_work(&rmn->work);
+	struct amdgpu_mn *amn = container_of(mn, struct amdgpu_mn, mn);
+
+	INIT_WORK(&amn->work, amdgpu_mn_destroy);
+	schedule_work(&amn->work);
 }
 
 
 /**
- * amdgpu_mn_lock - take the write side lock for this mn
+ * amdgpu_mn_lock - take the write side lock for this notifier
+ *
+ * @mn: our notifier
  */
 void amdgpu_mn_lock(struct amdgpu_mn *mn)
 {
@@ -119,7 +163,9 @@ void amdgpu_mn_lock(struct amdgpu_mn *mn)
 }
 
 /**
- * amdgpu_mn_unlock - drop the write side lock for this mn
+ * amdgpu_mn_unlock - drop the write side lock for this notifier
+ *
+ * @mn: our notifier
  */
 void amdgpu_mn_unlock(struct amdgpu_mn *mn)
 {
@@ -128,40 +174,38 @@ void amdgpu_mn_unlock(struct amdgpu_mn *mn)
 }
 
 /**
- * amdgpu_mn_read_lock - take the rmn read lock
- *
- * @rmn: our notifier
+ * amdgpu_mn_read_lock - take the read side lock for this notifier
  *
- * Take the rmn read side lock.
+ * @amn: our notifier
  */
-static void amdgpu_mn_read_lock(struct amdgpu_mn *rmn)
+static void amdgpu_mn_read_lock(struct amdgpu_mn *amn)
 {
-	mutex_lock(&rmn->read_lock);
-	if (atomic_inc_return(&rmn->recursion) == 1)
-		down_read_non_owner(&rmn->lock);
-	mutex_unlock(&rmn->read_lock);
+	mutex_lock(&amn->read_lock);
+	if (atomic_inc_return(&amn->recursion) == 1)
+		down_read_non_owner(&amn->lock);
+	mutex_unlock(&amn->read_lock);
 }
 
 /**
- * amdgpu_mn_read_unlock - drop the rmn read lock
+ * amdgpu_mn_read_unlock - drop the read side lock for this notifier
  *
- * @rmn: our notifier
- *
- * Drop the rmn read side lock.
+ * @amn: our notifier
  */
-static void amdgpu_mn_read_unlock(struct amdgpu_mn *rmn)
+static void amdgpu_mn_read_unlock(struct amdgpu_mn *amn)
 {
-	if (atomic_dec_return(&rmn->recursion) == 0)
-		up_read_non_owner(&rmn->lock);
+	if (atomic_dec_return(&amn->recursion) == 0)
+		up_read_non_owner(&amn->lock);
 }
 
 /**
  * amdgpu_mn_invalidate_node - unmap all BOs of a node
  *
  * @node: the node with the BOs to unmap
+ * @start: start of address range affected
+ * @end: end of address range affected
  *
- * We block for all BOs and unmap them by move them
- * into system domain again.
+ * Block for operations on BOs to finish and mark pages as accessed and
+ * potentially dirty.
  */
 static void amdgpu_mn_invalidate_node(struct amdgpu_mn_node *node,
 				      unsigned long start,
@@ -185,30 +229,30 @@ static void amdgpu_mn_invalidate_node(struct amdgpu_mn_node *node,
 }
 
 /**
- * amdgpu_mn_invalidate_range_start - callback to notify about mm change
+ * amdgpu_mn_invalidate_range_start_gfx - callback to notify about mm change
  *
  * @mn: our notifier
- * @mn: the mm this callback is about
+ * @mm: the mm this callback is about
  * @start: start of updated range
  * @end: end of updated range
  *
- * We block for all BOs between start and end to be idle and
- * unmap them by move them into system domain again.
+ * Block for operations on BOs to finish and mark pages as accessed and
+ * potentially dirty.
  */
-static void amdgpu_mn_invalidate_range_start(struct mmu_notifier *mn,
-					     struct mm_struct *mm,
-					     unsigned long start,
-					     unsigned long end)
+static void amdgpu_mn_invalidate_range_start_gfx(struct mmu_notifier *mn,
+						 struct mm_struct *mm,
+						 unsigned long start,
+						 unsigned long end)
 {
-	struct amdgpu_mn *rmn = container_of(mn, struct amdgpu_mn, mn);
+	struct amdgpu_mn *amn = container_of(mn, struct amdgpu_mn, mn);
 	struct interval_tree_node *it;
 
 	/* notification is exclusive, but interval is inclusive */
 	end -= 1;
 
-	amdgpu_mn_read_lock(rmn);
+	amdgpu_mn_read_lock(amn);
 
-	it = interval_tree_iter_first(&rmn->objects, start, end);
+	it = interval_tree_iter_first(&amn->objects, start, end);
 	while (it) {
 		struct amdgpu_mn_node *node;
 
@@ -220,10 +264,53 @@ static void amdgpu_mn_invalidate_range_start(struct mmu_notifier *mn,
 }
 
 /**
+ * amdgpu_mn_invalidate_range_start_hsa - callback to notify about mm change
+ *
+ * @mn: our notifier
+ * @mm: the mm this callback is about
+ * @start: start of updated range
+ * @end: end of updated range
+ *
+ * We temporarily evict all BOs between start and end. This
+ * necessitates evicting all user-mode queues of the process. The BOs
+ * are restorted in amdgpu_mn_invalidate_range_end_hsa.
+ */
+static void amdgpu_mn_invalidate_range_start_hsa(struct mmu_notifier *mn,
+						 struct mm_struct *mm,
+						 unsigned long start,
+						 unsigned long end)
+{
+	struct amdgpu_mn *amn = container_of(mn, struct amdgpu_mn, mn);
+	struct interval_tree_node *it;
+
+	/* notification is exclusive, but interval is inclusive */
+	end -= 1;
+
+	amdgpu_mn_read_lock(amn);
+
+	it = interval_tree_iter_first(&amn->objects, start, end);
+	while (it) {
+		struct amdgpu_mn_node *node;
+		struct amdgpu_bo *bo;
+
+		node = container_of(it, struct amdgpu_mn_node, it);
+		it = interval_tree_iter_next(it, start, end);
+
+		list_for_each_entry(bo, &node->bos, mn_list) {
+			struct kgd_mem *mem = bo->kfd_bo;
+
+			if (amdgpu_ttm_tt_affect_userptr(bo->tbo.ttm,
+							 start, end))
+				amdgpu_amdkfd_evict_userptr(mem, mm);
+		}
+	}
+}
+
+/**
  * amdgpu_mn_invalidate_range_end - callback to notify about mm change
  *
  * @mn: our notifier
- * @mn: the mm this callback is about
+ * @mm: the mm this callback is about
  * @start: start of updated range
  * @end: end of updated range
  *
@@ -234,28 +321,44 @@ static void amdgpu_mn_invalidate_range_end(struct mmu_notifier *mn,
 					   unsigned long start,
 					   unsigned long end)
 {
-	struct amdgpu_mn *rmn = container_of(mn, struct amdgpu_mn, mn);
+	struct amdgpu_mn *amn = container_of(mn, struct amdgpu_mn, mn);
 
-	amdgpu_mn_read_unlock(rmn);
+	amdgpu_mn_read_unlock(amn);
 }
 
-static const struct mmu_notifier_ops amdgpu_mn_ops = {
-	.release = amdgpu_mn_release,
-	.invalidate_range_start = amdgpu_mn_invalidate_range_start,
-	.invalidate_range_end = amdgpu_mn_invalidate_range_end,
+static const struct mmu_notifier_ops amdgpu_mn_ops[] = {
+	[AMDGPU_MN_TYPE_GFX] = {
+		.release = amdgpu_mn_release,
+		.invalidate_range_start = amdgpu_mn_invalidate_range_start_gfx,
+		.invalidate_range_end = amdgpu_mn_invalidate_range_end,
+	},
+	[AMDGPU_MN_TYPE_HSA] = {
+		.release = amdgpu_mn_release,
+		.invalidate_range_start = amdgpu_mn_invalidate_range_start_hsa,
+		.invalidate_range_end = amdgpu_mn_invalidate_range_end,
+	},
 };
 
+/* Low bits of any reasonable mm pointer will be unused due to struct
+ * alignment. Use these bits to make a unique key from the mm pointer
+ * and notifier type.
+ */
+#define AMDGPU_MN_KEY(mm, type) ((unsigned long)(mm) + (type))
+
 /**
  * amdgpu_mn_get - create notifier context
  *
  * @adev: amdgpu device pointer
+ * @type: type of MMU notifier context
  *
  * Creates a notifier context for current->mm.
  */
-struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev)
+struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev,
+				enum amdgpu_mn_type type)
 {
 	struct mm_struct *mm = current->mm;
-	struct amdgpu_mn *rmn;
+	struct amdgpu_mn *amn;
+	unsigned long key = AMDGPU_MN_KEY(mm, type);
 	int r;
 
 	mutex_lock(&adev->mn_lock);
@@ -264,40 +367,41 @@ struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev)
 		return ERR_PTR(-EINTR);
 	}
 
-	hash_for_each_possible(adev->mn_hash, rmn, node, (unsigned long)mm)
-		if (rmn->mm == mm)
+	hash_for_each_possible(adev->mn_hash, amn, node, key)
+		if (AMDGPU_MN_KEY(amn->mm, amn->type) == key)
 			goto release_locks;
 
-	rmn = kzalloc(sizeof(*rmn), GFP_KERNEL);
-	if (!rmn) {
-		rmn = ERR_PTR(-ENOMEM);
+	amn = kzalloc(sizeof(*amn), GFP_KERNEL);
+	if (!amn) {
+		amn = ERR_PTR(-ENOMEM);
 		goto release_locks;
 	}
 
-	rmn->adev = adev;
-	rmn->mm = mm;
-	rmn->mn.ops = &amdgpu_mn_ops;
-	init_rwsem(&rmn->lock);
-	rmn->objects = RB_ROOT_CACHED;
-	mutex_init(&rmn->read_lock);
-	atomic_set(&rmn->recursion, 0);
+	amn->adev = adev;
+	amn->mm = mm;
+	init_rwsem(&amn->lock);
+	amn->type = type;
+	amn->mn.ops = &amdgpu_mn_ops[type];
+	amn->objects = RB_ROOT_CACHED;
+	mutex_init(&amn->read_lock);
+	atomic_set(&amn->recursion, 0);
 
-	r = __mmu_notifier_register(&rmn->mn, mm);
+	r = __mmu_notifier_register(&amn->mn, mm);
 	if (r)
-		goto free_rmn;
+		goto free_amn;
 
-	hash_add(adev->mn_hash, &rmn->node, (unsigned long)mm);
+	hash_add(adev->mn_hash, &amn->node, AMDGPU_MN_KEY(mm, type));
 
 release_locks:
 	up_write(&mm->mmap_sem);
 	mutex_unlock(&adev->mn_lock);
 
-	return rmn;
+	return amn;
 
-free_rmn:
+free_amn:
 	up_write(&mm->mmap_sem);
 	mutex_unlock(&adev->mn_lock);
-	kfree(rmn);
+	kfree(amn);
 
 	return ERR_PTR(r);
 }
@@ -315,37 +419,40 @@ int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr)
 {
 	unsigned long end = addr + amdgpu_bo_size(bo) - 1;
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
-	struct amdgpu_mn *rmn;
-	struct amdgpu_mn_node *node = NULL;
+	enum amdgpu_mn_type type =
+		bo->kfd_bo ? AMDGPU_MN_TYPE_HSA : AMDGPU_MN_TYPE_GFX;
+	struct amdgpu_mn *amn;
+	struct amdgpu_mn_node *node = NULL, *new_node;
 	struct list_head bos;
 	struct interval_tree_node *it;
 
-	rmn = amdgpu_mn_get(adev);
-	if (IS_ERR(rmn))
-		return PTR_ERR(rmn);
+	amn = amdgpu_mn_get(adev, type);
+	if (IS_ERR(amn))
+		return PTR_ERR(amn);
+
+	new_node = kmalloc(sizeof(*new_node), GFP_KERNEL);
+	if (!new_node)
+		return -ENOMEM;
 
 	INIT_LIST_HEAD(&bos);
 
-	down_write(&rmn->lock);
+	down_write(&amn->lock);
 
-	while ((it = interval_tree_iter_first(&rmn->objects, addr, end))) {
+	while ((it = interval_tree_iter_first(&amn->objects, addr, end))) {
 		kfree(node);
 		node = container_of(it, struct amdgpu_mn_node, it);
-		interval_tree_remove(&node->it, &rmn->objects);
+		interval_tree_remove(&node->it, &amn->objects);
 		addr = min(it->start, addr);
 		end = max(it->last, end);
 		list_splice(&node->bos, &bos);
 	}
 
-	if (!node) {
-		node = kmalloc(sizeof(struct amdgpu_mn_node), GFP_KERNEL);
-		if (!node) {
-			up_write(&rmn->lock);
-			return -ENOMEM;
-		}
-	}
+	if (!node)
+		node = new_node;
+	else
+		kfree(new_node);
 
-	bo->mn = rmn;
+	bo->mn = amn;
 
 	node->it.start = addr;
 	node->it.last = end;
@@ -353,9 +460,9 @@ int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr)
 	list_splice(&bos, &node->bos);
 	list_add(&bo->mn_list, &node->bos);
 
-	interval_tree_insert(&node->it, &rmn->objects);
+	interval_tree_insert(&node->it, &amn->objects);
 
-	up_write(&rmn->lock);
+	up_write(&amn->lock);
 
 	return 0;
 }
@@ -370,18 +477,18 @@ int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr)
 void amdgpu_mn_unregister(struct amdgpu_bo *bo)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
-	struct amdgpu_mn *rmn;
+	struct amdgpu_mn *amn;
 	struct list_head *head;
 
 	mutex_lock(&adev->mn_lock);
 
-	rmn = bo->mn;
-	if (rmn == NULL) {
+	amn = bo->mn;
+	if (amn == NULL) {
 		mutex_unlock(&adev->mn_lock);
 		return;
 	}
 
-	down_write(&rmn->lock);
+	down_write(&amn->lock);
 
 	/* save the next list entry for later */
 	head = bo->mn_list.next;
@@ -391,12 +498,13 @@ void amdgpu_mn_unregister(struct amdgpu_bo *bo)
 
 	if (list_empty(head)) {
 		struct amdgpu_mn_node *node;
+
 		node = container_of(head, struct amdgpu_mn_node, bos);
-		interval_tree_remove(&node->it, &rmn->objects);
+		interval_tree_remove(&node->it, &amn->objects);
 		kfree(node);
 	}
 
-	up_write(&rmn->lock);
+	up_write(&amn->lock);
 	mutex_unlock(&adev->mn_lock);
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h
index d0095a3793b8..eb0f432f78fe 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h
@@ -29,16 +29,23 @@
  */
 struct amdgpu_mn;
 
+enum amdgpu_mn_type {
+	AMDGPU_MN_TYPE_GFX,
+	AMDGPU_MN_TYPE_HSA,
+};
+
 #if defined(CONFIG_MMU_NOTIFIER)
 void amdgpu_mn_lock(struct amdgpu_mn *mn);
 void amdgpu_mn_unlock(struct amdgpu_mn *mn);
-struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev);
+struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev,
+				enum amdgpu_mn_type type);
 int amdgpu_mn_register(struct amdgpu_bo *bo, unsigned long addr);
 void amdgpu_mn_unregister(struct amdgpu_bo *bo);
 #else
 static inline void amdgpu_mn_lock(struct amdgpu_mn *mn) {}
 static inline void amdgpu_mn_unlock(struct amdgpu_mn *mn) {}
-static inline struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev)
+static inline struct amdgpu_mn *amdgpu_mn_get(struct amdgpu_device *adev,
+					      enum amdgpu_mn_type type)
 {
 	return NULL;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h
index d6416ee52e32..b9e9e8b02fb7 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mode.h
@@ -308,7 +308,6 @@ struct amdgpu_display_funcs {
 
 struct amdgpu_framebuffer {
 	struct drm_framebuffer base;
-	struct drm_gem_object *obj;
 
 	/* caching for later use */
 	uint64_t address;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
index 6d08cde8443c..b12526ce1a9d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
@@ -38,6 +38,19 @@
 #include "amdgpu_trace.h"
 #include "amdgpu_amdkfd.h"
 
+/**
+ * DOC: amdgpu_object
+ *
+ * This defines the interfaces to operate on an &amdgpu_bo buffer object which
+ * represents memory used by driver (VRAM, system memory, etc.). The driver
+ * provides DRM/GEM APIs to userspace. DRM/GEM APIs then use these interfaces
+ * to create/destroy/set buffer object which are then managed by the kernel TTM
+ * memory manager.
+ * The interfaces are also used internally by kernel clients, including gfx,
+ * uvd, etc. for kernel managed allocations used by the GPU.
+ *
+ */
+
 static bool amdgpu_need_backup(struct amdgpu_device *adev)
 {
 	if (adev->flags & AMD_IS_APU)
@@ -50,11 +63,35 @@ static bool amdgpu_need_backup(struct amdgpu_device *adev)
 	return true;
 }
 
+/**
+ * amdgpu_bo_subtract_pin_size - Remove BO from pin_size accounting
+ *
+ * @bo: &amdgpu_bo buffer object
+ *
+ * This function is called when a BO stops being pinned, and updates the
+ * &amdgpu_device pin_size values accordingly.
+ */
+static void amdgpu_bo_subtract_pin_size(struct amdgpu_bo *bo)
+{
+	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
+
+	if (bo->tbo.mem.mem_type == TTM_PL_VRAM) {
+		atomic64_sub(amdgpu_bo_size(bo), &adev->vram_pin_size);
+		atomic64_sub(amdgpu_vram_mgr_bo_visible_size(bo),
+			     &adev->visible_pin_size);
+	} else if (bo->tbo.mem.mem_type == TTM_PL_TT) {
+		atomic64_sub(amdgpu_bo_size(bo), &adev->gart_pin_size);
+	}
+}
+
 static void amdgpu_ttm_bo_destroy(struct ttm_buffer_object *tbo)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(tbo->bdev);
 	struct amdgpu_bo *bo = ttm_to_amdgpu_bo(tbo);
 
+	if (WARN_ON_ONCE(bo->pin_count > 0))
+		amdgpu_bo_subtract_pin_size(bo);
+
 	if (bo->kfd_bo)
 		amdgpu_amdkfd_unreserve_system_memory_limit(bo);
 
@@ -73,6 +110,16 @@ static void amdgpu_ttm_bo_destroy(struct ttm_buffer_object *tbo)
 	kfree(bo);
 }
 
+/**
+ * amdgpu_ttm_bo_is_amdgpu_bo - check if the buffer object is an &amdgpu_bo
+ * @bo: buffer object to be checked
+ *
+ * Uses destroy function associated with the object to determine if this is
+ * an &amdgpu_bo.
+ *
+ * Returns:
+ * true if the object belongs to &amdgpu_bo, false if not.
+ */
 bool amdgpu_ttm_bo_is_amdgpu_bo(struct ttm_buffer_object *bo)
 {
 	if (bo->destroy == &amdgpu_ttm_bo_destroy)
@@ -80,6 +127,14 @@ bool amdgpu_ttm_bo_is_amdgpu_bo(struct ttm_buffer_object *bo)
 	return false;
 }
 
+/**
+ * amdgpu_ttm_placement_from_domain - set buffer's placement
+ * @abo: &amdgpu_bo buffer object whose placement is to be set
+ * @domain: requested domain
+ *
+ * Sets buffer's placement according to requested domain and the buffer's
+ * flags.
+ */
 void amdgpu_ttm_placement_from_domain(struct amdgpu_bo *abo, u32 domain)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(abo->tbo.bdev);
@@ -184,21 +239,29 @@ void amdgpu_ttm_placement_from_domain(struct amdgpu_bo *abo, u32 domain)
  *
  * Note: For bo_ptr new BO is only created if bo_ptr points to NULL.
  *
- * Returns 0 on success, negative error code otherwise.
+ * Returns:
+ * 0 on success, negative error code otherwise.
  */
 int amdgpu_bo_create_reserved(struct amdgpu_device *adev,
 			      unsigned long size, int align,
 			      u32 domain, struct amdgpu_bo **bo_ptr,
 			      u64 *gpu_addr, void **cpu_addr)
 {
+	struct amdgpu_bo_param bp;
 	bool free = false;
 	int r;
 
+	memset(&bp, 0, sizeof(bp));
+	bp.size = size;
+	bp.byte_align = align;
+	bp.domain = domain;
+	bp.flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
+		AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS;
+	bp.type = ttm_bo_type_kernel;
+	bp.resv = NULL;
+
 	if (!*bo_ptr) {
-		r = amdgpu_bo_create(adev, size, align, domain,
-				     AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
-				     AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS,
-				     ttm_bo_type_kernel, NULL, bo_ptr);
+		r = amdgpu_bo_create(adev, &bp, bo_ptr);
 		if (r) {
 			dev_err(adev->dev, "(%d) failed to allocate kernel bo\n",
 				r);
@@ -213,22 +276,33 @@ int amdgpu_bo_create_reserved(struct amdgpu_device *adev,
 		goto error_free;
 	}
 
-	r = amdgpu_bo_pin(*bo_ptr, domain, gpu_addr);
+	r = amdgpu_bo_pin(*bo_ptr, domain);
 	if (r) {
 		dev_err(adev->dev, "(%d) kernel bo pin failed\n", r);
 		goto error_unreserve;
 	}
 
+	r = amdgpu_ttm_alloc_gart(&(*bo_ptr)->tbo);
+	if (r) {
+		dev_err(adev->dev, "%p bind failed\n", *bo_ptr);
+		goto error_unpin;
+	}
+
+	if (gpu_addr)
+		*gpu_addr = amdgpu_bo_gpu_offset(*bo_ptr);
+
 	if (cpu_addr) {
 		r = amdgpu_bo_kmap(*bo_ptr, cpu_addr);
 		if (r) {
 			dev_err(adev->dev, "(%d) kernel bo map failed\n", r);
-			goto error_unreserve;
+			goto error_unpin;
 		}
 	}
 
 	return 0;
 
+error_unpin:
+	amdgpu_bo_unpin(*bo_ptr);
 error_unreserve:
 	amdgpu_bo_unreserve(*bo_ptr);
 
@@ -254,7 +328,8 @@ error_free:
  *
  * Note: For bo_ptr new BO is only created if bo_ptr points to NULL.
  *
- * Returns 0 on success, negative error code otherwise.
+ * Returns:
+ * 0 on success, negative error code otherwise.
  */
 int amdgpu_bo_create_kernel(struct amdgpu_device *adev,
 			    unsigned long size, int align,
@@ -278,6 +353,8 @@ int amdgpu_bo_create_kernel(struct amdgpu_device *adev,
  * amdgpu_bo_free_kernel - free BO for kernel use
  *
  * @bo: amdgpu BO to free
+ * @gpu_addr: pointer to where the BO's GPU memory space address was stored
+ * @cpu_addr: pointer to where the BO's CPU memory space address was stored
  *
  * unmaps and unpin a BO for kernel internal use.
  */
@@ -341,27 +418,25 @@ fail:
 	return false;
 }
 
-static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
-			       int byte_align, u32 domain,
-			       u64 flags, enum ttm_bo_type type,
-			       struct reservation_object *resv,
+static int amdgpu_bo_do_create(struct amdgpu_device *adev,
+			       struct amdgpu_bo_param *bp,
 			       struct amdgpu_bo **bo_ptr)
 {
 	struct ttm_operation_ctx ctx = {
-		.interruptible = (type != ttm_bo_type_kernel),
+		.interruptible = (bp->type != ttm_bo_type_kernel),
 		.no_wait_gpu = false,
-		.resv = resv,
+		.resv = bp->resv,
 		.flags = TTM_OPT_FLAG_ALLOW_RES_EVICT
 	};
 	struct amdgpu_bo *bo;
-	unsigned long page_align;
+	unsigned long page_align, size = bp->size;
 	size_t acc_size;
 	int r;
 
-	page_align = roundup(byte_align, PAGE_SIZE) >> PAGE_SHIFT;
+	page_align = roundup(bp->byte_align, PAGE_SIZE) >> PAGE_SHIFT;
 	size = ALIGN(size, PAGE_SIZE);
 
-	if (!amdgpu_bo_validate_size(adev, size, domain))
+	if (!amdgpu_bo_validate_size(adev, size, bp->domain))
 		return -ENOMEM;
 
 	*bo_ptr = NULL;
@@ -375,18 +450,14 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
 	drm_gem_private_object_init(adev->ddev, &bo->gem_base, size);
 	INIT_LIST_HEAD(&bo->shadow_list);
 	INIT_LIST_HEAD(&bo->va);
-	bo->preferred_domains = domain & (AMDGPU_GEM_DOMAIN_VRAM |
-					 AMDGPU_GEM_DOMAIN_GTT |
-					 AMDGPU_GEM_DOMAIN_CPU |
-					 AMDGPU_GEM_DOMAIN_GDS |
-					 AMDGPU_GEM_DOMAIN_GWS |
-					 AMDGPU_GEM_DOMAIN_OA);
+	bo->preferred_domains = bp->preferred_domain ? bp->preferred_domain :
+		bp->domain;
 	bo->allowed_domains = bo->preferred_domains;
-	if (type != ttm_bo_type_kernel &&
+	if (bp->type != ttm_bo_type_kernel &&
 	    bo->allowed_domains == AMDGPU_GEM_DOMAIN_VRAM)
 		bo->allowed_domains |= AMDGPU_GEM_DOMAIN_GTT;
 
-	bo->flags = flags;
+	bo->flags = bp->flags;
 
 #ifdef CONFIG_X86_32
 	/* XXX: Write-combined CPU mappings of GTT seem broken on 32-bit
@@ -417,15 +488,17 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
 #endif
 
 	bo->tbo.bdev = &adev->mman.bdev;
-	amdgpu_ttm_placement_from_domain(bo, domain);
+	amdgpu_ttm_placement_from_domain(bo, bp->domain);
+	if (bp->type == ttm_bo_type_kernel)
+		bo->tbo.priority = 1;
 
-	r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, type,
+	r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, bp->type,
 				 &bo->placement, page_align, &ctx, acc_size,
-				 NULL, resv, &amdgpu_ttm_bo_destroy);
+				 NULL, bp->resv, &amdgpu_ttm_bo_destroy);
 	if (unlikely(r != 0))
 		return r;
 
-	if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size &&
+	if (!amdgpu_gmc_vram_full_visible(&adev->gmc) &&
 	    bo->tbo.mem.mem_type == TTM_PL_VRAM &&
 	    bo->tbo.mem.start < adev->gmc.visible_vram_size >> PAGE_SHIFT)
 		amdgpu_cs_report_moved_bytes(adev, ctx.bytes_moved,
@@ -433,10 +506,7 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
 	else
 		amdgpu_cs_report_moved_bytes(adev, ctx.bytes_moved, 0);
 
-	if (type == ttm_bo_type_kernel)
-		bo->tbo.priority = 1;
-
-	if (flags & AMDGPU_GEM_CREATE_VRAM_CLEARED &&
+	if (bp->flags & AMDGPU_GEM_CREATE_VRAM_CLEARED &&
 	    bo->tbo.mem.placement & TTM_PL_FLAG_VRAM) {
 		struct dma_fence *fence;
 
@@ -449,20 +519,20 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev, unsigned long size,
 		bo->tbo.moving = dma_fence_get(fence);
 		dma_fence_put(fence);
 	}
-	if (!resv)
+	if (!bp->resv)
 		amdgpu_bo_unreserve(bo);
 	*bo_ptr = bo;
 
 	trace_amdgpu_bo_create(bo);
 
 	/* Treat CPU_ACCESS_REQUIRED only as a hint if given by UMD */
-	if (type == ttm_bo_type_device)
+	if (bp->type == ttm_bo_type_device)
 		bo->flags &= ~AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;
 
 	return 0;
 
 fail_unreserve:
-	if (!resv)
+	if (!bp->resv)
 		ww_mutex_unlock(&bo->tbo.resv->lock);
 	amdgpu_bo_unref(&bo);
 	return r;
@@ -472,16 +542,22 @@ static int amdgpu_bo_create_shadow(struct amdgpu_device *adev,
 				   unsigned long size, int byte_align,
 				   struct amdgpu_bo *bo)
 {
+	struct amdgpu_bo_param bp;
 	int r;
 
 	if (bo->shadow)
 		return 0;
 
-	r = amdgpu_bo_do_create(adev, size, byte_align, AMDGPU_GEM_DOMAIN_GTT,
-				AMDGPU_GEM_CREATE_CPU_GTT_USWC |
-				AMDGPU_GEM_CREATE_SHADOW,
-				ttm_bo_type_kernel,
-				bo->tbo.resv, &bo->shadow);
+	memset(&bp, 0, sizeof(bp));
+	bp.size = size;
+	bp.byte_align = byte_align;
+	bp.domain = AMDGPU_GEM_DOMAIN_GTT;
+	bp.flags = AMDGPU_GEM_CREATE_CPU_GTT_USWC |
+		AMDGPU_GEM_CREATE_SHADOW;
+	bp.type = ttm_bo_type_kernel;
+	bp.resv = bo->tbo.resv;
+
+	r = amdgpu_bo_do_create(adev, &bp, &bo->shadow);
 	if (!r) {
 		bo->shadow->parent = amdgpu_bo_ref(bo);
 		mutex_lock(&adev->shadow_list_lock);
@@ -492,28 +568,40 @@ static int amdgpu_bo_create_shadow(struct amdgpu_device *adev,
 	return r;
 }
 
-int amdgpu_bo_create(struct amdgpu_device *adev, unsigned long size,
-		     int byte_align, u32 domain,
-		     u64 flags, enum ttm_bo_type type,
-		     struct reservation_object *resv,
+/**
+ * amdgpu_bo_create - create an &amdgpu_bo buffer object
+ * @adev: amdgpu device object
+ * @bp: parameters to be used for the buffer object
+ * @bo_ptr: pointer to the buffer object pointer
+ *
+ * Creates an &amdgpu_bo buffer object; and if requested, also creates a
+ * shadow object.
+ * Shadow object is used to backup the original buffer object, and is always
+ * in GTT.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
+int amdgpu_bo_create(struct amdgpu_device *adev,
+		     struct amdgpu_bo_param *bp,
 		     struct amdgpu_bo **bo_ptr)
 {
-	uint64_t parent_flags = flags & ~AMDGPU_GEM_CREATE_SHADOW;
+	u64 flags = bp->flags;
 	int r;
 
-	r = amdgpu_bo_do_create(adev, size, byte_align, domain,
-				parent_flags, type, resv, bo_ptr);
+	bp->flags = bp->flags & ~AMDGPU_GEM_CREATE_SHADOW;
+	r = amdgpu_bo_do_create(adev, bp, bo_ptr);
 	if (r)
 		return r;
 
 	if ((flags & AMDGPU_GEM_CREATE_SHADOW) && amdgpu_need_backup(adev)) {
-		if (!resv)
+		if (!bp->resv)
 			WARN_ON(reservation_object_lock((*bo_ptr)->tbo.resv,
 							NULL));
 
-		r = amdgpu_bo_create_shadow(adev, size, byte_align, (*bo_ptr));
+		r = amdgpu_bo_create_shadow(adev, bp->size, bp->byte_align, (*bo_ptr));
 
-		if (!resv)
+		if (!bp->resv)
 			reservation_object_unlock((*bo_ptr)->tbo.resv);
 
 		if (r)
@@ -523,6 +611,21 @@ int amdgpu_bo_create(struct amdgpu_device *adev, unsigned long size,
 	return r;
 }
 
+/**
+ * amdgpu_bo_backup_to_shadow - Backs up an &amdgpu_bo buffer object
+ * @adev: amdgpu device object
+ * @ring: amdgpu_ring for the engine handling the buffer operations
+ * @bo: &amdgpu_bo buffer to be backed up
+ * @resv: reservation object with embedded fence
+ * @fence: dma_fence associated with the operation
+ * @direct: whether to submit the job directly
+ *
+ * Copies an &amdgpu_bo buffer object to its shadow object.
+ * Not used for now.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
 int amdgpu_bo_backup_to_shadow(struct amdgpu_device *adev,
 			       struct amdgpu_ring *ring,
 			       struct amdgpu_bo *bo,
@@ -555,6 +658,18 @@ err:
 	return r;
 }
 
+/**
+ * amdgpu_bo_validate - validate an &amdgpu_bo buffer object
+ * @bo: pointer to the buffer object
+ *
+ * Sets placement according to domain; and changes placement and caching
+ * policy of the buffer object according to the placement.
+ * This is used for validating shadow bos.  It calls ttm_bo_validate() to
+ * make sure the buffer is resident where it needs to be.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
 int amdgpu_bo_validate(struct amdgpu_bo *bo)
 {
 	struct ttm_operation_ctx ctx = { false, false };
@@ -577,6 +692,22 @@ retry:
 	return r;
 }
 
+/**
+ * amdgpu_bo_restore_from_shadow - restore an &amdgpu_bo buffer object
+ * @adev: amdgpu device object
+ * @ring: amdgpu_ring for the engine handling the buffer operations
+ * @bo: &amdgpu_bo buffer to be restored
+ * @resv: reservation object with embedded fence
+ * @fence: dma_fence associated with the operation
+ * @direct: whether to submit the job directly
+ *
+ * Copies a buffer object's shadow content back to the object.
+ * This is used for recovering a buffer from its shadow in case of a gpu
+ * reset where vram context may be lost.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
 int amdgpu_bo_restore_from_shadow(struct amdgpu_device *adev,
 				  struct amdgpu_ring *ring,
 				  struct amdgpu_bo *bo,
@@ -609,6 +740,17 @@ err:
 	return r;
 }
 
+/**
+ * amdgpu_bo_kmap - map an &amdgpu_bo buffer object
+ * @bo: &amdgpu_bo buffer object to be mapped
+ * @ptr: kernel virtual address to be returned
+ *
+ * Calls ttm_bo_kmap() to set up the kernel virtual mapping; calls
+ * amdgpu_bo_kptr() to get the kernel virtual address.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
 int amdgpu_bo_kmap(struct amdgpu_bo *bo, void **ptr)
 {
 	void *kptr;
@@ -639,6 +781,15 @@ int amdgpu_bo_kmap(struct amdgpu_bo *bo, void **ptr)
 	return 0;
 }
 
+/**
+ * amdgpu_bo_kptr - returns a kernel virtual address of the buffer object
+ * @bo: &amdgpu_bo buffer object
+ *
+ * Calls ttm_kmap_obj_virtual() to get the kernel virtual address
+ *
+ * Returns:
+ * the virtual address of a buffer object area.
+ */
 void *amdgpu_bo_kptr(struct amdgpu_bo *bo)
 {
 	bool is_iomem;
@@ -646,12 +797,27 @@ void *amdgpu_bo_kptr(struct amdgpu_bo *bo)
 	return ttm_kmap_obj_virtual(&bo->kmap, &is_iomem);
 }
 
+/**
+ * amdgpu_bo_kunmap - unmap an &amdgpu_bo buffer object
+ * @bo: &amdgpu_bo buffer object to be unmapped
+ *
+ * Unmaps a kernel map set up by amdgpu_bo_kmap().
+ */
 void amdgpu_bo_kunmap(struct amdgpu_bo *bo)
 {
 	if (bo->kmap.bo)
 		ttm_bo_kunmap(&bo->kmap);
 }
 
+/**
+ * amdgpu_bo_ref - reference an &amdgpu_bo buffer object
+ * @bo: &amdgpu_bo buffer object
+ *
+ * References the contained &ttm_buffer_object.
+ *
+ * Returns:
+ * a refcounted pointer to the &amdgpu_bo buffer object.
+ */
 struct amdgpu_bo *amdgpu_bo_ref(struct amdgpu_bo *bo)
 {
 	if (bo == NULL)
@@ -661,6 +827,12 @@ struct amdgpu_bo *amdgpu_bo_ref(struct amdgpu_bo *bo)
 	return bo;
 }
 
+/**
+ * amdgpu_bo_unref - unreference an &amdgpu_bo buffer object
+ * @bo: &amdgpu_bo buffer object
+ *
+ * Unreferences the contained &ttm_buffer_object and clear the pointer
+ */
 void amdgpu_bo_unref(struct amdgpu_bo **bo)
 {
 	struct ttm_buffer_object *tbo;
@@ -674,9 +846,30 @@ void amdgpu_bo_unref(struct amdgpu_bo **bo)
 		*bo = NULL;
 }
 
+/**
+ * amdgpu_bo_pin_restricted - pin an &amdgpu_bo buffer object
+ * @bo: &amdgpu_bo buffer object to be pinned
+ * @domain: domain to be pinned to
+ * @min_offset: the start of requested address range
+ * @max_offset: the end of requested address range
+ *
+ * Pins the buffer object according to requested domain and address range. If
+ * the memory is unbound gart memory, binds the pages into gart table. Adjusts
+ * pin_count and pin_size accordingly.
+ *
+ * Pinning means to lock pages in memory along with keeping them at a fixed
+ * offset. It is required when a buffer can not be moved, for example, when
+ * a display buffer is being scanned out.
+ *
+ * Compared with amdgpu_bo_pin(), this function gives more flexibility on
+ * where to pin a buffer if there are specific restrictions on where a buffer
+ * must be located.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
 int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain,
-			     u64 min_offset, u64 max_offset,
-			     u64 *gpu_addr)
+			     u64 min_offset, u64 max_offset)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
 	struct ttm_operation_ctx ctx = { false, false };
@@ -689,8 +882,17 @@ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain,
 		return -EINVAL;
 
 	/* A shared bo cannot be migrated to VRAM */
-	if (bo->prime_shared_count && (domain == AMDGPU_GEM_DOMAIN_VRAM))
-		return -EINVAL;
+	if (bo->prime_shared_count) {
+		if (domain & AMDGPU_GEM_DOMAIN_GTT)
+			domain = AMDGPU_GEM_DOMAIN_GTT;
+		else
+			return -EINVAL;
+	}
+
+	/* This assumes only APU display buffers are pinned with (VRAM|GTT).
+	 * See function amdgpu_display_supported_domains()
+	 */
+	domain = amdgpu_bo_get_preferred_pin_domain(adev, domain);
 
 	if (bo->pin_count) {
 		uint32_t mem_type = bo->tbo.mem.mem_type;
@@ -699,8 +901,6 @@ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain,
 			return -EINVAL;
 
 		bo->pin_count++;
-		if (gpu_addr)
-			*gpu_addr = amdgpu_bo_gpu_offset(bo);
 
 		if (max_offset != 0) {
 			u64 domain_start = bo->tbo.bdev->man[mem_type].gpu_offset;
@@ -736,34 +936,48 @@ int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain,
 		goto error;
 	}
 
-	r = amdgpu_ttm_alloc_gart(&bo->tbo);
-	if (unlikely(r)) {
-		dev_err(adev->dev, "%p bind failed\n", bo);
-		goto error;
-	}
-
 	bo->pin_count = 1;
-	if (gpu_addr != NULL)
-		*gpu_addr = amdgpu_bo_gpu_offset(bo);
 
 	domain = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type);
 	if (domain == AMDGPU_GEM_DOMAIN_VRAM) {
-		adev->vram_pin_size += amdgpu_bo_size(bo);
-		if (bo->flags & AMDGPU_GEM_CREATE_NO_CPU_ACCESS)
-			adev->invisible_pin_size += amdgpu_bo_size(bo);
+		atomic64_add(amdgpu_bo_size(bo), &adev->vram_pin_size);
+		atomic64_add(amdgpu_vram_mgr_bo_visible_size(bo),
+			     &adev->visible_pin_size);
 	} else if (domain == AMDGPU_GEM_DOMAIN_GTT) {
-		adev->gart_pin_size += amdgpu_bo_size(bo);
+		atomic64_add(amdgpu_bo_size(bo), &adev->gart_pin_size);
 	}
 
 error:
 	return r;
 }
 
-int amdgpu_bo_pin(struct amdgpu_bo *bo, u32 domain, u64 *gpu_addr)
+/**
+ * amdgpu_bo_pin - pin an &amdgpu_bo buffer object
+ * @bo: &amdgpu_bo buffer object to be pinned
+ * @domain: domain to be pinned to
+ *
+ * A simple wrapper to amdgpu_bo_pin_restricted().
+ * Provides a simpler API for buffers that do not have any strict restrictions
+ * on where a buffer must be located.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
+int amdgpu_bo_pin(struct amdgpu_bo *bo, u32 domain)
 {
-	return amdgpu_bo_pin_restricted(bo, domain, 0, 0, gpu_addr);
+	return amdgpu_bo_pin_restricted(bo, domain, 0, 0);
 }
 
+/**
+ * amdgpu_bo_unpin - unpin an &amdgpu_bo buffer object
+ * @bo: &amdgpu_bo buffer object to be unpinned
+ *
+ * Decreases the pin_count, and clears the flags if pin_count reaches 0.
+ * Changes placement and pin size accordingly.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
 int amdgpu_bo_unpin(struct amdgpu_bo *bo)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
@@ -777,28 +991,30 @@ int amdgpu_bo_unpin(struct amdgpu_bo *bo)
 	bo->pin_count--;
 	if (bo->pin_count)
 		return 0;
+
+	amdgpu_bo_subtract_pin_size(bo);
+
 	for (i = 0; i < bo->placement.num_placement; i++) {
 		bo->placements[i].lpfn = 0;
 		bo->placements[i].flags &= ~TTM_PL_FLAG_NO_EVICT;
 	}
 	r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
-	if (unlikely(r)) {
+	if (unlikely(r))
 		dev_err(adev->dev, "%p validate failed for unpin\n", bo);
-		goto error;
-	}
-
-	if (bo->tbo.mem.mem_type == TTM_PL_VRAM) {
-		adev->vram_pin_size -= amdgpu_bo_size(bo);
-		if (bo->flags & AMDGPU_GEM_CREATE_NO_CPU_ACCESS)
-			adev->invisible_pin_size -= amdgpu_bo_size(bo);
-	} else if (bo->tbo.mem.mem_type == TTM_PL_TT) {
-		adev->gart_pin_size -= amdgpu_bo_size(bo);
-	}
 
-error:
 	return r;
 }
 
+/**
+ * amdgpu_bo_evict_vram - evict VRAM buffers
+ * @adev: amdgpu device object
+ *
+ * Evicts all VRAM buffers on the lru list of the memory type.
+ * Mainly used for evicting vram at suspend time.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
 int amdgpu_bo_evict_vram(struct amdgpu_device *adev)
 {
 	/* late 2.6.33 fix IGP hibernate - we need pm ops to do this correct */
@@ -821,6 +1037,15 @@ static const char *amdgpu_vram_names[] = {
 	"DDR4",
 };
 
+/**
+ * amdgpu_bo_init - initialize memory manager
+ * @adev: amdgpu device object
+ *
+ * Calls amdgpu_ttm_init() to initialize amdgpu memory manager.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
 int amdgpu_bo_init(struct amdgpu_device *adev)
 {
 	/* reserve PAT memory space to WC for VRAM */
@@ -838,6 +1063,29 @@ int amdgpu_bo_init(struct amdgpu_device *adev)
 	return amdgpu_ttm_init(adev);
 }
 
+/**
+ * amdgpu_bo_late_init - late init
+ * @adev: amdgpu device object
+ *
+ * Calls amdgpu_ttm_late_init() to free resources used earlier during
+ * initialization.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
+int amdgpu_bo_late_init(struct amdgpu_device *adev)
+{
+	amdgpu_ttm_late_init(adev);
+
+	return 0;
+}
+
+/**
+ * amdgpu_bo_fini - tear down memory manager
+ * @adev: amdgpu device object
+ *
+ * Reverses amdgpu_bo_init() to tear down memory manager.
+ */
 void amdgpu_bo_fini(struct amdgpu_device *adev)
 {
 	amdgpu_ttm_fini(adev);
@@ -845,12 +1093,33 @@ void amdgpu_bo_fini(struct amdgpu_device *adev)
 	arch_io_free_memtype_wc(adev->gmc.aper_base, adev->gmc.aper_size);
 }
 
+/**
+ * amdgpu_bo_fbdev_mmap - mmap fbdev memory
+ * @bo: &amdgpu_bo buffer object
+ * @vma: vma as input from the fbdev mmap method
+ *
+ * Calls ttm_fbdev_mmap() to mmap fbdev memory if it is backed by a bo.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
 int amdgpu_bo_fbdev_mmap(struct amdgpu_bo *bo,
 			     struct vm_area_struct *vma)
 {
 	return ttm_fbdev_mmap(vma, &bo->tbo);
 }
 
+/**
+ * amdgpu_bo_set_tiling_flags - set tiling flags
+ * @bo: &amdgpu_bo buffer object
+ * @tiling_flags: new flags
+ *
+ * Sets buffer object's tiling flags with the new one. Used by GEM ioctl or
+ * kernel driver to set the tiling flags on a buffer.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
 int amdgpu_bo_set_tiling_flags(struct amdgpu_bo *bo, u64 tiling_flags)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
@@ -863,6 +1132,14 @@ int amdgpu_bo_set_tiling_flags(struct amdgpu_bo *bo, u64 tiling_flags)
 	return 0;
 }
 
+/**
+ * amdgpu_bo_get_tiling_flags - get tiling flags
+ * @bo: &amdgpu_bo buffer object
+ * @tiling_flags: returned flags
+ *
+ * Gets buffer object's tiling flags. Used by GEM ioctl or kernel driver to
+ * set the tiling flags on a buffer.
+ */
 void amdgpu_bo_get_tiling_flags(struct amdgpu_bo *bo, u64 *tiling_flags)
 {
 	lockdep_assert_held(&bo->tbo.resv->lock.base);
@@ -871,6 +1148,19 @@ void amdgpu_bo_get_tiling_flags(struct amdgpu_bo *bo, u64 *tiling_flags)
 		*tiling_flags = bo->tiling_flags;
 }
 
+/**
+ * amdgpu_bo_set_metadata - set metadata
+ * @bo: &amdgpu_bo buffer object
+ * @metadata: new metadata
+ * @metadata_size: size of the new metadata
+ * @flags: flags of the new metadata
+ *
+ * Sets buffer object's metadata, its size and flags.
+ * Used via GEM ioctl.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
 int amdgpu_bo_set_metadata (struct amdgpu_bo *bo, void *metadata,
 			    uint32_t metadata_size, uint64_t flags)
 {
@@ -900,6 +1190,21 @@ int amdgpu_bo_set_metadata (struct amdgpu_bo *bo, void *metadata,
 	return 0;
 }
 
+/**
+ * amdgpu_bo_get_metadata - get metadata
+ * @bo: &amdgpu_bo buffer object
+ * @buffer: returned metadata
+ * @buffer_size: size of the buffer
+ * @metadata_size: size of the returned metadata
+ * @flags: flags of the returned metadata
+ *
+ * Gets buffer object's metadata, its size and flags. buffer_size shall not be
+ * less than metadata_size.
+ * Used via GEM ioctl.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
 int amdgpu_bo_get_metadata(struct amdgpu_bo *bo, void *buffer,
 			   size_t buffer_size, uint32_t *metadata_size,
 			   uint64_t *flags)
@@ -923,6 +1228,16 @@ int amdgpu_bo_get_metadata(struct amdgpu_bo *bo, void *buffer,
 	return 0;
 }
 
+/**
+ * amdgpu_bo_move_notify - notification about a memory move
+ * @bo: pointer to a buffer object
+ * @evict: if this move is evicting the buffer from the graphics address space
+ * @new_mem: new information of the bufer object
+ *
+ * Marks the corresponding &amdgpu_bo buffer object as invalid, also performs
+ * bookkeeping.
+ * TTM driver callback which is called when ttm moves a buffer.
+ */
 void amdgpu_bo_move_notify(struct ttm_buffer_object *bo,
 			   bool evict,
 			   struct ttm_mem_reg *new_mem)
@@ -951,6 +1266,17 @@ void amdgpu_bo_move_notify(struct ttm_buffer_object *bo,
 	trace_amdgpu_ttm_bo_move(abo, new_mem->mem_type, old_mem->mem_type);
 }
 
+/**
+ * amdgpu_bo_fault_reserve_notify - notification about a memory fault
+ * @bo: pointer to a buffer object
+ *
+ * Notifies the driver we are taking a fault on this BO and have reserved it,
+ * also performs bookkeeping.
+ * TTM driver callback for dealing with vm faults.
+ *
+ * Returns:
+ * 0 for success or a negative error code on failure.
+ */
 int amdgpu_bo_fault_reserve_notify(struct ttm_buffer_object *bo)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->bdev);
@@ -1024,10 +1350,11 @@ void amdgpu_bo_fence(struct amdgpu_bo *bo, struct dma_fence *fence,
  * amdgpu_bo_gpu_offset - return GPU offset of bo
  * @bo:	amdgpu object for which we query the offset
  *
- * Returns current GPU offset of the object.
- *
  * Note: object should either be pinned or reserved when calling this
  * function, it might be useful to add check for this for debugging.
+ *
+ * Returns:
+ * current GPU offset of the object.
  */
 u64 amdgpu_bo_gpu_offset(struct amdgpu_bo *bo)
 {
@@ -1042,3 +1369,22 @@ u64 amdgpu_bo_gpu_offset(struct amdgpu_bo *bo)
 
 	return bo->tbo.offset;
 }
+
+/**
+ * amdgpu_bo_get_preferred_pin_domain - get preferred domain for scanout
+ * @adev: amdgpu device object
+ * @domain: allowed :ref:`memory domains <amdgpu_memory_domains>`
+ *
+ * Returns:
+ * Which of the allowed domains is preferred for pinning the BO for scanout.
+ */
+uint32_t amdgpu_bo_get_preferred_pin_domain(struct amdgpu_device *adev,
+					    uint32_t domain)
+{
+	if (domain == (AMDGPU_GEM_DOMAIN_VRAM | AMDGPU_GEM_DOMAIN_GTT)) {
+		domain = AMDGPU_GEM_DOMAIN_VRAM;
+		if (adev->gmc.real_vram_size <= AMDGPU_SG_THRESHOLD)
+			domain = AMDGPU_GEM_DOMAIN_GTT;
+	}
+	return domain;
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
index 546f77cb7882..9c3e29a04eb1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
@@ -33,6 +33,16 @@
 
 #define AMDGPU_BO_INVALID_OFFSET	LONG_MAX
 
+struct amdgpu_bo_param {
+	unsigned long			size;
+	int				byte_align;
+	u32				domain;
+	u32				preferred_domain;
+	u64				flags;
+	enum ttm_bo_type		type;
+	struct reservation_object	*resv;
+};
+
 /* bo virtual addresses in a vm */
 struct amdgpu_bo_va_mapping {
 	struct amdgpu_bo_va		*bo_va;
@@ -196,6 +206,27 @@ static inline bool amdgpu_bo_gpu_accessible(struct amdgpu_bo *bo)
 }
 
 /**
+ * amdgpu_bo_in_cpu_visible_vram - check if BO is (partly) in visible VRAM
+ */
+static inline bool amdgpu_bo_in_cpu_visible_vram(struct amdgpu_bo *bo)
+{
+	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
+	unsigned fpfn = adev->gmc.visible_vram_size >> PAGE_SHIFT;
+	struct drm_mm_node *node = bo->tbo.mem.mm_node;
+	unsigned long pages_left;
+
+	if (bo->tbo.mem.mem_type != TTM_PL_VRAM)
+		return false;
+
+	for (pages_left = bo->tbo.mem.num_pages; pages_left;
+	     pages_left -= node->size, node++)
+		if (node->start < fpfn)
+			return true;
+
+	return false;
+}
+
+/**
  * amdgpu_bo_explicit_sync - return whether the bo is explicitly synced
  */
 static inline bool amdgpu_bo_explicit_sync(struct amdgpu_bo *bo)
@@ -203,10 +234,8 @@ static inline bool amdgpu_bo_explicit_sync(struct amdgpu_bo *bo)
 	return bo->flags & AMDGPU_GEM_CREATE_EXPLICIT_SYNC;
 }
 
-int amdgpu_bo_create(struct amdgpu_device *adev, unsigned long size,
-		     int byte_align, u32 domain,
-		     u64 flags, enum ttm_bo_type type,
-		     struct reservation_object *resv,
+int amdgpu_bo_create(struct amdgpu_device *adev,
+		     struct amdgpu_bo_param *bp,
 		     struct amdgpu_bo **bo_ptr);
 int amdgpu_bo_create_reserved(struct amdgpu_device *adev,
 			      unsigned long size, int align,
@@ -223,13 +252,13 @@ void *amdgpu_bo_kptr(struct amdgpu_bo *bo);
 void amdgpu_bo_kunmap(struct amdgpu_bo *bo);
 struct amdgpu_bo *amdgpu_bo_ref(struct amdgpu_bo *bo);
 void amdgpu_bo_unref(struct amdgpu_bo **bo);
-int amdgpu_bo_pin(struct amdgpu_bo *bo, u32 domain, u64 *gpu_addr);
+int amdgpu_bo_pin(struct amdgpu_bo *bo, u32 domain);
 int amdgpu_bo_pin_restricted(struct amdgpu_bo *bo, u32 domain,
-			     u64 min_offset, u64 max_offset,
-			     u64 *gpu_addr);
+			     u64 min_offset, u64 max_offset);
 int amdgpu_bo_unpin(struct amdgpu_bo *bo);
 int amdgpu_bo_evict_vram(struct amdgpu_device *adev);
 int amdgpu_bo_init(struct amdgpu_device *adev);
+int amdgpu_bo_late_init(struct amdgpu_device *adev);
 void amdgpu_bo_fini(struct amdgpu_device *adev);
 int amdgpu_bo_fbdev_mmap(struct amdgpu_bo *bo,
 				struct vm_area_struct *vma);
@@ -259,7 +288,8 @@ int amdgpu_bo_restore_from_shadow(struct amdgpu_device *adev,
 				  struct reservation_object *resv,
 				  struct dma_fence **fence,
 				  bool direct);
-
+uint32_t amdgpu_bo_get_preferred_pin_domain(struct amdgpu_device *adev,
+					    uint32_t domain);
 
 /*
  * sub allocation
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index 361975cf45a9..15a1192c1ec5 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -68,15 +68,49 @@ void amdgpu_pm_acpi_event_handler(struct amdgpu_device *adev)
 	if (adev->pm.dpm_enabled) {
 		mutex_lock(&adev->pm.mutex);
 		if (power_supply_is_system_supplied() > 0)
-			adev->pm.dpm.ac_power = true;
+			adev->pm.ac_power = true;
 		else
-			adev->pm.dpm.ac_power = false;
+			adev->pm.ac_power = false;
 		if (adev->powerplay.pp_funcs->enable_bapm)
-			amdgpu_dpm_enable_bapm(adev, adev->pm.dpm.ac_power);
+			amdgpu_dpm_enable_bapm(adev, adev->pm.ac_power);
 		mutex_unlock(&adev->pm.mutex);
 	}
 }
 
+/**
+ * DOC: power_dpm_state
+ *
+ * The power_dpm_state file is a legacy interface and is only provided for
+ * backwards compatibility. The amdgpu driver provides a sysfs API for adjusting
+ * certain power related parameters.  The file power_dpm_state is used for this.
+ * It accepts the following arguments:
+ *
+ * - battery
+ *
+ * - balanced
+ *
+ * - performance
+ *
+ * battery
+ *
+ * On older GPUs, the vbios provided a special power state for battery
+ * operation.  Selecting battery switched to this state.  This is no
+ * longer provided on newer GPUs so the option does nothing in that case.
+ *
+ * balanced
+ *
+ * On older GPUs, the vbios provided a special power state for balanced
+ * operation.  Selecting balanced switched to this state.  This is no
+ * longer provided on newer GPUs so the option does nothing in that case.
+ *
+ * performance
+ *
+ * On older GPUs, the vbios provided a special power state for performance
+ * operation.  Selecting performance switched to this state.  This is no
+ * longer provided on newer GPUs so the option does nothing in that case.
+ *
+ */
+
 static ssize_t amdgpu_get_dpm_state(struct device *dev,
 				    struct device_attribute *attr,
 				    char *buf)
@@ -131,6 +165,66 @@ fail:
 	return count;
 }
 
+
+/**
+ * DOC: power_dpm_force_performance_level
+ *
+ * The amdgpu driver provides a sysfs API for adjusting certain power
+ * related parameters.  The file power_dpm_force_performance_level is
+ * used for this.  It accepts the following arguments:
+ *
+ * - auto
+ *
+ * - low
+ *
+ * - high
+ *
+ * - manual
+ *
+ * - profile_standard
+ *
+ * - profile_min_sclk
+ *
+ * - profile_min_mclk
+ *
+ * - profile_peak
+ *
+ * auto
+ *
+ * When auto is selected, the driver will attempt to dynamically select
+ * the optimal power profile for current conditions in the driver.
+ *
+ * low
+ *
+ * When low is selected, the clocks are forced to the lowest power state.
+ *
+ * high
+ *
+ * When high is selected, the clocks are forced to the highest power state.
+ *
+ * manual
+ *
+ * When manual is selected, the user can manually adjust which power states
+ * are enabled for each clock domain via the sysfs pp_dpm_mclk, pp_dpm_sclk,
+ * and pp_dpm_pcie files and adjust the power state transition heuristics
+ * via the pp_power_profile_mode sysfs file.
+ *
+ * profile_standard
+ * profile_min_sclk
+ * profile_min_mclk
+ * profile_peak
+ *
+ * When the profiling modes are selected, clock and power gating are
+ * disabled and the clocks are set for different profiling cases. This
+ * mode is recommended for profiling specific work loads where you do
+ * not want clock or power gating for clock fluctuation to interfere
+ * with your results. profile_standard sets the clocks to a fixed clock
+ * level which varies from asic to asic.  profile_min_sclk forces the sclk
+ * to the lowest level.  profile_min_mclk forces the mclk to the lowest level.
+ * profile_peak sets all clocks (mclk, sclk, pcie) to the highest levels.
+ *
+ */
+
 static ssize_t amdgpu_get_dpm_forced_performance_level(struct device *dev,
 						struct device_attribute *attr,
 								char *buf)
@@ -324,6 +418,17 @@ fail:
 	return count;
 }
 
+/**
+ * DOC: pp_table
+ *
+ * The amdgpu driver provides a sysfs API for uploading new powerplay
+ * tables.  The file pp_table is used for this.  Reading the file
+ * will dump the current power play table.  Writing to the file
+ * will attempt to upload a new powerplay table and re-initialize
+ * powerplay using that new table.
+ *
+ */
+
 static ssize_t amdgpu_get_pp_table(struct device *dev,
 		struct device_attribute *attr,
 		char *buf)
@@ -360,6 +465,32 @@ static ssize_t amdgpu_set_pp_table(struct device *dev,
 	return count;
 }
 
+/**
+ * DOC: pp_od_clk_voltage
+ *
+ * The amdgpu driver provides a sysfs API for adjusting the clocks and voltages
+ * in each power level within a power state.  The pp_od_clk_voltage is used for
+ * this.
+ *
+ * Reading the file will display:
+ *
+ * - a list of engine clock levels and voltages labeled OD_SCLK
+ *
+ * - a list of memory clock levels and voltages labeled OD_MCLK
+ *
+ * - a list of valid ranges for sclk, mclk, and voltage labeled OD_RANGE
+ *
+ * To manually adjust these settings, first select manual using
+ * power_dpm_force_performance_level. Enter a new value for each
+ * level by writing a string that contains "s/m level clock voltage" to
+ * the file.  E.g., "s 1 500 820" will update sclk level 1 to be 500 MHz
+ * at 820 mV; "m 0 350 810" will update mclk level 0 to be 350 MHz at
+ * 810 mV.  When you have edited all of the states as needed, write
+ * "c" (commit) to the file to commit your changes.  If you want to reset to the
+ * default power levels, write "r" (reset) to the file to reset them.
+ *
+ */
+
 static ssize_t amdgpu_set_pp_od_clk_voltage(struct device *dev,
 		struct device_attribute *attr,
 		const char *buf,
@@ -437,6 +568,7 @@ static ssize_t amdgpu_get_pp_od_clk_voltage(struct device *dev,
 	if (adev->powerplay.pp_funcs->print_clock_levels) {
 		size = amdgpu_dpm_print_clock_levels(adev, OD_SCLK, buf);
 		size += amdgpu_dpm_print_clock_levels(adev, OD_MCLK, buf+size);
+		size += amdgpu_dpm_print_clock_levels(adev, OD_RANGE, buf+size);
 		return size;
 	} else {
 		return snprintf(buf, PAGE_SIZE, "\n");
@@ -444,6 +576,23 @@ static ssize_t amdgpu_get_pp_od_clk_voltage(struct device *dev,
 
 }
 
+/**
+ * DOC: pp_dpm_sclk pp_dpm_mclk pp_dpm_pcie
+ *
+ * The amdgpu driver provides a sysfs API for adjusting what power levels
+ * are enabled for a given power state.  The files pp_dpm_sclk, pp_dpm_mclk,
+ * and pp_dpm_pcie are used for this.
+ *
+ * Reading back the files will show you the available power levels within
+ * the power state and the clock information for those levels.
+ *
+ * To manually adjust these states, first select manual using
+ * power_dpm_force_performance_level.
+ * Secondly,Enter a new value for each level by inputing a string that
+ * contains " echo xx xx xx > pp_dpm_sclk/mclk/pcie"
+ * E.g., echo 4 5 6 to > pp_dpm_sclk will enable sclk levels 4, 5, and 6.
+ */
+
 static ssize_t amdgpu_get_pp_dpm_sclk(struct device *dev,
 		struct device_attribute *attr,
 		char *buf)
@@ -457,6 +606,42 @@ static ssize_t amdgpu_get_pp_dpm_sclk(struct device *dev,
 		return snprintf(buf, PAGE_SIZE, "\n");
 }
 
+/*
+ * Worst case: 32 bits individually specified, in octal at 12 characters
+ * per line (+1 for \n).
+ */
+#define AMDGPU_MASK_BUF_MAX	(32 * 13)
+
+static ssize_t amdgpu_read_mask(const char *buf, size_t count, uint32_t *mask)
+{
+	int ret;
+	long level;
+	char *sub_str = NULL;
+	char *tmp;
+	char buf_cpy[AMDGPU_MASK_BUF_MAX + 1];
+	const char delimiter[3] = {' ', '\n', '\0'};
+	size_t bytes;
+
+	*mask = 0;
+
+	bytes = min(count, sizeof(buf_cpy) - 1);
+	memcpy(buf_cpy, buf, bytes);
+	buf_cpy[bytes] = '\0';
+	tmp = buf_cpy;
+	while (tmp[0]) {
+		sub_str = strsep(&tmp, delimiter);
+		if (strlen(sub_str)) {
+			ret = kstrtol(sub_str, 0, &level);
+			if (ret)
+				return -EINVAL;
+			*mask |= 1 << level;
+		} else
+			break;
+	}
+
+	return 0;
+}
+
 static ssize_t amdgpu_set_pp_dpm_sclk(struct device *dev,
 		struct device_attribute *attr,
 		const char *buf,
@@ -465,28 +650,15 @@ static ssize_t amdgpu_set_pp_dpm_sclk(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = ddev->dev_private;
 	int ret;
-	long level;
-	uint32_t i, mask = 0;
-	char sub_str[2];
-
-	for (i = 0; i < strlen(buf); i++) {
-		if (*(buf + i) == '\n')
-			continue;
-		sub_str[0] = *(buf + i);
-		sub_str[1] = '\0';
-		ret = kstrtol(sub_str, 0, &level);
+	uint32_t mask = 0;
 
-		if (ret) {
-			count = -EINVAL;
-			goto fail;
-		}
-		mask |= 1 << level;
-	}
+	ret = amdgpu_read_mask(buf, count, &mask);
+	if (ret)
+		return ret;
 
 	if (adev->powerplay.pp_funcs->force_clock_level)
 		amdgpu_dpm_force_clock_level(adev, PP_SCLK, mask);
 
-fail:
 	return count;
 }
 
@@ -511,27 +683,15 @@ static ssize_t amdgpu_set_pp_dpm_mclk(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = ddev->dev_private;
 	int ret;
-	long level;
-	uint32_t i, mask = 0;
-	char sub_str[2];
+	uint32_t mask = 0;
 
-	for (i = 0; i < strlen(buf); i++) {
-		if (*(buf + i) == '\n')
-			continue;
-		sub_str[0] = *(buf + i);
-		sub_str[1] = '\0';
-		ret = kstrtol(sub_str, 0, &level);
+	ret = amdgpu_read_mask(buf, count, &mask);
+	if (ret)
+		return ret;
 
-		if (ret) {
-			count = -EINVAL;
-			goto fail;
-		}
-		mask |= 1 << level;
-	}
 	if (adev->powerplay.pp_funcs->force_clock_level)
 		amdgpu_dpm_force_clock_level(adev, PP_MCLK, mask);
 
-fail:
 	return count;
 }
 
@@ -556,27 +716,15 @@ static ssize_t amdgpu_set_pp_dpm_pcie(struct device *dev,
 	struct drm_device *ddev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = ddev->dev_private;
 	int ret;
-	long level;
-	uint32_t i, mask = 0;
-	char sub_str[2];
+	uint32_t mask = 0;
 
-	for (i = 0; i < strlen(buf); i++) {
-		if (*(buf + i) == '\n')
-			continue;
-		sub_str[0] = *(buf + i);
-		sub_str[1] = '\0';
-		ret = kstrtol(sub_str, 0, &level);
+	ret = amdgpu_read_mask(buf, count, &mask);
+	if (ret)
+		return ret;
 
-		if (ret) {
-			count = -EINVAL;
-			goto fail;
-		}
-		mask |= 1 << level;
-	}
 	if (adev->powerplay.pp_funcs->force_clock_level)
 		amdgpu_dpm_force_clock_level(adev, PP_PCIE, mask);
 
-fail:
 	return count;
 }
 
@@ -668,6 +816,26 @@ fail:
 	return count;
 }
 
+/**
+ * DOC: pp_power_profile_mode
+ *
+ * The amdgpu driver provides a sysfs API for adjusting the heuristics
+ * related to switching between power levels in a power state.  The file
+ * pp_power_profile_mode is used for this.
+ *
+ * Reading this file outputs a list of all of the predefined power profiles
+ * and the relevant heuristics settings for that profile.
+ *
+ * To select a profile or create a custom profile, first select manual using
+ * power_dpm_force_performance_level.  Writing the number of a predefined
+ * profile to pp_power_profile_mode will enable those heuristics.  To
+ * create a custom set of heuristics, write a string of numbers to the file
+ * starting with the number of the custom profile along with a setting
+ * for each heuristic parameter.  Due to differences across asic families
+ * the heuristic parameters vary from family to family.
+ *
+ */
+
 static ssize_t amdgpu_get_pp_power_profile_mode(struct device *dev,
 		struct device_attribute *attr,
 		char *buf)
@@ -734,6 +902,36 @@ fail:
 	return -EINVAL;
 }
 
+/**
+ * DOC: busy_percent
+ *
+ * The amdgpu driver provides a sysfs API for reading how busy the GPU
+ * is as a percentage.  The file gpu_busy_percent is used for this.
+ * The SMU firmware computes a percentage of load based on the
+ * aggregate activity level in the IP cores.
+ */
+static ssize_t amdgpu_get_busy_percent(struct device *dev,
+		struct device_attribute *attr,
+		char *buf)
+{
+	struct drm_device *ddev = dev_get_drvdata(dev);
+	struct amdgpu_device *adev = ddev->dev_private;
+	int r, value, size = sizeof(value);
+
+	/* sanity check PP is enabled */
+	if (!(adev->powerplay.pp_funcs &&
+	      adev->powerplay.pp_funcs->read_sensor))
+		return -EINVAL;
+
+	/* read the IP busy sensor */
+	r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_LOAD,
+				   (void *)&value, &size);
+	if (r)
+		return r;
+
+	return snprintf(buf, PAGE_SIZE, "%d\n", value);
+}
+
 static DEVICE_ATTR(power_dpm_state, S_IRUGO | S_IWUSR, amdgpu_get_dpm_state, amdgpu_set_dpm_state);
 static DEVICE_ATTR(power_dpm_force_performance_level, S_IRUGO | S_IWUSR,
 		   amdgpu_get_dpm_forced_performance_level,
@@ -767,6 +965,8 @@ static DEVICE_ATTR(pp_power_profile_mode, S_IRUGO | S_IWUSR,
 static DEVICE_ATTR(pp_od_clk_voltage, S_IRUGO | S_IWUSR,
 		amdgpu_get_pp_od_clk_voltage,
 		amdgpu_set_pp_od_clk_voltage);
+static DEVICE_ATTR(gpu_busy_percent, S_IRUGO,
+		amdgpu_get_busy_percent, NULL);
 
 static ssize_t amdgpu_hwmon_show_temp(struct device *dev,
 				      struct device_attribute *attr,
@@ -1020,8 +1220,8 @@ static ssize_t amdgpu_hwmon_show_power_avg(struct device *dev,
 {
 	struct amdgpu_device *adev = dev_get_drvdata(dev);
 	struct drm_device *ddev = adev->ddev;
-	struct pp_gpu_power query = {0};
-	int r, size = sizeof(query);
+	u32 query = 0;
+	int r, size = sizeof(u32);
 	unsigned uw;
 
 	/* Can't get power when the card is off */
@@ -1041,7 +1241,7 @@ static ssize_t amdgpu_hwmon_show_power_avg(struct device *dev,
 		return r;
 
 	/* convert to microwatts */
-	uw = (query.average_gpu_power >> 8) * 1000000;
+	uw = (query >> 8) * 1000000 + (query & 0xff) * 1000;
 
 	return snprintf(buf, PAGE_SIZE, "%u\n", uw);
 }
@@ -1109,6 +1309,62 @@ static ssize_t amdgpu_hwmon_set_power_cap(struct device *dev,
 	return count;
 }
 
+
+/**
+ * DOC: hwmon
+ *
+ * The amdgpu driver exposes the following sensor interfaces:
+ *
+ * - GPU temperature (via the on-die sensor)
+ *
+ * - GPU voltage
+ *
+ * - Northbridge voltage (APUs only)
+ *
+ * - GPU power
+ *
+ * - GPU fan
+ *
+ * hwmon interfaces for GPU temperature:
+ *
+ * - temp1_input: the on die GPU temperature in millidegrees Celsius
+ *
+ * - temp1_crit: temperature critical max value in millidegrees Celsius
+ *
+ * - temp1_crit_hyst: temperature hysteresis for critical limit in millidegrees Celsius
+ *
+ * hwmon interfaces for GPU voltage:
+ *
+ * - in0_input: the voltage on the GPU in millivolts
+ *
+ * - in1_input: the voltage on the Northbridge in millivolts
+ *
+ * hwmon interfaces for GPU power:
+ *
+ * - power1_average: average power used by the GPU in microWatts
+ *
+ * - power1_cap_min: minimum cap supported in microWatts
+ *
+ * - power1_cap_max: maximum cap supported in microWatts
+ *
+ * - power1_cap: selected power cap in microWatts
+ *
+ * hwmon interfaces for GPU fan:
+ *
+ * - pwm1: pulse width modulation fan level (0-255)
+ *
+ * - pwm1_enable: pulse width modulation fan control method (0: no fan speed control, 1: manual fan speed control using pwm interface, 2: automatic fan speed control)
+ *
+ * - pwm1_min: pulse width modulation fan control minimum level (0)
+ *
+ * - pwm1_max: pulse width modulation fan control maximum level (255)
+ *
+ * - fan1_input: fan speed in RPM
+ *
+ * You can use hwmon tools like sensors to view this information on your system.
+ *
+ */
+
 static SENSOR_DEVICE_ATTR(temp1_input, S_IRUGO, amdgpu_hwmon_show_temp, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp1_crit, S_IRUGO, amdgpu_hwmon_show_temp_thresh, NULL, 0);
 static SENSOR_DEVICE_ATTR(temp1_crit_hyst, S_IRUGO, amdgpu_hwmon_show_temp_thresh, NULL, 1);
@@ -1153,19 +1409,14 @@ static umode_t hwmon_attributes_visible(struct kobject *kobj,
 	struct amdgpu_device *adev = dev_get_drvdata(dev);
 	umode_t effective_mode = attr->mode;
 
-	/* handle non-powerplay limitations */
-	if (!adev->powerplay.pp_handle) {
-		/* Skip fan attributes if fan is not present */
-		if (adev->pm.no_fan &&
-		    (attr == &sensor_dev_attr_pwm1.dev_attr.attr ||
-		     attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr ||
-		     attr == &sensor_dev_attr_pwm1_max.dev_attr.attr ||
-		     attr == &sensor_dev_attr_pwm1_min.dev_attr.attr))
-			return 0;
-		/* requires powerplay */
-		if (attr == &sensor_dev_attr_fan1_input.dev_attr.attr)
-			return 0;
-	}
+
+	/* Skip fan attributes if fan is not present */
+	if (adev->pm.no_fan && (attr == &sensor_dev_attr_pwm1.dev_attr.attr ||
+	    attr == &sensor_dev_attr_pwm1_enable.dev_attr.attr ||
+	    attr == &sensor_dev_attr_pwm1_max.dev_attr.attr ||
+	    attr == &sensor_dev_attr_pwm1_min.dev_attr.attr ||
+	    attr == &sensor_dev_attr_fan1_input.dev_attr.attr))
+		return 0;
 
 	/* Skip limit attributes if DPM is not enabled */
 	if (!adev->pm.dpm_enabled &&
@@ -1462,10 +1713,10 @@ static void amdgpu_dpm_change_power_state_locked(struct amdgpu_device *adev)
 
 void amdgpu_dpm_enable_uvd(struct amdgpu_device *adev, bool enable)
 {
-	if (adev->powerplay.pp_funcs->powergate_uvd) {
+	if (adev->powerplay.pp_funcs->set_powergating_by_smu) {
 		/* enable/disable UVD */
 		mutex_lock(&adev->pm.mutex);
-		amdgpu_dpm_powergate_uvd(adev, !enable);
+		amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_UVD, !enable);
 		mutex_unlock(&adev->pm.mutex);
 	} else {
 		if (enable) {
@@ -1484,10 +1735,10 @@ void amdgpu_dpm_enable_uvd(struct amdgpu_device *adev, bool enable)
 
 void amdgpu_dpm_enable_vce(struct amdgpu_device *adev, bool enable)
 {
-	if (adev->powerplay.pp_funcs->powergate_vce) {
+	if (adev->powerplay.pp_funcs->set_powergating_by_smu) {
 		/* enable/disable VCE */
 		mutex_lock(&adev->pm.mutex);
-		amdgpu_dpm_powergate_vce(adev, !enable);
+		amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_VCE, !enable);
 		mutex_unlock(&adev->pm.mutex);
 	} else {
 		if (enable) {
@@ -1619,6 +1870,13 @@ int amdgpu_pm_sysfs_init(struct amdgpu_device *adev)
 				"pp_od_clk_voltage\n");
 		return ret;
 	}
+	ret = device_create_file(adev->dev,
+			&dev_attr_gpu_busy_percent);
+	if (ret) {
+		DRM_ERROR("failed to create device file	"
+				"gpu_busy_level\n");
+		return ret;
+	}
 	ret = amdgpu_debugfs_pm_init(adev);
 	if (ret) {
 		DRM_ERROR("Failed to register debugfs file for dpm!\n");
@@ -1654,13 +1912,11 @@ void amdgpu_pm_sysfs_fini(struct amdgpu_device *adev)
 			&dev_attr_pp_power_profile_mode);
 	device_remove_file(adev->dev,
 			&dev_attr_pp_od_clk_voltage);
+	device_remove_file(adev->dev, &dev_attr_gpu_busy_percent);
 }
 
 void amdgpu_pm_compute_clocks(struct amdgpu_device *adev)
 {
-	struct drm_device *ddev = adev->ddev;
-	struct drm_crtc *crtc;
-	struct amdgpu_crtc *amdgpu_crtc;
 	int i = 0;
 
 	if (!adev->pm.dpm_enabled)
@@ -1675,30 +1931,35 @@ void amdgpu_pm_compute_clocks(struct amdgpu_device *adev)
 			amdgpu_fence_wait_empty(ring);
 	}
 
+	mutex_lock(&adev->pm.mutex);
+	/* update battery/ac status */
+	if (power_supply_is_system_supplied() > 0)
+		adev->pm.ac_power = true;
+	else
+		adev->pm.ac_power = false;
+	mutex_unlock(&adev->pm.mutex);
+
 	if (adev->powerplay.pp_funcs->dispatch_tasks) {
+		if (!amdgpu_device_has_dc_support(adev)) {
+			mutex_lock(&adev->pm.mutex);
+			amdgpu_dpm_get_active_displays(adev);
+			adev->pm.pm_display_cfg.num_display = adev->pm.dpm.new_active_crtc_count;
+			adev->pm.pm_display_cfg.vrefresh = amdgpu_dpm_get_vrefresh(adev);
+			adev->pm.pm_display_cfg.min_vblank_time = amdgpu_dpm_get_vblank_time(adev);
+			/* we have issues with mclk switching with refresh rates over 120 hz on the non-DC code. */
+			if (adev->pm.pm_display_cfg.vrefresh > 120)
+				adev->pm.pm_display_cfg.min_vblank_time = 0;
+			if (adev->powerplay.pp_funcs->display_configuration_change)
+				adev->powerplay.pp_funcs->display_configuration_change(
+								adev->powerplay.pp_handle,
+								&adev->pm.pm_display_cfg);
+			mutex_unlock(&adev->pm.mutex);
+		}
 		amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_DISPLAY_CONFIG_CHANGE, NULL);
 	} else {
 		mutex_lock(&adev->pm.mutex);
-		adev->pm.dpm.new_active_crtcs = 0;
-		adev->pm.dpm.new_active_crtc_count = 0;
-		if (adev->mode_info.num_crtc && adev->mode_info.mode_config_initialized) {
-			list_for_each_entry(crtc,
-					    &ddev->mode_config.crtc_list, head) {
-				amdgpu_crtc = to_amdgpu_crtc(crtc);
-				if (amdgpu_crtc->enabled) {
-					adev->pm.dpm.new_active_crtcs |= (1 << amdgpu_crtc->crtc_id);
-					adev->pm.dpm.new_active_crtc_count++;
-				}
-			}
-		}
-		/* update battery/ac status */
-		if (power_supply_is_system_supplied() > 0)
-			adev->pm.dpm.ac_power = true;
-		else
-			adev->pm.dpm.ac_power = false;
-
+		amdgpu_dpm_get_active_displays(adev);
 		amdgpu_dpm_change_power_state_locked(adev);
-
 		mutex_unlock(&adev->pm.mutex);
 	}
 }
@@ -1711,7 +1972,7 @@ void amdgpu_pm_compute_clocks(struct amdgpu_device *adev)
 static int amdgpu_debugfs_pm_info_pp(struct seq_file *m, struct amdgpu_device *adev)
 {
 	uint32_t value;
-	struct pp_gpu_power query = {0};
+	uint32_t query = 0;
 	int size;
 
 	/* sanity check PP is enabled */
@@ -1734,17 +1995,9 @@ static int amdgpu_debugfs_pm_info_pp(struct seq_file *m, struct amdgpu_device *a
 		seq_printf(m, "\t%u mV (VDDGFX)\n", value);
 	if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VDDNB, (void *)&value, &size))
 		seq_printf(m, "\t%u mV (VDDNB)\n", value);
-	size = sizeof(query);
-	if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_POWER, (void *)&query, &size)) {
-		seq_printf(m, "\t%u.%u W (VDDC)\n", query.vddc_power >> 8,
-				query.vddc_power & 0xff);
-		seq_printf(m, "\t%u.%u W (VDDCI)\n", query.vddci_power >> 8,
-				query.vddci_power & 0xff);
-		seq_printf(m, "\t%u.%u W (max GPU)\n", query.max_gpu_power >> 8,
-				query.max_gpu_power & 0xff);
-		seq_printf(m, "\t%u.%u W (average GPU)\n", query.average_gpu_power >> 8,
-				query.average_gpu_power & 0xff);
-	}
+	size = sizeof(uint32_t);
+	if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GPU_POWER, (void *)&query, &size))
+		seq_printf(m, "\t%u.%u W (average GPU)\n", query >> 8, query & 0xff);
 	size = sizeof(value);
 	seq_printf(m, "\n");
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c
index 4b584cb75bf4..3ed02f472003 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_prime.c
@@ -23,6 +23,14 @@
  *
  * Authors: Alex Deucher
  */
+
+/**
+ * DOC: PRIME Buffer Sharing
+ *
+ * The following callback implementations are used for :ref:`sharing GEM buffer
+ * objects between different devices via PRIME <prime_buffer_sharing>`.
+ */
+
 #include <drm/drmP.h>
 
 #include "amdgpu.h"
@@ -32,6 +40,14 @@
 
 static const struct dma_buf_ops amdgpu_dmabuf_ops;
 
+/**
+ * amdgpu_gem_prime_get_sg_table - &drm_driver.gem_prime_get_sg_table
+ * implementation
+ * @obj: GEM buffer object
+ *
+ * Returns:
+ * A scatter/gather table for the pinned pages of the buffer object's memory.
+ */
 struct sg_table *amdgpu_gem_prime_get_sg_table(struct drm_gem_object *obj)
 {
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
@@ -40,6 +56,15 @@ struct sg_table *amdgpu_gem_prime_get_sg_table(struct drm_gem_object *obj)
 	return drm_prime_pages_to_sg(bo->tbo.ttm->pages, npages);
 }
 
+/**
+ * amdgpu_gem_prime_vmap - &dma_buf_ops.vmap implementation
+ * @obj: GEM buffer object
+ *
+ * Sets up an in-kernel virtual mapping of the buffer object's memory.
+ *
+ * Returns:
+ * The virtual address of the mapping or an error pointer.
+ */
 void *amdgpu_gem_prime_vmap(struct drm_gem_object *obj)
 {
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
@@ -53,6 +78,13 @@ void *amdgpu_gem_prime_vmap(struct drm_gem_object *obj)
 	return bo->dma_buf_vmap.virtual;
 }
 
+/**
+ * amdgpu_gem_prime_vunmap - &dma_buf_ops.vunmap implementation
+ * @obj: GEM buffer object
+ * @vaddr: virtual address (unused)
+ *
+ * Tears down the in-kernel virtual mapping of the buffer object's memory.
+ */
 void amdgpu_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr)
 {
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
@@ -60,6 +92,17 @@ void amdgpu_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr)
 	ttm_bo_kunmap(&bo->dma_buf_vmap);
 }
 
+/**
+ * amdgpu_gem_prime_mmap - &drm_driver.gem_prime_mmap implementation
+ * @obj: GEM buffer object
+ * @vma: virtual memory area
+ *
+ * Sets up a userspace mapping of the buffer object's memory in the given
+ * virtual memory area.
+ *
+ * Returns:
+ * 0 on success or negative error code.
+ */
 int amdgpu_gem_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma)
 {
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
@@ -94,6 +137,19 @@ int amdgpu_gem_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma
 	return ret;
 }
 
+/**
+ * amdgpu_gem_prime_import_sg_table - &drm_driver.gem_prime_import_sg_table
+ * implementation
+ * @dev: DRM device
+ * @attach: DMA-buf attachment
+ * @sg: Scatter/gather table
+ *
+ * Import shared DMA buffer memory exported by another device.
+ *
+ * Returns:
+ * A new GEM buffer object of the given DRM device, representing the memory
+ * described by the given DMA-buf attachment and scatter/gather table.
+ */
 struct drm_gem_object *
 amdgpu_gem_prime_import_sg_table(struct drm_device *dev,
 				 struct dma_buf_attachment *attach,
@@ -102,12 +158,18 @@ amdgpu_gem_prime_import_sg_table(struct drm_device *dev,
 	struct reservation_object *resv = attach->dmabuf->resv;
 	struct amdgpu_device *adev = dev->dev_private;
 	struct amdgpu_bo *bo;
+	struct amdgpu_bo_param bp;
 	int ret;
 
+	memset(&bp, 0, sizeof(bp));
+	bp.size = attach->dmabuf->size;
+	bp.byte_align = PAGE_SIZE;
+	bp.domain = AMDGPU_GEM_DOMAIN_CPU;
+	bp.flags = 0;
+	bp.type = ttm_bo_type_sg;
+	bp.resv = resv;
 	ww_mutex_lock(&resv->lock, NULL);
-	ret = amdgpu_bo_create(adev, attach->dmabuf->size, PAGE_SIZE,
-			       AMDGPU_GEM_DOMAIN_CPU, 0, ttm_bo_type_sg,
-			       resv, &bo);
+	ret = amdgpu_bo_create(adev, &bp, &bo);
 	if (ret)
 		goto error;
 
@@ -126,8 +188,19 @@ error:
 	return ERR_PTR(ret);
 }
 
+/**
+ * amdgpu_gem_map_attach - &dma_buf_ops.attach implementation
+ * @dma_buf: shared DMA buffer
+ * @attach: DMA-buf attachment
+ *
+ * Makes sure that the shared DMA buffer can be accessed by the target device.
+ * For now, simply pins it to the GTT domain, where it should be accessible by
+ * all DMA devices.
+ *
+ * Returns:
+ * 0 on success or negative error code.
+ */
 static int amdgpu_gem_map_attach(struct dma_buf *dma_buf,
-				 struct device *target_dev,
 				 struct dma_buf_attachment *attach)
 {
 	struct drm_gem_object *obj = dma_buf->priv;
@@ -135,7 +208,7 @@ static int amdgpu_gem_map_attach(struct dma_buf *dma_buf,
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
 	long r;
 
-	r = drm_gem_map_attach(dma_buf, target_dev, attach);
+	r = drm_gem_map_attach(dma_buf, attach);
 	if (r)
 		return r;
 
@@ -159,7 +232,7 @@ static int amdgpu_gem_map_attach(struct dma_buf *dma_buf,
 	}
 
 	/* pin buffer into GTT */
-	r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT, NULL);
+	r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_GTT);
 	if (r)
 		goto error_unreserve;
 
@@ -175,6 +248,14 @@ error_detach:
 	return r;
 }
 
+/**
+ * amdgpu_gem_map_detach - &dma_buf_ops.detach implementation
+ * @dma_buf: shared DMA buffer
+ * @attach: DMA-buf attachment
+ *
+ * This is called when a shared DMA buffer no longer needs to be accessible by
+ * the other device. For now, simply unpins the buffer from GTT.
+ */
 static void amdgpu_gem_map_detach(struct dma_buf *dma_buf,
 				  struct dma_buf_attachment *attach)
 {
@@ -196,6 +277,13 @@ error:
 	drm_gem_map_detach(dma_buf, attach);
 }
 
+/**
+ * amdgpu_gem_prime_res_obj - &drm_driver.gem_prime_res_obj implementation
+ * @obj: GEM buffer object
+ *
+ * Returns:
+ * The buffer object's reservation object.
+ */
 struct reservation_object *amdgpu_gem_prime_res_obj(struct drm_gem_object *obj)
 {
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
@@ -203,13 +291,25 @@ struct reservation_object *amdgpu_gem_prime_res_obj(struct drm_gem_object *obj)
 	return bo->tbo.resv;
 }
 
+/**
+ * amdgpu_gem_begin_cpu_access - &dma_buf_ops.begin_cpu_access implementation
+ * @dma_buf: shared DMA buffer
+ * @direction: direction of DMA transfer
+ *
+ * This is called before CPU access to the shared DMA buffer's memory. If it's
+ * a read access, the buffer is moved to the GTT domain if possible, for optimal
+ * CPU read performance.
+ *
+ * Returns:
+ * 0 on success or negative error code.
+ */
 static int amdgpu_gem_begin_cpu_access(struct dma_buf *dma_buf,
 				       enum dma_data_direction direction)
 {
 	struct amdgpu_bo *bo = gem_to_amdgpu_bo(dma_buf->priv);
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
 	struct ttm_operation_ctx ctx = { true, false };
-	u32 domain = amdgpu_display_framebuffer_domains(adev);
+	u32 domain = amdgpu_display_supported_domains(adev);
 	int ret;
 	bool reads = (direction == DMA_BIDIRECTIONAL ||
 		      direction == DMA_FROM_DEVICE);
@@ -239,14 +339,24 @@ static const struct dma_buf_ops amdgpu_dmabuf_ops = {
 	.release = drm_gem_dmabuf_release,
 	.begin_cpu_access = amdgpu_gem_begin_cpu_access,
 	.map = drm_gem_dmabuf_kmap,
-	.map_atomic = drm_gem_dmabuf_kmap_atomic,
 	.unmap = drm_gem_dmabuf_kunmap,
-	.unmap_atomic = drm_gem_dmabuf_kunmap_atomic,
 	.mmap = drm_gem_dmabuf_mmap,
 	.vmap = drm_gem_dmabuf_vmap,
 	.vunmap = drm_gem_dmabuf_vunmap,
 };
 
+/**
+ * amdgpu_gem_prime_export - &drm_driver.gem_prime_export implementation
+ * @dev: DRM device
+ * @gobj: GEM buffer object
+ * @flags: flags like DRM_CLOEXEC and DRM_RDWR
+ *
+ * The main work is done by the &drm_gem_prime_export helper, which in turn
+ * uses &amdgpu_gem_prime_res_obj.
+ *
+ * Returns:
+ * Shared DMA buffer representing the GEM buffer object from the given device.
+ */
 struct dma_buf *amdgpu_gem_prime_export(struct drm_device *dev,
 					struct drm_gem_object *gobj,
 					int flags)
@@ -267,6 +377,17 @@ struct dma_buf *amdgpu_gem_prime_export(struct drm_device *dev,
 	return buf;
 }
 
+/**
+ * amdgpu_gem_prime_import - &drm_driver.gem_prime_import implementation
+ * @dev: DRM device
+ * @dma_buf: Shared DMA buffer
+ *
+ * The main work is done by the &drm_gem_prime_import helper, which in turn
+ * uses &amdgpu_gem_prime_import_sg_table.
+ *
+ * Returns:
+ * GEM buffer object representing the shared DMA buffer for the given device.
+ */
 struct drm_gem_object *amdgpu_gem_prime_import(struct drm_device *dev,
 					    struct dma_buf *dma_buf)
 {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
index c7d43e064fc7..9f1a5bd39ae8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
@@ -52,6 +52,7 @@ static int psp_sw_init(void *handle)
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		psp_v3_1_set_psp_funcs(psp);
 		break;
 	case CHIP_RAVEN:
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c
index 262c1267249e..ea9850c9224d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_queue_mgr.c
@@ -66,6 +66,8 @@ static int amdgpu_identity_map(struct amdgpu_device *adev,
 			       u32 ring,
 			       struct amdgpu_ring **out_ring)
 {
+	u32 instance;
+
 	switch (mapper->hw_ip) {
 	case AMDGPU_HW_IP_GFX:
 		*out_ring = &adev->gfx.gfx_ring[ring];
@@ -77,13 +79,16 @@ static int amdgpu_identity_map(struct amdgpu_device *adev,
 		*out_ring = &adev->sdma.instance[ring].ring;
 		break;
 	case AMDGPU_HW_IP_UVD:
-		*out_ring = &adev->uvd.ring;
+		instance = ring;
+		*out_ring = &adev->uvd.inst[instance].ring;
 		break;
 	case AMDGPU_HW_IP_VCE:
 		*out_ring = &adev->vce.ring[ring];
 		break;
 	case AMDGPU_HW_IP_UVD_ENC:
-		*out_ring = &adev->uvd.ring_enc[ring];
+		instance = ring / adev->uvd.num_enc_rings;
+		*out_ring =
+		&adev->uvd.inst[instance].ring_enc[ring%adev->uvd.num_enc_rings];
 		break;
 	case AMDGPU_HW_IP_VCN_DEC:
 		*out_ring = &adev->vcn.ring_dec;
@@ -91,6 +96,9 @@ static int amdgpu_identity_map(struct amdgpu_device *adev,
 	case AMDGPU_HW_IP_VCN_ENC:
 		*out_ring = &adev->vcn.ring_enc[ring];
 		break;
+	case AMDGPU_HW_IP_VCN_JPEG:
+		*out_ring = &adev->vcn.ring_jpeg;
+		break;
 	default:
 		*out_ring = NULL;
 		DRM_ERROR("unknown HW IP type: %d\n", mapper->hw_ip);
@@ -240,13 +248,14 @@ int amdgpu_queue_mgr_map(struct amdgpu_device *adev,
 		ip_num_rings = adev->sdma.num_instances;
 		break;
 	case AMDGPU_HW_IP_UVD:
-		ip_num_rings = 1;
+		ip_num_rings = adev->uvd.num_uvd_inst;
 		break;
 	case AMDGPU_HW_IP_VCE:
 		ip_num_rings = adev->vce.num_rings;
 		break;
 	case AMDGPU_HW_IP_UVD_ENC:
-		ip_num_rings = adev->uvd.num_enc_rings;
+		ip_num_rings =
+			adev->uvd.num_enc_rings * adev->uvd.num_uvd_inst;
 		break;
 	case AMDGPU_HW_IP_VCN_DEC:
 		ip_num_rings = 1;
@@ -254,6 +263,9 @@ int amdgpu_queue_mgr_map(struct amdgpu_device *adev,
 	case AMDGPU_HW_IP_VCN_ENC:
 		ip_num_rings = adev->vcn.num_enc_rings;
 		break;
+	case AMDGPU_HW_IP_VCN_JPEG:
+		ip_num_rings = 1;
+		break;
 	default:
 		DRM_DEBUG("unknown ip type: %d\n", hw_ip);
 		return -EINVAL;
@@ -281,6 +293,7 @@ int amdgpu_queue_mgr_map(struct amdgpu_device *adev,
 	case AMDGPU_HW_IP_UVD_ENC:
 	case AMDGPU_HW_IP_VCN_DEC:
 	case AMDGPU_HW_IP_VCN_ENC:
+	case AMDGPU_HW_IP_VCN_JPEG:
 		r = amdgpu_identity_map(adev, mapper, ring, out_ring);
 		break;
 	case AMDGPU_HW_IP_DMA:
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
index d5f526f38e50..93794a85f83d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.c
@@ -211,7 +211,8 @@ void amdgpu_ring_priority_get(struct amdgpu_ring *ring,
 	if (!ring->funcs->set_priority)
 		return;
 
-	atomic_inc(&ring->num_jobs[priority]);
+	if (atomic_inc_return(&ring->num_jobs[priority]) <= 0)
+		return;
 
 	mutex_lock(&ring->priority_mutex);
 	if (priority <= ring->priority)
@@ -304,7 +305,7 @@ int amdgpu_ring_init(struct amdgpu_device *adev, struct amdgpu_ring *ring,
 		0xffffffffffffffff : ring->buf_mask;
 	/* Allocate ring buffer */
 	if (ring->ring_obj == NULL) {
-		r = amdgpu_bo_create_kernel(adev, ring->ring_size, PAGE_SIZE,
+		r = amdgpu_bo_create_kernel(adev, ring->ring_size + ring->funcs->extra_dw, PAGE_SIZE,
 					    AMDGPU_GEM_DOMAIN_GTT,
 					    &ring->ring_obj,
 					    &ring->gpu_addr,
@@ -362,6 +363,7 @@ void amdgpu_ring_fini(struct amdgpu_ring *ring)
 
 	dma_fence_put(ring->vmid_wait);
 	ring->vmid_wait = NULL;
+	ring->me = 0;
 
 	ring->adev->rings[ring->idx] = NULL;
 }
@@ -459,6 +461,26 @@ void amdgpu_ring_lru_touch(struct amdgpu_device *adev, struct amdgpu_ring *ring)
 	spin_unlock(&adev->ring_lru_list_lock);
 }
 
+/**
+ * amdgpu_ring_emit_reg_write_reg_wait_helper - ring helper
+ *
+ * @adev: amdgpu_device pointer
+ * @reg0: register to write
+ * @reg1: register to wait on
+ * @ref: reference value to write/wait on
+ * @mask: mask to wait on
+ *
+ * Helper for rings that don't support write and wait in a
+ * single oneshot packet.
+ */
+void amdgpu_ring_emit_reg_write_reg_wait_helper(struct amdgpu_ring *ring,
+						uint32_t reg0, uint32_t reg1,
+						uint32_t ref, uint32_t mask)
+{
+	amdgpu_ring_emit_wreg(ring, reg0, ref);
+	amdgpu_ring_emit_reg_wait(ring, reg1, mask, mask);
+}
+
 /*
  * Debugfs info
  */
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
index 1a5911882657..5018c0b6bf1a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h
@@ -29,7 +29,7 @@
 #include <drm/drm_print.h>
 
 /* max number of rings */
-#define AMDGPU_MAX_RINGS		18
+#define AMDGPU_MAX_RINGS		21
 #define AMDGPU_MAX_GFX_RINGS		1
 #define AMDGPU_MAX_COMPUTE_RINGS	8
 #define AMDGPU_MAX_VCE_RINGS		3
@@ -42,6 +42,9 @@
 
 #define AMDGPU_FENCE_FLAG_64BIT         (1 << 0)
 #define AMDGPU_FENCE_FLAG_INT           (1 << 1)
+#define AMDGPU_FENCE_FLAG_TC_WB_ONLY    (1 << 2)
+
+#define to_amdgpu_ring(s) container_of((s), struct amdgpu_ring, sched)
 
 enum amdgpu_ring_type {
 	AMDGPU_RING_TYPE_GFX,
@@ -52,7 +55,8 @@ enum amdgpu_ring_type {
 	AMDGPU_RING_TYPE_KIQ,
 	AMDGPU_RING_TYPE_UVD_ENC,
 	AMDGPU_RING_TYPE_VCN_DEC,
-	AMDGPU_RING_TYPE_VCN_ENC
+	AMDGPU_RING_TYPE_VCN_ENC,
+	AMDGPU_RING_TYPE_VCN_JPEG
 };
 
 struct amdgpu_device;
@@ -90,7 +94,8 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring,
 				   unsigned irq_type);
 void amdgpu_fence_driver_suspend(struct amdgpu_device *adev);
 void amdgpu_fence_driver_resume(struct amdgpu_device *adev);
-int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **fence);
+int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **fence,
+		      unsigned flags);
 int amdgpu_fence_emit_polling(struct amdgpu_ring *ring, uint32_t *s);
 void amdgpu_fence_process(struct amdgpu_ring *ring);
 int amdgpu_fence_wait_empty(struct amdgpu_ring *ring);
@@ -110,6 +115,7 @@ struct amdgpu_ring_funcs {
 	u32			nop;
 	bool			support_64bit_ptrs;
 	unsigned		vmhub;
+	unsigned		extra_dw;
 
 	/* ring read/write ptr handling */
 	u64 (*get_rptr)(struct amdgpu_ring *ring);
@@ -154,6 +160,9 @@ struct amdgpu_ring_funcs {
 	void (*emit_wreg)(struct amdgpu_ring *ring, uint32_t reg, uint32_t val);
 	void (*emit_reg_wait)(struct amdgpu_ring *ring, uint32_t reg,
 			      uint32_t val, uint32_t mask);
+	void (*emit_reg_write_reg_wait)(struct amdgpu_ring *ring,
+					uint32_t reg0, uint32_t reg1,
+					uint32_t ref, uint32_t mask);
 	void (*emit_tmz)(struct amdgpu_ring *ring, bool start);
 	/* priority functions */
 	void (*set_priority) (struct amdgpu_ring *ring,
@@ -228,6 +237,10 @@ int amdgpu_ring_lru_get(struct amdgpu_device *adev, int type,
 			int *blacklist, int num_blacklist,
 			bool lru_pipe_order, struct amdgpu_ring **ring);
 void amdgpu_ring_lru_touch(struct amdgpu_device *adev, struct amdgpu_ring *ring);
+void amdgpu_ring_emit_reg_write_reg_wait_helper(struct amdgpu_ring *ring,
+						uint32_t reg0, uint32_t val0,
+						uint32_t reg1, uint32_t val1);
+
 static inline void amdgpu_ring_clear_ring(struct amdgpu_ring *ring)
 {
 	int i = 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c
index 2dbe87591f81..8904e62dca7a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_test.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /*
  * Copyright 2009 VMware, Inc.
  *
@@ -33,6 +34,7 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev)
 	struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring;
 	struct amdgpu_bo *vram_obj = NULL;
 	struct amdgpu_bo **gtt_obj = NULL;
+	struct amdgpu_bo_param bp;
 	uint64_t gart_addr, vram_addr;
 	unsigned n, size;
 	int i, r;
@@ -52,15 +54,21 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev)
 		n -= adev->irq.ih.ring_size;
 	n /= size;
 
-	gtt_obj = kzalloc(n * sizeof(*gtt_obj), GFP_KERNEL);
+	gtt_obj = kcalloc(n, sizeof(*gtt_obj), GFP_KERNEL);
 	if (!gtt_obj) {
 		DRM_ERROR("Failed to allocate %d pointers\n", n);
 		r = 1;
 		goto out_cleanup;
 	}
-
-	r = amdgpu_bo_create(adev, size, PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM, 0,
-			     ttm_bo_type_kernel, NULL, &vram_obj);
+	memset(&bp, 0, sizeof(bp));
+	bp.size = size;
+	bp.byte_align = PAGE_SIZE;
+	bp.domain = AMDGPU_GEM_DOMAIN_VRAM;
+	bp.flags = 0;
+	bp.type = ttm_bo_type_kernel;
+	bp.resv = NULL;
+
+	r = amdgpu_bo_create(adev, &bp, &vram_obj);
 	if (r) {
 		DRM_ERROR("Failed to create VRAM object\n");
 		goto out_cleanup;
@@ -68,20 +76,20 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev)
 	r = amdgpu_bo_reserve(vram_obj, false);
 	if (unlikely(r != 0))
 		goto out_unref;
-	r = amdgpu_bo_pin(vram_obj, AMDGPU_GEM_DOMAIN_VRAM, &vram_addr);
+	r = amdgpu_bo_pin(vram_obj, AMDGPU_GEM_DOMAIN_VRAM);
 	if (r) {
 		DRM_ERROR("Failed to pin VRAM object\n");
 		goto out_unres;
 	}
+	vram_addr = amdgpu_bo_gpu_offset(vram_obj);
 	for (i = 0; i < n; i++) {
 		void *gtt_map, *vram_map;
 		void **gart_start, **gart_end;
 		void **vram_start, **vram_end;
 		struct dma_fence *fence = NULL;
 
-		r = amdgpu_bo_create(adev, size, PAGE_SIZE,
-				     AMDGPU_GEM_DOMAIN_GTT, 0,
-				     ttm_bo_type_kernel, NULL, gtt_obj + i);
+		bp.domain = AMDGPU_GEM_DOMAIN_GTT;
+		r = amdgpu_bo_create(adev, &bp, gtt_obj + i);
 		if (r) {
 			DRM_ERROR("Failed to create GTT object %d\n", i);
 			goto out_lclean;
@@ -90,11 +98,17 @@ static void amdgpu_do_test_moves(struct amdgpu_device *adev)
 		r = amdgpu_bo_reserve(gtt_obj[i], false);
 		if (unlikely(r != 0))
 			goto out_lclean_unref;
-		r = amdgpu_bo_pin(gtt_obj[i], AMDGPU_GEM_DOMAIN_GTT, &gart_addr);
+		r = amdgpu_bo_pin(gtt_obj[i], AMDGPU_GEM_DOMAIN_GTT);
 		if (r) {
 			DRM_ERROR("Failed to pin GTT object %d\n", i);
 			goto out_lclean_unres;
 		}
+		r = amdgpu_ttm_alloc_gart(&gtt_obj[i]->tbo);
+		if (r) {
+			DRM_ERROR("%p bind failed\n", gtt_obj[i]);
+			goto out_lclean_unpin;
+		}
+		gart_addr = amdgpu_bo_gpu_offset(gtt_obj[i]);
 
 		r = amdgpu_bo_kmap(gtt_obj[i], &gtt_map);
 		if (r) {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
index 532263ab6e16..76920035eb22 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_trace.h
@@ -150,10 +150,10 @@ TRACE_EVENT(amdgpu_cs,
 
 	    TP_fast_assign(
 			   __entry->bo_list = p->bo_list;
-			   __entry->ring = p->job->ring->idx;
+			   __entry->ring = p->ring->idx;
 			   __entry->dw = p->job->ibs[i].length_dw;
 			   __entry->fences = amdgpu_fence_count_emitted(
-				p->job->ring);
+				p->ring);
 			   ),
 	    TP_printk("bo_list=%p, ring=%u, dw=%u, fences=%u",
 		      __entry->bo_list, __entry->ring, __entry->dw,
@@ -178,7 +178,7 @@ TRACE_EVENT(amdgpu_cs_ioctl,
 			   __assign_str(timeline, AMDGPU_JOB_GET_TIMELINE_NAME(job))
 			   __entry->context = job->base.s_fence->finished.context;
 			   __entry->seqno = job->base.s_fence->finished.seqno;
-			   __entry->ring_name = job->ring->name;
+			   __entry->ring_name = to_amdgpu_ring(job->base.sched)->name;
 			   __entry->num_ibs = job->num_ibs;
 			   ),
 	    TP_printk("sched_job=%llu, timeline=%s, context=%u, seqno=%u, ring_name=%s, num_ibs=%u",
@@ -203,7 +203,7 @@ TRACE_EVENT(amdgpu_sched_run_job,
 			   __assign_str(timeline, AMDGPU_JOB_GET_TIMELINE_NAME(job))
 			   __entry->context = job->base.s_fence->finished.context;
 			   __entry->seqno = job->base.s_fence->finished.seqno;
-			   __entry->ring_name = job->ring->name;
+			   __entry->ring_name = to_amdgpu_ring(job->base.sched)->name;
 			   __entry->num_ibs = job->num_ibs;
 			   ),
 	    TP_printk("sched_job=%llu, timeline=%s, context=%u, seqno=%u, ring_name=%s, num_ibs=%u",
@@ -275,7 +275,7 @@ TRACE_EVENT(amdgpu_vm_bo_unmap,
 			     ),
 
 	    TP_fast_assign(
-			   __entry->bo = bo_va->base.bo;
+			   __entry->bo = bo_va ? bo_va->base.bo : NULL;
 			   __entry->start = mapping->start;
 			   __entry->last = mapping->last;
 			   __entry->offset = mapping->offset;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 205da3ff9cd0..13977ea6a097 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -63,24 +63,52 @@ static void amdgpu_ttm_debugfs_fini(struct amdgpu_device *adev);
 /*
  * Global memory.
  */
+
+/**
+ * amdgpu_ttm_mem_global_init - Initialize and acquire reference to
+ * memory object
+ *
+ * @ref: Object for initialization.
+ *
+ * This is called by drm_global_item_ref() when an object is being
+ * initialized.
+ */
 static int amdgpu_ttm_mem_global_init(struct drm_global_reference *ref)
 {
 	return ttm_mem_global_init(ref->object);
 }
 
+/**
+ * amdgpu_ttm_mem_global_release - Drop reference to a memory object
+ *
+ * @ref: Object being removed
+ *
+ * This is called by drm_global_item_unref() when an object is being
+ * released.
+ */
 static void amdgpu_ttm_mem_global_release(struct drm_global_reference *ref)
 {
 	ttm_mem_global_release(ref->object);
 }
 
+/**
+ * amdgpu_ttm_global_init - Initialize global TTM memory reference
+ * 							structures.
+ *
+ * @adev:  	AMDGPU device for which the global structures need to be
+ *			registered.
+ *
+ * This is called as part of the AMDGPU ttm init from amdgpu_ttm_init()
+ * during bring up.
+ */
 static int amdgpu_ttm_global_init(struct amdgpu_device *adev)
 {
 	struct drm_global_reference *global_ref;
-	struct amdgpu_ring *ring;
-	struct drm_sched_rq *rq;
 	int r;
 
+	/* ensure reference is false in case init fails */
 	adev->mman.mem_global_referenced = false;
+
 	global_ref = &adev->mman.mem_global_ref;
 	global_ref->global_type = DRM_GLOBAL_TTM_MEM;
 	global_ref->size = sizeof(struct ttm_mem_global);
@@ -108,21 +136,10 @@ static int amdgpu_ttm_global_init(struct amdgpu_device *adev)
 
 	mutex_init(&adev->mman.gtt_window_lock);
 
-	ring = adev->mman.buffer_funcs_ring;
-	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_KERNEL];
-	r = drm_sched_entity_init(&ring->sched, &adev->mman.entity,
-				  rq, amdgpu_sched_jobs, NULL);
-	if (r) {
-		DRM_ERROR("Failed setting up TTM BO move run queue.\n");
-		goto error_entity;
-	}
-
 	adev->mman.mem_global_referenced = true;
 
 	return 0;
 
-error_entity:
-	drm_global_item_unref(&adev->mman.bo_global_ref.ref);
 error_bo:
 	drm_global_item_unref(&adev->mman.mem_global_ref);
 error_mem:
@@ -132,8 +149,6 @@ error_mem:
 static void amdgpu_ttm_global_fini(struct amdgpu_device *adev)
 {
 	if (adev->mman.mem_global_referenced) {
-		drm_sched_entity_fini(adev->mman.entity.sched,
-				      &adev->mman.entity);
 		mutex_destroy(&adev->mman.gtt_window_lock);
 		drm_global_item_unref(&adev->mman.bo_global_ref.ref);
 		drm_global_item_unref(&adev->mman.mem_global_ref);
@@ -146,6 +161,18 @@ static int amdgpu_invalidate_caches(struct ttm_bo_device *bdev, uint32_t flags)
 	return 0;
 }
 
+/**
+ * amdgpu_init_mem_type - 	Initialize a memory manager for a specific
+ * 							type of memory request.
+ *
+ * @bdev:	The TTM BO device object (contains a reference to
+ * 			amdgpu_device)
+ * @type:	The type of memory requested
+ * @man:
+ *
+ * This is called by ttm_bo_init_mm() when a buffer object is being
+ * initialized.
+ */
 static int amdgpu_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 				struct ttm_mem_type_manager *man)
 {
@@ -161,6 +188,7 @@ static int amdgpu_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 		man->default_caching = TTM_PL_FLAG_CACHED;
 		break;
 	case TTM_PL_TT:
+		/* GTT memory  */
 		man->func = &amdgpu_gtt_mgr_func;
 		man->gpu_offset = adev->gmc.gart_start;
 		man->available_caching = TTM_PL_MASK_CACHING;
@@ -193,6 +221,14 @@ static int amdgpu_init_mem_type(struct ttm_bo_device *bdev, uint32_t type,
 	return 0;
 }
 
+/**
+ * amdgpu_evict_flags - Compute placement flags
+ *
+ * @bo: The buffer object to evict
+ * @placement: Possible destination(s) for evicted BO
+ *
+ * Fill in placement data when ttm_bo_evict() is called
+ */
 static void amdgpu_evict_flags(struct ttm_buffer_object *bo,
 				struct ttm_placement *placement)
 {
@@ -204,12 +240,14 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo,
 		.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_SYSTEM
 	};
 
+	/* Don't handle scatter gather BOs */
 	if (bo->type == ttm_bo_type_sg) {
 		placement->num_placement = 0;
 		placement->num_busy_placement = 0;
 		return;
 	}
 
+	/* Object isn't an AMDGPU object so ignore */
 	if (!amdgpu_ttm_bo_is_amdgpu_bo(bo)) {
 		placement->placement = &placements;
 		placement->busy_placement = &placements;
@@ -217,26 +255,16 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo,
 		placement->num_busy_placement = 1;
 		return;
 	}
+
 	abo = ttm_to_amdgpu_bo(bo);
 	switch (bo->mem.mem_type) {
 	case TTM_PL_VRAM:
 		if (!adev->mman.buffer_funcs_enabled) {
+			/* Move to system memory */
 			amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_CPU);
-		} else if (adev->gmc.visible_vram_size < adev->gmc.real_vram_size &&
-			   !(abo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED)) {
-			unsigned fpfn = adev->gmc.visible_vram_size >> PAGE_SHIFT;
-			struct drm_mm_node *node = bo->mem.mm_node;
-			unsigned long pages_left;
-
-			for (pages_left = bo->mem.num_pages;
-			     pages_left;
-			     pages_left -= node->size, node++) {
-				if (node->start < fpfn)
-					break;
-			}
-
-			if (!pages_left)
-				goto gtt;
+		} else if (!amdgpu_gmc_vram_full_visible(&adev->gmc) &&
+			   !(abo->flags & AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED) &&
+			   amdgpu_bo_in_cpu_visible_vram(abo)) {
 
 			/* Try evicting to the CPU inaccessible part of VRAM
 			 * first, but only set GTT as busy placement, so this
@@ -245,12 +273,12 @@ static void amdgpu_evict_flags(struct ttm_buffer_object *bo,
 			 */
 			amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_VRAM |
 							 AMDGPU_GEM_DOMAIN_GTT);
-			abo->placements[0].fpfn = fpfn;
+			abo->placements[0].fpfn = adev->gmc.visible_vram_size >> PAGE_SHIFT;
 			abo->placements[0].lpfn = 0;
 			abo->placement.busy_placement = &abo->placements[1];
 			abo->placement.num_busy_placement = 1;
 		} else {
-gtt:
+			/* Move to GTT memory */
 			amdgpu_ttm_placement_from_domain(abo, AMDGPU_GEM_DOMAIN_GTT);
 		}
 		break;
@@ -261,6 +289,15 @@ gtt:
 	*placement = abo->placement;
 }
 
+/**
+ * amdgpu_verify_access - Verify access for a mmap call
+ *
+ * @bo:		The buffer object to map
+ * @filp:	The file pointer from the process performing the mmap
+ *
+ * This is called by ttm_bo_mmap() to verify whether a process
+ * has the right to mmap a BO to their process space.
+ */
 static int amdgpu_verify_access(struct ttm_buffer_object *bo, struct file *filp)
 {
 	struct amdgpu_bo *abo = ttm_to_amdgpu_bo(bo);
@@ -278,6 +315,15 @@ static int amdgpu_verify_access(struct ttm_buffer_object *bo, struct file *filp)
 					  filp->private_data);
 }
 
+/**
+ * amdgpu_move_null - Register memory for a buffer object
+ *
+ * @bo:			The bo to assign the memory to
+ * @new_mem:	The memory to be assigned.
+ *
+ * Assign the memory from new_mem to the memory of the buffer object
+ * bo.
+ */
 static void amdgpu_move_null(struct ttm_buffer_object *bo,
 			     struct ttm_mem_reg *new_mem)
 {
@@ -288,6 +334,10 @@ static void amdgpu_move_null(struct ttm_buffer_object *bo,
 	new_mem->mm_node = NULL;
 }
 
+/**
+ * amdgpu_mm_node_addr -	Compute the GPU relative offset of a GTT
+ * 							buffer.
+ */
 static uint64_t amdgpu_mm_node_addr(struct ttm_buffer_object *bo,
 				    struct drm_mm_node *mm_node,
 				    struct ttm_mem_reg *mem)
@@ -302,9 +352,10 @@ static uint64_t amdgpu_mm_node_addr(struct ttm_buffer_object *bo,
 }
 
 /**
- * amdgpu_find_mm_node - Helper function finds the drm_mm_node
- *  corresponding to @offset. It also modifies the offset to be
- *  within the drm_mm_node returned
+ * amdgpu_find_mm_node -	Helper function finds the drm_mm_node
+ *  						corresponding to @offset. It also modifies
+ * 							the offset to be within the drm_mm_node
+ * 							returned
  */
 static struct drm_mm_node *amdgpu_find_mm_node(struct ttm_mem_reg *mem,
 					       unsigned long *offset)
@@ -443,7 +494,12 @@ error:
 	return r;
 }
 
-
+/**
+ * amdgpu_move_blit - Copy an entire buffer to another buffer
+ *
+ * This is a helper called by amdgpu_bo_move() and
+ * amdgpu_move_vram_ram() to help move buffers to and from VRAM.
+ */
 static int amdgpu_move_blit(struct ttm_buffer_object *bo,
 			    bool evict, bool no_wait_gpu,
 			    struct ttm_mem_reg *new_mem,
@@ -478,6 +534,11 @@ error:
 	return r;
 }
 
+/**
+ * amdgpu_move_vram_ram - Copy VRAM buffer to RAM buffer
+ *
+ * Called by amdgpu_bo_move().
+ */
 static int amdgpu_move_vram_ram(struct ttm_buffer_object *bo, bool evict,
 				struct ttm_operation_ctx *ctx,
 				struct ttm_mem_reg *new_mem)
@@ -490,6 +551,8 @@ static int amdgpu_move_vram_ram(struct ttm_buffer_object *bo, bool evict,
 	int r;
 
 	adev = amdgpu_ttm_adev(bo->bdev);
+
+	/* create space/pages for new_mem in GTT space */
 	tmp_mem = *new_mem;
 	tmp_mem.mm_node = NULL;
 	placement.num_placement = 1;
@@ -504,25 +567,36 @@ static int amdgpu_move_vram_ram(struct ttm_buffer_object *bo, bool evict,
 		return r;
 	}
 
+	/* set caching flags */
 	r = ttm_tt_set_placement_caching(bo->ttm, tmp_mem.placement);
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
 
+	/* Bind the memory to the GTT space */
 	r = ttm_tt_bind(bo->ttm, &tmp_mem, ctx);
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
+
+	/* blit VRAM to GTT */
 	r = amdgpu_move_blit(bo, true, ctx->no_wait_gpu, &tmp_mem, old_mem);
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
+
+	/* move BO (in tmp_mem) to new_mem */
 	r = ttm_bo_move_ttm(bo, ctx, new_mem);
 out_cleanup:
 	ttm_bo_mem_put(bo, &tmp_mem);
 	return r;
 }
 
+/**
+ * amdgpu_move_ram_vram - Copy buffer from RAM to VRAM
+ *
+ * Called by amdgpu_bo_move().
+ */
 static int amdgpu_move_ram_vram(struct ttm_buffer_object *bo, bool evict,
 				struct ttm_operation_ctx *ctx,
 				struct ttm_mem_reg *new_mem)
@@ -535,6 +609,8 @@ static int amdgpu_move_ram_vram(struct ttm_buffer_object *bo, bool evict,
 	int r;
 
 	adev = amdgpu_ttm_adev(bo->bdev);
+
+	/* make space in GTT for old_mem buffer */
 	tmp_mem = *new_mem;
 	tmp_mem.mm_node = NULL;
 	placement.num_placement = 1;
@@ -548,10 +624,14 @@ static int amdgpu_move_ram_vram(struct ttm_buffer_object *bo, bool evict,
 	if (unlikely(r)) {
 		return r;
 	}
+
+	/* move/bind old memory to GTT space */
 	r = ttm_bo_move_ttm(bo, ctx, &tmp_mem);
 	if (unlikely(r)) {
 		goto out_cleanup;
 	}
+
+	/* copy to VRAM */
 	r = amdgpu_move_blit(bo, true, ctx->no_wait_gpu, new_mem, old_mem);
 	if (unlikely(r)) {
 		goto out_cleanup;
@@ -561,6 +641,11 @@ out_cleanup:
 	return r;
 }
 
+/**
+ * amdgpu_bo_move - Move a buffer object to a new memory location
+ *
+ * Called by ttm_bo_handle_move_mem()
+ */
 static int amdgpu_bo_move(struct ttm_buffer_object *bo, bool evict,
 			  struct ttm_operation_ctx *ctx,
 			  struct ttm_mem_reg *new_mem)
@@ -626,6 +711,11 @@ memcpy:
 	return 0;
 }
 
+/**
+ * amdgpu_ttm_io_mem_reserve - Reserve a block of memory during a fault
+ *
+ * Called by ttm_mem_io_reserve() ultimately via ttm_bo_vm_fault()
+ */
 static int amdgpu_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem)
 {
 	struct ttm_mem_type_manager *man = &bdev->man[mem->mem_type];
@@ -695,7 +785,7 @@ struct amdgpu_ttm_tt {
 	struct ttm_dma_tt	ttm;
 	u64			offset;
 	uint64_t		userptr;
-	struct mm_struct	*usermm;
+	struct task_struct	*usertask;
 	uint32_t		userflags;
 	spinlock_t              guptasklock;
 	struct list_head        guptasks;
@@ -703,17 +793,29 @@ struct amdgpu_ttm_tt {
 	uint32_t		last_set_pages;
 };
 
+/**
+ * amdgpu_ttm_tt_get_user_pages - 	Pin pages of memory pointed to
+ * 									by a USERPTR pointer to memory
+ *
+ * Called by amdgpu_gem_userptr_ioctl() and amdgpu_cs_parser_bos().
+ * This provides a wrapper around the get_user_pages() call to provide
+ * device accessible pages that back user memory.
+ */
 int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
+	struct mm_struct *mm = gtt->usertask->mm;
 	unsigned int flags = 0;
 	unsigned pinned = 0;
 	int r;
 
+	if (!mm) /* Happens during process shutdown */
+		return -ESRCH;
+
 	if (!(gtt->userflags & AMDGPU_GEM_USERPTR_READONLY))
 		flags |= FOLL_WRITE;
 
-	down_read(&current->mm->mmap_sem);
+	down_read(&mm->mmap_sem);
 
 	if (gtt->userflags & AMDGPU_GEM_USERPTR_ANONONLY) {
 		/* check that we only use anonymous memory
@@ -721,13 +823,14 @@ int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages)
 		unsigned long end = gtt->userptr + ttm->num_pages * PAGE_SIZE;
 		struct vm_area_struct *vma;
 
-		vma = find_vma(gtt->usermm, gtt->userptr);
+		vma = find_vma(mm, gtt->userptr);
 		if (!vma || vma->vm_file || vma->vm_end < end) {
-			up_read(&current->mm->mmap_sem);
+			up_read(&mm->mmap_sem);
 			return -EPERM;
 		}
 	}
 
+	/* loop enough times using contiguous pages of memory */
 	do {
 		unsigned num_pages = ttm->num_pages - pinned;
 		uint64_t userptr = gtt->userptr + pinned * PAGE_SIZE;
@@ -739,7 +842,12 @@ int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages)
 		list_add(&guptask.list, &gtt->guptasks);
 		spin_unlock(&gtt->guptasklock);
 
-		r = get_user_pages(userptr, num_pages, flags, p, NULL);
+		if (mm == current->mm)
+			r = get_user_pages(userptr, num_pages, flags, p, NULL);
+		else
+			r = get_user_pages_remote(gtt->usertask,
+					mm, userptr, num_pages,
+					flags, p, NULL, NULL);
 
 		spin_lock(&gtt->guptasklock);
 		list_del(&guptask.list);
@@ -752,15 +860,23 @@ int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages)
 
 	} while (pinned < ttm->num_pages);
 
-	up_read(&current->mm->mmap_sem);
+	up_read(&mm->mmap_sem);
 	return 0;
 
 release_pages:
 	release_pages(pages, pinned);
-	up_read(&current->mm->mmap_sem);
+	up_read(&mm->mmap_sem);
 	return r;
 }
 
+/**
+ * amdgpu_ttm_tt_set_user_pages - 	Copy pages in, putting old pages
+ * 									as necessary.
+ *
+ * Called by amdgpu_cs_list_validate().  This creates the page list
+ * that backs user memory and will ultimately be mapped into the device
+ * address space.
+ */
 void amdgpu_ttm_tt_set_user_pages(struct ttm_tt *ttm, struct page **pages)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
@@ -775,6 +891,11 @@ void amdgpu_ttm_tt_set_user_pages(struct ttm_tt *ttm, struct page **pages)
 	}
 }
 
+/**
+ * amdgpu_ttm_tt_mark_user_page - Mark pages as dirty
+ *
+ * Called while unpinning userptr pages
+ */
 void amdgpu_ttm_tt_mark_user_pages(struct ttm_tt *ttm)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
@@ -793,7 +914,12 @@ void amdgpu_ttm_tt_mark_user_pages(struct ttm_tt *ttm)
 	}
 }
 
-/* prepare the sg table with the user pages */
+/**
+ * amdgpu_ttm_tt_pin_userptr - 	prepare the sg table with the
+ * 								user pages
+ *
+ * Called by amdgpu_ttm_backend_bind()
+ **/
 static int amdgpu_ttm_tt_pin_userptr(struct ttm_tt *ttm)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(ttm->bdev);
@@ -805,17 +931,20 @@ static int amdgpu_ttm_tt_pin_userptr(struct ttm_tt *ttm)
 	enum dma_data_direction direction = write ?
 		DMA_BIDIRECTIONAL : DMA_TO_DEVICE;
 
+	/* Allocate an SG array and squash pages into it */
 	r = sg_alloc_table_from_pages(ttm->sg, ttm->pages, ttm->num_pages, 0,
 				      ttm->num_pages << PAGE_SHIFT,
 				      GFP_KERNEL);
 	if (r)
 		goto release_sg;
 
+	/* Map SG to device */
 	r = -ENOMEM;
 	nents = dma_map_sg(adev->dev, ttm->sg->sgl, ttm->sg->nents, direction);
 	if (nents != ttm->sg->nents)
 		goto release_sg;
 
+	/* convert SG to linear array of pages and dma addresses */
 	drm_prime_sg_to_page_addr_arrays(ttm->sg, ttm->pages,
 					 gtt->ttm.dma_address, ttm->num_pages);
 
@@ -826,6 +955,9 @@ release_sg:
 	return r;
 }
 
+/**
+ * amdgpu_ttm_tt_unpin_userptr - Unpin and unmap userptr pages
+ */
 static void amdgpu_ttm_tt_unpin_userptr(struct ttm_tt *ttm)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(ttm->bdev);
@@ -839,14 +971,60 @@ static void amdgpu_ttm_tt_unpin_userptr(struct ttm_tt *ttm)
 	if (!ttm->sg->sgl)
 		return;
 
-	/* free the sg table and pages again */
+	/* unmap the pages mapped to the device */
 	dma_unmap_sg(adev->dev, ttm->sg->sgl, ttm->sg->nents, direction);
 
+	/* mark the pages as dirty */
 	amdgpu_ttm_tt_mark_user_pages(ttm);
 
 	sg_free_table(ttm->sg);
 }
 
+int amdgpu_ttm_gart_bind(struct amdgpu_device *adev,
+				struct ttm_buffer_object *tbo,
+				uint64_t flags)
+{
+	struct amdgpu_bo *abo = ttm_to_amdgpu_bo(tbo);
+	struct ttm_tt *ttm = tbo->ttm;
+	struct amdgpu_ttm_tt *gtt = (void *)ttm;
+	int r;
+
+	if (abo->flags & AMDGPU_GEM_CREATE_MQD_GFX9) {
+		uint64_t page_idx = 1;
+
+		r = amdgpu_gart_bind(adev, gtt->offset, page_idx,
+				ttm->pages, gtt->ttm.dma_address, flags);
+		if (r)
+			goto gart_bind_fail;
+
+		/* Patch mtype of the second part BO */
+		flags &=  ~AMDGPU_PTE_MTYPE_MASK;
+		flags |= AMDGPU_PTE_MTYPE(AMDGPU_MTYPE_NC);
+
+		r = amdgpu_gart_bind(adev,
+				gtt->offset + (page_idx << PAGE_SHIFT),
+				ttm->num_pages - page_idx,
+				&ttm->pages[page_idx],
+				&(gtt->ttm.dma_address[page_idx]), flags);
+	} else {
+		r = amdgpu_gart_bind(adev, gtt->offset, ttm->num_pages,
+				     ttm->pages, gtt->ttm.dma_address, flags);
+	}
+
+gart_bind_fail:
+	if (r)
+		DRM_ERROR("failed to bind %lu pages at 0x%08llX\n",
+			  ttm->num_pages, gtt->offset);
+
+	return r;
+}
+
+/**
+ * amdgpu_ttm_backend_bind - Bind GTT memory
+ *
+ * Called by ttm_tt_bind() on behalf of ttm_bo_handle_move_mem().
+ * This handles binding GTT memory to the device address space.
+ */
 static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm,
 				   struct ttm_mem_reg *bo_mem)
 {
@@ -877,7 +1055,10 @@ static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm,
 		return 0;
 	}
 
+	/* compute PTE flags relevant to this BO memory */
 	flags = amdgpu_ttm_tt_pte_flags(adev, ttm, bo_mem);
+
+	/* bind pages into GART page tables */
 	gtt->offset = (u64)bo_mem->start << PAGE_SHIFT;
 	r = amdgpu_gart_bind(adev, gtt->offset, ttm->num_pages,
 		ttm->pages, gtt->ttm.dma_address, flags);
@@ -888,6 +1069,9 @@ static int amdgpu_ttm_backend_bind(struct ttm_tt *ttm,
 	return r;
 }
 
+/**
+ * amdgpu_ttm_alloc_gart - Allocate GART memory for buffer object
+ */
 int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->bdev);
@@ -903,6 +1087,7 @@ int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo)
 	    amdgpu_gtt_mgr_has_gart_addr(&bo->mem))
 		return 0;
 
+	/* allocate GTT space */
 	tmp = bo->mem;
 	tmp.mm_node = NULL;
 	placement.num_placement = 1;
@@ -918,10 +1103,12 @@ int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo)
 	if (unlikely(r))
 		return r;
 
+	/* compute PTE flags for this buffer object */
 	flags = amdgpu_ttm_tt_pte_flags(adev, bo->ttm, &tmp);
+
+	/* Bind pages */
 	gtt->offset = (u64)tmp.start << PAGE_SHIFT;
-	r = amdgpu_gart_bind(adev, gtt->offset, bo->ttm->num_pages,
-			     bo->ttm->pages, gtt->ttm.dma_address, flags);
+	r = amdgpu_ttm_gart_bind(adev, bo, flags);
 	if (unlikely(r)) {
 		ttm_bo_mem_put(bo, &tmp);
 		return r;
@@ -935,31 +1122,40 @@ int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo)
 	return 0;
 }
 
+/**
+ * amdgpu_ttm_recover_gart - Rebind GTT pages
+ *
+ * Called by amdgpu_gtt_mgr_recover() from amdgpu_device_reset() to
+ * rebind GTT pages during a GPU reset.
+ */
 int amdgpu_ttm_recover_gart(struct ttm_buffer_object *tbo)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(tbo->bdev);
-	struct amdgpu_ttm_tt *gtt = (void *)tbo->ttm;
 	uint64_t flags;
 	int r;
 
-	if (!gtt)
+	if (!tbo->ttm)
 		return 0;
 
-	flags = amdgpu_ttm_tt_pte_flags(adev, &gtt->ttm.ttm, &tbo->mem);
-	r = amdgpu_gart_bind(adev, gtt->offset, gtt->ttm.ttm.num_pages,
-			     gtt->ttm.ttm.pages, gtt->ttm.dma_address, flags);
-	if (r)
-		DRM_ERROR("failed to bind %lu pages at 0x%08llX\n",
-			  gtt->ttm.ttm.num_pages, gtt->offset);
+	flags = amdgpu_ttm_tt_pte_flags(adev, tbo->ttm, &tbo->mem);
+	r = amdgpu_ttm_gart_bind(adev, tbo, flags);
+
 	return r;
 }
 
+/**
+ * amdgpu_ttm_backend_unbind - Unbind GTT mapped pages
+ *
+ * Called by ttm_tt_unbind() on behalf of ttm_bo_move_ttm() and
+ * ttm_tt_destroy().
+ */
 static int amdgpu_ttm_backend_unbind(struct ttm_tt *ttm)
 {
 	struct amdgpu_device *adev = amdgpu_ttm_adev(ttm->bdev);
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
 	int r;
 
+	/* if the pages have userptr pinning then clear that first */
 	if (gtt->userptr)
 		amdgpu_ttm_tt_unpin_userptr(ttm);
 
@@ -978,6 +1174,9 @@ static void amdgpu_ttm_backend_destroy(struct ttm_tt *ttm)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
 
+	if (gtt->usertask)
+		put_task_struct(gtt->usertask);
+
 	ttm_dma_tt_fini(&gtt->ttm);
 	kfree(gtt);
 }
@@ -988,6 +1187,13 @@ static struct ttm_backend_func amdgpu_backend_func = {
 	.destroy = &amdgpu_ttm_backend_destroy,
 };
 
+/**
+ * amdgpu_ttm_tt_create - Create a ttm_tt object for a given BO
+ *
+ * @bo: The buffer object to create a GTT ttm_tt object around
+ *
+ * Called by ttm_tt_create().
+ */
 static struct ttm_tt *amdgpu_ttm_tt_create(struct ttm_buffer_object *bo,
 					   uint32_t page_flags)
 {
@@ -1001,6 +1207,8 @@ static struct ttm_tt *amdgpu_ttm_tt_create(struct ttm_buffer_object *bo,
 		return NULL;
 	}
 	gtt->ttm.ttm.func = &amdgpu_backend_func;
+
+	/* allocate space for the uninitialized page entries */
 	if (ttm_sg_tt_init(&gtt->ttm, bo, page_flags)) {
 		kfree(gtt);
 		return NULL;
@@ -1008,6 +1216,12 @@ static struct ttm_tt *amdgpu_ttm_tt_create(struct ttm_buffer_object *bo,
 	return &gtt->ttm.ttm;
 }
 
+/**
+ * amdgpu_ttm_tt_populate - Map GTT pages visible to the device
+ *
+ * Map the pages of a ttm_tt object to an address space visible
+ * to the underlying device.
+ */
 static int amdgpu_ttm_tt_populate(struct ttm_tt *ttm,
 			struct ttm_operation_ctx *ctx)
 {
@@ -1015,6 +1229,7 @@ static int amdgpu_ttm_tt_populate(struct ttm_tt *ttm,
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
 	bool slave = !!(ttm->page_flags & TTM_PAGE_FLAG_SG);
 
+	/* user pages are bound by amdgpu_ttm_tt_pin_userptr() */
 	if (gtt && gtt->userptr) {
 		ttm->sg = kzalloc(sizeof(struct sg_table), GFP_KERNEL);
 		if (!ttm->sg)
@@ -1039,9 +1254,17 @@ static int amdgpu_ttm_tt_populate(struct ttm_tt *ttm,
 	}
 #endif
 
+	/* fall back to generic helper to populate the page array
+	 * and map them to the device */
 	return ttm_populate_and_map_pages(adev->dev, &gtt->ttm, ctx);
 }
 
+/**
+ * amdgpu_ttm_tt_unpopulate - unmap GTT pages and unpopulate page arrays
+ *
+ * Unmaps pages of a ttm_tt object from the device address space and
+ * unpopulates the page array backing it.
+ */
 static void amdgpu_ttm_tt_unpopulate(struct ttm_tt *ttm)
 {
 	struct amdgpu_device *adev;
@@ -1067,9 +1290,21 @@ static void amdgpu_ttm_tt_unpopulate(struct ttm_tt *ttm)
 	}
 #endif
 
+	/* fall back to generic helper to unmap and unpopulate array */
 	ttm_unmap_and_unpopulate_pages(adev->dev, &gtt->ttm);
 }
 
+/**
+ * amdgpu_ttm_tt_set_userptr -	Initialize userptr GTT ttm_tt
+ * 								for the current task
+ *
+ * @ttm: The ttm_tt object to bind this userptr object to
+ * @addr:  The address in the current tasks VM space to use
+ * @flags: Requirements of userptr object.
+ *
+ * Called by amdgpu_gem_userptr_ioctl() to bind userptr pages
+ * to current task
+ */
 int amdgpu_ttm_tt_set_userptr(struct ttm_tt *ttm, uint64_t addr,
 			      uint32_t flags)
 {
@@ -1079,8 +1314,13 @@ int amdgpu_ttm_tt_set_userptr(struct ttm_tt *ttm, uint64_t addr,
 		return -EINVAL;
 
 	gtt->userptr = addr;
-	gtt->usermm = current->mm;
 	gtt->userflags = flags;
+
+	if (gtt->usertask)
+		put_task_struct(gtt->usertask);
+	gtt->usertask = current->group_leader;
+	get_task_struct(gtt->usertask);
+
 	spin_lock_init(&gtt->guptasklock);
 	INIT_LIST_HEAD(&gtt->guptasks);
 	atomic_set(&gtt->mmu_invalidations, 0);
@@ -1089,6 +1329,9 @@ int amdgpu_ttm_tt_set_userptr(struct ttm_tt *ttm, uint64_t addr,
 	return 0;
 }
 
+/**
+ * amdgpu_ttm_tt_get_usermm - Return memory manager for ttm_tt object
+ */
 struct mm_struct *amdgpu_ttm_tt_get_usermm(struct ttm_tt *ttm)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
@@ -1096,9 +1339,18 @@ struct mm_struct *amdgpu_ttm_tt_get_usermm(struct ttm_tt *ttm)
 	if (gtt == NULL)
 		return NULL;
 
-	return gtt->usermm;
+	if (gtt->usertask == NULL)
+		return NULL;
+
+	return gtt->usertask->mm;
 }
 
+/**
+ * amdgpu_ttm_tt_affect_userptr -	Determine if a ttm_tt object lays
+ * 									inside an address range for the
+ * 									current task.
+ *
+ */
 bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start,
 				  unsigned long end)
 {
@@ -1109,10 +1361,16 @@ bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start,
 	if (gtt == NULL || !gtt->userptr)
 		return false;
 
+	/* Return false if no part of the ttm_tt object lies within
+	 * the range
+	 */
 	size = (unsigned long)gtt->ttm.ttm.num_pages * PAGE_SIZE;
 	if (gtt->userptr > end || gtt->userptr + size <= start)
 		return false;
 
+	/* Search the lists of tasks that hold this mapping and see
+	 * if current is one of them.  If it is return false.
+	 */
 	spin_lock(&gtt->guptasklock);
 	list_for_each_entry(entry, &gtt->guptasks, list) {
 		if (entry->task == current) {
@@ -1127,6 +1385,10 @@ bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start,
 	return true;
 }
 
+/**
+ * amdgpu_ttm_tt_userptr_invalidated -	Has the ttm_tt object been
+ * 										invalidated?
+ */
 bool amdgpu_ttm_tt_userptr_invalidated(struct ttm_tt *ttm,
 				       int *last_invalidated)
 {
@@ -1137,6 +1399,12 @@ bool amdgpu_ttm_tt_userptr_invalidated(struct ttm_tt *ttm,
 	return prev_invalidated != *last_invalidated;
 }
 
+/**
+ * amdgpu_ttm_tt_userptr_needs_pages -	Have the pages backing this
+ * 										ttm_tt object been invalidated
+ * 										since the last time they've
+ * 										been set?
+ */
 bool amdgpu_ttm_tt_userptr_needs_pages(struct ttm_tt *ttm)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
@@ -1147,6 +1415,9 @@ bool amdgpu_ttm_tt_userptr_needs_pages(struct ttm_tt *ttm)
 	return atomic_read(&gtt->mmu_invalidations) != gtt->last_set_pages;
 }
 
+/**
+ * amdgpu_ttm_tt_is_readonly - Is the ttm_tt object read only?
+ */
 bool amdgpu_ttm_tt_is_readonly(struct ttm_tt *ttm)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
@@ -1157,6 +1428,12 @@ bool amdgpu_ttm_tt_is_readonly(struct ttm_tt *ttm)
 	return !!(gtt->userflags & AMDGPU_GEM_USERPTR_READONLY);
 }
 
+/**
+ * amdgpu_ttm_tt_pte_flags - Compute PTE flags for ttm_tt object
+ *
+ * @ttm: The ttm_tt object to compute the flags for
+ * @mem: The memory registry backing this ttm_tt object
+ */
 uint64_t amdgpu_ttm_tt_pte_flags(struct amdgpu_device *adev, struct ttm_tt *ttm,
 				 struct ttm_mem_reg *mem)
 {
@@ -1181,6 +1458,16 @@ uint64_t amdgpu_ttm_tt_pte_flags(struct amdgpu_device *adev, struct ttm_tt *ttm,
 	return flags;
 }
 
+/**
+ * amdgpu_ttm_bo_eviction_valuable -	Check to see if we can evict
+ * 										a buffer object.
+ *
+ * Return true if eviction is sensible.  Called by
+ * ttm_mem_evict_first() on behalf of ttm_bo_mem_force_space()
+ * which tries to evict buffer objects until it can find space
+ * for a new object and by ttm_bo_force_list_clean() which is
+ * used to clean out a memory space.
+ */
 static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo,
 					    const struct ttm_place *place)
 {
@@ -1227,6 +1514,19 @@ static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo,
 	return ttm_bo_eviction_valuable(bo, place);
 }
 
+/**
+ * amdgpu_ttm_access_memory -	Read or Write memory that backs a
+ * 								buffer object.
+ *
+ * @bo:  The buffer object to read/write
+ * @offset:  Offset into buffer object
+ * @buf:  Secondary buffer to write/read from
+ * @len: Length in bytes of access
+ * @write:  true if writing
+ *
+ * This is used to access VRAM that backs a buffer object via MMIO
+ * access for debugging purposes.
+ */
 static int amdgpu_ttm_access_memory(struct ttm_buffer_object *bo,
 				    unsigned long offset,
 				    void *buf, int len, int write)
@@ -1329,6 +1629,7 @@ static void amdgpu_ttm_fw_reserve_vram_fini(struct amdgpu_device *adev)
 static int amdgpu_ttm_fw_reserve_vram_init(struct amdgpu_device *adev)
 {
 	struct ttm_operation_ctx ctx = { false, false };
+	struct amdgpu_bo_param bp;
 	int r = 0;
 	int i;
 	u64 vram_size = adev->gmc.visible_vram_size;
@@ -1336,17 +1637,21 @@ static int amdgpu_ttm_fw_reserve_vram_init(struct amdgpu_device *adev)
 	u64 size = adev->fw_vram_usage.size;
 	struct amdgpu_bo *bo;
 
+	memset(&bp, 0, sizeof(bp));
+	bp.size = adev->fw_vram_usage.size;
+	bp.byte_align = PAGE_SIZE;
+	bp.domain = AMDGPU_GEM_DOMAIN_VRAM;
+	bp.flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
+		AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS;
+	bp.type = ttm_bo_type_kernel;
+	bp.resv = NULL;
 	adev->fw_vram_usage.va = NULL;
 	adev->fw_vram_usage.reserved_bo = NULL;
 
 	if (adev->fw_vram_usage.size > 0 &&
 		adev->fw_vram_usage.size <= vram_size) {
 
-		r = amdgpu_bo_create(adev, adev->fw_vram_usage.size, PAGE_SIZE,
-				     AMDGPU_GEM_DOMAIN_VRAM,
-				     AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
-				     AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS,
-				     ttm_bo_type_kernel, NULL,
+		r = amdgpu_bo_create(adev, &bp,
 				     &adev->fw_vram_usage.reserved_bo);
 		if (r)
 			goto error_create;
@@ -1375,7 +1680,7 @@ static int amdgpu_ttm_fw_reserve_vram_init(struct amdgpu_device *adev)
 			AMDGPU_GEM_DOMAIN_VRAM,
 			adev->fw_vram_usage.start_offset,
 			(adev->fw_vram_usage.start_offset +
-			adev->fw_vram_usage.size), NULL);
+			adev->fw_vram_usage.size));
 		if (r)
 			goto error_pin;
 		r = amdgpu_bo_kmap(adev->fw_vram_usage.reserved_bo,
@@ -1398,13 +1703,22 @@ error_create:
 	adev->fw_vram_usage.reserved_bo = NULL;
 	return r;
 }
-
+/**
+ * amdgpu_ttm_init -	Init the memory management (ttm) as well as
+ * 						various gtt/vram related fields.
+ *
+ * This initializes all of the memory space pools that the TTM layer
+ * will need such as the GTT space (system memory mapped to the device),
+ * VRAM (on-board memory), and on-chip memories (GDS, GWS, OA) which
+ * can be mapped per VMID.
+ */
 int amdgpu_ttm_init(struct amdgpu_device *adev)
 {
 	uint64_t gtt_size;
 	int r;
 	u64 vis_vram_limit;
 
+	/* initialize global references for vram/gtt */
 	r = amdgpu_ttm_global_init(adev);
 	if (r) {
 		return r;
@@ -1425,6 +1739,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	/* We opt to avoid OOM on system pages allocations */
 	adev->mman.bdev.no_retry = true;
 
+	/* Initialize VRAM pool with all of VRAM divided into pages */
 	r = ttm_bo_init_mm(&adev->mman.bdev, TTM_PL_VRAM,
 				adev->gmc.real_vram_size >> PAGE_SHIFT);
 	if (r) {
@@ -1454,15 +1769,23 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 		return r;
 	}
 
-	r = amdgpu_bo_create_kernel(adev, adev->gmc.stolen_size, PAGE_SIZE,
-				    AMDGPU_GEM_DOMAIN_VRAM,
-				    &adev->stolen_vga_memory,
-				    NULL, NULL);
-	if (r)
-		return r;
+	/* allocate memory as required for VGA
+	 * This is used for VGA emulation and pre-OS scanout buffers to
+	 * avoid display artifacts while transitioning between pre-OS
+	 * and driver.  */
+	if (adev->gmc.stolen_size) {
+		r = amdgpu_bo_create_kernel(adev, adev->gmc.stolen_size, PAGE_SIZE,
+					    AMDGPU_GEM_DOMAIN_VRAM,
+					    &adev->stolen_vga_memory,
+					    NULL, NULL);
+		if (r)
+			return r;
+	}
 	DRM_INFO("amdgpu: %uM of VRAM memory ready\n",
 		 (unsigned) (adev->gmc.real_vram_size / (1024 * 1024)));
 
+	/* Compute GTT size, either bsaed on 3/4th the size of RAM size
+	 * or whatever the user passed on module init */
 	if (amdgpu_gtt_size == -1) {
 		struct sysinfo si;
 
@@ -1473,6 +1796,8 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	}
 	else
 		gtt_size = (uint64_t)amdgpu_gtt_size << 20;
+
+	/* Initialize GTT memory pool */
 	r = ttm_bo_init_mm(&adev->mman.bdev, TTM_PL_TT, gtt_size >> PAGE_SHIFT);
 	if (r) {
 		DRM_ERROR("Failed initializing GTT heap.\n");
@@ -1481,6 +1806,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	DRM_INFO("amdgpu: %uM of GTT memory ready.\n",
 		 (unsigned)(gtt_size / (1024 * 1024)));
 
+	/* Initialize various on-chip memory pools */
 	adev->gds.mem.total_size = adev->gds.mem.total_size << AMDGPU_GDS_SHIFT;
 	adev->gds.mem.gfx_partition_size = adev->gds.mem.gfx_partition_size << AMDGPU_GDS_SHIFT;
 	adev->gds.mem.cs_partition_size = adev->gds.mem.cs_partition_size << AMDGPU_GDS_SHIFT;
@@ -1520,6 +1846,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 		}
 	}
 
+	/* Register debugfs entries for amdgpu_ttm */
 	r = amdgpu_ttm_debugfs_init(adev);
 	if (r) {
 		DRM_ERROR("Failed to init debugfs\n");
@@ -1528,13 +1855,25 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
 	return 0;
 }
 
+/**
+ * amdgpu_ttm_late_init -	Handle any late initialization for
+ * 							amdgpu_ttm
+ */
+void amdgpu_ttm_late_init(struct amdgpu_device *adev)
+{
+	/* return the VGA stolen memory (if any) back to VRAM */
+	amdgpu_bo_free_kernel(&adev->stolen_vga_memory, NULL, NULL);
+}
+
+/**
+ * amdgpu_ttm_fini - De-initialize the TTM memory pools
+ */
 void amdgpu_ttm_fini(struct amdgpu_device *adev)
 {
 	if (!adev->mman.initialized)
 		return;
 
 	amdgpu_ttm_debugfs_fini(adev);
-	amdgpu_bo_free_kernel(&adev->stolen_vga_memory, NULL, NULL);
 	amdgpu_ttm_fw_reserve_vram_fini(adev);
 	if (adev->mman.aper_base_kaddr)
 		iounmap(adev->mman.aper_base_kaddr);
@@ -1567,10 +1906,29 @@ void amdgpu_ttm_set_buffer_funcs_status(struct amdgpu_device *adev, bool enable)
 {
 	struct ttm_mem_type_manager *man = &adev->mman.bdev.man[TTM_PL_VRAM];
 	uint64_t size;
+	int r;
 
-	if (!adev->mman.initialized || adev->in_gpu_reset)
+	if (!adev->mman.initialized || adev->in_gpu_reset ||
+	    adev->mman.buffer_funcs_enabled == enable)
 		return;
 
+	if (enable) {
+		struct amdgpu_ring *ring;
+		struct drm_sched_rq *rq;
+
+		ring = adev->mman.buffer_funcs_ring;
+		rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_KERNEL];
+		r = drm_sched_entity_init(&adev->mman.entity, &rq, 1, NULL);
+		if (r) {
+			DRM_ERROR("Failed setting up TTM BO move entity (%d)\n",
+				  r);
+			return;
+		}
+	} else {
+		drm_sched_entity_destroy(adev->mman.entity.sched,
+					 &adev->mman.entity);
+	}
+
 	/* this just adjusts TTM size idea, which sets lpfn to the correct value */
 	if (enable)
 		size = adev->gmc.real_vram_size;
@@ -1648,7 +2006,7 @@ static int amdgpu_map_buffer(struct ttm_buffer_object *bo,
 	if (r)
 		goto error_free;
 
-	r = amdgpu_job_submit(job, ring, &adev->mman.entity,
+	r = amdgpu_job_submit(job, &adev->mman.entity,
 			      AMDGPU_FENCE_OWNER_UNDEFINED, &fence);
 	if (r)
 		goto error_free;
@@ -1717,24 +2075,19 @@ int amdgpu_copy_buffer(struct amdgpu_ring *ring, uint64_t src_offset,
 
 	amdgpu_ring_pad_ib(ring, &job->ibs[0]);
 	WARN_ON(job->ibs[0].length_dw > num_dw);
-	if (direct_submit) {
-		r = amdgpu_ib_schedule(ring, job->num_ibs, job->ibs,
-				       NULL, fence);
-		job->fence = dma_fence_get(*fence);
-		if (r)
-			DRM_ERROR("Error scheduling IBs (%d)\n", r);
-		amdgpu_job_free(job);
-	} else {
-		r = amdgpu_job_submit(job, ring, &adev->mman.entity,
+	if (direct_submit)
+		r = amdgpu_job_submit_direct(job, ring, fence);
+	else
+		r = amdgpu_job_submit(job, &adev->mman.entity,
 				      AMDGPU_FENCE_OWNER_UNDEFINED, fence);
-		if (r)
-			goto error_free;
-	}
+	if (r)
+		goto error_free;
 
 	return r;
 
 error_free:
 	amdgpu_job_free(job);
+	DRM_ERROR("Error scheduling IBs (%d)\n", r);
 	return r;
 }
 
@@ -1817,7 +2170,7 @@ int amdgpu_fill_buffer(struct amdgpu_bo *bo,
 
 	amdgpu_ring_pad_ib(ring, &job->ibs[0]);
 	WARN_ON(job->ibs[0].length_dw > num_dw);
-	r = amdgpu_job_submit(job, ring, &adev->mman.entity,
+	r = amdgpu_job_submit(job, &adev->mman.entity,
 			      AMDGPU_FENCE_OWNER_UNDEFINED, fence);
 	if (r)
 		goto error_free;
@@ -1856,6 +2209,11 @@ static const struct drm_info_list amdgpu_ttm_debugfs_list[] = {
 #endif
 };
 
+/**
+ * amdgpu_ttm_vram_read - Linear read access to VRAM
+ *
+ * Accesses VRAM via MMIO for debugging purposes.
+ */
 static ssize_t amdgpu_ttm_vram_read(struct file *f, char __user *buf,
 				    size_t size, loff_t *pos)
 {
@@ -1895,6 +2253,11 @@ static ssize_t amdgpu_ttm_vram_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/**
+ * amdgpu_ttm_vram_write - Linear write access to VRAM
+ *
+ * Accesses VRAM via MMIO for debugging purposes.
+ */
 static ssize_t amdgpu_ttm_vram_write(struct file *f, const char __user *buf,
 				    size_t size, loff_t *pos)
 {
@@ -1943,6 +2306,9 @@ static const struct file_operations amdgpu_ttm_vram_fops = {
 
 #ifdef CONFIG_DRM_AMDGPU_GART_DEBUGFS
 
+/**
+ * amdgpu_ttm_gtt_read - Linear read access to GTT memory
+ */
 static ssize_t amdgpu_ttm_gtt_read(struct file *f, char __user *buf,
 				   size_t size, loff_t *pos)
 {
@@ -1990,6 +2356,13 @@ static const struct file_operations amdgpu_ttm_gtt_fops = {
 
 #endif
 
+/**
+ * amdgpu_iomem_read - Virtual read access to GPU mapped memory
+ *
+ * This function is used to read memory that has been mapped to the
+ * GPU and the known addresses are not physical addresses but instead
+ * bus addresses (e.g., what you'd put in an IB or ring buffer).
+ */
 static ssize_t amdgpu_iomem_read(struct file *f, char __user *buf,
 				 size_t size, loff_t *pos)
 {
@@ -1998,6 +2371,7 @@ static ssize_t amdgpu_iomem_read(struct file *f, char __user *buf,
 	ssize_t result = 0;
 	int r;
 
+	/* retrieve the IOMMU domain if any for this device */
 	dom = iommu_get_domain_for_dev(adev->dev);
 
 	while (size) {
@@ -2010,6 +2384,10 @@ static ssize_t amdgpu_iomem_read(struct file *f, char __user *buf,
 
 		bytes = bytes < size ? bytes : size;
 
+		/* Translate the bus address to a physical address.  If
+		 * the domain is NULL it means there is no IOMMU active
+		 * and the address translation is the identity
+		 */
 		addr = dom ? iommu_iova_to_phys(dom, addr) : addr;
 
 		pfn = addr >> PAGE_SHIFT;
@@ -2034,6 +2412,13 @@ static ssize_t amdgpu_iomem_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/**
+ * amdgpu_iomem_write - Virtual write access to GPU mapped memory
+ *
+ * This function is used to write memory that has been mapped to the
+ * GPU and the known addresses are not physical addresses but instead
+ * bus addresses (e.g., what you'd put in an IB or ring buffer).
+ */
 static ssize_t amdgpu_iomem_write(struct file *f, const char __user *buf,
 				 size_t size, loff_t *pos)
 {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
index 6ea7de863041..8b3cc6687769 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
@@ -73,10 +73,12 @@ bool amdgpu_gtt_mgr_has_gart_addr(struct ttm_mem_reg *mem);
 uint64_t amdgpu_gtt_mgr_usage(struct ttm_mem_type_manager *man);
 int amdgpu_gtt_mgr_recover(struct ttm_mem_type_manager *man);
 
+u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo);
 uint64_t amdgpu_vram_mgr_usage(struct ttm_mem_type_manager *man);
 uint64_t amdgpu_vram_mgr_vis_usage(struct ttm_mem_type_manager *man);
 
 int amdgpu_ttm_init(struct amdgpu_device *adev);
+void amdgpu_ttm_late_init(struct amdgpu_device *adev);
 void amdgpu_ttm_fini(struct amdgpu_device *adev);
 void amdgpu_ttm_set_buffer_funcs_status(struct amdgpu_device *adev,
 					bool enable);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
index 5916cc25e28b..f55f72a37ca8 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c
@@ -161,8 +161,38 @@ void amdgpu_ucode_print_rlc_hdr(const struct common_firmware_header *hdr)
 			  le32_to_cpu(rlc_hdr->reg_list_format_separate_array_offset_bytes));
 		DRM_DEBUG("reg_list_separate_size_bytes: %u\n",
 			  le32_to_cpu(rlc_hdr->reg_list_separate_size_bytes));
-		DRM_DEBUG("reg_list_separate_size_bytes: %u\n",
-			  le32_to_cpu(rlc_hdr->reg_list_separate_size_bytes));
+		DRM_DEBUG("reg_list_separate_array_offset_bytes: %u\n",
+			  le32_to_cpu(rlc_hdr->reg_list_separate_array_offset_bytes));
+		if (version_minor == 1) {
+			const struct rlc_firmware_header_v2_1 *v2_1 =
+				container_of(rlc_hdr, struct rlc_firmware_header_v2_1, v2_0);
+			DRM_DEBUG("reg_list_format_direct_reg_list_length: %u\n",
+				  le32_to_cpu(v2_1->reg_list_format_direct_reg_list_length));
+			DRM_DEBUG("save_restore_list_cntl_ucode_ver: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_cntl_ucode_ver));
+			DRM_DEBUG("save_restore_list_cntl_feature_ver: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_cntl_feature_ver));
+			DRM_DEBUG("save_restore_list_cntl_size_bytes %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_cntl_size_bytes));
+			DRM_DEBUG("save_restore_list_cntl_offset_bytes: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_cntl_offset_bytes));
+			DRM_DEBUG("save_restore_list_gpm_ucode_ver: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_gpm_ucode_ver));
+			DRM_DEBUG("save_restore_list_gpm_feature_ver: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_gpm_feature_ver));
+			DRM_DEBUG("save_restore_list_gpm_size_bytes %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_gpm_size_bytes));
+			DRM_DEBUG("save_restore_list_gpm_offset_bytes: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_gpm_offset_bytes));
+			DRM_DEBUG("save_restore_list_srm_ucode_ver: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_srm_ucode_ver));
+			DRM_DEBUG("save_restore_list_srm_feature_ver: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_srm_feature_ver));
+			DRM_DEBUG("save_restore_list_srm_size_bytes %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_srm_size_bytes));
+			DRM_DEBUG("save_restore_list_srm_offset_bytes: %u\n",
+				  le32_to_cpu(v2_1->save_restore_list_srm_offset_bytes));
+		}
 	} else {
 		DRM_ERROR("Unknown RLC ucode version: %u.%u\n", version_major, version_minor);
 	}
@@ -265,6 +295,7 @@ amdgpu_ucode_get_load_type(struct amdgpu_device *adev, int load_type)
 	case CHIP_POLARIS10:
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 		if (!load_type)
 			return AMDGPU_FW_LOAD_DIRECT;
 		else
@@ -276,6 +307,8 @@ amdgpu_ucode_get_load_type(struct amdgpu_device *adev, int load_type)
 			return AMDGPU_FW_LOAD_DIRECT;
 		else
 			return AMDGPU_FW_LOAD_PSP;
+	case CHIP_VEGA20:
+		return AMDGPU_FW_LOAD_DIRECT;
 	default:
 		DRM_ERROR("Unknown firmware load type\n");
 	}
@@ -307,7 +340,10 @@ static int amdgpu_ucode_init_single_fw(struct amdgpu_device *adev,
 	    (ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC1 &&
 	     ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC2 &&
 	     ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC1_JT &&
-	     ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC2_JT)) {
+	     ucode->ucode_id != AMDGPU_UCODE_ID_CP_MEC2_JT &&
+	     ucode->ucode_id != AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL &&
+	     ucode->ucode_id != AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM &&
+	     ucode->ucode_id != AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM)) {
 		ucode->ucode_size = le32_to_cpu(header->ucode_size_bytes);
 
 		memcpy(ucode->kaddr, (void *)((uint8_t *)ucode->fw->data +
@@ -329,6 +365,18 @@ static int amdgpu_ucode_init_single_fw(struct amdgpu_device *adev,
 					      le32_to_cpu(header->ucode_array_offset_bytes) +
 					      le32_to_cpu(cp_hdr->jt_offset) * 4),
 		       ucode->ucode_size);
+	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL) {
+		ucode->ucode_size = adev->gfx.rlc.save_restore_list_cntl_size_bytes;
+		memcpy(ucode->kaddr, adev->gfx.rlc.save_restore_list_cntl,
+		       ucode->ucode_size);
+	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM) {
+		ucode->ucode_size = adev->gfx.rlc.save_restore_list_gpm_size_bytes;
+		memcpy(ucode->kaddr, adev->gfx.rlc.save_restore_list_gpm,
+		       ucode->ucode_size);
+	} else if (ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM) {
+		ucode->ucode_size = adev->gfx.rlc.save_restore_list_srm_size_bytes;
+		memcpy(ucode->kaddr, adev->gfx.rlc.save_restore_list_srm,
+		       ucode->ucode_size);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h
index 30b5500dc152..08e38579af24 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h
@@ -98,6 +98,24 @@ struct rlc_firmware_header_v2_0 {
 	uint32_t reg_list_separate_array_offset_bytes; /* payload offset from the start of the header */
 };
 
+/* version_major=2, version_minor=1 */
+struct rlc_firmware_header_v2_1 {
+	struct rlc_firmware_header_v2_0 v2_0;
+	uint32_t reg_list_format_direct_reg_list_length; /* length of direct reg list format array */
+	uint32_t save_restore_list_cntl_ucode_ver;
+	uint32_t save_restore_list_cntl_feature_ver;
+	uint32_t save_restore_list_cntl_size_bytes;
+	uint32_t save_restore_list_cntl_offset_bytes;
+	uint32_t save_restore_list_gpm_ucode_ver;
+	uint32_t save_restore_list_gpm_feature_ver;
+	uint32_t save_restore_list_gpm_size_bytes;
+	uint32_t save_restore_list_gpm_offset_bytes;
+	uint32_t save_restore_list_srm_ucode_ver;
+	uint32_t save_restore_list_srm_feature_ver;
+	uint32_t save_restore_list_srm_size_bytes;
+	uint32_t save_restore_list_srm_offset_bytes;
+};
+
 /* version_major=1, version_minor=0 */
 struct sdma_firmware_header_v1_0 {
 	struct common_firmware_header header;
@@ -148,6 +166,7 @@ union amdgpu_firmware_header {
 	struct gfx_firmware_header_v1_0 gfx;
 	struct rlc_firmware_header_v1_0 rlc;
 	struct rlc_firmware_header_v2_0 rlc_v2_0;
+	struct rlc_firmware_header_v2_1 rlc_v2_1;
 	struct sdma_firmware_header_v1_0 sdma;
 	struct sdma_firmware_header_v1_1 sdma_v1_1;
 	struct gpu_info_firmware_header_v1_0 gpu_info;
@@ -168,6 +187,9 @@ enum AMDGPU_UCODE_ID {
 	AMDGPU_UCODE_ID_CP_MEC2,
 	AMDGPU_UCODE_ID_CP_MEC2_JT,
 	AMDGPU_UCODE_ID_RLC_G,
+	AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL,
+	AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM,
+	AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM,
 	AMDGPU_UCODE_ID_STORAGE,
 	AMDGPU_UCODE_ID_SMC,
 	AMDGPU_UCODE_ID_UVD,
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index 627542b22ae4..80b5c453f8c1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -53,11 +53,11 @@
 
 /* Firmware Names */
 #ifdef CONFIG_DRM_AMDGPU_CIK
-#define FIRMWARE_BONAIRE	"radeon/bonaire_uvd.bin"
-#define FIRMWARE_KABINI	"radeon/kabini_uvd.bin"
-#define FIRMWARE_KAVERI	"radeon/kaveri_uvd.bin"
-#define FIRMWARE_HAWAII	"radeon/hawaii_uvd.bin"
-#define FIRMWARE_MULLINS	"radeon/mullins_uvd.bin"
+#define FIRMWARE_BONAIRE	"amdgpu/bonaire_uvd.bin"
+#define FIRMWARE_KABINI	"amdgpu/kabini_uvd.bin"
+#define FIRMWARE_KAVERI	"amdgpu/kaveri_uvd.bin"
+#define FIRMWARE_HAWAII	"amdgpu/hawaii_uvd.bin"
+#define FIRMWARE_MULLINS	"amdgpu/mullins_uvd.bin"
 #endif
 #define FIRMWARE_TONGA		"amdgpu/tonga_uvd.bin"
 #define FIRMWARE_CARRIZO	"amdgpu/carrizo_uvd.bin"
@@ -66,15 +66,18 @@
 #define FIRMWARE_POLARIS10	"amdgpu/polaris10_uvd.bin"
 #define FIRMWARE_POLARIS11	"amdgpu/polaris11_uvd.bin"
 #define FIRMWARE_POLARIS12	"amdgpu/polaris12_uvd.bin"
+#define FIRMWARE_VEGAM		"amdgpu/vegam_uvd.bin"
 
 #define FIRMWARE_VEGA10		"amdgpu/vega10_uvd.bin"
 #define FIRMWARE_VEGA12		"amdgpu/vega12_uvd.bin"
+#define FIRMWARE_VEGA20		"amdgpu/vega20_uvd.bin"
 
-#define mmUVD_GPCOM_VCPU_DATA0_VEGA10 (0x03c4 + 0x7e00)
-#define mmUVD_GPCOM_VCPU_DATA1_VEGA10 (0x03c5 + 0x7e00)
-#define mmUVD_GPCOM_VCPU_CMD_VEGA10 (0x03c3 + 0x7e00)
-#define mmUVD_NO_OP_VEGA10 (0x03ff + 0x7e00)
-#define mmUVD_ENGINE_CNTL_VEGA10 (0x03c6 + 0x7e00)
+/* These are common relative offsets for all asics, from uvd_7_0_offset.h,  */
+#define UVD_GPCOM_VCPU_CMD		0x03c3
+#define UVD_GPCOM_VCPU_DATA0	0x03c4
+#define UVD_GPCOM_VCPU_DATA1	0x03c5
+#define UVD_NO_OP				0x03ff
+#define UVD_BASE_SI				0x3800
 
 /**
  * amdgpu_uvd_cs_ctx - Command submission parser context
@@ -109,9 +112,11 @@ MODULE_FIRMWARE(FIRMWARE_STONEY);
 MODULE_FIRMWARE(FIRMWARE_POLARIS10);
 MODULE_FIRMWARE(FIRMWARE_POLARIS11);
 MODULE_FIRMWARE(FIRMWARE_POLARIS12);
+MODULE_FIRMWARE(FIRMWARE_VEGAM);
 
 MODULE_FIRMWARE(FIRMWARE_VEGA10);
 MODULE_FIRMWARE(FIRMWARE_VEGA12);
+MODULE_FIRMWARE(FIRMWARE_VEGA20);
 
 static void amdgpu_uvd_idle_work_handler(struct work_struct *work);
 
@@ -122,8 +127,8 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 	unsigned long bo_size;
 	const char *fw_name;
 	const struct common_firmware_header *hdr;
-	unsigned version_major, version_minor, family_id;
-	int i, r;
+	unsigned family_id;
+	int i, j, r;
 
 	INIT_DELAYED_WORK(&adev->uvd.idle_work, amdgpu_uvd_idle_work_handler);
 
@@ -172,6 +177,12 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 	case CHIP_VEGA12:
 		fw_name = FIRMWARE_VEGA12;
 		break;
+	case CHIP_VEGAM:
+		fw_name = FIRMWARE_VEGAM;
+		break;
+	case CHIP_VEGA20:
+		fw_name = FIRMWARE_VEGA20;
+		break;
 	default:
 		return -EINVAL;
 	}
@@ -197,52 +208,70 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 
 	hdr = (const struct common_firmware_header *)adev->uvd.fw->data;
 	family_id = le32_to_cpu(hdr->ucode_version) & 0xff;
-	version_major = (le32_to_cpu(hdr->ucode_version) >> 24) & 0xff;
-	version_minor = (le32_to_cpu(hdr->ucode_version) >> 8) & 0xff;
-	DRM_INFO("Found UVD firmware Version: %hu.%hu Family ID: %hu\n",
-		version_major, version_minor, family_id);
-
-	/*
-	 * Limit the number of UVD handles depending on microcode major
-	 * and minor versions. The firmware version which has 40 UVD
-	 * instances support is 1.80. So all subsequent versions should
-	 * also have the same support.
-	 */
-	if ((version_major > 0x01) ||
-	    ((version_major == 0x01) && (version_minor >= 0x50)))
-		adev->uvd.max_handles = AMDGPU_MAX_UVD_HANDLES;
 
-	adev->uvd.fw_version = ((version_major << 24) | (version_minor << 16) |
-				(family_id << 8));
+	if (adev->asic_type < CHIP_VEGA20) {
+		unsigned version_major, version_minor;
+
+		version_major = (le32_to_cpu(hdr->ucode_version) >> 24) & 0xff;
+		version_minor = (le32_to_cpu(hdr->ucode_version) >> 8) & 0xff;
+		DRM_INFO("Found UVD firmware Version: %hu.%hu Family ID: %hu\n",
+			version_major, version_minor, family_id);
+
+		/*
+		 * Limit the number of UVD handles depending on microcode major
+		 * and minor versions. The firmware version which has 40 UVD
+		 * instances support is 1.80. So all subsequent versions should
+		 * also have the same support.
+		 */
+		if ((version_major > 0x01) ||
+		    ((version_major == 0x01) && (version_minor >= 0x50)))
+			adev->uvd.max_handles = AMDGPU_MAX_UVD_HANDLES;
+
+		adev->uvd.fw_version = ((version_major << 24) | (version_minor << 16) |
+					(family_id << 8));
+
+		if ((adev->asic_type == CHIP_POLARIS10 ||
+		     adev->asic_type == CHIP_POLARIS11) &&
+		    (adev->uvd.fw_version < FW_1_66_16))
+			DRM_ERROR("POLARIS10/11 UVD firmware version %hu.%hu is too old.\n",
+				  version_major, version_minor);
+	} else {
+		unsigned int enc_major, enc_minor, dec_minor;
+
+		dec_minor = (le32_to_cpu(hdr->ucode_version) >> 8) & 0xff;
+		enc_minor = (le32_to_cpu(hdr->ucode_version) >> 24) & 0x3f;
+		enc_major = (le32_to_cpu(hdr->ucode_version) >> 30) & 0x3;
+		DRM_INFO("Found UVD firmware ENC: %hu.%hu DEC: .%hu Family ID: %hu\n",
+			enc_major, enc_minor, dec_minor, family_id);
 
-	if ((adev->asic_type == CHIP_POLARIS10 ||
-	     adev->asic_type == CHIP_POLARIS11) &&
-	    (adev->uvd.fw_version < FW_1_66_16))
-		DRM_ERROR("POLARIS10/11 UVD firmware version %hu.%hu is too old.\n",
-			  version_major, version_minor);
+		adev->uvd.max_handles = AMDGPU_MAX_UVD_HANDLES;
+
+		adev->uvd.fw_version = le32_to_cpu(hdr->ucode_version);
+	}
 
 	bo_size = AMDGPU_UVD_STACK_SIZE + AMDGPU_UVD_HEAP_SIZE
 		  +  AMDGPU_UVD_SESSION_SIZE * adev->uvd.max_handles;
 	if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP)
 		bo_size += AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8);
 
-	r = amdgpu_bo_create_kernel(adev, bo_size, PAGE_SIZE,
-				    AMDGPU_GEM_DOMAIN_VRAM, &adev->uvd.vcpu_bo,
-				    &adev->uvd.gpu_addr, &adev->uvd.cpu_addr);
-	if (r) {
-		dev_err(adev->dev, "(%d) failed to allocate UVD bo\n", r);
-		return r;
+	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
+
+		r = amdgpu_bo_create_kernel(adev, bo_size, PAGE_SIZE,
+					    AMDGPU_GEM_DOMAIN_VRAM, &adev->uvd.inst[j].vcpu_bo,
+					    &adev->uvd.inst[j].gpu_addr, &adev->uvd.inst[j].cpu_addr);
+		if (r) {
+			dev_err(adev->dev, "(%d) failed to allocate UVD bo\n", r);
+			return r;
+		}
 	}
 
-	ring = &adev->uvd.ring;
+	ring = &adev->uvd.inst[0].ring;
 	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
-	r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity,
-				  rq, amdgpu_sched_jobs, NULL);
-	if (r != 0) {
-		DRM_ERROR("Failed setting up UVD run queue.\n");
+	r = drm_sched_entity_init(&adev->uvd.entity, &rq, 1, NULL);
+	if (r) {
+		DRM_ERROR("Failed setting up UVD kernel entity.\n");
 		return r;
 	}
-
 	for (i = 0; i < adev->uvd.max_handles; ++i) {
 		atomic_set(&adev->uvd.handles[i], 0);
 		adev->uvd.filp[i] = NULL;
@@ -274,20 +303,23 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
 
 int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
 {
-	int i;
-	kfree(adev->uvd.saved_bo);
+	int i, j;
 
-	drm_sched_entity_fini(&adev->uvd.ring.sched, &adev->uvd.entity);
+	drm_sched_entity_destroy(&adev->uvd.inst->ring.sched,
+				 &adev->uvd.entity);
 
-	amdgpu_bo_free_kernel(&adev->uvd.vcpu_bo,
-			      &adev->uvd.gpu_addr,
-			      (void **)&adev->uvd.cpu_addr);
+	for (j = 0; j < adev->uvd.num_uvd_inst; ++j) {
+		kfree(adev->uvd.inst[j].saved_bo);
 
-	amdgpu_ring_fini(&adev->uvd.ring);
+		amdgpu_bo_free_kernel(&adev->uvd.inst[j].vcpu_bo,
+				      &adev->uvd.inst[j].gpu_addr,
+				      (void **)&adev->uvd.inst[j].cpu_addr);
 
-	for (i = 0; i < AMDGPU_MAX_UVD_ENC_RINGS; ++i)
-		amdgpu_ring_fini(&adev->uvd.ring_enc[i]);
+		amdgpu_ring_fini(&adev->uvd.inst[j].ring);
 
+		for (i = 0; i < AMDGPU_MAX_UVD_ENC_RINGS; ++i)
+			amdgpu_ring_fini(&adev->uvd.inst[j].ring_enc[i]);
+	}
 	release_firmware(adev->uvd.fw);
 
 	return 0;
@@ -297,10 +329,7 @@ int amdgpu_uvd_suspend(struct amdgpu_device *adev)
 {
 	unsigned size;
 	void *ptr;
-	int i;
-
-	if (adev->uvd.vcpu_bo == NULL)
-		return 0;
+	int i, j;
 
 	cancel_delayed_work_sync(&adev->uvd.idle_work);
 
@@ -314,15 +343,19 @@ int amdgpu_uvd_suspend(struct amdgpu_device *adev)
 			return 0;
 	}
 
-	size = amdgpu_bo_size(adev->uvd.vcpu_bo);
-	ptr = adev->uvd.cpu_addr;
+	for (j = 0; j < adev->uvd.num_uvd_inst; ++j) {
+		if (adev->uvd.inst[j].vcpu_bo == NULL)
+			continue;
 
-	adev->uvd.saved_bo = kmalloc(size, GFP_KERNEL);
-	if (!adev->uvd.saved_bo)
-		return -ENOMEM;
+		size = amdgpu_bo_size(adev->uvd.inst[j].vcpu_bo);
+		ptr = adev->uvd.inst[j].cpu_addr;
 
-	memcpy_fromio(adev->uvd.saved_bo, ptr, size);
+		adev->uvd.inst[j].saved_bo = kmalloc(size, GFP_KERNEL);
+		if (!adev->uvd.inst[j].saved_bo)
+			return -ENOMEM;
 
+		memcpy_fromio(adev->uvd.inst[j].saved_bo, ptr, size);
+	}
 	return 0;
 }
 
@@ -330,51 +363,54 @@ int amdgpu_uvd_resume(struct amdgpu_device *adev)
 {
 	unsigned size;
 	void *ptr;
+	int i;
 
-	if (adev->uvd.vcpu_bo == NULL)
-		return -EINVAL;
+	for (i = 0; i < adev->uvd.num_uvd_inst; i++) {
+		if (adev->uvd.inst[i].vcpu_bo == NULL)
+			return -EINVAL;
 
-	size = amdgpu_bo_size(adev->uvd.vcpu_bo);
-	ptr = adev->uvd.cpu_addr;
+		size = amdgpu_bo_size(adev->uvd.inst[i].vcpu_bo);
+		ptr = adev->uvd.inst[i].cpu_addr;
 
-	if (adev->uvd.saved_bo != NULL) {
-		memcpy_toio(ptr, adev->uvd.saved_bo, size);
-		kfree(adev->uvd.saved_bo);
-		adev->uvd.saved_bo = NULL;
-	} else {
-		const struct common_firmware_header *hdr;
-		unsigned offset;
-
-		hdr = (const struct common_firmware_header *)adev->uvd.fw->data;
-		if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) {
-			offset = le32_to_cpu(hdr->ucode_array_offset_bytes);
-			memcpy_toio(adev->uvd.cpu_addr, adev->uvd.fw->data + offset,
-				    le32_to_cpu(hdr->ucode_size_bytes));
-			size -= le32_to_cpu(hdr->ucode_size_bytes);
-			ptr += le32_to_cpu(hdr->ucode_size_bytes);
+		if (adev->uvd.inst[i].saved_bo != NULL) {
+			memcpy_toio(ptr, adev->uvd.inst[i].saved_bo, size);
+			kfree(adev->uvd.inst[i].saved_bo);
+			adev->uvd.inst[i].saved_bo = NULL;
+		} else {
+			const struct common_firmware_header *hdr;
+			unsigned offset;
+
+			hdr = (const struct common_firmware_header *)adev->uvd.fw->data;
+			if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) {
+				offset = le32_to_cpu(hdr->ucode_array_offset_bytes);
+				memcpy_toio(adev->uvd.inst[i].cpu_addr, adev->uvd.fw->data + offset,
+					    le32_to_cpu(hdr->ucode_size_bytes));
+				size -= le32_to_cpu(hdr->ucode_size_bytes);
+				ptr += le32_to_cpu(hdr->ucode_size_bytes);
+			}
+			memset_io(ptr, 0, size);
+			/* to restore uvd fence seq */
+			amdgpu_fence_driver_force_completion(&adev->uvd.inst[i].ring);
 		}
-		memset_io(ptr, 0, size);
-		/* to restore uvd fence seq */
-		amdgpu_fence_driver_force_completion(&adev->uvd.ring);
 	}
-
 	return 0;
 }
 
 void amdgpu_uvd_free_handles(struct amdgpu_device *adev, struct drm_file *filp)
 {
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst[0].ring;
 	int i, r;
 
 	for (i = 0; i < adev->uvd.max_handles; ++i) {
 		uint32_t handle = atomic_read(&adev->uvd.handles[i]);
+
 		if (handle != 0 && adev->uvd.filp[i] == filp) {
 			struct dma_fence *fence;
 
-			r = amdgpu_uvd_get_destroy_msg(ring, handle,
-						       false, &fence);
+			r = amdgpu_uvd_get_destroy_msg(ring, handle, false,
+						       &fence);
 			if (r) {
-				DRM_ERROR("Error destroying UVD (%d)!\n", r);
+				DRM_ERROR("Error destroying UVD %d!\n", r);
 				continue;
 			}
 
@@ -665,7 +701,7 @@ static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx,
 
 	r = amdgpu_bo_kmap(bo, &ptr);
 	if (r) {
-		DRM_ERROR("Failed mapping the UVD message (%ld)!\n", r);
+		DRM_ERROR("Failed mapping the UVD) message (%ld)!\n", r);
 		return r;
 	}
 
@@ -687,7 +723,8 @@ static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx,
 		/* try to alloc a new handle */
 		for (i = 0; i < adev->uvd.max_handles; ++i) {
 			if (atomic_read(&adev->uvd.handles[i]) == handle) {
-				DRM_ERROR("Handle 0x%x already in use!\n", handle);
+				DRM_ERROR(")Handle 0x%x already in use!\n",
+					  handle);
 				return -EINVAL;
 			}
 
@@ -800,7 +837,7 @@ static int amdgpu_uvd_cs_pass2(struct amdgpu_uvd_cs_ctx *ctx)
 		}
 
 		if ((cmd == 0 || cmd == 0x3) &&
-		    (start >> 28) != (ctx->parser->adev->uvd.gpu_addr >> 28)) {
+		    (start >> 28) != (ctx->parser->adev->uvd.inst->gpu_addr >> 28)) {
 			DRM_ERROR("msg/fb buffer %LX-%LX out of 256MB segment!\n",
 				  start, end);
 			return -EINVAL;
@@ -968,6 +1005,8 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
 	uint64_t addr;
 	long r;
 	int i;
+	unsigned offset_idx = 0;
+	unsigned offset[3] = { UVD_BASE_SI, 0, 0 };
 
 	amdgpu_bo_kunmap(bo);
 	amdgpu_bo_unpin(bo);
@@ -987,17 +1026,16 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
 		goto err;
 
 	if (adev->asic_type >= CHIP_VEGA10) {
-		data[0] = PACKET0(mmUVD_GPCOM_VCPU_DATA0_VEGA10, 0);
-		data[1] = PACKET0(mmUVD_GPCOM_VCPU_DATA1_VEGA10, 0);
-		data[2] = PACKET0(mmUVD_GPCOM_VCPU_CMD_VEGA10, 0);
-		data[3] = PACKET0(mmUVD_NO_OP_VEGA10, 0);
-	} else {
-		data[0] = PACKET0(mmUVD_GPCOM_VCPU_DATA0, 0);
-		data[1] = PACKET0(mmUVD_GPCOM_VCPU_DATA1, 0);
-		data[2] = PACKET0(mmUVD_GPCOM_VCPU_CMD, 0);
-		data[3] = PACKET0(mmUVD_NO_OP, 0);
+		offset_idx = 1 + ring->me;
+		offset[1] = adev->reg_offset[UVD_HWIP][0][1];
+		offset[2] = adev->reg_offset[UVD_HWIP][1][1];
 	}
 
+	data[0] = PACKET0(offset[offset_idx] + UVD_GPCOM_VCPU_DATA0, 0);
+	data[1] = PACKET0(offset[offset_idx] + UVD_GPCOM_VCPU_DATA1, 0);
+	data[2] = PACKET0(offset[offset_idx] + UVD_GPCOM_VCPU_CMD, 0);
+	data[3] = PACKET0(offset[offset_idx] + UVD_NO_OP, 0);
+
 	ib = &job->ibs[0];
 	addr = amdgpu_bo_gpu_offset(bo);
 	ib->ptr[0] = data[0];
@@ -1021,19 +1059,16 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
 		if (r < 0)
 			goto err_free;
 
-		r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f);
-		job->fence = dma_fence_get(f);
+		r = amdgpu_job_submit_direct(job, ring, &f);
 		if (r)
 			goto err_free;
-
-		amdgpu_job_free(job);
 	} else {
 		r = amdgpu_sync_resv(adev, &job->sync, bo->tbo.resv,
 				     AMDGPU_FENCE_OWNER_UNDEFINED, false);
 		if (r)
 			goto err_free;
 
-		r = amdgpu_job_submit(job, ring, &adev->uvd.entity,
+		r = amdgpu_job_submit(job, &adev->uvd.entity,
 				      AMDGPU_FENCE_OWNER_UNDEFINED, &f);
 		if (r)
 			goto err_free;
@@ -1122,7 +1157,14 @@ static void amdgpu_uvd_idle_work_handler(struct work_struct *work)
 {
 	struct amdgpu_device *adev =
 		container_of(work, struct amdgpu_device, uvd.idle_work.work);
-	unsigned fences = amdgpu_fence_count_emitted(&adev->uvd.ring);
+	unsigned fences = 0, i, j;
+
+	for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+		fences += amdgpu_fence_count_emitted(&adev->uvd.inst[i].ring);
+		for (j = 0; j < adev->uvd.num_enc_rings; ++j) {
+			fences += amdgpu_fence_count_emitted(&adev->uvd.inst[i].ring_enc[j]);
+		}
+	}
 
 	if (fences == 0) {
 		if (adev->pm.dpm_enabled) {
@@ -1179,27 +1221,28 @@ int amdgpu_uvd_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 {
 	struct dma_fence *fence;
 	long r;
+	uint32_t ip_instance = ring->me;
 
 	r = amdgpu_uvd_get_create_msg(ring, 1, NULL);
 	if (r) {
-		DRM_ERROR("amdgpu: failed to get create msg (%ld).\n", r);
+		DRM_ERROR("amdgpu: (%d)failed to get create msg (%ld).\n", ip_instance, r);
 		goto error;
 	}
 
 	r = amdgpu_uvd_get_destroy_msg(ring, 1, true, &fence);
 	if (r) {
-		DRM_ERROR("amdgpu: failed to get destroy ib (%ld).\n", r);
+		DRM_ERROR("amdgpu: (%d)failed to get destroy ib (%ld).\n", ip_instance, r);
 		goto error;
 	}
 
 	r = dma_fence_wait_timeout(fence, false, timeout);
 	if (r == 0) {
-		DRM_ERROR("amdgpu: IB test timed out.\n");
+		DRM_ERROR("amdgpu: (%d)IB test timed out.\n", ip_instance);
 		r = -ETIMEDOUT;
 	} else if (r < 0) {
-		DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r);
+		DRM_ERROR("amdgpu: (%d)fence wait failed (%ld).\n", ip_instance, r);
 	} else {
-		DRM_DEBUG("ib test on ring %d succeeded\n",  ring->idx);
+		DRM_DEBUG("ib test on (%d)ring %d succeeded\n", ip_instance, ring->idx);
 		r = 0;
 	}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h
index 32ea20b99e53..66872286ab12 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.h
@@ -31,30 +31,36 @@
 #define AMDGPU_UVD_SESSION_SIZE		(50*1024)
 #define AMDGPU_UVD_FIRMWARE_OFFSET	256
 
+#define AMDGPU_MAX_UVD_INSTANCES			2
+
 #define AMDGPU_UVD_FIRMWARE_SIZE(adev)    \
 	(AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(((const struct common_firmware_header *)(adev)->uvd.fw->data)->ucode_size_bytes) + \
 			       8) - AMDGPU_UVD_FIRMWARE_OFFSET)
 
-struct amdgpu_uvd {
+struct amdgpu_uvd_inst {
 	struct amdgpu_bo	*vcpu_bo;
 	void			*cpu_addr;
 	uint64_t		gpu_addr;
-	unsigned		fw_version;
 	void			*saved_bo;
-	unsigned		max_handles;
-	atomic_t		handles[AMDGPU_MAX_UVD_HANDLES];
-	struct drm_file		*filp[AMDGPU_MAX_UVD_HANDLES];
-	struct delayed_work	idle_work;
-	const struct firmware	*fw;	/* UVD firmware */
 	struct amdgpu_ring	ring;
 	struct amdgpu_ring	ring_enc[AMDGPU_MAX_UVD_ENC_RINGS];
 	struct amdgpu_irq_src	irq;
+	uint32_t                srbm_soft_reset;
+};
+
+struct amdgpu_uvd {
+	const struct firmware	*fw;	/* UVD firmware */
+	unsigned		fw_version;
+	unsigned		max_handles;
+	unsigned		num_enc_rings;
+	uint8_t			num_uvd_inst;
 	bool			address_64_bit;
 	bool			use_ctx_buf;
+	struct amdgpu_uvd_inst	inst[AMDGPU_MAX_UVD_INSTANCES];
+	struct drm_file		*filp[AMDGPU_MAX_UVD_HANDLES];
+	atomic_t		handles[AMDGPU_MAX_UVD_HANDLES];
 	struct drm_sched_entity entity;
-	struct drm_sched_entity entity_enc;
-	uint32_t                srbm_soft_reset;
-	unsigned		num_enc_rings;
+	struct delayed_work	idle_work;
 };
 
 int amdgpu_uvd_sw_init(struct amdgpu_device *adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
index a33804bd3314..86182c966ed6 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vce.c
@@ -40,22 +40,24 @@
 
 /* Firmware Names */
 #ifdef CONFIG_DRM_AMDGPU_CIK
-#define FIRMWARE_BONAIRE	"radeon/bonaire_vce.bin"
-#define FIRMWARE_KABINI	"radeon/kabini_vce.bin"
-#define FIRMWARE_KAVERI	"radeon/kaveri_vce.bin"
-#define FIRMWARE_HAWAII	"radeon/hawaii_vce.bin"
-#define FIRMWARE_MULLINS	"radeon/mullins_vce.bin"
+#define FIRMWARE_BONAIRE	"amdgpu/bonaire_vce.bin"
+#define FIRMWARE_KABINI	"amdgpu/kabini_vce.bin"
+#define FIRMWARE_KAVERI	"amdgpu/kaveri_vce.bin"
+#define FIRMWARE_HAWAII	"amdgpu/hawaii_vce.bin"
+#define FIRMWARE_MULLINS	"amdgpu/mullins_vce.bin"
 #endif
 #define FIRMWARE_TONGA		"amdgpu/tonga_vce.bin"
 #define FIRMWARE_CARRIZO	"amdgpu/carrizo_vce.bin"
 #define FIRMWARE_FIJI		"amdgpu/fiji_vce.bin"
 #define FIRMWARE_STONEY		"amdgpu/stoney_vce.bin"
 #define FIRMWARE_POLARIS10	"amdgpu/polaris10_vce.bin"
-#define FIRMWARE_POLARIS11         "amdgpu/polaris11_vce.bin"
-#define FIRMWARE_POLARIS12         "amdgpu/polaris12_vce.bin"
+#define FIRMWARE_POLARIS11	"amdgpu/polaris11_vce.bin"
+#define FIRMWARE_POLARIS12	"amdgpu/polaris12_vce.bin"
+#define FIRMWARE_VEGAM		"amdgpu/vegam_vce.bin"
 
 #define FIRMWARE_VEGA10		"amdgpu/vega10_vce.bin"
 #define FIRMWARE_VEGA12		"amdgpu/vega12_vce.bin"
+#define FIRMWARE_VEGA20		"amdgpu/vega20_vce.bin"
 
 #ifdef CONFIG_DRM_AMDGPU_CIK
 MODULE_FIRMWARE(FIRMWARE_BONAIRE);
@@ -71,9 +73,11 @@ MODULE_FIRMWARE(FIRMWARE_STONEY);
 MODULE_FIRMWARE(FIRMWARE_POLARIS10);
 MODULE_FIRMWARE(FIRMWARE_POLARIS11);
 MODULE_FIRMWARE(FIRMWARE_POLARIS12);
+MODULE_FIRMWARE(FIRMWARE_VEGAM);
 
 MODULE_FIRMWARE(FIRMWARE_VEGA10);
 MODULE_FIRMWARE(FIRMWARE_VEGA12);
+MODULE_FIRMWARE(FIRMWARE_VEGA20);
 
 static void amdgpu_vce_idle_work_handler(struct work_struct *work);
 
@@ -132,12 +136,18 @@ int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size)
 	case CHIP_POLARIS12:
 		fw_name = FIRMWARE_POLARIS12;
 		break;
+	case CHIP_VEGAM:
+		fw_name = FIRMWARE_VEGAM;
+		break;
 	case CHIP_VEGA10:
 		fw_name = FIRMWARE_VEGA10;
 		break;
 	case CHIP_VEGA12:
 		fw_name = FIRMWARE_VEGA12;
 		break;
+	case CHIP_VEGA20:
+		fw_name = FIRMWARE_VEGA20;
+		break;
 
 	default:
 		return -EINVAL;
@@ -180,8 +190,7 @@ int amdgpu_vce_sw_init(struct amdgpu_device *adev, unsigned long size)
 
 	ring = &adev->vce.ring[0];
 	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
-	r = drm_sched_entity_init(&ring->sched, &adev->vce.entity,
-				  rq, amdgpu_sched_jobs, NULL);
+	r = drm_sched_entity_init(&adev->vce.entity, &rq, 1, NULL);
 	if (r != 0) {
 		DRM_ERROR("Failed setting up VCE run queue.\n");
 		return r;
@@ -212,7 +221,7 @@ int amdgpu_vce_sw_fini(struct amdgpu_device *adev)
 	if (adev->vce.vcpu_bo == NULL)
 		return 0;
 
-	drm_sched_entity_fini(&adev->vce.ring[0].sched, &adev->vce.entity);
+	drm_sched_entity_destroy(&adev->vce.ring[0].sched, &adev->vce.entity);
 
 	amdgpu_bo_free_kernel(&adev->vce.vcpu_bo, &adev->vce.gpu_addr,
 		(void **)&adev->vce.cpu_addr);
@@ -460,12 +469,10 @@ int amdgpu_vce_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
 	for (i = ib->length_dw; i < ib_size_dw; ++i)
 		ib->ptr[i] = 0x0;
 
-	r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f);
-	job->fence = dma_fence_get(f);
+	r = amdgpu_job_submit_direct(job, ring, &f);
 	if (r)
 		goto err;
 
-	amdgpu_job_free(job);
 	if (fence)
 		*fence = dma_fence_get(f);
 	dma_fence_put(f);
@@ -522,19 +529,13 @@ int amdgpu_vce_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
 	for (i = ib->length_dw; i < ib_size_dw; ++i)
 		ib->ptr[i] = 0x0;
 
-	if (direct) {
-		r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f);
-		job->fence = dma_fence_get(f);
-		if (r)
-			goto err;
-
-		amdgpu_job_free(job);
-	} else {
-		r = amdgpu_job_submit(job, ring, &ring->adev->vce.entity,
+	if (direct)
+		r = amdgpu_job_submit_direct(job, ring, &f);
+	else
+		r = amdgpu_job_submit(job, &ring->adev->vce.entity,
 				      AMDGPU_FENCE_OWNER_UNDEFINED, &f);
-		if (r)
-			goto err;
-	}
+	if (r)
+		goto err;
 
 	if (fence)
 		*fence = dma_fence_get(f);
@@ -755,6 +756,18 @@ int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx)
 			if (r)
 				goto out;
 			break;
+
+		case 0x0500000d: /* MV buffer */
+			r = amdgpu_vce_validate_bo(p, ib_idx, idx + 3,
+							idx + 2, 0, 0);
+			if (r)
+				goto out;
+
+			r = amdgpu_vce_validate_bo(p, ib_idx, idx + 8,
+							idx + 7, 0, 0);
+			if (r)
+				goto out;
+			break;
 		}
 
 		idx += len / 4;
@@ -860,6 +873,18 @@ int amdgpu_vce_ring_parse_cs(struct amdgpu_cs_parser *p, uint32_t ib_idx)
 				goto out;
 			break;
 
+		case 0x0500000d: /* MV buffer */
+			r = amdgpu_vce_cs_reloc(p, ib_idx, idx + 3,
+							idx + 2, *size, 0);
+			if (r)
+				goto out;
+
+			r = amdgpu_vce_cs_reloc(p, ib_idx, idx + 8,
+							idx + 7, *size / 12, 0);
+			if (r)
+				goto out;
+			break;
+
 		default:
 			DRM_ERROR("invalid VCE command (0x%x)!\n", cmd);
 			r = -EINVAL;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
index 58e495330b38..798648a19710 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.c
@@ -49,12 +49,10 @@ static void amdgpu_vcn_idle_work_handler(struct work_struct *work);
 
 int amdgpu_vcn_sw_init(struct amdgpu_device *adev)
 {
-	struct amdgpu_ring *ring;
-	struct drm_sched_rq *rq;
 	unsigned long bo_size;
 	const char *fw_name;
 	const struct common_firmware_header *hdr;
-	unsigned version_major, version_minor, family_id;
+	unsigned char fw_check;
 	int r;
 
 	INIT_DELAYED_WORK(&adev->vcn.idle_work, amdgpu_vcn_idle_work_handler);
@@ -84,12 +82,34 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev)
 	}
 
 	hdr = (const struct common_firmware_header *)adev->vcn.fw->data;
-	family_id = le32_to_cpu(hdr->ucode_version) & 0xff;
-	version_major = (le32_to_cpu(hdr->ucode_version) >> 24) & 0xff;
-	version_minor = (le32_to_cpu(hdr->ucode_version) >> 8) & 0xff;
-	DRM_INFO("Found VCN firmware Version: %hu.%hu Family ID: %hu\n",
-		version_major, version_minor, family_id);
+	adev->vcn.fw_version = le32_to_cpu(hdr->ucode_version);
+
+	/* Bit 20-23, it is encode major and non-zero for new naming convention.
+	 * This field is part of version minor and DRM_DISABLED_FLAG in old naming
+	 * convention. Since the l:wq!atest version minor is 0x5B and DRM_DISABLED_FLAG
+	 * is zero in old naming convention, this field is always zero so far.
+	 * These four bits are used to tell which naming convention is present.
+	 */
+	fw_check = (le32_to_cpu(hdr->ucode_version) >> 20) & 0xf;
+	if (fw_check) {
+		unsigned int dec_ver, enc_major, enc_minor, vep, fw_rev;
+
+		fw_rev = le32_to_cpu(hdr->ucode_version) & 0xfff;
+		enc_minor = (le32_to_cpu(hdr->ucode_version) >> 12) & 0xff;
+		enc_major = fw_check;
+		dec_ver = (le32_to_cpu(hdr->ucode_version) >> 24) & 0xf;
+		vep = (le32_to_cpu(hdr->ucode_version) >> 28) & 0xf;
+		DRM_INFO("Found VCN firmware Version ENC: %hu.%hu DEC: %hu VEP: %hu Revision: %hu\n",
+			enc_major, enc_minor, dec_ver, vep, fw_rev);
+	} else {
+		unsigned int version_major, version_minor, family_id;
 
+		family_id = le32_to_cpu(hdr->ucode_version) & 0xff;
+		version_major = (le32_to_cpu(hdr->ucode_version) >> 24) & 0xff;
+		version_minor = (le32_to_cpu(hdr->ucode_version) >> 8) & 0xff;
+		DRM_INFO("Found VCN firmware Version: %hu.%hu Family ID: %hu\n",
+			version_major, version_minor, family_id);
+	}
 
 	bo_size = AMDGPU_GPU_PAGE_ALIGN(le32_to_cpu(hdr->ucode_size_bytes) + 8)
 		  +  AMDGPU_VCN_STACK_SIZE + AMDGPU_VCN_HEAP_SIZE
@@ -102,24 +122,6 @@ int amdgpu_vcn_sw_init(struct amdgpu_device *adev)
 		return r;
 	}
 
-	ring = &adev->vcn.ring_dec;
-	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
-	r = drm_sched_entity_init(&ring->sched, &adev->vcn.entity_dec,
-				  rq, amdgpu_sched_jobs, NULL);
-	if (r != 0) {
-		DRM_ERROR("Failed setting up VCN dec run queue.\n");
-		return r;
-	}
-
-	ring = &adev->vcn.ring_enc[0];
-	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
-	r = drm_sched_entity_init(&ring->sched, &adev->vcn.entity_enc,
-				  rq, amdgpu_sched_jobs, NULL);
-	if (r != 0) {
-		DRM_ERROR("Failed setting up VCN enc run queue.\n");
-		return r;
-	}
-
 	return 0;
 }
 
@@ -129,10 +131,6 @@ int amdgpu_vcn_sw_fini(struct amdgpu_device *adev)
 
 	kfree(adev->vcn.saved_bo);
 
-	drm_sched_entity_fini(&adev->vcn.ring_dec.sched, &adev->vcn.entity_dec);
-
-	drm_sched_entity_fini(&adev->vcn.ring_enc[0].sched, &adev->vcn.entity_enc);
-
 	amdgpu_bo_free_kernel(&adev->vcn.vcpu_bo,
 			      &adev->vcn.gpu_addr,
 			      (void **)&adev->vcn.cpu_addr);
@@ -142,6 +140,8 @@ int amdgpu_vcn_sw_fini(struct amdgpu_device *adev)
 	for (i = 0; i < adev->vcn.num_enc_rings; ++i)
 		amdgpu_ring_fini(&adev->vcn.ring_enc[i]);
 
+	amdgpu_ring_fini(&adev->vcn.ring_jpeg);
+
 	release_firmware(adev->vcn.fw);
 
 	return 0;
@@ -205,13 +205,20 @@ static void amdgpu_vcn_idle_work_handler(struct work_struct *work)
 	struct amdgpu_device *adev =
 		container_of(work, struct amdgpu_device, vcn.idle_work.work);
 	unsigned fences = amdgpu_fence_count_emitted(&adev->vcn.ring_dec);
+	unsigned i;
+
+	for (i = 0; i < adev->vcn.num_enc_rings; ++i) {
+		fences += amdgpu_fence_count_emitted(&adev->vcn.ring_enc[i]);
+	}
+
+	fences += amdgpu_fence_count_emitted(&adev->vcn.ring_jpeg);
 
 	if (fences == 0) {
-		if (adev->pm.dpm_enabled) {
-			/* might be used when with pg/cg
+		if (adev->pm.dpm_enabled)
 			amdgpu_dpm_enable_uvd(adev, false);
-			*/
-		}
+		else
+			amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCN,
+							       AMD_PG_STATE_GATE);
 	} else {
 		schedule_delayed_work(&adev->vcn.idle_work, VCN_IDLE_TIMEOUT);
 	}
@@ -222,10 +229,12 @@ void amdgpu_vcn_ring_begin_use(struct amdgpu_ring *ring)
 	struct amdgpu_device *adev = ring->adev;
 	bool set_clocks = !cancel_delayed_work_sync(&adev->vcn.idle_work);
 
-	if (set_clocks && adev->pm.dpm_enabled) {
-		/* might be used when with pg/cg
-		amdgpu_dpm_enable_uvd(adev, true);
-		*/
+	if (set_clocks) {
+		if (adev->pm.dpm_enabled)
+			amdgpu_dpm_enable_uvd(adev, true);
+		else
+			amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_VCN,
+							       AMD_PG_STATE_UNGATE);
 	}
 }
 
@@ -271,7 +280,7 @@ int amdgpu_vcn_dec_ring_test_ring(struct amdgpu_ring *ring)
 }
 
 static int amdgpu_vcn_dec_send_msg(struct amdgpu_ring *ring,
-				   struct amdgpu_bo *bo, bool direct,
+				   struct amdgpu_bo *bo,
 				   struct dma_fence **fence)
 {
 	struct amdgpu_device *adev = ring->adev;
@@ -299,19 +308,9 @@ static int amdgpu_vcn_dec_send_msg(struct amdgpu_ring *ring,
 	}
 	ib->length_dw = 16;
 
-	if (direct) {
-		r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f);
-		job->fence = dma_fence_get(f);
-		if (r)
-			goto err_free;
-
-		amdgpu_job_free(job);
-	} else {
-		r = amdgpu_job_submit(job, ring, &adev->vcn.entity_dec,
-				      AMDGPU_FENCE_OWNER_UNDEFINED, &f);
-		if (r)
-			goto err_free;
-	}
+	r = amdgpu_job_submit_direct(job, ring, &f);
+	if (r)
+		goto err_free;
 
 	amdgpu_bo_fence(bo, f, false);
 	amdgpu_bo_unreserve(bo);
@@ -363,11 +362,11 @@ static int amdgpu_vcn_dec_get_create_msg(struct amdgpu_ring *ring, uint32_t hand
 	for (i = 14; i < 1024; ++i)
 		msg[i] = cpu_to_le32(0x0);
 
-	return amdgpu_vcn_dec_send_msg(ring, bo, true, fence);
+	return amdgpu_vcn_dec_send_msg(ring, bo, fence);
 }
 
 static int amdgpu_vcn_dec_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
-			       bool direct, struct dma_fence **fence)
+			       struct dma_fence **fence)
 {
 	struct amdgpu_device *adev = ring->adev;
 	struct amdgpu_bo *bo = NULL;
@@ -389,7 +388,7 @@ static int amdgpu_vcn_dec_get_destroy_msg(struct amdgpu_ring *ring, uint32_t han
 	for (i = 6; i < 1024; ++i)
 		msg[i] = cpu_to_le32(0x0);
 
-	return amdgpu_vcn_dec_send_msg(ring, bo, direct, fence);
+	return amdgpu_vcn_dec_send_msg(ring, bo, fence);
 }
 
 int amdgpu_vcn_dec_ring_test_ib(struct amdgpu_ring *ring, long timeout)
@@ -403,7 +402,7 @@ int amdgpu_vcn_dec_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 		goto error;
 	}
 
-	r = amdgpu_vcn_dec_get_destroy_msg(ring, 1, true, &fence);
+	r = amdgpu_vcn_dec_get_destroy_msg(ring, 1, &fence);
 	if (r) {
 		DRM_ERROR("amdgpu: failed to get destroy ib (%ld).\n", r);
 		goto error;
@@ -497,12 +496,10 @@ static int amdgpu_vcn_enc_get_create_msg(struct amdgpu_ring *ring, uint32_t hand
 	for (i = ib->length_dw; i < ib_size_dw; ++i)
 		ib->ptr[i] = 0x0;
 
-	r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f);
-	job->fence = dma_fence_get(f);
+	r = amdgpu_job_submit_direct(job, ring, &f);
 	if (r)
 		goto err;
 
-	amdgpu_job_free(job);
 	if (fence)
 		*fence = dma_fence_get(f);
 	dma_fence_put(f);
@@ -551,12 +548,10 @@ static int amdgpu_vcn_enc_get_destroy_msg(struct amdgpu_ring *ring, uint32_t han
 	for (i = ib->length_dw; i < ib_size_dw; ++i)
 		ib->ptr[i] = 0x0;
 
-	r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f);
-	job->fence = dma_fence_get(f);
+	r = amdgpu_job_submit_direct(job, ring, &f);
 	if (r)
 		goto err;
 
-	amdgpu_job_free(job);
 	if (fence)
 		*fence = dma_fence_get(f);
 	dma_fence_put(f);
@@ -599,3 +594,127 @@ error:
 	dma_fence_put(fence);
 	return r;
 }
+
+int amdgpu_vcn_jpeg_ring_test_ring(struct amdgpu_ring *ring)
+{
+	struct amdgpu_device *adev = ring->adev;
+	uint32_t tmp = 0;
+	unsigned i;
+	int r;
+
+	WREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_CONTEXT_ID), 0xCAFEDEAD);
+	r = amdgpu_ring_alloc(ring, 3);
+
+	if (r) {
+		DRM_ERROR("amdgpu: cp failed to lock ring %d (%d).\n",
+				  ring->idx, r);
+		return r;
+	}
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_CONTEXT_ID), 0, 0, 0));
+	amdgpu_ring_write(ring, 0xDEADBEEF);
+	amdgpu_ring_commit(ring);
+
+	for (i = 0; i < adev->usec_timeout; i++) {
+		tmp = RREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_CONTEXT_ID));
+		if (tmp == 0xDEADBEEF)
+			break;
+		DRM_UDELAY(1);
+	}
+
+	if (i < adev->usec_timeout) {
+		DRM_DEBUG("ring test on %d succeeded in %d usecs\n",
+				  ring->idx, i);
+	} else {
+		DRM_ERROR("amdgpu: ring %d test failed (0x%08X)\n",
+				  ring->idx, tmp);
+		r = -EINVAL;
+	}
+
+	return r;
+}
+
+static int amdgpu_vcn_jpeg_set_reg(struct amdgpu_ring *ring, uint32_t handle,
+		struct dma_fence **fence)
+{
+	struct amdgpu_device *adev = ring->adev;
+	struct amdgpu_job *job;
+	struct amdgpu_ib *ib;
+	struct dma_fence *f = NULL;
+	const unsigned ib_size_dw = 16;
+	int i, r;
+
+	r = amdgpu_job_alloc_with_ib(ring->adev, ib_size_dw * 4, &job);
+	if (r)
+		return r;
+
+	ib = &job->ibs[0];
+
+	ib->ptr[0] = PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JPEG_PITCH), 0, 0, PACKETJ_TYPE0);
+	ib->ptr[1] = 0xDEADBEEF;
+	for (i = 2; i < 16; i += 2) {
+		ib->ptr[i] = PACKETJ(0, 0, 0, PACKETJ_TYPE6);
+		ib->ptr[i+1] = 0;
+	}
+	ib->length_dw = 16;
+
+	r = amdgpu_job_submit_direct(job, ring, &f);
+	if (r)
+		goto err;
+
+	if (fence)
+		*fence = dma_fence_get(f);
+	dma_fence_put(f);
+
+	return 0;
+
+err:
+	amdgpu_job_free(job);
+	return r;
+}
+
+int amdgpu_vcn_jpeg_ring_test_ib(struct amdgpu_ring *ring, long timeout)
+{
+	struct amdgpu_device *adev = ring->adev;
+	uint32_t tmp = 0;
+	unsigned i;
+	struct dma_fence *fence = NULL;
+	long r = 0;
+
+	r = amdgpu_vcn_jpeg_set_reg(ring, 1, &fence);
+	if (r) {
+		DRM_ERROR("amdgpu: failed to set jpeg register (%ld).\n", r);
+		goto error;
+	}
+
+	r = dma_fence_wait_timeout(fence, false, timeout);
+	if (r == 0) {
+		DRM_ERROR("amdgpu: IB test timed out.\n");
+		r = -ETIMEDOUT;
+		goto error;
+	} else if (r < 0) {
+		DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r);
+		goto error;
+	} else
+		r = 0;
+
+	for (i = 0; i < adev->usec_timeout; i++) {
+		tmp = RREG32(SOC15_REG_OFFSET(UVD, 0, mmUVD_JPEG_PITCH));
+		if (tmp == 0xDEADBEEF)
+			break;
+		DRM_UDELAY(1);
+	}
+
+	if (i < adev->usec_timeout)
+		DRM_DEBUG("ib test on ring %d succeeded\n", ring->idx);
+	else {
+		DRM_ERROR("ib test failed (0x%08X)\n", tmp);
+		r = -EINVAL;
+	}
+
+	dma_fence_put(fence);
+
+error:
+	return r;
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
index 2fd7db891689..0b0b8638d73f 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
@@ -45,6 +45,17 @@
 #define VCN_ENC_CMD_REG_WRITE		0x0000000b
 #define VCN_ENC_CMD_REG_WAIT		0x0000000c
 
+enum engine_status_constants {
+	UVD_PGFSM_STATUS__UVDM_UVDU_PWR_ON = 0x2AAAA0,
+	UVD_PGFSM_CONFIG__UVDM_UVDU_PWR_ON = 0x00000002,
+	UVD_STATUS__UVD_BUSY = 0x00000004,
+	GB_ADDR_CONFIG_DEFAULT = 0x26010011,
+	UVD_STATUS__IDLE = 0x2,
+	UVD_STATUS__BUSY = 0x5,
+	UVD_POWER_STATUS__UVD_POWER_STATUS_TILES_OFF = 0x1,
+	UVD_STATUS__RBC_BUSY = 0x1,
+};
+
 struct amdgpu_vcn {
 	struct amdgpu_bo	*vcpu_bo;
 	void			*cpu_addr;
@@ -55,9 +66,8 @@ struct amdgpu_vcn {
 	const struct firmware	*fw;	/* VCN firmware */
 	struct amdgpu_ring	ring_dec;
 	struct amdgpu_ring	ring_enc[AMDGPU_VCN_MAX_ENC_RINGS];
+	struct amdgpu_ring	ring_jpeg;
 	struct amdgpu_irq_src	irq;
-	struct drm_sched_entity entity_dec;
-	struct drm_sched_entity entity_enc;
 	unsigned		num_enc_rings;
 };
 
@@ -74,4 +84,7 @@ int amdgpu_vcn_dec_ring_test_ib(struct amdgpu_ring *ring, long timeout);
 int amdgpu_vcn_enc_ring_test_ring(struct amdgpu_ring *ring);
 int amdgpu_vcn_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout);
 
+int amdgpu_vcn_jpeg_ring_test_ring(struct amdgpu_ring *ring);
+int amdgpu_vcn_jpeg_ring_test_ib(struct amdgpu_ring *ring, long timeout);
+
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
index da55a78d7380..098dd1ba751a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c
@@ -33,9 +33,11 @@
 #include "amdgpu.h"
 #include "amdgpu_trace.h"
 #include "amdgpu_amdkfd.h"
+#include "amdgpu_gmc.h"
 
-/*
- * GPUVM
+/**
+ * DOC: GPUVM
+ *
  * GPUVM is similar to the legacy gart on older asics, however
  * rather than there being a single global gart table
  * for the entire GPU, there are multiple VM page tables active
@@ -63,43 +65,123 @@ INTERVAL_TREE_DEFINE(struct amdgpu_bo_va_mapping, rb, uint64_t, __subtree_last,
 #undef START
 #undef LAST
 
-/* Local structure. Encapsulate some VM table update parameters to reduce
+/**
+ * struct amdgpu_pte_update_params - Local structure
+ *
+ * Encapsulate some VM table update parameters to reduce
  * the number of function parameters
+ *
  */
 struct amdgpu_pte_update_params {
-	/* amdgpu device we do this update for */
+
+	/**
+	 * @adev: amdgpu device we do this update for
+	 */
 	struct amdgpu_device *adev;
-	/* optional amdgpu_vm we do this update for */
+
+	/**
+	 * @vm: optional amdgpu_vm we do this update for
+	 */
 	struct amdgpu_vm *vm;
-	/* address where to copy page table entries from */
+
+	/**
+	 * @src: address where to copy page table entries from
+	 */
 	uint64_t src;
-	/* indirect buffer to fill with commands */
+
+	/**
+	 * @ib: indirect buffer to fill with commands
+	 */
 	struct amdgpu_ib *ib;
-	/* Function which actually does the update */
+
+	/**
+	 * @func: Function which actually does the update
+	 */
 	void (*func)(struct amdgpu_pte_update_params *params,
 		     struct amdgpu_bo *bo, uint64_t pe,
 		     uint64_t addr, unsigned count, uint32_t incr,
 		     uint64_t flags);
-	/* The next two are used during VM update by CPU
-	 *  DMA addresses to use for mapping
-	 *  Kernel pointer of PD/PT BO that needs to be updated
+	/**
+	 * @pages_addr:
+	 *
+	 * DMA addresses to use for mapping, used during VM update by CPU
 	 */
 	dma_addr_t *pages_addr;
+
+	/**
+	 * @kptr:
+	 *
+	 * Kernel pointer of PD/PT BO that needs to be updated,
+	 * used during VM update by CPU
+	 */
 	void *kptr;
 };
 
-/* Helper to disable partial resident texture feature from a fence callback */
+/**
+ * struct amdgpu_prt_cb - Helper to disable partial resident texture feature from a fence callback
+ */
 struct amdgpu_prt_cb {
+
+	/**
+	 * @adev: amdgpu device
+	 */
 	struct amdgpu_device *adev;
+
+	/**
+	 * @cb: callback
+	 */
 	struct dma_fence_cb cb;
 };
 
 /**
+ * amdgpu_vm_bo_base_init - Adds bo to the list of bos associated with the vm
+ *
+ * @base: base structure for tracking BO usage in a VM
+ * @vm: vm to which bo is to be added
+ * @bo: amdgpu buffer object
+ *
+ * Initialize a bo_va_base structure and add it to the appropriate lists
+ *
+ */
+static void amdgpu_vm_bo_base_init(struct amdgpu_vm_bo_base *base,
+				   struct amdgpu_vm *vm,
+				   struct amdgpu_bo *bo)
+{
+	base->vm = vm;
+	base->bo = bo;
+	INIT_LIST_HEAD(&base->bo_list);
+	INIT_LIST_HEAD(&base->vm_status);
+
+	if (!bo)
+		return;
+	list_add_tail(&base->bo_list, &bo->va);
+
+	if (bo->tbo.type == ttm_bo_type_kernel)
+		list_move(&base->vm_status, &vm->relocated);
+
+	if (bo->tbo.resv != vm->root.base.bo->tbo.resv)
+		return;
+
+	if (bo->preferred_domains &
+	    amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type))
+		return;
+
+	/*
+	 * we checked all the prerequisites, but it looks like this per vm bo
+	 * is currently evicted. add the bo to the evicted list to make sure it
+	 * is validated on next vm use to avoid fault.
+	 * */
+	list_move_tail(&base->vm_status, &vm->evicted);
+}
+
+/**
  * amdgpu_vm_level_shift - return the addr shift for each level
  *
  * @adev: amdgpu_device pointer
+ * @level: VMPT level
  *
- * Returns the number of bits the pfn needs to be right shifted for a level.
+ * Returns:
+ * The number of bits the pfn needs to be right shifted for a level.
  */
 static unsigned amdgpu_vm_level_shift(struct amdgpu_device *adev,
 				      unsigned level)
@@ -127,8 +209,10 @@ static unsigned amdgpu_vm_level_shift(struct amdgpu_device *adev,
  * amdgpu_vm_num_entries - return the number of entries in a PD/PT
  *
  * @adev: amdgpu_device pointer
+ * @level: VMPT level
  *
- * Calculate the number of entries in a page directory or page table.
+ * Returns:
+ * The number of entries in a page directory or page table.
  */
 static unsigned amdgpu_vm_num_entries(struct amdgpu_device *adev,
 				      unsigned level)
@@ -151,8 +235,10 @@ static unsigned amdgpu_vm_num_entries(struct amdgpu_device *adev,
  * amdgpu_vm_bo_size - returns the size of the BOs in bytes
  *
  * @adev: amdgpu_device pointer
+ * @level: VMPT level
  *
- * Calculate the size of the BO for a page directory or page table in bytes.
+ * Returns:
+ * The size of the BO for a page directory or page table in bytes.
  */
 static unsigned amdgpu_vm_bo_size(struct amdgpu_device *adev, unsigned level)
 {
@@ -190,30 +276,25 @@ void amdgpu_vm_get_pd_bo(struct amdgpu_vm *vm,
  * @param: parameter for the validation callback
  *
  * Validate the page table BOs on command submission if neccessary.
+ *
+ * Returns:
+ * Validation result.
  */
 int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 			      int (*validate)(void *p, struct amdgpu_bo *bo),
 			      void *param)
 {
 	struct ttm_bo_global *glob = adev->mman.bdev.glob;
-	int r;
+	struct amdgpu_vm_bo_base *bo_base, *tmp;
+	int r = 0;
 
-	spin_lock(&vm->status_lock);
-	while (!list_empty(&vm->evicted)) {
-		struct amdgpu_vm_bo_base *bo_base;
-		struct amdgpu_bo *bo;
+	list_for_each_entry_safe(bo_base, tmp, &vm->evicted, vm_status) {
+		struct amdgpu_bo *bo = bo_base->bo;
 
-		bo_base = list_first_entry(&vm->evicted,
-					   struct amdgpu_vm_bo_base,
-					   vm_status);
-		spin_unlock(&vm->status_lock);
-
-		bo = bo_base->bo;
-		BUG_ON(!bo);
 		if (bo->parent) {
 			r = validate(param, bo);
 			if (r)
-				return r;
+				break;
 
 			spin_lock(&glob->lru_lock);
 			ttm_bo_move_to_lru_tail(&bo->tbo);
@@ -222,22 +303,29 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 			spin_unlock(&glob->lru_lock);
 		}
 
-		if (bo->tbo.type == ttm_bo_type_kernel &&
-		    vm->use_cpu_for_update) {
-			r = amdgpu_bo_kmap(bo, NULL);
-			if (r)
-				return r;
-		}
-
-		spin_lock(&vm->status_lock);
-		if (bo->tbo.type != ttm_bo_type_kernel)
+		if (bo->tbo.type != ttm_bo_type_kernel) {
+			spin_lock(&vm->moved_lock);
 			list_move(&bo_base->vm_status, &vm->moved);
-		else
+			spin_unlock(&vm->moved_lock);
+		} else {
 			list_move(&bo_base->vm_status, &vm->relocated);
+		}
 	}
-	spin_unlock(&vm->status_lock);
 
-	return 0;
+	spin_lock(&glob->lru_lock);
+	list_for_each_entry(bo_base, &vm->idle, vm_status) {
+		struct amdgpu_bo *bo = bo_base->bo;
+
+		if (!bo->parent)
+			continue;
+
+		ttm_bo_move_to_lru_tail(&bo->tbo);
+		if (bo->shadow)
+			ttm_bo_move_to_lru_tail(&bo->shadow->tbo);
+	}
+	spin_unlock(&glob->lru_lock);
+
+	return r;
 }
 
 /**
@@ -246,26 +334,28 @@ int amdgpu_vm_validate_pt_bos(struct amdgpu_device *adev, struct amdgpu_vm *vm,
  * @vm: VM to check
  *
  * Check if all VM PDs/PTs are ready for updates
+ *
+ * Returns:
+ * True if eviction list is empty.
  */
 bool amdgpu_vm_ready(struct amdgpu_vm *vm)
 {
-	bool ready;
-
-	spin_lock(&vm->status_lock);
-	ready = list_empty(&vm->evicted);
-	spin_unlock(&vm->status_lock);
-
-	return ready;
+	return list_empty(&vm->evicted);
 }
 
 /**
  * amdgpu_vm_clear_bo - initially clear the PDs/PTs
  *
  * @adev: amdgpu_device pointer
+ * @vm: VM to clear BO from
  * @bo: BO to clear
  * @level: level this BO is at
+ * @pte_support_ats: indicate ATS support from PTE
  *
  * Root PD needs to be reserved when calling this.
+ *
+ * Returns:
+ * 0 on success, errno otherwise.
  */
 static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
 			      struct amdgpu_vm *vm, struct amdgpu_bo *bo,
@@ -335,8 +425,8 @@ static int amdgpu_vm_clear_bo(struct amdgpu_device *adev,
 	if (r)
 		goto error_free;
 
-	r = amdgpu_job_submit(job, ring, &vm->entity,
-			      AMDGPU_FENCE_OWNER_UNDEFINED, &fence);
+	r = amdgpu_job_submit(job, &vm->entity, AMDGPU_FENCE_OWNER_UNDEFINED,
+			      &fence);
 	if (r)
 		goto error_free;
 
@@ -361,10 +451,16 @@ error:
  *
  * @adev: amdgpu_device pointer
  * @vm: requested vm
+ * @parent: parent PT
  * @saddr: start of the address range
  * @eaddr: end of the address range
+ * @level: VMPT level
+ * @ats: indicate ATS support from PTE
  *
  * Make sure the page directories and page tables are allocated
+ *
+ * Returns:
+ * 0 on success, errno otherwise.
  */
 static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev,
 				  struct amdgpu_vm *vm,
@@ -412,11 +508,16 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev,
 		struct amdgpu_bo *pt;
 
 		if (!entry->base.bo) {
-			r = amdgpu_bo_create(adev,
-					     amdgpu_vm_bo_size(adev, level),
-					     AMDGPU_GPU_PAGE_SIZE,
-					     AMDGPU_GEM_DOMAIN_VRAM, flags,
-					     ttm_bo_type_kernel, resv, &pt);
+			struct amdgpu_bo_param bp;
+
+			memset(&bp, 0, sizeof(bp));
+			bp.size = amdgpu_vm_bo_size(adev, level);
+			bp.byte_align = AMDGPU_GPU_PAGE_SIZE;
+			bp.domain = AMDGPU_GEM_DOMAIN_VRAM;
+			bp.flags = flags;
+			bp.type = ttm_bo_type_kernel;
+			bp.resv = resv;
+			r = amdgpu_bo_create(adev, &bp, &pt);
 			if (r)
 				return r;
 
@@ -441,12 +542,7 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev,
 			*/
 			pt->parent = amdgpu_bo_ref(parent->base.bo);
 
-			entry->base.vm = vm;
-			entry->base.bo = pt;
-			list_add_tail(&entry->base.bo_list, &pt->va);
-			spin_lock(&vm->status_lock);
-			list_add(&entry->base.vm_status, &vm->relocated);
-			spin_unlock(&vm->status_lock);
+			amdgpu_vm_bo_base_init(&entry->base, vm, pt);
 		}
 
 		if (level < AMDGPU_VM_PTB) {
@@ -472,6 +568,9 @@ static int amdgpu_vm_alloc_levels(struct amdgpu_device *adev,
  * @size: Size from start address we need.
  *
  * Make sure the page tables are allocated.
+ *
+ * Returns:
+ * 0 on success, errno otherwise.
  */
 int amdgpu_vm_alloc_pts(struct amdgpu_device *adev,
 			struct amdgpu_vm *vm,
@@ -537,6 +636,15 @@ void amdgpu_vm_check_compute_bug(struct amdgpu_device *adev)
 	}
 }
 
+/**
+ * amdgpu_vm_need_pipeline_sync - Check if pipe sync is needed for job.
+ *
+ * @ring: ring on which the job will be submitted
+ * @job: job to submit
+ *
+ * Returns:
+ * True if sync is needed.
+ */
 bool amdgpu_vm_need_pipeline_sync(struct amdgpu_ring *ring,
 				  struct amdgpu_job *job)
 {
@@ -564,19 +672,17 @@ bool amdgpu_vm_need_pipeline_sync(struct amdgpu_ring *ring,
 	return vm_flush_needed || gds_switch_needed;
 }
 
-static bool amdgpu_vm_is_large_bar(struct amdgpu_device *adev)
-{
-	return (adev->gmc.real_vram_size == adev->gmc.visible_vram_size);
-}
-
 /**
  * amdgpu_vm_flush - hardware flush the vm
  *
  * @ring: ring to use for flush
- * @vmid: vmid number to use
- * @pd_addr: address of the page directory
+ * @job:  related job
+ * @need_pipe_sync: is pipe sync needed
  *
  * Emit a VM flush when it is necessary.
+ *
+ * Returns:
+ * 0 on success, errno otherwise.
  */
 int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_pipe_sync)
 {
@@ -628,7 +734,7 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_
 		amdgpu_gmc_emit_pasid_mapping(ring, job->vmid, job->pasid);
 
 	if (vm_flush_needed || pasid_mapping_needed) {
-		r = amdgpu_fence_emit(ring, &fence);
+		r = amdgpu_fence_emit(ring, &fence, 0);
 		if (r)
 			return r;
 	}
@@ -684,6 +790,9 @@ int amdgpu_vm_flush(struct amdgpu_ring *ring, struct amdgpu_job *job, bool need_
  * Returns the found bo_va or NULL if none is found
  *
  * Object has to be reserved!
+ *
+ * Returns:
+ * Found bo_va or NULL.
  */
 struct amdgpu_bo_va *amdgpu_vm_bo_find(struct amdgpu_vm *vm,
 				       struct amdgpu_bo *bo)
@@ -765,7 +874,10 @@ static void amdgpu_vm_do_copy_ptes(struct amdgpu_pte_update_params *params,
  * @addr: the unmapped addr
  *
  * Look up the physical address of the page that the pte resolves
- * to and return the pointer for the page table entry.
+ * to.
+ *
+ * Returns:
+ * The pointer for the page table entry.
  */
 static uint64_t amdgpu_vm_map_gart(const dma_addr_t *pages_addr, uint64_t addr)
 {
@@ -818,6 +930,17 @@ static void amdgpu_vm_cpu_set_ptes(struct amdgpu_pte_update_params *params,
 	}
 }
 
+
+/**
+ * amdgpu_vm_wait_pd - Wait for PT BOs to be free.
+ *
+ * @adev: amdgpu_device pointer
+ * @vm: related vm
+ * @owner: fence owner
+ *
+ * Returns:
+ * 0 on success, errno otherwise.
+ */
 static int amdgpu_vm_wait_pd(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 			     void *owner)
 {
@@ -871,7 +994,10 @@ static void amdgpu_vm_update_pde(struct amdgpu_pte_update_params *params,
 /*
  * amdgpu_vm_invalidate_level - mark all PD levels as invalid
  *
+ * @adev: amdgpu_device pointer
+ * @vm: related vm
  * @parent: parent PD
+ * @level: VMPT level
  *
  * Mark all PD level as invalid after an error.
  */
@@ -893,10 +1019,8 @@ static void amdgpu_vm_invalidate_level(struct amdgpu_device *adev,
 		if (!entry->base.bo)
 			continue;
 
-		spin_lock(&vm->status_lock);
-		if (list_empty(&entry->base.vm_status))
-			list_add(&entry->base.vm_status, &vm->relocated);
-		spin_unlock(&vm->status_lock);
+		if (!entry->base.moved)
+			list_move(&entry->base.vm_status, &vm->relocated);
 		amdgpu_vm_invalidate_level(adev, vm, entry, level + 1);
 	}
 }
@@ -908,7 +1032,9 @@ static void amdgpu_vm_invalidate_level(struct amdgpu_device *adev,
  * @vm: requested vm
  *
  * Makes sure all directories are up to date.
- * Returns 0 for success, error for failure.
+ *
+ * Returns:
+ * 0 for success, error for failure.
  */
 int amdgpu_vm_update_directories(struct amdgpu_device *adev,
 				 struct amdgpu_vm *vm)
@@ -926,6 +1052,14 @@ restart:
 	params.adev = adev;
 
 	if (vm->use_cpu_for_update) {
+		struct amdgpu_vm_bo_base *bo_base;
+
+		list_for_each_entry(bo_base, &vm->relocated, vm_status) {
+			r = amdgpu_bo_kmap(bo_base->bo, NULL);
+			if (unlikely(r))
+				return r;
+		}
+
 		r = amdgpu_vm_wait_pd(adev, vm, AMDGPU_FENCE_OWNER_VM);
 		if (unlikely(r))
 			return r;
@@ -941,7 +1075,6 @@ restart:
 		params.func = amdgpu_vm_do_set_ptes;
 	}
 
-	spin_lock(&vm->status_lock);
 	while (!list_empty(&vm->relocated)) {
 		struct amdgpu_vm_bo_base *bo_base, *parent;
 		struct amdgpu_vm_pt *pt, *entry;
@@ -950,14 +1083,12 @@ restart:
 		bo_base = list_first_entry(&vm->relocated,
 					   struct amdgpu_vm_bo_base,
 					   vm_status);
+		bo_base->moved = false;
 		list_del_init(&bo_base->vm_status);
-		spin_unlock(&vm->status_lock);
 
 		bo = bo_base->bo->parent;
-		if (!bo) {
-			spin_lock(&vm->status_lock);
+		if (!bo)
 			continue;
-		}
 
 		parent = list_first_entry(&bo->va, struct amdgpu_vm_bo_base,
 					  bo_list);
@@ -966,12 +1097,10 @@ restart:
 
 		amdgpu_vm_update_pde(&params, vm, pt, entry);
 
-		spin_lock(&vm->status_lock);
 		if (!vm->use_cpu_for_update &&
 		    (ndw - params.ib->length_dw) < 32)
 			break;
 	}
-	spin_unlock(&vm->status_lock);
 
 	if (vm->use_cpu_for_update) {
 		/* Flush HDP */
@@ -991,8 +1120,8 @@ restart:
 		amdgpu_sync_resv(adev, &job->sync, root->tbo.resv,
 				 AMDGPU_FENCE_OWNER_VM, false);
 		WARN_ON(params.ib->length_dw > ndw);
-		r = amdgpu_job_submit(job, ring, &vm->entity,
-				      AMDGPU_FENCE_OWNER_VM, &fence);
+		r = amdgpu_job_submit(job, &vm->entity, AMDGPU_FENCE_OWNER_VM,
+				      &fence);
 		if (r)
 			goto error;
 
@@ -1074,9 +1203,7 @@ static void amdgpu_vm_handle_huge_pages(struct amdgpu_pte_update_params *p,
 		if (entry->huge) {
 			/* Add the entry to the relocated list to update it. */
 			entry->huge = false;
-			spin_lock(&p->vm->status_lock);
 			list_move(&entry->base.vm_status, &p->vm->relocated);
-			spin_unlock(&p->vm->status_lock);
 		}
 		return;
 	}
@@ -1094,14 +1221,15 @@ static void amdgpu_vm_handle_huge_pages(struct amdgpu_pte_update_params *p,
  * amdgpu_vm_update_ptes - make sure that page tables are valid
  *
  * @params: see amdgpu_pte_update_params definition
- * @vm: requested vm
  * @start: start of GPU address range
  * @end: end of GPU address range
  * @dst: destination address to map to, the next dst inside the function
  * @flags: mapping flags
  *
  * Update the page tables in the range @start - @end.
- * Returns 0 for success, -EINVAL for failure.
+ *
+ * Returns:
+ * 0 for success, -EINVAL for failure.
  */
 static int amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params,
 				  uint64_t start, uint64_t end,
@@ -1155,7 +1283,9 @@ static int amdgpu_vm_update_ptes(struct amdgpu_pte_update_params *params,
  * @end: last PTE to handle
  * @dst: addr those PTEs should point to
  * @flags: hw mapping flags
- * Returns 0 for success, -EINVAL for failure.
+ *
+ * Returns:
+ * 0 for success, -EINVAL for failure.
  */
 static int amdgpu_vm_frag_ptes(struct amdgpu_pte_update_params	*params,
 				uint64_t start, uint64_t end,
@@ -1227,7 +1357,9 @@ static int amdgpu_vm_frag_ptes(struct amdgpu_pte_update_params	*params,
  * @fence: optional resulting fence
  *
  * Fill in the page table entries between @start and @last.
- * Returns 0 for success, -EINVAL for failure.
+ *
+ * Returns:
+ * 0 for success, -EINVAL for failure.
  */
 static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev,
 				       struct dma_fence *exclusive,
@@ -1303,7 +1435,10 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev,
 		ndw += ncmds * 10;
 
 		/* extra commands for begin/end fragments */
-		ndw += 2 * 10 * adev->vm_manager.fragment_size;
+		if (vm->root.base.bo->shadow)
+		        ndw += 2 * 10 * adev->vm_manager.fragment_size * 2;
+		else
+		        ndw += 2 * 10 * adev->vm_manager.fragment_size;
 
 		params.func = amdgpu_vm_do_set_ptes;
 	}
@@ -1350,8 +1485,7 @@ static int amdgpu_vm_bo_update_mapping(struct amdgpu_device *adev,
 
 	amdgpu_ring_pad_ib(ring, params.ib);
 	WARN_ON(params.ib->length_dw > ndw);
-	r = amdgpu_job_submit(job, ring, &vm->entity,
-			      AMDGPU_FENCE_OWNER_VM, &f);
+	r = amdgpu_job_submit(job, &vm->entity, AMDGPU_FENCE_OWNER_VM, &f);
 	if (r)
 		goto error_free;
 
@@ -1379,7 +1513,9 @@ error_free:
  *
  * Split the mapping into smaller chunks so that each update fits
  * into a SDMA IB.
- * Returns 0 for success, -EINVAL for failure.
+ *
+ * Returns:
+ * 0 for success, -EINVAL for failure.
  */
 static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev,
 				      struct dma_fence *exclusive,
@@ -1432,7 +1568,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev,
 		if (nodes) {
 			addr = nodes->start << PAGE_SHIFT;
 			max_entries = (nodes->size - pfn) *
-				(PAGE_SIZE / AMDGPU_GPU_PAGE_SIZE);
+				AMDGPU_GPU_PAGES_IN_CPU_PAGE;
 		} else {
 			addr = 0;
 			max_entries = S64_MAX;
@@ -1442,7 +1578,9 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev,
 			uint64_t count;
 
 			max_entries = min(max_entries, 16ull * 1024ull);
-			for (count = 1; count < max_entries; ++count) {
+			for (count = 1;
+			     count < max_entries / AMDGPU_GPU_PAGES_IN_CPU_PAGE;
+			     ++count) {
 				uint64_t idx = pfn + count;
 
 				if (pages_addr[idx] !=
@@ -1455,7 +1593,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev,
 				dma_addr = pages_addr;
 			} else {
 				addr = pages_addr[pfn];
-				max_entries = count;
+				max_entries = count * AMDGPU_GPU_PAGES_IN_CPU_PAGE;
 			}
 
 		} else if (flags & AMDGPU_PTE_VALID) {
@@ -1470,7 +1608,7 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev,
 		if (r)
 			return r;
 
-		pfn += last - start + 1;
+		pfn += (last - start + 1) / AMDGPU_GPU_PAGES_IN_CPU_PAGE;
 		if (nodes && nodes->size == pfn) {
 			pfn = 0;
 			++nodes;
@@ -1490,7 +1628,9 @@ static int amdgpu_vm_bo_split_mapping(struct amdgpu_device *adev,
  * @clear: if true clear the entries
  *
  * Fill in the page table entries for @bo_va.
- * Returns 0 for success, -EINVAL for failure.
+ *
+ * Returns:
+ * 0 for success, -EINVAL for failure.
  */
 int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 			struct amdgpu_bo_va *bo_va,
@@ -1506,18 +1646,17 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 	uint64_t flags;
 	int r;
 
-	if (clear || !bo_va->base.bo) {
+	if (clear || !bo) {
 		mem = NULL;
 		nodes = NULL;
 		exclusive = NULL;
 	} else {
 		struct ttm_dma_tt *ttm;
 
-		mem = &bo_va->base.bo->tbo.mem;
+		mem = &bo->tbo.mem;
 		nodes = mem->mm_node;
 		if (mem->mem_type == TTM_PL_TT) {
-			ttm = container_of(bo_va->base.bo->tbo.ttm,
-					   struct ttm_dma_tt, ttm);
+			ttm = container_of(bo->tbo.ttm, struct ttm_dma_tt, ttm);
 			pages_addr = ttm->dma_address;
 		}
 		exclusive = reservation_object_get_excl(bo->tbo.resv);
@@ -1555,9 +1694,22 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 		amdgpu_asic_flush_hdp(adev, NULL);
 	}
 
-	spin_lock(&vm->status_lock);
+	spin_lock(&vm->moved_lock);
 	list_del_init(&bo_va->base.vm_status);
-	spin_unlock(&vm->status_lock);
+	spin_unlock(&vm->moved_lock);
+
+	/* If the BO is not in its preferred location add it back to
+	 * the evicted list so that it gets validated again on the
+	 * next command submission.
+	 */
+	if (bo && bo->tbo.resv == vm->root.base.bo->tbo.resv) {
+		uint32_t mem_type = bo->tbo.mem.mem_type;
+
+		if (!(bo->preferred_domains & amdgpu_mem_type_to_domain(mem_type)))
+			list_add_tail(&bo_va->base.vm_status, &vm->evicted);
+		else
+			list_add(&bo_va->base.vm_status, &vm->idle);
+	}
 
 	list_splice_init(&bo_va->invalids, &bo_va->valids);
 	bo_va->cleared = clear;
@@ -1572,6 +1724,8 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev,
 
 /**
  * amdgpu_vm_update_prt_state - update the global PRT state
+ *
+ * @adev: amdgpu_device pointer
  */
 static void amdgpu_vm_update_prt_state(struct amdgpu_device *adev)
 {
@@ -1586,6 +1740,8 @@ static void amdgpu_vm_update_prt_state(struct amdgpu_device *adev)
 
 /**
  * amdgpu_vm_prt_get - add a PRT user
+ *
+ * @adev: amdgpu_device pointer
  */
 static void amdgpu_vm_prt_get(struct amdgpu_device *adev)
 {
@@ -1598,6 +1754,8 @@ static void amdgpu_vm_prt_get(struct amdgpu_device *adev)
 
 /**
  * amdgpu_vm_prt_put - drop a PRT user
+ *
+ * @adev: amdgpu_device pointer
  */
 static void amdgpu_vm_prt_put(struct amdgpu_device *adev)
 {
@@ -1607,6 +1765,9 @@ static void amdgpu_vm_prt_put(struct amdgpu_device *adev)
 
 /**
  * amdgpu_vm_prt_cb - callback for updating the PRT status
+ *
+ * @fence: fence for the callback
+ * @_cb: the callback function
  */
 static void amdgpu_vm_prt_cb(struct dma_fence *fence, struct dma_fence_cb *_cb)
 {
@@ -1618,6 +1779,9 @@ static void amdgpu_vm_prt_cb(struct dma_fence *fence, struct dma_fence_cb *_cb)
 
 /**
  * amdgpu_vm_add_prt_cb - add callback for updating the PRT status
+ *
+ * @adev: amdgpu_device pointer
+ * @fence: fence for the callback
  */
 static void amdgpu_vm_add_prt_cb(struct amdgpu_device *adev,
 				 struct dma_fence *fence)
@@ -1709,9 +1873,11 @@ static void amdgpu_vm_prt_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm)
  * or if an error occurred)
  *
  * Make sure all freed BOs are cleared in the PT.
- * Returns 0 for success.
- *
  * PTs have to be reserved and mutex must be locked!
+ *
+ * Returns:
+ * 0 for success.
+ *
  */
 int amdgpu_vm_clear_freed(struct amdgpu_device *adev,
 			  struct amdgpu_vm *vm,
@@ -1756,29 +1922,29 @@ int amdgpu_vm_clear_freed(struct amdgpu_device *adev,
  *
  * @adev: amdgpu_device pointer
  * @vm: requested vm
- * @sync: sync object to add fences to
  *
  * Make sure all BOs which are moved are updated in the PTs.
- * Returns 0 for success.
+ *
+ * Returns:
+ * 0 for success.
  *
  * PTs have to be reserved!
  */
 int amdgpu_vm_handle_moved(struct amdgpu_device *adev,
 			   struct amdgpu_vm *vm)
 {
+	struct amdgpu_bo_va *bo_va, *tmp;
+	struct list_head moved;
 	bool clear;
-	int r = 0;
-
-	spin_lock(&vm->status_lock);
-	while (!list_empty(&vm->moved)) {
-		struct amdgpu_bo_va *bo_va;
-		struct reservation_object *resv;
+	int r;
 
-		bo_va = list_first_entry(&vm->moved,
-			struct amdgpu_bo_va, base.vm_status);
-		spin_unlock(&vm->status_lock);
+	INIT_LIST_HEAD(&moved);
+	spin_lock(&vm->moved_lock);
+	list_splice_init(&vm->moved, &moved);
+	spin_unlock(&vm->moved_lock);
 
-		resv = bo_va->base.bo->tbo.resv;
+	list_for_each_entry_safe(bo_va, tmp, &moved, base.vm_status) {
+		struct reservation_object *resv = bo_va->base.bo->tbo.resv;
 
 		/* Per VM BOs never need to bo cleared in the page tables */
 		if (resv == vm->root.base.bo->tbo.resv)
@@ -1791,17 +1957,19 @@ int amdgpu_vm_handle_moved(struct amdgpu_device *adev,
 			clear = true;
 
 		r = amdgpu_vm_bo_update(adev, bo_va, clear);
-		if (r)
+		if (r) {
+			spin_lock(&vm->moved_lock);
+			list_splice(&moved, &vm->moved);
+			spin_unlock(&vm->moved_lock);
 			return r;
+		}
 
 		if (!clear && resv != vm->root.base.bo->tbo.resv)
 			reservation_object_unlock(resv);
 
-		spin_lock(&vm->status_lock);
 	}
-	spin_unlock(&vm->status_lock);
 
-	return r;
+	return 0;
 }
 
 /**
@@ -1813,7 +1981,9 @@ int amdgpu_vm_handle_moved(struct amdgpu_device *adev,
  *
  * Add @bo into the requested vm.
  * Add @bo to the list of bos associated with the vm
- * Returns newly added bo_va or NULL for failure
+ *
+ * Returns:
+ * Newly added bo_va or NULL for failure
  *
  * Object has to be reserved!
  */
@@ -1827,36 +1997,12 @@ struct amdgpu_bo_va *amdgpu_vm_bo_add(struct amdgpu_device *adev,
 	if (bo_va == NULL) {
 		return NULL;
 	}
-	bo_va->base.vm = vm;
-	bo_va->base.bo = bo;
-	INIT_LIST_HEAD(&bo_va->base.bo_list);
-	INIT_LIST_HEAD(&bo_va->base.vm_status);
+	amdgpu_vm_bo_base_init(&bo_va->base, vm, bo);
 
 	bo_va->ref_count = 1;
 	INIT_LIST_HEAD(&bo_va->valids);
 	INIT_LIST_HEAD(&bo_va->invalids);
 
-	if (!bo)
-		return bo_va;
-
-	list_add_tail(&bo_va->base.bo_list, &bo->va);
-
-	if (bo->tbo.resv != vm->root.base.bo->tbo.resv)
-		return bo_va;
-
-	if (bo->preferred_domains &
-	    amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type))
-		return bo_va;
-
-	/*
-	 * We checked all the prerequisites, but it looks like this per VM BO
-	 * is currently evicted. add the BO to the evicted list to make sure it
-	 * is validated on next VM use to avoid fault.
-	 * */
-	spin_lock(&vm->status_lock);
-	list_move_tail(&bo_va->base.vm_status, &vm->evicted);
-	spin_unlock(&vm->status_lock);
-
 	return bo_va;
 }
 
@@ -1884,11 +2030,11 @@ static void amdgpu_vm_bo_insert_map(struct amdgpu_device *adev,
 	if (mapping->flags & AMDGPU_PTE_PRT)
 		amdgpu_vm_prt_get(adev);
 
-	if (bo && bo->tbo.resv == vm->root.base.bo->tbo.resv) {
-		spin_lock(&vm->status_lock);
-		if (list_empty(&bo_va->base.vm_status))
-			list_add(&bo_va->base.vm_status, &vm->moved);
-		spin_unlock(&vm->status_lock);
+	if (bo && bo->tbo.resv == vm->root.base.bo->tbo.resv &&
+	    !bo_va->base.moved) {
+		spin_lock(&vm->moved_lock);
+		list_move(&bo_va->base.vm_status, &vm->moved);
+		spin_unlock(&vm->moved_lock);
 	}
 	trace_amdgpu_vm_bo_map(bo_va, mapping);
 }
@@ -1900,10 +2046,13 @@ static void amdgpu_vm_bo_insert_map(struct amdgpu_device *adev,
  * @bo_va: bo_va to store the address
  * @saddr: where to map the BO
  * @offset: requested offset in the BO
+ * @size: BO size in bytes
  * @flags: attributes of pages (read/write/valid/etc.)
  *
  * Add a mapping of the BO at the specefied addr into the VM.
- * Returns 0 for success, error for failure.
+ *
+ * Returns:
+ * 0 for success, error for failure.
  *
  * Object has to be reserved and unreserved outside!
  */
@@ -1961,11 +2110,14 @@ int amdgpu_vm_bo_map(struct amdgpu_device *adev,
  * @bo_va: bo_va to store the address
  * @saddr: where to map the BO
  * @offset: requested offset in the BO
+ * @size: BO size in bytes
  * @flags: attributes of pages (read/write/valid/etc.)
  *
  * Add a mapping of the BO at the specefied addr into the VM. Replace existing
  * mappings as we do so.
- * Returns 0 for success, error for failure.
+ *
+ * Returns:
+ * 0 for success, error for failure.
  *
  * Object has to be reserved and unreserved outside!
  */
@@ -2022,7 +2174,9 @@ int amdgpu_vm_bo_replace_map(struct amdgpu_device *adev,
  * @saddr: where to the BO is mapped
  *
  * Remove a mapping of the BO at the specefied addr from the VM.
- * Returns 0 for success, error for failure.
+ *
+ * Returns:
+ * 0 for success, error for failure.
  *
  * Object has to be reserved and unreserved outside!
  */
@@ -2076,7 +2230,9 @@ int amdgpu_vm_bo_unmap(struct amdgpu_device *adev,
  * @size: size of the range
  *
  * Remove all mappings in a range, split them as appropriate.
- * Returns 0 for success, error for failure.
+ *
+ * Returns:
+ * 0 for success, error for failure.
  */
 int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
 				struct amdgpu_vm *vm,
@@ -2112,7 +2268,8 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
 			before->last = saddr - 1;
 			before->offset = tmp->offset;
 			before->flags = tmp->flags;
-			list_add(&before->list, &tmp->list);
+			before->bo_va = tmp->bo_va;
+			list_add(&before->list, &tmp->bo_va->invalids);
 		}
 
 		/* Remember mapping split at the end */
@@ -2122,7 +2279,8 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
 			after->offset = tmp->offset;
 			after->offset += after->start - tmp->start;
 			after->flags = tmp->flags;
-			list_add(&after->list, &tmp->list);
+			after->bo_va = tmp->bo_va;
+			list_add(&after->list, &tmp->bo_va->invalids);
 		}
 
 		list_del(&tmp->list);
@@ -2171,8 +2329,13 @@ int amdgpu_vm_bo_clear_mappings(struct amdgpu_device *adev,
  * amdgpu_vm_bo_lookup_mapping - find mapping by address
  *
  * @vm: the requested VM
+ * @addr: the address
  *
  * Find a mapping by it's address.
+ *
+ * Returns:
+ * The amdgpu_bo_va_mapping matching for addr or NULL
+ *
  */
 struct amdgpu_bo_va_mapping *amdgpu_vm_bo_lookup_mapping(struct amdgpu_vm *vm,
 							 uint64_t addr)
@@ -2198,9 +2361,9 @@ void amdgpu_vm_bo_rmv(struct amdgpu_device *adev,
 
 	list_del(&bo_va->base.bo_list);
 
-	spin_lock(&vm->status_lock);
+	spin_lock(&vm->moved_lock);
 	list_del(&bo_va->base.vm_status);
-	spin_unlock(&vm->status_lock);
+	spin_unlock(&vm->moved_lock);
 
 	list_for_each_entry_safe(mapping, next, &bo_va->valids, list) {
 		list_del(&mapping->list);
@@ -2224,8 +2387,8 @@ void amdgpu_vm_bo_rmv(struct amdgpu_device *adev,
  * amdgpu_vm_bo_invalidate - mark the bo as invalid
  *
  * @adev: amdgpu_device pointer
- * @vm: requested vm
  * @bo: amdgpu buffer object
+ * @evicted: is the BO evicted
  *
  * Mark @bo as invalid.
  */
@@ -2234,36 +2397,45 @@ void amdgpu_vm_bo_invalidate(struct amdgpu_device *adev,
 {
 	struct amdgpu_vm_bo_base *bo_base;
 
+	/* shadow bo doesn't have bo base, its validation needs its parent */
+	if (bo->parent && bo->parent->shadow == bo)
+		bo = bo->parent;
+
 	list_for_each_entry(bo_base, &bo->va, bo_list) {
 		struct amdgpu_vm *vm = bo_base->vm;
+		bool was_moved = bo_base->moved;
 
 		bo_base->moved = true;
 		if (evicted && bo->tbo.resv == vm->root.base.bo->tbo.resv) {
-			spin_lock(&bo_base->vm->status_lock);
 			if (bo->tbo.type == ttm_bo_type_kernel)
 				list_move(&bo_base->vm_status, &vm->evicted);
 			else
 				list_move_tail(&bo_base->vm_status,
 					       &vm->evicted);
-			spin_unlock(&bo_base->vm->status_lock);
 			continue;
 		}
 
-		if (bo->tbo.type == ttm_bo_type_kernel) {
-			spin_lock(&bo_base->vm->status_lock);
-			if (list_empty(&bo_base->vm_status))
-				list_add(&bo_base->vm_status, &vm->relocated);
-			spin_unlock(&bo_base->vm->status_lock);
+		if (was_moved)
 			continue;
-		}
 
-		spin_lock(&bo_base->vm->status_lock);
-		if (list_empty(&bo_base->vm_status))
-			list_add(&bo_base->vm_status, &vm->moved);
-		spin_unlock(&bo_base->vm->status_lock);
+		if (bo->tbo.type == ttm_bo_type_kernel) {
+			list_move(&bo_base->vm_status, &vm->relocated);
+		} else {
+			spin_lock(&bo_base->vm->moved_lock);
+			list_move(&bo_base->vm_status, &vm->moved);
+			spin_unlock(&bo_base->vm->moved_lock);
+		}
 	}
 }
 
+/**
+ * amdgpu_vm_get_block_size - calculate VM page table size as power of two
+ *
+ * @vm_size: VM size
+ *
+ * Returns:
+ * VM page table as power of two
+ */
 static uint32_t amdgpu_vm_get_block_size(uint64_t vm_size)
 {
 	/* Total bits covered by PD + PTs */
@@ -2282,6 +2454,10 @@ static uint32_t amdgpu_vm_get_block_size(uint64_t vm_size)
  *
  * @adev: amdgpu_device pointer
  * @vm_size: the default vm size if it's set auto
+ * @fragment_size_default: Default PTE fragment size
+ * @max_level: max VMPT level
+ * @max_bits: max address space size in bits
+ *
  */
 void amdgpu_vm_adjust_size(struct amdgpu_device *adev, uint32_t vm_size,
 			   uint32_t fragment_size_default, unsigned max_level,
@@ -2349,12 +2525,18 @@ void amdgpu_vm_adjust_size(struct amdgpu_device *adev, uint32_t vm_size,
  * @adev: amdgpu_device pointer
  * @vm: requested vm
  * @vm_context: Indicates if it GFX or Compute context
+ * @pasid: Process address space identifier
  *
  * Init @vm fields.
+ *
+ * Returns:
+ * 0 for success, error for failure.
  */
 int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		   int vm_context, unsigned int pasid)
 {
+	struct amdgpu_bo_param bp;
+	struct amdgpu_bo *root;
 	const unsigned align = min(AMDGPU_VM_PTB_ALIGN_SIZE,
 		AMDGPU_VM_PTE_COUNT(adev) * 8);
 	unsigned ring_instance;
@@ -2367,10 +2549,11 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	vm->va = RB_ROOT_CACHED;
 	for (i = 0; i < AMDGPU_MAX_VMHUBS; i++)
 		vm->reserved_vmid[i] = NULL;
-	spin_lock_init(&vm->status_lock);
 	INIT_LIST_HEAD(&vm->evicted);
 	INIT_LIST_HEAD(&vm->relocated);
+	spin_lock_init(&vm->moved_lock);
 	INIT_LIST_HEAD(&vm->moved);
+	INIT_LIST_HEAD(&vm->idle);
 	INIT_LIST_HEAD(&vm->freed);
 
 	/* create scheduler entity for page table updates */
@@ -2379,8 +2562,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	ring_instance %= adev->vm_manager.vm_pte_num_rings;
 	ring = adev->vm_manager.vm_pte_rings[ring_instance];
 	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_KERNEL];
-	r = drm_sched_entity_init(&ring->sched, &vm->entity,
-				  rq, amdgpu_sched_jobs, NULL);
+	r = drm_sched_entity_init(&vm->entity, &rq, 1, NULL);
 	if (r)
 		return r;
 
@@ -2398,7 +2580,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	}
 	DRM_DEBUG_DRIVER("VM update mode is %s\n",
 			 vm->use_cpu_for_update ? "CPU" : "SDMA");
-	WARN_ONCE((vm->use_cpu_for_update & !amdgpu_vm_is_large_bar(adev)),
+	WARN_ONCE((vm->use_cpu_for_update & !amdgpu_gmc_vram_full_visible(&adev->gmc)),
 		  "CPU update of VM recommended only for large BAR system\n");
 	vm->last_update = NULL;
 
@@ -2409,24 +2591,28 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		flags |= AMDGPU_GEM_CREATE_SHADOW;
 
 	size = amdgpu_vm_bo_size(adev, adev->vm_manager.root_level);
-	r = amdgpu_bo_create(adev, size, align, AMDGPU_GEM_DOMAIN_VRAM, flags,
-			     ttm_bo_type_kernel, NULL, &vm->root.base.bo);
+	memset(&bp, 0, sizeof(bp));
+	bp.size = size;
+	bp.byte_align = align;
+	bp.domain = AMDGPU_GEM_DOMAIN_VRAM;
+	bp.flags = flags;
+	bp.type = ttm_bo_type_kernel;
+	bp.resv = NULL;
+	r = amdgpu_bo_create(adev, &bp, &root);
 	if (r)
 		goto error_free_sched_entity;
 
-	r = amdgpu_bo_reserve(vm->root.base.bo, true);
+	r = amdgpu_bo_reserve(root, true);
 	if (r)
 		goto error_free_root;
 
-	r = amdgpu_vm_clear_bo(adev, vm, vm->root.base.bo,
+	r = amdgpu_vm_clear_bo(adev, vm, root,
 			       adev->vm_manager.root_level,
 			       vm->pte_support_ats);
 	if (r)
 		goto error_unreserve;
 
-	vm->root.base.vm = vm;
-	list_add_tail(&vm->root.base.bo_list, &vm->root.base.bo->va);
-	list_add_tail(&vm->root.base.vm_status, &vm->evicted);
+	amdgpu_vm_bo_base_init(&vm->root.base, vm, root);
 	amdgpu_bo_unreserve(vm->root.base.bo);
 
 	if (pasid) {
@@ -2456,7 +2642,7 @@ error_free_root:
 	vm->root.base.bo = NULL;
 
 error_free_sched_entity:
-	drm_sched_entity_fini(&ring->sched, &vm->entity);
+	drm_sched_entity_destroy(&ring->sched, &vm->entity);
 
 	return r;
 }
@@ -2464,6 +2650,9 @@ error_free_sched_entity:
 /**
  * amdgpu_vm_make_compute - Turn a GFX VM into a compute VM
  *
+ * @adev: amdgpu_device pointer
+ * @vm: requested vm
+ *
  * This only works on GFX VMs that don't have any BOs added and no
  * page tables allocated yet.
  *
@@ -2476,7 +2665,8 @@ error_free_sched_entity:
  * setting. May leave behind an unused shadow BO for the page
  * directory when switching from SDMA updates to CPU updates.
  *
- * Returns 0 for success, -errno for errors.
+ * Returns:
+ * 0 for success, -errno for errors.
  */
 int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm)
 {
@@ -2510,7 +2700,7 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm)
 	vm->pte_support_ats = pte_support_ats;
 	DRM_DEBUG_DRIVER("VM update mode is %s\n",
 			 vm->use_cpu_for_update ? "CPU" : "SDMA");
-	WARN_ONCE((vm->use_cpu_for_update & !amdgpu_vm_is_large_bar(adev)),
+	WARN_ONCE((vm->use_cpu_for_update & !amdgpu_gmc_vram_full_visible(&adev->gmc)),
 		  "CPU update of VM recommended only for large BAR system\n");
 
 	if (vm->pasid) {
@@ -2589,7 +2779,7 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm)
 		spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, flags);
 	}
 
-	drm_sched_entity_fini(vm->entity.sched, &vm->entity);
+	drm_sched_entity_destroy(vm->entity.sched, &vm->entity);
 
 	if (!RB_EMPTY_ROOT(&vm->va.rb_root)) {
 		dev_err(adev->dev, "still active bo inside vm\n");
@@ -2631,8 +2821,10 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm)
  * @adev: amdgpu_device pointer
  * @pasid: PASID do identify the VM
  *
- * This function is expected to be called in interrupt context. Returns
- * true if there was fault credit, false otherwise
+ * This function is expected to be called in interrupt context.
+ *
+ * Returns:
+ * True if there was fault credit, false otherwise
  */
 bool amdgpu_vm_pasid_fault_credit(struct amdgpu_device *adev,
 				  unsigned int pasid)
@@ -2686,7 +2878,7 @@ void amdgpu_vm_manager_init(struct amdgpu_device *adev)
 	 */
 #ifdef CONFIG_X86_64
 	if (amdgpu_vm_update_mode == -1) {
-		if (amdgpu_vm_is_large_bar(adev))
+		if (amdgpu_gmc_vram_full_visible(&adev->gmc))
 			adev->vm_manager.vm_update_mode =
 				AMDGPU_VM_USE_CPU_FOR_COMPUTE;
 		else
@@ -2716,6 +2908,16 @@ void amdgpu_vm_manager_fini(struct amdgpu_device *adev)
 	amdgpu_vmid_mgr_fini(adev);
 }
 
+/**
+ * amdgpu_vm_ioctl - Manages VMID reservation for vm hubs.
+ *
+ * @dev: drm device pointer
+ * @data: drm_amdgpu_vm
+ * @filp: drm file pointer
+ *
+ * Returns:
+ * 0 for success, -errno for errors.
+ */
 int amdgpu_vm_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 {
 	union drm_amdgpu_vm *args = data;
@@ -2739,3 +2941,42 @@ int amdgpu_vm_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 
 	return 0;
 }
+
+/**
+ * amdgpu_vm_get_task_info - Extracts task info for a PASID.
+ *
+ * @dev: drm device pointer
+ * @pasid: PASID identifier for VM
+ * @task_info: task_info to fill.
+ */
+void amdgpu_vm_get_task_info(struct amdgpu_device *adev, unsigned int pasid,
+			 struct amdgpu_task_info *task_info)
+{
+	struct amdgpu_vm *vm;
+
+	spin_lock(&adev->vm_manager.pasid_lock);
+
+	vm = idr_find(&adev->vm_manager.pasid_idr, pasid);
+	if (vm)
+		*task_info = vm->task_info;
+
+	spin_unlock(&adev->vm_manager.pasid_lock);
+}
+
+/**
+ * amdgpu_vm_set_task_info - Sets VMs task info.
+ *
+ * @vm: vm for which to set the info
+ */
+void amdgpu_vm_set_task_info(struct amdgpu_vm *vm)
+{
+	if (!vm->task_info.pid) {
+		vm->task_info.pid = current->pid;
+		get_task_comm(vm->task_info.task_name, current);
+
+		if (current->group_leader->mm == current->mm) {
+			vm->task_info.tgid = current->group_leader->pid;
+			get_task_comm(vm->task_info.process_name, current->group_leader);
+		}
+	}
+}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
index 30f080364c97..d416f895233d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h
@@ -75,11 +75,12 @@ struct amdgpu_bo_list_entry;
 /* PDE Block Fragment Size for VEGA10 */
 #define AMDGPU_PDE_BFS(a)	((uint64_t)a << 59)
 
-/* VEGA10 only */
+
+/* For GFX9 */
 #define AMDGPU_PTE_MTYPE(a)    ((uint64_t)a << 57)
 #define AMDGPU_PTE_MTYPE_MASK	AMDGPU_PTE_MTYPE(3ULL)
 
-/* For Raven */
+#define AMDGPU_MTYPE_NC 0
 #define AMDGPU_MTYPE_CC 2
 
 #define AMDGPU_PTE_DEFAULT_ATC  (AMDGPU_PTE_SYSTEM      \
@@ -163,13 +164,18 @@ struct amdgpu_vm_pt {
 #define AMDGPU_VM_FAULT_PASID(fault) ((u64)(fault) >> 48)
 #define AMDGPU_VM_FAULT_ADDR(fault)  ((u64)(fault) & 0xfffffffff000ULL)
 
+
+struct amdgpu_task_info {
+	char	process_name[TASK_COMM_LEN];
+	char	task_name[TASK_COMM_LEN];
+	pid_t	pid;
+	pid_t	tgid;
+};
+
 struct amdgpu_vm {
 	/* tree of virtual addresses mapped */
 	struct rb_root_cached	va;
 
-	/* protecting invalidated */
-	spinlock_t		status_lock;
-
 	/* BOs who needs a validation */
 	struct list_head	evicted;
 
@@ -178,6 +184,10 @@ struct amdgpu_vm {
 
 	/* BOs moved, but not yet updated in the PT */
 	struct list_head	moved;
+	spinlock_t		moved_lock;
+
+	/* All BOs of this VM not currently in the state machine */
+	struct list_head	idle;
 
 	/* BO mappings freed, but not yet updated in the PT */
 	struct list_head	freed;
@@ -186,9 +196,6 @@ struct amdgpu_vm {
 	struct amdgpu_vm_pt     root;
 	struct dma_fence	*last_update;
 
-	/* protecting freed */
-	spinlock_t		freed_lock;
-
 	/* Scheduler entity for page table updates */
 	struct drm_sched_entity	entity;
 
@@ -216,6 +223,9 @@ struct amdgpu_vm {
 
 	/* Valid while the PD is reserved or fenced */
 	uint64_t		pd_phys_addr;
+
+	/* Some basic info about the task */
+	struct amdgpu_task_info task_info;
 };
 
 struct amdgpu_vm_manager {
@@ -318,4 +328,9 @@ bool amdgpu_vm_need_pipeline_sync(struct amdgpu_ring *ring,
 				  struct amdgpu_job *job);
 void amdgpu_vm_check_compute_bug(struct amdgpu_device *adev);
 
+void amdgpu_vm_get_task_info(struct amdgpu_device *adev, unsigned int pasid,
+			 struct amdgpu_task_info *task_info);
+
+void amdgpu_vm_set_task_info(struct amdgpu_vm *vm);
+
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
index 9aca653bec07..9cfa8a9ada92 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vram_mgr.c
@@ -97,6 +97,34 @@ static u64 amdgpu_vram_mgr_vis_size(struct amdgpu_device *adev,
 }
 
 /**
+ * amdgpu_vram_mgr_bo_visible_size - CPU visible BO size
+ *
+ * @bo: &amdgpu_bo buffer object (must be in VRAM)
+ *
+ * Returns:
+ * How much of the given &amdgpu_bo buffer object lies in CPU visible VRAM.
+ */
+u64 amdgpu_vram_mgr_bo_visible_size(struct amdgpu_bo *bo)
+{
+	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
+	struct ttm_mem_reg *mem = &bo->tbo.mem;
+	struct drm_mm_node *nodes = mem->mm_node;
+	unsigned pages = mem->num_pages;
+	u64 usage;
+
+	if (amdgpu_gmc_vram_full_visible(&adev->gmc))
+		return amdgpu_bo_size(bo);
+
+	if (mem->start >= adev->gmc.visible_vram_size >> PAGE_SHIFT)
+		return 0;
+
+	for (usage = 0; nodes && pages; pages -= nodes->size, nodes++)
+		usage += amdgpu_vram_mgr_vis_size(adev, nodes);
+
+	return usage;
+}
+
+/**
  * amdgpu_vram_mgr_new - allocate new ranges
  *
  * @man: TTM memory type manager
@@ -135,7 +163,8 @@ static int amdgpu_vram_mgr_new(struct ttm_mem_type_manager *man,
 		num_nodes = DIV_ROUND_UP(mem->num_pages, pages_per_node);
 	}
 
-	nodes = kcalloc(num_nodes, sizeof(*nodes), GFP_KERNEL);
+	nodes = kvmalloc_array(num_nodes, sizeof(*nodes),
+			       GFP_KERNEL | __GFP_ZERO);
 	if (!nodes)
 		return -ENOMEM;
 
@@ -190,7 +219,7 @@ error:
 		drm_mm_remove_node(&nodes[i]);
 	spin_unlock(&mgr->lock);
 
-	kfree(nodes);
+	kvfree(nodes);
 	return r == -ENOSPC ? 0 : r;
 }
 
@@ -229,7 +258,7 @@ static void amdgpu_vram_mgr_del(struct ttm_mem_type_manager *man,
 	atomic64_sub(usage, &mgr->usage);
 	atomic64_sub(vis_usage, &mgr->vis_usage);
 
-	kfree(mem->mm_node);
+	kvfree(mem->mm_node);
 	mem->mm_node = NULL;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/atom.c b/drivers/gpu/drm/amd/amdgpu/atom.c
index 69500a8b4e2d..e9934de1b9cf 100644
--- a/drivers/gpu/drm/amd/amdgpu/atom.c
+++ b/drivers/gpu/drm/amd/amdgpu/atom.c
@@ -1221,7 +1221,7 @@ static int amdgpu_atom_execute_table_locked(struct atom_context *ctx, int index,
 	ectx.abort = false;
 	ectx.last_jump = 0;
 	if (ws)
-		ectx.ws = kzalloc(4 * ws, GFP_KERNEL);
+		ectx.ws = kcalloc(4, ws, GFP_KERNEL);
 	else
 		ectx.ws = NULL;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
index 47ef3e6e7178..d2469453dca2 100644
--- a/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/ci_dpm.c
@@ -49,10 +49,10 @@
 #include "gmc/gmc_7_1_d.h"
 #include "gmc/gmc_7_1_sh_mask.h"
 
-MODULE_FIRMWARE("radeon/bonaire_smc.bin");
-MODULE_FIRMWARE("radeon/bonaire_k_smc.bin");
-MODULE_FIRMWARE("radeon/hawaii_smc.bin");
-MODULE_FIRMWARE("radeon/hawaii_k_smc.bin");
+MODULE_FIRMWARE("amdgpu/bonaire_smc.bin");
+MODULE_FIRMWARE("amdgpu/bonaire_k_smc.bin");
+MODULE_FIRMWARE("amdgpu/hawaii_smc.bin");
+MODULE_FIRMWARE("amdgpu/hawaii_k_smc.bin");
 
 #define MC_CG_ARB_FREQ_F0           0x0a
 #define MC_CG_ARB_FREQ_F1           0x0b
@@ -951,12 +951,12 @@ static void ci_apply_state_adjust_rules(struct amdgpu_device *adev,
 	else
 		pi->battery_state = false;
 
-	if (adev->pm.dpm.ac_power)
+	if (adev->pm.ac_power)
 		max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_ac;
 	else
 		max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_dc;
 
-	if (adev->pm.dpm.ac_power == false) {
+	if (adev->pm.ac_power == false) {
 		for (i = 0; i < ps->performance_level_count; i++) {
 			if (ps->performance_levels[i].mclk > max_limits->mclk)
 				ps->performance_levels[i].mclk = max_limits->mclk;
@@ -4078,7 +4078,7 @@ static int ci_enable_uvd_dpm(struct amdgpu_device *adev, bool enable)
 	const struct amdgpu_clock_and_voltage_limits *max_limits;
 	int i;
 
-	if (adev->pm.dpm.ac_power)
+	if (adev->pm.ac_power)
 		max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_ac;
 	else
 		max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_dc;
@@ -4127,7 +4127,7 @@ static int ci_enable_vce_dpm(struct amdgpu_device *adev, bool enable)
 	const struct amdgpu_clock_and_voltage_limits *max_limits;
 	int i;
 
-	if (adev->pm.dpm.ac_power)
+	if (adev->pm.ac_power)
 		max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_ac;
 	else
 		max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_dc;
@@ -4160,7 +4160,7 @@ static int ci_enable_samu_dpm(struct amdgpu_device *adev, bool enable)
 	const struct amdgpu_clock_and_voltage_limits *max_limits;
 	int i;
 
-	if (adev->pm.dpm.ac_power)
+	if (adev->pm.ac_power)
 		max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_ac;
 	else
 		max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_dc;
@@ -4191,7 +4191,7 @@ static int ci_enable_acp_dpm(struct amdgpu_device *adev, bool enable)
 	const struct amdgpu_clock_and_voltage_limits *max_limits;
 	int i;
 
-	if (adev->pm.dpm.ac_power)
+	if (adev->pm.ac_power)
 		max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_ac;
 	else
 		max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_dc;
@@ -5679,8 +5679,9 @@ static int ci_parse_power_table(struct amdgpu_device *adev)
 		(mode_info->atom_context->bios + data_offset +
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 
-	adev->pm.dpm.ps = kzalloc(sizeof(struct amdgpu_ps) *
-				  state_array->ucNumEntries, GFP_KERNEL);
+	adev->pm.dpm.ps = kcalloc(state_array->ucNumEntries,
+				  sizeof(struct amdgpu_ps),
+				  GFP_KERNEL);
 	if (!adev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
@@ -5814,7 +5815,7 @@ static int ci_dpm_init_microcode(struct amdgpu_device *adev)
 	default: BUG();
 	}
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", chip_name);
+	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_smc.bin", chip_name);
 	err = request_firmware(&adev->pm.fw, fw_name, adev->dev);
 	if (err)
 		goto out;
@@ -5845,8 +5846,7 @@ static int ci_dpm_init(struct amdgpu_device *adev)
 	adev->pm.dpm.priv = pi;
 
 	pi->sys_pcie_mask =
-		(adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_MASK) >>
-		CAIL_PCIE_LINK_SPEED_SUPPORT_SHIFT;
+		adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_MASK;
 
 	pi->force_pcie_gen = AMDGPU_PCIE_GEN_INVALID;
 
@@ -5903,7 +5903,7 @@ static int ci_dpm_init(struct amdgpu_device *adev)
 	pi->pcie_dpm_key_disabled = 0;
 	pi->thermal_sclk_dpm_enabled = 0;
 
-	if (amdgpu_pp_feature_mask & SCLK_DEEP_SLEEP_MASK)
+	if (adev->powerplay.pp_feature & PP_SCLK_DEEP_SLEEP_MASK)
 		pi->caps_sclk_ds = true;
 	else
 		pi->caps_sclk_ds = false;
@@ -5927,7 +5927,9 @@ static int ci_dpm_init(struct amdgpu_device *adev)
 	ci_set_private_data_variables_based_on_pptable(adev);
 
 	adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries =
-		kzalloc(4 * sizeof(struct amdgpu_clock_voltage_dependency_entry), GFP_KERNEL);
+		kcalloc(4,
+			sizeof(struct amdgpu_clock_voltage_dependency_entry),
+			GFP_KERNEL);
 	if (!adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries) {
 		ci_dpm_fini(adev);
 		return -ENOMEM;
@@ -6255,7 +6257,7 @@ static int ci_dpm_late_init(void *handle)
 	int ret;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!amdgpu_dpm)
+	if (!adev->pm.dpm_enabled)
 		return 0;
 
 	/* init the sysfs and debugfs files late */
@@ -6764,6 +6766,19 @@ static int ci_dpm_read_sensor(void *handle, int idx,
 	}
 }
 
+static int ci_set_powergating_by_smu(void *handle,
+				uint32_t block_type, bool gate)
+{
+	switch (block_type) {
+	case AMD_IP_BLOCK_TYPE_UVD:
+		ci_dpm_powergate_uvd(handle, gate);
+		break;
+	default:
+		break;
+	}
+	return 0;
+}
+
 static const struct amd_ip_funcs ci_dpm_ip_funcs = {
 	.name = "ci_dpm",
 	.early_init = ci_dpm_early_init,
@@ -6801,7 +6816,7 @@ static const struct amd_pm_funcs ci_dpm_funcs = {
 	.debugfs_print_current_performance_level = &ci_dpm_debugfs_print_current_performance_level,
 	.force_performance_level = &ci_dpm_force_performance_level,
 	.vblank_too_short = &ci_dpm_vblank_too_short,
-	.powergate_uvd = &ci_dpm_powergate_uvd,
+	.set_powergating_by_smu = &ci_set_powergating_by_smu,
 	.set_fan_control_mode = &ci_dpm_set_fan_control_mode,
 	.get_fan_control_mode = &ci_dpm_get_fan_control_mode,
 	.set_fan_speed_percent = &ci_dpm_set_fan_speed_percent,
diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c
index 0df22030e713..702e257a483f 100644
--- a/drivers/gpu/drm/amd/amdgpu/cik.c
+++ b/drivers/gpu/drm/amd/amdgpu/cik.c
@@ -1735,6 +1735,12 @@ static void cik_invalidate_hdp(struct amdgpu_device *adev,
 	}
 }
 
+static bool cik_need_full_reset(struct amdgpu_device *adev)
+{
+	/* change this when we support soft reset */
+	return true;
+}
+
 static const struct amdgpu_asic_funcs cik_asic_funcs =
 {
 	.read_disabled_bios = &cik_read_disabled_bios,
@@ -1748,6 +1754,7 @@ static const struct amdgpu_asic_funcs cik_asic_funcs =
 	.get_config_memsize = &cik_get_config_memsize,
 	.flush_hdp = &cik_flush_hdp,
 	.invalidate_hdp = &cik_invalidate_hdp,
+	.need_full_reset = &cik_need_full_reset,
 };
 
 static int cik_common_early_init(void *handle)
@@ -1996,9 +2003,9 @@ int cik_set_ip_blocks(struct amdgpu_device *adev)
 		amdgpu_device_ip_block_add(adev, &gmc_v7_0_ip_block);
 		amdgpu_device_ip_block_add(adev, &cik_ih_ip_block);
 		if (amdgpu_dpm == -1)
-			amdgpu_device_ip_block_add(adev, &ci_smu_ip_block);
-		else
 			amdgpu_device_ip_block_add(adev, &pp_smu_ip_block);
+		else
+			amdgpu_device_ip_block_add(adev, &ci_smu_ip_block);
 		if (adev->enable_virtual_display)
 			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
@@ -2017,9 +2024,9 @@ int cik_set_ip_blocks(struct amdgpu_device *adev)
 		amdgpu_device_ip_block_add(adev, &gmc_v7_0_ip_block);
 		amdgpu_device_ip_block_add(adev, &cik_ih_ip_block);
 		if (amdgpu_dpm == -1)
-			amdgpu_device_ip_block_add(adev, &ci_smu_ip_block);
-		else
 			amdgpu_device_ip_block_add(adev, &pp_smu_ip_block);
+		else
+			amdgpu_device_ip_block_add(adev, &ci_smu_ip_block);
 		if (adev->enable_virtual_display)
 			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
diff --git a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c
index a7576255cc30..d0fa2aac2388 100644
--- a/drivers/gpu/drm/amd/amdgpu/cik_sdma.c
+++ b/drivers/gpu/drm/amd/amdgpu/cik_sdma.c
@@ -54,16 +54,16 @@ static void cik_sdma_set_buffer_funcs(struct amdgpu_device *adev);
 static void cik_sdma_set_vm_pte_funcs(struct amdgpu_device *adev);
 static int cik_sdma_soft_reset(void *handle);
 
-MODULE_FIRMWARE("radeon/bonaire_sdma.bin");
-MODULE_FIRMWARE("radeon/bonaire_sdma1.bin");
-MODULE_FIRMWARE("radeon/hawaii_sdma.bin");
-MODULE_FIRMWARE("radeon/hawaii_sdma1.bin");
-MODULE_FIRMWARE("radeon/kaveri_sdma.bin");
-MODULE_FIRMWARE("radeon/kaveri_sdma1.bin");
-MODULE_FIRMWARE("radeon/kabini_sdma.bin");
-MODULE_FIRMWARE("radeon/kabini_sdma1.bin");
-MODULE_FIRMWARE("radeon/mullins_sdma.bin");
-MODULE_FIRMWARE("radeon/mullins_sdma1.bin");
+MODULE_FIRMWARE("amdgpu/bonaire_sdma.bin");
+MODULE_FIRMWARE("amdgpu/bonaire_sdma1.bin");
+MODULE_FIRMWARE("amdgpu/hawaii_sdma.bin");
+MODULE_FIRMWARE("amdgpu/hawaii_sdma1.bin");
+MODULE_FIRMWARE("amdgpu/kaveri_sdma.bin");
+MODULE_FIRMWARE("amdgpu/kaveri_sdma1.bin");
+MODULE_FIRMWARE("amdgpu/kabini_sdma.bin");
+MODULE_FIRMWARE("amdgpu/kabini_sdma1.bin");
+MODULE_FIRMWARE("amdgpu/mullins_sdma.bin");
+MODULE_FIRMWARE("amdgpu/mullins_sdma1.bin");
 
 u32 amdgpu_cik_gpu_check_soft_reset(struct amdgpu_device *adev);
 
@@ -132,9 +132,9 @@ static int cik_sdma_init_microcode(struct amdgpu_device *adev)
 
 	for (i = 0; i < adev->sdma.num_instances; i++) {
 		if (i == 0)
-			snprintf(fw_name, sizeof(fw_name), "radeon/%s_sdma.bin", chip_name);
+			snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_sdma.bin", chip_name);
 		else
-			snprintf(fw_name, sizeof(fw_name), "radeon/%s_sdma1.bin", chip_name);
+			snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_sdma1.bin", chip_name);
 		err = request_firmware(&adev->sdma.instance[i].fw, fw_name, adev->dev);
 		if (err)
 			goto out;
@@ -177,9 +177,8 @@ static uint64_t cik_sdma_ring_get_rptr(struct amdgpu_ring *ring)
 static uint64_t cik_sdma_ring_get_wptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
-	u32 me = (ring == &adev->sdma.instance[0].ring) ? 0 : 1;
 
-	return (RREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[me]) & 0x3fffc) >> 2;
+	return (RREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[ring->me]) & 0x3fffc) >> 2;
 }
 
 /**
@@ -192,9 +191,8 @@ static uint64_t cik_sdma_ring_get_wptr(struct amdgpu_ring *ring)
 static void cik_sdma_ring_set_wptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
-	u32 me = (ring == &adev->sdma.instance[0].ring) ? 0 : 1;
 
-	WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[me],
+	WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[ring->me],
 		       	(lower_32_bits(ring->wptr) << 2) & 0x3fffc);
 }
 
@@ -248,7 +246,7 @@ static void cik_sdma_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 			  SDMA_POLL_REG_MEM_EXTRA_FUNC(3)); /* == */
 	u32 ref_and_mask;
 
-	if (ring == &ring->adev->sdma.instance[0].ring)
+	if (ring->me == 0)
 		ref_and_mask = GPU_HDP_FLUSH_DONE__SDMA0_MASK;
 	else
 		ref_and_mask = GPU_HDP_FLUSH_DONE__SDMA1_MASK;
@@ -1290,8 +1288,10 @@ static void cik_sdma_set_ring_funcs(struct amdgpu_device *adev)
 {
 	int i;
 
-	for (i = 0; i < adev->sdma.num_instances; i++)
+	for (i = 0; i < adev->sdma.num_instances; i++) {
 		adev->sdma.instance[i].ring.funcs = &cik_sdma_ring_funcs;
+		adev->sdma.instance[i].ring.me = i;
+	}
 }
 
 static const struct amdgpu_irq_src_funcs cik_sdma_trap_irq_funcs = {
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
index 452f88ea46a2..308f9f238bc1 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v10_0.c
@@ -41,6 +41,8 @@
 #include "gmc/gmc_8_1_d.h"
 #include "gmc/gmc_8_1_sh_mask.h"
 
+#include "ivsrcid/ivsrcid_vislands30.h"
+
 static void dce_v10_0_set_display_funcs(struct amdgpu_device *adev);
 static void dce_v10_0_set_irq_funcs(struct amdgpu_device *adev);
 
@@ -1823,7 +1825,6 @@ static int dce_v10_0_crtc_do_set_base(struct drm_crtc *crtc,
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct amdgpu_device *adev = dev->dev_private;
-	struct amdgpu_framebuffer *amdgpu_fb;
 	struct drm_framebuffer *target_fb;
 	struct drm_gem_object *obj;
 	struct amdgpu_bo *abo;
@@ -1842,32 +1843,28 @@ static int dce_v10_0_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	} else {
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
 	/* If atomic, assume fb object is pinned & idle & fenced and
 	 * just update base pointers
 	 */
-	obj = amdgpu_fb->obj;
+	obj = target_fb->obj[0];
 	abo = gem_to_amdgpu_bo(obj);
 	r = amdgpu_bo_reserve(abo, false);
 	if (unlikely(r != 0))
 		return r;
 
-	if (atomic) {
-		fb_location = amdgpu_bo_gpu_offset(abo);
-	} else {
-		r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM, &fb_location);
+	if (!atomic) {
+		r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM);
 		if (unlikely(r != 0)) {
 			amdgpu_bo_unreserve(abo);
 			return -EINVAL;
 		}
 	}
+	fb_location = amdgpu_bo_gpu_offset(abo);
 
 	amdgpu_bo_get_tiling_flags(abo, &tiling_flags);
 	amdgpu_bo_unreserve(abo);
@@ -2043,8 +2040,7 @@ static int dce_v10_0_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(mmMASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 0);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r != 0))
 			return r;
@@ -2375,13 +2371,14 @@ static int dce_v10_0_crtc_cursor_set2(struct drm_crtc *crtc,
 		return ret;
 	}
 
-	ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM, &amdgpu_crtc->cursor_addr);
+	ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM);
 	amdgpu_bo_unreserve(aobj);
 	if (ret) {
 		DRM_ERROR("Failed to pin new cursor BO (%d)\n", ret);
 		drm_gem_object_put_unlocked(obj);
 		return ret;
 	}
+	amdgpu_crtc->cursor_addr = amdgpu_bo_gpu_offset(aobj);
 
 	dce_v10_0_lock_cursor(crtc, true);
 
@@ -2526,11 +2523,9 @@ static void dce_v10_0_crtc_disable(struct drm_crtc *crtc)
 	dce_v10_0_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct amdgpu_framebuffer *amdgpu_fb;
 		struct amdgpu_bo *abo;
 
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(crtc->primary->fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve abo before unpin\n");
@@ -2744,14 +2739,14 @@ static int dce_v10_0_sw_init(void *handle)
 			return r;
 	}
 
-	for (i = 8; i < 20; i += 2) {
+	for (i = VISLANDS30_IV_SRCID_D1_GRPH_PFLIP; i < 20; i += 2) {
 		r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, i, &adev->pageflip_irq);
 		if (r)
 			return r;
 	}
 
 	/* HPD hotplug */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 42, &adev->hpd_irq);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_HOTPLUG_DETECT_A, &adev->hpd_irq);
 	if (r)
 		return r;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
index a7c1c584a191..76dfb76f7900 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c
@@ -41,6 +41,8 @@
 #include "gmc/gmc_8_1_d.h"
 #include "gmc/gmc_8_1_sh_mask.h"
 
+#include "ivsrcid/ivsrcid_vislands30.h"
+
 static void dce_v11_0_set_display_funcs(struct amdgpu_device *adev);
 static void dce_v11_0_set_irq_funcs(struct amdgpu_device *adev);
 
@@ -173,6 +175,7 @@ static void dce_v11_0_init_golden_registers(struct amdgpu_device *adev)
 							ARRAY_SIZE(polaris11_golden_settings_a11));
 		break;
 	case CHIP_POLARIS10:
+	case CHIP_VEGAM:
 		amdgpu_device_program_register_sequence(adev,
 							polaris10_golden_settings_a11,
 							ARRAY_SIZE(polaris10_golden_settings_a11));
@@ -473,6 +476,7 @@ static int dce_v11_0_get_num_crtc (struct amdgpu_device *adev)
 		num_crtc = 2;
 		break;
 	case CHIP_POLARIS10:
+	case CHIP_VEGAM:
 		num_crtc = 6;
 		break;
 	case CHIP_POLARIS11:
@@ -1445,6 +1449,7 @@ static int dce_v11_0_audio_init(struct amdgpu_device *adev)
 		adev->mode_info.audio.num_pins = 7;
 		break;
 	case CHIP_POLARIS10:
+	case CHIP_VEGAM:
 		adev->mode_info.audio.num_pins = 8;
 		break;
 	case CHIP_POLARIS11:
@@ -1862,7 +1867,6 @@ static int dce_v11_0_crtc_do_set_base(struct drm_crtc *crtc,
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct amdgpu_device *adev = dev->dev_private;
-	struct amdgpu_framebuffer *amdgpu_fb;
 	struct drm_framebuffer *target_fb;
 	struct drm_gem_object *obj;
 	struct amdgpu_bo *abo;
@@ -1881,32 +1885,28 @@ static int dce_v11_0_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	} else {
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
 	/* If atomic, assume fb object is pinned & idle & fenced and
 	 * just update base pointers
 	 */
-	obj = amdgpu_fb->obj;
+	obj = target_fb->obj[0];
 	abo = gem_to_amdgpu_bo(obj);
 	r = amdgpu_bo_reserve(abo, false);
 	if (unlikely(r != 0))
 		return r;
 
-	if (atomic) {
-		fb_location = amdgpu_bo_gpu_offset(abo);
-	} else {
-		r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM, &fb_location);
+	if (!atomic) {
+		r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM);
 		if (unlikely(r != 0)) {
 			amdgpu_bo_unreserve(abo);
 			return -EINVAL;
 		}
 	}
+	fb_location = amdgpu_bo_gpu_offset(abo);
 
 	amdgpu_bo_get_tiling_flags(abo, &tiling_flags);
 	amdgpu_bo_unreserve(abo);
@@ -2082,8 +2082,7 @@ static int dce_v11_0_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(mmCRTC_MASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 0);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r != 0))
 			return r;
@@ -2253,7 +2252,8 @@ static u32 dce_v11_0_pick_pll(struct drm_crtc *crtc)
 
 	if ((adev->asic_type == CHIP_POLARIS10) ||
 	    (adev->asic_type == CHIP_POLARIS11) ||
-	    (adev->asic_type == CHIP_POLARIS12)) {
+	    (adev->asic_type == CHIP_POLARIS12) ||
+	    (adev->asic_type == CHIP_VEGAM)) {
 		struct amdgpu_encoder *amdgpu_encoder =
 			to_amdgpu_encoder(amdgpu_crtc->encoder);
 		struct amdgpu_encoder_atom_dig *dig = amdgpu_encoder->enc_priv;
@@ -2450,13 +2450,14 @@ static int dce_v11_0_crtc_cursor_set2(struct drm_crtc *crtc,
 		return ret;
 	}
 
-	ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM, &amdgpu_crtc->cursor_addr);
+	ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM);
 	amdgpu_bo_unreserve(aobj);
 	if (ret) {
 		DRM_ERROR("Failed to pin new cursor BO (%d)\n", ret);
 		drm_gem_object_put_unlocked(obj);
 		return ret;
 	}
+	amdgpu_crtc->cursor_addr = amdgpu_bo_gpu_offset(aobj);
 
 	dce_v11_0_lock_cursor(crtc, true);
 
@@ -2601,11 +2602,9 @@ static void dce_v11_0_crtc_disable(struct drm_crtc *crtc)
 	dce_v11_0_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct amdgpu_framebuffer *amdgpu_fb;
 		struct amdgpu_bo *abo;
 
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(crtc->primary->fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve abo before unpin\n");
@@ -2673,7 +2672,8 @@ static int dce_v11_0_crtc_mode_set(struct drm_crtc *crtc,
 
 	if ((adev->asic_type == CHIP_POLARIS10) ||
 	    (adev->asic_type == CHIP_POLARIS11) ||
-	    (adev->asic_type == CHIP_POLARIS12)) {
+	    (adev->asic_type == CHIP_POLARIS12) ||
+	    (adev->asic_type == CHIP_VEGAM)) {
 		struct amdgpu_encoder *amdgpu_encoder =
 			to_amdgpu_encoder(amdgpu_crtc->encoder);
 		int encoder_mode =
@@ -2830,6 +2830,7 @@ static int dce_v11_0_early_init(void *handle)
 		adev->mode_info.num_dig = 9;
 		break;
 	case CHIP_POLARIS10:
+	case CHIP_VEGAM:
 		adev->mode_info.num_hpd = 6;
 		adev->mode_info.num_dig = 6;
 		break;
@@ -2859,14 +2860,14 @@ static int dce_v11_0_sw_init(void *handle)
 			return r;
 	}
 
-	for (i = 8; i < 20; i += 2) {
+	for (i = VISLANDS30_IV_SRCID_D1_GRPH_PFLIP; i < 20; i += 2) {
 		r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, i, &adev->pageflip_irq);
 		if (r)
 			return r;
 	}
 
 	/* HPD hotplug */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 42, &adev->hpd_irq);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_HOTPLUG_DETECT_A, &adev->hpd_irq);
 	if (r)
 		return r;
 
@@ -2949,7 +2950,8 @@ static int dce_v11_0_hw_init(void *handle)
 	amdgpu_atombios_encoder_init_dig(adev);
 	if ((adev->asic_type == CHIP_POLARIS10) ||
 	    (adev->asic_type == CHIP_POLARIS11) ||
-	    (adev->asic_type == CHIP_POLARIS12)) {
+	    (adev->asic_type == CHIP_POLARIS12) ||
+	    (adev->asic_type == CHIP_VEGAM)) {
 		amdgpu_atombios_crtc_set_dce_clock(adev, adev->clock.default_dispclk,
 						   DCE_CLOCK_TYPE_DISPCLK, ATOM_GCK_DFS);
 		amdgpu_atombios_crtc_set_dce_clock(adev, 0,
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
index 9f67b7fd3487..c9adc627305d 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v6_0.c
@@ -1780,7 +1780,6 @@ static int dce_v6_0_crtc_do_set_base(struct drm_crtc *crtc,
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct amdgpu_device *adev = dev->dev_private;
-	struct amdgpu_framebuffer *amdgpu_fb;
 	struct drm_framebuffer *target_fb;
 	struct drm_gem_object *obj;
 	struct amdgpu_bo *abo;
@@ -1798,32 +1797,28 @@ static int dce_v6_0_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	} else {
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
 	/* If atomic, assume fb object is pinned & idle & fenced and
 	 * just update base pointers
 	 */
-	obj = amdgpu_fb->obj;
+	obj = target_fb->obj[0];
 	abo = gem_to_amdgpu_bo(obj);
 	r = amdgpu_bo_reserve(abo, false);
 	if (unlikely(r != 0))
 		return r;
 
-	if (atomic) {
-		fb_location = amdgpu_bo_gpu_offset(abo);
-	} else {
-		r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM, &fb_location);
+	if (!atomic) {
+		r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM);
 		if (unlikely(r != 0)) {
 			amdgpu_bo_unreserve(abo);
 			return -EINVAL;
 		}
 	}
+	fb_location = amdgpu_bo_gpu_offset(abo);
 
 	amdgpu_bo_get_tiling_flags(abo, &tiling_flags);
 	amdgpu_bo_unreserve(abo);
@@ -1978,8 +1973,7 @@ static int dce_v6_0_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(mmMASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 0);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r != 0))
 			return r;
@@ -2268,13 +2262,14 @@ static int dce_v6_0_crtc_cursor_set2(struct drm_crtc *crtc,
 		return ret;
 	}
 
-	ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM, &amdgpu_crtc->cursor_addr);
+	ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM);
 	amdgpu_bo_unreserve(aobj);
 	if (ret) {
 		DRM_ERROR("Failed to pin new cursor BO (%d)\n", ret);
 		drm_gem_object_put_unlocked(obj);
 		return ret;
 	}
+	amdgpu_crtc->cursor_addr = amdgpu_bo_gpu_offset(aobj);
 
 	dce_v6_0_lock_cursor(crtc, true);
 
@@ -2414,11 +2409,9 @@ static void dce_v6_0_crtc_disable(struct drm_crtc *crtc)
 	dce_v6_0_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct amdgpu_framebuffer *amdgpu_fb;
 		struct amdgpu_bo *abo;
 
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(crtc->primary->fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve abo before unpin\n");
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
index f55422cbd77a..50cd03beac7d 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_v8_0.c
@@ -1754,7 +1754,6 @@ static int dce_v8_0_crtc_do_set_base(struct drm_crtc *crtc,
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct amdgpu_device *adev = dev->dev_private;
-	struct amdgpu_framebuffer *amdgpu_fb;
 	struct drm_framebuffer *target_fb;
 	struct drm_gem_object *obj;
 	struct amdgpu_bo *abo;
@@ -1773,32 +1772,28 @@ static int dce_v8_0_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	} else {
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
 	/* If atomic, assume fb object is pinned & idle & fenced and
 	 * just update base pointers
 	 */
-	obj = amdgpu_fb->obj;
+	obj = target_fb->obj[0];
 	abo = gem_to_amdgpu_bo(obj);
 	r = amdgpu_bo_reserve(abo, false);
 	if (unlikely(r != 0))
 		return r;
 
-	if (atomic) {
-		fb_location = amdgpu_bo_gpu_offset(abo);
-	} else {
-		r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM, &fb_location);
+	if (!atomic) {
+		r = amdgpu_bo_pin(abo, AMDGPU_GEM_DOMAIN_VRAM);
 		if (unlikely(r != 0)) {
 			amdgpu_bo_unreserve(abo);
 			return -EINVAL;
 		}
 	}
+	fb_location = amdgpu_bo_gpu_offset(abo);
 
 	amdgpu_bo_get_tiling_flags(abo, &tiling_flags);
 	amdgpu_bo_unreserve(abo);
@@ -1955,8 +1950,7 @@ static int dce_v8_0_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(mmMASTER_UPDATE_MODE + amdgpu_crtc->crtc_offset, 0);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		amdgpu_fb = to_amdgpu_framebuffer(fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r != 0))
 			return r;
@@ -2279,13 +2273,14 @@ static int dce_v8_0_crtc_cursor_set2(struct drm_crtc *crtc,
 		return ret;
 	}
 
-	ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM, &amdgpu_crtc->cursor_addr);
+	ret = amdgpu_bo_pin(aobj, AMDGPU_GEM_DOMAIN_VRAM);
 	amdgpu_bo_unreserve(aobj);
 	if (ret) {
 		DRM_ERROR("Failed to pin new cursor BO (%d)\n", ret);
 		drm_gem_object_put_unlocked(obj);
 		return ret;
 	}
+	amdgpu_crtc->cursor_addr = amdgpu_bo_gpu_offset(aobj);
 
 	dce_v8_0_lock_cursor(crtc, true);
 
@@ -2430,11 +2425,9 @@ static void dce_v8_0_crtc_disable(struct drm_crtc *crtc)
 	dce_v8_0_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct amdgpu_framebuffer *amdgpu_fb;
 		struct amdgpu_bo *abo;
 
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(crtc->primary->fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve abo before unpin\n");
diff --git a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
index b51f05dc9582..15257634a53a 100644
--- a/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
+++ b/drivers/gpu/drm/amd/amdgpu/dce_virtual.c
@@ -36,6 +36,7 @@
 #include "dce_v10_0.h"
 #include "dce_v11_0.h"
 #include "dce_virtual.h"
+#include "ivsrcid/ivsrcid_vislands30.h"
 
 #define DCE_VIRTUAL_VBLANK_PERIOD 16666666
 
@@ -168,11 +169,9 @@ static void dce_virtual_crtc_disable(struct drm_crtc *crtc)
 	dce_virtual_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct amdgpu_framebuffer *amdgpu_fb;
 		struct amdgpu_bo *abo;
 
-		amdgpu_fb = to_amdgpu_framebuffer(crtc->primary->fb);
-		abo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+		abo = gem_to_amdgpu_bo(crtc->primary->fb->obj[0]);
 		r = amdgpu_bo_reserve(abo, true);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve abo before unpin\n");
@@ -271,25 +270,18 @@ static int dce_virtual_early_init(void *handle)
 static struct drm_encoder *
 dce_virtual_encoder(struct drm_connector *connector)
 {
-	int enc_id = connector->encoder_ids[0];
 	struct drm_encoder *encoder;
 	int i;
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-
-		encoder = drm_encoder_find(connector->dev, NULL, connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
-
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		if (encoder->encoder_type == DRM_MODE_ENCODER_VIRTUAL)
 			return encoder;
 	}
 
 	/* pick the first one */
-	if (enc_id)
-		return drm_encoder_find(connector->dev, NULL, enc_id);
+	drm_connector_for_each_possible_encoder(connector, encoder, i)
+		return encoder;
+
 	return NULL;
 }
 
@@ -329,7 +321,7 @@ static int dce_virtual_get_modes(struct drm_connector *connector)
 	return 0;
 }
 
-static int dce_virtual_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status dce_virtual_mode_valid(struct drm_connector *connector,
 				  struct drm_display_mode *mode)
 {
 	return MODE_OK;
@@ -380,7 +372,7 @@ static int dce_virtual_sw_init(void *handle)
 	int r, i;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 229, &adev->crtc_irq);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_SMU_DISP_TIMER2_TRIGGER, &adev->crtc_irq);
 	if (r)
 		return r;
 
@@ -462,8 +454,9 @@ static int dce_virtual_hw_init(void *handle)
 		break;
 	case CHIP_CARRIZO:
 	case CHIP_STONEY:
-	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
+	case CHIP_VEGAM:
 		dce_v11_0_disable_dce(adev);
 		break;
 	case CHIP_TOPAZ:
@@ -474,6 +467,7 @@ static int dce_virtual_hw_init(void *handle)
 		break;
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		break;
 	default:
 		DRM_ERROR("Virtual display unsupported ASIC type: 0x%X\n", adev->asic_type);
@@ -634,7 +628,7 @@ static int dce_virtual_connector_encoder_init(struct amdgpu_device *adev,
 	drm_connector_register(connector);
 
 	/* link them */
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/df_v1_7.c b/drivers/gpu/drm/amd/amdgpu/df_v1_7.c
new file mode 100644
index 000000000000..9935371db7ce
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/df_v1_7.c
@@ -0,0 +1,120 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "amdgpu.h"
+#include "df_v1_7.h"
+
+#include "df/df_1_7_default.h"
+#include "df/df_1_7_offset.h"
+#include "df/df_1_7_sh_mask.h"
+
+static u32 df_v1_7_channel_number[] = {1, 2, 0, 4, 0, 8, 0, 16, 2};
+
+static void df_v1_7_init (struct amdgpu_device *adev)
+{
+}
+
+static void df_v1_7_enable_broadcast_mode(struct amdgpu_device *adev,
+                                          bool enable)
+{
+	u32 tmp;
+
+	if (enable) {
+		tmp = RREG32_SOC15(DF, 0, mmFabricConfigAccessControl);
+		tmp &= ~FabricConfigAccessControl__CfgRegInstAccEn_MASK;
+		WREG32_SOC15(DF, 0, mmFabricConfigAccessControl, tmp);
+	} else
+		WREG32_SOC15(DF, 0, mmFabricConfigAccessControl,
+			     mmFabricConfigAccessControl_DEFAULT);
+}
+
+static u32 df_v1_7_get_fb_channel_number(struct amdgpu_device *adev)
+{
+	u32 tmp;
+
+	tmp = RREG32_SOC15(DF, 0, mmDF_CS_AON0_DramBaseAddress0);
+	tmp &= DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK;
+	tmp >>= DF_CS_AON0_DramBaseAddress0__IntLvNumChan__SHIFT;
+
+	return tmp;
+}
+
+static u32 df_v1_7_get_hbm_channel_number(struct amdgpu_device *adev)
+{
+	int fb_channel_number;
+
+	fb_channel_number = adev->df_funcs->get_fb_channel_number(adev);
+
+	return df_v1_7_channel_number[fb_channel_number];
+}
+
+static void df_v1_7_update_medium_grain_clock_gating(struct amdgpu_device *adev,
+						     bool enable)
+{
+	u32 tmp;
+
+	/* Put DF on broadcast mode */
+	adev->df_funcs->enable_broadcast_mode(adev, true);
+
+	if (enable && (adev->cg_flags & AMD_CG_SUPPORT_DF_MGCG)) {
+		tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater);
+		tmp &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK;
+		tmp |= DF_V1_7_MGCG_ENABLE_15_CYCLE_DELAY;
+		WREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater, tmp);
+	} else {
+		tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater);
+		tmp &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK;
+		tmp |= DF_V1_7_MGCG_DISABLE;
+		WREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater, tmp);
+	}
+
+	/* Exit boradcast mode */
+	adev->df_funcs->enable_broadcast_mode(adev, false);
+}
+
+static void df_v1_7_get_clockgating_state(struct amdgpu_device *adev,
+					  u32 *flags)
+{
+	u32 tmp;
+
+	/* AMD_CG_SUPPORT_DF_MGCG */
+	tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater);
+	if (tmp & DF_V1_7_MGCG_ENABLE_15_CYCLE_DELAY)
+		*flags |= AMD_CG_SUPPORT_DF_MGCG;
+}
+
+static void df_v1_7_enable_ecc_force_par_wr_rmw(struct amdgpu_device *adev,
+						bool enable)
+{
+	WREG32_FIELD15(DF, 0, DF_CS_AON0_CoherentSlaveModeCtrlA0,
+		       ForceParWrRMW, enable);
+}
+
+const struct amdgpu_df_funcs df_v1_7_funcs = {
+	.init = df_v1_7_init,
+	.enable_broadcast_mode = df_v1_7_enable_broadcast_mode,
+	.get_fb_channel_number = df_v1_7_get_fb_channel_number,
+	.get_hbm_channel_number = df_v1_7_get_hbm_channel_number,
+	.update_medium_grain_clock_gating = df_v1_7_update_medium_grain_clock_gating,
+	.get_clockgating_state = df_v1_7_get_clockgating_state,
+	.enable_ecc_force_par_wr_rmw = df_v1_7_enable_ecc_force_par_wr_rmw,
+};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegk110.c b/drivers/gpu/drm/amd/amdgpu/df_v1_7.h
index d8bdd246c8ed..74621104c487 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegk110.c
+++ b/drivers/gpu/drm/amd/amdgpu/df_v1_7.h
@@ -1,5 +1,5 @@
 /*
- * Copyright 2015 Red Hat Inc.
+ * Copyright 2018 Advanced Micro Devices, Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -19,20 +19,22 @@
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
  *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
 
-#include <nvif/class.h>
+#ifndef __DF_V1_7_H__
+#define __DF_V1_7_H__
 
-const struct nv50_disp_dmac_oclass
-gk110_disp_base_oclass = {
-	.base.oclass = GK110_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &gf119_disp_dmac_func,
-	.mthd = &gf119_disp_base_chan_mthd,
-	.chid = 1,
+#include "soc15_common.h"
+enum DF_V1_7_MGCG
+{
+	DF_V1_7_MGCG_DISABLE = 0,
+	DF_V1_7_MGCG_ENABLE_00_CYCLE_DELAY =1,
+	DF_V1_7_MGCG_ENABLE_01_CYCLE_DELAY =2,
+	DF_V1_7_MGCG_ENABLE_15_CYCLE_DELAY =13,
+	DF_V1_7_MGCG_ENABLE_31_CYCLE_DELAY =14,
+	DF_V1_7_MGCG_ENABLE_63_CYCLE_DELAY =15
 };
+
+extern const struct amdgpu_df_funcs df_v1_7_funcs;
+
+#endif
diff --git a/drivers/gpu/drm/amd/amdgpu/df_v3_6.c b/drivers/gpu/drm/amd/amdgpu/df_v3_6.c
new file mode 100644
index 000000000000..d5ebe566809b
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/df_v3_6.c
@@ -0,0 +1,116 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "amdgpu.h"
+#include "df_v3_6.h"
+
+#include "df/df_3_6_default.h"
+#include "df/df_3_6_offset.h"
+#include "df/df_3_6_sh_mask.h"
+
+static u32 df_v3_6_channel_number[] = {1, 2, 0, 4, 0, 8, 0,
+				       16, 32, 0, 0, 0, 2, 4, 8};
+
+static void df_v3_6_init(struct amdgpu_device *adev)
+{
+}
+
+static void df_v3_6_enable_broadcast_mode(struct amdgpu_device *adev,
+					  bool enable)
+{
+	u32 tmp;
+
+	if (enable) {
+		tmp = RREG32_SOC15(DF, 0, mmFabricConfigAccessControl);
+		tmp &= ~FabricConfigAccessControl__CfgRegInstAccEn_MASK;
+		WREG32_SOC15(DF, 0, mmFabricConfigAccessControl, tmp);
+	} else
+		WREG32_SOC15(DF, 0, mmFabricConfigAccessControl,
+			     mmFabricConfigAccessControl_DEFAULT);
+}
+
+static u32 df_v3_6_get_fb_channel_number(struct amdgpu_device *adev)
+{
+	u32 tmp;
+
+	tmp = RREG32_SOC15(DF, 0, mmDF_CS_UMC_AON0_DramBaseAddress0);
+	tmp &= DF_CS_UMC_AON0_DramBaseAddress0__IntLvNumChan_MASK;
+	tmp >>= DF_CS_UMC_AON0_DramBaseAddress0__IntLvNumChan__SHIFT;
+
+	return tmp;
+}
+
+static u32 df_v3_6_get_hbm_channel_number(struct amdgpu_device *adev)
+{
+	int fb_channel_number;
+
+	fb_channel_number = adev->df_funcs->get_fb_channel_number(adev);
+	if (fb_channel_number >= ARRAY_SIZE(df_v3_6_channel_number))
+		fb_channel_number = 0;
+
+	return df_v3_6_channel_number[fb_channel_number];
+}
+
+static void df_v3_6_update_medium_grain_clock_gating(struct amdgpu_device *adev,
+						     bool enable)
+{
+	u32 tmp;
+
+	/* Put DF on broadcast mode */
+	adev->df_funcs->enable_broadcast_mode(adev, true);
+
+	if (enable && (adev->cg_flags & AMD_CG_SUPPORT_DF_MGCG)) {
+		tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater);
+		tmp &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK;
+		tmp |= DF_V3_6_MGCG_ENABLE_15_CYCLE_DELAY;
+		WREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater, tmp);
+	} else {
+		tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater);
+		tmp &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK;
+		tmp |= DF_V3_6_MGCG_DISABLE;
+		WREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater, tmp);
+	}
+
+	/* Exit broadcast mode */
+	adev->df_funcs->enable_broadcast_mode(adev, false);
+}
+
+static void df_v3_6_get_clockgating_state(struct amdgpu_device *adev,
+					  u32 *flags)
+{
+	u32 tmp;
+
+	/* AMD_CG_SUPPORT_DF_MGCG */
+	tmp = RREG32_SOC15(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater);
+	if (tmp & DF_V3_6_MGCG_ENABLE_15_CYCLE_DELAY)
+		*flags |= AMD_CG_SUPPORT_DF_MGCG;
+}
+
+const struct amdgpu_df_funcs df_v3_6_funcs = {
+	.init = df_v3_6_init,
+	.enable_broadcast_mode = df_v3_6_enable_broadcast_mode,
+	.get_fb_channel_number = df_v3_6_get_fb_channel_number,
+	.get_hbm_channel_number = df_v3_6_get_hbm_channel_number,
+	.update_medium_grain_clock_gating =
+			df_v3_6_update_medium_grain_clock_gating,
+	.get_clockgating_state = df_v3_6_get_clockgating_state,
+};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegk104.c b/drivers/gpu/drm/amd/amdgpu/df_v3_6.h
index 780a1d973634..e79c58e5efcb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegk104.c
+++ b/drivers/gpu/drm/amd/amdgpu/df_v3_6.h
@@ -1,5 +1,5 @@
 /*
- * Copyright 2015 Red Hat Inc.
+ * Copyright 2018 Advanced Micro Devices, Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -19,20 +19,22 @@
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
  *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
 
-#include <nvif/class.h>
+#ifndef __DF_V3_6_H__
+#define __DF_V3_6_H__
 
-const struct nv50_disp_dmac_oclass
-gk104_disp_base_oclass = {
-	.base.oclass = GK104_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &gf119_disp_dmac_func,
-	.mthd = &gf119_disp_base_chan_mthd,
-	.chid = 1,
+#include "soc15_common.h"
+
+enum DF_V3_6_MGCG {
+	DF_V3_6_MGCG_DISABLE = 0,
+	DF_V3_6_MGCG_ENABLE_00_CYCLE_DELAY = 1,
+	DF_V3_6_MGCG_ENABLE_01_CYCLE_DELAY = 2,
+	DF_V3_6_MGCG_ENABLE_15_CYCLE_DELAY = 13,
+	DF_V3_6_MGCG_ENABLE_31_CYCLE_DELAY = 14,
+	DF_V3_6_MGCG_ENABLE_63_CYCLE_DELAY = 15
 };
+
+extern const struct amdgpu_df_funcs df_v3_6_funcs;
+
+#endif
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
index cd6bf291a853..de184a886057 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
@@ -44,30 +44,30 @@ static void gfx_v6_0_set_ring_funcs(struct amdgpu_device *adev);
 static void gfx_v6_0_set_irq_funcs(struct amdgpu_device *adev);
 static void gfx_v6_0_get_cu_info(struct amdgpu_device *adev);
 
-MODULE_FIRMWARE("radeon/tahiti_pfp.bin");
-MODULE_FIRMWARE("radeon/tahiti_me.bin");
-MODULE_FIRMWARE("radeon/tahiti_ce.bin");
-MODULE_FIRMWARE("radeon/tahiti_rlc.bin");
-
-MODULE_FIRMWARE("radeon/pitcairn_pfp.bin");
-MODULE_FIRMWARE("radeon/pitcairn_me.bin");
-MODULE_FIRMWARE("radeon/pitcairn_ce.bin");
-MODULE_FIRMWARE("radeon/pitcairn_rlc.bin");
-
-MODULE_FIRMWARE("radeon/verde_pfp.bin");
-MODULE_FIRMWARE("radeon/verde_me.bin");
-MODULE_FIRMWARE("radeon/verde_ce.bin");
-MODULE_FIRMWARE("radeon/verde_rlc.bin");
-
-MODULE_FIRMWARE("radeon/oland_pfp.bin");
-MODULE_FIRMWARE("radeon/oland_me.bin");
-MODULE_FIRMWARE("radeon/oland_ce.bin");
-MODULE_FIRMWARE("radeon/oland_rlc.bin");
-
-MODULE_FIRMWARE("radeon/hainan_pfp.bin");
-MODULE_FIRMWARE("radeon/hainan_me.bin");
-MODULE_FIRMWARE("radeon/hainan_ce.bin");
-MODULE_FIRMWARE("radeon/hainan_rlc.bin");
+MODULE_FIRMWARE("amdgpu/tahiti_pfp.bin");
+MODULE_FIRMWARE("amdgpu/tahiti_me.bin");
+MODULE_FIRMWARE("amdgpu/tahiti_ce.bin");
+MODULE_FIRMWARE("amdgpu/tahiti_rlc.bin");
+
+MODULE_FIRMWARE("amdgpu/pitcairn_pfp.bin");
+MODULE_FIRMWARE("amdgpu/pitcairn_me.bin");
+MODULE_FIRMWARE("amdgpu/pitcairn_ce.bin");
+MODULE_FIRMWARE("amdgpu/pitcairn_rlc.bin");
+
+MODULE_FIRMWARE("amdgpu/verde_pfp.bin");
+MODULE_FIRMWARE("amdgpu/verde_me.bin");
+MODULE_FIRMWARE("amdgpu/verde_ce.bin");
+MODULE_FIRMWARE("amdgpu/verde_rlc.bin");
+
+MODULE_FIRMWARE("amdgpu/oland_pfp.bin");
+MODULE_FIRMWARE("amdgpu/oland_me.bin");
+MODULE_FIRMWARE("amdgpu/oland_ce.bin");
+MODULE_FIRMWARE("amdgpu/oland_rlc.bin");
+
+MODULE_FIRMWARE("amdgpu/hainan_pfp.bin");
+MODULE_FIRMWARE("amdgpu/hainan_me.bin");
+MODULE_FIRMWARE("amdgpu/hainan_ce.bin");
+MODULE_FIRMWARE("amdgpu/hainan_rlc.bin");
 
 static u32 gfx_v6_0_get_csb_size(struct amdgpu_device *adev);
 static void gfx_v6_0_get_csb_buffer(struct amdgpu_device *adev, volatile u32 *buffer);
@@ -335,7 +335,7 @@ static int gfx_v6_0_init_microcode(struct amdgpu_device *adev)
 	default: BUG();
 	}
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_pfp.bin", chip_name);
+	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_pfp.bin", chip_name);
 	err = request_firmware(&adev->gfx.pfp_fw, fw_name, adev->dev);
 	if (err)
 		goto out;
@@ -346,7 +346,7 @@ static int gfx_v6_0_init_microcode(struct amdgpu_device *adev)
 	adev->gfx.pfp_fw_version = le32_to_cpu(cp_hdr->header.ucode_version);
 	adev->gfx.pfp_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version);
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_me.bin", chip_name);
+	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_me.bin", chip_name);
 	err = request_firmware(&adev->gfx.me_fw, fw_name, adev->dev);
 	if (err)
 		goto out;
@@ -357,7 +357,7 @@ static int gfx_v6_0_init_microcode(struct amdgpu_device *adev)
 	adev->gfx.me_fw_version = le32_to_cpu(cp_hdr->header.ucode_version);
 	adev->gfx.me_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version);
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_ce.bin", chip_name);
+	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_ce.bin", chip_name);
 	err = request_firmware(&adev->gfx.ce_fw, fw_name, adev->dev);
 	if (err)
 		goto out;
@@ -368,7 +368,7 @@ static int gfx_v6_0_init_microcode(struct amdgpu_device *adev)
 	adev->gfx.ce_fw_version = le32_to_cpu(cp_hdr->header.ucode_version);
 	adev->gfx.ce_feature_version = le32_to_cpu(cp_hdr->ucode_feature_version);
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_rlc.bin", chip_name);
+	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_rlc.bin", chip_name);
 	err = request_firmware(&adev->gfx.rlc_fw, fw_name, adev->dev);
 	if (err)
 		goto out;
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
index 42b6144c1fd5..95452c5a9df6 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c
@@ -57,36 +57,36 @@ static void gfx_v7_0_set_ring_funcs(struct amdgpu_device *adev);
 static void gfx_v7_0_set_irq_funcs(struct amdgpu_device *adev);
 static void gfx_v7_0_set_gds_init(struct amdgpu_device *adev);
 
-MODULE_FIRMWARE("radeon/bonaire_pfp.bin");
-MODULE_FIRMWARE("radeon/bonaire_me.bin");
-MODULE_FIRMWARE("radeon/bonaire_ce.bin");
-MODULE_FIRMWARE("radeon/bonaire_rlc.bin");
-MODULE_FIRMWARE("radeon/bonaire_mec.bin");
-
-MODULE_FIRMWARE("radeon/hawaii_pfp.bin");
-MODULE_FIRMWARE("radeon/hawaii_me.bin");
-MODULE_FIRMWARE("radeon/hawaii_ce.bin");
-MODULE_FIRMWARE("radeon/hawaii_rlc.bin");
-MODULE_FIRMWARE("radeon/hawaii_mec.bin");
-
-MODULE_FIRMWARE("radeon/kaveri_pfp.bin");
-MODULE_FIRMWARE("radeon/kaveri_me.bin");
-MODULE_FIRMWARE("radeon/kaveri_ce.bin");
-MODULE_FIRMWARE("radeon/kaveri_rlc.bin");
-MODULE_FIRMWARE("radeon/kaveri_mec.bin");
-MODULE_FIRMWARE("radeon/kaveri_mec2.bin");
-
-MODULE_FIRMWARE("radeon/kabini_pfp.bin");
-MODULE_FIRMWARE("radeon/kabini_me.bin");
-MODULE_FIRMWARE("radeon/kabini_ce.bin");
-MODULE_FIRMWARE("radeon/kabini_rlc.bin");
-MODULE_FIRMWARE("radeon/kabini_mec.bin");
-
-MODULE_FIRMWARE("radeon/mullins_pfp.bin");
-MODULE_FIRMWARE("radeon/mullins_me.bin");
-MODULE_FIRMWARE("radeon/mullins_ce.bin");
-MODULE_FIRMWARE("radeon/mullins_rlc.bin");
-MODULE_FIRMWARE("radeon/mullins_mec.bin");
+MODULE_FIRMWARE("amdgpu/bonaire_pfp.bin");
+MODULE_FIRMWARE("amdgpu/bonaire_me.bin");
+MODULE_FIRMWARE("amdgpu/bonaire_ce.bin");
+MODULE_FIRMWARE("amdgpu/bonaire_rlc.bin");
+MODULE_FIRMWARE("amdgpu/bonaire_mec.bin");
+
+MODULE_FIRMWARE("amdgpu/hawaii_pfp.bin");
+MODULE_FIRMWARE("amdgpu/hawaii_me.bin");
+MODULE_FIRMWARE("amdgpu/hawaii_ce.bin");
+MODULE_FIRMWARE("amdgpu/hawaii_rlc.bin");
+MODULE_FIRMWARE("amdgpu/hawaii_mec.bin");
+
+MODULE_FIRMWARE("amdgpu/kaveri_pfp.bin");
+MODULE_FIRMWARE("amdgpu/kaveri_me.bin");
+MODULE_FIRMWARE("amdgpu/kaveri_ce.bin");
+MODULE_FIRMWARE("amdgpu/kaveri_rlc.bin");
+MODULE_FIRMWARE("amdgpu/kaveri_mec.bin");
+MODULE_FIRMWARE("amdgpu/kaveri_mec2.bin");
+
+MODULE_FIRMWARE("amdgpu/kabini_pfp.bin");
+MODULE_FIRMWARE("amdgpu/kabini_me.bin");
+MODULE_FIRMWARE("amdgpu/kabini_ce.bin");
+MODULE_FIRMWARE("amdgpu/kabini_rlc.bin");
+MODULE_FIRMWARE("amdgpu/kabini_mec.bin");
+
+MODULE_FIRMWARE("amdgpu/mullins_pfp.bin");
+MODULE_FIRMWARE("amdgpu/mullins_me.bin");
+MODULE_FIRMWARE("amdgpu/mullins_ce.bin");
+MODULE_FIRMWARE("amdgpu/mullins_rlc.bin");
+MODULE_FIRMWARE("amdgpu/mullins_mec.bin");
 
 static const struct amdgpu_gds_reg_offset amdgpu_gds_reg_offset[] =
 {
@@ -925,7 +925,7 @@ static int gfx_v7_0_init_microcode(struct amdgpu_device *adev)
 	default: BUG();
 	}
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_pfp.bin", chip_name);
+	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_pfp.bin", chip_name);
 	err = request_firmware(&adev->gfx.pfp_fw, fw_name, adev->dev);
 	if (err)
 		goto out;
@@ -933,7 +933,7 @@ static int gfx_v7_0_init_microcode(struct amdgpu_device *adev)
 	if (err)
 		goto out;
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_me.bin", chip_name);
+	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_me.bin", chip_name);
 	err = request_firmware(&adev->gfx.me_fw, fw_name, adev->dev);
 	if (err)
 		goto out;
@@ -941,7 +941,7 @@ static int gfx_v7_0_init_microcode(struct amdgpu_device *adev)
 	if (err)
 		goto out;
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_ce.bin", chip_name);
+	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_ce.bin", chip_name);
 	err = request_firmware(&adev->gfx.ce_fw, fw_name, adev->dev);
 	if (err)
 		goto out;
@@ -949,7 +949,7 @@ static int gfx_v7_0_init_microcode(struct amdgpu_device *adev)
 	if (err)
 		goto out;
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_mec.bin", chip_name);
+	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mec.bin", chip_name);
 	err = request_firmware(&adev->gfx.mec_fw, fw_name, adev->dev);
 	if (err)
 		goto out;
@@ -958,7 +958,7 @@ static int gfx_v7_0_init_microcode(struct amdgpu_device *adev)
 		goto out;
 
 	if (adev->asic_type == CHIP_KAVERI) {
-		snprintf(fw_name, sizeof(fw_name), "radeon/%s_mec2.bin", chip_name);
+		snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mec2.bin", chip_name);
 		err = request_firmware(&adev->gfx.mec2_fw, fw_name, adev->dev);
 		if (err)
 			goto out;
@@ -967,7 +967,7 @@ static int gfx_v7_0_init_microcode(struct amdgpu_device *adev)
 			goto out;
 	}
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_rlc.bin", chip_name);
+	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_rlc.bin", chip_name);
 	err = request_firmware(&adev->gfx.rlc_fw, fw_name, adev->dev);
 	if (err)
 		goto out;
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index e14263fca1c9..5cd45210113f 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -51,6 +51,8 @@
 
 #include "smu/smu_7_1_3_d.h"
 
+#include "ivsrcid/ivsrcid_vislands30.h"
+
 #define GFX8_NUM_GFX_RINGS     1
 #define GFX8_MEC_HPD_SIZE 2048
 
@@ -125,18 +127,6 @@ MODULE_FIRMWARE("amdgpu/fiji_mec.bin");
 MODULE_FIRMWARE("amdgpu/fiji_mec2.bin");
 MODULE_FIRMWARE("amdgpu/fiji_rlc.bin");
 
-MODULE_FIRMWARE("amdgpu/polaris11_ce.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_ce_2.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_pfp.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_pfp_2.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_me.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_me_2.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_mec.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_mec_2.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_mec2.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_mec2_2.bin");
-MODULE_FIRMWARE("amdgpu/polaris11_rlc.bin");
-
 MODULE_FIRMWARE("amdgpu/polaris10_ce.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_ce_2.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_pfp.bin");
@@ -149,6 +139,18 @@ MODULE_FIRMWARE("amdgpu/polaris10_mec2.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_mec2_2.bin");
 MODULE_FIRMWARE("amdgpu/polaris10_rlc.bin");
 
+MODULE_FIRMWARE("amdgpu/polaris11_ce.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_ce_2.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_pfp.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_pfp_2.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_me.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_me_2.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_mec.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_mec_2.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_mec2.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_mec2_2.bin");
+MODULE_FIRMWARE("amdgpu/polaris11_rlc.bin");
+
 MODULE_FIRMWARE("amdgpu/polaris12_ce.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_ce_2.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_pfp.bin");
@@ -161,6 +163,13 @@ MODULE_FIRMWARE("amdgpu/polaris12_mec2.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_mec2_2.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_rlc.bin");
 
+MODULE_FIRMWARE("amdgpu/vegam_ce.bin");
+MODULE_FIRMWARE("amdgpu/vegam_pfp.bin");
+MODULE_FIRMWARE("amdgpu/vegam_me.bin");
+MODULE_FIRMWARE("amdgpu/vegam_mec.bin");
+MODULE_FIRMWARE("amdgpu/vegam_mec2.bin");
+MODULE_FIRMWARE("amdgpu/vegam_rlc.bin");
+
 static const struct amdgpu_gds_reg_offset amdgpu_gds_reg_offset[] =
 {
 	{mmGDS_VMID0_BASE, mmGDS_VMID0_SIZE, mmGDS_GWS_VMID0, mmGDS_OA_VMID0},
@@ -292,6 +301,37 @@ static const u32 tonga_mgcg_cgcg_init[] =
 	mmCP_MEM_SLP_CNTL, 0x00000001, 0x00000001,
 };
 
+static const u32 golden_settings_vegam_a11[] =
+{
+	mmCB_HW_CONTROL, 0x0001f3cf, 0x00007208,
+	mmCB_HW_CONTROL_2, 0x0f000000, 0x0d000000,
+	mmCB_HW_CONTROL_3, 0x000001ff, 0x00000040,
+	mmDB_DEBUG2, 0xf00fffff, 0x00000400,
+	mmPA_SC_ENHANCE, 0xffffffff, 0x20000001,
+	mmPA_SC_LINE_STIPPLE_STATE, 0x0000ff0f, 0x00000000,
+	mmPA_SC_RASTER_CONFIG, 0x3f3fffff, 0x3a00161a,
+	mmPA_SC_RASTER_CONFIG_1, 0x0000003f, 0x0000002e,
+	mmRLC_CGCG_CGLS_CTRL, 0x00000003, 0x0001003c,
+	mmRLC_CGCG_CGLS_CTRL_3D, 0xffffffff, 0x0001003c,
+	mmSQ_CONFIG, 0x07f80000, 0x01180000,
+	mmTA_CNTL_AUX, 0x000f000f, 0x000b0000,
+	mmTCC_CTRL, 0x00100000, 0xf31fff7f,
+	mmTCP_ADDR_CONFIG, 0x000003ff, 0x000000f7,
+	mmTCP_CHAN_STEER_HI, 0xffffffff, 0x00000000,
+	mmTCP_CHAN_STEER_LO, 0xffffffff, 0x32761054,
+	mmVGT_RESET_DEBUG, 0x00000004, 0x00000004,
+};
+
+static const u32 vegam_golden_common_all[] =
+{
+	mmGRBM_GFX_INDEX, 0xffffffff, 0xe0000000,
+	mmGB_ADDR_CONFIG, 0xffffffff, 0x22011003,
+	mmSPI_RESOURCE_RESERVE_CU_0, 0xffffffff, 0x00000800,
+	mmSPI_RESOURCE_RESERVE_CU_1, 0xffffffff, 0x00000800,
+	mmSPI_RESOURCE_RESERVE_EN_CU_0, 0xffffffff, 0x00FF7FBF,
+	mmSPI_RESOURCE_RESERVE_EN_CU_1, 0xffffffff, 0x00FF7FAF,
+};
+
 static const u32 golden_settings_polaris11_a11[] =
 {
 	mmCB_HW_CONTROL, 0x0000f3cf, 0x00007208,
@@ -666,6 +706,17 @@ static const u32 stoney_mgcg_cgcg_init[] =
 	mmCGTS_SM_CTRL_REG, 0xffffffff, 0x96940200,
 };
 
+
+static const char * const sq_edc_source_names[] = {
+	"SQ_EDC_INFO_SOURCE_INVALID: No EDC error has occurred",
+	"SQ_EDC_INFO_SOURCE_INST: EDC source is Instruction Fetch",
+	"SQ_EDC_INFO_SOURCE_SGPR: EDC source is SGPR or SQC data return",
+	"SQ_EDC_INFO_SOURCE_VGPR: EDC source is VGPR",
+	"SQ_EDC_INFO_SOURCE_LDS: EDC source is LDS",
+	"SQ_EDC_INFO_SOURCE_GDS: EDC source is GDS",
+	"SQ_EDC_INFO_SOURCE_TA: EDC source is TA",
+};
+
 static void gfx_v8_0_set_ring_funcs(struct amdgpu_device *adev);
 static void gfx_v8_0_set_irq_funcs(struct amdgpu_device *adev);
 static void gfx_v8_0_set_gds_init(struct amdgpu_device *adev);
@@ -712,6 +763,14 @@ static void gfx_v8_0_init_golden_registers(struct amdgpu_device *adev)
 							tonga_golden_common_all,
 							ARRAY_SIZE(tonga_golden_common_all));
 		break;
+	case CHIP_VEGAM:
+		amdgpu_device_program_register_sequence(adev,
+							golden_settings_vegam_a11,
+							ARRAY_SIZE(golden_settings_vegam_a11));
+		amdgpu_device_program_register_sequence(adev,
+							vegam_golden_common_all,
+							ARRAY_SIZE(vegam_golden_common_all));
+		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
 		amdgpu_device_program_register_sequence(adev,
@@ -820,26 +879,32 @@ static int gfx_v8_0_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 	struct amdgpu_device *adev = ring->adev;
 	struct amdgpu_ib ib;
 	struct dma_fence *f = NULL;
-	uint32_t scratch;
-	uint32_t tmp = 0;
+
+	unsigned int index;
+	uint64_t gpu_addr;
+	uint32_t tmp;
 	long r;
 
-	r = amdgpu_gfx_scratch_get(adev, &scratch);
+	r = amdgpu_device_wb_get(adev, &index);
 	if (r) {
-		DRM_ERROR("amdgpu: failed to get scratch reg (%ld).\n", r);
+		dev_err(adev->dev, "(%ld) failed to allocate wb slot\n", r);
 		return r;
 	}
-	WREG32(scratch, 0xCAFEDEAD);
+
+	gpu_addr = adev->wb.gpu_addr + (index * 4);
+	adev->wb.wb[index] = cpu_to_le32(0xCAFEDEAD);
 	memset(&ib, 0, sizeof(ib));
-	r = amdgpu_ib_get(adev, NULL, 256, &ib);
+	r = amdgpu_ib_get(adev, NULL, 16, &ib);
 	if (r) {
 		DRM_ERROR("amdgpu: failed to get ib (%ld).\n", r);
 		goto err1;
 	}
-	ib.ptr[0] = PACKET3(PACKET3_SET_UCONFIG_REG, 1);
-	ib.ptr[1] = ((scratch - PACKET3_SET_UCONFIG_REG_START));
-	ib.ptr[2] = 0xDEADBEEF;
-	ib.length_dw = 3;
+	ib.ptr[0] = PACKET3(PACKET3_WRITE_DATA, 3);
+	ib.ptr[1] = WRITE_DATA_DST_SEL(5) | WR_CONFIRM;
+	ib.ptr[2] = lower_32_bits(gpu_addr);
+	ib.ptr[3] = upper_32_bits(gpu_addr);
+	ib.ptr[4] = 0xDEADBEEF;
+	ib.length_dw = 5;
 
 	r = amdgpu_ib_schedule(ring, 1, &ib, NULL, &f);
 	if (r)
@@ -854,20 +919,21 @@ static int gfx_v8_0_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 		DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r);
 		goto err2;
 	}
-	tmp = RREG32(scratch);
+
+	tmp = adev->wb.wb[index];
 	if (tmp == 0xDEADBEEF) {
 		DRM_DEBUG("ib test on ring %d succeeded\n", ring->idx);
 		r = 0;
 	} else {
-		DRM_ERROR("amdgpu: ib test failed (scratch(0x%04X)=0x%08X)\n",
-			  scratch, tmp);
+		DRM_ERROR("ib test on ring %d failed\n", ring->idx);
 		r = -EINVAL;
 	}
+
 err2:
 	amdgpu_ib_free(adev, &ib, NULL);
 	dma_fence_put(f);
 err1:
-	amdgpu_gfx_scratch_free(adev, scratch);
+	amdgpu_device_wb_free(adev, index);
 	return r;
 }
 
@@ -918,17 +984,20 @@ static int gfx_v8_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_FIJI:
 		chip_name = "fiji";
 		break;
-	case CHIP_POLARIS11:
-		chip_name = "polaris11";
+	case CHIP_STONEY:
+		chip_name = "stoney";
 		break;
 	case CHIP_POLARIS10:
 		chip_name = "polaris10";
 		break;
+	case CHIP_POLARIS11:
+		chip_name = "polaris11";
+		break;
 	case CHIP_POLARIS12:
 		chip_name = "polaris12";
 		break;
-	case CHIP_STONEY:
-		chip_name = "stoney";
+	case CHIP_VEGAM:
+		chip_name = "vegam";
 		break;
 	default:
 		BUG();
@@ -1770,6 +1839,7 @@ static int gfx_v8_0_gpu_early_init(struct amdgpu_device *adev)
 		gb_addr_config = POLARIS11_GB_ADDR_CONFIG_GOLDEN;
 		break;
 	case CHIP_POLARIS10:
+	case CHIP_VEGAM:
 		ret = amdgpu_atombios_get_gfx_info(adev);
 		if (ret)
 			return ret;
@@ -1949,6 +2019,8 @@ static int gfx_v8_0_compute_ring_init(struct amdgpu_device *adev, int ring_id,
 	return 0;
 }
 
+static void gfx_v8_0_sq_irq_work_func(struct work_struct *work);
+
 static int gfx_v8_0_sw_init(void *handle)
 {
 	int i, j, k, r, ring_id;
@@ -1957,12 +2029,13 @@ static int gfx_v8_0_sw_init(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	switch (adev->asic_type) {
-	case CHIP_FIJI:
 	case CHIP_TONGA:
+	case CHIP_CARRIZO:
+	case CHIP_FIJI:
+	case CHIP_POLARIS10:
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
-	case CHIP_POLARIS10:
-	case CHIP_CARRIZO:
+	case CHIP_VEGAM:
 		adev->gfx.mec.num_mec = 2;
 		break;
 	case CHIP_TOPAZ:
@@ -1976,27 +2049,43 @@ static int gfx_v8_0_sw_init(void *handle)
 	adev->gfx.mec.num_queue_per_pipe = 8;
 
 	/* KIQ event */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 178, &adev->gfx.kiq.irq);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_CP_INT_IB2, &adev->gfx.kiq.irq);
 	if (r)
 		return r;
 
 	/* EOP Event */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 181, &adev->gfx.eop_irq);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_CP_END_OF_PIPE, &adev->gfx.eop_irq);
 	if (r)
 		return r;
 
 	/* Privileged reg */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 184,
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_CP_PRIV_REG_FAULT,
 			      &adev->gfx.priv_reg_irq);
 	if (r)
 		return r;
 
 	/* Privileged inst */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 185,
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_CP_PRIV_INSTR_FAULT,
 			      &adev->gfx.priv_inst_irq);
 	if (r)
 		return r;
 
+	/* Add CP EDC/ECC irq  */
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_CP_ECC_ERROR,
+			      &adev->gfx.cp_ecc_error_irq);
+	if (r)
+		return r;
+
+	/* SQ interrupts. */
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_SQ_INTERRUPT_MSG,
+			      &adev->gfx.sq_irq);
+	if (r) {
+		DRM_ERROR("amdgpu_irq_add() for SQ failed: %d\n", r);
+		return r;
+	}
+
+	INIT_WORK(&adev->gfx.sq_work.work, gfx_v8_0_sq_irq_work_func);
+
 	adev->gfx.gfx_current_status = AMDGPU_GFX_NORMAL_MODE;
 
 	gfx_v8_0_scratch_init(adev);
@@ -2323,6 +2412,7 @@ static void gfx_v8_0_tiling_mode_table_init(struct amdgpu_device *adev)
 
 		break;
 	case CHIP_FIJI:
+	case CHIP_VEGAM:
 		modearray[0] = (ARRAY_MODE(ARRAY_2D_TILED_THIN1) |
 				PIPE_CONFIG(ADDR_SURF_P16_32x32_16x16) |
 				TILE_SPLIT(ADDR_SURF_TILE_SPLIT_64B) |
@@ -3504,6 +3594,7 @@ gfx_v8_0_raster_config(struct amdgpu_device *adev, u32 *rconf, u32 *rconf1)
 {
 	switch (adev->asic_type) {
 	case CHIP_FIJI:
+	case CHIP_VEGAM:
 		*rconf |= RB_MAP_PKR0(2) | RB_MAP_PKR1(2) |
 			  RB_XSEL2(1) | PKR_MAP(2) |
 			  PKR_XSEL(1) | PKR_YSEL(1) |
@@ -4071,7 +4162,8 @@ static void gfx_v8_0_init_pg(struct amdgpu_device *adev)
 		gfx_v8_0_init_power_gating(adev);
 		WREG32(mmRLC_PG_ALWAYS_ON_CU_MASK, adev->gfx.cu_info.ao_cu_mask);
 	} else if ((adev->asic_type == CHIP_POLARIS11) ||
-		   (adev->asic_type == CHIP_POLARIS12)) {
+		   (adev->asic_type == CHIP_POLARIS12) ||
+		   (adev->asic_type == CHIP_VEGAM)) {
 		gfx_v8_0_init_csb(adev);
 		gfx_v8_0_init_save_restore_list(adev);
 		gfx_v8_0_enable_save_restore_machine(adev);
@@ -4146,7 +4238,8 @@ static int gfx_v8_0_rlc_resume(struct amdgpu_device *adev)
 	WREG32(mmRLC_CGCG_CGLS_CTRL, tmp);
 	if (adev->asic_type == CHIP_POLARIS11 ||
 	    adev->asic_type == CHIP_POLARIS10 ||
-	    adev->asic_type == CHIP_POLARIS12) {
+	    adev->asic_type == CHIP_POLARIS12 ||
+	    adev->asic_type == CHIP_VEGAM) {
 		tmp = RREG32(mmRLC_CGCG_CGLS_CTRL_3D);
 		tmp &= ~0x3;
 		WREG32(mmRLC_CGCG_CGLS_CTRL_3D, tmp);
@@ -5056,6 +5149,10 @@ static int gfx_v8_0_hw_fini(void *handle)
 	amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0);
 	amdgpu_irq_put(adev, &adev->gfx.priv_inst_irq, 0);
 
+	amdgpu_irq_put(adev, &adev->gfx.cp_ecc_error_irq, 0);
+
+	amdgpu_irq_put(adev, &adev->gfx.sq_irq, 0);
+
 	/* disable KCQ to avoid CPC touch memory not valid anymore */
 	for (i = 0; i < adev->gfx.num_compute_rings; i++)
 		gfx_v8_0_kcq_disable(&adev->gfx.kiq.ring, &adev->gfx.compute_ring[i]);
@@ -5487,9 +5584,19 @@ static int gfx_v8_0_late_init(void *handle)
 	if (r)
 		return r;
 
-	amdgpu_device_ip_set_powergating_state(adev,
-					       AMD_IP_BLOCK_TYPE_GFX,
-					       AMD_PG_STATE_GATE);
+	r = amdgpu_irq_get(adev, &adev->gfx.cp_ecc_error_irq, 0);
+	if (r) {
+		DRM_ERROR("amdgpu_irq_get() failed to get IRQ for EDC, r: %d.\n", r);
+		return r;
+	}
+
+	r = amdgpu_irq_get(adev, &adev->gfx.sq_irq, 0);
+	if (r) {
+		DRM_ERROR(
+			"amdgpu_irq_get() failed to get IRQ for SQ, r: %d.\n",
+			r);
+		return r;
+	}
 
 	return 0;
 }
@@ -5497,13 +5604,12 @@ static int gfx_v8_0_late_init(void *handle)
 static void gfx_v8_0_enable_gfx_static_mg_power_gating(struct amdgpu_device *adev,
 						       bool enable)
 {
-	if ((adev->asic_type == CHIP_POLARIS11) ||
-	    (adev->asic_type == CHIP_POLARIS12))
+	if (((adev->asic_type == CHIP_POLARIS11) ||
+	    (adev->asic_type == CHIP_POLARIS12) ||
+	    (adev->asic_type == CHIP_VEGAM)) &&
+	    adev->powerplay.pp_funcs->set_powergating_by_smu)
 		/* Send msg to SMU via Powerplay */
-		amdgpu_device_ip_set_powergating_state(adev,
-						       AMD_IP_BLOCK_TYPE_SMC,
-						       enable ?
-						       AMD_PG_STATE_GATE : AMD_PG_STATE_UNGATE);
+		amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_GFX, enable);
 
 	WREG32_FIELD(RLC_PG_CNTL, STATIC_PER_CU_PG_ENABLE, enable ? 1 : 0);
 }
@@ -5588,6 +5694,7 @@ static int gfx_v8_0_set_powergating_state(void *handle,
 		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 		if ((adev->pg_flags & AMD_PG_SUPPORT_GFX_SMG) && enable)
 			gfx_v8_0_enable_gfx_static_mg_power_gating(adev, true);
 		else
@@ -6154,6 +6261,7 @@ static int gfx_v8_0_set_clockgating_state(void *handle,
 	case CHIP_POLARIS10:
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 		gfx_v8_0_polaris_update_gfx_clock_gating(adev, state);
 		break;
 	default:
@@ -6729,6 +6837,77 @@ static int gfx_v8_0_set_eop_interrupt_state(struct amdgpu_device *adev,
 	return 0;
 }
 
+static int gfx_v8_0_set_cp_ecc_int_state(struct amdgpu_device *adev,
+					 struct amdgpu_irq_src *source,
+					 unsigned int type,
+					 enum amdgpu_interrupt_state state)
+{
+	int enable_flag;
+
+	switch (state) {
+	case AMDGPU_IRQ_STATE_DISABLE:
+		enable_flag = 0;
+		break;
+
+	case AMDGPU_IRQ_STATE_ENABLE:
+		enable_flag = 1;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	WREG32_FIELD(CP_INT_CNTL, CP_ECC_ERROR_INT_ENABLE, enable_flag);
+	WREG32_FIELD(CP_INT_CNTL_RING0, CP_ECC_ERROR_INT_ENABLE, enable_flag);
+	WREG32_FIELD(CP_INT_CNTL_RING1, CP_ECC_ERROR_INT_ENABLE, enable_flag);
+	WREG32_FIELD(CP_INT_CNTL_RING2, CP_ECC_ERROR_INT_ENABLE, enable_flag);
+	WREG32_FIELD(CPC_INT_CNTL, CP_ECC_ERROR_INT_ENABLE, enable_flag);
+	WREG32_FIELD(CP_ME1_PIPE0_INT_CNTL, CP_ECC_ERROR_INT_ENABLE,
+		     enable_flag);
+	WREG32_FIELD(CP_ME1_PIPE1_INT_CNTL, CP_ECC_ERROR_INT_ENABLE,
+		     enable_flag);
+	WREG32_FIELD(CP_ME1_PIPE2_INT_CNTL, CP_ECC_ERROR_INT_ENABLE,
+		     enable_flag);
+	WREG32_FIELD(CP_ME1_PIPE3_INT_CNTL, CP_ECC_ERROR_INT_ENABLE,
+		     enable_flag);
+	WREG32_FIELD(CP_ME2_PIPE0_INT_CNTL, CP_ECC_ERROR_INT_ENABLE,
+		     enable_flag);
+	WREG32_FIELD(CP_ME2_PIPE1_INT_CNTL, CP_ECC_ERROR_INT_ENABLE,
+		     enable_flag);
+	WREG32_FIELD(CP_ME2_PIPE2_INT_CNTL, CP_ECC_ERROR_INT_ENABLE,
+		     enable_flag);
+	WREG32_FIELD(CP_ME2_PIPE3_INT_CNTL, CP_ECC_ERROR_INT_ENABLE,
+		     enable_flag);
+
+	return 0;
+}
+
+static int gfx_v8_0_set_sq_int_state(struct amdgpu_device *adev,
+				     struct amdgpu_irq_src *source,
+				     unsigned int type,
+				     enum amdgpu_interrupt_state state)
+{
+	int enable_flag;
+
+	switch (state) {
+	case AMDGPU_IRQ_STATE_DISABLE:
+		enable_flag = 1;
+		break;
+
+	case AMDGPU_IRQ_STATE_ENABLE:
+		enable_flag = 0;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	WREG32_FIELD(SQ_INTERRUPT_MSG_CTRL, STALL,
+		     enable_flag);
+
+	return 0;
+}
+
 static int gfx_v8_0_eop_irq(struct amdgpu_device *adev,
 			    struct amdgpu_irq_src *source,
 			    struct amdgpu_iv_entry *entry)
@@ -6779,6 +6958,114 @@ static int gfx_v8_0_priv_inst_irq(struct amdgpu_device *adev,
 	return 0;
 }
 
+static int gfx_v8_0_cp_ecc_error_irq(struct amdgpu_device *adev,
+				     struct amdgpu_irq_src *source,
+				     struct amdgpu_iv_entry *entry)
+{
+	DRM_ERROR("CP EDC/ECC error detected.");
+	return 0;
+}
+
+static void gfx_v8_0_parse_sq_irq(struct amdgpu_device *adev, unsigned ih_data)
+{
+	u32 enc, se_id, sh_id, cu_id;
+	char type[20];
+	int sq_edc_source = -1;
+
+	enc = REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_CMN, ENCODING);
+	se_id = REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_CMN, SE_ID);
+
+	switch (enc) {
+		case 0:
+			DRM_INFO("SQ general purpose intr detected:"
+					"se_id %d, immed_overflow %d, host_reg_overflow %d,"
+					"host_cmd_overflow %d, cmd_timestamp %d,"
+					"reg_timestamp %d, thread_trace_buff_full %d,"
+					"wlt %d, thread_trace %d.\n",
+					se_id,
+					REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, IMMED_OVERFLOW),
+					REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, HOST_REG_OVERFLOW),
+					REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, HOST_CMD_OVERFLOW),
+					REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, CMD_TIMESTAMP),
+					REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, REG_TIMESTAMP),
+					REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, THREAD_TRACE_BUF_FULL),
+					REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, WLT),
+					REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_AUTO, THREAD_TRACE)
+					);
+			break;
+		case 1:
+		case 2:
+
+			cu_id = REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_WAVE, CU_ID);
+			sh_id = REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_WAVE, SH_ID);
+
+			/*
+			 * This function can be called either directly from ISR
+			 * or from BH in which case we can access SQ_EDC_INFO
+			 * instance
+			 */
+			if (in_task()) {
+				mutex_lock(&adev->grbm_idx_mutex);
+				gfx_v8_0_select_se_sh(adev, se_id, sh_id, cu_id);
+
+				sq_edc_source = REG_GET_FIELD(RREG32(mmSQ_EDC_INFO), SQ_EDC_INFO, SOURCE);
+
+				gfx_v8_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff);
+				mutex_unlock(&adev->grbm_idx_mutex);
+			}
+
+			if (enc == 1)
+				sprintf(type, "instruction intr");
+			else
+				sprintf(type, "EDC/ECC error");
+
+			DRM_INFO(
+				"SQ %s detected: "
+					"se_id %d, sh_id %d, cu_id %d, simd_id %d, wave_id %d, vm_id %d "
+					"trap %s, sq_ed_info.source %s.\n",
+					type, se_id, sh_id, cu_id,
+					REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_WAVE, SIMD_ID),
+					REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_WAVE, WAVE_ID),
+					REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_WAVE, VM_ID),
+					REG_GET_FIELD(ih_data, SQ_INTERRUPT_WORD_WAVE, PRIV) ? "true" : "false",
+					(sq_edc_source != -1) ? sq_edc_source_names[sq_edc_source] : "unavailable"
+				);
+			break;
+		default:
+			DRM_ERROR("SQ invalid encoding type\n.");
+	}
+}
+
+static void gfx_v8_0_sq_irq_work_func(struct work_struct *work)
+{
+
+	struct amdgpu_device *adev = container_of(work, struct amdgpu_device, gfx.sq_work.work);
+	struct sq_work *sq_work = container_of(work, struct sq_work, work);
+
+	gfx_v8_0_parse_sq_irq(adev, sq_work->ih_data);
+}
+
+static int gfx_v8_0_sq_irq(struct amdgpu_device *adev,
+			   struct amdgpu_irq_src *source,
+			   struct amdgpu_iv_entry *entry)
+{
+	unsigned ih_data = entry->src_data[0];
+
+	/*
+	 * Try to submit work so SQ_EDC_INFO can be accessed from
+	 * BH. If previous work submission hasn't finished yet
+	 * just print whatever info is possible directly from the ISR.
+	 */
+	if (work_pending(&adev->gfx.sq_work.work)) {
+		gfx_v8_0_parse_sq_irq(adev, ih_data);
+	} else {
+		adev->gfx.sq_work.ih_data = ih_data;
+		schedule_work(&adev->gfx.sq_work.work);
+	}
+
+	return 0;
+}
+
 static int gfx_v8_0_kiq_set_interrupt_state(struct amdgpu_device *adev,
 					    struct amdgpu_irq_src *src,
 					    unsigned int type,
@@ -6979,6 +7266,16 @@ static const struct amdgpu_irq_src_funcs gfx_v8_0_kiq_irq_funcs = {
 	.process = gfx_v8_0_kiq_irq,
 };
 
+static const struct amdgpu_irq_src_funcs gfx_v8_0_cp_ecc_error_irq_funcs = {
+	.set = gfx_v8_0_set_cp_ecc_int_state,
+	.process = gfx_v8_0_cp_ecc_error_irq,
+};
+
+static const struct amdgpu_irq_src_funcs gfx_v8_0_sq_irq_funcs = {
+	.set = gfx_v8_0_set_sq_int_state,
+	.process = gfx_v8_0_sq_irq,
+};
+
 static void gfx_v8_0_set_irq_funcs(struct amdgpu_device *adev)
 {
 	adev->gfx.eop_irq.num_types = AMDGPU_CP_IRQ_LAST;
@@ -6992,6 +7289,12 @@ static void gfx_v8_0_set_irq_funcs(struct amdgpu_device *adev)
 
 	adev->gfx.kiq.irq.num_types = AMDGPU_CP_KIQ_IRQ_LAST;
 	adev->gfx.kiq.irq.funcs = &gfx_v8_0_kiq_irq_funcs;
+
+	adev->gfx.cp_ecc_error_irq.num_types = 1;
+	adev->gfx.cp_ecc_error_irq.funcs = &gfx_v8_0_cp_ecc_error_irq_funcs;
+
+	adev->gfx.sq_irq.num_types = 1;
+	adev->gfx.sq_irq.funcs = &gfx_v8_0_sq_irq_funcs;
 }
 
 static void gfx_v8_0_set_rlc_funcs(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index 9d39fd5b1822..9ab39117cc4e 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -27,6 +27,7 @@
 #include "amdgpu_gfx.h"
 #include "soc15.h"
 #include "soc15d.h"
+#include "amdgpu_atomfirmware.h"
 
 #include "gc/gc_9_0_offset.h"
 #include "gc/gc_9_0_sh_mask.h"
@@ -37,11 +38,12 @@
 #include "clearstate_gfx9.h"
 #include "v9_structs.h"
 
+#include "ivsrcid/gfx/irqsrcs_gfx_9_0.h"
+
 #define GFX9_NUM_GFX_RINGS     1
 #define GFX9_MEC_HPD_SIZE 2048
 #define RLCG_UCODE_LOADING_START_ADDRESS 0x00002000L
 #define RLC_SAVE_RESTORE_ADDR_STARTING_OFFSET 0x00000000L
-#define GFX9_RLC_FORMAT_DIRECT_REG_LIST_LENGTH 34
 
 #define mmPWR_MISC_CNTL_STATUS					0x0183
 #define mmPWR_MISC_CNTL_STATUS_BASE_IDX				0
@@ -64,6 +66,13 @@ MODULE_FIRMWARE("amdgpu/vega12_mec.bin");
 MODULE_FIRMWARE("amdgpu/vega12_mec2.bin");
 MODULE_FIRMWARE("amdgpu/vega12_rlc.bin");
 
+MODULE_FIRMWARE("amdgpu/vega20_ce.bin");
+MODULE_FIRMWARE("amdgpu/vega20_pfp.bin");
+MODULE_FIRMWARE("amdgpu/vega20_me.bin");
+MODULE_FIRMWARE("amdgpu/vega20_mec.bin");
+MODULE_FIRMWARE("amdgpu/vega20_mec2.bin");
+MODULE_FIRMWARE("amdgpu/vega20_rlc.bin");
+
 MODULE_FIRMWARE("amdgpu/raven_ce.bin");
 MODULE_FIRMWARE("amdgpu/raven_pfp.bin");
 MODULE_FIRMWARE("amdgpu/raven_me.bin");
@@ -73,40 +82,59 @@ MODULE_FIRMWARE("amdgpu/raven_rlc.bin");
 
 static const struct soc15_reg_golden golden_settings_gc_9_0[] =
 {
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPC_UTCL1_CNTL, 0x08000000, 0x08000080),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPF_UTCL1_CNTL, 0x08000000, 0x08000080),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPG_UTCL1_CNTL, 0x08000000, 0x08000080),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG2, 0xf00fffff, 0x00000420),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_GPU_ID, 0x0000000f, 0x00000000),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmIA_UTCL1_CNTL, 0x08000000, 0x08000080),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_BINNER_EVENT_CNTL_3, 0x00000003, 0x82400024),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE, 0x3fffffff, 0x00000001),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_LINE_STIPPLE_STATE, 0x0000ff0f, 0x00000000),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_0, 0x08000000, 0x08000080),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_1, 0x08000000, 0x08000080),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_2, 0x08000000, 0x08000080),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_PREWALKER_UTCL1_CNTL, 0x08000000, 0x08000080),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_SPM_UTCL1_CNTL, 0x08000000, 0x08000080),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSH_MEM_CONFIG, 0x00001000, 0x00001000),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_CONFIG_CNTL_1, 0x0000000f, 0x01000107),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_CU_0, 0x0007ffff, 0x00000800),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_CU_1, 0x0007ffff, 0x00000800),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_EN_CU_0, 0x01ffffff, 0x0000ff87),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_RESOURCE_RESERVE_EN_CU_1, 0x01ffffff, 0x0000ff8f),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQC_CONFIG, 0x03000000, 0x020a2000),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTA_CNTL_AUX, 0xfffffeef, 0x010b0000),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_HI, 0xffffffff, 0x4a2c0e68),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTCP_CHAN_STEER_LO, 0xffffffff, 0xb5d3f197),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_CACHE_INVALIDATION, 0x3fff3af3, 0x19200000),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_GS_MAX_WAVE_ID, 0x00000fff, 0x000003ff),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmWD_UTCL1_CNTL, 0x08000000, 0x08000080)
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmVGT_GS_MAX_WAVE_ID, 0x00000fff, 0x000003ff)
 };
 
 static const struct soc15_reg_golden golden_settings_gc_9_0_vg10[] =
 {
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL, 0x0000f000, 0x00012107),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_3, 0x30000000, 0x10000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPC_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPF_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCPG_UTCL1_CNTL, 0x08000000, 0x08000080),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG, 0xffff77ff, 0x2a114042),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG_READ, 0xffff77ff, 0x2a114042),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmIA_UTCL1_CNTL, 0x08000000, 0x08000080),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE_1, 0x00008000, 0x00048000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_0, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_1, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_GPM_UTCL1_CNTL_2, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_PREWALKER_UTCL1_CNTL, 0x08000000, 0x08000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRLC_SPM_UTCL1_CNTL, 0x08000000, 0x08000080),
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRMI_UTCL1_CNTL2, 0x00030000, 0x00020000),
-	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTD_CNTL, 0x00001800, 0x00000800)
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_CONFIG_CNTL_1, 0x0000000f, 0x01000107),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTD_CNTL, 0x00001800, 0x00000800),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmWD_UTCL1_CNTL, 0x08000000, 0x08000080)
+};
+
+static const struct soc15_reg_golden golden_settings_gc_9_0_vg20[] =
+{
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_DCC_CONFIG, 0x0f000080, 0x04000080),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_2, 0x0f000000, 0x0a000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_3, 0x30000000, 0x10000000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG, 0xf3e777ff, 0x22014042),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG_READ, 0xf3e777ff, 0x22014042),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG2, 0x00003e00, 0x00000400),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE_1, 0xff840000, 0x04040000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmRMI_UTCL1_CNTL2, 0x00030000, 0x00030000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmSPI_CONFIG_CNTL_1, 0xffff010f, 0x01000107),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTA_CNTL_AUX, 0x000b0000, 0x000b0000),
+	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTD_CNTL, 0x01000000, 0x01000000)
 };
 
 static const struct soc15_reg_golden golden_settings_gc_9_1[] =
@@ -185,6 +213,30 @@ static const struct soc15_reg_golden golden_settings_gc_9_2_1_vg12[] =
 	SOC15_REG_GOLDEN_VALUE(GC, 0, mmTD_CNTL, 0x01bd9f33, 0x01000000)
 };
 
+static const u32 GFX_RLC_SRM_INDEX_CNTL_ADDR_OFFSETS[] =
+{
+	mmRLC_SRM_INDEX_CNTL_ADDR_0 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+	mmRLC_SRM_INDEX_CNTL_ADDR_1 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+	mmRLC_SRM_INDEX_CNTL_ADDR_2 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+	mmRLC_SRM_INDEX_CNTL_ADDR_3 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+	mmRLC_SRM_INDEX_CNTL_ADDR_4 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+	mmRLC_SRM_INDEX_CNTL_ADDR_5 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+	mmRLC_SRM_INDEX_CNTL_ADDR_6 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+	mmRLC_SRM_INDEX_CNTL_ADDR_7 - mmRLC_SRM_INDEX_CNTL_ADDR_0,
+};
+
+static const u32 GFX_RLC_SRM_INDEX_CNTL_DATA_OFFSETS[] =
+{
+	mmRLC_SRM_INDEX_CNTL_DATA_0 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+	mmRLC_SRM_INDEX_CNTL_DATA_1 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+	mmRLC_SRM_INDEX_CNTL_DATA_2 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+	mmRLC_SRM_INDEX_CNTL_DATA_3 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+	mmRLC_SRM_INDEX_CNTL_DATA_4 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+	mmRLC_SRM_INDEX_CNTL_DATA_5 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+	mmRLC_SRM_INDEX_CNTL_DATA_6 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+	mmRLC_SRM_INDEX_CNTL_DATA_7 - mmRLC_SRM_INDEX_CNTL_DATA_0,
+};
+
 #define VEGA10_GB_ADDR_CONFIG_GOLDEN 0x2a114042
 #define VEGA12_GB_ADDR_CONFIG_GOLDEN 0x24104041
 #define RAVEN_GB_ADDR_CONFIG_GOLDEN 0x24000042
@@ -218,6 +270,14 @@ static void gfx_v9_0_init_golden_registers(struct amdgpu_device *adev)
 						golden_settings_gc_9_2_1_vg12,
 						ARRAY_SIZE(golden_settings_gc_9_2_1_vg12));
 		break;
+	case CHIP_VEGA20:
+		soc15_program_register_sequence(adev,
+						golden_settings_gc_9_0,
+						ARRAY_SIZE(golden_settings_gc_9_0));
+		soc15_program_register_sequence(adev,
+						golden_settings_gc_9_0_vg20,
+						ARRAY_SIZE(golden_settings_gc_9_0_vg20));
+		break;
 	case CHIP_RAVEN:
 		soc15_program_register_sequence(adev,
 						 golden_settings_gc_9_1,
@@ -401,6 +461,27 @@ static void gfx_v9_0_free_microcode(struct amdgpu_device *adev)
 	kfree(adev->gfx.rlc.register_list_format);
 }
 
+static void gfx_v9_0_init_rlc_ext_microcode(struct amdgpu_device *adev)
+{
+	const struct rlc_firmware_header_v2_1 *rlc_hdr;
+
+	rlc_hdr = (const struct rlc_firmware_header_v2_1 *)adev->gfx.rlc_fw->data;
+	adev->gfx.rlc_srlc_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_cntl_ucode_ver);
+	adev->gfx.rlc_srlc_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_cntl_feature_ver);
+	adev->gfx.rlc.save_restore_list_cntl_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_cntl_size_bytes);
+	adev->gfx.rlc.save_restore_list_cntl = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_cntl_offset_bytes);
+	adev->gfx.rlc_srlg_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_gpm_ucode_ver);
+	adev->gfx.rlc_srlg_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_gpm_feature_ver);
+	adev->gfx.rlc.save_restore_list_gpm_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_gpm_size_bytes);
+	adev->gfx.rlc.save_restore_list_gpm = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_gpm_offset_bytes);
+	adev->gfx.rlc_srls_fw_version = le32_to_cpu(rlc_hdr->save_restore_list_srm_ucode_ver);
+	adev->gfx.rlc_srls_feature_version = le32_to_cpu(rlc_hdr->save_restore_list_srm_feature_ver);
+	adev->gfx.rlc.save_restore_list_srm_size_bytes = le32_to_cpu(rlc_hdr->save_restore_list_srm_size_bytes);
+	adev->gfx.rlc.save_restore_list_srm = (u8 *)rlc_hdr + le32_to_cpu(rlc_hdr->save_restore_list_srm_offset_bytes);
+	adev->gfx.rlc.reg_list_format_direct_reg_list_length =
+			le32_to_cpu(rlc_hdr->reg_list_format_direct_reg_list_length);
+}
+
 static int gfx_v9_0_init_microcode(struct amdgpu_device *adev)
 {
 	const char *chip_name;
@@ -412,6 +493,8 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev)
 	const struct rlc_firmware_header_v2_0 *rlc_hdr;
 	unsigned int *tmp = NULL;
 	unsigned int i = 0;
+	uint16_t version_major;
+	uint16_t version_minor;
 
 	DRM_DEBUG("\n");
 
@@ -422,6 +505,9 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_VEGA12:
 		chip_name = "vega12";
 		break;
+	case CHIP_VEGA20:
+		chip_name = "vega20";
+		break;
 	case CHIP_RAVEN:
 		chip_name = "raven";
 		break;
@@ -468,6 +554,12 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev)
 		goto out;
 	err = amdgpu_ucode_validate(adev->gfx.rlc_fw);
 	rlc_hdr = (const struct rlc_firmware_header_v2_0 *)adev->gfx.rlc_fw->data;
+
+	version_major = le16_to_cpu(rlc_hdr->header.header_version_major);
+	version_minor = le16_to_cpu(rlc_hdr->header.header_version_minor);
+	if (version_major == 2 && version_minor == 1)
+		adev->gfx.rlc.is_rlc_v2_1 = true;
+
 	adev->gfx.rlc_fw_version = le32_to_cpu(rlc_hdr->header.ucode_version);
 	adev->gfx.rlc_feature_version = le32_to_cpu(rlc_hdr->ucode_feature_version);
 	adev->gfx.rlc.save_and_restore_offset =
@@ -508,6 +600,9 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev)
 	for (i = 0 ; i < (rlc_hdr->reg_list_size_bytes >> 2); i++)
 		adev->gfx.rlc.register_restore[i] = le32_to_cpu(tmp[i]);
 
+	if (adev->gfx.rlc.is_rlc_v2_1)
+		gfx_v9_0_init_rlc_ext_microcode(adev);
+
 	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mec.bin", chip_name);
 	err = request_firmware(&adev->gfx.mec_fw, fw_name, adev->dev);
 	if (err)
@@ -566,6 +661,29 @@ static int gfx_v9_0_init_microcode(struct amdgpu_device *adev)
 		adev->firmware.fw_size +=
 			ALIGN(le32_to_cpu(header->ucode_size_bytes), PAGE_SIZE);
 
+		if (adev->gfx.rlc.is_rlc_v2_1 &&
+		    adev->gfx.rlc.save_restore_list_cntl_size_bytes &&
+		    adev->gfx.rlc.save_restore_list_gpm_size_bytes &&
+		    adev->gfx.rlc.save_restore_list_srm_size_bytes) {
+			info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL];
+			info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL;
+			info->fw = adev->gfx.rlc_fw;
+			adev->firmware.fw_size +=
+				ALIGN(adev->gfx.rlc.save_restore_list_cntl_size_bytes, PAGE_SIZE);
+
+			info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM];
+			info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM;
+			info->fw = adev->gfx.rlc_fw;
+			adev->firmware.fw_size +=
+				ALIGN(adev->gfx.rlc.save_restore_list_gpm_size_bytes, PAGE_SIZE);
+
+			info = &adev->firmware.ucode[AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM];
+			info->ucode_id = AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM;
+			info->fw = adev->gfx.rlc_fw;
+			adev->firmware.fw_size +=
+				ALIGN(adev->gfx.rlc.save_restore_list_srm_size_bytes, PAGE_SIZE);
+		}
+
 		info = &adev->firmware.ucode[AMDGPU_UCODE_ID_CP_MEC1];
 		info->ucode_id = AMDGPU_UCODE_ID_CP_MEC1;
 		info->fw = adev->gfx.mec_fw;
@@ -841,6 +959,7 @@ static int gfx_v9_0_rlc_init(struct amdgpu_device *adev)
 		dst_ptr = adev->gfx.rlc.cs_ptr;
 		gfx_v9_0_get_csb_buffer(adev, dst_ptr);
 		amdgpu_bo_kunmap(adev->gfx.rlc.clear_state_obj);
+		amdgpu_bo_unpin(adev->gfx.rlc.clear_state_obj);
 		amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj);
 	}
 
@@ -869,6 +988,39 @@ static int gfx_v9_0_rlc_init(struct amdgpu_device *adev)
 	return 0;
 }
 
+static int gfx_v9_0_csb_vram_pin(struct amdgpu_device *adev)
+{
+	int r;
+
+	r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, false);
+	if (unlikely(r != 0))
+		return r;
+
+	r = amdgpu_bo_pin(adev->gfx.rlc.clear_state_obj,
+			AMDGPU_GEM_DOMAIN_VRAM);
+	if (!r)
+		adev->gfx.rlc.clear_state_gpu_addr =
+			amdgpu_bo_gpu_offset(adev->gfx.rlc.clear_state_obj);
+
+	amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj);
+
+	return r;
+}
+
+static void gfx_v9_0_csb_vram_unpin(struct amdgpu_device *adev)
+{
+	int r;
+
+	if (!adev->gfx.rlc.clear_state_obj)
+		return;
+
+	r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, true);
+	if (likely(r == 0)) {
+		amdgpu_bo_unpin(adev->gfx.rlc.clear_state_obj);
+		amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj);
+	}
+}
+
 static void gfx_v9_0_mec_fini(struct amdgpu_device *adev)
 {
 	amdgpu_bo_free_kernel(&adev->gfx.mec.hpd_eop_obj, NULL, NULL);
@@ -1013,9 +1165,10 @@ static const struct amdgpu_gfx_funcs gfx_v9_0_gfx_funcs = {
 	.select_me_pipe_q = &gfx_v9_0_select_me_pipe_q
 };
 
-static void gfx_v9_0_gpu_early_init(struct amdgpu_device *adev)
+static int gfx_v9_0_gpu_early_init(struct amdgpu_device *adev)
 {
 	u32 gb_addr_config;
+	int err;
 
 	adev->gfx.funcs = &gfx_v9_0_gfx_funcs;
 
@@ -1037,6 +1190,20 @@ static void gfx_v9_0_gpu_early_init(struct amdgpu_device *adev)
 		gb_addr_config = VEGA12_GB_ADDR_CONFIG_GOLDEN;
 		DRM_INFO("fix gfx.config for vega12\n");
 		break;
+	case CHIP_VEGA20:
+		adev->gfx.config.max_hw_contexts = 8;
+		adev->gfx.config.sc_prim_fifo_size_frontend = 0x20;
+		adev->gfx.config.sc_prim_fifo_size_backend = 0x100;
+		adev->gfx.config.sc_hiz_tile_fifo_size = 0x30;
+		adev->gfx.config.sc_earlyz_tile_fifo_size = 0x4C0;
+		gb_addr_config = RREG32_SOC15(GC, 0, mmGB_ADDR_CONFIG);
+		gb_addr_config &= ~0xf3e777ff;
+		gb_addr_config |= 0x22014042;
+		/* check vbios table if gpu info is not available */
+		err = amdgpu_atomfirmware_get_gfx_info(adev);
+		if (err)
+			return err;
+		break;
 	case CHIP_RAVEN:
 		adev->gfx.config.max_hw_contexts = 8;
 		adev->gfx.config.sc_prim_fifo_size_frontend = 0x20;
@@ -1086,6 +1253,8 @@ static void gfx_v9_0_gpu_early_init(struct amdgpu_device *adev)
 					adev->gfx.config.gb_addr_config,
 					GB_ADDR_CONFIG,
 					PIPE_INTERLEAVE_SIZE));
+
+	return 0;
 }
 
 static int gfx_v9_0_ngg_create_buf(struct amdgpu_device *adev,
@@ -1319,6 +1488,7 @@ static int gfx_v9_0_sw_init(void *handle)
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 	case CHIP_RAVEN:
 		adev->gfx.mec.num_mec = 2;
 		break;
@@ -1331,23 +1501,23 @@ static int gfx_v9_0_sw_init(void *handle)
 	adev->gfx.mec.num_queue_per_pipe = 8;
 
 	/* KIQ event */
-	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, 178, &adev->gfx.kiq.irq);
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, GFX_9_0__SRCID__CP_IB2_INTERRUPT_PKT, &adev->gfx.kiq.irq);
 	if (r)
 		return r;
 
 	/* EOP Event */
-	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, 181, &adev->gfx.eop_irq);
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, GFX_9_0__SRCID__CP_EOP_INTERRUPT, &adev->gfx.eop_irq);
 	if (r)
 		return r;
 
 	/* Privileged reg */
-	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, 184,
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, GFX_9_0__SRCID__CP_PRIV_REG_FAULT,
 			      &adev->gfx.priv_reg_irq);
 	if (r)
 		return r;
 
 	/* Privileged inst */
-	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, 185,
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_GRBM_CP, GFX_9_0__SRCID__CP_PRIV_INSTR_FAULT,
 			      &adev->gfx.priv_inst_irq);
 	if (r)
 		return r;
@@ -1446,7 +1616,9 @@ static int gfx_v9_0_sw_init(void *handle)
 
 	adev->gfx.ce_ram_size = 0x8000;
 
-	gfx_v9_0_gpu_early_init(adev);
+	r = gfx_v9_0_gpu_early_init(adev);
+	if (r)
+		return r;
 
 	r = gfx_v9_0_ngg_init(adev);
 	if (r)
@@ -1600,6 +1772,7 @@ static void gfx_v9_0_gpu_init(struct amdgpu_device *adev)
 
 	gfx_v9_0_setup_rb(adev);
 	gfx_v9_0_get_cu_info(adev, &adev->gfx.cu_info);
+	adev->gfx.config.db_debug2 = RREG32_SOC15(GC, 0, mmDB_DEBUG2);
 
 	/* XXX SH_MEM regs */
 	/* where to put LDS, scratch, GPUVM in FSA64 space */
@@ -1616,7 +1789,10 @@ static void gfx_v9_0_gpu_init(struct amdgpu_device *adev)
 			tmp = REG_SET_FIELD(0, SH_MEM_CONFIG, ALIGNMENT_MODE,
 					    SH_MEM_ALIGNMENT_MODE_UNALIGNED);
 			WREG32_SOC15(GC, 0, mmSH_MEM_CONFIG, tmp);
-			tmp = adev->gmc.shared_aperture_start >> 48;
+			tmp = REG_SET_FIELD(0, SH_MEM_BASES, PRIVATE_BASE,
+				(adev->gmc.private_aperture_start >> 48));
+			tmp = REG_SET_FIELD(tmp, SH_MEM_BASES, SHARED_BASE,
+				(adev->gmc.shared_aperture_start >> 48));
 			WREG32_SOC15(GC, 0, mmSH_MEM_BASES, tmp);
 		}
 	}
@@ -1708,55 +1884,44 @@ static void gfx_v9_0_init_csb(struct amdgpu_device *adev)
 			adev->gfx.rlc.clear_state_size);
 }
 
-static void gfx_v9_0_parse_ind_reg_list(int *register_list_format,
+static void gfx_v9_1_parse_ind_reg_list(int *register_list_format,
 				int indirect_offset,
 				int list_size,
 				int *unique_indirect_regs,
-				int *unique_indirect_reg_count,
-				int max_indirect_reg_count,
+				int unique_indirect_reg_count,
 				int *indirect_start_offsets,
 				int *indirect_start_offsets_count,
-				int max_indirect_start_offsets_count)
+				int max_start_offsets_count)
 {
 	int idx;
-	bool new_entry = true;
 
 	for (; indirect_offset < list_size; indirect_offset++) {
+		WARN_ON(*indirect_start_offsets_count >= max_start_offsets_count);
+		indirect_start_offsets[*indirect_start_offsets_count] = indirect_offset;
+		*indirect_start_offsets_count = *indirect_start_offsets_count + 1;
+
+		while (register_list_format[indirect_offset] != 0xFFFFFFFF) {
+			indirect_offset += 2;
+
+			/* look for the matching indice */
+			for (idx = 0; idx < unique_indirect_reg_count; idx++) {
+				if (unique_indirect_regs[idx] ==
+					register_list_format[indirect_offset] ||
+					!unique_indirect_regs[idx])
+					break;
+			}
 
-		if (new_entry) {
-			new_entry = false;
-			indirect_start_offsets[*indirect_start_offsets_count] = indirect_offset;
-			*indirect_start_offsets_count = *indirect_start_offsets_count + 1;
-			BUG_ON(*indirect_start_offsets_count >= max_indirect_start_offsets_count);
-		}
-
-		if (register_list_format[indirect_offset] == 0xFFFFFFFF) {
-			new_entry = true;
-			continue;
-		}
-
-		indirect_offset += 2;
+			BUG_ON(idx >= unique_indirect_reg_count);
 
-		/* look for the matching indice */
-		for (idx = 0; idx < *unique_indirect_reg_count; idx++) {
-			if (unique_indirect_regs[idx] ==
-				register_list_format[indirect_offset])
-				break;
-		}
+			if (!unique_indirect_regs[idx])
+				unique_indirect_regs[idx] = register_list_format[indirect_offset];
 
-		if (idx >= *unique_indirect_reg_count) {
-			unique_indirect_regs[*unique_indirect_reg_count] =
-				register_list_format[indirect_offset];
-			idx = *unique_indirect_reg_count;
-			*unique_indirect_reg_count = *unique_indirect_reg_count + 1;
-			BUG_ON(*unique_indirect_reg_count >= max_indirect_reg_count);
+			indirect_offset++;
 		}
-
-		register_list_format[indirect_offset] = idx;
 	}
 }
 
-static int gfx_v9_0_init_rlc_save_restore_list(struct amdgpu_device *adev)
+static int gfx_v9_1_init_rlc_save_restore_list(struct amdgpu_device *adev)
 {
 	int unique_indirect_regs[] = {0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0};
 	int unique_indirect_reg_count = 0;
@@ -1765,7 +1930,7 @@ static int gfx_v9_0_init_rlc_save_restore_list(struct amdgpu_device *adev)
 	int indirect_start_offsets_count = 0;
 
 	int list_size = 0;
-	int i = 0;
+	int i = 0, j = 0;
 	u32 tmp = 0;
 
 	u32 *register_list_format =
@@ -1776,15 +1941,15 @@ static int gfx_v9_0_init_rlc_save_restore_list(struct amdgpu_device *adev)
 		adev->gfx.rlc.reg_list_format_size_bytes);
 
 	/* setup unique_indirect_regs array and indirect_start_offsets array */
-	gfx_v9_0_parse_ind_reg_list(register_list_format,
-				GFX9_RLC_FORMAT_DIRECT_REG_LIST_LENGTH,
-				adev->gfx.rlc.reg_list_format_size_bytes >> 2,
-				unique_indirect_regs,
-				&unique_indirect_reg_count,
-				ARRAY_SIZE(unique_indirect_regs),
-				indirect_start_offsets,
-				&indirect_start_offsets_count,
-				ARRAY_SIZE(indirect_start_offsets));
+	unique_indirect_reg_count = ARRAY_SIZE(unique_indirect_regs);
+	gfx_v9_1_parse_ind_reg_list(register_list_format,
+				    adev->gfx.rlc.reg_list_format_direct_reg_list_length,
+				    adev->gfx.rlc.reg_list_format_size_bytes >> 2,
+				    unique_indirect_regs,
+				    unique_indirect_reg_count,
+				    indirect_start_offsets,
+				    &indirect_start_offsets_count,
+				    ARRAY_SIZE(indirect_start_offsets));
 
 	/* enable auto inc in case it is disabled */
 	tmp = RREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_CNTL));
@@ -1798,19 +1963,37 @@ static int gfx_v9_0_init_rlc_save_restore_list(struct amdgpu_device *adev)
 		WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_ARAM_DATA),
 			adev->gfx.rlc.register_restore[i]);
 
-	/* load direct register */
-	WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_ARAM_ADDR), 0);
-	for (i = 0; i < adev->gfx.rlc.reg_list_size_bytes >> 2; i++)
-		WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_ARAM_DATA),
-			adev->gfx.rlc.register_restore[i]);
-
 	/* load indirect register */
 	WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_SCRATCH_ADDR),
 		adev->gfx.rlc.reg_list_format_start);
-	for (i = 0; i < adev->gfx.rlc.reg_list_format_size_bytes >> 2; i++)
+
+	/* direct register portion */
+	for (i = 0; i < adev->gfx.rlc.reg_list_format_direct_reg_list_length; i++)
 		WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_SCRATCH_DATA),
 			register_list_format[i]);
 
+	/* indirect register portion */
+	while (i < (adev->gfx.rlc.reg_list_format_size_bytes >> 2)) {
+		if (register_list_format[i] == 0xFFFFFFFF) {
+			WREG32_SOC15(GC, 0, mmRLC_GPM_SCRATCH_DATA, register_list_format[i++]);
+			continue;
+		}
+
+		WREG32_SOC15(GC, 0, mmRLC_GPM_SCRATCH_DATA, register_list_format[i++]);
+		WREG32_SOC15(GC, 0, mmRLC_GPM_SCRATCH_DATA, register_list_format[i++]);
+
+		for (j = 0; j < unique_indirect_reg_count; j++) {
+			if (register_list_format[i] == unique_indirect_regs[j]) {
+				WREG32_SOC15(GC, 0, mmRLC_GPM_SCRATCH_DATA, j);
+				break;
+			}
+		}
+
+		BUG_ON(j >= unique_indirect_reg_count);
+
+		i++;
+	}
+
 	/* set save/restore list size */
 	list_size = adev->gfx.rlc.reg_list_size_bytes >> 2;
 	list_size = list_size >> 1;
@@ -1823,14 +2006,19 @@ static int gfx_v9_0_init_rlc_save_restore_list(struct amdgpu_device *adev)
 		adev->gfx.rlc.starting_offsets_start);
 	for (i = 0; i < ARRAY_SIZE(indirect_start_offsets); i++)
 		WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_SCRATCH_DATA),
-			indirect_start_offsets[i]);
+		       indirect_start_offsets[i]);
 
 	/* load unique indirect regs*/
 	for (i = 0; i < ARRAY_SIZE(unique_indirect_regs); i++) {
-		WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_INDEX_CNTL_ADDR_0) + i,
-			unique_indirect_regs[i] & 0x3FFFF);
-		WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_INDEX_CNTL_DATA_0) + i,
-			unique_indirect_regs[i] >> 20);
+		if (unique_indirect_regs[i] != 0) {
+			WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_INDEX_CNTL_ADDR_0)
+			       + GFX_RLC_SRM_INDEX_CNTL_ADDR_OFFSETS[i],
+			       unique_indirect_regs[i] & 0x3FFFF);
+
+			WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_SRM_INDEX_CNTL_DATA_0)
+			       + GFX_RLC_SRM_INDEX_CNTL_DATA_OFFSETS[i],
+			       unique_indirect_regs[i] >> 20);
+		}
 	}
 
 	kfree(register_list_format);
@@ -2010,34 +2198,26 @@ static void gfx_v9_0_enable_gfx_dynamic_mg_power_gating(struct amdgpu_device *ad
 
 static void gfx_v9_0_init_pg(struct amdgpu_device *adev)
 {
+	gfx_v9_0_init_csb(adev);
+
+	/*
+	 * Rlc save restore list is workable since v2_1.
+	 * And it's needed by gfxoff feature.
+	 */
+	if (adev->gfx.rlc.is_rlc_v2_1) {
+		gfx_v9_1_init_rlc_save_restore_list(adev);
+		gfx_v9_0_enable_save_restore_machine(adev);
+	}
+
 	if (adev->pg_flags & (AMD_PG_SUPPORT_GFX_PG |
 			      AMD_PG_SUPPORT_GFX_SMG |
 			      AMD_PG_SUPPORT_GFX_DMG |
 			      AMD_PG_SUPPORT_CP |
 			      AMD_PG_SUPPORT_GDS |
 			      AMD_PG_SUPPORT_RLC_SMU_HS)) {
-		gfx_v9_0_init_csb(adev);
-		gfx_v9_0_init_rlc_save_restore_list(adev);
-		gfx_v9_0_enable_save_restore_machine(adev);
-
-		if (adev->asic_type == CHIP_RAVEN) {
-			WREG32(mmRLC_JUMP_TABLE_RESTORE,
-				adev->gfx.rlc.cp_table_gpu_addr >> 8);
-			gfx_v9_0_init_gfx_power_gating(adev);
-
-			if (adev->pg_flags & AMD_PG_SUPPORT_RLC_SMU_HS) {
-				gfx_v9_0_enable_sck_slow_down_on_power_up(adev, true);
-				gfx_v9_0_enable_sck_slow_down_on_power_down(adev, true);
-			} else {
-				gfx_v9_0_enable_sck_slow_down_on_power_up(adev, false);
-				gfx_v9_0_enable_sck_slow_down_on_power_down(adev, false);
-			}
-
-			if (adev->pg_flags & AMD_PG_SUPPORT_CP)
-				gfx_v9_0_enable_cp_power_gating(adev, true);
-			else
-				gfx_v9_0_enable_cp_power_gating(adev, false);
-		}
+		WREG32(mmRLC_JUMP_TABLE_RESTORE,
+		       adev->gfx.rlc.cp_table_gpu_addr >> 8);
+		gfx_v9_0_init_gfx_power_gating(adev);
 	}
 }
 
@@ -2126,9 +2306,6 @@ static int gfx_v9_0_rlc_resume(struct amdgpu_device *adev)
 	/* disable CG */
 	WREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL, 0);
 
-	/* disable PG */
-	WREG32_SOC15(GC, 0, mmRLC_PG_CNTL, 0);
-
 	gfx_v9_0_rlc_reset(adev);
 
 	gfx_v9_0_init_pg(adev);
@@ -2990,6 +3167,10 @@ static int gfx_v9_0_hw_init(void *handle)
 
 	gfx_v9_0_gpu_init(adev);
 
+	r = gfx_v9_0_csb_vram_pin(adev);
+	if (r)
+		return r;
+
 	r = gfx_v9_0_rlc_resume(adev);
 	if (r)
 		return r;
@@ -3061,6 +3242,9 @@ static int gfx_v9_0_hw_fini(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	int i;
 
+	amdgpu_device_ip_set_powergating_state(adev, AMD_IP_BLOCK_TYPE_GFX,
+					       AMD_PG_STATE_UNGATE);
+
 	amdgpu_irq_put(adev, &adev->gfx.priv_reg_irq, 0);
 	amdgpu_irq_put(adev, &adev->gfx.priv_inst_irq, 0);
 
@@ -3095,6 +3279,8 @@ static int gfx_v9_0_hw_fini(void *handle)
 	gfx_v9_0_cp_enable(adev, false);
 	gfx_v9_0_rlc_stop(adev);
 
+	gfx_v9_0_csb_vram_unpin(adev);
+
 	return 0;
 }
 
@@ -3339,8 +3525,7 @@ static void gfx_v9_0_exit_rlc_safe_mode(struct amdgpu_device *adev)
 static void gfx_v9_0_update_gfx_cg_power_gating(struct amdgpu_device *adev,
 						bool enable)
 {
-	/* TODO: double check if we need to perform under safe mdoe */
-	/* gfx_v9_0_enter_rlc_safe_mode(adev); */
+	gfx_v9_0_enter_rlc_safe_mode(adev);
 
 	if ((adev->pg_flags & AMD_PG_SUPPORT_GFX_PG) && enable) {
 		gfx_v9_0_enable_gfx_cg_power_gating(adev, true);
@@ -3351,7 +3536,7 @@ static void gfx_v9_0_update_gfx_cg_power_gating(struct amdgpu_device *adev,
 		gfx_v9_0_enable_gfx_pipeline_powergating(adev, false);
 	}
 
-	/* gfx_v9_0_exit_rlc_safe_mode(adev); */
+	gfx_v9_0_exit_rlc_safe_mode(adev);
 }
 
 static void gfx_v9_0_update_gfx_mg_power_gating(struct amdgpu_device *adev,
@@ -3382,8 +3567,11 @@ static void gfx_v9_0_update_medium_grain_clock_gating(struct amdgpu_device *adev
 	if (enable && (adev->cg_flags & AMD_CG_SUPPORT_GFX_MGCG)) {
 		/* 1 - RLC_CGTT_MGCG_OVERRIDE */
 		def = data = RREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE);
-		data &= ~(RLC_CGTT_MGCG_OVERRIDE__CPF_CGTT_SCLK_OVERRIDE_MASK |
-			  RLC_CGTT_MGCG_OVERRIDE__GRBM_CGTT_SCLK_OVERRIDE_MASK |
+
+		if (adev->asic_type != CHIP_VEGA12)
+			data &= ~RLC_CGTT_MGCG_OVERRIDE__CPF_CGTT_SCLK_OVERRIDE_MASK;
+
+		data &= ~(RLC_CGTT_MGCG_OVERRIDE__GRBM_CGTT_SCLK_OVERRIDE_MASK |
 			  RLC_CGTT_MGCG_OVERRIDE__GFXIP_MGCG_OVERRIDE_MASK |
 			  RLC_CGTT_MGCG_OVERRIDE__GFXIP_MGLS_OVERRIDE_MASK);
 
@@ -3413,11 +3601,15 @@ static void gfx_v9_0_update_medium_grain_clock_gating(struct amdgpu_device *adev
 	} else {
 		/* 1 - MGCG_OVERRIDE */
 		def = data = RREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE);
-		data |= (RLC_CGTT_MGCG_OVERRIDE__CPF_CGTT_SCLK_OVERRIDE_MASK |
-			 RLC_CGTT_MGCG_OVERRIDE__RLC_CGTT_SCLK_OVERRIDE_MASK |
+
+		if (adev->asic_type != CHIP_VEGA12)
+			data |= RLC_CGTT_MGCG_OVERRIDE__CPF_CGTT_SCLK_OVERRIDE_MASK;
+
+		data |= (RLC_CGTT_MGCG_OVERRIDE__RLC_CGTT_SCLK_OVERRIDE_MASK |
 			 RLC_CGTT_MGCG_OVERRIDE__GRBM_CGTT_SCLK_OVERRIDE_MASK |
 			 RLC_CGTT_MGCG_OVERRIDE__GFXIP_MGCG_OVERRIDE_MASK |
 			 RLC_CGTT_MGCG_OVERRIDE__GFXIP_MGLS_OVERRIDE_MASK);
+
 		if (def != data)
 			WREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE, data);
 
@@ -3453,9 +3645,11 @@ static void gfx_v9_0_update_3d_clock_gating(struct amdgpu_device *adev,
 		/* update CGCG and CGLS override bits */
 		if (def != data)
 			WREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE, data);
-		/* enable 3Dcgcg FSM(0x0020003f) */
+
+		/* enable 3Dcgcg FSM(0x0000363f) */
 		def = RREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL_3D);
-		data = (0x2000 << RLC_CGCG_CGLS_CTRL_3D__CGCG_GFX_IDLE_THRESHOLD__SHIFT) |
+
+		data = (0x36 << RLC_CGCG_CGLS_CTRL_3D__CGCG_GFX_IDLE_THRESHOLD__SHIFT) |
 			RLC_CGCG_CGLS_CTRL_3D__CGCG_EN_MASK;
 		if (adev->cg_flags & AMD_CG_SUPPORT_GFX_3D_CGLS)
 			data |= (0x000F << RLC_CGCG_CGLS_CTRL_3D__CGLS_REP_COMPANSAT_DELAY__SHIFT) |
@@ -3502,9 +3696,10 @@ static void gfx_v9_0_update_coarse_grain_clock_gating(struct amdgpu_device *adev
 		if (def != data)
 			WREG32_SOC15(GC, 0, mmRLC_CGTT_MGCG_OVERRIDE, data);
 
-		/* enable cgcg FSM(0x0020003F) */
+		/* enable cgcg FSM(0x0000363F) */
 		def = RREG32_SOC15(GC, 0, mmRLC_CGCG_CGLS_CTRL);
-		data = (0x2000 << RLC_CGCG_CGLS_CTRL__CGCG_GFX_IDLE_THRESHOLD__SHIFT) |
+
+		data = (0x36 << RLC_CGCG_CGLS_CTRL__CGCG_GFX_IDLE_THRESHOLD__SHIFT) |
 			RLC_CGCG_CGLS_CTRL__CGCG_EN_MASK;
 		if (adev->cg_flags & AMD_CG_SUPPORT_GFX_CGLS)
 			data |= (0x000F << RLC_CGCG_CGLS_CTRL__CGLS_REP_COMPANSAT_DELAY__SHIFT) |
@@ -3586,6 +3781,15 @@ static int gfx_v9_0_set_powergating_state(void *handle,
 
 		/* update mgcg state */
 		gfx_v9_0_update_gfx_mg_power_gating(adev, enable);
+
+		/* set gfx off through smu */
+		if (enable && adev->powerplay.pp_funcs->set_powergating_by_smu)
+			amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_GFX, true);
+		break;
+	case CHIP_VEGA12:
+		/* set gfx off through smu */
+		if (enable && adev->powerplay.pp_funcs->set_powergating_by_smu)
+			amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_GFX, true);
 		break;
 	default:
 		break;
@@ -3605,6 +3809,7 @@ static int gfx_v9_0_set_clockgating_state(void *handle,
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 	case CHIP_RAVEN:
 		gfx_v9_0_update_gfx_clock_gating(adev,
 						 state == AMD_CG_STATE_GATE ? true : false);
@@ -3742,7 +3947,7 @@ static void gfx_v9_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
 	}
 
 	amdgpu_ring_write(ring, header);
-BUG_ON(ib->gpu_addr & 0x3); /* Dword align */
+	BUG_ON(ib->gpu_addr & 0x3); /* Dword align */
 	amdgpu_ring_write(ring,
 #ifdef __BIG_ENDIAN
 		(2 << 0) |
@@ -3774,13 +3979,16 @@ static void gfx_v9_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr,
 {
 	bool write64bit = flags & AMDGPU_FENCE_FLAG_64BIT;
 	bool int_sel = flags & AMDGPU_FENCE_FLAG_INT;
+	bool writeback = flags & AMDGPU_FENCE_FLAG_TC_WB_ONLY;
 
 	/* RELEASE_MEM - flush caches, send int */
 	amdgpu_ring_write(ring, PACKET3(PACKET3_RELEASE_MEM, 6));
-	amdgpu_ring_write(ring, (EOP_TCL1_ACTION_EN |
-				 EOP_TC_ACTION_EN |
-				 EOP_TC_WB_ACTION_EN |
-				 EOP_TC_MD_ACTION_EN |
+	amdgpu_ring_write(ring, ((writeback ? (EOP_TC_WB_ACTION_EN |
+					       EOP_TC_NC_ACTION_EN) :
+					      (EOP_TCL1_ACTION_EN |
+					       EOP_TC_ACTION_EN |
+					       EOP_TC_WB_ACTION_EN |
+					       EOP_TC_MD_ACTION_EN)) |
 				 EVENT_TYPE(CACHE_FLUSH_AND_INV_TS_EVENT) |
 				 EVENT_INDEX(5)));
 	amdgpu_ring_write(ring, DATA_SEL(write64bit ? 2 : 1) | INT_SEL(int_sel ? 2 : 0));
@@ -4137,6 +4345,20 @@ static void gfx_v9_0_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
 	gfx_v9_0_wait_reg_mem(ring, 0, 0, 0, reg, 0, val, mask, 0x20);
 }
 
+static void gfx_v9_0_ring_emit_reg_write_reg_wait(struct amdgpu_ring *ring,
+						  uint32_t reg0, uint32_t reg1,
+						  uint32_t ref, uint32_t mask)
+{
+	int usepfp = (ring->funcs->type == AMDGPU_RING_TYPE_GFX);
+
+	if (amdgpu_sriov_vf(ring->adev))
+		gfx_v9_0_wait_reg_mem(ring, usepfp, 0, 1, reg0, reg1,
+				      ref, mask, 0x20);
+	else
+		amdgpu_ring_emit_reg_write_reg_wait_helper(ring, reg0, reg1,
+							   ref, mask);
+}
+
 static void gfx_v9_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev,
 						 enum amdgpu_interrupt_state state)
 {
@@ -4458,6 +4680,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_gfx = {
 	.emit_tmz = gfx_v9_0_ring_emit_tmz,
 	.emit_wreg = gfx_v9_0_ring_emit_wreg,
 	.emit_reg_wait = gfx_v9_0_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = gfx_v9_0_ring_emit_reg_write_reg_wait,
 };
 
 static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_compute = {
@@ -4492,6 +4715,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_compute = {
 	.set_priority = gfx_v9_0_ring_set_priority_compute,
 	.emit_wreg = gfx_v9_0_ring_emit_wreg,
 	.emit_reg_wait = gfx_v9_0_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = gfx_v9_0_ring_emit_reg_write_reg_wait,
 };
 
 static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_kiq = {
@@ -4522,6 +4746,7 @@ static const struct amdgpu_ring_funcs gfx_v9_0_ring_funcs_kiq = {
 	.emit_rreg = gfx_v9_0_ring_emit_rreg,
 	.emit_wreg = gfx_v9_0_ring_emit_wreg,
 	.emit_reg_wait = gfx_v9_0_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = gfx_v9_0_ring_emit_reg_write_reg_wait,
 };
 
 static void gfx_v9_0_set_ring_funcs(struct amdgpu_device *adev)
@@ -4577,6 +4802,7 @@ static void gfx_v9_0_set_rlc_funcs(struct amdgpu_device *adev)
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 	case CHIP_RAVEN:
 		adev->gfx.rlc.funcs = &gfx_v9_0_rlc_funcs;
 		break;
@@ -4686,6 +4912,7 @@ static int gfx_v9_0_get_cu_info(struct amdgpu_device *adev,
 
 	cu_info->number = active_cu_number;
 	cu_info->ao_cu_mask = ao_cu_mask;
+	cu_info->simd_per_cu = NUM_SIMD_PER_CU;
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
index 5617cf62c566..75317f283c69 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c
@@ -41,11 +41,11 @@ static void gmc_v6_0_set_gmc_funcs(struct amdgpu_device *adev);
 static void gmc_v6_0_set_irq_funcs(struct amdgpu_device *adev);
 static int gmc_v6_0_wait_for_idle(void *handle);
 
-MODULE_FIRMWARE("radeon/tahiti_mc.bin");
-MODULE_FIRMWARE("radeon/pitcairn_mc.bin");
-MODULE_FIRMWARE("radeon/verde_mc.bin");
-MODULE_FIRMWARE("radeon/oland_mc.bin");
-MODULE_FIRMWARE("radeon/si58_mc.bin");
+MODULE_FIRMWARE("amdgpu/tahiti_mc.bin");
+MODULE_FIRMWARE("amdgpu/pitcairn_mc.bin");
+MODULE_FIRMWARE("amdgpu/verde_mc.bin");
+MODULE_FIRMWARE("amdgpu/oland_mc.bin");
+MODULE_FIRMWARE("amdgpu/si58_mc.bin");
 
 #define MC_SEQ_MISC0__MT__MASK   0xf0000000
 #define MC_SEQ_MISC0__MT__GDDR1  0x10000000
@@ -134,9 +134,9 @@ static int gmc_v6_0_init_microcode(struct amdgpu_device *adev)
 		is_58_fw = true;
 
 	if (is_58_fw)
-		snprintf(fw_name, sizeof(fw_name), "radeon/si58_mc.bin");
+		snprintf(fw_name, sizeof(fw_name), "amdgpu/si58_mc.bin");
 	else
-		snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", chip_name);
+		snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mc.bin", chip_name);
 	err = request_firmware(&adev->gmc.fw, fw_name, adev->dev);
 	if (err)
 		goto out;
@@ -819,12 +819,33 @@ static int gmc_v6_0_late_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	amdgpu_bo_late_init(adev);
+
 	if (amdgpu_vm_fault_stop != AMDGPU_VM_FAULT_STOP_ALWAYS)
 		return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0);
 	else
 		return 0;
 }
 
+static unsigned gmc_v6_0_get_vbios_fb_size(struct amdgpu_device *adev)
+{
+	u32 d1vga_control = RREG32(mmD1VGA_CONTROL);
+	unsigned size;
+
+	if (REG_GET_FIELD(d1vga_control, D1VGA_CONTROL, D1VGA_MODE_ENABLE)) {
+		size = 9 * 1024 * 1024; /* reserve 8MB for vga emulator and 1 MB for FB */
+	} else {
+		u32 viewport = RREG32(mmVIEWPORT_SIZE);
+		size = (REG_GET_FIELD(viewport, VIEWPORT_SIZE, VIEWPORT_HEIGHT) *
+			REG_GET_FIELD(viewport, VIEWPORT_SIZE, VIEWPORT_WIDTH) *
+			4);
+	}
+	/* return 0 if the pre-OS buffer uses up most of vram */
+	if ((adev->gmc.real_vram_size - size) < (8 * 1024 * 1024))
+		return 0;
+	return size;
+}
+
 static int gmc_v6_0_sw_init(void *handle)
 {
 	int r;
@@ -851,8 +872,6 @@ static int gmc_v6_0_sw_init(void *handle)
 
 	adev->gmc.mc_mask = 0xffffffffffULL;
 
-	adev->gmc.stolen_size = 256 * 1024;
-
 	adev->need_dma32 = false;
 	dma_bits = adev->need_dma32 ? 32 : 40;
 	r = pci_set_dma_mask(adev->pdev, DMA_BIT_MASK(dma_bits));
@@ -878,6 +897,8 @@ static int gmc_v6_0_sw_init(void *handle)
 	if (r)
 		return r;
 
+	adev->gmc.stolen_size = gmc_v6_0_get_vbios_fb_size(adev);
+
 	r = amdgpu_bo_init(adev);
 	if (r)
 		return r;
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
index 80054f36e487..10920f0bd85f 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v7_0.c
@@ -43,12 +43,14 @@
 
 #include "amdgpu_atombios.h"
 
+#include "ivsrcid/ivsrcid_vislands30.h"
+
 static void gmc_v7_0_set_gmc_funcs(struct amdgpu_device *adev);
 static void gmc_v7_0_set_irq_funcs(struct amdgpu_device *adev);
 static int gmc_v7_0_wait_for_idle(void *handle);
 
-MODULE_FIRMWARE("radeon/bonaire_mc.bin");
-MODULE_FIRMWARE("radeon/hawaii_mc.bin");
+MODULE_FIRMWARE("amdgpu/bonaire_mc.bin");
+MODULE_FIRMWARE("amdgpu/hawaii_mc.bin");
 MODULE_FIRMWARE("amdgpu/topaz_mc.bin");
 
 static const u32 golden_settings_iceland_a11[] =
@@ -147,10 +149,7 @@ static int gmc_v7_0_init_microcode(struct amdgpu_device *adev)
 	default: BUG();
 	}
 
-	if (adev->asic_type == CHIP_TOPAZ)
-		snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mc.bin", chip_name);
-	else
-		snprintf(fw_name, sizeof(fw_name), "radeon/%s_mc.bin", chip_name);
+	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_mc.bin", chip_name);
 
 	err = request_firmware(&adev->gmc.fw, fw_name, adev->dev);
 	if (err)
@@ -958,12 +957,33 @@ static int gmc_v7_0_late_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	amdgpu_bo_late_init(adev);
+
 	if (amdgpu_vm_fault_stop != AMDGPU_VM_FAULT_STOP_ALWAYS)
 		return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0);
 	else
 		return 0;
 }
 
+static unsigned gmc_v7_0_get_vbios_fb_size(struct amdgpu_device *adev)
+{
+	u32 d1vga_control = RREG32(mmD1VGA_CONTROL);
+	unsigned size;
+
+	if (REG_GET_FIELD(d1vga_control, D1VGA_CONTROL, D1VGA_MODE_ENABLE)) {
+		size = 9 * 1024 * 1024; /* reserve 8MB for vga emulator and 1 MB for FB */
+	} else {
+		u32 viewport = RREG32(mmVIEWPORT_SIZE);
+		size = (REG_GET_FIELD(viewport, VIEWPORT_SIZE, VIEWPORT_HEIGHT) *
+			REG_GET_FIELD(viewport, VIEWPORT_SIZE, VIEWPORT_WIDTH) *
+			4);
+	}
+	/* return 0 if the pre-OS buffer uses up most of vram */
+	if ((adev->gmc.real_vram_size - size) < (8 * 1024 * 1024))
+		return 0;
+	return size;
+}
+
 static int gmc_v7_0_sw_init(void *handle)
 {
 	int r;
@@ -978,11 +998,11 @@ static int gmc_v7_0_sw_init(void *handle)
 		adev->gmc.vram_type = gmc_v7_0_convert_vram_type(tmp);
 	}
 
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 146, &adev->gmc.vm_fault);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_GFX_PAGE_INV_FAULT, &adev->gmc.vm_fault);
 	if (r)
 		return r;
 
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 147, &adev->gmc.vm_fault);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_GFX_MEM_PROT_FAULT, &adev->gmc.vm_fault);
 	if (r)
 		return r;
 
@@ -998,8 +1018,6 @@ static int gmc_v7_0_sw_init(void *handle)
 	 */
 	adev->gmc.mc_mask = 0xffffffffffULL; /* 40 bit MC */
 
-	adev->gmc.stolen_size = 256 * 1024;
-
 	/* set DMA mask + need_dma32 flags.
 	 * PCIE - can handle 40-bits.
 	 * IGP - can handle 40-bits
@@ -1030,6 +1048,8 @@ static int gmc_v7_0_sw_init(void *handle)
 	if (r)
 		return r;
 
+	adev->gmc.stolen_size = gmc_v7_0_get_vbios_fb_size(adev);
+
 	/* Memory manager */
 	r = amdgpu_bo_init(adev);
 	if (r)
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
index d71d4cb68f9c..75f3ffb2891e 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v8_0.c
@@ -44,6 +44,7 @@
 
 #include "amdgpu_atombios.h"
 
+#include "ivsrcid/ivsrcid_vislands30.h"
 
 static void gmc_v8_0_set_gmc_funcs(struct amdgpu_device *adev);
 static void gmc_v8_0_set_irq_funcs(struct amdgpu_device *adev);
@@ -138,6 +139,7 @@ static void gmc_v8_0_init_golden_registers(struct amdgpu_device *adev)
 		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 		amdgpu_device_program_register_sequence(adev,
 							golden_settings_polaris11_a11,
 							ARRAY_SIZE(golden_settings_polaris11_a11));
@@ -231,6 +233,7 @@ static int gmc_v8_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_FIJI:
 	case CHIP_CARRIZO:
 	case CHIP_STONEY:
+	case CHIP_VEGAM:
 		return 0;
 	default: BUG();
 	}
@@ -567,9 +570,10 @@ static int gmc_v8_0_mc_init(struct amdgpu_device *adev)
 	/* set the gart size */
 	if (amdgpu_gart_size == -1) {
 		switch (adev->asic_type) {
-		case CHIP_POLARIS11: /* all engines support GPUVM */
 		case CHIP_POLARIS10: /* all engines support GPUVM */
+		case CHIP_POLARIS11: /* all engines support GPUVM */
 		case CHIP_POLARIS12: /* all engines support GPUVM */
+		case CHIP_VEGAM:     /* all engines support GPUVM */
 		default:
 			adev->gmc.gart_size = 256ULL << 20;
 			break;
@@ -1049,12 +1053,33 @@ static int gmc_v8_0_late_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
+	amdgpu_bo_late_init(adev);
+
 	if (amdgpu_vm_fault_stop != AMDGPU_VM_FAULT_STOP_ALWAYS)
 		return amdgpu_irq_get(adev, &adev->gmc.vm_fault, 0);
 	else
 		return 0;
 }
 
+static unsigned gmc_v8_0_get_vbios_fb_size(struct amdgpu_device *adev)
+{
+	u32 d1vga_control = RREG32(mmD1VGA_CONTROL);
+	unsigned size;
+
+	if (REG_GET_FIELD(d1vga_control, D1VGA_CONTROL, D1VGA_MODE_ENABLE)) {
+		size = 9 * 1024 * 1024; /* reserve 8MB for vga emulator and 1 MB for FB */
+	} else {
+		u32 viewport = RREG32(mmVIEWPORT_SIZE);
+		size = (REG_GET_FIELD(viewport, VIEWPORT_SIZE, VIEWPORT_HEIGHT) *
+			REG_GET_FIELD(viewport, VIEWPORT_SIZE, VIEWPORT_WIDTH) *
+			4);
+	}
+	/* return 0 if the pre-OS buffer uses up most of vram */
+	if ((adev->gmc.real_vram_size - size) < (8 * 1024 * 1024))
+		return 0;
+	return size;
+}
+
 #define mmMC_SEQ_MISC0_FIJI 0xA71
 
 static int gmc_v8_0_sw_init(void *handle)
@@ -1068,7 +1093,8 @@ static int gmc_v8_0_sw_init(void *handle)
 	} else {
 		u32 tmp;
 
-		if (adev->asic_type == CHIP_FIJI)
+		if ((adev->asic_type == CHIP_FIJI) ||
+		    (adev->asic_type == CHIP_VEGAM))
 			tmp = RREG32(mmMC_SEQ_MISC0_FIJI);
 		else
 			tmp = RREG32(mmMC_SEQ_MISC0);
@@ -1076,11 +1102,11 @@ static int gmc_v8_0_sw_init(void *handle)
 		adev->gmc.vram_type = gmc_v8_0_convert_vram_type(tmp);
 	}
 
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 146, &adev->gmc.vm_fault);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_GFX_PAGE_INV_FAULT, &adev->gmc.vm_fault);
 	if (r)
 		return r;
 
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 147, &adev->gmc.vm_fault);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_GFX_MEM_PROT_FAULT, &adev->gmc.vm_fault);
 	if (r)
 		return r;
 
@@ -1096,8 +1122,6 @@ static int gmc_v8_0_sw_init(void *handle)
 	 */
 	adev->gmc.mc_mask = 0xffffffffffULL; /* 40 bit MC */
 
-	adev->gmc.stolen_size = 256 * 1024;
-
 	/* set DMA mask + need_dma32 flags.
 	 * PCIE - can handle 40-bits.
 	 * IGP - can handle 40-bits
@@ -1128,6 +1152,8 @@ static int gmc_v8_0_sw_init(void *handle)
 	if (r)
 		return r;
 
+	adev->gmc.stolen_size = gmc_v8_0_get_vbios_fb_size(adev);
+
 	/* Memory manager */
 	r = amdgpu_bo_init(adev);
 	if (r)
@@ -1422,8 +1448,13 @@ static int gmc_v8_0_process_interrupt(struct amdgpu_device *adev,
 		gmc_v8_0_set_fault_enable_default(adev, false);
 
 	if (printk_ratelimit()) {
-		dev_err(adev->dev, "GPU fault detected: %d 0x%08x\n",
-			entry->src_id, entry->src_data[0]);
+		struct amdgpu_task_info task_info = { 0 };
+
+		amdgpu_vm_get_task_info(adev, entry->pasid, &task_info);
+
+		dev_err(adev->dev, "GPU fault detected: %d 0x%08x for process %s pid %d thread %s pid %d\n",
+			entry->src_id, entry->src_data[0], task_info.process_name,
+			task_info.tgid, task_info.task_name, task_info.pid);
 		dev_err(adev->dev, "  VM_CONTEXT1_PROTECTION_FAULT_ADDR   0x%08X\n",
 			addr);
 		dev_err(adev->dev, "  VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x%08X\n",
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index e687363900bb..9df94b45d17d 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -43,19 +43,15 @@
 #include "gfxhub_v1_0.h"
 #include "mmhub_v1_0.h"
 
-#define mmDF_CS_AON0_DramBaseAddress0                                                                  0x0044
-#define mmDF_CS_AON0_DramBaseAddress0_BASE_IDX                                                         0
-//DF_CS_AON0_DramBaseAddress0
-#define DF_CS_AON0_DramBaseAddress0__AddrRngVal__SHIFT                                                        0x0
-#define DF_CS_AON0_DramBaseAddress0__LgcyMmioHoleEn__SHIFT                                                    0x1
-#define DF_CS_AON0_DramBaseAddress0__IntLvNumChan__SHIFT                                                      0x4
-#define DF_CS_AON0_DramBaseAddress0__IntLvAddrSel__SHIFT                                                      0x8
-#define DF_CS_AON0_DramBaseAddress0__DramBaseAddr__SHIFT                                                      0xc
-#define DF_CS_AON0_DramBaseAddress0__AddrRngVal_MASK                                                          0x00000001L
-#define DF_CS_AON0_DramBaseAddress0__LgcyMmioHoleEn_MASK                                                      0x00000002L
-#define DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK                                                        0x000000F0L
-#define DF_CS_AON0_DramBaseAddress0__IntLvAddrSel_MASK                                                        0x00000700L
-#define DF_CS_AON0_DramBaseAddress0__DramBaseAddr_MASK                                                        0xFFFFF000L
+#include "ivsrcid/vmc/irqsrcs_vmc_1_0.h"
+
+/* add these here since we already include dce12 headers and these are for DCN */
+#define mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION                                                          0x055d
+#define mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION_BASE_IDX                                                 2
+#define HUBP0_DCSURF_PRI_VIEWPORT_DIMENSION__PRI_VIEWPORT_WIDTH__SHIFT                                        0x0
+#define HUBP0_DCSURF_PRI_VIEWPORT_DIMENSION__PRI_VIEWPORT_HEIGHT__SHIFT                                       0x10
+#define HUBP0_DCSURF_PRI_VIEWPORT_DIMENSION__PRI_VIEWPORT_WIDTH_MASK                                          0x00003FFFL
+#define HUBP0_DCSURF_PRI_VIEWPORT_DIMENSION__PRI_VIEWPORT_HEIGHT_MASK                                         0x3FFF0000L
 
 /* XXX Move this macro to VEGA10 header file, which is like vid.h for VI.*/
 #define AMDGPU_NUM_OF_VMIDS			8
@@ -263,11 +259,16 @@ static int gmc_v9_0_process_interrupt(struct amdgpu_device *adev,
 	}
 
 	if (printk_ratelimit()) {
+		struct amdgpu_task_info task_info = { 0 };
+
+		amdgpu_vm_get_task_info(adev, entry->pasid, &task_info);
+
 		dev_err(adev->dev,
-			"[%s] VMC page fault (src_id:%u ring:%u vmid:%u pasid:%u)\n",
+			"[%s] VMC page fault (src_id:%u ring:%u vmid:%u pasid:%u, for process %s pid %d thread %s pid %d\n)\n",
 			entry->vmid_src ? "mmhub" : "gfxhub",
 			entry->src_id, entry->ring_id, entry->vmid,
-			entry->pasid);
+			entry->pasid, task_info.process_name, task_info.tgid,
+			task_info.task_name, task_info.pid);
 		dev_err(adev->dev, "  at page 0x%016llx from %d\n",
 			addr, entry->client_id);
 		if (!amdgpu_sriov_vf(adev))
@@ -385,11 +386,9 @@ static uint64_t gmc_v9_0_emit_flush_gpu_tlb(struct amdgpu_ring *ring,
 	amdgpu_ring_emit_wreg(ring, hub->ctx0_ptb_addr_hi32 + (2 * vmid),
 			      upper_32_bits(pd_addr));
 
-	amdgpu_ring_emit_wreg(ring, hub->vm_inv_eng0_req + eng, req);
-
-	/* wait for the invalidate to complete */
-	amdgpu_ring_emit_reg_wait(ring, hub->vm_inv_eng0_ack + eng,
-				  1 << vmid, 1 << vmid);
+	amdgpu_ring_emit_reg_write_reg_wait(ring, hub->vm_inv_eng0_req + eng,
+					    hub->vm_inv_eng0_ack + eng,
+					    req, 1 << vmid);
 
 	return pd_addr;
 }
@@ -556,8 +555,7 @@ static int gmc_v9_0_early_init(void *handle)
 	adev->gmc.shared_aperture_start = 0x2000000000000000ULL;
 	adev->gmc.shared_aperture_end =
 		adev->gmc.shared_aperture_start + (4ULL << 30) - 1;
-	adev->gmc.private_aperture_start =
-		adev->gmc.shared_aperture_end + 1;
+	adev->gmc.private_aperture_start = 0x1000000000000000ULL;
 	adev->gmc.private_aperture_end =
 		adev->gmc.private_aperture_start + (4ULL << 30) - 1;
 
@@ -659,6 +657,11 @@ static int gmc_v9_0_late_init(void *handle)
 	unsigned i;
 	int r;
 
+	/*
+	 * TODO - Uncomment once GART corruption issue is fixed.
+	 */
+	/* amdgpu_bo_late_init(adev); */
+
 	for(i = 0; i < adev->num_rings; ++i) {
 		struct amdgpu_ring *ring = adev->rings[i];
 		unsigned vmhub = ring->funcs->vmhub;
@@ -679,6 +682,7 @@ static int gmc_v9_0_late_init(void *handle)
 			DRM_INFO("ECC is active.\n");
 		} else if (r == 0) {
 			DRM_INFO("ECC is not present.\n");
+			adev->df_funcs->enable_ecc_force_par_wr_rmw(adev, false);
 		} else {
 			DRM_ERROR("gmc_v9_0_ecc_available() failed. r: %d\n", r);
 			return r;
@@ -697,10 +701,7 @@ static void gmc_v9_0_vram_gtt_location(struct amdgpu_device *adev,
 	amdgpu_device_vram_location(adev, &adev->gmc, base);
 	amdgpu_device_gart_location(adev, mc);
 	/* base offset of vram pages */
-	if (adev->flags & AMD_IS_APU)
-		adev->vm_manager.vram_base_offset = gfxhub_v1_0_get_mc_fb_offset(adev);
-	else
-		adev->vm_manager.vram_base_offset = 0;
+	adev->vm_manager.vram_base_offset = gfxhub_v1_0_get_mc_fb_offset(adev);
 }
 
 /**
@@ -714,7 +715,6 @@ static void gmc_v9_0_vram_gtt_location(struct amdgpu_device *adev,
  */
 static int gmc_v9_0_mc_init(struct amdgpu_device *adev)
 {
-	u32 tmp;
 	int chansize, numchan;
 	int r;
 
@@ -727,39 +727,7 @@ static int gmc_v9_0_mc_init(struct amdgpu_device *adev)
 		else
 			chansize = 128;
 
-		tmp = RREG32_SOC15(DF, 0, mmDF_CS_AON0_DramBaseAddress0);
-		tmp &= DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK;
-		tmp >>= DF_CS_AON0_DramBaseAddress0__IntLvNumChan__SHIFT;
-		switch (tmp) {
-		case 0:
-		default:
-			numchan = 1;
-			break;
-		case 1:
-			numchan = 2;
-			break;
-		case 2:
-			numchan = 0;
-			break;
-		case 3:
-			numchan = 4;
-			break;
-		case 4:
-			numchan = 0;
-			break;
-		case 5:
-			numchan = 8;
-			break;
-		case 6:
-			numchan = 0;
-			break;
-		case 7:
-			numchan = 16;
-			break;
-		case 8:
-			numchan = 2;
-			break;
-		}
+		numchan = adev->df_funcs->get_hbm_channel_number(adev);
 		adev->gmc.vram_width = numchan * chansize;
 	}
 
@@ -792,6 +760,7 @@ static int gmc_v9_0_mc_init(struct amdgpu_device *adev)
 		switch (adev->asic_type) {
 		case CHIP_VEGA10:  /* all engines support GPUVM */
 		case CHIP_VEGA12:  /* all engines support GPUVM */
+		case CHIP_VEGA20:
 		default:
 			adev->gmc.gart_size = 512ULL << 20;
 			break;
@@ -826,6 +795,52 @@ static int gmc_v9_0_gart_init(struct amdgpu_device *adev)
 	return amdgpu_gart_table_vram_alloc(adev);
 }
 
+static unsigned gmc_v9_0_get_vbios_fb_size(struct amdgpu_device *adev)
+{
+#if 0
+	u32 d1vga_control = RREG32_SOC15(DCE, 0, mmD1VGA_CONTROL);
+#endif
+	unsigned size;
+
+	/*
+	 * TODO Remove once GART corruption is resolved
+	 * Check related code in gmc_v9_0_sw_fini
+	 * */
+	size = 9 * 1024 * 1024;
+
+#if 0
+	if (REG_GET_FIELD(d1vga_control, D1VGA_CONTROL, D1VGA_MODE_ENABLE)) {
+		size = 9 * 1024 * 1024; /* reserve 8MB for vga emulator and 1 MB for FB */
+	} else {
+		u32 viewport;
+
+		switch (adev->asic_type) {
+		case CHIP_RAVEN:
+			viewport = RREG32_SOC15(DCE, 0, mmHUBP0_DCSURF_PRI_VIEWPORT_DIMENSION);
+			size = (REG_GET_FIELD(viewport,
+					      HUBP0_DCSURF_PRI_VIEWPORT_DIMENSION, PRI_VIEWPORT_HEIGHT) *
+				REG_GET_FIELD(viewport,
+					      HUBP0_DCSURF_PRI_VIEWPORT_DIMENSION, PRI_VIEWPORT_WIDTH) *
+				4);
+			break;
+		case CHIP_VEGA10:
+		case CHIP_VEGA12:
+		default:
+			viewport = RREG32_SOC15(DCE, 0, mmSCL0_VIEWPORT_SIZE);
+			size = (REG_GET_FIELD(viewport, SCL0_VIEWPORT_SIZE, VIEWPORT_HEIGHT) *
+				REG_GET_FIELD(viewport, SCL0_VIEWPORT_SIZE, VIEWPORT_WIDTH) *
+				4);
+			break;
+		}
+	}
+	/* return 0 if the pre-OS buffer uses up most of vram */
+	if ((adev->gmc.real_vram_size - size) < (8 * 1024 * 1024))
+		return 0;
+
+#endif
+	return size;
+}
+
 static int gmc_v9_0_sw_init(void *handle)
 {
 	int r;
@@ -851,6 +866,7 @@ static int gmc_v9_0_sw_init(void *handle)
 		break;
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		/*
 		 * To fulfill 4-level page support,
 		 * vm size is 256TB (48bit), maximum size of Vega10,
@@ -863,9 +879,9 @@ static int gmc_v9_0_sw_init(void *handle)
 	}
 
 	/* This interrupt is VMC page fault.*/
-	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VMC, 0,
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VMC, VMC_1_0__SRCID__VM_FAULT,
 				&adev->gmc.vm_fault);
-	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UTCL2, 0,
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UTCL2, UTCL2_1_0__SRCID__FAULT,
 				&adev->gmc.vm_fault);
 
 	if (r)
@@ -877,12 +893,6 @@ static int gmc_v9_0_sw_init(void *handle)
 	 */
 	adev->gmc.mc_mask = 0xffffffffffffULL; /* 48 bit MC */
 
-	/*
-	 * It needs to reserve 8M stolen memory for vega10
-	 * TODO: Figure out how to avoid that...
-	 */
-	adev->gmc.stolen_size = 8 * 1024 * 1024;
-
 	/* set DMA mask + need_dma32 flags.
 	 * PCIE - can handle 44-bits.
 	 * IGP - can handle 44-bits
@@ -907,6 +917,8 @@ static int gmc_v9_0_sw_init(void *handle)
 	if (r)
 		return r;
 
+	adev->gmc.stolen_size = gmc_v9_0_get_vbios_fb_size(adev);
+
 	/* Memory manager */
 	r = amdgpu_bo_init(adev);
 	if (r)
@@ -950,6 +962,18 @@ static int gmc_v9_0_sw_fini(void *handle)
 	amdgpu_gem_force_release(adev);
 	amdgpu_vm_manager_fini(adev);
 	gmc_v9_0_gart_fini(adev);
+
+	/*
+	* TODO:
+	* Currently there is a bug where some memory client outside
+	* of the driver writes to first 8M of VRAM on S3 resume,
+	* this overrides GART which by default gets placed in first 8M and
+	* causes VM_FAULTS once GTT is accessed.
+	* Keep the stolen memory reservation until the while this is not solved.
+	* Also check code in gmc_v9_0_get_vbios_fb_size and gmc_v9_0_late_init
+	*/
+	amdgpu_bo_free_kernel(&adev->stolen_vga_memory, NULL, NULL);
+
 	amdgpu_bo_fini(adev);
 
 	return 0;
@@ -960,6 +984,7 @@ static void gmc_v9_0_init_golden_registers(struct amdgpu_device *adev)
 
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
+	case CHIP_VEGA20:
 		soc15_program_register_sequence(adev,
 						golden_settings_mmhub_1_0_0,
 						ARRAY_SIZE(golden_settings_mmhub_1_0_0));
diff --git a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
index 26ba984ab2b7..3f57f6463dc8 100644
--- a/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/kv_dpm.c
@@ -1921,7 +1921,7 @@ static int kv_dpm_set_power_state(void *handle)
 	int ret;
 
 	if (pi->bapm_enable) {
-		ret = amdgpu_kv_smc_bapm_enable(adev, adev->pm.dpm.ac_power);
+		ret = amdgpu_kv_smc_bapm_enable(adev, adev->pm.ac_power);
 		if (ret) {
 			DRM_ERROR("amdgpu_kv_smc_bapm_enable failed\n");
 			return ret;
@@ -2727,8 +2727,9 @@ static int kv_parse_power_table(struct amdgpu_device *adev)
 		(mode_info->atom_context->bios + data_offset +
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 
-	adev->pm.dpm.ps = kzalloc(sizeof(struct amdgpu_ps) *
-				  state_array->ucNumEntries, GFP_KERNEL);
+	adev->pm.dpm.ps = kcalloc(state_array->ucNumEntries,
+				  sizeof(struct amdgpu_ps),
+				  GFP_KERNEL);
 	if (!adev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
@@ -2817,7 +2818,7 @@ static int kv_dpm_init(struct amdgpu_device *adev)
 		pi->caps_tcp_ramping = true;
 	}
 
-	if (amdgpu_pp_feature_mask & SCLK_DEEP_SLEEP_MASK)
+	if (adev->powerplay.pp_feature & PP_SCLK_DEEP_SLEEP_MASK)
 		pi->caps_sclk_ds = true;
 	else
 		pi->caps_sclk_ds = false;
@@ -2974,7 +2975,7 @@ static int kv_dpm_late_init(void *handle)
 	/* powerdown unused blocks for now */
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!amdgpu_dpm)
+	if (!adev->pm.dpm_enabled)
 		return 0;
 
 	kv_dpm_powergate_acp(adev, true);
@@ -3305,6 +3306,19 @@ static int kv_dpm_read_sensor(void *handle, int idx,
 	}
 }
 
+static int kv_set_powergating_by_smu(void *handle,
+				uint32_t block_type, bool gate)
+{
+	switch (block_type) {
+	case AMD_IP_BLOCK_TYPE_UVD:
+		kv_dpm_powergate_uvd(handle, gate);
+		break;
+	default:
+		break;
+	}
+	return 0;
+}
+
 static const struct amd_ip_funcs kv_dpm_ip_funcs = {
 	.name = "kv_dpm",
 	.early_init = kv_dpm_early_init,
@@ -3341,7 +3355,7 @@ static const struct amd_pm_funcs kv_dpm_funcs = {
 	.print_power_state = &kv_dpm_print_power_state,
 	.debugfs_print_current_performance_level = &kv_dpm_debugfs_print_current_performance_level,
 	.force_performance_level = &kv_dpm_force_performance_level,
-	.powergate_uvd = &kv_dpm_powergate_uvd,
+	.set_powergating_by_smu = kv_set_powergating_by_smu,
 	.enable_bapm = &kv_dpm_enable_bapm,
 	.get_vce_clock_state = amdgpu_get_vce_clock_state,
 	.check_state_equal = kv_check_state_equal,
diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c
index 43f925773b57..e70a0d4d6db4 100644
--- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c
@@ -471,8 +471,8 @@ void mmhub_v1_0_update_power_gating(struct amdgpu_device *adev,
 						RENG_EXECUTE_ON_REG_UPDATE, 1);
 		WREG32_SOC15(MMHUB, 0, mmPCTL1_RENG_EXECUTE, pctl1_reng_execute);
 
-		if (adev->powerplay.pp_funcs->set_mmhub_powergating_by_smu)
-			amdgpu_dpm_set_mmhub_powergating_by_smu(adev);
+		if (adev->powerplay.pp_funcs->set_powergating_by_smu)
+			amdgpu_dpm_set_powergating_by_smu(adev, AMD_IP_BLOCK_TYPE_GMC, true);
 
 	} else {
 		pctl0_reng_execute = REG_SET_FIELD(pctl0_reng_execute,
@@ -734,6 +734,7 @@ int mmhub_v1_0_set_clockgating(struct amdgpu_device *adev,
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 	case CHIP_RAVEN:
 		mmhub_v1_0_update_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
diff --git a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
index 493348672475..078f70faedcb 100644
--- a/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
+++ b/drivers/gpu/drm/amd/amdgpu/mxgpu_ai.c
@@ -260,8 +260,10 @@ static void xgpu_ai_mailbox_flr_work(struct work_struct *work)
 	} while (timeout > 1);
 
 flr_done:
-	if (locked)
+	if (locked) {
+		adev->in_gpu_reset = 0;
 		mutex_unlock(&adev->lock_reset);
+	}
 
 	/* Trigger recovery for world switch failure if no TDR */
 	if (amdgpu_lockup_timeout == 0)
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c
index df34dc79d444..365517c0121e 100644
--- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c
@@ -34,10 +34,19 @@
 #define smnCPM_CONTROL                                                                                  0x11180460
 #define smnPCIE_CNTL2                                                                                   0x11180070
 
+/* vega20 */
+#define mmRCC_DEV0_EPF0_STRAP0_VG20                                                                         0x0011
+#define mmRCC_DEV0_EPF0_STRAP0_VG20_BASE_IDX                                                                2
+
 static u32 nbio_v7_0_get_rev_id(struct amdgpu_device *adev)
 {
         u32 tmp = RREG32_SOC15(NBIO, 0, mmRCC_DEV0_EPF0_STRAP0);
 
+	if (adev->asic_type == CHIP_VEGA20)
+		tmp = RREG32_SOC15(NBIO, 0, mmRCC_DEV0_EPF0_STRAP0_VG20);
+	else
+		tmp = RREG32_SOC15(NBIO, 0, mmRCC_DEV0_EPF0_STRAP0);
+
 	tmp &= RCC_DEV0_EPF0_STRAP0__STRAP_ATI_REV_ID_DEV0_F0_MASK;
 	tmp >>= RCC_DEV0_EPF0_STRAP0__STRAP_ATI_REV_ID_DEV0_F0__SHIFT;
 
@@ -75,10 +84,14 @@ static void nbio_v7_0_sdma_doorbell_range(struct amdgpu_device *adev, int instan
 			SOC15_REG_OFFSET(NBIO, 0, mmBIF_SDMA1_DOORBELL_RANGE);
 
 	u32 doorbell_range = RREG32(reg);
+	u32 range = 2;
+
+	if (adev->asic_type == CHIP_VEGA20)
+		range = 8;
 
 	if (use_doorbell) {
 		doorbell_range = REG_SET_FIELD(doorbell_range, BIF_SDMA0_DOORBELL_RANGE, OFFSET, doorbell_index);
-		doorbell_range = REG_SET_FIELD(doorbell_range, BIF_SDMA0_DOORBELL_RANGE, SIZE, 2);
+		doorbell_range = REG_SET_FIELD(doorbell_range, BIF_SDMA0_DOORBELL_RANGE, SIZE, range);
 	} else
 		doorbell_range = REG_SET_FIELD(doorbell_range, BIF_SDMA0_DOORBELL_RANGE, SIZE, 0);
 
@@ -133,6 +146,9 @@ static void nbio_v7_0_update_medium_grain_clock_gating(struct amdgpu_device *ade
 {
 	uint32_t def, data;
 
+	if (adev->asic_type == CHIP_VEGA20)
+		return;
+
 	/* NBIF_MGCG_CTRL_LCLK */
 	def = data = RREG32_PCIE(smnNBIF_MGCG_CTRL_LCLK);
 
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h b/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h
index 8da6da90b1c9..0cf48d26c676 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h
+++ b/drivers/gpu/drm/amd/amdgpu/psp_gfx_if.h
@@ -40,11 +40,20 @@ enum psp_gfx_crtl_cmd_id
     GFX_CTRL_CMD_ID_INIT_GPCOM_RING = 0x00020000,   /* initialize GPCOM ring */
     GFX_CTRL_CMD_ID_DESTROY_RINGS   = 0x00030000,   /* destroy rings */
     GFX_CTRL_CMD_ID_CAN_INIT_RINGS  = 0x00040000,   /* is it allowed to initialized the rings */
+    GFX_CTRL_CMD_ID_ENABLE_INT      = 0x00050000,   /* enable PSP-to-Gfx interrupt */
+    GFX_CTRL_CMD_ID_DISABLE_INT     = 0x00060000,   /* disable PSP-to-Gfx interrupt */
+    GFX_CTRL_CMD_ID_MODE1_RST       = 0x00070000,   /* trigger the Mode 1 reset */
 
     GFX_CTRL_CMD_ID_MAX             = 0x000F0000,   /* max command ID */
 };
 
 
+/*-----------------------------------------------------------------------------
+    NOTE:   All physical addresses used in this interface are actually
+            GPU Virtual Addresses.
+*/
+
+
 /* Control registers of the TEE Gfx interface. These are located in
 *  SRBM-to-PSP mailbox registers (total 8 registers).
 */
@@ -55,8 +64,8 @@ struct psp_gfx_ctrl
     volatile uint32_t   rbi_rptr;         /* +8   Read pointer (index) of RBI ring */
     volatile uint32_t   gpcom_wptr;       /* +12  Write pointer (index) of GPCOM ring */
     volatile uint32_t   gpcom_rptr;       /* +16  Read pointer (index) of GPCOM ring */
-    volatile uint32_t   ring_addr_lo;     /* +20  bits [31:0] of physical address of ring buffer */
-    volatile uint32_t   ring_addr_hi;     /* +24  bits [63:32] of physical address of ring buffer */
+    volatile uint32_t   ring_addr_lo;     /* +20  bits [31:0] of GPU Virtual of ring buffer (VMID=0)*/
+    volatile uint32_t   ring_addr_hi;     /* +24  bits [63:32] of GPU Virtual of ring buffer (VMID=0) */
     volatile uint32_t   ring_buf_size;    /* +28  Ring buffer size (in bytes) */
 
 };
@@ -78,6 +87,8 @@ enum psp_gfx_cmd_id
     GFX_CMD_ID_LOAD_ASD     = 0x00000004,   /* load ASD Driver */
     GFX_CMD_ID_SETUP_TMR    = 0x00000005,   /* setup TMR region */
     GFX_CMD_ID_LOAD_IP_FW   = 0x00000006,   /* load HW IP FW */
+    GFX_CMD_ID_DESTROY_TMR  = 0x00000007,   /* destroy TMR region */
+    GFX_CMD_ID_SAVE_RESTORE = 0x00000008,   /* save/restore HW IP FW */
 
 };
 
@@ -85,11 +96,11 @@ enum psp_gfx_cmd_id
 /* Command to load Trusted Application binary into PSP OS. */
 struct psp_gfx_cmd_load_ta
 {
-    uint32_t        app_phy_addr_lo;        /* bits [31:0] of the physical address of the TA binary (must be 4 KB aligned) */
-    uint32_t        app_phy_addr_hi;        /* bits [63:32] of the physical address of the TA binary */
+    uint32_t        app_phy_addr_lo;        /* bits [31:0] of the GPU Virtual address of the TA binary (must be 4 KB aligned) */
+    uint32_t        app_phy_addr_hi;        /* bits [63:32] of the GPU Virtual address of the TA binary */
     uint32_t        app_len;                /* length of the TA binary in bytes */
-    uint32_t        cmd_buf_phy_addr_lo;    /* bits [31:0] of the physical address of CMD buffer (must be 4 KB aligned) */
-    uint32_t        cmd_buf_phy_addr_hi;    /* bits [63:32] of the physical address of CMD buffer */
+    uint32_t        cmd_buf_phy_addr_lo;    /* bits [31:0] of the GPU Virtual address of CMD buffer (must be 4 KB aligned) */
+    uint32_t        cmd_buf_phy_addr_hi;    /* bits [63:32] of the GPU Virtual address of CMD buffer */
     uint32_t        cmd_buf_len;            /* length of the CMD buffer in bytes; must be multiple of 4 KB */
 
     /* Note: CmdBufLen can be set to 0. In this case no persistent CMD buffer is provided
@@ -111,8 +122,8 @@ struct psp_gfx_cmd_unload_ta
 */
 struct psp_gfx_buf_desc
 {
-    uint32_t        buf_phy_addr_lo;       /* bits [31:0] of physical address of the buffer (must be 4 KB aligned) */
-    uint32_t        buf_phy_addr_hi;       /* bits [63:32] of physical address of the buffer */
+    uint32_t        buf_phy_addr_lo;       /* bits [31:0] of GPU Virtual address of the buffer (must be 4 KB aligned) */
+    uint32_t        buf_phy_addr_hi;       /* bits [63:32] of GPU Virtual address of the buffer */
     uint32_t        buf_size;              /* buffer size in bytes (must be multiple of 4 KB and no bigger than 64 MB) */
 
 };
@@ -145,8 +156,8 @@ struct psp_gfx_cmd_invoke_cmd
 /* Command to setup TMR region. */
 struct psp_gfx_cmd_setup_tmr
 {
-    uint32_t        buf_phy_addr_lo;       /* bits [31:0] of physical address of TMR buffer (must be 4 KB aligned) */
-    uint32_t        buf_phy_addr_hi;       /* bits [63:32] of physical address of TMR buffer */
+    uint32_t        buf_phy_addr_lo;       /* bits [31:0] of GPU Virtual address of TMR buffer (must be 4 KB aligned) */
+    uint32_t        buf_phy_addr_hi;       /* bits [63:32] of GPU Virtual address of TMR buffer */
     uint32_t        buf_size;              /* buffer size in bytes (must be multiple of 4 KB) */
 
 };
@@ -174,18 +185,32 @@ enum psp_gfx_fw_type
     GFX_FW_TYPE_ISP         = 16,
     GFX_FW_TYPE_ACP         = 17,
     GFX_FW_TYPE_SMU         = 18,
+    GFX_FW_TYPE_MMSCH       = 19,
+    GFX_FW_TYPE_RLC_RESTORE_LIST_GPM_MEM        = 20,
+    GFX_FW_TYPE_RLC_RESTORE_LIST_SRM_MEM        = 21,
+    GFX_FW_TYPE_RLC_RESTORE_LIST_CNTL           = 22,
+    GFX_FW_TYPE_MAX         = 23
 };
 
 /* Command to load HW IP FW. */
 struct psp_gfx_cmd_load_ip_fw
 {
-    uint32_t                fw_phy_addr_lo;    /* bits [31:0] of physical address of FW location (must be 4 KB aligned) */
-    uint32_t                fw_phy_addr_hi;    /* bits [63:32] of physical address of FW location */
+    uint32_t                fw_phy_addr_lo;    /* bits [31:0] of GPU Virtual address of FW location (must be 4 KB aligned) */
+    uint32_t                fw_phy_addr_hi;    /* bits [63:32] of GPU Virtual address of FW location */
     uint32_t                fw_size;           /* FW buffer size in bytes */
     enum psp_gfx_fw_type    fw_type;           /* FW type */
 
 };
 
+/* Command to save/restore HW IP FW. */
+struct psp_gfx_cmd_save_restore_ip_fw
+{
+    uint32_t                save_fw;              /* if set, command is used for saving fw otherwise for resetoring*/
+    uint32_t                save_restore_addr_lo; /* bits [31:0] of FB address of GART memory used as save/restore buffer (must be 4 KB aligned) */
+    uint32_t                save_restore_addr_hi; /* bits [63:32] of FB address of GART memory used as save/restore buffer */
+    uint32_t                buf_size;             /* Size of the save/restore buffer in bytes */
+    enum psp_gfx_fw_type    fw_type;              /* FW type */
+};
 
 /* All GFX ring buffer commands. */
 union psp_gfx_commands
@@ -195,7 +220,7 @@ union psp_gfx_commands
     struct psp_gfx_cmd_invoke_cmd       cmd_invoke_cmd;
     struct psp_gfx_cmd_setup_tmr        cmd_setup_tmr;
     struct psp_gfx_cmd_load_ip_fw       cmd_load_ip_fw;
-
+    struct psp_gfx_cmd_save_restore_ip_fw cmd_save_restore_ip_fw;
 };
 
 
@@ -226,8 +251,8 @@ struct psp_gfx_cmd_resp
 
     /* These fields are used for RBI only. They are all 0 in GPCOM commands
     */
-    uint32_t        resp_buf_addr_lo;   /* +12 bits [31:0] of physical address of response buffer (must be 4 KB aligned) */
-    uint32_t        resp_buf_addr_hi;   /* +16 bits [63:32] of physical address of response buffer */
+    uint32_t        resp_buf_addr_lo;   /* +12 bits [31:0] of GPU Virtual address of response buffer (must be 4 KB aligned) */
+    uint32_t        resp_buf_addr_hi;   /* +16 bits [63:32] of GPU Virtual address of response buffer */
     uint32_t        resp_offset;        /* +20 offset within response buffer */
     uint32_t        resp_buf_size;      /* +24 total size of the response buffer in bytes */
 
@@ -251,19 +276,19 @@ struct psp_gfx_cmd_resp
 /* Structure of the Ring Buffer Frame */
 struct psp_gfx_rb_frame
 {
-    uint32_t    cmd_buf_addr_lo;    /* +0  bits [31:0] of physical address of command buffer (must be 4 KB aligned) */
-    uint32_t    cmd_buf_addr_hi;    /* +4  bits [63:32] of physical address of command buffer */
+    uint32_t    cmd_buf_addr_lo;    /* +0  bits [31:0] of GPU Virtual address of command buffer (must be 4 KB aligned) */
+    uint32_t    cmd_buf_addr_hi;    /* +4  bits [63:32] of GPU Virtual address of command buffer */
     uint32_t    cmd_buf_size;       /* +8  command buffer size in bytes */
-    uint32_t    fence_addr_lo;      /* +12 bits [31:0] of physical address of Fence for this frame */
-    uint32_t    fence_addr_hi;      /* +16 bits [63:32] of physical address of Fence for this frame */
+    uint32_t    fence_addr_lo;      /* +12 bits [31:0] of GPU Virtual address of Fence for this frame */
+    uint32_t    fence_addr_hi;      /* +16 bits [63:32] of GPU Virtual address of Fence for this frame */
     uint32_t    fence_value;        /* +20 Fence value */
     uint32_t    sid_lo;             /* +24 bits [31:0] of SID value (used only for RBI frames) */
     uint32_t    sid_hi;             /* +28 bits [63:32] of SID value (used only for RBI frames) */
     uint8_t     vmid;               /* +32 VMID value used for mapping of all addresses for this frame */
     uint8_t     frame_type;         /* +33 1: destory context frame, 0: all other frames; used only for RBI frames */
     uint8_t     reserved1[2];       /* +34 reserved, must be 0 */
-    uint32_t    reserved2[7];       /* +40 reserved, must be 0 */
-    /* total 64 bytes */
+    uint32_t    reserved2[7];       /* +36 reserved, must be 0 */
+                /* total 64 bytes */
 };
 
 #endif /* _PSP_TEE_GFX_IF_H_ */
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
index 8873d833a7f7..0ff136d02d9b 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c
@@ -70,6 +70,15 @@ psp_v10_0_get_fw_type(struct amdgpu_firmware_info *ucode, enum psp_gfx_fw_type *
 	case AMDGPU_UCODE_ID_RLC_G:
 		*type = GFX_FW_TYPE_RLC_G;
 		break;
+	case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL:
+		*type = GFX_FW_TYPE_RLC_RESTORE_LIST_CNTL;
+		break;
+	case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM:
+		*type = GFX_FW_TYPE_RLC_RESTORE_LIST_GPM_MEM;
+		break;
+	case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM:
+		*type = GFX_FW_TYPE_RLC_RESTORE_LIST_SRM_MEM;
+		break;
 	case AMDGPU_UCODE_ID_SMC:
 		*type = GFX_FW_TYPE_SMU;
 		break;
diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
index 196e75def1f2..727071fee6f6 100644
--- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
+++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c
@@ -41,9 +41,14 @@ MODULE_FIRMWARE("amdgpu/vega10_sos.bin");
 MODULE_FIRMWARE("amdgpu/vega10_asd.bin");
 MODULE_FIRMWARE("amdgpu/vega12_sos.bin");
 MODULE_FIRMWARE("amdgpu/vega12_asd.bin");
+MODULE_FIRMWARE("amdgpu/vega20_sos.bin");
+MODULE_FIRMWARE("amdgpu/vega20_asd.bin");
+
 
 #define smnMP1_FIRMWARE_FLAGS 0x3010028
 
+static uint32_t sos_old_versions[] = {1517616, 1510592, 1448594, 1446554};
+
 static int
 psp_v3_1_get_fw_type(struct amdgpu_firmware_info *ucode, enum psp_gfx_fw_type *type)
 {
@@ -207,12 +212,31 @@ static int psp_v3_1_bootloader_load_sysdrv(struct psp_context *psp)
 	return ret;
 }
 
+static bool psp_v3_1_match_version(struct amdgpu_device *adev, uint32_t ver)
+{
+	int i;
+
+	if (ver == adev->psp.sos_fw_version)
+		return true;
+
+	/*
+	 * Double check if the latest four legacy versions.
+	 * If yes, it is still the right version.
+	 */
+	for (i = 0; i < sizeof(sos_old_versions) / sizeof(uint32_t); i++) {
+		if (sos_old_versions[i] == adev->psp.sos_fw_version)
+			return true;
+	}
+
+	return false;
+}
+
 static int psp_v3_1_bootloader_load_sos(struct psp_context *psp)
 {
 	int ret;
 	unsigned int psp_gfxdrv_command_reg = 0;
 	struct amdgpu_device *adev = psp->adev;
-	uint32_t sol_reg;
+	uint32_t sol_reg, ver;
 
 	/* Check sOS sign of life register to confirm sys driver and sOS
 	 * are already been loaded.
@@ -245,6 +269,10 @@ static int psp_v3_1_bootloader_load_sos(struct psp_context *psp)
 			   RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_81),
 			   0, true);
 
+	ver = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_58);
+	if (!psp_v3_1_match_version(adev, ver))
+		DRM_WARN("SOS version doesn't match\n");
+
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c
index c7190c39c4f5..15ae4bc9c072 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v2_4.c
@@ -44,6 +44,8 @@
 
 #include "iceland_sdma_pkt_open.h"
 
+#include "ivsrcid/ivsrcid_vislands30.h"
+
 static void sdma_v2_4_set_ring_funcs(struct amdgpu_device *adev);
 static void sdma_v2_4_set_buffer_funcs(struct amdgpu_device *adev);
 static void sdma_v2_4_set_vm_pte_funcs(struct amdgpu_device *adev);
@@ -202,8 +204,7 @@ static uint64_t sdma_v2_4_ring_get_rptr(struct amdgpu_ring *ring)
 static uint64_t sdma_v2_4_ring_get_wptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
-	int me = (ring == &ring->adev->sdma.instance[0].ring) ? 0 : 1;
-	u32 wptr = RREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[me]) >> 2;
+	u32 wptr = RREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[ring->me]) >> 2;
 
 	return wptr;
 }
@@ -218,9 +219,8 @@ static uint64_t sdma_v2_4_ring_get_wptr(struct amdgpu_ring *ring)
 static void sdma_v2_4_ring_set_wptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
-	int me = (ring == &ring->adev->sdma.instance[0].ring) ? 0 : 1;
 
-	WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[me], lower_32_bits(ring->wptr) << 2);
+	WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[ring->me], lower_32_bits(ring->wptr) << 2);
 }
 
 static void sdma_v2_4_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
@@ -273,7 +273,7 @@ static void sdma_v2_4_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 {
 	u32 ref_and_mask = 0;
 
-	if (ring == &ring->adev->sdma.instance[0].ring)
+	if (ring->me == 0)
 		ref_and_mask = REG_SET_FIELD(ref_and_mask, GPU_HDP_FLUSH_DONE, SDMA0, 1);
 	else
 		ref_and_mask = REG_SET_FIELD(ref_and_mask, GPU_HDP_FLUSH_DONE, SDMA1, 1);
@@ -898,7 +898,7 @@ static int sdma_v2_4_sw_init(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	/* SDMA trap event */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 224,
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_SDMA_TRAP,
 			      &adev->sdma.trap_irq);
 	if (r)
 		return r;
@@ -910,7 +910,7 @@ static int sdma_v2_4_sw_init(void *handle)
 		return r;
 
 	/* SDMA Privileged inst */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 247,
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_SDMA_SRBM_WRITE,
 			      &adev->sdma.illegal_inst_irq);
 	if (r)
 		return r;
@@ -1213,8 +1213,10 @@ static void sdma_v2_4_set_ring_funcs(struct amdgpu_device *adev)
 {
 	int i;
 
-	for (i = 0; i < adev->sdma.num_instances; i++)
+	for (i = 0; i < adev->sdma.num_instances; i++) {
 		adev->sdma.instance[i].ring.funcs = &sdma_v2_4_ring_funcs;
+		adev->sdma.instance[i].ring.me = i;
+	}
 }
 
 static const struct amdgpu_irq_src_funcs sdma_v2_4_trap_irq_funcs = {
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
index be20a387d961..1e07ff274d73 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v3_0.c
@@ -44,6 +44,8 @@
 
 #include "tonga_sdma_pkt_open.h"
 
+#include "ivsrcid/ivsrcid_vislands30.h"
+
 static void sdma_v3_0_set_ring_funcs(struct amdgpu_device *adev);
 static void sdma_v3_0_set_buffer_funcs(struct amdgpu_device *adev);
 static void sdma_v3_0_set_vm_pte_funcs(struct amdgpu_device *adev);
@@ -62,6 +64,8 @@ MODULE_FIRMWARE("amdgpu/polaris11_sdma.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_sdma1.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_sdma.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_sdma1.bin");
+MODULE_FIRMWARE("amdgpu/vegam_sdma.bin");
+MODULE_FIRMWARE("amdgpu/vegam_sdma1.bin");
 
 
 static const u32 sdma_offsets[SDMA_MAX_INSTANCE] =
@@ -209,6 +213,7 @@ static void sdma_v3_0_init_golden_registers(struct amdgpu_device *adev)
 		break;
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 		amdgpu_device_program_register_sequence(adev,
 							golden_settings_polaris11_a11,
 							ARRAY_SIZE(golden_settings_polaris11_a11));
@@ -275,15 +280,18 @@ static int sdma_v3_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_FIJI:
 		chip_name = "fiji";
 		break;
-	case CHIP_POLARIS11:
-		chip_name = "polaris11";
-		break;
 	case CHIP_POLARIS10:
 		chip_name = "polaris10";
 		break;
+	case CHIP_POLARIS11:
+		chip_name = "polaris11";
+		break;
 	case CHIP_POLARIS12:
 		chip_name = "polaris12";
 		break;
+	case CHIP_VEGAM:
+		chip_name = "vegam";
+		break;
 	case CHIP_CARRIZO:
 		chip_name = "carrizo";
 		break;
@@ -359,9 +367,7 @@ static uint64_t sdma_v3_0_ring_get_wptr(struct amdgpu_ring *ring)
 		/* XXX check if swapping is necessary on BE */
 		wptr = ring->adev->wb.wb[ring->wptr_offs] >> 2;
 	} else {
-		int me = (ring == &ring->adev->sdma.instance[0].ring) ? 0 : 1;
-
-		wptr = RREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[me]) >> 2;
+		wptr = RREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[ring->me]) >> 2;
 	}
 
 	return wptr;
@@ -388,9 +394,7 @@ static void sdma_v3_0_ring_set_wptr(struct amdgpu_ring *ring)
 
 		WRITE_ONCE(*wb, (lower_32_bits(ring->wptr) << 2));
 	} else {
-		int me = (ring == &ring->adev->sdma.instance[0].ring) ? 0 : 1;
-
-		WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[me], lower_32_bits(ring->wptr) << 2);
+		WREG32(mmSDMA0_GFX_RB_WPTR + sdma_offsets[ring->me], lower_32_bits(ring->wptr) << 2);
 	}
 }
 
@@ -444,7 +448,7 @@ static void sdma_v3_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 {
 	u32 ref_and_mask = 0;
 
-	if (ring == &ring->adev->sdma.instance[0].ring)
+	if (ring->me == 0)
 		ref_and_mask = REG_SET_FIELD(ref_and_mask, GPU_HDP_FLUSH_DONE, SDMA0, 1);
 	else
 		ref_and_mask = REG_SET_FIELD(ref_and_mask, GPU_HDP_FLUSH_DONE, SDMA1, 1);
@@ -1173,7 +1177,7 @@ static int sdma_v3_0_sw_init(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	/* SDMA trap event */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 224,
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_SDMA_TRAP,
 			      &adev->sdma.trap_irq);
 	if (r)
 		return r;
@@ -1185,7 +1189,7 @@ static int sdma_v3_0_sw_init(void *handle)
 		return r;
 
 	/* SDMA Privileged inst */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 247,
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_SDMA_SRBM_WRITE,
 			      &adev->sdma.illegal_inst_irq);
 	if (r)
 		return r;
@@ -1649,8 +1653,10 @@ static void sdma_v3_0_set_ring_funcs(struct amdgpu_device *adev)
 {
 	int i;
 
-	for (i = 0; i < adev->sdma.num_instances; i++)
+	for (i = 0; i < adev->sdma.num_instances; i++) {
 		adev->sdma.instance[i].ring.funcs = &sdma_v3_0_ring_funcs;
+		adev->sdma.instance[i].ring.me = i;
+	}
 }
 
 static const struct amdgpu_irq_src_funcs sdma_v3_0_trap_irq_funcs = {
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
index 399f876f9cad..e7ca4623cfb9 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
@@ -38,10 +38,15 @@
 #include "soc15.h"
 #include "vega10_sdma_pkt_open.h"
 
+#include "ivsrcid/sdma0/irqsrcs_sdma0_4_0.h"
+#include "ivsrcid/sdma1/irqsrcs_sdma1_4_0.h"
+
 MODULE_FIRMWARE("amdgpu/vega10_sdma.bin");
 MODULE_FIRMWARE("amdgpu/vega10_sdma1.bin");
 MODULE_FIRMWARE("amdgpu/vega12_sdma.bin");
 MODULE_FIRMWARE("amdgpu/vega12_sdma1.bin");
+MODULE_FIRMWARE("amdgpu/vega20_sdma.bin");
+MODULE_FIRMWARE("amdgpu/vega20_sdma1.bin");
 MODULE_FIRMWARE("amdgpu/raven_sdma.bin");
 
 #define SDMA0_POWER_CNTL__ON_OFF_CONDITION_HOLD_TIME_MASK  0x000000F8L
@@ -107,6 +112,28 @@ static const struct soc15_reg_golden golden_settings_sdma_4_1[] =
 	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_UTCL1_PAGE, 0x000003ff, 0x000003c0)
 };
 
+static const struct soc15_reg_golden golden_settings_sdma_4_2[] =
+{
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_CHICKEN_BITS, 0xfe931f07, 0x02831d07),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_CLK_CTRL, 0xffffffff, 0x3f000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG, 0x0000773f, 0x00004002),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG_READ, 0x0000773f, 0x00004002),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GFX_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_PAGE_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC0_RB_WPTR_POLL_CNTL, 0xfffffff0, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_RLC1_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_UTCL1_PAGE, 0x000003ff, 0x000003c0),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_CHICKEN_BITS, 0xfe931f07, 0x02831d07),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_CLK_CTRL, 0xffffffff, 0x3f000100),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_GB_ADDR_CONFIG, 0x0000773f, 0x00004002),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_GB_ADDR_CONFIG_READ, 0x0000773f, 0x00004002),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_GFX_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_PAGE_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_RLC0_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_RLC1_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000),
+	SOC15_REG_GOLDEN_VALUE(SDMA1, 0, mmSDMA1_UTCL1_PAGE, 0x000003ff, 0x000003c0)
+};
+
 static const struct soc15_reg_golden golden_settings_sdma_rv1[] =
 {
 	SOC15_REG_GOLDEN_VALUE(SDMA0, 0, mmSDMA0_GB_ADDR_CONFIG, 0x0018773f, 0x00000002),
@@ -139,6 +166,11 @@ static void sdma_v4_0_init_golden_registers(struct amdgpu_device *adev)
 						golden_settings_sdma_vg12,
 						ARRAY_SIZE(golden_settings_sdma_vg12));
 		break;
+	case CHIP_VEGA20:
+		soc15_program_register_sequence(adev,
+						golden_settings_sdma_4_2,
+						ARRAY_SIZE(golden_settings_sdma_4_2));
+		break;
 	case CHIP_RAVEN:
 		soc15_program_register_sequence(adev,
 						 golden_settings_sdma_4_1,
@@ -182,6 +214,9 @@ static int sdma_v4_0_init_microcode(struct amdgpu_device *adev)
 	case CHIP_VEGA12:
 		chip_name = "vega12";
 		break;
+	case CHIP_VEGA20:
+		chip_name = "vega20";
+		break;
 	case CHIP_RAVEN:
 		chip_name = "raven";
 		break;
@@ -264,13 +299,12 @@ static uint64_t sdma_v4_0_ring_get_wptr(struct amdgpu_ring *ring)
 		DRM_DEBUG("wptr/doorbell before shift == 0x%016llx\n", wptr);
 	} else {
 		u32 lowbit, highbit;
-		int me = (ring == &adev->sdma.instance[0].ring) ? 0 : 1;
 
-		lowbit = RREG32(sdma_v4_0_get_reg_offset(adev, me, mmSDMA0_GFX_RB_WPTR)) >> 2;
-		highbit = RREG32(sdma_v4_0_get_reg_offset(adev, me, mmSDMA0_GFX_RB_WPTR_HI)) >> 2;
+		lowbit = RREG32(sdma_v4_0_get_reg_offset(adev, ring->me, mmSDMA0_GFX_RB_WPTR)) >> 2;
+		highbit = RREG32(sdma_v4_0_get_reg_offset(adev, ring->me, mmSDMA0_GFX_RB_WPTR_HI)) >> 2;
 
 		DRM_DEBUG("wptr [%i]high== 0x%08x low==0x%08x\n",
-				me, highbit, lowbit);
+				ring->me, highbit, lowbit);
 		wptr = highbit;
 		wptr = wptr << 32;
 		wptr |= lowbit;
@@ -307,17 +341,15 @@ static void sdma_v4_0_ring_set_wptr(struct amdgpu_ring *ring)
 				ring->doorbell_index, ring->wptr << 2);
 		WDOORBELL64(ring->doorbell_index, ring->wptr << 2);
 	} else {
-		int me = (ring == &ring->adev->sdma.instance[0].ring) ? 0 : 1;
-
 		DRM_DEBUG("Not using doorbell -- "
 				"mmSDMA%i_GFX_RB_WPTR == 0x%08x "
 				"mmSDMA%i_GFX_RB_WPTR_HI == 0x%08x\n",
-				me,
+				ring->me,
 				lower_32_bits(ring->wptr << 2),
-				me,
+				ring->me,
 				upper_32_bits(ring->wptr << 2));
-		WREG32(sdma_v4_0_get_reg_offset(adev, me, mmSDMA0_GFX_RB_WPTR), lower_32_bits(ring->wptr << 2));
-		WREG32(sdma_v4_0_get_reg_offset(adev, me, mmSDMA0_GFX_RB_WPTR_HI), upper_32_bits(ring->wptr << 2));
+		WREG32(sdma_v4_0_get_reg_offset(adev, ring->me, mmSDMA0_GFX_RB_WPTR), lower_32_bits(ring->wptr << 2));
+		WREG32(sdma_v4_0_get_reg_offset(adev, ring->me, mmSDMA0_GFX_RB_WPTR_HI), upper_32_bits(ring->wptr << 2));
 	}
 }
 
@@ -360,6 +392,31 @@ static void sdma_v4_0_ring_emit_ib(struct amdgpu_ring *ring,
 
 }
 
+static void sdma_v4_0_wait_reg_mem(struct amdgpu_ring *ring,
+				   int mem_space, int hdp,
+				   uint32_t addr0, uint32_t addr1,
+				   uint32_t ref, uint32_t mask,
+				   uint32_t inv)
+{
+	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) |
+			  SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(hdp) |
+			  SDMA_PKT_POLL_REGMEM_HEADER_MEM_POLL(mem_space) |
+			  SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* == */
+	if (mem_space) {
+		/* memory */
+		amdgpu_ring_write(ring, addr0);
+		amdgpu_ring_write(ring, addr1);
+	} else {
+		/* registers */
+		amdgpu_ring_write(ring, addr0 << 2);
+		amdgpu_ring_write(ring, addr1 << 2);
+	}
+	amdgpu_ring_write(ring, ref); /* reference */
+	amdgpu_ring_write(ring, mask); /* mask */
+	amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) |
+			  SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(inv)); /* retry count, poll interval */
+}
+
 /**
  * sdma_v4_0_ring_emit_hdp_flush - emit an hdp flush on the DMA ring
  *
@@ -373,20 +430,15 @@ static void sdma_v4_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 	u32 ref_and_mask = 0;
 	const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio_funcs->hdp_flush_reg;
 
-	if (ring == &ring->adev->sdma.instance[0].ring)
+	if (ring->me == 0)
 		ref_and_mask = nbio_hf_reg->ref_and_mask_sdma0;
 	else
 		ref_and_mask = nbio_hf_reg->ref_and_mask_sdma1;
 
-	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) |
-			  SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(1) |
-			  SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* == */
-	amdgpu_ring_write(ring, (adev->nbio_funcs->get_hdp_flush_done_offset(adev)) << 2);
-	amdgpu_ring_write(ring, (adev->nbio_funcs->get_hdp_flush_req_offset(adev)) << 2);
-	amdgpu_ring_write(ring, ref_and_mask); /* reference */
-	amdgpu_ring_write(ring, ref_and_mask); /* mask */
-	amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) |
-			  SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10)); /* retry count, poll interval */
+	sdma_v4_0_wait_reg_mem(ring, 0, 1,
+			       adev->nbio_funcs->get_hdp_flush_done_offset(adev),
+			       adev->nbio_funcs->get_hdp_flush_req_offset(adev),
+			       ref_and_mask, ref_and_mask, 10);
 }
 
 /**
@@ -1114,16 +1166,10 @@ static void sdma_v4_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring)
 	uint64_t addr = ring->fence_drv.gpu_addr;
 
 	/* wait for idle */
-	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) |
-			  SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(0) |
-			  SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3) | /* equal */
-			  SDMA_PKT_POLL_REGMEM_HEADER_MEM_POLL(1));
-	amdgpu_ring_write(ring, addr & 0xfffffffc);
-	amdgpu_ring_write(ring, upper_32_bits(addr) & 0xffffffff);
-	amdgpu_ring_write(ring, seq); /* reference */
-	amdgpu_ring_write(ring, 0xffffffff); /* mask */
-	amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) |
-			  SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(4)); /* retry count, poll interval */
+	sdma_v4_0_wait_reg_mem(ring, 1, 0,
+			       addr & 0xfffffffc,
+			       upper_32_bits(addr) & 0xffffffff,
+			       seq, 0xffffffff, 4);
 }
 
 
@@ -1154,15 +1200,7 @@ static void sdma_v4_0_ring_emit_wreg(struct amdgpu_ring *ring,
 static void sdma_v4_0_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
 					 uint32_t val, uint32_t mask)
 {
-	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_POLL_REGMEM) |
-			  SDMA_PKT_POLL_REGMEM_HEADER_HDP_FLUSH(0) |
-			  SDMA_PKT_POLL_REGMEM_HEADER_FUNC(3)); /* equal */
-	amdgpu_ring_write(ring, reg << 2);
-	amdgpu_ring_write(ring, 0);
-	amdgpu_ring_write(ring, val); /* reference */
-	amdgpu_ring_write(ring, mask); /* mask */
-	amdgpu_ring_write(ring, SDMA_PKT_POLL_REGMEM_DW5_RETRY_COUNT(0xfff) |
-			  SDMA_PKT_POLL_REGMEM_DW5_INTERVAL(10));
+	sdma_v4_0_wait_reg_mem(ring, 0, 0, reg, 0, val, mask, 10);
 }
 
 static int sdma_v4_0_early_init(void *handle)
@@ -1190,13 +1228,13 @@ static int sdma_v4_0_sw_init(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	/* SDMA trap event */
-	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_SDMA0, 224,
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_SDMA0, SDMA0_4_0__SRCID__SDMA_TRAP,
 			      &adev->sdma.trap_irq);
 	if (r)
 		return r;
 
 	/* SDMA trap event */
-	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_SDMA1, 224,
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_SDMA1, SDMA1_4_0__SRCID__SDMA_TRAP,
 			      &adev->sdma.trap_irq);
 	if (r)
 		return r;
@@ -1510,6 +1548,7 @@ static int sdma_v4_0_set_clockgating_state(void *handle,
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 	case CHIP_RAVEN:
 		sdma_v4_0_update_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
@@ -1605,14 +1644,17 @@ static const struct amdgpu_ring_funcs sdma_v4_0_ring_funcs = {
 	.pad_ib = sdma_v4_0_ring_pad_ib,
 	.emit_wreg = sdma_v4_0_ring_emit_wreg,
 	.emit_reg_wait = sdma_v4_0_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
 };
 
 static void sdma_v4_0_set_ring_funcs(struct amdgpu_device *adev)
 {
 	int i;
 
-	for (i = 0; i < adev->sdma.num_instances; i++)
+	for (i = 0; i < adev->sdma.num_instances; i++) {
 		adev->sdma.instance[i].ring.funcs = &sdma_v4_0_ring_funcs;
+		adev->sdma.instance[i].ring.me = i;
+	}
 }
 
 static const struct amdgpu_irq_src_funcs sdma_v4_0_trap_irq_funcs = {
diff --git a/drivers/gpu/drm/amd/amdgpu/si.c b/drivers/gpu/drm/amd/amdgpu/si.c
index a675ec6d2811..c364ef94cc36 100644
--- a/drivers/gpu/drm/amd/amdgpu/si.c
+++ b/drivers/gpu/drm/amd/amdgpu/si.c
@@ -1252,6 +1252,12 @@ static void si_invalidate_hdp(struct amdgpu_device *adev,
 	}
 }
 
+static bool si_need_full_reset(struct amdgpu_device *adev)
+{
+	/* change this when we support soft reset */
+	return true;
+}
+
 static int si_get_pcie_lanes(struct amdgpu_device *adev)
 {
 	u32 link_width_cntl;
@@ -1332,6 +1338,7 @@ static const struct amdgpu_asic_funcs si_asic_funcs =
 	.get_config_memsize = &si_get_config_memsize,
 	.flush_hdp = &si_flush_hdp,
 	.invalidate_hdp = &si_invalidate_hdp,
+	.need_full_reset = &si_need_full_reset,
 };
 
 static uint32_t si_get_rev_id(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/si_dpm.c b/drivers/gpu/drm/amd/amdgpu/si_dpm.c
index 797d505bf9ee..db327b412562 100644
--- a/drivers/gpu/drm/amd/amdgpu/si_dpm.c
+++ b/drivers/gpu/drm/amd/amdgpu/si_dpm.c
@@ -56,16 +56,16 @@
 
 #define BIOS_SCRATCH_4                                    0x5cd
 
-MODULE_FIRMWARE("radeon/tahiti_smc.bin");
-MODULE_FIRMWARE("radeon/pitcairn_smc.bin");
-MODULE_FIRMWARE("radeon/pitcairn_k_smc.bin");
-MODULE_FIRMWARE("radeon/verde_smc.bin");
-MODULE_FIRMWARE("radeon/verde_k_smc.bin");
-MODULE_FIRMWARE("radeon/oland_smc.bin");
-MODULE_FIRMWARE("radeon/oland_k_smc.bin");
-MODULE_FIRMWARE("radeon/hainan_smc.bin");
-MODULE_FIRMWARE("radeon/hainan_k_smc.bin");
-MODULE_FIRMWARE("radeon/banks_k_2_smc.bin");
+MODULE_FIRMWARE("amdgpu/tahiti_smc.bin");
+MODULE_FIRMWARE("amdgpu/pitcairn_smc.bin");
+MODULE_FIRMWARE("amdgpu/pitcairn_k_smc.bin");
+MODULE_FIRMWARE("amdgpu/verde_smc.bin");
+MODULE_FIRMWARE("amdgpu/verde_k_smc.bin");
+MODULE_FIRMWARE("amdgpu/oland_smc.bin");
+MODULE_FIRMWARE("amdgpu/oland_k_smc.bin");
+MODULE_FIRMWARE("amdgpu/hainan_smc.bin");
+MODULE_FIRMWARE("amdgpu/hainan_k_smc.bin");
+MODULE_FIRMWARE("amdgpu/banks_k_2_smc.bin");
 
 static const struct amd_pm_funcs si_dpm_funcs;
 
@@ -3480,7 +3480,7 @@ static void si_apply_state_adjust_rules(struct amdgpu_device *adev,
 		disable_sclk_switching = true;
 	}
 
-	if (adev->pm.dpm.ac_power)
+	if (adev->pm.ac_power)
 		max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_ac;
 	else
 		max_limits = &adev->pm.dpm.dyn_state.max_clock_voltage_on_dc;
@@ -3489,7 +3489,7 @@ static void si_apply_state_adjust_rules(struct amdgpu_device *adev,
 		if (ps->performance_levels[i].vddc > ps->performance_levels[i+1].vddc)
 			ps->performance_levels[i].vddc = ps->performance_levels[i+1].vddc;
 	}
-	if (adev->pm.dpm.ac_power == false) {
+	if (adev->pm.ac_power == false) {
 		for (i = 0; i < ps->performance_level_count; i++) {
 			if (ps->performance_levels[i].mclk > max_limits->mclk)
 				ps->performance_levels[i].mclk = max_limits->mclk;
@@ -7242,8 +7242,9 @@ static int si_parse_power_table(struct amdgpu_device *adev)
 		(mode_info->atom_context->bios + data_offset +
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 
-	adev->pm.dpm.ps = kzalloc(sizeof(struct amdgpu_ps) *
-				  state_array->ucNumEntries, GFP_KERNEL);
+	adev->pm.dpm.ps = kcalloc(state_array->ucNumEntries,
+				  sizeof(struct amdgpu_ps),
+				  GFP_KERNEL);
 	if (!adev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
@@ -7317,8 +7318,7 @@ static int si_dpm_init(struct amdgpu_device *adev)
 	pi = &eg_pi->rv7xx;
 
 	si_pi->sys_pcie_mask =
-		(adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_MASK) >>
-		CAIL_PCIE_LINK_SPEED_SUPPORT_SHIFT;
+		adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_MASK;
 	si_pi->force_pcie_gen = AMDGPU_PCIE_GEN_INVALID;
 	si_pi->boot_pcie_gen = si_get_current_pcie_speed(adev);
 
@@ -7346,7 +7346,9 @@ static int si_dpm_init(struct amdgpu_device *adev)
 		return ret;
 
 	adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries =
-		kzalloc(4 * sizeof(struct amdgpu_clock_voltage_dependency_entry), GFP_KERNEL);
+		kcalloc(4,
+			sizeof(struct amdgpu_clock_voltage_dependency_entry),
+			GFP_KERNEL);
 	if (!adev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries) {
 		amdgpu_free_extended_power_table(adev);
 		return -ENOMEM;
@@ -7580,7 +7582,7 @@ static int si_dpm_late_init(void *handle)
 	int ret;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!amdgpu_dpm)
+	if (!adev->pm.dpm_enabled)
 		return 0;
 
 	ret = si_set_temperature_range(adev);
@@ -7664,7 +7666,7 @@ static int si_dpm_init_microcode(struct amdgpu_device *adev)
 	default: BUG();
 	}
 
-	snprintf(fw_name, sizeof(fw_name), "radeon/%s_smc.bin", chip_name);
+	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_smc.bin", chip_name);
 	err = request_firmware(&adev->pm.fw, fw_name, adev->dev);
 	if (err)
 		goto out;
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index 51cf8a30f6c2..83f2717fcf81 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -41,8 +41,6 @@
 #include "sdma1/sdma1_4_0_offset.h"
 #include "hdp/hdp_4_0_offset.h"
 #include "hdp/hdp_4_0_sh_mask.h"
-#include "mp/mp_9_0_offset.h"
-#include "mp/mp_9_0_sh_mask.h"
 #include "smuio/smuio_9_0_offset.h"
 #include "smuio/smuio_9_0_sh_mask.h"
 
@@ -52,6 +50,8 @@
 #include "gmc_v9_0.h"
 #include "gfxhub_v1_0.h"
 #include "mmhub_v1_0.h"
+#include "df_v1_7.h"
+#include "df_v3_6.h"
 #include "vega10_ih.h"
 #include "sdma_v4_0.h"
 #include "uvd_v7_0.h"
@@ -60,33 +60,6 @@
 #include "dce_virtual.h"
 #include "mxgpu_ai.h"
 
-#define mmFabricConfigAccessControl                                                                    0x0410
-#define mmFabricConfigAccessControl_BASE_IDX                                                           0
-#define mmFabricConfigAccessControl_DEFAULT                                      0x00000000
-//FabricConfigAccessControl
-#define FabricConfigAccessControl__CfgRegInstAccEn__SHIFT                                                     0x0
-#define FabricConfigAccessControl__CfgRegInstAccRegLock__SHIFT                                                0x1
-#define FabricConfigAccessControl__CfgRegInstID__SHIFT                                                        0x10
-#define FabricConfigAccessControl__CfgRegInstAccEn_MASK                                                       0x00000001L
-#define FabricConfigAccessControl__CfgRegInstAccRegLock_MASK                                                  0x00000002L
-#define FabricConfigAccessControl__CfgRegInstID_MASK                                                          0x00FF0000L
-
-
-#define mmDF_PIE_AON0_DfGlobalClkGater                                                                 0x00fc
-#define mmDF_PIE_AON0_DfGlobalClkGater_BASE_IDX                                                        0
-//DF_PIE_AON0_DfGlobalClkGater
-#define DF_PIE_AON0_DfGlobalClkGater__MGCGMode__SHIFT                                                         0x0
-#define DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK                                                           0x0000000FL
-
-enum {
-	DF_MGCG_DISABLE = 0,
-	DF_MGCG_ENABLE_00_CYCLE_DELAY =1,
-	DF_MGCG_ENABLE_01_CYCLE_DELAY =2,
-	DF_MGCG_ENABLE_15_CYCLE_DELAY =13,
-	DF_MGCG_ENABLE_31_CYCLE_DELAY =14,
-	DF_MGCG_ENABLE_63_CYCLE_DELAY =15
-};
-
 #define mmMP0_MISC_CGTT_CTRL0                                                                   0x01b9
 #define mmMP0_MISC_CGTT_CTRL0_BASE_IDX                                                          0
 #define mmMP0_MISC_LIGHT_SLEEP_CTRL                                                             0x01ba
@@ -313,6 +286,7 @@ static struct soc15_allowed_register_entry soc15_allowed_read_registers[] = {
 	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPC_STALLED_STAT1)},
 	{ SOC15_REG_ENTRY(GC, 0, mmCP_CPC_STATUS)},
 	{ SOC15_REG_ENTRY(GC, 0, mmGB_ADDR_CONFIG)},
+	{ SOC15_REG_ENTRY(GC, 0, mmDB_DEBUG2)},
 };
 
 static uint32_t soc15_read_indexed_register(struct amdgpu_device *adev, u32 se_num,
@@ -341,6 +315,8 @@ static uint32_t soc15_get_register_value(struct amdgpu_device *adev,
 	} else {
 		if (reg_offset == SOC15_REG_OFFSET(GC, 0, mmGB_ADDR_CONFIG))
 			return adev->gfx.config.gb_addr_config;
+		else if (reg_offset == SOC15_REG_OFFSET(GC, 0, mmDB_DEBUG2))
+			return adev->gfx.config.db_debug2;
 		return RREG32(reg_offset);
 	}
 }
@@ -512,15 +488,24 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 	case CHIP_RAVEN:
 		vega10_reg_base_init(adev);
 		break;
+	case CHIP_VEGA20:
+		vega20_reg_base_init(adev);
+		break;
 	default:
 		return -EINVAL;
 	}
 
 	if (adev->flags & AMD_IS_APU)
 		adev->nbio_funcs = &nbio_v7_0_funcs;
+	else if (adev->asic_type == CHIP_VEGA20)
+		adev->nbio_funcs = &nbio_v7_0_funcs;
 	else
 		adev->nbio_funcs = &nbio_v6_1_funcs;
 
+	if (adev->asic_type == CHIP_VEGA20)
+		adev->df_funcs = &df_v3_6_funcs;
+	else
+		adev->df_funcs = &df_v1_7_funcs;
 	adev->nbio_funcs->detect_hw_virt(adev);
 
 	if (amdgpu_sriov_vf(adev))
@@ -529,12 +514,15 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev)
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		amdgpu_device_ip_block_add(adev, &vega10_common_ip_block);
 		amdgpu_device_ip_block_add(adev, &gmc_v9_0_ip_block);
 		amdgpu_device_ip_block_add(adev, &vega10_ih_ip_block);
-		amdgpu_device_ip_block_add(adev, &psp_v3_1_ip_block);
-		if (!amdgpu_sriov_vf(adev))
-			amdgpu_device_ip_block_add(adev, &pp_smu_ip_block);
+		if (adev->asic_type != CHIP_VEGA20) {
+			amdgpu_device_ip_block_add(adev, &psp_v3_1_ip_block);
+			if (!amdgpu_sriov_vf(adev))
+				amdgpu_device_ip_block_add(adev, &pp_smu_ip_block);
+		}
 		if (adev->enable_virtual_display || amdgpu_sriov_vf(adev))
 			amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block);
 #if defined(CONFIG_DRM_AMD_DC)
@@ -593,6 +581,12 @@ static void soc15_invalidate_hdp(struct amdgpu_device *adev,
 			HDP, 0, mmHDP_READ_CACHE_INVALIDATE), 1);
 }
 
+static bool soc15_need_full_reset(struct amdgpu_device *adev)
+{
+	/* change this when we implement soft reset */
+	return true;
+}
+
 static const struct amdgpu_asic_funcs soc15_asic_funcs =
 {
 	.read_disabled_bios = &soc15_read_disabled_bios,
@@ -606,6 +600,7 @@ static const struct amdgpu_asic_funcs soc15_asic_funcs =
 	.get_config_memsize = &soc15_get_config_memsize,
 	.flush_hdp = &soc15_flush_hdp,
 	.invalidate_hdp = &soc15_invalidate_hdp,
+	.need_full_reset = &soc15_need_full_reset,
 };
 
 static int soc15_common_early_init(void *handle)
@@ -675,6 +670,28 @@ static int soc15_common_early_init(void *handle)
 		adev->pg_flags = 0;
 		adev->external_rev_id = adev->rev_id + 0x14;
 		break;
+	case CHIP_VEGA20:
+		adev->cg_flags = AMD_CG_SUPPORT_GFX_MGCG |
+			AMD_CG_SUPPORT_GFX_MGLS |
+			AMD_CG_SUPPORT_GFX_CGCG |
+			AMD_CG_SUPPORT_GFX_CGLS |
+			AMD_CG_SUPPORT_GFX_3D_CGCG |
+			AMD_CG_SUPPORT_GFX_3D_CGLS |
+			AMD_CG_SUPPORT_GFX_CP_LS |
+			AMD_CG_SUPPORT_MC_LS |
+			AMD_CG_SUPPORT_MC_MGCG |
+			AMD_CG_SUPPORT_SDMA_MGCG |
+			AMD_CG_SUPPORT_SDMA_LS |
+			AMD_CG_SUPPORT_BIF_MGCG |
+			AMD_CG_SUPPORT_BIF_LS |
+			AMD_CG_SUPPORT_HDP_MGCG |
+			AMD_CG_SUPPORT_HDP_LS |
+			AMD_CG_SUPPORT_ROM_MGCG |
+			AMD_CG_SUPPORT_VCE_MGCG |
+			AMD_CG_SUPPORT_UVD_MGCG;
+		adev->pg_flags = 0;
+		adev->external_rev_id = adev->rev_id + 0x28;
+		break;
 	case CHIP_RAVEN:
 		adev->cg_flags = AMD_CG_SUPPORT_GFX_MGCG |
 			AMD_CG_SUPPORT_GFX_MGLS |
@@ -694,8 +711,15 @@ static int soc15_common_early_init(void *handle)
 			AMD_CG_SUPPORT_MC_MGCG |
 			AMD_CG_SUPPORT_MC_LS |
 			AMD_CG_SUPPORT_SDMA_MGCG |
-			AMD_CG_SUPPORT_SDMA_LS;
-		adev->pg_flags = AMD_PG_SUPPORT_SDMA;
+			AMD_CG_SUPPORT_SDMA_LS |
+			AMD_CG_SUPPORT_VCN_MGCG;
+
+		adev->pg_flags = AMD_PG_SUPPORT_SDMA | AMD_PG_SUPPORT_VCN;
+
+		if (adev->powerplay.pp_feature & PP_GFXOFF_MASK)
+			adev->pg_flags |= AMD_PG_SUPPORT_GFX_PG |
+				AMD_PG_SUPPORT_CP |
+				AMD_PG_SUPPORT_RLC_SMU_HS;
 
 		adev->external_rev_id = 0x1;
 		break;
@@ -871,32 +895,6 @@ static void soc15_update_rom_medium_grain_clock_gating(struct amdgpu_device *ade
 		WREG32(SOC15_REG_OFFSET(SMUIO, 0, mmCGTT_ROM_CLK_CTRL0), data);
 }
 
-static void soc15_update_df_medium_grain_clock_gating(struct amdgpu_device *adev,
-						       bool enable)
-{
-	uint32_t data;
-
-	/* Put DF on broadcast mode */
-	data = RREG32(SOC15_REG_OFFSET(DF, 0, mmFabricConfigAccessControl));
-	data &= ~FabricConfigAccessControl__CfgRegInstAccEn_MASK;
-	WREG32(SOC15_REG_OFFSET(DF, 0, mmFabricConfigAccessControl), data);
-
-	if (enable && (adev->cg_flags & AMD_CG_SUPPORT_DF_MGCG)) {
-		data = RREG32(SOC15_REG_OFFSET(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater));
-		data &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK;
-		data |= DF_MGCG_ENABLE_15_CYCLE_DELAY;
-		WREG32(SOC15_REG_OFFSET(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater), data);
-	} else {
-		data = RREG32(SOC15_REG_OFFSET(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater));
-		data &= ~DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK;
-		data |= DF_MGCG_DISABLE;
-		WREG32(SOC15_REG_OFFSET(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater), data);
-	}
-
-	WREG32(SOC15_REG_OFFSET(DF, 0, mmFabricConfigAccessControl),
-	       mmFabricConfigAccessControl_DEFAULT);
-}
-
 static int soc15_common_set_clockgating_state(void *handle,
 					    enum amd_clockgating_state state)
 {
@@ -908,6 +906,7 @@ static int soc15_common_set_clockgating_state(void *handle,
 	switch (adev->asic_type) {
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		adev->nbio_funcs->update_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
 		adev->nbio_funcs->update_medium_grain_light_sleep(adev,
@@ -920,7 +919,7 @@ static int soc15_common_set_clockgating_state(void *handle,
 				state == AMD_CG_STATE_GATE ? true : false);
 		soc15_update_rom_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
-		soc15_update_df_medium_grain_clock_gating(adev,
+		adev->df_funcs->update_medium_grain_clock_gating(adev,
 				state == AMD_CG_STATE_GATE ? true : false);
 		break;
 	case CHIP_RAVEN:
@@ -973,10 +972,7 @@ static void soc15_common_get_clockgating_state(void *handle, u32 *flags)
 	if (!(data & CGTT_ROM_CLK_CTRL0__SOFT_OVERRIDE0_MASK))
 		*flags |= AMD_CG_SUPPORT_ROM_MGCG;
 
-	/* AMD_CG_SUPPORT_DF_MGCG */
-	data = RREG32(SOC15_REG_OFFSET(DF, 0, mmDF_PIE_AON0_DfGlobalClkGater));
-	if (data & DF_MGCG_ENABLE_15_CYCLE_DELAY)
-		*flags |= AMD_CG_SUPPORT_DF_MGCG;
+	adev->df_funcs->get_clockgating_state(adev, flags);
 }
 
 static int soc15_common_set_powergating_state(void *handle,
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.h b/drivers/gpu/drm/amd/amdgpu/soc15.h
index f70da8a29f86..1f714b7af520 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.h
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.h
@@ -55,5 +55,6 @@ void soc15_program_register_sequence(struct amdgpu_device *adev,
 					     const u32 array_size);
 
 int vega10_reg_base_init(struct amdgpu_device *adev);
+int vega20_reg_base_init(struct amdgpu_device *adev);
 
 #endif
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15_common.h b/drivers/gpu/drm/amd/amdgpu/soc15_common.h
index def865067edd..0942f492d2e1 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15_common.h
+++ b/drivers/gpu/drm/amd/amdgpu/soc15_common.h
@@ -47,6 +47,21 @@
 #define WREG32_SOC15_OFFSET(ip, inst, reg, offset, value) \
 	WREG32((adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg) + offset, value)
 
+#define SOC15_WAIT_ON_RREG(ip, inst, reg, expected_value, mask, ret) \
+	do {							\
+		uint32_t tmp_ = RREG32(adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg); \
+		uint32_t loop = adev->usec_timeout;		\
+		while ((tmp_ & (mask)) != (expected_value)) {	\
+			udelay(2);				\
+			tmp_ = RREG32(adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg); \
+			loop--;					\
+			if (!loop) {				\
+				ret = -ETIMEDOUT;		\
+				break;				\
+			}					\
+		}						\
+	} while (0)
+
 #endif
 
 
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15d.h b/drivers/gpu/drm/amd/amdgpu/soc15d.h
index 7f408f85fdb6..edfe50821cd9 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15d.h
+++ b/drivers/gpu/drm/amd/amdgpu/soc15d.h
@@ -53,6 +53,29 @@
 
 #define PACKET3_COMPUTE(op, n) (PACKET3(op, n) | 1 << 1)
 
+#define	PACKETJ_CONDITION_CHECK0	0
+#define	PACKETJ_CONDITION_CHECK1	1
+#define	PACKETJ_CONDITION_CHECK2	2
+#define	PACKETJ_CONDITION_CHECK3	3
+#define	PACKETJ_CONDITION_CHECK4	4
+#define	PACKETJ_CONDITION_CHECK5	5
+#define	PACKETJ_CONDITION_CHECK6	6
+#define	PACKETJ_CONDITION_CHECK7	7
+
+#define	PACKETJ_TYPE0	0
+#define	PACKETJ_TYPE1	1
+#define	PACKETJ_TYPE2	2
+#define	PACKETJ_TYPE3	3
+#define	PACKETJ_TYPE4	4
+#define	PACKETJ_TYPE5	5
+#define	PACKETJ_TYPE6	6
+#define	PACKETJ_TYPE7	7
+
+#define PACKETJ(reg, r, cond, type)	((reg & 0x3FFFF) |			\
+			 ((r & 0x3F) << 18) |			\
+			 ((cond & 0xF) << 24) |				\
+			 ((type & 0xF) << 28))
+
 /* Packet 3 types */
 #define	PACKET3_NOP					0x10
 #define	PACKET3_SET_BASE				0x11
@@ -159,6 +182,7 @@
 #define		EOP_TC_WB_ACTION_EN                     (1 << 15) /* L2 */
 #define		EOP_TCL1_ACTION_EN                      (1 << 16)
 #define		EOP_TC_ACTION_EN                        (1 << 17) /* L2 */
+#define		EOP_TC_NC_ACTION_EN			(1 << 19)
 #define		EOP_TC_MD_ACTION_EN			(1 << 21) /* L2 metadata */
 
 #define		DATA_SEL(x)                             ((x) << 29)
@@ -268,6 +292,11 @@
 			 * x=1: tmz_end
 			 */
 
+#define	PACKET3_INVALIDATE_TLBS				0x98
+#              define PACKET3_INVALIDATE_TLBS_DST_SEL(x)     ((x) << 0)
+#              define PACKET3_INVALIDATE_TLBS_ALL_HUB(x)     ((x) << 4)
+#              define PACKET3_INVALIDATE_TLBS_PASID(x)       ((x) << 5)
+#              define PACKET3_INVALIDATE_TLBS_FLUSH_TYPE(x)  ((x) << 29)
 #define PACKET3_SET_RESOURCES				0xA0
 /* 1. header
  * 2. CONTROL
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
index 948bb9437757..6fed3d7797a8 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v4_2.c
@@ -93,6 +93,7 @@ static void uvd_v4_2_ring_set_wptr(struct amdgpu_ring *ring)
 static int uvd_v4_2_early_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	adev->uvd.num_uvd_inst = 1;
 
 	uvd_v4_2_set_ring_funcs(adev);
 	uvd_v4_2_set_irq_funcs(adev);
@@ -107,7 +108,7 @@ static int uvd_v4_2_sw_init(void *handle)
 	int r;
 
 	/* UVD TRAP */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 124, &adev->uvd.irq);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 124, &adev->uvd.inst->irq);
 	if (r)
 		return r;
 
@@ -119,9 +120,9 @@ static int uvd_v4_2_sw_init(void *handle)
 	if (r)
 		return r;
 
-	ring = &adev->uvd.ring;
+	ring = &adev->uvd.inst->ring;
 	sprintf(ring->name, "uvd");
-	r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0);
+	r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst->irq, 0);
 
 	return r;
 }
@@ -150,7 +151,7 @@ static void uvd_v4_2_enable_mgcg(struct amdgpu_device *adev,
 static int uvd_v4_2_hw_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	uint32_t tmp;
 	int r;
 
@@ -208,7 +209,7 @@ done:
 static int uvd_v4_2_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 
 	if (RREG32(mmUVD_STATUS) != 0)
 		uvd_v4_2_stop(adev);
@@ -251,7 +252,7 @@ static int uvd_v4_2_resume(void *handle)
  */
 static int uvd_v4_2_start(struct amdgpu_device *adev)
 {
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	uint32_t rb_bufsz;
 	int i, j, r;
 	u32 tmp;
@@ -523,6 +524,18 @@ static void uvd_v4_2_ring_emit_ib(struct amdgpu_ring *ring,
 	amdgpu_ring_write(ring, ib->length_dw);
 }
 
+static void uvd_v4_2_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
+{
+	int i;
+
+	WARN_ON(ring->wptr % 2 || count % 2);
+
+	for (i = 0; i < count / 2; i++) {
+		amdgpu_ring_write(ring, PACKET0(mmUVD_NO_OP, 0));
+		amdgpu_ring_write(ring, 0);
+	}
+}
+
 /**
  * uvd_v4_2_mc_resume - memory controller programming
  *
@@ -536,7 +549,7 @@ static void uvd_v4_2_mc_resume(struct amdgpu_device *adev)
 	uint32_t size;
 
 	/* programm the VCPU memory controller bits 0-27 */
-	addr = (adev->uvd.gpu_addr + AMDGPU_UVD_FIRMWARE_OFFSET) >> 3;
+	addr = (adev->uvd.inst->gpu_addr + AMDGPU_UVD_FIRMWARE_OFFSET) >> 3;
 	size = AMDGPU_UVD_FIRMWARE_SIZE(adev) >> 3;
 	WREG32(mmUVD_VCPU_CACHE_OFFSET0, addr);
 	WREG32(mmUVD_VCPU_CACHE_SIZE0, size);
@@ -553,11 +566,11 @@ static void uvd_v4_2_mc_resume(struct amdgpu_device *adev)
 	WREG32(mmUVD_VCPU_CACHE_SIZE2, size);
 
 	/* bits 28-31 */
-	addr = (adev->uvd.gpu_addr >> 28) & 0xF;
+	addr = (adev->uvd.inst->gpu_addr >> 28) & 0xF;
 	WREG32(mmUVD_LMI_ADDR_EXT, (addr << 12) | (addr << 0));
 
 	/* bits 32-39 */
-	addr = (adev->uvd.gpu_addr >> 32) & 0xFF;
+	addr = (adev->uvd.inst->gpu_addr >> 32) & 0xFF;
 	WREG32(mmUVD_LMI_EXT40_ADDR, addr | (0x9 << 16) | (0x1 << 31));
 
 	WREG32(mmUVD_UDEC_ADDR_CONFIG, adev->gfx.config.gb_addr_config);
@@ -664,7 +677,7 @@ static int uvd_v4_2_process_interrupt(struct amdgpu_device *adev,
 				      struct amdgpu_iv_entry *entry)
 {
 	DRM_DEBUG("IH: UVD TRAP\n");
-	amdgpu_fence_process(&adev->uvd.ring);
+	amdgpu_fence_process(&adev->uvd.inst->ring);
 	return 0;
 }
 
@@ -688,7 +701,7 @@ static int uvd_v4_2_set_powergating_state(void *handle,
 
 	if (state == AMD_PG_STATE_GATE) {
 		uvd_v4_2_stop(adev);
-		if (adev->pg_flags & AMD_PG_SUPPORT_UVD && amdgpu_dpm == 0) {
+		if (adev->pg_flags & AMD_PG_SUPPORT_UVD && !adev->pm.dpm_enabled) {
 			if (!(RREG32_SMC(ixCURRENT_PG_STATUS) &
 				CURRENT_PG_STATUS__UVD_PG_STATUS_MASK)) {
 				WREG32(mmUVD_PGFSM_CONFIG, (UVD_PGFSM_CONFIG__UVD_PGFSM_FSM_ADDR_MASK   |
@@ -699,7 +712,7 @@ static int uvd_v4_2_set_powergating_state(void *handle,
 		}
 		return 0;
 	} else {
-		if (adev->pg_flags & AMD_PG_SUPPORT_UVD && amdgpu_dpm == 0) {
+		if (adev->pg_flags & AMD_PG_SUPPORT_UVD && !adev->pm.dpm_enabled) {
 			if (RREG32_SMC(ixCURRENT_PG_STATUS) &
 				CURRENT_PG_STATUS__UVD_PG_STATUS_MASK) {
 				WREG32(mmUVD_PGFSM_CONFIG, (UVD_PGFSM_CONFIG__UVD_PGFSM_FSM_ADDR_MASK   |
@@ -732,7 +745,6 @@ static const struct amd_ip_funcs uvd_v4_2_ip_funcs = {
 static const struct amdgpu_ring_funcs uvd_v4_2_ring_funcs = {
 	.type = AMDGPU_RING_TYPE_UVD,
 	.align_mask = 0xf,
-	.nop = PACKET0(mmUVD_NO_OP, 0),
 	.support_64bit_ptrs = false,
 	.get_rptr = uvd_v4_2_ring_get_rptr,
 	.get_wptr = uvd_v4_2_ring_get_wptr,
@@ -745,7 +757,7 @@ static const struct amdgpu_ring_funcs uvd_v4_2_ring_funcs = {
 	.emit_fence = uvd_v4_2_ring_emit_fence,
 	.test_ring = uvd_v4_2_ring_test_ring,
 	.test_ib = amdgpu_uvd_ring_test_ib,
-	.insert_nop = amdgpu_ring_insert_nop,
+	.insert_nop = uvd_v4_2_ring_insert_nop,
 	.pad_ib = amdgpu_ring_generic_pad_ib,
 	.begin_use = amdgpu_uvd_ring_begin_use,
 	.end_use = amdgpu_uvd_ring_end_use,
@@ -753,7 +765,7 @@ static const struct amdgpu_ring_funcs uvd_v4_2_ring_funcs = {
 
 static void uvd_v4_2_set_ring_funcs(struct amdgpu_device *adev)
 {
-	adev->uvd.ring.funcs = &uvd_v4_2_ring_funcs;
+	adev->uvd.inst->ring.funcs = &uvd_v4_2_ring_funcs;
 }
 
 static const struct amdgpu_irq_src_funcs uvd_v4_2_irq_funcs = {
@@ -763,8 +775,8 @@ static const struct amdgpu_irq_src_funcs uvd_v4_2_irq_funcs = {
 
 static void uvd_v4_2_set_irq_funcs(struct amdgpu_device *adev)
 {
-	adev->uvd.irq.num_types = 1;
-	adev->uvd.irq.funcs = &uvd_v4_2_irq_funcs;
+	adev->uvd.inst->irq.num_types = 1;
+	adev->uvd.inst->irq.funcs = &uvd_v4_2_irq_funcs;
 }
 
 const struct amdgpu_ip_block_version uvd_v4_2_ip_block =
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
index 6445d55e7d5a..aeaa1ca46a99 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v5_0.c
@@ -35,6 +35,7 @@
 #include "vi.h"
 #include "smu/smu_7_1_2_d.h"
 #include "smu/smu_7_1_2_sh_mask.h"
+#include "ivsrcid/ivsrcid_vislands30.h"
 
 static void uvd_v5_0_set_ring_funcs(struct amdgpu_device *adev);
 static void uvd_v5_0_set_irq_funcs(struct amdgpu_device *adev);
@@ -89,6 +90,7 @@ static void uvd_v5_0_ring_set_wptr(struct amdgpu_ring *ring)
 static int uvd_v5_0_early_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	adev->uvd.num_uvd_inst = 1;
 
 	uvd_v5_0_set_ring_funcs(adev);
 	uvd_v5_0_set_irq_funcs(adev);
@@ -103,7 +105,7 @@ static int uvd_v5_0_sw_init(void *handle)
 	int r;
 
 	/* UVD TRAP */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 124, &adev->uvd.irq);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_UVD_SYSTEM_MESSAGE, &adev->uvd.inst->irq);
 	if (r)
 		return r;
 
@@ -115,9 +117,9 @@ static int uvd_v5_0_sw_init(void *handle)
 	if (r)
 		return r;
 
-	ring = &adev->uvd.ring;
+	ring = &adev->uvd.inst->ring;
 	sprintf(ring->name, "uvd");
-	r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0);
+	r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst->irq, 0);
 
 	return r;
 }
@@ -144,7 +146,7 @@ static int uvd_v5_0_sw_fini(void *handle)
 static int uvd_v5_0_hw_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	uint32_t tmp;
 	int r;
 
@@ -204,7 +206,7 @@ done:
 static int uvd_v5_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 
 	if (RREG32(mmUVD_STATUS) != 0)
 		uvd_v5_0_stop(adev);
@@ -253,9 +255,9 @@ static void uvd_v5_0_mc_resume(struct amdgpu_device *adev)
 
 	/* programm memory controller bits 0-27 */
 	WREG32(mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
-			lower_32_bits(adev->uvd.gpu_addr));
+			lower_32_bits(adev->uvd.inst->gpu_addr));
 	WREG32(mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
-			upper_32_bits(adev->uvd.gpu_addr));
+			upper_32_bits(adev->uvd.inst->gpu_addr));
 
 	offset = AMDGPU_UVD_FIRMWARE_OFFSET;
 	size = AMDGPU_UVD_FIRMWARE_SIZE(adev);
@@ -287,7 +289,7 @@ static void uvd_v5_0_mc_resume(struct amdgpu_device *adev)
  */
 static int uvd_v5_0_start(struct amdgpu_device *adev)
 {
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	uint32_t rb_bufsz, tmp;
 	uint32_t lmi_swap_cntl;
 	uint32_t mp_swap_cntl;
@@ -540,6 +542,18 @@ static void uvd_v5_0_ring_emit_ib(struct amdgpu_ring *ring,
 	amdgpu_ring_write(ring, ib->length_dw);
 }
 
+static void uvd_v5_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
+{
+	int i;
+
+	WARN_ON(ring->wptr % 2 || count % 2);
+
+	for (i = 0; i < count / 2; i++) {
+		amdgpu_ring_write(ring, PACKET0(mmUVD_NO_OP, 0));
+		amdgpu_ring_write(ring, 0);
+	}
+}
+
 static bool uvd_v5_0_is_idle(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
@@ -586,7 +600,7 @@ static int uvd_v5_0_process_interrupt(struct amdgpu_device *adev,
 				      struct amdgpu_iv_entry *entry)
 {
 	DRM_DEBUG("IH: UVD TRAP\n");
-	amdgpu_fence_process(&adev->uvd.ring);
+	amdgpu_fence_process(&adev->uvd.inst->ring);
 	return 0;
 }
 
@@ -840,7 +854,6 @@ static const struct amd_ip_funcs uvd_v5_0_ip_funcs = {
 static const struct amdgpu_ring_funcs uvd_v5_0_ring_funcs = {
 	.type = AMDGPU_RING_TYPE_UVD,
 	.align_mask = 0xf,
-	.nop = PACKET0(mmUVD_NO_OP, 0),
 	.support_64bit_ptrs = false,
 	.get_rptr = uvd_v5_0_ring_get_rptr,
 	.get_wptr = uvd_v5_0_ring_get_wptr,
@@ -853,7 +866,7 @@ static const struct amdgpu_ring_funcs uvd_v5_0_ring_funcs = {
 	.emit_fence = uvd_v5_0_ring_emit_fence,
 	.test_ring = uvd_v5_0_ring_test_ring,
 	.test_ib = amdgpu_uvd_ring_test_ib,
-	.insert_nop = amdgpu_ring_insert_nop,
+	.insert_nop = uvd_v5_0_ring_insert_nop,
 	.pad_ib = amdgpu_ring_generic_pad_ib,
 	.begin_use = amdgpu_uvd_ring_begin_use,
 	.end_use = amdgpu_uvd_ring_end_use,
@@ -861,7 +874,7 @@ static const struct amdgpu_ring_funcs uvd_v5_0_ring_funcs = {
 
 static void uvd_v5_0_set_ring_funcs(struct amdgpu_device *adev)
 {
-	adev->uvd.ring.funcs = &uvd_v5_0_ring_funcs;
+	adev->uvd.inst->ring.funcs = &uvd_v5_0_ring_funcs;
 }
 
 static const struct amdgpu_irq_src_funcs uvd_v5_0_irq_funcs = {
@@ -871,8 +884,8 @@ static const struct amdgpu_irq_src_funcs uvd_v5_0_irq_funcs = {
 
 static void uvd_v5_0_set_irq_funcs(struct amdgpu_device *adev)
 {
-	adev->uvd.irq.num_types = 1;
-	adev->uvd.irq.funcs = &uvd_v5_0_irq_funcs;
+	adev->uvd.inst->irq.num_types = 1;
+	adev->uvd.inst->irq.funcs = &uvd_v5_0_irq_funcs;
 }
 
 const struct amdgpu_ip_block_version uvd_v5_0_ip_block =
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
index f26f515db2fb..598dbeaba636 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v6_0.c
@@ -36,6 +36,7 @@
 #include "bif/bif_5_1_d.h"
 #include "gmc/gmc_8_1_d.h"
 #include "vi.h"
+#include "ivsrcid/ivsrcid_vislands30.h"
 
 /* Polaris10/11/12 firmware version */
 #define FW_1_130_16 ((1 << 24) | (130 << 16) | (16 << 8))
@@ -62,7 +63,7 @@ static void uvd_v6_0_enable_mgcg(struct amdgpu_device *adev,
 static inline bool uvd_v6_0_enc_support(struct amdgpu_device *adev)
 {
 	return ((adev->asic_type >= CHIP_POLARIS10) &&
-			(adev->asic_type <= CHIP_POLARIS12) &&
+			(adev->asic_type <= CHIP_VEGAM) &&
 			(!adev->uvd.fw_version || adev->uvd.fw_version >= FW_1_130_16));
 }
 
@@ -91,7 +92,7 @@ static uint64_t uvd_v6_0_enc_ring_get_rptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	if (ring == &adev->uvd.ring_enc[0])
+	if (ring == &adev->uvd.inst->ring_enc[0])
 		return RREG32(mmUVD_RB_RPTR);
 	else
 		return RREG32(mmUVD_RB_RPTR2);
@@ -121,7 +122,7 @@ static uint64_t uvd_v6_0_enc_ring_get_wptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	if (ring == &adev->uvd.ring_enc[0])
+	if (ring == &adev->uvd.inst->ring_enc[0])
 		return RREG32(mmUVD_RB_WPTR);
 	else
 		return RREG32(mmUVD_RB_WPTR2);
@@ -152,7 +153,7 @@ static void uvd_v6_0_enc_ring_set_wptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	if (ring == &adev->uvd.ring_enc[0])
+	if (ring == &adev->uvd.inst->ring_enc[0])
 		WREG32(mmUVD_RB_WPTR,
 			lower_32_bits(ring->wptr));
 	else
@@ -247,12 +248,10 @@ static int uvd_v6_0_enc_get_create_msg(struct amdgpu_ring *ring, uint32_t handle
 	for (i = ib->length_dw; i < ib_size_dw; ++i)
 		ib->ptr[i] = 0x0;
 
-	r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f);
-	job->fence = dma_fence_get(f);
+	r = amdgpu_job_submit_direct(job, ring, &f);
 	if (r)
 		goto err;
 
-	amdgpu_job_free(job);
 	if (fence)
 		*fence = dma_fence_get(f);
 	dma_fence_put(f);
@@ -311,19 +310,13 @@ static int uvd_v6_0_enc_get_destroy_msg(struct amdgpu_ring *ring,
 	for (i = ib->length_dw; i < ib_size_dw; ++i)
 		ib->ptr[i] = 0x0;
 
-	if (direct) {
-		r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f);
-		job->fence = dma_fence_get(f);
-		if (r)
-			goto err;
-
-		amdgpu_job_free(job);
-	} else {
-		r = amdgpu_job_submit(job, ring, &ring->adev->vce.entity,
+	if (direct)
+		r = amdgpu_job_submit_direct(job, ring, &f);
+	else
+		r = amdgpu_job_submit(job, &ring->adev->vce.entity,
 				      AMDGPU_FENCE_OWNER_UNDEFINED, &f);
-		if (r)
-			goto err;
-	}
+	if (r)
+		goto err;
 
 	if (fence)
 		*fence = dma_fence_get(f);
@@ -375,6 +368,7 @@ error:
 static int uvd_v6_0_early_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	adev->uvd.num_uvd_inst = 1;
 
 	if (!(adev->flags & AMD_IS_APU) &&
 	    (RREG32_SMC(ixCC_HARVEST_FUSES) & CC_HARVEST_FUSES__UVD_DISABLE_MASK))
@@ -399,14 +393,14 @@ static int uvd_v6_0_sw_init(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	/* UVD TRAP */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 124, &adev->uvd.irq);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_UVD_SYSTEM_MESSAGE, &adev->uvd.inst->irq);
 	if (r)
 		return r;
 
 	/* UVD ENC TRAP */
 	if (uvd_v6_0_enc_support(adev)) {
 		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-			r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, i + 119, &adev->uvd.irq);
+			r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, i + VISLANDS30_IV_SRCID_UVD_ENC_GEN_PURP, &adev->uvd.inst->irq);
 			if (r)
 				return r;
 		}
@@ -418,39 +412,29 @@ static int uvd_v6_0_sw_init(void *handle)
 
 	if (!uvd_v6_0_enc_support(adev)) {
 		for (i = 0; i < adev->uvd.num_enc_rings; ++i)
-			adev->uvd.ring_enc[i].funcs = NULL;
+			adev->uvd.inst->ring_enc[i].funcs = NULL;
 
-		adev->uvd.irq.num_types = 1;
+		adev->uvd.inst->irq.num_types = 1;
 		adev->uvd.num_enc_rings = 0;
 
 		DRM_INFO("UVD ENC is disabled\n");
-	} else {
-		struct drm_sched_rq *rq;
-		ring = &adev->uvd.ring_enc[0];
-		rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
-		r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity_enc,
-					  rq, amdgpu_sched_jobs, NULL);
-		if (r) {
-			DRM_ERROR("Failed setting up UVD ENC run queue.\n");
-			return r;
-		}
 	}
 
 	r = amdgpu_uvd_resume(adev);
 	if (r)
 		return r;
 
-	ring = &adev->uvd.ring;
+	ring = &adev->uvd.inst->ring;
 	sprintf(ring->name, "uvd");
-	r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0);
+	r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst->irq, 0);
 	if (r)
 		return r;
 
 	if (uvd_v6_0_enc_support(adev)) {
 		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-			ring = &adev->uvd.ring_enc[i];
+			ring = &adev->uvd.inst->ring_enc[i];
 			sprintf(ring->name, "uvd_enc%d", i);
-			r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0);
+			r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst->irq, 0);
 			if (r)
 				return r;
 		}
@@ -469,10 +453,8 @@ static int uvd_v6_0_sw_fini(void *handle)
 		return r;
 
 	if (uvd_v6_0_enc_support(adev)) {
-		drm_sched_entity_fini(&adev->uvd.ring_enc[0].sched, &adev->uvd.entity_enc);
-
 		for (i = 0; i < adev->uvd.num_enc_rings; ++i)
-			amdgpu_ring_fini(&adev->uvd.ring_enc[i]);
+			amdgpu_ring_fini(&adev->uvd.inst->ring_enc[i]);
 	}
 
 	return amdgpu_uvd_sw_fini(adev);
@@ -488,7 +470,7 @@ static int uvd_v6_0_sw_fini(void *handle)
 static int uvd_v6_0_hw_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	uint32_t tmp;
 	int i, r;
 
@@ -532,7 +514,7 @@ static int uvd_v6_0_hw_init(void *handle)
 
 	if (uvd_v6_0_enc_support(adev)) {
 		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-			ring = &adev->uvd.ring_enc[i];
+			ring = &adev->uvd.inst->ring_enc[i];
 			ring->ready = true;
 			r = amdgpu_ring_test_ring(ring);
 			if (r) {
@@ -563,7 +545,7 @@ done:
 static int uvd_v6_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 
 	if (RREG32(mmUVD_STATUS) != 0)
 		uvd_v6_0_stop(adev);
@@ -611,9 +593,9 @@ static void uvd_v6_0_mc_resume(struct amdgpu_device *adev)
 
 	/* programm memory controller bits 0-27 */
 	WREG32(mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
-			lower_32_bits(adev->uvd.gpu_addr));
+			lower_32_bits(adev->uvd.inst->gpu_addr));
 	WREG32(mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
-			upper_32_bits(adev->uvd.gpu_addr));
+			upper_32_bits(adev->uvd.inst->gpu_addr));
 
 	offset = AMDGPU_UVD_FIRMWARE_OFFSET;
 	size = AMDGPU_UVD_FIRMWARE_SIZE(adev);
@@ -726,7 +708,7 @@ static void cz_set_uvd_clock_gating_branches(struct amdgpu_device *adev,
  */
 static int uvd_v6_0_start(struct amdgpu_device *adev)
 {
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring = &adev->uvd.inst->ring;
 	uint32_t rb_bufsz, tmp;
 	uint32_t lmi_swap_cntl;
 	uint32_t mp_swap_cntl;
@@ -866,14 +848,14 @@ static int uvd_v6_0_start(struct amdgpu_device *adev)
 	WREG32_FIELD(UVD_RBC_RB_CNTL, RB_NO_FETCH, 0);
 
 	if (uvd_v6_0_enc_support(adev)) {
-		ring = &adev->uvd.ring_enc[0];
+		ring = &adev->uvd.inst->ring_enc[0];
 		WREG32(mmUVD_RB_RPTR, lower_32_bits(ring->wptr));
 		WREG32(mmUVD_RB_WPTR, lower_32_bits(ring->wptr));
 		WREG32(mmUVD_RB_BASE_LO, ring->gpu_addr);
 		WREG32(mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
 		WREG32(mmUVD_RB_SIZE, ring->ring_size / 4);
 
-		ring = &adev->uvd.ring_enc[1];
+		ring = &adev->uvd.inst->ring_enc[1];
 		WREG32(mmUVD_RB_RPTR2, lower_32_bits(ring->wptr));
 		WREG32(mmUVD_RB_WPTR2, lower_32_bits(ring->wptr));
 		WREG32(mmUVD_RB_BASE_LO2, ring->gpu_addr);
@@ -964,6 +946,16 @@ static void uvd_v6_0_enc_ring_emit_fence(struct amdgpu_ring *ring, u64 addr,
 }
 
 /**
+ * uvd_v6_0_ring_emit_hdp_flush - skip HDP flushing
+ *
+ * @ring: amdgpu_ring pointer
+ */
+static void uvd_v6_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
+{
+	/* The firmware doesn't seem to like touching registers at this point. */
+}
+
+/**
  * uvd_v6_0_ring_test_ring - register write test
  *
  * @ring: amdgpu_ring pointer
@@ -1089,6 +1081,18 @@ static void uvd_v6_0_ring_emit_pipeline_sync(struct amdgpu_ring *ring)
 	amdgpu_ring_write(ring, 0xE);
 }
 
+static void uvd_v6_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
+{
+	int i;
+
+	WARN_ON(ring->wptr % 2 || count % 2);
+
+	for (i = 0; i < count / 2; i++) {
+		amdgpu_ring_write(ring, PACKET0(mmUVD_NO_OP, 0));
+		amdgpu_ring_write(ring, 0);
+	}
+}
+
 static void uvd_v6_0_enc_ring_emit_pipeline_sync(struct amdgpu_ring *ring)
 {
 	uint32_t seq = ring->fence_drv.sync_seq;
@@ -1148,10 +1152,10 @@ static bool uvd_v6_0_check_soft_reset(void *handle)
 		srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset, SRBM_SOFT_RESET, SOFT_RESET_UVD, 1);
 
 	if (srbm_soft_reset) {
-		adev->uvd.srbm_soft_reset = srbm_soft_reset;
+		adev->uvd.inst->srbm_soft_reset = srbm_soft_reset;
 		return true;
 	} else {
-		adev->uvd.srbm_soft_reset = 0;
+		adev->uvd.inst->srbm_soft_reset = 0;
 		return false;
 	}
 }
@@ -1160,7 +1164,7 @@ static int uvd_v6_0_pre_soft_reset(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!adev->uvd.srbm_soft_reset)
+	if (!adev->uvd.inst->srbm_soft_reset)
 		return 0;
 
 	uvd_v6_0_stop(adev);
@@ -1172,9 +1176,9 @@ static int uvd_v6_0_soft_reset(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	u32 srbm_soft_reset;
 
-	if (!adev->uvd.srbm_soft_reset)
+	if (!adev->uvd.inst->srbm_soft_reset)
 		return 0;
-	srbm_soft_reset = adev->uvd.srbm_soft_reset;
+	srbm_soft_reset = adev->uvd.inst->srbm_soft_reset;
 
 	if (srbm_soft_reset) {
 		u32 tmp;
@@ -1202,7 +1206,7 @@ static int uvd_v6_0_post_soft_reset(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!adev->uvd.srbm_soft_reset)
+	if (!adev->uvd.inst->srbm_soft_reset)
 		return 0;
 
 	mdelay(5);
@@ -1228,17 +1232,17 @@ static int uvd_v6_0_process_interrupt(struct amdgpu_device *adev,
 
 	switch (entry->src_id) {
 	case 124:
-		amdgpu_fence_process(&adev->uvd.ring);
+		amdgpu_fence_process(&adev->uvd.inst->ring);
 		break;
 	case 119:
 		if (likely(uvd_v6_0_enc_support(adev)))
-			amdgpu_fence_process(&adev->uvd.ring_enc[0]);
+			amdgpu_fence_process(&adev->uvd.inst->ring_enc[0]);
 		else
 			int_handled = false;
 		break;
 	case 120:
 		if (likely(uvd_v6_0_enc_support(adev)))
-			amdgpu_fence_process(&adev->uvd.ring_enc[1]);
+			amdgpu_fence_process(&adev->uvd.inst->ring_enc[1]);
 		else
 			int_handled = false;
 		break;
@@ -1521,22 +1525,22 @@ static const struct amd_ip_funcs uvd_v6_0_ip_funcs = {
 static const struct amdgpu_ring_funcs uvd_v6_0_ring_phys_funcs = {
 	.type = AMDGPU_RING_TYPE_UVD,
 	.align_mask = 0xf,
-	.nop = PACKET0(mmUVD_NO_OP, 0),
 	.support_64bit_ptrs = false,
 	.get_rptr = uvd_v6_0_ring_get_rptr,
 	.get_wptr = uvd_v6_0_ring_get_wptr,
 	.set_wptr = uvd_v6_0_ring_set_wptr,
 	.parse_cs = amdgpu_uvd_ring_parse_cs,
 	.emit_frame_size =
-		6 + 6 + /* hdp flush / invalidate */
+		6 + /* hdp invalidate */
 		10 + /* uvd_v6_0_ring_emit_pipeline_sync */
 		14, /* uvd_v6_0_ring_emit_fence x1 no user fence */
 	.emit_ib_size = 8, /* uvd_v6_0_ring_emit_ib */
 	.emit_ib = uvd_v6_0_ring_emit_ib,
 	.emit_fence = uvd_v6_0_ring_emit_fence,
+	.emit_hdp_flush = uvd_v6_0_ring_emit_hdp_flush,
 	.test_ring = uvd_v6_0_ring_test_ring,
 	.test_ib = amdgpu_uvd_ring_test_ib,
-	.insert_nop = amdgpu_ring_insert_nop,
+	.insert_nop = uvd_v6_0_ring_insert_nop,
 	.pad_ib = amdgpu_ring_generic_pad_ib,
 	.begin_use = amdgpu_uvd_ring_begin_use,
 	.end_use = amdgpu_uvd_ring_end_use,
@@ -1546,13 +1550,12 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_phys_funcs = {
 static const struct amdgpu_ring_funcs uvd_v6_0_ring_vm_funcs = {
 	.type = AMDGPU_RING_TYPE_UVD,
 	.align_mask = 0xf,
-	.nop = PACKET0(mmUVD_NO_OP, 0),
 	.support_64bit_ptrs = false,
 	.get_rptr = uvd_v6_0_ring_get_rptr,
 	.get_wptr = uvd_v6_0_ring_get_wptr,
 	.set_wptr = uvd_v6_0_ring_set_wptr,
 	.emit_frame_size =
-		6 + 6 + /* hdp flush / invalidate */
+		6 + /* hdp invalidate */
 		10 + /* uvd_v6_0_ring_emit_pipeline_sync */
 		VI_FLUSH_GPU_TLB_NUM_WREG * 6 + 8 + /* uvd_v6_0_ring_emit_vm_flush */
 		14 + 14, /* uvd_v6_0_ring_emit_fence x2 vm fence */
@@ -1561,9 +1564,10 @@ static const struct amdgpu_ring_funcs uvd_v6_0_ring_vm_funcs = {
 	.emit_fence = uvd_v6_0_ring_emit_fence,
 	.emit_vm_flush = uvd_v6_0_ring_emit_vm_flush,
 	.emit_pipeline_sync = uvd_v6_0_ring_emit_pipeline_sync,
+	.emit_hdp_flush = uvd_v6_0_ring_emit_hdp_flush,
 	.test_ring = uvd_v6_0_ring_test_ring,
 	.test_ib = amdgpu_uvd_ring_test_ib,
-	.insert_nop = amdgpu_ring_insert_nop,
+	.insert_nop = uvd_v6_0_ring_insert_nop,
 	.pad_ib = amdgpu_ring_generic_pad_ib,
 	.begin_use = amdgpu_uvd_ring_begin_use,
 	.end_use = amdgpu_uvd_ring_end_use,
@@ -1600,10 +1604,10 @@ static const struct amdgpu_ring_funcs uvd_v6_0_enc_ring_vm_funcs = {
 static void uvd_v6_0_set_ring_funcs(struct amdgpu_device *adev)
 {
 	if (adev->asic_type >= CHIP_POLARIS10) {
-		adev->uvd.ring.funcs = &uvd_v6_0_ring_vm_funcs;
+		adev->uvd.inst->ring.funcs = &uvd_v6_0_ring_vm_funcs;
 		DRM_INFO("UVD is enabled in VM mode\n");
 	} else {
-		adev->uvd.ring.funcs = &uvd_v6_0_ring_phys_funcs;
+		adev->uvd.inst->ring.funcs = &uvd_v6_0_ring_phys_funcs;
 		DRM_INFO("UVD is enabled in physical mode\n");
 	}
 }
@@ -1613,7 +1617,7 @@ static void uvd_v6_0_set_enc_ring_funcs(struct amdgpu_device *adev)
 	int i;
 
 	for (i = 0; i < adev->uvd.num_enc_rings; ++i)
-		adev->uvd.ring_enc[i].funcs = &uvd_v6_0_enc_ring_vm_funcs;
+		adev->uvd.inst->ring_enc[i].funcs = &uvd_v6_0_enc_ring_vm_funcs;
 
 	DRM_INFO("UVD ENC is enabled in VM mode\n");
 }
@@ -1626,11 +1630,11 @@ static const struct amdgpu_irq_src_funcs uvd_v6_0_irq_funcs = {
 static void uvd_v6_0_set_irq_funcs(struct amdgpu_device *adev)
 {
 	if (uvd_v6_0_enc_support(adev))
-		adev->uvd.irq.num_types = adev->uvd.num_enc_rings + 1;
+		adev->uvd.inst->irq.num_types = adev->uvd.num_enc_rings + 1;
 	else
-		adev->uvd.irq.num_types = 1;
+		adev->uvd.inst->irq.num_types = 1;
 
-	adev->uvd.irq.funcs = &uvd_v6_0_irq_funcs;
+	adev->uvd.inst->irq.funcs = &uvd_v6_0_irq_funcs;
 }
 
 const struct amdgpu_ip_block_version uvd_v6_0_ip_block =
diff --git a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
index eddc57f3b72a..db5f3d78ab12 100644
--- a/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/uvd_v7_0.c
@@ -39,6 +39,9 @@
 #include "hdp/hdp_4_0_offset.h"
 #include "mmhub/mmhub_1_0_offset.h"
 #include "mmhub/mmhub_1_0_sh_mask.h"
+#include "ivsrcid/uvd/irqsrcs_uvd_7_0.h"
+
+#define UVD7_MAX_HW_INSTANCES_VEGA20			2
 
 static void uvd_v7_0_set_ring_funcs(struct amdgpu_device *adev);
 static void uvd_v7_0_set_enc_ring_funcs(struct amdgpu_device *adev);
@@ -47,6 +50,11 @@ static int uvd_v7_0_start(struct amdgpu_device *adev);
 static void uvd_v7_0_stop(struct amdgpu_device *adev);
 static int uvd_v7_0_sriov_start(struct amdgpu_device *adev);
 
+static int amdgpu_ih_clientid_uvds[] = {
+	SOC15_IH_CLIENTID_UVD,
+	SOC15_IH_CLIENTID_UVD1
+};
+
 /**
  * uvd_v7_0_ring_get_rptr - get read pointer
  *
@@ -58,7 +66,7 @@ static uint64_t uvd_v7_0_ring_get_rptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	return RREG32_SOC15(UVD, 0, mmUVD_RBC_RB_RPTR);
+	return RREG32_SOC15(UVD, ring->me, mmUVD_RBC_RB_RPTR);
 }
 
 /**
@@ -72,10 +80,10 @@ static uint64_t uvd_v7_0_enc_ring_get_rptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	if (ring == &adev->uvd.ring_enc[0])
-		return RREG32_SOC15(UVD, 0, mmUVD_RB_RPTR);
+	if (ring == &adev->uvd.inst[ring->me].ring_enc[0])
+		return RREG32_SOC15(UVD, ring->me, mmUVD_RB_RPTR);
 	else
-		return RREG32_SOC15(UVD, 0, mmUVD_RB_RPTR2);
+		return RREG32_SOC15(UVD, ring->me, mmUVD_RB_RPTR2);
 }
 
 /**
@@ -89,7 +97,7 @@ static uint64_t uvd_v7_0_ring_get_wptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	return RREG32_SOC15(UVD, 0, mmUVD_RBC_RB_WPTR);
+	return RREG32_SOC15(UVD, ring->me, mmUVD_RBC_RB_WPTR);
 }
 
 /**
@@ -106,10 +114,10 @@ static uint64_t uvd_v7_0_enc_ring_get_wptr(struct amdgpu_ring *ring)
 	if (ring->use_doorbell)
 		return adev->wb.wb[ring->wptr_offs];
 
-	if (ring == &adev->uvd.ring_enc[0])
-		return RREG32_SOC15(UVD, 0, mmUVD_RB_WPTR);
+	if (ring == &adev->uvd.inst[ring->me].ring_enc[0])
+		return RREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR);
 	else
-		return RREG32_SOC15(UVD, 0, mmUVD_RB_WPTR2);
+		return RREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR2);
 }
 
 /**
@@ -123,7 +131,7 @@ static void uvd_v7_0_ring_set_wptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	WREG32_SOC15(UVD, 0, mmUVD_RBC_RB_WPTR, lower_32_bits(ring->wptr));
+	WREG32_SOC15(UVD, ring->me, mmUVD_RBC_RB_WPTR, lower_32_bits(ring->wptr));
 }
 
 /**
@@ -144,11 +152,11 @@ static void uvd_v7_0_enc_ring_set_wptr(struct amdgpu_ring *ring)
 		return;
 	}
 
-	if (ring == &adev->uvd.ring_enc[0])
-		WREG32_SOC15(UVD, 0, mmUVD_RB_WPTR,
+	if (ring == &adev->uvd.inst[ring->me].ring_enc[0])
+		WREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR,
 			lower_32_bits(ring->wptr));
 	else
-		WREG32_SOC15(UVD, 0, mmUVD_RB_WPTR2,
+		WREG32_SOC15(UVD, ring->me, mmUVD_RB_WPTR2,
 			lower_32_bits(ring->wptr));
 }
 
@@ -170,8 +178,8 @@ static int uvd_v7_0_enc_ring_test_ring(struct amdgpu_ring *ring)
 
 	r = amdgpu_ring_alloc(ring, 16);
 	if (r) {
-		DRM_ERROR("amdgpu: uvd enc failed to lock ring %d (%d).\n",
-			  ring->idx, r);
+		DRM_ERROR("amdgpu: uvd enc failed to lock (%d)ring %d (%d).\n",
+			  ring->me, ring->idx, r);
 		return r;
 	}
 	amdgpu_ring_write(ring, HEVC_ENC_CMD_END);
@@ -184,11 +192,11 @@ static int uvd_v7_0_enc_ring_test_ring(struct amdgpu_ring *ring)
 	}
 
 	if (i < adev->usec_timeout) {
-		DRM_DEBUG("ring test on %d succeeded in %d usecs\n",
-			 ring->idx, i);
+		DRM_DEBUG("(%d)ring test on %d succeeded in %d usecs\n",
+			 ring->me, ring->idx, i);
 	} else {
-		DRM_ERROR("amdgpu: ring %d test failed\n",
-			  ring->idx);
+		DRM_ERROR("amdgpu: (%d)ring %d test failed\n",
+			  ring->me, ring->idx);
 		r = -ETIMEDOUT;
 	}
 
@@ -242,12 +250,10 @@ static int uvd_v7_0_enc_get_create_msg(struct amdgpu_ring *ring, uint32_t handle
 	for (i = ib->length_dw; i < ib_size_dw; ++i)
 		ib->ptr[i] = 0x0;
 
-	r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f);
-	job->fence = dma_fence_get(f);
+	r = amdgpu_job_submit_direct(job, ring, &f);
 	if (r)
 		goto err;
 
-	amdgpu_job_free(job);
 	if (fence)
 		*fence = dma_fence_get(f);
 	dma_fence_put(f);
@@ -305,19 +311,13 @@ int uvd_v7_0_enc_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
 	for (i = ib->length_dw; i < ib_size_dw; ++i)
 		ib->ptr[i] = 0x0;
 
-	if (direct) {
-		r = amdgpu_ib_schedule(ring, 1, ib, NULL, &f);
-		job->fence = dma_fence_get(f);
-		if (r)
-			goto err;
-
-		amdgpu_job_free(job);
-	} else {
-		r = amdgpu_job_submit(job, ring, &ring->adev->vce.entity,
+	if (direct)
+		r = amdgpu_job_submit_direct(job, ring, &f);
+	else
+		r = amdgpu_job_submit(job, &ring->adev->vce.entity,
 				      AMDGPU_FENCE_OWNER_UNDEFINED, &f);
-		if (r)
-			goto err;
-	}
+	if (r)
+		goto err;
 
 	if (fence)
 		*fence = dma_fence_get(f);
@@ -342,24 +342,24 @@ static int uvd_v7_0_enc_ring_test_ib(struct amdgpu_ring *ring, long timeout)
 
 	r = uvd_v7_0_enc_get_create_msg(ring, 1, NULL);
 	if (r) {
-		DRM_ERROR("amdgpu: failed to get create msg (%ld).\n", r);
+		DRM_ERROR("amdgpu: (%d)failed to get create msg (%ld).\n", ring->me, r);
 		goto error;
 	}
 
 	r = uvd_v7_0_enc_get_destroy_msg(ring, 1, true, &fence);
 	if (r) {
-		DRM_ERROR("amdgpu: failed to get destroy ib (%ld).\n", r);
+		DRM_ERROR("amdgpu: (%d)failed to get destroy ib (%ld).\n", ring->me, r);
 		goto error;
 	}
 
 	r = dma_fence_wait_timeout(fence, false, timeout);
 	if (r == 0) {
-		DRM_ERROR("amdgpu: IB test timed out.\n");
+		DRM_ERROR("amdgpu: (%d)IB test timed out.\n", ring->me);
 		r = -ETIMEDOUT;
 	} else if (r < 0) {
-		DRM_ERROR("amdgpu: fence wait failed (%ld).\n", r);
+		DRM_ERROR("amdgpu: (%d)fence wait failed (%ld).\n", ring->me, r);
 	} else {
-		DRM_DEBUG("ib test on ring %d succeeded\n", ring->idx);
+		DRM_DEBUG("ib test on (%d)ring %d succeeded\n", ring->me, ring->idx);
 		r = 0;
 	}
 error:
@@ -370,6 +370,10 @@ error:
 static int uvd_v7_0_early_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	if (adev->asic_type == CHIP_VEGA20)
+		adev->uvd.num_uvd_inst = UVD7_MAX_HW_INSTANCES_VEGA20;
+	else
+		adev->uvd.num_uvd_inst = 1;
 
 	if (amdgpu_sriov_vf(adev))
 		adev->uvd.num_enc_rings = 1;
@@ -385,20 +389,21 @@ static int uvd_v7_0_early_init(void *handle)
 static int uvd_v7_0_sw_init(void *handle)
 {
 	struct amdgpu_ring *ring;
-	struct drm_sched_rq *rq;
-	int i, r;
+	int i, j, r;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	/* UVD TRAP */
-	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UVD, 124, &adev->uvd.irq);
-	if (r)
-		return r;
-
-	/* UVD ENC TRAP */
-	for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-		r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_UVD, i + 119, &adev->uvd.irq);
+	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
+		/* UVD TRAP */
+		r = amdgpu_irq_add_id(adev, amdgpu_ih_clientid_uvds[j], UVD_7_0__SRCID__UVD_SYSTEM_MESSAGE_INTERRUPT, &adev->uvd.inst[j].irq);
 		if (r)
 			return r;
+
+		/* UVD ENC TRAP */
+		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
+			r = amdgpu_irq_add_id(adev, amdgpu_ih_clientid_uvds[j], i + UVD_7_0__SRCID__UVD_ENC_GEN_PURP, &adev->uvd.inst[j].irq);
+			if (r)
+				return r;
+		}
 	}
 
 	r = amdgpu_uvd_sw_init(adev);
@@ -415,43 +420,37 @@ static int uvd_v7_0_sw_init(void *handle)
 		DRM_INFO("PSP loading UVD firmware\n");
 	}
 
-	ring = &adev->uvd.ring_enc[0];
-	rq = &ring->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
-	r = drm_sched_entity_init(&ring->sched, &adev->uvd.entity_enc,
-				  rq, amdgpu_sched_jobs, NULL);
-	if (r) {
-		DRM_ERROR("Failed setting up UVD ENC run queue.\n");
-		return r;
-	}
-
 	r = amdgpu_uvd_resume(adev);
 	if (r)
 		return r;
-	if (!amdgpu_sriov_vf(adev)) {
-		ring = &adev->uvd.ring;
-		sprintf(ring->name, "uvd");
-		r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0);
-		if (r)
-			return r;
-	}
 
-	for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-		ring = &adev->uvd.ring_enc[i];
-		sprintf(ring->name, "uvd_enc%d", i);
-		if (amdgpu_sriov_vf(adev)) {
-			ring->use_doorbell = true;
-
-			/* currently only use the first enconding ring for
-			 * sriov, so set unused location for other unused rings.
-			 */
-			if (i == 0)
-				ring->doorbell_index = AMDGPU_DOORBELL64_UVD_RING0_1 * 2;
-			else
-				ring->doorbell_index = AMDGPU_DOORBELL64_UVD_RING2_3 * 2 + 1;
+	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
+		if (!amdgpu_sriov_vf(adev)) {
+			ring = &adev->uvd.inst[j].ring;
+			sprintf(ring->name, "uvd<%d>", j);
+			r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst[j].irq, 0);
+			if (r)
+				return r;
+		}
+
+		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
+			ring = &adev->uvd.inst[j].ring_enc[i];
+			sprintf(ring->name, "uvd_enc%d<%d>", i, j);
+			if (amdgpu_sriov_vf(adev)) {
+				ring->use_doorbell = true;
+
+				/* currently only use the first enconding ring for
+				 * sriov, so set unused location for other unused rings.
+				 */
+				if (i == 0)
+					ring->doorbell_index = AMDGPU_DOORBELL64_UVD_RING0_1 * 2;
+				else
+					ring->doorbell_index = AMDGPU_DOORBELL64_UVD_RING2_3 * 2 + 1;
+			}
+			r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.inst[j].irq, 0);
+			if (r)
+				return r;
 		}
-		r = amdgpu_ring_init(adev, ring, 512, &adev->uvd.irq, 0);
-		if (r)
-			return r;
 	}
 
 	r = amdgpu_virt_alloc_mm_table(adev);
@@ -463,7 +462,7 @@ static int uvd_v7_0_sw_init(void *handle)
 
 static int uvd_v7_0_sw_fini(void *handle)
 {
-	int i, r;
+	int i, j, r;
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	amdgpu_virt_free_mm_table(adev);
@@ -472,11 +471,10 @@ static int uvd_v7_0_sw_fini(void *handle)
 	if (r)
 		return r;
 
-	drm_sched_entity_fini(&adev->uvd.ring_enc[0].sched, &adev->uvd.entity_enc);
-
-	for (i = 0; i < adev->uvd.num_enc_rings; ++i)
-		amdgpu_ring_fini(&adev->uvd.ring_enc[i]);
-
+	for (j = 0; j < adev->uvd.num_uvd_inst; ++j) {
+		for (i = 0; i < adev->uvd.num_enc_rings; ++i)
+			amdgpu_ring_fini(&adev->uvd.inst[j].ring_enc[i]);
+	}
 	return amdgpu_uvd_sw_fini(adev);
 }
 
@@ -490,9 +488,9 @@ static int uvd_v7_0_sw_fini(void *handle)
 static int uvd_v7_0_hw_init(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring;
 	uint32_t tmp;
-	int i, r;
+	int i, j, r;
 
 	if (amdgpu_sriov_vf(adev))
 		r = uvd_v7_0_sriov_start(adev);
@@ -501,57 +499,60 @@ static int uvd_v7_0_hw_init(void *handle)
 	if (r)
 		goto done;
 
-	if (!amdgpu_sriov_vf(adev)) {
-		ring->ready = true;
-		r = amdgpu_ring_test_ring(ring);
-		if (r) {
-			ring->ready = false;
-			goto done;
+	for (j = 0; j < adev->uvd.num_uvd_inst; ++j) {
+		ring = &adev->uvd.inst[j].ring;
+
+		if (!amdgpu_sriov_vf(adev)) {
+			ring->ready = true;
+			r = amdgpu_ring_test_ring(ring);
+			if (r) {
+				ring->ready = false;
+				goto done;
+			}
+
+			r = amdgpu_ring_alloc(ring, 10);
+			if (r) {
+				DRM_ERROR("amdgpu: (%d)ring failed to lock UVD ring (%d).\n", j, r);
+				goto done;
+			}
+
+			tmp = PACKET0(SOC15_REG_OFFSET(UVD, j,
+				mmUVD_SEMA_WAIT_FAULT_TIMEOUT_CNTL), 0);
+			amdgpu_ring_write(ring, tmp);
+			amdgpu_ring_write(ring, 0xFFFFF);
+
+			tmp = PACKET0(SOC15_REG_OFFSET(UVD, j,
+				mmUVD_SEMA_WAIT_INCOMPLETE_TIMEOUT_CNTL), 0);
+			amdgpu_ring_write(ring, tmp);
+			amdgpu_ring_write(ring, 0xFFFFF);
+
+			tmp = PACKET0(SOC15_REG_OFFSET(UVD, j,
+				mmUVD_SEMA_SIGNAL_INCOMPLETE_TIMEOUT_CNTL), 0);
+			amdgpu_ring_write(ring, tmp);
+			amdgpu_ring_write(ring, 0xFFFFF);
+
+			/* Clear timeout status bits */
+			amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, j,
+				mmUVD_SEMA_TIMEOUT_STATUS), 0));
+			amdgpu_ring_write(ring, 0x8);
+
+			amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, j,
+				mmUVD_SEMA_CNTL), 0));
+			amdgpu_ring_write(ring, 3);
+
+			amdgpu_ring_commit(ring);
 		}
 
-		r = amdgpu_ring_alloc(ring, 10);
-		if (r) {
-			DRM_ERROR("amdgpu: ring failed to lock UVD ring (%d).\n", r);
-			goto done;
-		}
-
-		tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0,
-			mmUVD_SEMA_WAIT_FAULT_TIMEOUT_CNTL), 0);
-		amdgpu_ring_write(ring, tmp);
-		amdgpu_ring_write(ring, 0xFFFFF);
-
-		tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0,
-			mmUVD_SEMA_WAIT_INCOMPLETE_TIMEOUT_CNTL), 0);
-		amdgpu_ring_write(ring, tmp);
-		amdgpu_ring_write(ring, 0xFFFFF);
-
-		tmp = PACKET0(SOC15_REG_OFFSET(UVD, 0,
-			mmUVD_SEMA_SIGNAL_INCOMPLETE_TIMEOUT_CNTL), 0);
-		amdgpu_ring_write(ring, tmp);
-		amdgpu_ring_write(ring, 0xFFFFF);
-
-		/* Clear timeout status bits */
-		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0,
-			mmUVD_SEMA_TIMEOUT_STATUS), 0));
-		amdgpu_ring_write(ring, 0x8);
-
-		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0,
-			mmUVD_SEMA_CNTL), 0));
-		amdgpu_ring_write(ring, 3);
-
-		amdgpu_ring_commit(ring);
-	}
-
-	for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
-		ring = &adev->uvd.ring_enc[i];
-		ring->ready = true;
-		r = amdgpu_ring_test_ring(ring);
-		if (r) {
-			ring->ready = false;
-			goto done;
+		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
+			ring = &adev->uvd.inst[j].ring_enc[i];
+			ring->ready = true;
+			r = amdgpu_ring_test_ring(ring);
+			if (r) {
+				ring->ready = false;
+				goto done;
+			}
 		}
 	}
-
 done:
 	if (!r)
 		DRM_INFO("UVD and UVD ENC initialized successfully.\n");
@@ -569,7 +570,7 @@ done:
 static int uvd_v7_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	int i;
 
 	if (!amdgpu_sriov_vf(adev))
 		uvd_v7_0_stop(adev);
@@ -578,7 +579,8 @@ static int uvd_v7_0_hw_fini(void *handle)
 		DRM_DEBUG("For SRIOV client, shouldn't do anything.\n");
 	}
 
-	ring->ready = false;
+	for (i = 0; i < adev->uvd.num_uvd_inst; ++i)
+		adev->uvd.inst[i].ring.ready = false;
 
 	return 0;
 }
@@ -618,48 +620,51 @@ static void uvd_v7_0_mc_resume(struct amdgpu_device *adev)
 {
 	uint32_t size = AMDGPU_UVD_FIRMWARE_SIZE(adev);
 	uint32_t offset;
+	int i;
 
-	if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
-		WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
-			lower_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
-		WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
-			upper_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
-		offset = 0;
-	} else {
-		WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
-			lower_32_bits(adev->uvd.gpu_addr));
-		WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
-			upper_32_bits(adev->uvd.gpu_addr));
-		offset = size;
-	}
+	for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+		if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
+			WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
+				lower_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
+			WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
+				upper_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
+			offset = 0;
+		} else {
+			WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW,
+				lower_32_bits(adev->uvd.inst[i].gpu_addr));
+			WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH,
+				upper_32_bits(adev->uvd.inst[i].gpu_addr));
+			offset = size;
+		}
 
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_OFFSET0,
-				AMDGPU_UVD_FIRMWARE_OFFSET >> 3);
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_SIZE0, size);
-
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW,
-			lower_32_bits(adev->uvd.gpu_addr + offset));
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH,
-			upper_32_bits(adev->uvd.gpu_addr + offset));
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_OFFSET1, (1 << 21));
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_SIZE1, AMDGPU_UVD_HEAP_SIZE);
-
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW,
-			lower_32_bits(adev->uvd.gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH,
-			upper_32_bits(adev->uvd.gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_OFFSET2, (2 << 21));
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CACHE_SIZE2,
-			AMDGPU_UVD_STACK_SIZE + (AMDGPU_UVD_SESSION_SIZE * 40));
-
-	WREG32_SOC15(UVD, 0, mmUVD_UDEC_ADDR_CONFIG,
-			adev->gfx.config.gb_addr_config);
-	WREG32_SOC15(UVD, 0, mmUVD_UDEC_DB_ADDR_CONFIG,
-			adev->gfx.config.gb_addr_config);
-	WREG32_SOC15(UVD, 0, mmUVD_UDEC_DBW_ADDR_CONFIG,
-			adev->gfx.config.gb_addr_config);
-
-	WREG32_SOC15(UVD, 0, mmUVD_GP_SCRATCH4, adev->uvd.max_handles);
+		WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_OFFSET0,
+					AMDGPU_UVD_FIRMWARE_OFFSET >> 3);
+		WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_SIZE0, size);
+
+		WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW,
+				lower_32_bits(adev->uvd.inst[i].gpu_addr + offset));
+		WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH,
+				upper_32_bits(adev->uvd.inst[i].gpu_addr + offset));
+		WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_OFFSET1, (1 << 21));
+		WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_SIZE1, AMDGPU_UVD_HEAP_SIZE);
+
+		WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW,
+				lower_32_bits(adev->uvd.inst[i].gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
+		WREG32_SOC15(UVD, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH,
+				upper_32_bits(adev->uvd.inst[i].gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
+		WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_OFFSET2, (2 << 21));
+		WREG32_SOC15(UVD, i, mmUVD_VCPU_CACHE_SIZE2,
+				AMDGPU_UVD_STACK_SIZE + (AMDGPU_UVD_SESSION_SIZE * 40));
+
+		WREG32_SOC15(UVD, i, mmUVD_UDEC_ADDR_CONFIG,
+				adev->gfx.config.gb_addr_config);
+		WREG32_SOC15(UVD, i, mmUVD_UDEC_DB_ADDR_CONFIG,
+				adev->gfx.config.gb_addr_config);
+		WREG32_SOC15(UVD, i, mmUVD_UDEC_DBW_ADDR_CONFIG,
+				adev->gfx.config.gb_addr_config);
+
+		WREG32_SOC15(UVD, i, mmUVD_GP_SCRATCH4, adev->uvd.max_handles);
+	}
 }
 
 static int uvd_v7_0_mmsch_start(struct amdgpu_device *adev,
@@ -669,6 +674,7 @@ static int uvd_v7_0_mmsch_start(struct amdgpu_device *adev,
 	uint64_t addr = table->gpu_addr;
 	struct mmsch_v1_0_init_header *header = (struct mmsch_v1_0_init_header *)table->cpu_addr;
 	uint32_t size;
+	int i;
 
 	size = header->header_size + header->vce_table_size + header->uvd_table_size;
 
@@ -688,11 +694,12 @@ static int uvd_v7_0_mmsch_start(struct amdgpu_device *adev,
 	/* 4, set resp to zero */
 	WREG32_SOC15(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_RESP, 0);
 
-	WDOORBELL32(adev->uvd.ring_enc[0].doorbell_index, 0);
-	adev->wb.wb[adev->uvd.ring_enc[0].wptr_offs] = 0;
-	adev->uvd.ring_enc[0].wptr = 0;
-	adev->uvd.ring_enc[0].wptr_old = 0;
-
+	for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+		WDOORBELL32(adev->uvd.inst[i].ring_enc[0].doorbell_index, 0);
+		adev->wb.wb[adev->uvd.inst[i].ring_enc[0].wptr_offs] = 0;
+		adev->uvd.inst[i].ring_enc[0].wptr = 0;
+		adev->uvd.inst[i].ring_enc[0].wptr_old = 0;
+	}
 	/* 5, kick off the initialization and wait until VCE_MMSCH_VF_MAILBOX_RESP becomes non-zero */
 	WREG32_SOC15(VCE, 0, mmVCE_MMSCH_VF_MAILBOX_HOST, 0x10000001);
 
@@ -725,6 +732,7 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev)
 	struct mmsch_v1_0_cmd_end end = { {0} };
 	uint32_t *init_table = adev->virt.mm_table.cpu_addr;
 	struct mmsch_v1_0_init_header *header = (struct mmsch_v1_0_init_header *)init_table;
+	uint8_t i = 0;
 
 	direct_wt.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_WRITE;
 	direct_rd_mod_wt.cmd_header.command_type = MMSCH_COMMAND__DIRECT_REG_READ_MODIFY_WRITE;
@@ -742,120 +750,121 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev)
 
 		init_table += header->uvd_table_offset;
 
-		ring = &adev->uvd.ring;
-		ring->wptr = 0;
-		size = AMDGPU_GPU_PAGE_ALIGN(adev->uvd.fw->size + 4);
-
-		MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_STATUS),
-						   0xFFFFFFFF, 0x00000004);
-		/* mc resume*/
-		if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
-			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
-						    lower_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
-			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
-						    upper_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
-			offset = 0;
-		} else {
-			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
-						    lower_32_bits(adev->uvd.gpu_addr));
-			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
-						    upper_32_bits(adev->uvd.gpu_addr));
-			offset = size;
+		for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+			ring = &adev->uvd.inst[i].ring;
+			ring->wptr = 0;
+			size = AMDGPU_GPU_PAGE_ALIGN(adev->uvd.fw->size + 4);
+
+			MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_STATUS),
+							   0xFFFFFFFF, 0x00000004);
+			/* mc resume*/
+			if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) {
+				MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
+							    lower_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
+				MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
+							    upper_32_bits(adev->firmware.ucode[AMDGPU_UCODE_ID_UVD].mc_addr));
+				offset = 0;
+			} else {
+				MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW),
+							    lower_32_bits(adev->uvd.inst[i].gpu_addr));
+				MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH),
+							    upper_32_bits(adev->uvd.inst[i].gpu_addr));
+				offset = size;
+			}
+
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_OFFSET0),
+						    AMDGPU_UVD_FIRMWARE_OFFSET >> 3);
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_SIZE0), size);
+
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW),
+						    lower_32_bits(adev->uvd.inst[i].gpu_addr + offset));
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH),
+						    upper_32_bits(adev->uvd.inst[i].gpu_addr + offset));
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_OFFSET1), (1 << 21));
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_SIZE1), AMDGPU_UVD_HEAP_SIZE);
+
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW),
+						    lower_32_bits(adev->uvd.inst[i].gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH),
+						    upper_32_bits(adev->uvd.inst[i].gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_OFFSET2), (2 << 21));
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CACHE_SIZE2),
+						    AMDGPU_UVD_STACK_SIZE + (AMDGPU_UVD_SESSION_SIZE * 40));
+
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_GP_SCRATCH4), adev->uvd.max_handles);
+			/* mc resume end*/
+
+			/* disable clock gating */
+			MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_CGC_CTRL),
+							   ~UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK, 0);
+
+			/* disable interupt */
+			MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_MASTINT_EN),
+							   ~UVD_MASTINT_EN__VCPU_EN_MASK, 0);
+
+			/* stall UMC and register bus before resetting VCPU */
+			MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_CTRL2),
+							   ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK,
+							   UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+
+			/* put LMI, VCPU, RBC etc... into reset */
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_SOFT_RESET),
+						    (uint32_t)(UVD_SOFT_RESET__LMI_SOFT_RESET_MASK |
+							       UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK |
+							       UVD_SOFT_RESET__LBSI_SOFT_RESET_MASK |
+							       UVD_SOFT_RESET__RBC_SOFT_RESET_MASK |
+							       UVD_SOFT_RESET__CSM_SOFT_RESET_MASK |
+							       UVD_SOFT_RESET__CXW_SOFT_RESET_MASK |
+							       UVD_SOFT_RESET__TAP_SOFT_RESET_MASK |
+							       UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK));
+
+			/* initialize UVD memory controller */
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_CTRL),
+						    (uint32_t)((0x40 << UVD_LMI_CTRL__WRITE_CLEAN_TIMER__SHIFT) |
+							       UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK |
+							       UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK |
+							       UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK |
+							       UVD_LMI_CTRL__REQ_MODE_MASK |
+							       0x00100000L));
+
+			/* take all subblocks out of reset, except VCPU */
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_SOFT_RESET),
+						    UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
+
+			/* enable VCPU clock */
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_VCPU_CNTL),
+						    UVD_VCPU_CNTL__CLK_EN_MASK);
+
+			/* enable master interrupt */
+			MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_MASTINT_EN),
+							   ~(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK),
+							   (UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK));
+
+			/* clear the bit 4 of UVD_STATUS */
+			MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_STATUS),
+							   ~(2 << UVD_STATUS__VCPU_REPORT__SHIFT), 0);
+
+			/* force RBC into idle state */
+			size = order_base_2(ring->ring_size);
+			tmp = REG_SET_FIELD(0, UVD_RBC_RB_CNTL, RB_BUFSZ, size);
+			tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1);
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_RBC_RB_CNTL), tmp);
+
+			ring = &adev->uvd.inst[i].ring_enc[0];
+			ring->wptr = 0;
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_RB_BASE_LO), ring->gpu_addr);
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_RB_BASE_HI), upper_32_bits(ring->gpu_addr));
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_RB_SIZE), ring->ring_size / 4);
+
+			/* boot up the VCPU */
+			MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_SOFT_RESET), 0);
+
+			/* enable UMC */
+			MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_CTRL2),
+											   ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK, 0);
+
+			MMSCH_V1_0_INSERT_DIRECT_POLL(SOC15_REG_OFFSET(UVD, i, mmUVD_STATUS), 0x02, 0x02);
 		}
-
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_OFFSET0),
-					    AMDGPU_UVD_FIRMWARE_OFFSET >> 3);
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE0), size);
-
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_LOW),
-					    lower_32_bits(adev->uvd.gpu_addr + offset));
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE1_64BIT_BAR_HIGH),
-					    upper_32_bits(adev->uvd.gpu_addr + offset));
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_OFFSET1), (1 << 21));
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE1), AMDGPU_UVD_HEAP_SIZE);
-
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_LOW),
-					    lower_32_bits(adev->uvd.gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_VCPU_CACHE2_64BIT_BAR_HIGH),
-					    upper_32_bits(adev->uvd.gpu_addr + offset + AMDGPU_UVD_HEAP_SIZE));
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_OFFSET2), (2 << 21));
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CACHE_SIZE2),
-					    AMDGPU_UVD_STACK_SIZE + (AMDGPU_UVD_SESSION_SIZE * 40));
-
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_GP_SCRATCH4), adev->uvd.max_handles);
-		/* mc resume end*/
-
-		/* disable clock gating */
-		MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_CGC_CTRL),
-						   ~UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK, 0);
-
-		/* disable interupt */
-		MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MASTINT_EN),
-						   ~UVD_MASTINT_EN__VCPU_EN_MASK, 0);
-
-		/* stall UMC and register bus before resetting VCPU */
-		MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2),
-						   ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK,
-						   UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
-
-		/* put LMI, VCPU, RBC etc... into reset */
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET),
-					    (uint32_t)(UVD_SOFT_RESET__LMI_SOFT_RESET_MASK |
-						       UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK |
-						       UVD_SOFT_RESET__LBSI_SOFT_RESET_MASK |
-						       UVD_SOFT_RESET__RBC_SOFT_RESET_MASK |
-						       UVD_SOFT_RESET__CSM_SOFT_RESET_MASK |
-						       UVD_SOFT_RESET__CXW_SOFT_RESET_MASK |
-						       UVD_SOFT_RESET__TAP_SOFT_RESET_MASK |
-						       UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK));
-
-		/* initialize UVD memory controller */
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL),
-					    (uint32_t)((0x40 << UVD_LMI_CTRL__WRITE_CLEAN_TIMER__SHIFT) |
-						       UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK |
-						       UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK |
-						       UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK |
-						       UVD_LMI_CTRL__REQ_MODE_MASK |
-						       0x00100000L));
-
-		/* take all subblocks out of reset, except VCPU */
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET),
-					    UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
-
-		/* enable VCPU clock */
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_VCPU_CNTL),
-					    UVD_VCPU_CNTL__CLK_EN_MASK);
-
-		/* enable master interrupt */
-		MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_MASTINT_EN),
-						   ~(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK),
-						   (UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK));
-
-		/* clear the bit 4 of UVD_STATUS */
-		MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_STATUS),
-						   ~(2 << UVD_STATUS__VCPU_REPORT__SHIFT), 0);
-
-		/* force RBC into idle state */
-		size = order_base_2(ring->ring_size);
-		tmp = REG_SET_FIELD(0, UVD_RBC_RB_CNTL, RB_BUFSZ, size);
-		tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1);
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RBC_RB_CNTL), tmp);
-
-		ring = &adev->uvd.ring_enc[0];
-		ring->wptr = 0;
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_LO), ring->gpu_addr);
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_BASE_HI), upper_32_bits(ring->gpu_addr));
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_RB_SIZE), ring->ring_size / 4);
-
-		/* boot up the VCPU */
-		MMSCH_V1_0_INSERT_DIRECT_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET), 0);
-
-		/* enable UMC */
-		MMSCH_V1_0_INSERT_DIRECT_RD_MOD_WT(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2),
-										   ~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK, 0);
-
-		MMSCH_V1_0_INSERT_DIRECT_POLL(SOC15_REG_OFFSET(UVD, 0, mmUVD_STATUS), 0x02, 0x02);
-
 		/* add end packet */
 		memcpy((void *)init_table, &end, sizeof(struct mmsch_v1_0_cmd_end));
 		table_size += sizeof(struct mmsch_v1_0_cmd_end) / 4;
@@ -874,15 +883,17 @@ static int uvd_v7_0_sriov_start(struct amdgpu_device *adev)
  */
 static int uvd_v7_0_start(struct amdgpu_device *adev)
 {
-	struct amdgpu_ring *ring = &adev->uvd.ring;
+	struct amdgpu_ring *ring;
 	uint32_t rb_bufsz, tmp;
 	uint32_t lmi_swap_cntl;
 	uint32_t mp_swap_cntl;
-	int i, j, r;
+	int i, j, k, r;
 
-	/* disable DPG */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_POWER_STATUS), 0,
-			~UVD_POWER_STATUS__UVD_PG_MODE_MASK);
+	for (k = 0; k < adev->uvd.num_uvd_inst; ++k) {
+		/* disable DPG */
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_POWER_STATUS), 0,
+				~UVD_POWER_STATUS__UVD_PG_MODE_MASK);
+	}
 
 	/* disable byte swapping */
 	lmi_swap_cntl = 0;
@@ -890,157 +901,159 @@ static int uvd_v7_0_start(struct amdgpu_device *adev)
 
 	uvd_v7_0_mc_resume(adev);
 
-	/* disable clock gating */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_CGC_CTRL), 0,
-			~UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK);
-
-	/* disable interupt */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_MASTINT_EN), 0,
-			~UVD_MASTINT_EN__VCPU_EN_MASK);
-
-	/* stall UMC and register bus before resetting VCPU */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2),
-			UVD_LMI_CTRL2__STALL_ARB_UMC_MASK,
-			~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
-	mdelay(1);
-
-	/* put LMI, VCPU, RBC etc... into reset */
-	WREG32_SOC15(UVD, 0, mmUVD_SOFT_RESET,
-		UVD_SOFT_RESET__LMI_SOFT_RESET_MASK |
-		UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK |
-		UVD_SOFT_RESET__LBSI_SOFT_RESET_MASK |
-		UVD_SOFT_RESET__RBC_SOFT_RESET_MASK |
-		UVD_SOFT_RESET__CSM_SOFT_RESET_MASK |
-		UVD_SOFT_RESET__CXW_SOFT_RESET_MASK |
-		UVD_SOFT_RESET__TAP_SOFT_RESET_MASK |
-		UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK);
-	mdelay(5);
+	for (k = 0; k < adev->uvd.num_uvd_inst; ++k) {
+		ring = &adev->uvd.inst[k].ring;
+		/* disable clock gating */
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_CGC_CTRL), 0,
+				~UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK);
 
-	/* initialize UVD memory controller */
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_CTRL,
-		(0x40 << UVD_LMI_CTRL__WRITE_CLEAN_TIMER__SHIFT) |
-		UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK |
-		UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK |
-		UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK |
-		UVD_LMI_CTRL__REQ_MODE_MASK |
-		0x00100000L);
+		/* disable interupt */
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_MASTINT_EN), 0,
+				~UVD_MASTINT_EN__VCPU_EN_MASK);
+
+		/* stall UMC and register bus before resetting VCPU */
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_LMI_CTRL2),
+				UVD_LMI_CTRL2__STALL_ARB_UMC_MASK,
+				~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+		mdelay(1);
+
+		/* put LMI, VCPU, RBC etc... into reset */
+		WREG32_SOC15(UVD, k, mmUVD_SOFT_RESET,
+			UVD_SOFT_RESET__LMI_SOFT_RESET_MASK |
+			UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK |
+			UVD_SOFT_RESET__LBSI_SOFT_RESET_MASK |
+			UVD_SOFT_RESET__RBC_SOFT_RESET_MASK |
+			UVD_SOFT_RESET__CSM_SOFT_RESET_MASK |
+			UVD_SOFT_RESET__CXW_SOFT_RESET_MASK |
+			UVD_SOFT_RESET__TAP_SOFT_RESET_MASK |
+			UVD_SOFT_RESET__LMI_UMC_SOFT_RESET_MASK);
+		mdelay(5);
+
+		/* initialize UVD memory controller */
+		WREG32_SOC15(UVD, k, mmUVD_LMI_CTRL,
+			(0x40 << UVD_LMI_CTRL__WRITE_CLEAN_TIMER__SHIFT) |
+			UVD_LMI_CTRL__WRITE_CLEAN_TIMER_EN_MASK |
+			UVD_LMI_CTRL__DATA_COHERENCY_EN_MASK |
+			UVD_LMI_CTRL__VCPU_DATA_COHERENCY_EN_MASK |
+			UVD_LMI_CTRL__REQ_MODE_MASK |
+			0x00100000L);
 
 #ifdef __BIG_ENDIAN
-	/* swap (8 in 32) RB and IB */
-	lmi_swap_cntl = 0xa;
-	mp_swap_cntl = 0;
+		/* swap (8 in 32) RB and IB */
+		lmi_swap_cntl = 0xa;
+		mp_swap_cntl = 0;
 #endif
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_SWAP_CNTL, lmi_swap_cntl);
-	WREG32_SOC15(UVD, 0, mmUVD_MP_SWAP_CNTL, mp_swap_cntl);
-
-	WREG32_SOC15(UVD, 0, mmUVD_MPC_SET_MUXA0, 0x40c2040);
-	WREG32_SOC15(UVD, 0, mmUVD_MPC_SET_MUXA1, 0x0);
-	WREG32_SOC15(UVD, 0, mmUVD_MPC_SET_MUXB0, 0x40c2040);
-	WREG32_SOC15(UVD, 0, mmUVD_MPC_SET_MUXB1, 0x0);
-	WREG32_SOC15(UVD, 0, mmUVD_MPC_SET_ALU, 0);
-	WREG32_SOC15(UVD, 0, mmUVD_MPC_SET_MUX, 0x88);
-
-	/* take all subblocks out of reset, except VCPU */
-	WREG32_SOC15(UVD, 0, mmUVD_SOFT_RESET,
-			UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
-	mdelay(5);
+		WREG32_SOC15(UVD, k, mmUVD_LMI_SWAP_CNTL, lmi_swap_cntl);
+		WREG32_SOC15(UVD, k, mmUVD_MP_SWAP_CNTL, mp_swap_cntl);
 
-	/* enable VCPU clock */
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CNTL,
-			UVD_VCPU_CNTL__CLK_EN_MASK);
+		WREG32_SOC15(UVD, k, mmUVD_MPC_SET_MUXA0, 0x40c2040);
+		WREG32_SOC15(UVD, k, mmUVD_MPC_SET_MUXA1, 0x0);
+		WREG32_SOC15(UVD, k, mmUVD_MPC_SET_MUXB0, 0x40c2040);
+		WREG32_SOC15(UVD, k, mmUVD_MPC_SET_MUXB1, 0x0);
+		WREG32_SOC15(UVD, k, mmUVD_MPC_SET_ALU, 0);
+		WREG32_SOC15(UVD, k, mmUVD_MPC_SET_MUX, 0x88);
 
-	/* enable UMC */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2), 0,
-			~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+		/* take all subblocks out of reset, except VCPU */
+		WREG32_SOC15(UVD, k, mmUVD_SOFT_RESET,
+				UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
+		mdelay(5);
 
-	/* boot up the VCPU */
-	WREG32_SOC15(UVD, 0, mmUVD_SOFT_RESET, 0);
-	mdelay(10);
+		/* enable VCPU clock */
+		WREG32_SOC15(UVD, k, mmUVD_VCPU_CNTL,
+				UVD_VCPU_CNTL__CLK_EN_MASK);
+
+		/* enable UMC */
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_LMI_CTRL2), 0,
+				~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+
+		/* boot up the VCPU */
+		WREG32_SOC15(UVD, k, mmUVD_SOFT_RESET, 0);
+		mdelay(10);
 
-	for (i = 0; i < 10; ++i) {
-		uint32_t status;
+		for (i = 0; i < 10; ++i) {
+			uint32_t status;
 
-		for (j = 0; j < 100; ++j) {
-			status = RREG32_SOC15(UVD, 0, mmUVD_STATUS);
+			for (j = 0; j < 100; ++j) {
+				status = RREG32_SOC15(UVD, k, mmUVD_STATUS);
+				if (status & 2)
+					break;
+				mdelay(10);
+			}
+			r = 0;
 			if (status & 2)
 				break;
+
+			DRM_ERROR("UVD(%d) not responding, trying to reset the VCPU!!!\n", k);
+			WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_SOFT_RESET),
+					UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK,
+					~UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
+			mdelay(10);
+			WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_SOFT_RESET), 0,
+					~UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
 			mdelay(10);
+			r = -1;
 		}
-		r = 0;
-		if (status & 2)
-			break;
 
-		DRM_ERROR("UVD not responding, trying to reset the VCPU!!!\n");
-		WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET),
-				UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK,
-				~UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
-		mdelay(10);
-		WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_SOFT_RESET), 0,
-				~UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
-		mdelay(10);
-		r = -1;
-	}
-
-	if (r) {
-		DRM_ERROR("UVD not responding, giving up!!!\n");
-		return r;
-	}
-	/* enable master interrupt */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_MASTINT_EN),
-		(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK),
-		~(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK));
-
-	/* clear the bit 4 of UVD_STATUS */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_STATUS), 0,
-			~(2 << UVD_STATUS__VCPU_REPORT__SHIFT));
-
-	/* force RBC into idle state */
-	rb_bufsz = order_base_2(ring->ring_size);
-	tmp = REG_SET_FIELD(0, UVD_RBC_RB_CNTL, RB_BUFSZ, rb_bufsz);
-	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_BLKSZ, 1);
-	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1);
-	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_WPTR_POLL_EN, 0);
-	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_UPDATE, 1);
-	tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_RPTR_WR_EN, 1);
-	WREG32_SOC15(UVD, 0, mmUVD_RBC_RB_CNTL, tmp);
-
-	/* set the write pointer delay */
-	WREG32_SOC15(UVD, 0, mmUVD_RBC_RB_WPTR_CNTL, 0);
-
-	/* set the wb address */
-	WREG32_SOC15(UVD, 0, mmUVD_RBC_RB_RPTR_ADDR,
-			(upper_32_bits(ring->gpu_addr) >> 2));
-
-	/* programm the RB_BASE for ring buffer */
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW,
-			lower_32_bits(ring->gpu_addr));
-	WREG32_SOC15(UVD, 0, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH,
-			upper_32_bits(ring->gpu_addr));
-
-	/* Initialize the ring buffer's read and write pointers */
-	WREG32_SOC15(UVD, 0, mmUVD_RBC_RB_RPTR, 0);
-
-	ring->wptr = RREG32_SOC15(UVD, 0, mmUVD_RBC_RB_RPTR);
-	WREG32_SOC15(UVD, 0, mmUVD_RBC_RB_WPTR,
-			lower_32_bits(ring->wptr));
-
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_RBC_RB_CNTL), 0,
-			~UVD_RBC_RB_CNTL__RB_NO_FETCH_MASK);
-
-	ring = &adev->uvd.ring_enc[0];
-	WREG32_SOC15(UVD, 0, mmUVD_RB_RPTR, lower_32_bits(ring->wptr));
-	WREG32_SOC15(UVD, 0, mmUVD_RB_WPTR, lower_32_bits(ring->wptr));
-	WREG32_SOC15(UVD, 0, mmUVD_RB_BASE_LO, ring->gpu_addr);
-	WREG32_SOC15(UVD, 0, mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
-	WREG32_SOC15(UVD, 0, mmUVD_RB_SIZE, ring->ring_size / 4);
+		if (r) {
+			DRM_ERROR("UVD(%d) not responding, giving up!!!\n", k);
+			return r;
+		}
+		/* enable master interrupt */
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_MASTINT_EN),
+			(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK),
+			~(UVD_MASTINT_EN__VCPU_EN_MASK|UVD_MASTINT_EN__SYS_EN_MASK));
 
-	ring = &adev->uvd.ring_enc[1];
-	WREG32_SOC15(UVD, 0, mmUVD_RB_RPTR2, lower_32_bits(ring->wptr));
-	WREG32_SOC15(UVD, 0, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr));
-	WREG32_SOC15(UVD, 0, mmUVD_RB_BASE_LO2, ring->gpu_addr);
-	WREG32_SOC15(UVD, 0, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
-	WREG32_SOC15(UVD, 0, mmUVD_RB_SIZE2, ring->ring_size / 4);
+		/* clear the bit 4 of UVD_STATUS */
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_STATUS), 0,
+				~(2 << UVD_STATUS__VCPU_REPORT__SHIFT));
 
+		/* force RBC into idle state */
+		rb_bufsz = order_base_2(ring->ring_size);
+		tmp = REG_SET_FIELD(0, UVD_RBC_RB_CNTL, RB_BUFSZ, rb_bufsz);
+		tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_BLKSZ, 1);
+		tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_FETCH, 1);
+		tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_WPTR_POLL_EN, 0);
+		tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_NO_UPDATE, 1);
+		tmp = REG_SET_FIELD(tmp, UVD_RBC_RB_CNTL, RB_RPTR_WR_EN, 1);
+		WREG32_SOC15(UVD, k, mmUVD_RBC_RB_CNTL, tmp);
+
+		/* set the write pointer delay */
+		WREG32_SOC15(UVD, k, mmUVD_RBC_RB_WPTR_CNTL, 0);
+
+		/* set the wb address */
+		WREG32_SOC15(UVD, k, mmUVD_RBC_RB_RPTR_ADDR,
+				(upper_32_bits(ring->gpu_addr) >> 2));
+
+		/* programm the RB_BASE for ring buffer */
+		WREG32_SOC15(UVD, k, mmUVD_LMI_RBC_RB_64BIT_BAR_LOW,
+				lower_32_bits(ring->gpu_addr));
+		WREG32_SOC15(UVD, k, mmUVD_LMI_RBC_RB_64BIT_BAR_HIGH,
+				upper_32_bits(ring->gpu_addr));
+
+		/* Initialize the ring buffer's read and write pointers */
+		WREG32_SOC15(UVD, k, mmUVD_RBC_RB_RPTR, 0);
+
+		ring->wptr = RREG32_SOC15(UVD, k, mmUVD_RBC_RB_RPTR);
+		WREG32_SOC15(UVD, k, mmUVD_RBC_RB_WPTR,
+				lower_32_bits(ring->wptr));
+
+		WREG32_P(SOC15_REG_OFFSET(UVD, k, mmUVD_RBC_RB_CNTL), 0,
+				~UVD_RBC_RB_CNTL__RB_NO_FETCH_MASK);
+
+		ring = &adev->uvd.inst[k].ring_enc[0];
+		WREG32_SOC15(UVD, k, mmUVD_RB_RPTR, lower_32_bits(ring->wptr));
+		WREG32_SOC15(UVD, k, mmUVD_RB_WPTR, lower_32_bits(ring->wptr));
+		WREG32_SOC15(UVD, k, mmUVD_RB_BASE_LO, ring->gpu_addr);
+		WREG32_SOC15(UVD, k, mmUVD_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
+		WREG32_SOC15(UVD, k, mmUVD_RB_SIZE, ring->ring_size / 4);
+
+		ring = &adev->uvd.inst[k].ring_enc[1];
+		WREG32_SOC15(UVD, k, mmUVD_RB_RPTR2, lower_32_bits(ring->wptr));
+		WREG32_SOC15(UVD, k, mmUVD_RB_WPTR2, lower_32_bits(ring->wptr));
+		WREG32_SOC15(UVD, k, mmUVD_RB_BASE_LO2, ring->gpu_addr);
+		WREG32_SOC15(UVD, k, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
+		WREG32_SOC15(UVD, k, mmUVD_RB_SIZE2, ring->ring_size / 4);
+	}
 	return 0;
 }
 
@@ -1053,26 +1066,30 @@ static int uvd_v7_0_start(struct amdgpu_device *adev)
  */
 static void uvd_v7_0_stop(struct amdgpu_device *adev)
 {
-	/* force RBC into idle state */
-	WREG32_SOC15(UVD, 0, mmUVD_RBC_RB_CNTL, 0x11010101);
-
-	/* Stall UMC and register bus before resetting VCPU */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2),
-			UVD_LMI_CTRL2__STALL_ARB_UMC_MASK,
-			~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
-	mdelay(1);
-
-	/* put VCPU into reset */
-	WREG32_SOC15(UVD, 0, mmUVD_SOFT_RESET,
-			UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
-	mdelay(5);
+	uint8_t i = 0;
+
+	for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
+		/* force RBC into idle state */
+		WREG32_SOC15(UVD, i, mmUVD_RBC_RB_CNTL, 0x11010101);
 
-	/* disable VCPU clock */
-	WREG32_SOC15(UVD, 0, mmUVD_VCPU_CNTL, 0x0);
+		/* Stall UMC and register bus before resetting VCPU */
+		WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_CTRL2),
+				UVD_LMI_CTRL2__STALL_ARB_UMC_MASK,
+				~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+		mdelay(1);
 
-	/* Unstall UMC and register bus */
-	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2), 0,
-			~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+		/* put VCPU into reset */
+		WREG32_SOC15(UVD, i, mmUVD_SOFT_RESET,
+				UVD_SOFT_RESET__VCPU_SOFT_RESET_MASK);
+		mdelay(5);
+
+		/* disable VCPU clock */
+		WREG32_SOC15(UVD, i, mmUVD_VCPU_CNTL, 0x0);
+
+		/* Unstall UMC and register bus */
+		WREG32_P(SOC15_REG_OFFSET(UVD, i, mmUVD_LMI_CTRL2), 0,
+				~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
+	}
 }
 
 /**
@@ -1091,26 +1108,26 @@ static void uvd_v7_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq
 	WARN_ON(flags & AMDGPU_FENCE_FLAG_64BIT);
 
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_CONTEXT_ID), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_CONTEXT_ID), 0));
 	amdgpu_ring_write(ring, seq);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA0), 0));
 	amdgpu_ring_write(ring, addr & 0xffffffff);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA1), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA1), 0));
 	amdgpu_ring_write(ring, upper_32_bits(addr) & 0xff);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_CMD), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_CMD), 0));
 	amdgpu_ring_write(ring, 0);
 
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA0), 0));
 	amdgpu_ring_write(ring, 0);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA1), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA1), 0));
 	amdgpu_ring_write(ring, 0);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_CMD), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_CMD), 0));
 	amdgpu_ring_write(ring, 2);
 }
 
@@ -1136,6 +1153,16 @@ static void uvd_v7_0_enc_ring_emit_fence(struct amdgpu_ring *ring, u64 addr,
 }
 
 /**
+ * uvd_v7_0_ring_emit_hdp_flush - skip HDP flushing
+ *
+ * @ring: amdgpu_ring pointer
+ */
+static void uvd_v7_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
+{
+	/* The firmware doesn't seem to like touching registers at this point. */
+}
+
+/**
  * uvd_v7_0_ring_test_ring - register write test
  *
  * @ring: amdgpu_ring pointer
@@ -1149,30 +1176,30 @@ static int uvd_v7_0_ring_test_ring(struct amdgpu_ring *ring)
 	unsigned i;
 	int r;
 
-	WREG32_SOC15(UVD, 0, mmUVD_CONTEXT_ID, 0xCAFEDEAD);
+	WREG32_SOC15(UVD, ring->me, mmUVD_CONTEXT_ID, 0xCAFEDEAD);
 	r = amdgpu_ring_alloc(ring, 3);
 	if (r) {
-		DRM_ERROR("amdgpu: cp failed to lock ring %d (%d).\n",
-			  ring->idx, r);
+		DRM_ERROR("amdgpu: (%d)cp failed to lock ring %d (%d).\n",
+			  ring->me, ring->idx, r);
 		return r;
 	}
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_CONTEXT_ID), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_CONTEXT_ID), 0));
 	amdgpu_ring_write(ring, 0xDEADBEEF);
 	amdgpu_ring_commit(ring);
 	for (i = 0; i < adev->usec_timeout; i++) {
-		tmp = RREG32_SOC15(UVD, 0, mmUVD_CONTEXT_ID);
+		tmp = RREG32_SOC15(UVD, ring->me, mmUVD_CONTEXT_ID);
 		if (tmp == 0xDEADBEEF)
 			break;
 		DRM_UDELAY(1);
 	}
 
 	if (i < adev->usec_timeout) {
-		DRM_DEBUG("ring test on %d succeeded in %d usecs\n",
-			 ring->idx, i);
+		DRM_DEBUG("(%d)ring test on %d succeeded in %d usecs\n",
+			 ring->me, ring->idx, i);
 	} else {
-		DRM_ERROR("amdgpu: ring %d test failed (0x%08X)\n",
-			  ring->idx, tmp);
+		DRM_ERROR("(%d)amdgpu: ring %d test failed (0x%08X)\n",
+			  ring->me, ring->idx, tmp);
 		r = -EINVAL;
 	}
 	return r;
@@ -1193,17 +1220,17 @@ static void uvd_v7_0_ring_emit_ib(struct amdgpu_ring *ring,
 	struct amdgpu_device *adev = ring->adev;
 
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_RBC_IB_VMID), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_LMI_RBC_IB_VMID), 0));
 	amdgpu_ring_write(ring, vmid);
 
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_RBC_IB_64BIT_BAR_LOW), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_LMI_RBC_IB_64BIT_BAR_LOW), 0));
 	amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr));
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_RBC_IB_64BIT_BAR_HIGH), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_LMI_RBC_IB_64BIT_BAR_HIGH), 0));
 	amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr));
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_RBC_IB_SIZE), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_RBC_IB_SIZE), 0));
 	amdgpu_ring_write(ring, ib->length_dw);
 }
 
@@ -1231,13 +1258,13 @@ static void uvd_v7_0_ring_emit_wreg(struct amdgpu_ring *ring,
 	struct amdgpu_device *adev = ring->adev;
 
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA0), 0));
 	amdgpu_ring_write(ring, reg << 2);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA1), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA1), 0));
 	amdgpu_ring_write(ring, val);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_CMD), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_CMD), 0));
 	amdgpu_ring_write(ring, 8);
 }
 
@@ -1247,16 +1274,16 @@ static void uvd_v7_0_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
 	struct amdgpu_device *adev = ring->adev;
 
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA0), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA0), 0));
 	amdgpu_ring_write(ring, reg << 2);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_DATA1), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_DATA1), 0));
 	amdgpu_ring_write(ring, val);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GP_SCRATCH8), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GP_SCRATCH8), 0));
 	amdgpu_ring_write(ring, mask);
 	amdgpu_ring_write(ring,
-		PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_GPCOM_VCPU_CMD), 0));
+		PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_GPCOM_VCPU_CMD), 0));
 	amdgpu_ring_write(ring, 12);
 }
 
@@ -1277,12 +1304,15 @@ static void uvd_v7_0_ring_emit_vm_flush(struct amdgpu_ring *ring,
 
 static void uvd_v7_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
 {
-	int i;
 	struct amdgpu_device *adev = ring->adev;
+	int i;
 
-	for (i = 0; i < count; i++)
-		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_NO_OP), 0));
+	WARN_ON(ring->wptr % 2 || count % 2);
 
+	for (i = 0; i < count / 2; i++) {
+		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, ring->me, mmUVD_NO_OP), 0));
+		amdgpu_ring_write(ring, 0);
+	}
 }
 
 static void uvd_v7_0_enc_ring_insert_end(struct amdgpu_ring *ring)
@@ -1349,16 +1379,16 @@ static bool uvd_v7_0_check_soft_reset(void *handle)
 
 	if (REG_GET_FIELD(tmp, SRBM_STATUS, UVD_RQ_PENDING) ||
 	    REG_GET_FIELD(tmp, SRBM_STATUS, UVD_BUSY) ||
-	    (RREG32_SOC15(UVD, 0, mmUVD_STATUS) &
+	    (RREG32_SOC15(UVD, ring->me, mmUVD_STATUS) &
 		    AMDGPU_UVD_STATUS_BUSY_MASK))
 		srbm_soft_reset = REG_SET_FIELD(srbm_soft_reset,
 				SRBM_SOFT_RESET, SOFT_RESET_UVD, 1);
 
 	if (srbm_soft_reset) {
-		adev->uvd.srbm_soft_reset = srbm_soft_reset;
+		adev->uvd.inst[ring->me].srbm_soft_reset = srbm_soft_reset;
 		return true;
 	} else {
-		adev->uvd.srbm_soft_reset = 0;
+		adev->uvd.inst[ring->me].srbm_soft_reset = 0;
 		return false;
 	}
 }
@@ -1367,7 +1397,7 @@ static int uvd_v7_0_pre_soft_reset(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!adev->uvd.srbm_soft_reset)
+	if (!adev->uvd.inst[ring->me].srbm_soft_reset)
 		return 0;
 
 	uvd_v7_0_stop(adev);
@@ -1379,9 +1409,9 @@ static int uvd_v7_0_soft_reset(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	u32 srbm_soft_reset;
 
-	if (!adev->uvd.srbm_soft_reset)
+	if (!adev->uvd.inst[ring->me].srbm_soft_reset)
 		return 0;
-	srbm_soft_reset = adev->uvd.srbm_soft_reset;
+	srbm_soft_reset = adev->uvd.inst[ring->me].srbm_soft_reset;
 
 	if (srbm_soft_reset) {
 		u32 tmp;
@@ -1409,7 +1439,7 @@ static int uvd_v7_0_post_soft_reset(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
-	if (!adev->uvd.srbm_soft_reset)
+	if (!adev->uvd.inst[ring->me].srbm_soft_reset)
 		return 0;
 
 	mdelay(5);
@@ -1431,17 +1461,32 @@ static int uvd_v7_0_process_interrupt(struct amdgpu_device *adev,
 				      struct amdgpu_irq_src *source,
 				      struct amdgpu_iv_entry *entry)
 {
+	uint32_t ip_instance;
+
+	switch (entry->client_id) {
+	case SOC15_IH_CLIENTID_UVD:
+		ip_instance = 0;
+		break;
+	case SOC15_IH_CLIENTID_UVD1:
+		ip_instance = 1;
+		break;
+	default:
+		DRM_ERROR("Unhandled client id: %d\n", entry->client_id);
+		return 0;
+	}
+
 	DRM_DEBUG("IH: UVD TRAP\n");
+
 	switch (entry->src_id) {
 	case 124:
-		amdgpu_fence_process(&adev->uvd.ring);
+		amdgpu_fence_process(&adev->uvd.inst[ip_instance].ring);
 		break;
 	case 119:
-		amdgpu_fence_process(&adev->uvd.ring_enc[0]);
+		amdgpu_fence_process(&adev->uvd.inst[ip_instance].ring_enc[0]);
 		break;
 	case 120:
 		if (!amdgpu_sriov_vf(adev))
-			amdgpu_fence_process(&adev->uvd.ring_enc[1]);
+			amdgpu_fence_process(&adev->uvd.inst[ip_instance].ring_enc[1]);
 		break;
 	default:
 		DRM_ERROR("Unhandled interrupt: %d %d\n",
@@ -1457,9 +1502,9 @@ static void uvd_v7_0_set_sw_clock_gating(struct amdgpu_device *adev)
 {
 	uint32_t data, data1, data2, suvd_flags;
 
-	data = RREG32_SOC15(UVD, 0, mmUVD_CGC_CTRL);
-	data1 = RREG32_SOC15(UVD, 0, mmUVD_SUVD_CGC_GATE);
-	data2 = RREG32_SOC15(UVD, 0, mmUVD_SUVD_CGC_CTRL);
+	data = RREG32_SOC15(UVD, ring->me, mmUVD_CGC_CTRL);
+	data1 = RREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_GATE);
+	data2 = RREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_CTRL);
 
 	data &= ~(UVD_CGC_CTRL__CLK_OFF_DELAY_MASK |
 		  UVD_CGC_CTRL__CLK_GATE_DLY_TIMER_MASK);
@@ -1503,18 +1548,18 @@ static void uvd_v7_0_set_sw_clock_gating(struct amdgpu_device *adev)
 			UVD_SUVD_CGC_CTRL__SDB_MODE_MASK);
 	data1 |= suvd_flags;
 
-	WREG32_SOC15(UVD, 0, mmUVD_CGC_CTRL, data);
-	WREG32_SOC15(UVD, 0, mmUVD_CGC_GATE, 0);
-	WREG32_SOC15(UVD, 0, mmUVD_SUVD_CGC_GATE, data1);
-	WREG32_SOC15(UVD, 0, mmUVD_SUVD_CGC_CTRL, data2);
+	WREG32_SOC15(UVD, ring->me, mmUVD_CGC_CTRL, data);
+	WREG32_SOC15(UVD, ring->me, mmUVD_CGC_GATE, 0);
+	WREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_GATE, data1);
+	WREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_CTRL, data2);
 }
 
 static void uvd_v7_0_set_hw_clock_gating(struct amdgpu_device *adev)
 {
 	uint32_t data, data1, cgc_flags, suvd_flags;
 
-	data = RREG32_SOC15(UVD, 0, mmUVD_CGC_GATE);
-	data1 = RREG32_SOC15(UVD, 0, mmUVD_SUVD_CGC_GATE);
+	data = RREG32_SOC15(UVD, ring->me, mmUVD_CGC_GATE);
+	data1 = RREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_GATE);
 
 	cgc_flags = UVD_CGC_GATE__SYS_MASK |
 		UVD_CGC_GATE__UDEC_MASK |
@@ -1546,8 +1591,8 @@ static void uvd_v7_0_set_hw_clock_gating(struct amdgpu_device *adev)
 	data |= cgc_flags;
 	data1 |= suvd_flags;
 
-	WREG32_SOC15(UVD, 0, mmUVD_CGC_GATE, data);
-	WREG32_SOC15(UVD, 0, mmUVD_SUVD_CGC_GATE, data1);
+	WREG32_SOC15(UVD, ring->me, mmUVD_CGC_GATE, data);
+	WREG32_SOC15(UVD, ring->me, mmUVD_SUVD_CGC_GATE, data1);
 }
 
 static void uvd_v7_0_set_bypass_mode(struct amdgpu_device *adev, bool enable)
@@ -1606,7 +1651,7 @@ static int uvd_v7_0_set_powergating_state(void *handle,
 	if (!(adev->pg_flags & AMD_PG_SUPPORT_UVD))
 		return 0;
 
-	WREG32_SOC15(UVD, 0, mmUVD_POWER_STATUS, UVD_POWER_STATUS__UVD_PG_EN_MASK);
+	WREG32_SOC15(UVD, ring->me, mmUVD_POWER_STATUS, UVD_POWER_STATUS__UVD_PG_EN_MASK);
 
 	if (state == AMD_PG_STATE_GATE) {
 		uvd_v7_0_stop(adev);
@@ -1647,14 +1692,13 @@ const struct amd_ip_funcs uvd_v7_0_ip_funcs = {
 static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = {
 	.type = AMDGPU_RING_TYPE_UVD,
 	.align_mask = 0xf,
-	.nop = PACKET0(0x81ff, 0),
 	.support_64bit_ptrs = false,
 	.vmhub = AMDGPU_MMHUB,
 	.get_rptr = uvd_v7_0_ring_get_rptr,
 	.get_wptr = uvd_v7_0_ring_get_wptr,
 	.set_wptr = uvd_v7_0_ring_set_wptr,
 	.emit_frame_size =
-		6 + 6 + /* hdp flush / invalidate */
+		6 + /* hdp invalidate */
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
 		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +
 		8 + /* uvd_v7_0_ring_emit_vm_flush */
@@ -1663,6 +1707,7 @@ static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = {
 	.emit_ib = uvd_v7_0_ring_emit_ib,
 	.emit_fence = uvd_v7_0_ring_emit_fence,
 	.emit_vm_flush = uvd_v7_0_ring_emit_vm_flush,
+	.emit_hdp_flush = uvd_v7_0_ring_emit_hdp_flush,
 	.test_ring = uvd_v7_0_ring_test_ring,
 	.test_ib = amdgpu_uvd_ring_test_ib,
 	.insert_nop = uvd_v7_0_ring_insert_nop,
@@ -1671,6 +1716,7 @@ static const struct amdgpu_ring_funcs uvd_v7_0_ring_vm_funcs = {
 	.end_use = amdgpu_uvd_ring_end_use,
 	.emit_wreg = uvd_v7_0_ring_emit_wreg,
 	.emit_reg_wait = uvd_v7_0_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
 };
 
 static const struct amdgpu_ring_funcs uvd_v7_0_enc_ring_vm_funcs = {
@@ -1702,22 +1748,32 @@ static const struct amdgpu_ring_funcs uvd_v7_0_enc_ring_vm_funcs = {
 	.end_use = amdgpu_uvd_ring_end_use,
 	.emit_wreg = uvd_v7_0_enc_ring_emit_wreg,
 	.emit_reg_wait = uvd_v7_0_enc_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
 };
 
 static void uvd_v7_0_set_ring_funcs(struct amdgpu_device *adev)
 {
-	adev->uvd.ring.funcs = &uvd_v7_0_ring_vm_funcs;
-	DRM_INFO("UVD is enabled in VM mode\n");
+	int i;
+
+	for (i = 0; i < adev->uvd.num_uvd_inst; i++) {
+		adev->uvd.inst[i].ring.funcs = &uvd_v7_0_ring_vm_funcs;
+		adev->uvd.inst[i].ring.me = i;
+		DRM_INFO("UVD(%d) is enabled in VM mode\n", i);
+	}
 }
 
 static void uvd_v7_0_set_enc_ring_funcs(struct amdgpu_device *adev)
 {
-	int i;
+	int i, j;
 
-	for (i = 0; i < adev->uvd.num_enc_rings; ++i)
-		adev->uvd.ring_enc[i].funcs = &uvd_v7_0_enc_ring_vm_funcs;
+	for (j = 0; j < adev->uvd.num_uvd_inst; j++) {
+		for (i = 0; i < adev->uvd.num_enc_rings; ++i) {
+			adev->uvd.inst[j].ring_enc[i].funcs = &uvd_v7_0_enc_ring_vm_funcs;
+			adev->uvd.inst[j].ring_enc[i].me = j;
+		}
 
-	DRM_INFO("UVD ENC is enabled in VM mode\n");
+		DRM_INFO("UVD(%d) ENC is enabled in VM mode\n", j);
+	}
 }
 
 static const struct amdgpu_irq_src_funcs uvd_v7_0_irq_funcs = {
@@ -1727,8 +1783,12 @@ static const struct amdgpu_irq_src_funcs uvd_v7_0_irq_funcs = {
 
 static void uvd_v7_0_set_irq_funcs(struct amdgpu_device *adev)
 {
-	adev->uvd.irq.num_types = adev->uvd.num_enc_rings + 1;
-	adev->uvd.irq.funcs = &uvd_v7_0_irq_funcs;
+	int i;
+
+	for (i = 0; i < adev->uvd.num_uvd_inst; i++) {
+		adev->uvd.inst[i].irq.num_types = adev->uvd.num_enc_rings + 1;
+		adev->uvd.inst[i].irq.funcs = &uvd_v7_0_irq_funcs;
+	}
 }
 
 const struct amdgpu_ip_block_version uvd_v7_0_ip_block =
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
index 47f70827195b..d48e877b682e 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v2_0.c
@@ -56,7 +56,7 @@ static uint64_t vce_v2_0_ring_get_rptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	if (ring == &adev->vce.ring[0])
+	if (ring->me == 0)
 		return RREG32(mmVCE_RB_RPTR);
 	else
 		return RREG32(mmVCE_RB_RPTR2);
@@ -73,7 +73,7 @@ static uint64_t vce_v2_0_ring_get_wptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	if (ring == &adev->vce.ring[0])
+	if (ring->me == 0)
 		return RREG32(mmVCE_RB_WPTR);
 	else
 		return RREG32(mmVCE_RB_WPTR2);
@@ -90,7 +90,7 @@ static void vce_v2_0_ring_set_wptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	if (ring == &adev->vce.ring[0])
+	if (ring->me == 0)
 		WREG32(mmVCE_RB_WPTR, lower_32_bits(ring->wptr));
 	else
 		WREG32(mmVCE_RB_WPTR2, lower_32_bits(ring->wptr));
@@ -627,8 +627,10 @@ static void vce_v2_0_set_ring_funcs(struct amdgpu_device *adev)
 {
 	int i;
 
-	for (i = 0; i < adev->vce.num_rings; i++)
+	for (i = 0; i < adev->vce.num_rings; i++) {
 		adev->vce.ring[i].funcs = &vce_v2_0_ring_funcs;
+		adev->vce.ring[i].me = i;
+	}
 }
 
 static const struct amdgpu_irq_src_funcs vce_v2_0_irq_funcs = {
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
index 428d1928e44e..cc6ce6cc03f4 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v3_0.c
@@ -39,6 +39,7 @@
 #include "smu/smu_7_1_2_sh_mask.h"
 #include "gca/gfx_8_0_d.h"
 #include "gca/gfx_8_0_sh_mask.h"
+#include "ivsrcid/ivsrcid_vislands30.h"
 
 
 #define GRBM_GFX_INDEX__VCE_INSTANCE__SHIFT	0x04
@@ -86,9 +87,9 @@ static uint64_t vce_v3_0_ring_get_rptr(struct amdgpu_ring *ring)
 	else if (adev->vce.harvest_config == AMDGPU_VCE_HARVEST_VCE0)
 		WREG32(mmGRBM_GFX_INDEX, GET_VCE_INSTANCE(1));
 
-	if (ring == &adev->vce.ring[0])
+	if (ring->me == 0)
 		v = RREG32(mmVCE_RB_RPTR);
-	else if (ring == &adev->vce.ring[1])
+	else if (ring->me == 1)
 		v = RREG32(mmVCE_RB_RPTR2);
 	else
 		v = RREG32(mmVCE_RB_RPTR3);
@@ -118,9 +119,9 @@ static uint64_t vce_v3_0_ring_get_wptr(struct amdgpu_ring *ring)
 	else if (adev->vce.harvest_config == AMDGPU_VCE_HARVEST_VCE0)
 		WREG32(mmGRBM_GFX_INDEX, GET_VCE_INSTANCE(1));
 
-	if (ring == &adev->vce.ring[0])
+	if (ring->me == 0)
 		v = RREG32(mmVCE_RB_WPTR);
-	else if (ring == &adev->vce.ring[1])
+	else if (ring->me == 1)
 		v = RREG32(mmVCE_RB_WPTR2);
 	else
 		v = RREG32(mmVCE_RB_WPTR3);
@@ -149,9 +150,9 @@ static void vce_v3_0_ring_set_wptr(struct amdgpu_ring *ring)
 	else if (adev->vce.harvest_config == AMDGPU_VCE_HARVEST_VCE0)
 		WREG32(mmGRBM_GFX_INDEX, GET_VCE_INSTANCE(1));
 
-	if (ring == &adev->vce.ring[0])
+	if (ring->me == 0)
 		WREG32(mmVCE_RB_WPTR, lower_32_bits(ring->wptr));
-	else if (ring == &adev->vce.ring[1])
+	else if (ring->me == 1)
 		WREG32(mmVCE_RB_WPTR2, lower_32_bits(ring->wptr));
 	else
 		WREG32(mmVCE_RB_WPTR3, lower_32_bits(ring->wptr));
@@ -388,7 +389,8 @@ static unsigned vce_v3_0_get_harvest_config(struct amdgpu_device *adev)
 	default:
 		if ((adev->asic_type == CHIP_POLARIS10) ||
 		    (adev->asic_type == CHIP_POLARIS11) ||
-		    (adev->asic_type == CHIP_POLARIS12))
+		    (adev->asic_type == CHIP_POLARIS12) ||
+		    (adev->asic_type == CHIP_VEGAM))
 			return AMDGPU_VCE_HARVEST_VCE1;
 
 		return 0;
@@ -421,7 +423,7 @@ static int vce_v3_0_sw_init(void *handle)
 	int r, i;
 
 	/* VCE */
-	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 167, &adev->vce.irq);
+	r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, VISLANDS30_IV_SRCID_VCE_TRAP, &adev->vce.irq);
 	if (r)
 		return r;
 
@@ -467,8 +469,8 @@ static int vce_v3_0_hw_init(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	vce_v3_0_override_vce_clock_gating(adev, true);
-	if (!(adev->flags & AMD_IS_APU))
-		amdgpu_asic_set_vce_clocks(adev, 10000, 10000);
+
+	amdgpu_asic_set_vce_clocks(adev, 10000, 10000);
 
 	for (i = 0; i < adev->vce.num_rings; i++)
 		adev->vce.ring[i].ready = false;
@@ -899,7 +901,7 @@ static const struct amdgpu_ring_funcs vce_v3_0_ring_phys_funcs = {
 	.emit_frame_size =
 		4 + /* vce_v3_0_emit_pipeline_sync */
 		6, /* amdgpu_vce_ring_emit_fence x1 no user fence */
-	.emit_ib_size = 5, /* vce_v3_0_ring_emit_ib */
+	.emit_ib_size = 4, /* amdgpu_vce_ring_emit_ib */
 	.emit_ib = amdgpu_vce_ring_emit_ib,
 	.emit_fence = amdgpu_vce_ring_emit_fence,
 	.test_ring = amdgpu_vce_ring_test_ring,
@@ -923,7 +925,7 @@ static const struct amdgpu_ring_funcs vce_v3_0_ring_vm_funcs = {
 		6 + /* vce_v3_0_emit_vm_flush */
 		4 + /* vce_v3_0_emit_pipeline_sync */
 		6 + 6, /* amdgpu_vce_ring_emit_fence x2 vm fence */
-	.emit_ib_size = 4, /* amdgpu_vce_ring_emit_ib */
+	.emit_ib_size = 5, /* vce_v3_0_ring_emit_ib */
 	.emit_ib = vce_v3_0_ring_emit_ib,
 	.emit_vm_flush = vce_v3_0_emit_vm_flush,
 	.emit_pipeline_sync = vce_v3_0_emit_pipeline_sync,
@@ -941,12 +943,16 @@ static void vce_v3_0_set_ring_funcs(struct amdgpu_device *adev)
 	int i;
 
 	if (adev->asic_type >= CHIP_STONEY) {
-		for (i = 0; i < adev->vce.num_rings; i++)
+		for (i = 0; i < adev->vce.num_rings; i++) {
 			adev->vce.ring[i].funcs = &vce_v3_0_ring_vm_funcs;
+			adev->vce.ring[i].me = i;
+		}
 		DRM_INFO("VCE enabled in VM mode\n");
 	} else {
-		for (i = 0; i < adev->vce.num_rings; i++)
+		for (i = 0; i < adev->vce.num_rings; i++) {
 			adev->vce.ring[i].funcs = &vce_v3_0_ring_phys_funcs;
+			adev->vce.ring[i].me = i;
+		}
 		DRM_INFO("VCE enabled in physical mode\n");
 	}
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
index 73fd48d6c756..65f8860169e9 100644
--- a/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vce_v4_0.c
@@ -39,6 +39,8 @@
 #include "mmhub/mmhub_1_0_offset.h"
 #include "mmhub/mmhub_1_0_sh_mask.h"
 
+#include "ivsrcid/vce/irqsrcs_vce_4_0.h"
+
 #define VCE_STATUS_VCPU_REPORT_FW_LOADED_MASK	0x02
 
 #define VCE_V4_0_FW_SIZE	(384 * 1024)
@@ -60,9 +62,9 @@ static uint64_t vce_v4_0_ring_get_rptr(struct amdgpu_ring *ring)
 {
 	struct amdgpu_device *adev = ring->adev;
 
-	if (ring == &adev->vce.ring[0])
+	if (ring->me == 0)
 		return RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_RPTR));
-	else if (ring == &adev->vce.ring[1])
+	else if (ring->me == 1)
 		return RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_RPTR2));
 	else
 		return RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_RPTR3));
@@ -82,9 +84,9 @@ static uint64_t vce_v4_0_ring_get_wptr(struct amdgpu_ring *ring)
 	if (ring->use_doorbell)
 		return adev->wb.wb[ring->wptr_offs];
 
-	if (ring == &adev->vce.ring[0])
+	if (ring->me == 0)
 		return RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_WPTR));
-	else if (ring == &adev->vce.ring[1])
+	else if (ring->me == 1)
 		return RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_WPTR2));
 	else
 		return RREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_WPTR3));
@@ -108,10 +110,10 @@ static void vce_v4_0_ring_set_wptr(struct amdgpu_ring *ring)
 		return;
 	}
 
-	if (ring == &adev->vce.ring[0])
+	if (ring->me == 0)
 		WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_WPTR),
 			lower_32_bits(ring->wptr));
-	else if (ring == &adev->vce.ring[1])
+	else if (ring->me == 1)
 		WREG32(SOC15_REG_OFFSET(VCE, 0, mmVCE_RB_WPTR2),
 			lower_32_bits(ring->wptr));
 	else
@@ -1081,14 +1083,17 @@ static const struct amdgpu_ring_funcs vce_v4_0_ring_vm_funcs = {
 	.end_use = amdgpu_vce_ring_end_use,
 	.emit_wreg = vce_v4_0_emit_wreg,
 	.emit_reg_wait = vce_v4_0_emit_reg_wait,
+	.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
 };
 
 static void vce_v4_0_set_ring_funcs(struct amdgpu_device *adev)
 {
 	int i;
 
-	for (i = 0; i < adev->vce.num_rings; i++)
+	for (i = 0; i < adev->vce.num_rings; i++) {
 		adev->vce.ring[i].funcs = &vce_v4_0_ring_vm_funcs;
+		adev->vce.ring[i].me = i;
+	}
 	DRM_INFO("VCE enabled in VM mode\n");
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
index 8c132673bc79..2ce91a748c40 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c
@@ -35,11 +35,14 @@
 #include "mmhub/mmhub_9_1_offset.h"
 #include "mmhub/mmhub_9_1_sh_mask.h"
 
-static int vcn_v1_0_start(struct amdgpu_device *adev);
+#include "ivsrcid/vcn/irqsrcs_vcn_1_0.h"
+
 static int vcn_v1_0_stop(struct amdgpu_device *adev);
 static void vcn_v1_0_set_dec_ring_funcs(struct amdgpu_device *adev);
 static void vcn_v1_0_set_enc_ring_funcs(struct amdgpu_device *adev);
+static void vcn_v1_0_set_jpeg_ring_funcs(struct amdgpu_device *adev);
 static void vcn_v1_0_set_irq_funcs(struct amdgpu_device *adev);
+static void vcn_v1_0_jpeg_ring_set_patch_ring(struct amdgpu_ring *ring, uint32_t ptr);
 
 /**
  * vcn_v1_0_early_init - set function pointers
@@ -56,6 +59,7 @@ static int vcn_v1_0_early_init(void *handle)
 
 	vcn_v1_0_set_dec_ring_funcs(adev);
 	vcn_v1_0_set_enc_ring_funcs(adev);
+	vcn_v1_0_set_jpeg_ring_funcs(adev);
 	vcn_v1_0_set_irq_funcs(adev);
 
 	return 0;
@@ -75,18 +79,23 @@ static int vcn_v1_0_sw_init(void *handle)
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 
 	/* VCN DEC TRAP */
-	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VCN, 124, &adev->vcn.irq);
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VCN, VCN_1_0__SRCID__UVD_SYSTEM_MESSAGE_INTERRUPT, &adev->vcn.irq);
 	if (r)
 		return r;
 
 	/* VCN ENC TRAP */
 	for (i = 0; i < adev->vcn.num_enc_rings; ++i) {
-		r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VCN, i + 119,
+		r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VCN, i + VCN_1_0__SRCID__UVD_ENC_GENERAL_PURPOSE,
 					&adev->vcn.irq);
 		if (r)
 			return r;
 	}
 
+	/* VCN JPEG TRAP */
+	r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_VCN, 126, &adev->vcn.irq);
+	if (r)
+		return r;
+
 	r = amdgpu_vcn_sw_init(adev);
 	if (r)
 		return r;
@@ -109,6 +118,12 @@ static int vcn_v1_0_sw_init(void *handle)
 			return r;
 	}
 
+	ring = &adev->vcn.ring_jpeg;
+	sprintf(ring->name, "vcn_jpeg");
+	r = amdgpu_ring_init(adev, ring, 512, &adev->vcn.irq, 0);
+	if (r)
+		return r;
+
 	return r;
 }
 
@@ -146,10 +161,6 @@ static int vcn_v1_0_hw_init(void *handle)
 	struct amdgpu_ring *ring = &adev->vcn.ring_dec;
 	int i, r;
 
-	r = vcn_v1_0_start(adev);
-	if (r)
-		goto done;
-
 	ring->ready = true;
 	r = amdgpu_ring_test_ring(ring);
 	if (r) {
@@ -167,6 +178,14 @@ static int vcn_v1_0_hw_init(void *handle)
 		}
 	}
 
+	ring = &adev->vcn.ring_jpeg;
+	ring->ready = true;
+	r = amdgpu_ring_test_ring(ring);
+	if (r) {
+		ring->ready = false;
+		goto done;
+	}
+
 done:
 	if (!r)
 		DRM_INFO("VCN decode and encode initialized successfully.\n");
@@ -185,11 +204,9 @@ static int vcn_v1_0_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
 	struct amdgpu_ring *ring = &adev->vcn.ring_dec;
-	int r;
 
-	r = vcn_v1_0_stop(adev);
-	if (r)
-		return r;
+	if (RREG32_SOC15(VCN, 0, mmUVD_STATUS))
+		vcn_v1_0_stop(adev);
 
 	ring->ready = false;
 
@@ -288,14 +305,14 @@ static void vcn_v1_0_mc_resume(struct amdgpu_device *adev)
  *
  * Disable clock gating for VCN block
  */
-static void vcn_v1_0_disable_clock_gating(struct amdgpu_device *adev, bool sw)
+static void vcn_v1_0_disable_clock_gating(struct amdgpu_device *adev)
 {
 	uint32_t data;
 
 	/* JPEG disable CGC */
 	data = RREG32_SOC15(VCN, 0, mmJPEG_CGC_CTRL);
 
-	if (sw)
+	if (adev->cg_flags & AMD_CG_SUPPORT_VCN_MGCG)
 		data |= 1 << JPEG_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
 	else
 		data &= ~JPEG_CGC_CTRL__DYN_CLOCK_MODE_MASK;
@@ -310,7 +327,7 @@ static void vcn_v1_0_disable_clock_gating(struct amdgpu_device *adev, bool sw)
 
 	/* UVD disable CGC */
 	data = RREG32_SOC15(VCN, 0, mmUVD_CGC_CTRL);
-	if (sw)
+	if (adev->cg_flags & AMD_CG_SUPPORT_VCN_MGCG)
 		data |= 1 << UVD_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
 	else
 		data &= ~ UVD_CGC_CTRL__DYN_CLOCK_MODE_MASK;
@@ -415,13 +432,13 @@ static void vcn_v1_0_disable_clock_gating(struct amdgpu_device *adev, bool sw)
  *
  * Enable clock gating for VCN block
  */
-static void vcn_v1_0_enable_clock_gating(struct amdgpu_device *adev, bool sw)
+static void vcn_v1_0_enable_clock_gating(struct amdgpu_device *adev)
 {
 	uint32_t data = 0;
 
 	/* enable JPEG CGC */
 	data = RREG32_SOC15(VCN, 0, mmJPEG_CGC_CTRL);
-	if (sw)
+	if (adev->cg_flags & AMD_CG_SUPPORT_VCN_MGCG)
 		data |= 1 << JPEG_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
 	else
 		data |= 0 << JPEG_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
@@ -435,7 +452,7 @@ static void vcn_v1_0_enable_clock_gating(struct amdgpu_device *adev, bool sw)
 
 	/* enable UVD CGC */
 	data = RREG32_SOC15(VCN, 0, mmUVD_CGC_CTRL);
-	if (sw)
+	if (adev->cg_flags & AMD_CG_SUPPORT_VCN_MGCG)
 		data |= 1 << UVD_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
 	else
 		data |= 0 << UVD_CGC_CTRL__DYN_CLOCK_MODE__SHIFT;
@@ -480,6 +497,94 @@ static void vcn_v1_0_enable_clock_gating(struct amdgpu_device *adev, bool sw)
 	WREG32_SOC15(VCN, 0, mmUVD_SUVD_CGC_CTRL, data);
 }
 
+static void vcn_1_0_disable_static_power_gating(struct amdgpu_device *adev)
+{
+	uint32_t data = 0;
+	int ret;
+
+	if (adev->pg_flags & AMD_PG_SUPPORT_VCN) {
+		data = (1 << UVD_PGFSM_CONFIG__UVDM_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDU_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDF_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDC_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDB_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDIL_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDIR_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDTD_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDTE_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDE_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDW_PWR_CONFIG__SHIFT);
+
+		WREG32_SOC15(VCN, 0, mmUVD_PGFSM_CONFIG, data);
+		SOC15_WAIT_ON_RREG(VCN, 0, mmUVD_PGFSM_STATUS, UVD_PGFSM_STATUS__UVDM_UVDU_PWR_ON, 0xFFFFFF, ret);
+	} else {
+		data = (1 << UVD_PGFSM_CONFIG__UVDM_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDU_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDF_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDC_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDB_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDIL_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDIR_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDTD_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDTE_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDE_PWR_CONFIG__SHIFT
+			| 1 << UVD_PGFSM_CONFIG__UVDW_PWR_CONFIG__SHIFT);
+		WREG32_SOC15(VCN, 0, mmUVD_PGFSM_CONFIG, data);
+		SOC15_WAIT_ON_RREG(VCN, 0, mmUVD_PGFSM_STATUS, 0,  0xFFFFFFFF, ret);
+	}
+
+	/* polling UVD_PGFSM_STATUS to confirm UVDM_PWR_STATUS , UVDU_PWR_STATUS are 0 (power on) */
+
+	data = RREG32_SOC15(VCN, 0, mmUVD_POWER_STATUS);
+	data &= ~0x103;
+	if (adev->pg_flags & AMD_PG_SUPPORT_VCN)
+		data |= UVD_PGFSM_CONFIG__UVDM_UVDU_PWR_ON | UVD_POWER_STATUS__UVD_PG_EN_MASK;
+
+	WREG32_SOC15(VCN, 0, mmUVD_POWER_STATUS, data);
+}
+
+static void vcn_1_0_enable_static_power_gating(struct amdgpu_device *adev)
+{
+	uint32_t data = 0;
+	int ret;
+
+	if (adev->pg_flags & AMD_PG_SUPPORT_VCN) {
+		/* Before power off, this indicator has to be turned on */
+		data = RREG32_SOC15(VCN, 0, mmUVD_POWER_STATUS);
+		data &= ~UVD_POWER_STATUS__UVD_POWER_STATUS_MASK;
+		data |= UVD_POWER_STATUS__UVD_POWER_STATUS_TILES_OFF;
+		WREG32_SOC15(VCN, 0, mmUVD_POWER_STATUS, data);
+
+
+		data = (2 << UVD_PGFSM_CONFIG__UVDM_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDU_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDF_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDC_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDB_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDIL_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDIR_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDTD_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDTE_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDE_PWR_CONFIG__SHIFT
+			| 2 << UVD_PGFSM_CONFIG__UVDW_PWR_CONFIG__SHIFT);
+
+		WREG32_SOC15(VCN, 0, mmUVD_PGFSM_CONFIG, data);
+
+		data = (2 << UVD_PGFSM_STATUS__UVDM_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDU_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDF_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDC_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDB_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDIL_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDIR_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDTD_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDTE_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDE_PWR_STATUS__SHIFT
+			| 2 << UVD_PGFSM_STATUS__UVDW_PWR_STATUS__SHIFT);
+		SOC15_WAIT_ON_RREG(VCN, 0, mmUVD_PGFSM_STATUS, data, 0xFFFFFFFF, ret);
+	}
+}
+
 /**
  * vcn_v1_0_start - start VCN block
  *
@@ -497,10 +602,11 @@ static int vcn_v1_0_start(struct amdgpu_device *adev)
 	/* disable byte swapping */
 	lmi_swap_cntl = 0;
 
-	vcn_v1_0_mc_resume(adev);
-
+	vcn_1_0_disable_static_power_gating(adev);
 	/* disable clock gating */
-	vcn_v1_0_disable_clock_gating(adev, true);
+	vcn_v1_0_disable_clock_gating(adev);
+
+	vcn_v1_0_mc_resume(adev);
 
 	/* disable interupt */
 	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_MASTINT_EN), 0,
@@ -647,6 +753,22 @@ static int vcn_v1_0_start(struct amdgpu_device *adev)
 	WREG32_SOC15(UVD, 0, mmUVD_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
 	WREG32_SOC15(UVD, 0, mmUVD_RB_SIZE2, ring->ring_size / 4);
 
+	ring = &adev->vcn.ring_jpeg;
+	WREG32_SOC15(UVD, 0, mmUVD_LMI_JRBC_RB_VMID, 0);
+	WREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_CNTL, (0x00000001L | 0x00000002L));
+	WREG32_SOC15(UVD, 0, mmUVD_LMI_JRBC_RB_64BIT_BAR_LOW, lower_32_bits(ring->gpu_addr));
+	WREG32_SOC15(UVD, 0, mmUVD_LMI_JRBC_RB_64BIT_BAR_HIGH, upper_32_bits(ring->gpu_addr));
+	WREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_RPTR, 0);
+	WREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_WPTR, 0);
+	WREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_CNTL, 0x00000002L);
+
+	/* initialize wptr */
+	ring->wptr = RREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_WPTR);
+
+	/* copy patch commands to the jpeg ring */
+	vcn_v1_0_jpeg_ring_set_patch_ring(ring,
+		(ring->wptr + ring->max_dw * amdgpu_sched_hw_submission));
+
 	return 0;
 }
 
@@ -680,16 +802,45 @@ static int vcn_v1_0_stop(struct amdgpu_device *adev)
 	WREG32_P(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_CTRL2), 0,
 			~UVD_LMI_CTRL2__STALL_ARB_UMC_MASK);
 
-	/* enable clock gating */
-	vcn_v1_0_enable_clock_gating(adev, true);
+	WREG32_SOC15(VCN, 0, mmUVD_STATUS, 0);
 
+	vcn_v1_0_enable_clock_gating(adev);
+	vcn_1_0_enable_static_power_gating(adev);
 	return 0;
 }
 
+static bool vcn_v1_0_is_idle(void *handle)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+	return (RREG32_SOC15(VCN, 0, mmUVD_STATUS) == 0x2);
+}
+
+static int vcn_v1_0_wait_for_idle(void *handle)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	int ret = 0;
+
+	SOC15_WAIT_ON_RREG(VCN, 0, mmUVD_STATUS, 0x2, 0x2, ret);
+
+	return ret;
+}
+
 static int vcn_v1_0_set_clockgating_state(void *handle,
 					  enum amd_clockgating_state state)
 {
-	/* needed for driver unload*/
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+	bool enable = (state == AMD_CG_STATE_GATE) ? true : false;
+
+	if (enable) {
+		/* wait for STATUS to clear */
+		if (vcn_v1_0_is_idle(handle))
+			return -EBUSY;
+		vcn_v1_0_enable_clock_gating(adev);
+	} else {
+		/* disable HW gating and enable Sw gating */
+		vcn_v1_0_disable_clock_gating(adev);
+	}
 	return 0;
 }
 
@@ -1015,6 +1166,383 @@ static void vcn_v1_0_enc_ring_emit_wreg(struct amdgpu_ring *ring,
 	amdgpu_ring_write(ring, val);
 }
 
+
+/**
+ * vcn_v1_0_jpeg_ring_get_rptr - get read pointer
+ *
+ * @ring: amdgpu_ring pointer
+ *
+ * Returns the current hardware read pointer
+ */
+static uint64_t vcn_v1_0_jpeg_ring_get_rptr(struct amdgpu_ring *ring)
+{
+	struct amdgpu_device *adev = ring->adev;
+
+	return RREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_RPTR);
+}
+
+/**
+ * vcn_v1_0_jpeg_ring_get_wptr - get write pointer
+ *
+ * @ring: amdgpu_ring pointer
+ *
+ * Returns the current hardware write pointer
+ */
+static uint64_t vcn_v1_0_jpeg_ring_get_wptr(struct amdgpu_ring *ring)
+{
+	struct amdgpu_device *adev = ring->adev;
+
+	return RREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_WPTR);
+}
+
+/**
+ * vcn_v1_0_jpeg_ring_set_wptr - set write pointer
+ *
+ * @ring: amdgpu_ring pointer
+ *
+ * Commits the write pointer to the hardware
+ */
+static void vcn_v1_0_jpeg_ring_set_wptr(struct amdgpu_ring *ring)
+{
+	struct amdgpu_device *adev = ring->adev;
+
+	WREG32_SOC15(UVD, 0, mmUVD_JRBC_RB_WPTR, lower_32_bits(ring->wptr));
+}
+
+/**
+ * vcn_v1_0_jpeg_ring_insert_start - insert a start command
+ *
+ * @ring: amdgpu_ring pointer
+ *
+ * Write a start command to the ring.
+ */
+static void vcn_v1_0_jpeg_ring_insert_start(struct amdgpu_ring *ring)
+{
+	struct amdgpu_device *adev = ring->adev;
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_EXTERNAL_REG_BASE), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, 0x68e04);
+
+	amdgpu_ring_write(ring, PACKETJ(0, 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, 0x80010000);
+}
+
+/**
+ * vcn_v1_0_jpeg_ring_insert_end - insert a end command
+ *
+ * @ring: amdgpu_ring pointer
+ *
+ * Write a end command to the ring.
+ */
+static void vcn_v1_0_jpeg_ring_insert_end(struct amdgpu_ring *ring)
+{
+	struct amdgpu_device *adev = ring->adev;
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_EXTERNAL_REG_BASE), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, 0x68e04);
+
+	amdgpu_ring_write(ring, PACKETJ(0, 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, 0x00010000);
+}
+
+/**
+ * vcn_v1_0_jpeg_ring_emit_fence - emit an fence & trap command
+ *
+ * @ring: amdgpu_ring pointer
+ * @fence: fence to emit
+ *
+ * Write a fence and a trap command to the ring.
+ */
+static void vcn_v1_0_jpeg_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64 seq,
+				     unsigned flags)
+{
+	struct amdgpu_device *adev = ring->adev;
+
+	WARN_ON(flags & AMDGPU_FENCE_FLAG_64BIT);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JPEG_GPCOM_DATA0), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, seq);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JPEG_GPCOM_DATA1), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, seq);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_LOW), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, lower_32_bits(addr));
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_HIGH), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, upper_32_bits(addr));
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JPEG_GPCOM_CMD), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, 0x8);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JPEG_GPCOM_CMD), 0, PACKETJ_CONDITION_CHECK0, PACKETJ_TYPE4));
+	amdgpu_ring_write(ring, 0);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_COND_RD_TIMER), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, 0x01400200);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_REF_DATA), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, seq);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_LOW), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, lower_32_bits(addr));
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_HIGH), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, upper_32_bits(addr));
+
+	amdgpu_ring_write(ring,
+		PACKETJ(0, 0, PACKETJ_CONDITION_CHECK3, PACKETJ_TYPE2));
+	amdgpu_ring_write(ring, 0xffffffff);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_EXTERNAL_REG_BASE), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, 0x3fbc);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(0, 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, 0x1);
+}
+
+/**
+ * vcn_v1_0_jpeg_ring_emit_ib - execute indirect buffer
+ *
+ * @ring: amdgpu_ring pointer
+ * @ib: indirect buffer to execute
+ *
+ * Write ring commands to execute the indirect buffer.
+ */
+static void vcn_v1_0_jpeg_ring_emit_ib(struct amdgpu_ring *ring,
+				  struct amdgpu_ib *ib,
+				  unsigned vmid, bool ctx_switch)
+{
+	struct amdgpu_device *adev = ring->adev;
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_IB_VMID), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, (vmid | (vmid << 4)));
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JPEG_VMID), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, (vmid | (vmid << 4)));
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_IB_64BIT_BAR_LOW), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, lower_32_bits(ib->gpu_addr));
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_IB_64BIT_BAR_HIGH), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, upper_32_bits(ib->gpu_addr));
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_IB_SIZE), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, ib->length_dw);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_LOW), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, lower_32_bits(ring->gpu_addr));
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_HIGH), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, upper_32_bits(ring->gpu_addr));
+
+	amdgpu_ring_write(ring,
+		PACKETJ(0, 0, PACKETJ_CONDITION_CHECK0, PACKETJ_TYPE2));
+	amdgpu_ring_write(ring, 0);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_COND_RD_TIMER), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, 0x01400200);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_REF_DATA), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, 0x2);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_STATUS), 0, PACKETJ_CONDITION_CHECK3, PACKETJ_TYPE3));
+	amdgpu_ring_write(ring, 0x2);
+}
+
+static void vcn_v1_0_jpeg_ring_emit_reg_wait(struct amdgpu_ring *ring,
+					    uint32_t reg, uint32_t val,
+					    uint32_t mask)
+{
+	struct amdgpu_device *adev = ring->adev;
+	uint32_t reg_offset = (reg << 2);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_COND_RD_TIMER), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, 0x01400200);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_REF_DATA), 0, 0, PACKETJ_TYPE0));
+	amdgpu_ring_write(ring, val);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_EXTERNAL_REG_BASE), 0, 0, PACKETJ_TYPE0));
+	if (((reg_offset >= 0x1f800) && (reg_offset <= 0x21fff)) ||
+		((reg_offset >= 0x1e000) && (reg_offset <= 0x1e1ff))) {
+		amdgpu_ring_write(ring, 0);
+		amdgpu_ring_write(ring,
+			PACKETJ((reg_offset >> 2), 0, 0, PACKETJ_TYPE3));
+	} else {
+		amdgpu_ring_write(ring, reg_offset);
+		amdgpu_ring_write(ring,
+			PACKETJ(0, 0, 0, PACKETJ_TYPE3));
+	}
+	amdgpu_ring_write(ring, mask);
+}
+
+static void vcn_v1_0_jpeg_ring_emit_vm_flush(struct amdgpu_ring *ring,
+		unsigned vmid, uint64_t pd_addr)
+{
+	struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->funcs->vmhub];
+	uint32_t data0, data1, mask;
+
+	pd_addr = amdgpu_gmc_emit_flush_gpu_tlb(ring, vmid, pd_addr);
+
+	/* wait for register write */
+	data0 = hub->ctx0_ptb_addr_lo32 + vmid * 2;
+	data1 = lower_32_bits(pd_addr);
+	mask = 0xffffffff;
+	vcn_v1_0_jpeg_ring_emit_reg_wait(ring, data0, data1, mask);
+}
+
+static void vcn_v1_0_jpeg_ring_emit_wreg(struct amdgpu_ring *ring,
+					uint32_t reg, uint32_t val)
+{
+	struct amdgpu_device *adev = ring->adev;
+	uint32_t reg_offset = (reg << 2);
+
+	amdgpu_ring_write(ring,
+		PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_EXTERNAL_REG_BASE), 0, 0, PACKETJ_TYPE0));
+	if (((reg_offset >= 0x1f800) && (reg_offset <= 0x21fff)) ||
+			((reg_offset >= 0x1e000) && (reg_offset <= 0x1e1ff))) {
+		amdgpu_ring_write(ring, 0);
+		amdgpu_ring_write(ring,
+			PACKETJ((reg_offset >> 2), 0, 0, PACKETJ_TYPE0));
+	} else {
+		amdgpu_ring_write(ring, reg_offset);
+		amdgpu_ring_write(ring,
+			PACKETJ(0, 0, 0, PACKETJ_TYPE0));
+	}
+	amdgpu_ring_write(ring, val);
+}
+
+static void vcn_v1_0_jpeg_ring_nop(struct amdgpu_ring *ring, uint32_t count)
+{
+	int i;
+
+	WARN_ON(ring->wptr % 2 || count % 2);
+
+	for (i = 0; i < count / 2; i++) {
+		amdgpu_ring_write(ring, PACKETJ(0, 0, 0, PACKETJ_TYPE6));
+		amdgpu_ring_write(ring, 0);
+	}
+}
+
+static void vcn_v1_0_jpeg_ring_patch_wreg(struct amdgpu_ring *ring, uint32_t *ptr, uint32_t reg_offset, uint32_t val)
+{
+	struct amdgpu_device *adev = ring->adev;
+	ring->ring[(*ptr)++] = PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_EXTERNAL_REG_BASE), 0, 0, PACKETJ_TYPE0);
+	if (((reg_offset >= 0x1f800) && (reg_offset <= 0x21fff)) ||
+		((reg_offset >= 0x1e000) && (reg_offset <= 0x1e1ff))) {
+		ring->ring[(*ptr)++] = 0;
+		ring->ring[(*ptr)++] = PACKETJ((reg_offset >> 2), 0, 0, PACKETJ_TYPE0);
+	} else {
+		ring->ring[(*ptr)++] = reg_offset;
+		ring->ring[(*ptr)++] = PACKETJ(0, 0, 0, PACKETJ_TYPE0);
+	}
+	ring->ring[(*ptr)++] = val;
+}
+
+static void vcn_v1_0_jpeg_ring_set_patch_ring(struct amdgpu_ring *ring, uint32_t ptr)
+{
+	struct amdgpu_device *adev = ring->adev;
+
+	uint32_t reg, reg_offset, val, mask, i;
+
+	// 1st: program mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_LOW
+	reg = SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_LOW);
+	reg_offset = (reg << 2);
+	val = lower_32_bits(ring->gpu_addr);
+	vcn_v1_0_jpeg_ring_patch_wreg(ring, &ptr, reg_offset, val);
+
+	// 2nd: program mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_HIGH
+	reg = SOC15_REG_OFFSET(UVD, 0, mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_HIGH);
+	reg_offset = (reg << 2);
+	val = upper_32_bits(ring->gpu_addr);
+	vcn_v1_0_jpeg_ring_patch_wreg(ring, &ptr, reg_offset, val);
+
+	// 3rd to 5th: issue MEM_READ commands
+	for (i = 0; i <= 2; i++) {
+		ring->ring[ptr++] = PACKETJ(0, 0, 0, PACKETJ_TYPE2);
+		ring->ring[ptr++] = 0;
+	}
+
+	// 6th: program mmUVD_JRBC_RB_CNTL register to enable NO_FETCH and RPTR write ability
+	reg = SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_CNTL);
+	reg_offset = (reg << 2);
+	val = 0x13;
+	vcn_v1_0_jpeg_ring_patch_wreg(ring, &ptr, reg_offset, val);
+
+	// 7th: program mmUVD_JRBC_RB_REF_DATA
+	reg = SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_REF_DATA);
+	reg_offset = (reg << 2);
+	val = 0x1;
+	vcn_v1_0_jpeg_ring_patch_wreg(ring, &ptr, reg_offset, val);
+
+	// 8th: issue conditional register read mmUVD_JRBC_RB_CNTL
+	reg = SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_CNTL);
+	reg_offset = (reg << 2);
+	val = 0x1;
+	mask = 0x1;
+
+	ring->ring[ptr++] = PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_COND_RD_TIMER), 0, 0, PACKETJ_TYPE0);
+	ring->ring[ptr++] = 0x01400200;
+	ring->ring[ptr++] = PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_REF_DATA), 0, 0, PACKETJ_TYPE0);
+	ring->ring[ptr++] = val;
+	ring->ring[ptr++] = PACKETJ(SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_EXTERNAL_REG_BASE), 0, 0, PACKETJ_TYPE0);
+	if (((reg_offset >= 0x1f800) && (reg_offset <= 0x21fff)) ||
+		((reg_offset >= 0x1e000) && (reg_offset <= 0x1e1ff))) {
+		ring->ring[ptr++] = 0;
+		ring->ring[ptr++] = PACKETJ((reg_offset >> 2), 0, 0, PACKETJ_TYPE3);
+	} else {
+		ring->ring[ptr++] = reg_offset;
+		ring->ring[ptr++] = PACKETJ(0, 0, 0, PACKETJ_TYPE3);
+	}
+	ring->ring[ptr++] = mask;
+
+	//9th to 21st: insert no-op
+	for (i = 0; i <= 12; i++) {
+		ring->ring[ptr++] = PACKETJ(0, 0, 0, PACKETJ_TYPE6);
+		ring->ring[ptr++] = 0;
+	}
+
+	//22nd: reset mmUVD_JRBC_RB_RPTR
+	reg = SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_RPTR);
+	reg_offset = (reg << 2);
+	val = 0;
+	vcn_v1_0_jpeg_ring_patch_wreg(ring, &ptr, reg_offset, val);
+
+	//23rd: program mmUVD_JRBC_RB_CNTL to disable no_fetch
+	reg = SOC15_REG_OFFSET(UVD, 0, mmUVD_JRBC_RB_CNTL);
+	reg_offset = (reg << 2);
+	val = 0x12;
+	vcn_v1_0_jpeg_ring_patch_wreg(ring, &ptr, reg_offset, val);
+}
+
 static int vcn_v1_0_set_interrupt_state(struct amdgpu_device *adev,
 					struct amdgpu_irq_src *source,
 					unsigned type,
@@ -1039,6 +1567,9 @@ static int vcn_v1_0_process_interrupt(struct amdgpu_device *adev,
 	case 120:
 		amdgpu_fence_process(&adev->vcn.ring_enc[1]);
 		break;
+	case 126:
+		amdgpu_fence_process(&adev->vcn.ring_jpeg);
+		break;
 	default:
 		DRM_ERROR("Unhandled interrupt: %d %d\n",
 			  entry->src_id, entry->src_data[0]);
@@ -1048,16 +1579,36 @@ static int vcn_v1_0_process_interrupt(struct amdgpu_device *adev,
 	return 0;
 }
 
-static void vcn_v1_0_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
+static void vcn_v1_0_dec_ring_insert_nop(struct amdgpu_ring *ring, uint32_t count)
 {
-	int i;
 	struct amdgpu_device *adev = ring->adev;
+	int i;
 
-	for (i = 0; i < count; i++)
-		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_NO_OP), 0));
+	WARN_ON(ring->wptr % 2 || count % 2);
 
+	for (i = 0; i < count / 2; i++) {
+		amdgpu_ring_write(ring, PACKET0(SOC15_REG_OFFSET(UVD, 0, mmUVD_NO_OP), 0));
+		amdgpu_ring_write(ring, 0);
+	}
 }
 
+static int vcn_v1_0_set_powergating_state(void *handle,
+					  enum amd_powergating_state state)
+{
+	/* This doesn't actually powergate the VCN block.
+	 * That's done in the dpm code via the SMC.  This
+	 * just re-inits the block as necessary.  The actual
+	 * gating still happens in the dpm code.  We should
+	 * revisit this when there is a cleaner line between
+	 * the smc and the hw blocks
+	 */
+	struct amdgpu_device *adev = (struct amdgpu_device *)handle;
+
+	if (state == AMD_PG_STATE_GATE)
+		return vcn_v1_0_stop(adev);
+	else
+		return vcn_v1_0_start(adev);
+}
 
 static const struct amd_ip_funcs vcn_v1_0_ip_funcs = {
 	.name = "vcn_v1_0",
@@ -1069,20 +1620,19 @@ static const struct amd_ip_funcs vcn_v1_0_ip_funcs = {
 	.hw_fini = vcn_v1_0_hw_fini,
 	.suspend = vcn_v1_0_suspend,
 	.resume = vcn_v1_0_resume,
-	.is_idle = NULL /* vcn_v1_0_is_idle */,
-	.wait_for_idle = NULL /* vcn_v1_0_wait_for_idle */,
+	.is_idle = vcn_v1_0_is_idle,
+	.wait_for_idle = vcn_v1_0_wait_for_idle,
 	.check_soft_reset = NULL /* vcn_v1_0_check_soft_reset */,
 	.pre_soft_reset = NULL /* vcn_v1_0_pre_soft_reset */,
 	.soft_reset = NULL /* vcn_v1_0_soft_reset */,
 	.post_soft_reset = NULL /* vcn_v1_0_post_soft_reset */,
 	.set_clockgating_state = vcn_v1_0_set_clockgating_state,
-	.set_powergating_state = NULL /* vcn_v1_0_set_powergating_state */,
+	.set_powergating_state = vcn_v1_0_set_powergating_state,
 };
 
 static const struct amdgpu_ring_funcs vcn_v1_0_dec_ring_vm_funcs = {
 	.type = AMDGPU_RING_TYPE_VCN_DEC,
 	.align_mask = 0xf,
-	.nop = PACKET0(0x81ff, 0),
 	.support_64bit_ptrs = false,
 	.vmhub = AMDGPU_MMHUB,
 	.get_rptr = vcn_v1_0_dec_ring_get_rptr,
@@ -1101,7 +1651,7 @@ static const struct amdgpu_ring_funcs vcn_v1_0_dec_ring_vm_funcs = {
 	.emit_vm_flush = vcn_v1_0_dec_ring_emit_vm_flush,
 	.test_ring = amdgpu_vcn_dec_ring_test_ring,
 	.test_ib = amdgpu_vcn_dec_ring_test_ib,
-	.insert_nop = vcn_v1_0_ring_insert_nop,
+	.insert_nop = vcn_v1_0_dec_ring_insert_nop,
 	.insert_start = vcn_v1_0_dec_ring_insert_start,
 	.insert_end = vcn_v1_0_dec_ring_insert_end,
 	.pad_ib = amdgpu_ring_generic_pad_ib,
@@ -1109,6 +1659,7 @@ static const struct amdgpu_ring_funcs vcn_v1_0_dec_ring_vm_funcs = {
 	.end_use = amdgpu_vcn_ring_end_use,
 	.emit_wreg = vcn_v1_0_dec_ring_emit_wreg,
 	.emit_reg_wait = vcn_v1_0_dec_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
 };
 
 static const struct amdgpu_ring_funcs vcn_v1_0_enc_ring_vm_funcs = {
@@ -1139,6 +1690,40 @@ static const struct amdgpu_ring_funcs vcn_v1_0_enc_ring_vm_funcs = {
 	.end_use = amdgpu_vcn_ring_end_use,
 	.emit_wreg = vcn_v1_0_enc_ring_emit_wreg,
 	.emit_reg_wait = vcn_v1_0_enc_ring_emit_reg_wait,
+	.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
+};
+
+static const struct amdgpu_ring_funcs vcn_v1_0_jpeg_ring_vm_funcs = {
+	.type = AMDGPU_RING_TYPE_VCN_JPEG,
+	.align_mask = 0xf,
+	.nop = PACKET0(0x81ff, 0),
+	.support_64bit_ptrs = false,
+	.vmhub = AMDGPU_MMHUB,
+	.extra_dw = 64,
+	.get_rptr = vcn_v1_0_jpeg_ring_get_rptr,
+	.get_wptr = vcn_v1_0_jpeg_ring_get_wptr,
+	.set_wptr = vcn_v1_0_jpeg_ring_set_wptr,
+	.emit_frame_size =
+		6 + 6 + /* hdp invalidate / flush */
+		SOC15_FLUSH_GPU_TLB_NUM_WREG * 6 +
+		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 8 +
+		8 + /* vcn_v1_0_dec_ring_emit_vm_flush */
+		14 + 14 + /* vcn_v1_0_dec_ring_emit_fence x2 vm fence */
+		6,
+	.emit_ib_size = 22, /* vcn_v1_0_dec_ring_emit_ib */
+	.emit_ib = vcn_v1_0_jpeg_ring_emit_ib,
+	.emit_fence = vcn_v1_0_jpeg_ring_emit_fence,
+	.emit_vm_flush = vcn_v1_0_jpeg_ring_emit_vm_flush,
+	.test_ring = amdgpu_vcn_jpeg_ring_test_ring,
+	.test_ib = amdgpu_vcn_jpeg_ring_test_ib,
+	.insert_nop = vcn_v1_0_jpeg_ring_nop,
+	.insert_start = vcn_v1_0_jpeg_ring_insert_start,
+	.insert_end = vcn_v1_0_jpeg_ring_insert_end,
+	.pad_ib = amdgpu_ring_generic_pad_ib,
+	.begin_use = amdgpu_vcn_ring_begin_use,
+	.end_use = amdgpu_vcn_ring_end_use,
+	.emit_wreg = vcn_v1_0_jpeg_ring_emit_wreg,
+	.emit_reg_wait = vcn_v1_0_jpeg_ring_emit_reg_wait,
 };
 
 static void vcn_v1_0_set_dec_ring_funcs(struct amdgpu_device *adev)
@@ -1157,6 +1742,12 @@ static void vcn_v1_0_set_enc_ring_funcs(struct amdgpu_device *adev)
 	DRM_INFO("VCN encode is enabled in VM mode\n");
 }
 
+static void vcn_v1_0_set_jpeg_ring_funcs(struct amdgpu_device *adev)
+{
+	adev->vcn.ring_jpeg.funcs = &vcn_v1_0_jpeg_ring_vm_funcs;
+	DRM_INFO("VCN jpeg decode is enabled in VM mode\n");
+}
+
 static const struct amdgpu_irq_src_funcs vcn_v1_0_irq_funcs = {
 	.set = vcn_v1_0_set_interrupt_state,
 	.process = vcn_v1_0_process_interrupt,
diff --git a/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c b/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c
index 45aafca7f315..c5c9b2bc190d 100644
--- a/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c
+++ b/drivers/gpu/drm/amd/amdgpu/vega10_reg_init.c
@@ -51,6 +51,7 @@ int vega10_reg_base_init(struct amdgpu_device *adev)
 		adev->reg_offset[PWR_HWIP][i] = (uint32_t *)(&(PWR_BASE.instance[i]));
 		adev->reg_offset[NBIF_HWIP][i] = (uint32_t *)(&(NBIF_BASE.instance[i]));
 		adev->reg_offset[THM_HWIP][i] = (uint32_t *)(&(THM_BASE.instance[i]));
+		adev->reg_offset[CLK_HWIP][i] = (uint32_t *)(&(CLK_BASE.instance[i]));
 	}
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c b/drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c
new file mode 100644
index 000000000000..52778de93ab0
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdgpu/vega20_reg_init.c
@@ -0,0 +1,53 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "amdgpu.h"
+#include "soc15.h"
+
+#include "soc15_common.h"
+#include "soc15_hw_ip.h"
+#include "vega20_ip_offset.h"
+
+int vega20_reg_base_init(struct amdgpu_device *adev)
+{
+	/* HW has more IP blocks,  only initialized the blocke beend by our driver  */
+	uint32_t i;
+	for (i = 0 ; i < MAX_INSTANCE ; ++i) {
+		adev->reg_offset[GC_HWIP][i] = (uint32_t *)(&(GC_BASE.instance[i]));
+		adev->reg_offset[HDP_HWIP][i] = (uint32_t *)(&(HDP_BASE.instance[i]));
+		adev->reg_offset[MMHUB_HWIP][i] = (uint32_t *)(&(MMHUB_BASE.instance[i]));
+		adev->reg_offset[ATHUB_HWIP][i] = (uint32_t *)(&(ATHUB_BASE.instance[i]));
+		adev->reg_offset[NBIO_HWIP][i] = (uint32_t *)(&(NBIO_BASE.instance[i]));
+		adev->reg_offset[MP0_HWIP][i] = (uint32_t *)(&(MP0_BASE.instance[i]));
+		adev->reg_offset[UVD_HWIP][i] = (uint32_t *)(&(UVD_BASE.instance[i]));
+		adev->reg_offset[VCE_HWIP][i] = (uint32_t *)(&(VCE_BASE.instance[i]));
+		adev->reg_offset[DF_HWIP][i] = (uint32_t *)(&(DF_BASE.instance[i]));
+		adev->reg_offset[DCE_HWIP][i] = (uint32_t *)(&(DCE_BASE.instance[i]));
+		adev->reg_offset[OSSSYS_HWIP][i] = (uint32_t *)(&(OSSSYS_BASE.instance[i]));
+		adev->reg_offset[SDMA0_HWIP][i] = (uint32_t *)(&(SDMA0_BASE.instance[i]));
+		adev->reg_offset[SDMA1_HWIP][i] = (uint32_t *)(&(SDMA1_BASE.instance[i]));
+		adev->reg_offset[SMUIO_HWIP][i] = (uint32_t *)(&(SMUIO_BASE.instance[i]));
+	}
+	return 0;
+}
+
+
diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
index 126f1276d347..42c8ad105b05 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -305,9 +305,10 @@ static void vi_init_golden_registers(struct amdgpu_device *adev)
 							stoney_mgcg_cgcg_init,
 							ARRAY_SIZE(stoney_mgcg_cgcg_init));
 		break;
-	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 	default:
 		break;
 	}
@@ -728,33 +729,59 @@ static int vi_set_uvd_clock(struct amdgpu_device *adev, u32 clock,
 		return r;
 
 	tmp = RREG32_SMC(cntl_reg);
-	tmp &= ~(CG_DCLK_CNTL__DCLK_DIR_CNTL_EN_MASK |
-		CG_DCLK_CNTL__DCLK_DIVIDER_MASK);
+
+	if (adev->flags & AMD_IS_APU)
+		tmp &= ~CG_DCLK_CNTL__DCLK_DIVIDER_MASK;
+	else
+		tmp &= ~(CG_DCLK_CNTL__DCLK_DIR_CNTL_EN_MASK |
+				CG_DCLK_CNTL__DCLK_DIVIDER_MASK);
 	tmp |= dividers.post_divider;
 	WREG32_SMC(cntl_reg, tmp);
 
 	for (i = 0; i < 100; i++) {
-		if (RREG32_SMC(status_reg) & CG_DCLK_STATUS__DCLK_STATUS_MASK)
-			break;
+		tmp = RREG32_SMC(status_reg);
+		if (adev->flags & AMD_IS_APU) {
+			if (tmp & 0x10000)
+				break;
+		} else {
+			if (tmp & CG_DCLK_STATUS__DCLK_STATUS_MASK)
+				break;
+		}
 		mdelay(10);
 	}
 	if (i == 100)
 		return -ETIMEDOUT;
-
 	return 0;
 }
 
+#define ixGNB_CLK1_DFS_CNTL 0xD82200F0
+#define ixGNB_CLK1_STATUS   0xD822010C
+#define ixGNB_CLK2_DFS_CNTL 0xD8220110
+#define ixGNB_CLK2_STATUS   0xD822012C
+#define ixGNB_CLK3_DFS_CNTL 0xD8220130
+#define ixGNB_CLK3_STATUS   0xD822014C
+
 static int vi_set_uvd_clocks(struct amdgpu_device *adev, u32 vclk, u32 dclk)
 {
 	int r;
 
-	r = vi_set_uvd_clock(adev, vclk, ixCG_VCLK_CNTL, ixCG_VCLK_STATUS);
-	if (r)
-		return r;
+	if (adev->flags & AMD_IS_APU) {
+		r = vi_set_uvd_clock(adev, vclk, ixGNB_CLK2_DFS_CNTL, ixGNB_CLK2_STATUS);
+		if (r)
+			return r;
 
-	r = vi_set_uvd_clock(adev, dclk, ixCG_DCLK_CNTL, ixCG_DCLK_STATUS);
-	if (r)
-		return r;
+		r = vi_set_uvd_clock(adev, dclk, ixGNB_CLK1_DFS_CNTL, ixGNB_CLK1_STATUS);
+		if (r)
+			return r;
+	} else {
+		r = vi_set_uvd_clock(adev, vclk, ixCG_VCLK_CNTL, ixCG_VCLK_STATUS);
+		if (r)
+			return r;
+
+		r = vi_set_uvd_clock(adev, dclk, ixCG_DCLK_CNTL, ixCG_DCLK_STATUS);
+		if (r)
+			return r;
+	}
 
 	return 0;
 }
@@ -764,6 +791,22 @@ static int vi_set_vce_clocks(struct amdgpu_device *adev, u32 evclk, u32 ecclk)
 	int r, i;
 	struct atom_clock_dividers dividers;
 	u32 tmp;
+	u32 reg_ctrl;
+	u32 reg_status;
+	u32 status_mask;
+	u32 reg_mask;
+
+	if (adev->flags & AMD_IS_APU) {
+		reg_ctrl = ixGNB_CLK3_DFS_CNTL;
+		reg_status = ixGNB_CLK3_STATUS;
+		status_mask = 0x00010000;
+		reg_mask = CG_ECLK_CNTL__ECLK_DIVIDER_MASK;
+	} else {
+		reg_ctrl = ixCG_ECLK_CNTL;
+		reg_status = ixCG_ECLK_STATUS;
+		status_mask = CG_ECLK_STATUS__ECLK_STATUS_MASK;
+		reg_mask = CG_ECLK_CNTL__ECLK_DIR_CNTL_EN_MASK | CG_ECLK_CNTL__ECLK_DIVIDER_MASK;
+	}
 
 	r = amdgpu_atombios_get_clock_dividers(adev,
 					       COMPUTE_GPUCLK_INPUT_FLAG_DEFAULT_GPUCLK,
@@ -772,24 +815,25 @@ static int vi_set_vce_clocks(struct amdgpu_device *adev, u32 evclk, u32 ecclk)
 		return r;
 
 	for (i = 0; i < 100; i++) {
-		if (RREG32_SMC(ixCG_ECLK_STATUS) & CG_ECLK_STATUS__ECLK_STATUS_MASK)
+		if (RREG32_SMC(reg_status) & status_mask)
 			break;
 		mdelay(10);
 	}
+
 	if (i == 100)
 		return -ETIMEDOUT;
 
-	tmp = RREG32_SMC(ixCG_ECLK_CNTL);
-	tmp &= ~(CG_ECLK_CNTL__ECLK_DIR_CNTL_EN_MASK |
-		CG_ECLK_CNTL__ECLK_DIVIDER_MASK);
+	tmp = RREG32_SMC(reg_ctrl);
+	tmp &= ~reg_mask;
 	tmp |= dividers.post_divider;
-	WREG32_SMC(ixCG_ECLK_CNTL, tmp);
+	WREG32_SMC(reg_ctrl, tmp);
 
 	for (i = 0; i < 100; i++) {
-		if (RREG32_SMC(ixCG_ECLK_STATUS) & CG_ECLK_STATUS__ECLK_STATUS_MASK)
+		if (RREG32_SMC(reg_status) & status_mask)
 			break;
 		mdelay(10);
 	}
+
 	if (i == 100)
 		return -ETIMEDOUT;
 
@@ -876,6 +920,27 @@ static void vi_invalidate_hdp(struct amdgpu_device *adev,
 	}
 }
 
+static bool vi_need_full_reset(struct amdgpu_device *adev)
+{
+	switch (adev->asic_type) {
+	case CHIP_CARRIZO:
+	case CHIP_STONEY:
+		/* CZ has hang issues with full reset at the moment */
+		return false;
+	case CHIP_FIJI:
+	case CHIP_TONGA:
+		/* XXX: soft reset should work on fiji and tonga */
+		return true;
+	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
+	case CHIP_POLARIS12:
+	case CHIP_TOPAZ:
+	default:
+		/* change this when we support soft reset */
+		return true;
+	}
+}
+
 static const struct amdgpu_asic_funcs vi_asic_funcs =
 {
 	.read_disabled_bios = &vi_read_disabled_bios,
@@ -889,6 +954,7 @@ static const struct amdgpu_asic_funcs vi_asic_funcs =
 	.get_config_memsize = &vi_get_config_memsize,
 	.flush_hdp = &vi_flush_hdp,
 	.invalidate_hdp = &vi_invalidate_hdp,
+	.need_full_reset = &vi_need_full_reset,
 };
 
 #define CZ_REV_BRISTOL(rev)	 \
@@ -1031,6 +1097,30 @@ static int vi_common_early_init(void *handle)
 		adev->pg_flags = 0;
 		adev->external_rev_id = adev->rev_id + 0x64;
 		break;
+	case CHIP_VEGAM:
+		adev->cg_flags = 0;
+			/*AMD_CG_SUPPORT_GFX_MGCG |
+			AMD_CG_SUPPORT_GFX_RLC_LS |
+			AMD_CG_SUPPORT_GFX_CP_LS |
+			AMD_CG_SUPPORT_GFX_CGCG |
+			AMD_CG_SUPPORT_GFX_CGLS |
+			AMD_CG_SUPPORT_GFX_3D_CGCG |
+			AMD_CG_SUPPORT_GFX_3D_CGLS |
+			AMD_CG_SUPPORT_SDMA_MGCG |
+			AMD_CG_SUPPORT_SDMA_LS |
+			AMD_CG_SUPPORT_BIF_MGCG |
+			AMD_CG_SUPPORT_BIF_LS |
+			AMD_CG_SUPPORT_HDP_MGCG |
+			AMD_CG_SUPPORT_HDP_LS |
+			AMD_CG_SUPPORT_ROM_MGCG |
+			AMD_CG_SUPPORT_MC_MGCG |
+			AMD_CG_SUPPORT_MC_LS |
+			AMD_CG_SUPPORT_DRM_LS |
+			AMD_CG_SUPPORT_UVD_MGCG |
+			AMD_CG_SUPPORT_VCE_MGCG;*/
+		adev->pg_flags = 0;
+		adev->external_rev_id = adev->rev_id + 0x6E;
+		break;
 	case CHIP_CARRIZO:
 		adev->cg_flags = AMD_CG_SUPPORT_UVD_MGCG |
 			AMD_CG_SUPPORT_GFX_MGCG |
@@ -1273,11 +1363,11 @@ static int vi_common_set_clockgating_state_by_smu(void *handle,
 
 	if (adev->cg_flags & (AMD_CG_SUPPORT_MC_LS | AMD_CG_SUPPORT_MC_MGCG)) {
 		if (adev->cg_flags & AMD_CG_SUPPORT_MC_LS) {
-			pp_support_state = AMD_CG_SUPPORT_MC_LS;
+			pp_support_state = PP_STATE_SUPPORT_LS;
 			pp_state = PP_STATE_LS;
 		}
 		if (adev->cg_flags & AMD_CG_SUPPORT_MC_MGCG) {
-			pp_support_state |= AMD_CG_SUPPORT_MC_MGCG;
+			pp_support_state |= PP_STATE_SUPPORT_CG;
 			pp_state |= PP_STATE_CG;
 		}
 		if (state == AMD_CG_STATE_UNGATE)
@@ -1292,11 +1382,11 @@ static int vi_common_set_clockgating_state_by_smu(void *handle,
 
 	if (adev->cg_flags & (AMD_CG_SUPPORT_SDMA_LS | AMD_CG_SUPPORT_SDMA_MGCG)) {
 		if (adev->cg_flags & AMD_CG_SUPPORT_SDMA_LS) {
-			pp_support_state = AMD_CG_SUPPORT_SDMA_LS;
+			pp_support_state = PP_STATE_SUPPORT_LS;
 			pp_state = PP_STATE_LS;
 		}
 		if (adev->cg_flags & AMD_CG_SUPPORT_SDMA_MGCG) {
-			pp_support_state |= AMD_CG_SUPPORT_SDMA_MGCG;
+			pp_support_state |= PP_STATE_SUPPORT_CG;
 			pp_state |= PP_STATE_CG;
 		}
 		if (state == AMD_CG_STATE_UNGATE)
@@ -1311,11 +1401,11 @@ static int vi_common_set_clockgating_state_by_smu(void *handle,
 
 	if (adev->cg_flags & (AMD_CG_SUPPORT_HDP_LS | AMD_CG_SUPPORT_HDP_MGCG)) {
 		if (adev->cg_flags & AMD_CG_SUPPORT_HDP_LS) {
-			pp_support_state = AMD_CG_SUPPORT_HDP_LS;
+			pp_support_state = PP_STATE_SUPPORT_LS;
 			pp_state = PP_STATE_LS;
 		}
 		if (adev->cg_flags & AMD_CG_SUPPORT_HDP_MGCG) {
-			pp_support_state |= AMD_CG_SUPPORT_HDP_MGCG;
+			pp_support_state |= PP_STATE_SUPPORT_CG;
 			pp_state |= PP_STATE_CG;
 		}
 		if (state == AMD_CG_STATE_UNGATE)
@@ -1422,6 +1512,7 @@ static int vi_common_set_clockgating_state(void *handle,
 	case CHIP_POLARIS10:
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 		vi_common_set_clockgating_state_by_smu(adev, state);
 	default:
 		break;
@@ -1551,9 +1642,10 @@ int vi_set_ip_blocks(struct amdgpu_device *adev)
 			amdgpu_device_ip_block_add(adev, &vce_v3_0_ip_block);
 		}
 		break;
-	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 		amdgpu_device_ip_block_add(adev, &vi_common_ip_block);
 		amdgpu_device_ip_block_add(adev, &gmc_v8_1_ip_block);
 		amdgpu_device_ip_block_add(adev, &tonga_ih_ip_block);
diff --git a/drivers/gpu/drm/amd/amdkfd/Makefile b/drivers/gpu/drm/amd/amdkfd/Makefile
index 0d0242240c47..ffd096fffc1c 100644
--- a/drivers/gpu/drm/amd/amdkfd/Makefile
+++ b/drivers/gpu/drm/amd/amdkfd/Makefile
@@ -30,12 +30,14 @@ amdkfd-y	:= kfd_module.o kfd_device.o kfd_chardev.o kfd_topology.o \
 		kfd_pasid.o kfd_doorbell.o kfd_flat_memory.o \
 		kfd_process.o kfd_queue.o kfd_mqd_manager.o \
 		kfd_mqd_manager_cik.o kfd_mqd_manager_vi.o \
+		kfd_mqd_manager_v9.o \
 		kfd_kernel_queue.o kfd_kernel_queue_cik.o \
-		kfd_kernel_queue_vi.o kfd_packet_manager.o \
-		kfd_process_queue_manager.o kfd_device_queue_manager.o \
-		kfd_device_queue_manager_cik.o kfd_device_queue_manager_vi.o \
+		kfd_kernel_queue_vi.o kfd_kernel_queue_v9.o \
+		kfd_packet_manager.o kfd_process_queue_manager.o \
+		kfd_device_queue_manager.o kfd_device_queue_manager_cik.o \
+		kfd_device_queue_manager_vi.o kfd_device_queue_manager_v9.o \
 		kfd_interrupt.o kfd_events.o cik_event_interrupt.o \
-		kfd_dbgdev.o kfd_dbgmgr.o kfd_crat.o
+		kfd_int_process_v9.o kfd_dbgdev.o kfd_dbgmgr.o kfd_crat.o
 
 ifneq ($(CONFIG_AMD_IOMMU_V2),)
 amdkfd-y += kfd_iommu.o
diff --git a/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c b/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c
index 3d5ccb3755d4..49df6c791cfc 100644
--- a/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c
+++ b/drivers/gpu/drm/amd/amdkfd/cik_event_interrupt.c
@@ -27,18 +27,28 @@
 static bool cik_event_interrupt_isr(struct kfd_dev *dev,
 					const uint32_t *ih_ring_entry)
 {
-	unsigned int pasid;
 	const struct cik_ih_ring_entry *ihre =
 			(const struct cik_ih_ring_entry *)ih_ring_entry;
+	unsigned int vmid, pasid;
+
+	/* Only handle interrupts from KFD VMIDs */
+	vmid  = (ihre->ring_id & 0x0000ff00) >> 8;
+	if (vmid < dev->vm_info.first_vmid_kfd ||
+	    vmid > dev->vm_info.last_vmid_kfd)
+		return 0;
 
+	/* If there is no valid PASID, it's likely a firmware bug */
 	pasid = (ihre->ring_id & 0xffff0000) >> 16;
+	if (WARN_ONCE(pasid == 0, "FW bug: No PASID in KFD interrupt"))
+		return 0;
 
-	/* Do not process in ISR, just request it to be forwarded to WQ. */
-	return (pasid != 0) &&
-		(ihre->source_id == CIK_INTSRC_CP_END_OF_PIPE ||
+	/* Interrupt types we care about: various signals and faults.
+	 * They will be forwarded to a work queue (see below).
+	 */
+	return ihre->source_id == CIK_INTSRC_CP_END_OF_PIPE ||
 		ihre->source_id == CIK_INTSRC_SDMA_TRAP ||
 		ihre->source_id == CIK_INTSRC_SQ_INTERRUPT_MSG ||
-		ihre->source_id == CIK_INTSRC_CP_BAD_OPCODE);
+		ihre->source_id == CIK_INTSRC_CP_BAD_OPCODE;
 }
 
 static void cik_event_interrupt_wq(struct kfd_dev *dev,
diff --git a/drivers/gpu/drm/amd/amdkfd/cik_regs.h b/drivers/gpu/drm/amd/amdkfd/cik_regs.h
index 48769d12dd7b..37ce6dd65391 100644
--- a/drivers/gpu/drm/amd/amdkfd/cik_regs.h
+++ b/drivers/gpu/drm/amd/amdkfd/cik_regs.h
@@ -33,7 +33,8 @@
 #define	APE1_MTYPE(x)					((x) << 7)
 
 /* valid for both DEFAULT_MTYPE and APE1_MTYPE */
-#define	MTYPE_CACHED					0
+#define	MTYPE_CACHED_NV					0
+#define	MTYPE_CACHED					1
 #define	MTYPE_NONCACHED					3
 
 #define	DEFAULT_CP_HQD_PERSISTENT_STATE			(0x33U << 8)
diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
new file mode 100644
index 000000000000..f68aef02fc1f
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler.h
@@ -0,0 +1,560 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+static const uint32_t cwsr_trap_gfx8_hex[] = {
+	0xbf820001, 0xbf820125,
+	0xb8f4f802, 0x89748674,
+	0xb8f5f803, 0x8675ff75,
+	0x00000400, 0xbf850011,
+	0xc00a1e37, 0x00000000,
+	0xbf8c007f, 0x87777978,
+	0xbf840002, 0xb974f802,
+	0xbe801d78, 0xb8f5f803,
+	0x8675ff75, 0x000001ff,
+	0xbf850002, 0x80708470,
+	0x82718071, 0x8671ff71,
+	0x0000ffff, 0xb974f802,
+	0xbe801f70, 0xb8f5f803,
+	0x8675ff75, 0x00000100,
+	0xbf840006, 0xbefa0080,
+	0xb97a0203, 0x8671ff71,
+	0x0000ffff, 0x80f08870,
+	0x82f18071, 0xbefa0080,
+	0xb97a0283, 0xbef60068,
+	0xbef70069, 0xb8fa1c07,
+	0x8e7a9c7a, 0x87717a71,
+	0xb8fa03c7, 0x8e7a9b7a,
+	0x87717a71, 0xb8faf807,
+	0x867aff7a, 0x00007fff,
+	0xb97af807, 0xbef2007e,
+	0xbef3007f, 0xbefe0180,
+	0xbf900004, 0x877a8474,
+	0xb97af802, 0xbf8e0002,
+	0xbf88fffe, 0xbef8007e,
+	0x8679ff7f, 0x0000ffff,
+	0x8779ff79, 0x00040000,
+	0xbefa0080, 0xbefb00ff,
+	0x00807fac, 0x867aff7f,
+	0x08000000, 0x8f7a837a,
+	0x877b7a7b, 0x867aff7f,
+	0x70000000, 0x8f7a817a,
+	0x877b7a7b, 0xbeef007c,
+	0xbeee0080, 0xb8ee2a05,
+	0x806e816e, 0x8e6e8a6e,
+	0xb8fa1605, 0x807a817a,
+	0x8e7a867a, 0x806e7a6e,
+	0xbefa0084, 0xbefa00ff,
+	0x01000000, 0xbefe007c,
+	0xbefc006e, 0xc0611bfc,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xbefe007c,
+	0xbefc006e, 0xc0611c3c,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xbefe007c,
+	0xbefc006e, 0xc0611c7c,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xbefe007c,
+	0xbefc006e, 0xc0611cbc,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xbefe007c,
+	0xbefc006e, 0xc0611cfc,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xbefe007c,
+	0xbefc006e, 0xc0611d3c,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xb8f5f803,
+	0xbefe007c, 0xbefc006e,
+	0xc0611d7c, 0x0000007c,
+	0x806e846e, 0xbefc007e,
+	0xbefe007c, 0xbefc006e,
+	0xc0611dbc, 0x0000007c,
+	0x806e846e, 0xbefc007e,
+	0xbefe007c, 0xbefc006e,
+	0xc0611dfc, 0x0000007c,
+	0x806e846e, 0xbefc007e,
+	0xb8eff801, 0xbefe007c,
+	0xbefc006e, 0xc0611bfc,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xbefe007c,
+	0xbefc006e, 0xc0611b3c,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0xbefe007c,
+	0xbefc006e, 0xc0611b7c,
+	0x0000007c, 0x806e846e,
+	0xbefc007e, 0x867aff7f,
+	0x04000000, 0xbef30080,
+	0x8773737a, 0xb8ee2a05,
+	0x806e816e, 0x8e6e8a6e,
+	0xb8f51605, 0x80758175,
+	0x8e758475, 0x8e7a8275,
+	0xbefa00ff, 0x01000000,
+	0xbef60178, 0x80786e78,
+	0x82798079, 0xbefc0080,
+	0xbe802b00, 0xbe822b02,
+	0xbe842b04, 0xbe862b06,
+	0xbe882b08, 0xbe8a2b0a,
+	0xbe8c2b0c, 0xbe8e2b0e,
+	0xc06b003c, 0x00000000,
+	0xc06b013c, 0x00000010,
+	0xc06b023c, 0x00000020,
+	0xc06b033c, 0x00000030,
+	0x8078c078, 0x82798079,
+	0x807c907c, 0xbf0a757c,
+	0xbf85ffeb, 0xbef80176,
+	0xbeee0080, 0xbefe00c1,
+	0xbeff00c1, 0xbefa00ff,
+	0x01000000, 0xe0724000,
+	0x6e1e0000, 0xe0724100,
+	0x6e1e0100, 0xe0724200,
+	0x6e1e0200, 0xe0724300,
+	0x6e1e0300, 0xbefe00c1,
+	0xbeff00c1, 0xb8f54306,
+	0x8675c175, 0xbf84002c,
+	0xbf8a0000, 0x867aff73,
+	0x04000000, 0xbf840028,
+	0x8e758675, 0x8e758275,
+	0xbefa0075, 0xb8ee2a05,
+	0x806e816e, 0x8e6e8a6e,
+	0xb8fa1605, 0x807a817a,
+	0x8e7a867a, 0x806e7a6e,
+	0x806eff6e, 0x00000080,
+	0xbefa00ff, 0x01000000,
+	0xbefc0080, 0xd28c0002,
+	0x000100c1, 0xd28d0003,
+	0x000204c1, 0xd1060002,
+	0x00011103, 0x7e0602ff,
+	0x00000200, 0xbefc00ff,
+	0x00010000, 0xbe80007b,
+	0x867bff7b, 0xff7fffff,
+	0x877bff7b, 0x00058000,
+	0xd8ec0000, 0x00000002,
+	0xbf8c007f, 0xe0765000,
+	0x6e1e0002, 0x32040702,
+	0xd0c9006a, 0x0000eb02,
+	0xbf87fff7, 0xbefb0000,
+	0xbeee00ff, 0x00000400,
+	0xbefe00c1, 0xbeff00c1,
+	0xb8f52a05, 0x80758175,
+	0x8e758275, 0x8e7a8875,
+	0xbefa00ff, 0x01000000,
+	0xbefc0084, 0xbf0a757c,
+	0xbf840015, 0xbf11017c,
+	0x8075ff75, 0x00001000,
+	0x7e000300, 0x7e020301,
+	0x7e040302, 0x7e060303,
+	0xe0724000, 0x6e1e0000,
+	0xe0724100, 0x6e1e0100,
+	0xe0724200, 0x6e1e0200,
+	0xe0724300, 0x6e1e0300,
+	0x807c847c, 0x806eff6e,
+	0x00000400, 0xbf0a757c,
+	0xbf85ffef, 0xbf9c0000,
+	0xbf8200ca, 0xbef8007e,
+	0x8679ff7f, 0x0000ffff,
+	0x8779ff79, 0x00040000,
+	0xbefa0080, 0xbefb00ff,
+	0x00807fac, 0x8676ff7f,
+	0x08000000, 0x8f768376,
+	0x877b767b, 0x8676ff7f,
+	0x70000000, 0x8f768176,
+	0x877b767b, 0x8676ff7f,
+	0x04000000, 0xbf84001e,
+	0xbefe00c1, 0xbeff00c1,
+	0xb8f34306, 0x8673c173,
+	0xbf840019, 0x8e738673,
+	0x8e738273, 0xbefa0073,
+	0xb8f22a05, 0x80728172,
+	0x8e728a72, 0xb8f61605,
+	0x80768176, 0x8e768676,
+	0x80727672, 0x8072ff72,
+	0x00000080, 0xbefa00ff,
+	0x01000000, 0xbefc0080,
+	0xe0510000, 0x721e0000,
+	0xe0510100, 0x721e0000,
+	0x807cff7c, 0x00000200,
+	0x8072ff72, 0x00000200,
+	0xbf0a737c, 0xbf85fff6,
+	0xbef20080, 0xbefe00c1,
+	0xbeff00c1, 0xb8f32a05,
+	0x80738173, 0x8e738273,
+	0x8e7a8873, 0xbefa00ff,
+	0x01000000, 0xbef60072,
+	0x8072ff72, 0x00000400,
+	0xbefc0084, 0xbf11087c,
+	0x8073ff73, 0x00008000,
+	0xe0524000, 0x721e0000,
+	0xe0524100, 0x721e0100,
+	0xe0524200, 0x721e0200,
+	0xe0524300, 0x721e0300,
+	0xbf8c0f70, 0x7e000300,
+	0x7e020301, 0x7e040302,
+	0x7e060303, 0x807c847c,
+	0x8072ff72, 0x00000400,
+	0xbf0a737c, 0xbf85ffee,
+	0xbf9c0000, 0xe0524000,
+	0x761e0000, 0xe0524100,
+	0x761e0100, 0xe0524200,
+	0x761e0200, 0xe0524300,
+	0x761e0300, 0xb8f22a05,
+	0x80728172, 0x8e728a72,
+	0xb8f61605, 0x80768176,
+	0x8e768676, 0x80727672,
+	0x80f2c072, 0xb8f31605,
+	0x80738173, 0x8e738473,
+	0x8e7a8273, 0xbefa00ff,
+	0x01000000, 0xbefc0073,
+	0xc031003c, 0x00000072,
+	0x80f2c072, 0xbf8c007f,
+	0x80fc907c, 0xbe802d00,
+	0xbe822d02, 0xbe842d04,
+	0xbe862d06, 0xbe882d08,
+	0xbe8a2d0a, 0xbe8c2d0c,
+	0xbe8e2d0e, 0xbf06807c,
+	0xbf84fff1, 0xb8f22a05,
+	0x80728172, 0x8e728a72,
+	0xb8f61605, 0x80768176,
+	0x8e768676, 0x80727672,
+	0xbefa0084, 0xbefa00ff,
+	0x01000000, 0xc0211cfc,
+	0x00000072, 0x80728472,
+	0xc0211c3c, 0x00000072,
+	0x80728472, 0xc0211c7c,
+	0x00000072, 0x80728472,
+	0xc0211bbc, 0x00000072,
+	0x80728472, 0xc0211bfc,
+	0x00000072, 0x80728472,
+	0xc0211d3c, 0x00000072,
+	0x80728472, 0xc0211d7c,
+	0x00000072, 0x80728472,
+	0xc0211a3c, 0x00000072,
+	0x80728472, 0xc0211a7c,
+	0x00000072, 0x80728472,
+	0xc0211dfc, 0x00000072,
+	0x80728472, 0xc0211b3c,
+	0x00000072, 0x80728472,
+	0xc0211b7c, 0x00000072,
+	0x80728472, 0xbf8c007f,
+	0xbefc0073, 0xbefe006e,
+	0xbeff006f, 0x867375ff,
+	0x000003ff, 0xb9734803,
+	0x867375ff, 0xfffff800,
+	0x8f738b73, 0xb973a2c3,
+	0xb977f801, 0x8673ff71,
+	0xf0000000, 0x8f739c73,
+	0x8e739073, 0xbef60080,
+	0x87767376, 0x8673ff71,
+	0x08000000, 0x8f739b73,
+	0x8e738f73, 0x87767376,
+	0x8673ff74, 0x00800000,
+	0x8f739773, 0xb976f807,
+	0x8671ff71, 0x0000ffff,
+	0x86fe7e7e, 0x86ea6a6a,
+	0xb974f802, 0xbf8a0000,
+	0x95807370, 0xbf810000,
+};
+
+
+static const uint32_t cwsr_trap_gfx9_hex[] = {
+	0xbf820001, 0xbf82015a,
+	0xb8f8f802, 0x89788678,
+	0xb8f1f803, 0x866eff71,
+	0x00000400, 0xbf850034,
+	0x866eff71, 0x00000800,
+	0xbf850003, 0x866eff71,
+	0x00000100, 0xbf840008,
+	0x866eff78, 0x00002000,
+	0xbf840001, 0xbf810000,
+	0x8778ff78, 0x00002000,
+	0x80ec886c, 0x82ed806d,
+	0xb8eef807, 0x866fff6e,
+	0x001f8000, 0x8e6f8b6f,
+	0x8977ff77, 0xfc000000,
+	0x87776f77, 0x896eff6e,
+	0x001f8000, 0xb96ef807,
+	0xb8f0f812, 0xb8f1f813,
+	0x8ef08870, 0xc0071bb8,
+	0x00000000, 0xbf8cc07f,
+	0xc0071c38, 0x00000008,
+	0xbf8cc07f, 0x86ee6e6e,
+	0xbf840001, 0xbe801d6e,
+	0xb8f1f803, 0x8671ff71,
+	0x000001ff, 0xbf850002,
+	0x806c846c, 0x826d806d,
+	0x866dff6d, 0x0000ffff,
+	0x8f6e8b77, 0x866eff6e,
+	0x001f8000, 0xb96ef807,
+	0x86fe7e7e, 0x86ea6a6a,
+	0xb978f802, 0xbe801f6c,
+	0x866dff6d, 0x0000ffff,
+	0xbef00080, 0xb9700283,
+	0xb8f02407, 0x8e709c70,
+	0x876d706d, 0xb8f003c7,
+	0x8e709b70, 0x876d706d,
+	0xb8f0f807, 0x8670ff70,
+	0x00007fff, 0xb970f807,
+	0xbeee007e, 0xbeef007f,
+	0xbefe0180, 0xbf900004,
+	0x87708478, 0xb970f802,
+	0xbf8e0002, 0xbf88fffe,
+	0xb8f02a05, 0x80708170,
+	0x8e708a70, 0xb8f11605,
+	0x80718171, 0x8e718671,
+	0x80707170, 0x80707e70,
+	0x8271807f, 0x8671ff71,
+	0x0000ffff, 0xc0471cb8,
+	0x00000040, 0xbf8cc07f,
+	0xc04b1d38, 0x00000048,
+	0xbf8cc07f, 0xc0431e78,
+	0x00000058, 0xbf8cc07f,
+	0xc0471eb8, 0x0000005c,
+	0xbf8cc07f, 0xbef4007e,
+	0x8675ff7f, 0x0000ffff,
+	0x8775ff75, 0x00040000,
+	0xbef60080, 0xbef700ff,
+	0x00807fac, 0x8670ff7f,
+	0x08000000, 0x8f708370,
+	0x87777077, 0x8670ff7f,
+	0x70000000, 0x8f708170,
+	0x87777077, 0xbefb007c,
+	0xbefa0080, 0xb8fa2a05,
+	0x807a817a, 0x8e7a8a7a,
+	0xb8f01605, 0x80708170,
+	0x8e708670, 0x807a707a,
+	0xbef60084, 0xbef600ff,
+	0x01000000, 0xbefe007c,
+	0xbefc007a, 0xc0611efa,
+	0x0000007c, 0xbf8cc07f,
+	0x807a847a, 0xbefc007e,
+	0xbefe007c, 0xbefc007a,
+	0xc0611b3a, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xbefe007c,
+	0xbefc007a, 0xc0611b7a,
+	0x0000007c, 0xbf8cc07f,
+	0x807a847a, 0xbefc007e,
+	0xbefe007c, 0xbefc007a,
+	0xc0611bba, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xbefe007c,
+	0xbefc007a, 0xc0611bfa,
+	0x0000007c, 0xbf8cc07f,
+	0x807a847a, 0xbefc007e,
+	0xbefe007c, 0xbefc007a,
+	0xc0611e3a, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xb8f1f803,
+	0xbefe007c, 0xbefc007a,
+	0xc0611c7a, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xbefe007c,
+	0xbefc007a, 0xc0611a3a,
+	0x0000007c, 0xbf8cc07f,
+	0x807a847a, 0xbefc007e,
+	0xbefe007c, 0xbefc007a,
+	0xc0611a7a, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0xb8fbf801,
+	0xbefe007c, 0xbefc007a,
+	0xc0611efa, 0x0000007c,
+	0xbf8cc07f, 0x807a847a,
+	0xbefc007e, 0x8670ff7f,
+	0x04000000, 0xbeef0080,
+	0x876f6f70, 0xb8fa2a05,
+	0x807a817a, 0x8e7a8a7a,
+	0xb8f11605, 0x80718171,
+	0x8e718471, 0x8e768271,
+	0xbef600ff, 0x01000000,
+	0xbef20174, 0x80747a74,
+	0x82758075, 0xbefc0080,
+	0xbf800000, 0xbe802b00,
+	0xbe822b02, 0xbe842b04,
+	0xbe862b06, 0xbe882b08,
+	0xbe8a2b0a, 0xbe8c2b0c,
+	0xbe8e2b0e, 0xc06b003a,
+	0x00000000, 0xbf8cc07f,
+	0xc06b013a, 0x00000010,
+	0xbf8cc07f, 0xc06b023a,
+	0x00000020, 0xbf8cc07f,
+	0xc06b033a, 0x00000030,
+	0xbf8cc07f, 0x8074c074,
+	0x82758075, 0x807c907c,
+	0xbf0a717c, 0xbf85ffe7,
+	0xbef40172, 0xbefa0080,
+	0xbefe00c1, 0xbeff00c1,
+	0xbee80080, 0xbee90080,
+	0xbef600ff, 0x01000000,
+	0xe0724000, 0x7a1d0000,
+	0xe0724100, 0x7a1d0100,
+	0xe0724200, 0x7a1d0200,
+	0xe0724300, 0x7a1d0300,
+	0xbefe00c1, 0xbeff00c1,
+	0xb8f14306, 0x8671c171,
+	0xbf84002c, 0xbf8a0000,
+	0x8670ff6f, 0x04000000,
+	0xbf840028, 0x8e718671,
+	0x8e718271, 0xbef60071,
+	0xb8fa2a05, 0x807a817a,
+	0x8e7a8a7a, 0xb8f01605,
+	0x80708170, 0x8e708670,
+	0x807a707a, 0x807aff7a,
+	0x00000080, 0xbef600ff,
+	0x01000000, 0xbefc0080,
+	0xd28c0002, 0x000100c1,
+	0xd28d0003, 0x000204c1,
+	0xd1060002, 0x00011103,
+	0x7e0602ff, 0x00000200,
+	0xbefc00ff, 0x00010000,
+	0xbe800077, 0x8677ff77,
+	0xff7fffff, 0x8777ff77,
+	0x00058000, 0xd8ec0000,
+	0x00000002, 0xbf8cc07f,
+	0xe0765000, 0x7a1d0002,
+	0x68040702, 0xd0c9006a,
+	0x0000e302, 0xbf87fff7,
+	0xbef70000, 0xbefa00ff,
+	0x00000400, 0xbefe00c1,
+	0xbeff00c1, 0xb8f12a05,
+	0x80718171, 0x8e718271,
+	0x8e768871, 0xbef600ff,
+	0x01000000, 0xbefc0084,
+	0xbf0a717c, 0xbf840015,
+	0xbf11017c, 0x8071ff71,
+	0x00001000, 0x7e000300,
+	0x7e020301, 0x7e040302,
+	0x7e060303, 0xe0724000,
+	0x7a1d0000, 0xe0724100,
+	0x7a1d0100, 0xe0724200,
+	0x7a1d0200, 0xe0724300,
+	0x7a1d0300, 0x807c847c,
+	0x807aff7a, 0x00000400,
+	0xbf0a717c, 0xbf85ffef,
+	0xbf9c0000, 0xbf8200d9,
+	0xbef4007e, 0x8675ff7f,
+	0x0000ffff, 0x8775ff75,
+	0x00040000, 0xbef60080,
+	0xbef700ff, 0x00807fac,
+	0x866eff7f, 0x08000000,
+	0x8f6e836e, 0x87776e77,
+	0x866eff7f, 0x70000000,
+	0x8f6e816e, 0x87776e77,
+	0x866eff7f, 0x04000000,
+	0xbf84001e, 0xbefe00c1,
+	0xbeff00c1, 0xb8ef4306,
+	0x866fc16f, 0xbf840019,
+	0x8e6f866f, 0x8e6f826f,
+	0xbef6006f, 0xb8f82a05,
+	0x80788178, 0x8e788a78,
+	0xb8ee1605, 0x806e816e,
+	0x8e6e866e, 0x80786e78,
+	0x8078ff78, 0x00000080,
+	0xbef600ff, 0x01000000,
+	0xbefc0080, 0xe0510000,
+	0x781d0000, 0xe0510100,
+	0x781d0000, 0x807cff7c,
+	0x00000200, 0x8078ff78,
+	0x00000200, 0xbf0a6f7c,
+	0xbf85fff6, 0xbef80080,
+	0xbefe00c1, 0xbeff00c1,
+	0xb8ef2a05, 0x806f816f,
+	0x8e6f826f, 0x8e76886f,
+	0xbef600ff, 0x01000000,
+	0xbeee0078, 0x8078ff78,
+	0x00000400, 0xbefc0084,
+	0xbf11087c, 0x806fff6f,
+	0x00008000, 0xe0524000,
+	0x781d0000, 0xe0524100,
+	0x781d0100, 0xe0524200,
+	0x781d0200, 0xe0524300,
+	0x781d0300, 0xbf8c0f70,
+	0x7e000300, 0x7e020301,
+	0x7e040302, 0x7e060303,
+	0x807c847c, 0x8078ff78,
+	0x00000400, 0xbf0a6f7c,
+	0xbf85ffee, 0xbf9c0000,
+	0xe0524000, 0x6e1d0000,
+	0xe0524100, 0x6e1d0100,
+	0xe0524200, 0x6e1d0200,
+	0xe0524300, 0x6e1d0300,
+	0xb8f82a05, 0x80788178,
+	0x8e788a78, 0xb8ee1605,
+	0x806e816e, 0x8e6e866e,
+	0x80786e78, 0x80f8c078,
+	0xb8ef1605, 0x806f816f,
+	0x8e6f846f, 0x8e76826f,
+	0xbef600ff, 0x01000000,
+	0xbefc006f, 0xc031003a,
+	0x00000078, 0x80f8c078,
+	0xbf8cc07f, 0x80fc907c,
+	0xbf800000, 0xbe802d00,
+	0xbe822d02, 0xbe842d04,
+	0xbe862d06, 0xbe882d08,
+	0xbe8a2d0a, 0xbe8c2d0c,
+	0xbe8e2d0e, 0xbf06807c,
+	0xbf84fff0, 0xb8f82a05,
+	0x80788178, 0x8e788a78,
+	0xb8ee1605, 0x806e816e,
+	0x8e6e866e, 0x80786e78,
+	0xbef60084, 0xbef600ff,
+	0x01000000, 0xc0211bfa,
+	0x00000078, 0x80788478,
+	0xc0211b3a, 0x00000078,
+	0x80788478, 0xc0211b7a,
+	0x00000078, 0x80788478,
+	0xc0211eba, 0x00000078,
+	0x80788478, 0xc0211efa,
+	0x00000078, 0x80788478,
+	0xc0211c3a, 0x00000078,
+	0x80788478, 0xc0211c7a,
+	0x00000078, 0x80788478,
+	0xc0211a3a, 0x00000078,
+	0x80788478, 0xc0211a7a,
+	0x00000078, 0x80788478,
+	0xc0211cfa, 0x00000078,
+	0x80788478, 0xbf8cc07f,
+	0xbefc006f, 0xbefe007a,
+	0xbeff007b, 0x866f71ff,
+	0x000003ff, 0xb96f4803,
+	0x866f71ff, 0xfffff800,
+	0x8f6f8b6f, 0xb96fa2c3,
+	0xb973f801, 0xb8ee2a05,
+	0x806e816e, 0x8e6e8a6e,
+	0xb8ef1605, 0x806f816f,
+	0x8e6f866f, 0x806e6f6e,
+	0x806e746e, 0x826f8075,
+	0x866fff6f, 0x0000ffff,
+	0xc0071cb7, 0x00000040,
+	0xc00b1d37, 0x00000048,
+	0xc0031e77, 0x00000058,
+	0xc0071eb7, 0x0000005c,
+	0xbf8cc07f, 0x866fff6d,
+	0xf0000000, 0x8f6f9c6f,
+	0x8e6f906f, 0xbeee0080,
+	0x876e6f6e, 0x866fff6d,
+	0x08000000, 0x8f6f9b6f,
+	0x8e6f8f6f, 0x876e6f6e,
+	0x866fff70, 0x00800000,
+	0x8f6f976f, 0xb96ef807,
+	0x866dff6d, 0x0000ffff,
+	0x86fe7e7e, 0x86ea6a6a,
+	0xb970f802, 0xbf8a0000,
+	0x95806f6c, 0xbf810000,
+};
diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm
index 997a383dcb8b..a2a04bb64096 100644
--- a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx8.asm
@@ -20,9 +20,12 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
-#if 0
-HW (VI) source code for CWSR trap handler
-#Version 18 + multiple trap handler
+/* To compile this assembly code:
+ * PROJECT=vi ./sp3 cwsr_trap_handler_gfx8.asm -hex tmp.hex
+ */
+
+/* HW (VI) source code for CWSR trap handler */
+/* Version 18 + multiple trap handler */
 
 // this performance-optimal version was originally from Seven Xu at SRDC
 
@@ -98,6 +101,7 @@ var SWIZZLE_EN                      =   0                   //whether we use swi
 /**************************************************************************/
 var SQ_WAVE_STATUS_INST_ATC_SHIFT  = 23
 var SQ_WAVE_STATUS_INST_ATC_MASK   = 0x00800000
+var SQ_WAVE_STATUS_SPI_PRIO_SHIFT  = 1
 var SQ_WAVE_STATUS_SPI_PRIO_MASK   = 0x00000006
 
 var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT    = 12
@@ -149,7 +153,7 @@ var s_save_spi_init_lo              =   exec_lo
 var s_save_spi_init_hi              =   exec_hi
 
                                                 //tba_lo and tba_hi need to be saved/restored
-var s_save_pc_lo            =   ttmp0           //{TTMP1, TTMP0} = {3??h0,pc_rewind[3:0], HT[0],trapID[7:0], PC[47:0]}
+var s_save_pc_lo            =   ttmp0           //{TTMP1, TTMP0} = {3'h0,pc_rewind[3:0], HT[0],trapID[7:0], PC[47:0]}
 var s_save_pc_hi            =   ttmp1
 var s_save_exec_lo          =   ttmp2
 var s_save_exec_hi          =   ttmp3
@@ -319,6 +323,10 @@ end
         s_sendmsg   sendmsg(MSG_SAVEWAVE)  //send SPI a message and wait for SPI's write to EXEC
     end
 
+    // Set SPI_PRIO=2 to avoid starving instruction fetch in the waves we're waiting for.
+    s_or_b32 s_save_tmp, s_save_status, (2 << SQ_WAVE_STATUS_SPI_PRIO_SHIFT)
+    s_setreg_b32 hwreg(HW_REG_STATUS), s_save_tmp
+
   L_SLEEP:
     s_sleep 0x2                // sleep 1 (64clk) is not enough for 8 waves per SIMD, which will cause SQ hang, since the 7,8th wave could not get arbit to exec inst, while other waves are stuck into the sleep-loop and waiting for wrexec!=0
 
@@ -1007,8 +1015,6 @@ end
 
     s_waitcnt       lgkmcnt(0)                                                                                      //from now on, it is safe to restore STATUS and IB_STS
 
-    s_and_b32 s_restore_pc_hi, s_restore_pc_hi, 0x0000ffff      //pc[47:32]        //Do it here in order not to affect STATUS
-
     //for normal save & restore, the saved PC points to the next inst to execute, no adjustment needs to be made, otherwise:
     if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_RESTORE_NORMAL))
         s_add_u32 s_restore_pc_lo, s_restore_pc_lo, 8            //pc[31:0]+8     //two back-to-back s_trap are used (first for save and second for restore)
@@ -1044,6 +1050,7 @@ end
     s_lshr_b32      s_restore_m0, s_restore_m0, SQ_WAVE_STATUS_INST_ATC_SHIFT
     s_setreg_b32    hwreg(HW_REG_IB_STS),   s_restore_tmp
 
+    s_and_b32 s_restore_pc_hi, s_restore_pc_hi, 0x0000ffff      //pc[47:32]        //Do it here in order not to affect STATUS
     s_and_b64    exec, exec, exec  // Restore STATUS.EXECZ, not writable by s_setreg_b32
     s_and_b64    vcc, vcc, vcc  // Restore STATUS.VCCZ, not writable by s_setreg_b32
     s_setreg_b32    hwreg(HW_REG_STATUS),   s_restore_status     // SCC is included, which is changed by previous salu
@@ -1127,258 +1134,3 @@ end
 function get_hwreg_size_bytes
     return 128 //HWREG size 128 bytes
 end
-
-
-#endif
-
-static const uint32_t cwsr_trap_gfx8_hex[] = {
-	0xbf820001, 0xbf820123,
-	0xb8f4f802, 0x89748674,
-	0xb8f5f803, 0x8675ff75,
-	0x00000400, 0xbf850011,
-	0xc00a1e37, 0x00000000,
-	0xbf8c007f, 0x87777978,
-	0xbf840002, 0xb974f802,
-	0xbe801d78, 0xb8f5f803,
-	0x8675ff75, 0x000001ff,
-	0xbf850002, 0x80708470,
-	0x82718071, 0x8671ff71,
-	0x0000ffff, 0xb974f802,
-	0xbe801f70, 0xb8f5f803,
-	0x8675ff75, 0x00000100,
-	0xbf840006, 0xbefa0080,
-	0xb97a0203, 0x8671ff71,
-	0x0000ffff, 0x80f08870,
-	0x82f18071, 0xbefa0080,
-	0xb97a0283, 0xbef60068,
-	0xbef70069, 0xb8fa1c07,
-	0x8e7a9c7a, 0x87717a71,
-	0xb8fa03c7, 0x8e7a9b7a,
-	0x87717a71, 0xb8faf807,
-	0x867aff7a, 0x00007fff,
-	0xb97af807, 0xbef2007e,
-	0xbef3007f, 0xbefe0180,
-	0xbf900004, 0xbf8e0002,
-	0xbf88fffe, 0xbef8007e,
-	0x8679ff7f, 0x0000ffff,
-	0x8779ff79, 0x00040000,
-	0xbefa0080, 0xbefb00ff,
-	0x00807fac, 0x867aff7f,
-	0x08000000, 0x8f7a837a,
-	0x877b7a7b, 0x867aff7f,
-	0x70000000, 0x8f7a817a,
-	0x877b7a7b, 0xbeef007c,
-	0xbeee0080, 0xb8ee2a05,
-	0x806e816e, 0x8e6e8a6e,
-	0xb8fa1605, 0x807a817a,
-	0x8e7a867a, 0x806e7a6e,
-	0xbefa0084, 0xbefa00ff,
-	0x01000000, 0xbefe007c,
-	0xbefc006e, 0xc0611bfc,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xbefe007c,
-	0xbefc006e, 0xc0611c3c,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xbefe007c,
-	0xbefc006e, 0xc0611c7c,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xbefe007c,
-	0xbefc006e, 0xc0611cbc,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xbefe007c,
-	0xbefc006e, 0xc0611cfc,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xbefe007c,
-	0xbefc006e, 0xc0611d3c,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xb8f5f803,
-	0xbefe007c, 0xbefc006e,
-	0xc0611d7c, 0x0000007c,
-	0x806e846e, 0xbefc007e,
-	0xbefe007c, 0xbefc006e,
-	0xc0611dbc, 0x0000007c,
-	0x806e846e, 0xbefc007e,
-	0xbefe007c, 0xbefc006e,
-	0xc0611dfc, 0x0000007c,
-	0x806e846e, 0xbefc007e,
-	0xb8eff801, 0xbefe007c,
-	0xbefc006e, 0xc0611bfc,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xbefe007c,
-	0xbefc006e, 0xc0611b3c,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0xbefe007c,
-	0xbefc006e, 0xc0611b7c,
-	0x0000007c, 0x806e846e,
-	0xbefc007e, 0x867aff7f,
-	0x04000000, 0xbef30080,
-	0x8773737a, 0xb8ee2a05,
-	0x806e816e, 0x8e6e8a6e,
-	0xb8f51605, 0x80758175,
-	0x8e758475, 0x8e7a8275,
-	0xbefa00ff, 0x01000000,
-	0xbef60178, 0x80786e78,
-	0x82798079, 0xbefc0080,
-	0xbe802b00, 0xbe822b02,
-	0xbe842b04, 0xbe862b06,
-	0xbe882b08, 0xbe8a2b0a,
-	0xbe8c2b0c, 0xbe8e2b0e,
-	0xc06b003c, 0x00000000,
-	0xc06b013c, 0x00000010,
-	0xc06b023c, 0x00000020,
-	0xc06b033c, 0x00000030,
-	0x8078c078, 0x82798079,
-	0x807c907c, 0xbf0a757c,
-	0xbf85ffeb, 0xbef80176,
-	0xbeee0080, 0xbefe00c1,
-	0xbeff00c1, 0xbefa00ff,
-	0x01000000, 0xe0724000,
-	0x6e1e0000, 0xe0724100,
-	0x6e1e0100, 0xe0724200,
-	0x6e1e0200, 0xe0724300,
-	0x6e1e0300, 0xbefe00c1,
-	0xbeff00c1, 0xb8f54306,
-	0x8675c175, 0xbf84002c,
-	0xbf8a0000, 0x867aff73,
-	0x04000000, 0xbf840028,
-	0x8e758675, 0x8e758275,
-	0xbefa0075, 0xb8ee2a05,
-	0x806e816e, 0x8e6e8a6e,
-	0xb8fa1605, 0x807a817a,
-	0x8e7a867a, 0x806e7a6e,
-	0x806eff6e, 0x00000080,
-	0xbefa00ff, 0x01000000,
-	0xbefc0080, 0xd28c0002,
-	0x000100c1, 0xd28d0003,
-	0x000204c1, 0xd1060002,
-	0x00011103, 0x7e0602ff,
-	0x00000200, 0xbefc00ff,
-	0x00010000, 0xbe80007b,
-	0x867bff7b, 0xff7fffff,
-	0x877bff7b, 0x00058000,
-	0xd8ec0000, 0x00000002,
-	0xbf8c007f, 0xe0765000,
-	0x6e1e0002, 0x32040702,
-	0xd0c9006a, 0x0000eb02,
-	0xbf87fff7, 0xbefb0000,
-	0xbeee00ff, 0x00000400,
-	0xbefe00c1, 0xbeff00c1,
-	0xb8f52a05, 0x80758175,
-	0x8e758275, 0x8e7a8875,
-	0xbefa00ff, 0x01000000,
-	0xbefc0084, 0xbf0a757c,
-	0xbf840015, 0xbf11017c,
-	0x8075ff75, 0x00001000,
-	0x7e000300, 0x7e020301,
-	0x7e040302, 0x7e060303,
-	0xe0724000, 0x6e1e0000,
-	0xe0724100, 0x6e1e0100,
-	0xe0724200, 0x6e1e0200,
-	0xe0724300, 0x6e1e0300,
-	0x807c847c, 0x806eff6e,
-	0x00000400, 0xbf0a757c,
-	0xbf85ffef, 0xbf9c0000,
-	0xbf8200ca, 0xbef8007e,
-	0x8679ff7f, 0x0000ffff,
-	0x8779ff79, 0x00040000,
-	0xbefa0080, 0xbefb00ff,
-	0x00807fac, 0x8676ff7f,
-	0x08000000, 0x8f768376,
-	0x877b767b, 0x8676ff7f,
-	0x70000000, 0x8f768176,
-	0x877b767b, 0x8676ff7f,
-	0x04000000, 0xbf84001e,
-	0xbefe00c1, 0xbeff00c1,
-	0xb8f34306, 0x8673c173,
-	0xbf840019, 0x8e738673,
-	0x8e738273, 0xbefa0073,
-	0xb8f22a05, 0x80728172,
-	0x8e728a72, 0xb8f61605,
-	0x80768176, 0x8e768676,
-	0x80727672, 0x8072ff72,
-	0x00000080, 0xbefa00ff,
-	0x01000000, 0xbefc0080,
-	0xe0510000, 0x721e0000,
-	0xe0510100, 0x721e0000,
-	0x807cff7c, 0x00000200,
-	0x8072ff72, 0x00000200,
-	0xbf0a737c, 0xbf85fff6,
-	0xbef20080, 0xbefe00c1,
-	0xbeff00c1, 0xb8f32a05,
-	0x80738173, 0x8e738273,
-	0x8e7a8873, 0xbefa00ff,
-	0x01000000, 0xbef60072,
-	0x8072ff72, 0x00000400,
-	0xbefc0084, 0xbf11087c,
-	0x8073ff73, 0x00008000,
-	0xe0524000, 0x721e0000,
-	0xe0524100, 0x721e0100,
-	0xe0524200, 0x721e0200,
-	0xe0524300, 0x721e0300,
-	0xbf8c0f70, 0x7e000300,
-	0x7e020301, 0x7e040302,
-	0x7e060303, 0x807c847c,
-	0x8072ff72, 0x00000400,
-	0xbf0a737c, 0xbf85ffee,
-	0xbf9c0000, 0xe0524000,
-	0x761e0000, 0xe0524100,
-	0x761e0100, 0xe0524200,
-	0x761e0200, 0xe0524300,
-	0x761e0300, 0xb8f22a05,
-	0x80728172, 0x8e728a72,
-	0xb8f61605, 0x80768176,
-	0x8e768676, 0x80727672,
-	0x80f2c072, 0xb8f31605,
-	0x80738173, 0x8e738473,
-	0x8e7a8273, 0xbefa00ff,
-	0x01000000, 0xbefc0073,
-	0xc031003c, 0x00000072,
-	0x80f2c072, 0xbf8c007f,
-	0x80fc907c, 0xbe802d00,
-	0xbe822d02, 0xbe842d04,
-	0xbe862d06, 0xbe882d08,
-	0xbe8a2d0a, 0xbe8c2d0c,
-	0xbe8e2d0e, 0xbf06807c,
-	0xbf84fff1, 0xb8f22a05,
-	0x80728172, 0x8e728a72,
-	0xb8f61605, 0x80768176,
-	0x8e768676, 0x80727672,
-	0xbefa0084, 0xbefa00ff,
-	0x01000000, 0xc0211cfc,
-	0x00000072, 0x80728472,
-	0xc0211c3c, 0x00000072,
-	0x80728472, 0xc0211c7c,
-	0x00000072, 0x80728472,
-	0xc0211bbc, 0x00000072,
-	0x80728472, 0xc0211bfc,
-	0x00000072, 0x80728472,
-	0xc0211d3c, 0x00000072,
-	0x80728472, 0xc0211d7c,
-	0x00000072, 0x80728472,
-	0xc0211a3c, 0x00000072,
-	0x80728472, 0xc0211a7c,
-	0x00000072, 0x80728472,
-	0xc0211dfc, 0x00000072,
-	0x80728472, 0xc0211b3c,
-	0x00000072, 0x80728472,
-	0xc0211b7c, 0x00000072,
-	0x80728472, 0xbf8c007f,
-	0x8671ff71, 0x0000ffff,
-	0xbefc0073, 0xbefe006e,
-	0xbeff006f, 0x867375ff,
-	0x000003ff, 0xb9734803,
-	0x867375ff, 0xfffff800,
-	0x8f738b73, 0xb973a2c3,
-	0xb977f801, 0x8673ff71,
-	0xf0000000, 0x8f739c73,
-	0x8e739073, 0xbef60080,
-	0x87767376, 0x8673ff71,
-	0x08000000, 0x8f739b73,
-	0x8e738f73, 0x87767376,
-	0x8673ff74, 0x00800000,
-	0x8f739773, 0xb976f807,
-	0x86fe7e7e, 0x86ea6a6a,
-	0xb974f802, 0xbf8a0000,
-	0x95807370, 0xbf810000,
-};
-
diff --git a/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm
new file mode 100644
index 000000000000..998be96be736
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/cwsr_trap_handler_gfx9.asm
@@ -0,0 +1,1214 @@
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+/* To compile this assembly code:
+ * PROJECT=greenland ./sp3 cwsr_trap_handler_gfx9.asm -hex tmp.hex
+ */
+
+/* HW (GFX9) source code for CWSR trap handler */
+/* Version 18 + multiple trap handler */
+
+// this performance-optimal version was originally from Seven Xu at SRDC
+
+// Revison #18	 --...
+/* Rev History
+** #1. Branch from gc dv.   //gfxip/gfx9/main/src/test/suites/block/cs/sr/cs_trap_handler.sp3#1,#50, #51, #52-53(Skip, Already Fixed by PV), #54-56(merged),#57-58(mergerd, skiped-already fixed by PV)
+** #4. SR Memory Layout:
+**			 1. VGPR-SGPR-HWREG-{LDS}
+**			 2. tba_hi.bits.26 - reconfigured as the first wave in tg bits, for defer Save LDS for a threadgroup.. performance concern..
+** #5. Update: 1. Accurate g8sr_ts_save_d timestamp
+** #6. Update: 1. Fix s_barrier usage; 2. VGPR s/r using swizzle buffer?(NoNeed, already matched the swizzle pattern, more investigation)
+** #7. Update: 1. don't barrier if noLDS
+** #8. Branch: 1. Branch to ver#0, which is very similar to gc dv version
+**	       2. Fix SQ issue by s_sleep 2
+** #9. Update: 1. Fix scc restore failed issue, restore wave_status at last
+**	       2. optimize s_buffer save by burst 16sgprs...
+** #10. Update 1. Optimize restore sgpr by busrt 16 sgprs.
+** #11. Update 1. Add 2 more timestamp for debug version
+** #12. Update 1. Add VGPR SR using DWx4, some case improve and some case drop performance
+** #13. Integ  1. Always use MUBUF for PV trap shader...
+** #14. Update 1. s_buffer_store soft clause...
+** #15. Update 1. PERF - sclar write with glc:0/mtype0 to allow L2 combine. perf improvement a lot.
+** #16. Update 1. PRRF - UNROLL LDS_DMA got 2500cycle save in IP tree
+** #17. Update 1. FUNC - LDS_DMA has issues while ATC, replace with ds_read/buffer_store for save part[TODO restore part]
+**	       2. PERF - Save LDS before save VGPR to cover LDS save long latency...
+** #18. Update 1. FUNC - Implicitly estore STATUS.VCCZ, which is not writable by s_setreg_b32
+**	       2. FUNC - Handle non-CWSR traps
+*/
+
+var G8SR_WDMEM_HWREG_OFFSET = 0
+var G8SR_WDMEM_SGPR_OFFSET  = 128  // in bytes
+
+// Keep definition same as the app shader, These 2 time stamps are part of the app shader... Should before any Save and after restore.
+
+var G8SR_DEBUG_TIMESTAMP = 0
+var G8SR_DEBUG_TS_SAVE_D_OFFSET = 40*4	// ts_save_d timestamp offset relative to SGPR_SR_memory_offset
+var s_g8sr_ts_save_s	= s[34:35]   // save start
+var s_g8sr_ts_sq_save_msg  = s[36:37]	// The save shader send SAVEWAVE msg to spi
+var s_g8sr_ts_spi_wrexec   = s[38:39]	// the SPI write the sr address to SQ
+var s_g8sr_ts_save_d	= s[40:41]   // save end
+var s_g8sr_ts_restore_s = s[42:43]   // restore start
+var s_g8sr_ts_restore_d = s[44:45]   // restore end
+
+var G8SR_VGPR_SR_IN_DWX4 = 0
+var G8SR_SAVE_BUF_RSRC_WORD1_STRIDE_DWx4 = 0x00100000	 // DWx4 stride is 4*4Bytes
+var G8SR_RESTORE_BUF_RSRC_WORD1_STRIDE_DWx4  = G8SR_SAVE_BUF_RSRC_WORD1_STRIDE_DWx4
+
+
+/*************************************************************************/
+/*		    control on how to run the shader			 */
+/*************************************************************************/
+//any hack that needs to be made to run this code in EMU (either because various EMU code are not ready or no compute save & restore in EMU run)
+var EMU_RUN_HACK		    =	0
+var EMU_RUN_HACK_RESTORE_NORMAL	    =	0
+var EMU_RUN_HACK_SAVE_NORMAL_EXIT   =	0
+var EMU_RUN_HACK_SAVE_SINGLE_WAVE   =	0
+var EMU_RUN_HACK_SAVE_FIRST_TIME    =	0		    //for interrupted restore in which the first save is through EMU_RUN_HACK
+var SAVE_LDS			    =	1
+var WG_BASE_ADDR_LO		    =	0x9000a000
+var WG_BASE_ADDR_HI		    =	0x0
+var WAVE_SPACE			    =	0x5000		    //memory size that each wave occupies in workgroup state mem
+var CTX_SAVE_CONTROL		    =	0x0
+var CTX_RESTORE_CONTROL		    =	CTX_SAVE_CONTROL
+var SIM_RUN_HACK		    =	0		    //any hack that needs to be made to run this code in SIM (either because various RTL code are not ready or no compute save & restore in RTL run)
+var SGPR_SAVE_USE_SQC		    =	1		    //use SQC D$ to do the write
+var USE_MTBUF_INSTEAD_OF_MUBUF	    =	0		    //because TC EMU currently asserts on 0 of // overload DFMT field to carry 4 more bits of stride for MUBUF opcodes
+var SWIZZLE_EN			    =	0		    //whether we use swizzled buffer addressing
+var ACK_SQC_STORE		    =	1		    //workaround for suspected SQC store bug causing incorrect stores under concurrency
+
+/**************************************************************************/
+/*			variables					  */
+/**************************************************************************/
+var SQ_WAVE_STATUS_INST_ATC_SHIFT  = 23
+var SQ_WAVE_STATUS_INST_ATC_MASK   = 0x00800000
+var SQ_WAVE_STATUS_SPI_PRIO_SHIFT  = 1
+var SQ_WAVE_STATUS_SPI_PRIO_MASK   = 0x00000006
+var SQ_WAVE_STATUS_HALT_MASK       = 0x2000
+
+var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT	= 12
+var SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE	= 9
+var SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SHIFT	= 8
+var SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SIZE	= 6
+var SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SHIFT	= 24
+var SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SIZE	= 3			//FIXME	 sq.blk still has 4 bits at this time while SQ programming guide has 3 bits
+
+var SQ_WAVE_TRAPSTS_SAVECTX_MASK    =	0x400
+var SQ_WAVE_TRAPSTS_EXCE_MASK	    =	0x1FF			// Exception mask
+var SQ_WAVE_TRAPSTS_SAVECTX_SHIFT   =	10
+var SQ_WAVE_TRAPSTS_MEM_VIOL_MASK   =	0x100
+var SQ_WAVE_TRAPSTS_MEM_VIOL_SHIFT  =	8
+var SQ_WAVE_TRAPSTS_PRE_SAVECTX_MASK	=   0x3FF
+var SQ_WAVE_TRAPSTS_PRE_SAVECTX_SHIFT	=   0x0
+var SQ_WAVE_TRAPSTS_PRE_SAVECTX_SIZE	=   10
+var SQ_WAVE_TRAPSTS_POST_SAVECTX_MASK	=   0xFFFFF800
+var SQ_WAVE_TRAPSTS_POST_SAVECTX_SHIFT	=   11
+var SQ_WAVE_TRAPSTS_POST_SAVECTX_SIZE	=   21
+var SQ_WAVE_TRAPSTS_ILLEGAL_INST_MASK	=   0x800
+
+var SQ_WAVE_IB_STS_RCNT_SHIFT		=   16			//FIXME
+var SQ_WAVE_IB_STS_FIRST_REPLAY_SHIFT	=   15			//FIXME
+var SQ_WAVE_IB_STS_RCNT_FIRST_REPLAY_MASK	= 0x1F8000
+var SQ_WAVE_IB_STS_RCNT_FIRST_REPLAY_MASK_NEG	= 0x00007FFF	//FIXME
+
+var SQ_BUF_RSRC_WORD1_ATC_SHIFT	    =	24
+var SQ_BUF_RSRC_WORD3_MTYPE_SHIFT   =	27
+
+var TTMP11_SAVE_RCNT_FIRST_REPLAY_SHIFT	=   26			// bits [31:26] unused by SPI debug data
+var TTMP11_SAVE_RCNT_FIRST_REPLAY_MASK	=   0xFC000000
+
+/*	Save	    */
+var S_SAVE_BUF_RSRC_WORD1_STRIDE	=   0x00040000		//stride is 4 bytes
+var S_SAVE_BUF_RSRC_WORD3_MISC		=   0x00807FAC		//SQ_SEL_X/Y/Z/W, BUF_NUM_FORMAT_FLOAT, (0 for MUBUF stride[17:14] when ADD_TID_ENABLE and BUF_DATA_FORMAT_32 for MTBUF), ADD_TID_ENABLE
+
+var S_SAVE_SPI_INIT_ATC_MASK		=   0x08000000		//bit[27]: ATC bit
+var S_SAVE_SPI_INIT_ATC_SHIFT		=   27
+var S_SAVE_SPI_INIT_MTYPE_MASK		=   0x70000000		//bit[30:28]: Mtype
+var S_SAVE_SPI_INIT_MTYPE_SHIFT		=   28
+var S_SAVE_SPI_INIT_FIRST_WAVE_MASK	=   0x04000000		//bit[26]: FirstWaveInTG
+var S_SAVE_SPI_INIT_FIRST_WAVE_SHIFT	=   26
+
+var S_SAVE_PC_HI_RCNT_SHIFT		=   28			//FIXME	 check with Brian to ensure all fields other than PC[47:0] can be used
+var S_SAVE_PC_HI_RCNT_MASK		=   0xF0000000		//FIXME
+var S_SAVE_PC_HI_FIRST_REPLAY_SHIFT	=   27			//FIXME
+var S_SAVE_PC_HI_FIRST_REPLAY_MASK	=   0x08000000		//FIXME
+
+var s_save_spi_init_lo		    =	exec_lo
+var s_save_spi_init_hi		    =	exec_hi
+
+var s_save_pc_lo	    =	ttmp0		//{TTMP1, TTMP0} = {3'h0,pc_rewind[3:0], HT[0],trapID[7:0], PC[47:0]}
+var s_save_pc_hi	    =	ttmp1
+var s_save_exec_lo	    =	ttmp2
+var s_save_exec_hi	    =	ttmp3
+var s_save_tmp		    =	ttmp4
+var s_save_trapsts	    =	ttmp5		//not really used until the end of the SAVE routine
+var s_save_xnack_mask_lo    =	ttmp6
+var s_save_xnack_mask_hi    =	ttmp7
+var s_save_buf_rsrc0	    =	ttmp8
+var s_save_buf_rsrc1	    =	ttmp9
+var s_save_buf_rsrc2	    =	ttmp10
+var s_save_buf_rsrc3	    =	ttmp11
+var s_save_status	    =	ttmp12
+var s_save_mem_offset	    =	ttmp14
+var s_save_alloc_size	    =	s_save_trapsts		//conflict
+var s_save_m0		    =	ttmp15
+var s_save_ttmps_lo	    =	s_save_tmp		//no conflict
+var s_save_ttmps_hi	    =	s_save_trapsts		//no conflict
+
+/*	Restore	    */
+var S_RESTORE_BUF_RSRC_WORD1_STRIDE	    =	S_SAVE_BUF_RSRC_WORD1_STRIDE
+var S_RESTORE_BUF_RSRC_WORD3_MISC	    =	S_SAVE_BUF_RSRC_WORD3_MISC
+
+var S_RESTORE_SPI_INIT_ATC_MASK		    =	0x08000000	    //bit[27]: ATC bit
+var S_RESTORE_SPI_INIT_ATC_SHIFT	    =	27
+var S_RESTORE_SPI_INIT_MTYPE_MASK	    =	0x70000000	    //bit[30:28]: Mtype
+var S_RESTORE_SPI_INIT_MTYPE_SHIFT	    =	28
+var S_RESTORE_SPI_INIT_FIRST_WAVE_MASK	    =	0x04000000	    //bit[26]: FirstWaveInTG
+var S_RESTORE_SPI_INIT_FIRST_WAVE_SHIFT	    =	26
+
+var S_RESTORE_PC_HI_RCNT_SHIFT		    =	S_SAVE_PC_HI_RCNT_SHIFT
+var S_RESTORE_PC_HI_RCNT_MASK		    =	S_SAVE_PC_HI_RCNT_MASK
+var S_RESTORE_PC_HI_FIRST_REPLAY_SHIFT	    =	S_SAVE_PC_HI_FIRST_REPLAY_SHIFT
+var S_RESTORE_PC_HI_FIRST_REPLAY_MASK	    =	S_SAVE_PC_HI_FIRST_REPLAY_MASK
+
+var s_restore_spi_init_lo		    =	exec_lo
+var s_restore_spi_init_hi		    =	exec_hi
+
+var s_restore_mem_offset	=   ttmp12
+var s_restore_alloc_size	=   ttmp3
+var s_restore_tmp		=   ttmp2
+var s_restore_mem_offset_save	=   s_restore_tmp	//no conflict
+
+var s_restore_m0	    =	s_restore_alloc_size	//no conflict
+
+var s_restore_mode	    =	ttmp7
+
+var s_restore_pc_lo	    =	ttmp0
+var s_restore_pc_hi	    =	ttmp1
+var s_restore_exec_lo	    =	ttmp14
+var s_restore_exec_hi	    = 	ttmp15
+var s_restore_status	    =	ttmp4
+var s_restore_trapsts	    =	ttmp5
+var s_restore_xnack_mask_lo =	xnack_mask_lo
+var s_restore_xnack_mask_hi =	xnack_mask_hi
+var s_restore_buf_rsrc0	    =	ttmp8
+var s_restore_buf_rsrc1	    =	ttmp9
+var s_restore_buf_rsrc2	    =	ttmp10
+var s_restore_buf_rsrc3	    =	ttmp11
+var s_restore_ttmps_lo	    =	s_restore_tmp		//no conflict
+var s_restore_ttmps_hi	    =	s_restore_alloc_size	//no conflict
+
+/**************************************************************************/
+/*			trap handler entry points			  */
+/**************************************************************************/
+/* Shader Main*/
+
+shader main
+  asic(GFX9)
+  type(CS)
+
+
+    if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_RESTORE_NORMAL))		    //hack to use trap_id for determining save/restore
+	//FIXME VCCZ un-init assertion s_getreg_b32	s_save_status, hwreg(HW_REG_STATUS)	    //save STATUS since we will change SCC
+	s_and_b32 s_save_tmp, s_save_pc_hi, 0xffff0000		    //change SCC
+	s_cmp_eq_u32 s_save_tmp, 0x007e0000			    //Save: trap_id = 0x7e. Restore: trap_id = 0x7f.
+	s_cbranch_scc0 L_JUMP_TO_RESTORE			    //do not need to recover STATUS here  since we are going to RESTORE
+	//FIXME	 s_setreg_b32	hwreg(HW_REG_STATUS),	s_save_status	    //need to recover STATUS since we are going to SAVE
+	s_branch L_SKIP_RESTORE					    //NOT restore, SAVE actually
+    else
+	s_branch L_SKIP_RESTORE					    //NOT restore. might be a regular trap or save
+    end
+
+L_JUMP_TO_RESTORE:
+    s_branch L_RESTORE						    //restore
+
+L_SKIP_RESTORE:
+
+    s_getreg_b32    s_save_status, hwreg(HW_REG_STATUS)				    //save STATUS since we will change SCC
+    s_andn2_b32	    s_save_status, s_save_status, SQ_WAVE_STATUS_SPI_PRIO_MASK	    //check whether this is for save
+    s_getreg_b32    s_save_trapsts, hwreg(HW_REG_TRAPSTS)
+    s_and_b32       ttmp2, s_save_trapsts, SQ_WAVE_TRAPSTS_SAVECTX_MASK    //check whether this is for save
+    s_cbranch_scc1  L_SAVE					//this is the operation for save
+
+    // *********    Handle non-CWSR traps	*******************
+if (!EMU_RUN_HACK)
+    // Illegal instruction is a non-maskable exception which blocks context save.
+    // Halt the wavefront and return from the trap.
+    s_and_b32       ttmp2, s_save_trapsts, SQ_WAVE_TRAPSTS_ILLEGAL_INST_MASK
+    s_cbranch_scc1  L_HALT_WAVE
+
+    // If STATUS.MEM_VIOL is asserted then we cannot fetch from the TMA.
+    // Instead, halt the wavefront and return from the trap.
+    s_and_b32       ttmp2, s_save_trapsts, SQ_WAVE_TRAPSTS_MEM_VIOL_MASK
+    s_cbranch_scc0  L_FETCH_2ND_TRAP
+
+L_HALT_WAVE:
+    // If STATUS.HALT is set then this fault must come from SQC instruction fetch.
+    // We cannot prevent further faults so just terminate the wavefront.
+    s_and_b32       ttmp2, s_save_status, SQ_WAVE_STATUS_HALT_MASK
+    s_cbranch_scc0  L_NOT_ALREADY_HALTED
+    s_endpgm
+L_NOT_ALREADY_HALTED:
+    s_or_b32        s_save_status, s_save_status, SQ_WAVE_STATUS_HALT_MASK
+
+    // If the PC points to S_ENDPGM then context save will fail if STATUS.HALT is set.
+    // Rewind the PC to prevent this from occurring. The debugger compensates for this.
+    s_sub_u32       ttmp0, ttmp0, 0x8
+    s_subb_u32      ttmp1, ttmp1, 0x0
+
+L_FETCH_2ND_TRAP:
+    // Preserve and clear scalar XNACK state before issuing scalar reads.
+    // Save IB_STS.FIRST_REPLAY[15] and IB_STS.RCNT[20:16] into unused space ttmp11[31:26].
+    s_getreg_b32    ttmp2, hwreg(HW_REG_IB_STS)
+    s_and_b32       ttmp3, ttmp2, SQ_WAVE_IB_STS_RCNT_FIRST_REPLAY_MASK
+    s_lshl_b32      ttmp3, ttmp3, (TTMP11_SAVE_RCNT_FIRST_REPLAY_SHIFT - SQ_WAVE_IB_STS_FIRST_REPLAY_SHIFT)
+    s_andn2_b32     ttmp11, ttmp11, TTMP11_SAVE_RCNT_FIRST_REPLAY_MASK
+    s_or_b32        ttmp11, ttmp11, ttmp3
+
+    s_andn2_b32     ttmp2, ttmp2, SQ_WAVE_IB_STS_RCNT_FIRST_REPLAY_MASK
+    s_setreg_b32    hwreg(HW_REG_IB_STS), ttmp2
+
+    // Read second-level TBA/TMA from first-level TMA and jump if available.
+    // ttmp[2:5] and ttmp12 can be used (others hold SPI-initialized debug data)
+    // ttmp12 holds SQ_WAVE_STATUS
+    s_getreg_b32    ttmp4, hwreg(HW_REG_SQ_SHADER_TMA_LO)
+    s_getreg_b32    ttmp5, hwreg(HW_REG_SQ_SHADER_TMA_HI)
+    s_lshl_b64      [ttmp4, ttmp5], [ttmp4, ttmp5], 0x8
+    s_load_dwordx2  [ttmp2, ttmp3], [ttmp4, ttmp5], 0x0 glc:1 // second-level TBA
+    s_waitcnt       lgkmcnt(0)
+    s_load_dwordx2  [ttmp4, ttmp5], [ttmp4, ttmp5], 0x8 glc:1 // second-level TMA
+    s_waitcnt       lgkmcnt(0)
+    s_and_b64       [ttmp2, ttmp3], [ttmp2, ttmp3], [ttmp2, ttmp3]
+    s_cbranch_scc0  L_NO_NEXT_TRAP // second-level trap handler not been set
+    s_setpc_b64     [ttmp2, ttmp3] // jump to second-level trap handler
+
+L_NO_NEXT_TRAP:
+    s_getreg_b32    s_save_trapsts, hwreg(HW_REG_TRAPSTS)
+    s_and_b32	    s_save_trapsts, s_save_trapsts, SQ_WAVE_TRAPSTS_EXCE_MASK // Check whether it is an exception
+    s_cbranch_scc1  L_EXCP_CASE	  // Exception, jump back to the shader program directly.
+    s_add_u32	    ttmp0, ttmp0, 4   // S_TRAP case, add 4 to ttmp0
+    s_addc_u32	ttmp1, ttmp1, 0
+L_EXCP_CASE:
+    s_and_b32	ttmp1, ttmp1, 0xFFFF
+
+    // Restore SQ_WAVE_IB_STS.
+    s_lshr_b32      ttmp2, ttmp11, (TTMP11_SAVE_RCNT_FIRST_REPLAY_SHIFT - SQ_WAVE_IB_STS_FIRST_REPLAY_SHIFT)
+    s_and_b32       ttmp2, ttmp2, SQ_WAVE_IB_STS_RCNT_FIRST_REPLAY_MASK
+    s_setreg_b32    hwreg(HW_REG_IB_STS), ttmp2
+
+    // Restore SQ_WAVE_STATUS.
+    s_and_b64       exec, exec, exec // Restore STATUS.EXECZ, not writable by s_setreg_b32
+    s_and_b64       vcc, vcc, vcc    // Restore STATUS.VCCZ, not writable by s_setreg_b32
+    s_setreg_b32    hwreg(HW_REG_STATUS), s_save_status
+
+    s_rfe_b64       [ttmp0, ttmp1]
+end
+    // *********	End handling of non-CWSR traps	 *******************
+
+/**************************************************************************/
+/*			save routine					  */
+/**************************************************************************/
+
+L_SAVE:
+
+if G8SR_DEBUG_TIMESTAMP
+	s_memrealtime	s_g8sr_ts_save_s
+	s_waitcnt lgkmcnt(0)	     //FIXME, will cause xnack??
+end
+
+    s_and_b32	    s_save_pc_hi, s_save_pc_hi, 0x0000ffff    //pc[47:32]
+
+    s_mov_b32	    s_save_tmp, 0							    //clear saveCtx bit
+    s_setreg_b32    hwreg(HW_REG_TRAPSTS, SQ_WAVE_TRAPSTS_SAVECTX_SHIFT, 1), s_save_tmp	    //clear saveCtx bit
+
+    s_getreg_b32    s_save_tmp, hwreg(HW_REG_IB_STS, SQ_WAVE_IB_STS_RCNT_SHIFT, SQ_WAVE_IB_STS_RCNT_SIZE)		    //save RCNT
+    s_lshl_b32	    s_save_tmp, s_save_tmp, S_SAVE_PC_HI_RCNT_SHIFT
+    s_or_b32	    s_save_pc_hi, s_save_pc_hi, s_save_tmp
+    s_getreg_b32    s_save_tmp, hwreg(HW_REG_IB_STS, SQ_WAVE_IB_STS_FIRST_REPLAY_SHIFT, SQ_WAVE_IB_STS_FIRST_REPLAY_SIZE)   //save FIRST_REPLAY
+    s_lshl_b32	    s_save_tmp, s_save_tmp, S_SAVE_PC_HI_FIRST_REPLAY_SHIFT
+    s_or_b32	    s_save_pc_hi, s_save_pc_hi, s_save_tmp
+    s_getreg_b32    s_save_tmp, hwreg(HW_REG_IB_STS)					    //clear RCNT and FIRST_REPLAY in IB_STS
+    s_and_b32	    s_save_tmp, s_save_tmp, SQ_WAVE_IB_STS_RCNT_FIRST_REPLAY_MASK_NEG
+
+    s_setreg_b32    hwreg(HW_REG_IB_STS), s_save_tmp
+
+    /*	    inform SPI the readiness and wait for SPI's go signal */
+    s_mov_b32	    s_save_exec_lo, exec_lo						    //save EXEC and use EXEC for the go signal from SPI
+    s_mov_b32	    s_save_exec_hi, exec_hi
+    s_mov_b64	    exec,   0x0								    //clear EXEC to get ready to receive
+
+if G8SR_DEBUG_TIMESTAMP
+	s_memrealtime  s_g8sr_ts_sq_save_msg
+	s_waitcnt lgkmcnt(0)
+end
+
+    if (EMU_RUN_HACK)
+
+    else
+	s_sendmsg   sendmsg(MSG_SAVEWAVE)  //send SPI a message and wait for SPI's write to EXEC
+    end
+
+    // Set SPI_PRIO=2 to avoid starving instruction fetch in the waves we're waiting for.
+    s_or_b32 s_save_tmp, s_save_status, (2 << SQ_WAVE_STATUS_SPI_PRIO_SHIFT)
+    s_setreg_b32 hwreg(HW_REG_STATUS), s_save_tmp
+
+  L_SLEEP:
+    s_sleep 0x2		       // sleep 1 (64clk) is not enough for 8 waves per SIMD, which will cause SQ hang, since the 7,8th wave could not get arbit to exec inst, while other waves are stuck into the sleep-loop and waiting for wrexec!=0
+
+    if (EMU_RUN_HACK)
+
+    else
+	s_cbranch_execz L_SLEEP
+    end
+
+if G8SR_DEBUG_TIMESTAMP
+	s_memrealtime  s_g8sr_ts_spi_wrexec
+	s_waitcnt lgkmcnt(0)
+end
+
+    if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_SAVE_SINGLE_WAVE))
+	//calculate wd_addr using absolute thread id
+	v_readlane_b32 s_save_tmp, v9, 0
+	s_lshr_b32 s_save_tmp, s_save_tmp, 6
+	s_mul_i32 s_save_tmp, s_save_tmp, WAVE_SPACE
+	s_add_i32 s_save_spi_init_lo, s_save_tmp, WG_BASE_ADDR_LO
+	s_mov_b32 s_save_spi_init_hi, WG_BASE_ADDR_HI
+	s_and_b32 s_save_spi_init_hi, s_save_spi_init_hi, CTX_SAVE_CONTROL
+    else
+    end
+    if ((EMU_RUN_HACK) && (EMU_RUN_HACK_SAVE_SINGLE_WAVE))
+	s_add_i32 s_save_spi_init_lo, s_save_tmp, WG_BASE_ADDR_LO
+	s_mov_b32 s_save_spi_init_hi, WG_BASE_ADDR_HI
+	s_and_b32 s_save_spi_init_hi, s_save_spi_init_hi, CTX_SAVE_CONTROL
+    else
+    end
+
+    // Save trap temporaries 6-11, 13-15 initialized by SPI debug dispatch logic
+    // ttmp SR memory offset : size(VGPR)+size(SGPR)+0x40
+    get_vgpr_size_bytes(s_save_ttmps_lo)
+    get_sgpr_size_bytes(s_save_ttmps_hi)
+    s_add_u32	    s_save_ttmps_lo, s_save_ttmps_lo, s_save_ttmps_hi
+    s_add_u32	    s_save_ttmps_lo, s_save_ttmps_lo, s_save_spi_init_lo
+    s_addc_u32	    s_save_ttmps_hi, s_save_spi_init_hi, 0x0
+    s_and_b32	    s_save_ttmps_hi, s_save_ttmps_hi, 0xFFFF
+    s_store_dwordx2 [ttmp6, ttmp7], [s_save_ttmps_lo, s_save_ttmps_hi], 0x40 glc:1
+    ack_sqc_store_workaround()
+    s_store_dwordx4 [ttmp8, ttmp9, ttmp10, ttmp11], [s_save_ttmps_lo, s_save_ttmps_hi], 0x48 glc:1
+    ack_sqc_store_workaround()
+    s_store_dword   ttmp13, [s_save_ttmps_lo, s_save_ttmps_hi], 0x58 glc:1
+    ack_sqc_store_workaround()
+    s_store_dwordx2 [ttmp14, ttmp15], [s_save_ttmps_lo, s_save_ttmps_hi], 0x5C glc:1
+    ack_sqc_store_workaround()
+
+    /*	    setup Resource Contants    */
+    s_mov_b32	    s_save_buf_rsrc0,	s_save_spi_init_lo							//base_addr_lo
+    s_and_b32	    s_save_buf_rsrc1,	s_save_spi_init_hi, 0x0000FFFF						//base_addr_hi
+    s_or_b32	    s_save_buf_rsrc1,	s_save_buf_rsrc1,  S_SAVE_BUF_RSRC_WORD1_STRIDE
+    s_mov_b32	    s_save_buf_rsrc2,	0									//NUM_RECORDS initial value = 0 (in bytes) although not neccessarily inited
+    s_mov_b32	    s_save_buf_rsrc3,	S_SAVE_BUF_RSRC_WORD3_MISC
+    s_and_b32	    s_save_tmp,		s_save_spi_init_hi, S_SAVE_SPI_INIT_ATC_MASK
+    s_lshr_b32	    s_save_tmp,		s_save_tmp, (S_SAVE_SPI_INIT_ATC_SHIFT-SQ_BUF_RSRC_WORD1_ATC_SHIFT)	    //get ATC bit into position
+    s_or_b32	    s_save_buf_rsrc3,	s_save_buf_rsrc3,  s_save_tmp						//or ATC
+    s_and_b32	    s_save_tmp,		s_save_spi_init_hi, S_SAVE_SPI_INIT_MTYPE_MASK
+    s_lshr_b32	    s_save_tmp,		s_save_tmp, (S_SAVE_SPI_INIT_MTYPE_SHIFT-SQ_BUF_RSRC_WORD3_MTYPE_SHIFT)	    //get MTYPE bits into position
+    s_or_b32	    s_save_buf_rsrc3,	s_save_buf_rsrc3,  s_save_tmp						//or MTYPE
+
+    //FIXME  right now s_save_m0/s_save_mem_offset use tma_lo/tma_hi  (might need to save them before using them?)
+    s_mov_b32	    s_save_m0,		m0								    //save M0
+
+    /*	    global mem offset		*/
+    s_mov_b32	    s_save_mem_offset,	0x0									//mem offset initial value = 0
+
+
+
+
+    /*	    save HW registers	*/
+    //////////////////////////////
+
+  L_SAVE_HWREG:
+	// HWREG SR memory offset : size(VGPR)+size(SGPR)
+       get_vgpr_size_bytes(s_save_mem_offset)
+       get_sgpr_size_bytes(s_save_tmp)
+       s_add_u32 s_save_mem_offset, s_save_mem_offset, s_save_tmp
+
+
+    s_mov_b32	    s_save_buf_rsrc2, 0x4				//NUM_RECORDS	in bytes
+    if (SWIZZLE_EN)
+	s_add_u32	s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_save_buf_rsrc2,  0x1000000				    //NUM_RECORDS in bytes
+    end
+
+
+    write_hwreg_to_mem(s_save_m0, s_save_buf_rsrc0, s_save_mem_offset)			//M0
+
+    if ((EMU_RUN_HACK) && (EMU_RUN_HACK_SAVE_FIRST_TIME))
+	s_add_u32 s_save_pc_lo, s_save_pc_lo, 4		    //pc[31:0]+4
+	s_addc_u32 s_save_pc_hi, s_save_pc_hi, 0x0	    //carry bit over
+    end
+
+    write_hwreg_to_mem(s_save_pc_lo, s_save_buf_rsrc0, s_save_mem_offset)		    //PC
+    write_hwreg_to_mem(s_save_pc_hi, s_save_buf_rsrc0, s_save_mem_offset)
+    write_hwreg_to_mem(s_save_exec_lo, s_save_buf_rsrc0, s_save_mem_offset)		//EXEC
+    write_hwreg_to_mem(s_save_exec_hi, s_save_buf_rsrc0, s_save_mem_offset)
+    write_hwreg_to_mem(s_save_status, s_save_buf_rsrc0, s_save_mem_offset)		//STATUS
+
+    //s_save_trapsts conflicts with s_save_alloc_size
+    s_getreg_b32    s_save_trapsts, hwreg(HW_REG_TRAPSTS)
+    write_hwreg_to_mem(s_save_trapsts, s_save_buf_rsrc0, s_save_mem_offset)		//TRAPSTS
+
+    write_hwreg_to_mem(xnack_mask_lo, s_save_buf_rsrc0, s_save_mem_offset)	    //XNACK_MASK_LO
+    write_hwreg_to_mem(xnack_mask_hi, s_save_buf_rsrc0, s_save_mem_offset)	    //XNACK_MASK_HI
+
+    //use s_save_tmp would introduce conflict here between s_save_tmp and s_save_buf_rsrc2
+    s_getreg_b32    s_save_m0, hwreg(HW_REG_MODE)						    //MODE
+    write_hwreg_to_mem(s_save_m0, s_save_buf_rsrc0, s_save_mem_offset)
+
+
+
+    /*	    the first wave in the threadgroup	 */
+    s_and_b32	    s_save_tmp, s_save_spi_init_hi, S_SAVE_SPI_INIT_FIRST_WAVE_MASK	// extract fisrt wave bit
+    s_mov_b32	     s_save_exec_hi, 0x0
+    s_or_b32	     s_save_exec_hi, s_save_tmp, s_save_exec_hi				 // save first wave bit to s_save_exec_hi.bits[26]
+
+
+    /*		save SGPRs	*/
+	// Save SGPR before LDS save, then the s0 to s4 can be used during LDS save...
+    //////////////////////////////
+
+    // SGPR SR memory offset : size(VGPR)
+    get_vgpr_size_bytes(s_save_mem_offset)
+    // TODO, change RSRC word to rearrange memory layout for SGPRS
+
+    s_getreg_b32    s_save_alloc_size, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SIZE)		//spgr_size
+    s_add_u32	    s_save_alloc_size, s_save_alloc_size, 1
+    s_lshl_b32	    s_save_alloc_size, s_save_alloc_size, 4			    //Number of SGPRs = (sgpr_size + 1) * 16   (non-zero value)
+
+    if (SGPR_SAVE_USE_SQC)
+	s_lshl_b32	s_save_buf_rsrc2,   s_save_alloc_size, 2		    //NUM_RECORDS in bytes
+    else
+	s_lshl_b32	s_save_buf_rsrc2,   s_save_alloc_size, 8		    //NUM_RECORDS in bytes (64 threads)
+    end
+
+    if (SWIZZLE_EN)
+	s_add_u32	s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_save_buf_rsrc2,  0x1000000				    //NUM_RECORDS in bytes
+    end
+
+
+    // backup s_save_buf_rsrc0,1 to s_save_pc_lo/hi, since write_16sgpr_to_mem function will change the rsrc0
+    //s_mov_b64 s_save_pc_lo, s_save_buf_rsrc0
+    s_mov_b64 s_save_xnack_mask_lo, s_save_buf_rsrc0
+    s_add_u32 s_save_buf_rsrc0, s_save_buf_rsrc0, s_save_mem_offset
+    s_addc_u32 s_save_buf_rsrc1, s_save_buf_rsrc1, 0
+
+    s_mov_b32	    m0, 0x0			    //SGPR initial index value =0
+    s_nop	    0x0				    //Manually inserted wait states
+  L_SAVE_SGPR_LOOP:
+    // SGPR is allocated in 16 SGPR granularity
+    s_movrels_b64   s0, s0     //s0 = s[0+m0], s1 = s[1+m0]
+    s_movrels_b64   s2, s2     //s2 = s[2+m0], s3 = s[3+m0]
+    s_movrels_b64   s4, s4     //s4 = s[4+m0], s5 = s[5+m0]
+    s_movrels_b64   s6, s6     //s6 = s[6+m0], s7 = s[7+m0]
+    s_movrels_b64   s8, s8     //s8 = s[8+m0], s9 = s[9+m0]
+    s_movrels_b64   s10, s10   //s10 = s[10+m0], s11 = s[11+m0]
+    s_movrels_b64   s12, s12   //s12 = s[12+m0], s13 = s[13+m0]
+    s_movrels_b64   s14, s14   //s14 = s[14+m0], s15 = s[15+m0]
+
+    write_16sgpr_to_mem(s0, s_save_buf_rsrc0, s_save_mem_offset) //PV: the best performance should be using s_buffer_store_dwordx4
+    s_add_u32	    m0, m0, 16							    //next sgpr index
+    s_cmp_lt_u32    m0, s_save_alloc_size					    //scc = (m0 < s_save_alloc_size) ? 1 : 0
+    s_cbranch_scc1  L_SAVE_SGPR_LOOP					//SGPR save is complete?
+    // restore s_save_buf_rsrc0,1
+    //s_mov_b64 s_save_buf_rsrc0, s_save_pc_lo
+    s_mov_b64 s_save_buf_rsrc0, s_save_xnack_mask_lo
+
+
+
+
+    /*		save first 4 VGPR, then LDS save could use   */
+	// each wave will alloc 4 vgprs at least...
+    /////////////////////////////////////////////////////////////////////////////////////
+
+    s_mov_b32	    s_save_mem_offset, 0
+    s_mov_b32	    exec_lo, 0xFFFFFFFF						    //need every thread from now on
+    s_mov_b32	    exec_hi, 0xFFFFFFFF
+    s_mov_b32	    xnack_mask_lo, 0x0
+    s_mov_b32	    xnack_mask_hi, 0x0
+
+    if (SWIZZLE_EN)
+	s_add_u32	s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_save_buf_rsrc2,  0x1000000				    //NUM_RECORDS in bytes
+    end
+
+
+    // VGPR Allocated in 4-GPR granularity
+
+if G8SR_VGPR_SR_IN_DWX4
+	// the const stride for DWx4 is 4*4 bytes
+	s_and_b32 s_save_buf_rsrc1, s_save_buf_rsrc1, 0x0000FFFF   // reset const stride to 0
+	s_or_b32  s_save_buf_rsrc1, s_save_buf_rsrc1, G8SR_SAVE_BUF_RSRC_WORD1_STRIDE_DWx4  // const stride to 4*4 bytes
+
+	buffer_store_dwordx4 v0, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1
+
+	s_and_b32 s_save_buf_rsrc1, s_save_buf_rsrc1, 0x0000FFFF   // reset const stride to 0
+	s_or_b32  s_save_buf_rsrc1, s_save_buf_rsrc1, S_SAVE_BUF_RSRC_WORD1_STRIDE  // reset const stride to 4 bytes
+else
+	buffer_store_dword v0, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1
+	buffer_store_dword v1, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1  offset:256
+	buffer_store_dword v2, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1  offset:256*2
+	buffer_store_dword v3, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1  offset:256*3
+end
+
+
+
+    /*		save LDS	*/
+    //////////////////////////////
+
+  L_SAVE_LDS:
+
+	// Change EXEC to all threads...
+    s_mov_b32	    exec_lo, 0xFFFFFFFF	  //need every thread from now on
+    s_mov_b32	    exec_hi, 0xFFFFFFFF
+
+    s_getreg_b32    s_save_alloc_size, hwreg(HW_REG_LDS_ALLOC,SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT,SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE)		    //lds_size
+    s_and_b32	    s_save_alloc_size, s_save_alloc_size, 0xFFFFFFFF		    //lds_size is zero?
+    s_cbranch_scc0  L_SAVE_LDS_DONE									       //no lds used? jump to L_SAVE_DONE
+
+    s_barrier		    //LDS is used? wait for other waves in the same TG
+    s_and_b32	    s_save_tmp, s_save_exec_hi, S_SAVE_SPI_INIT_FIRST_WAVE_MASK		       //exec is still used here
+    s_cbranch_scc0  L_SAVE_LDS_DONE
+
+	// first wave do LDS save;
+
+    s_lshl_b32	    s_save_alloc_size, s_save_alloc_size, 6			    //LDS size in dwords = lds_size * 64dw
+    s_lshl_b32	    s_save_alloc_size, s_save_alloc_size, 2			    //LDS size in bytes
+    s_mov_b32	    s_save_buf_rsrc2,  s_save_alloc_size			    //NUM_RECORDS in bytes
+
+    // LDS at offset: size(VGPR)+SIZE(SGPR)+SIZE(HWREG)
+    //
+    get_vgpr_size_bytes(s_save_mem_offset)
+    get_sgpr_size_bytes(s_save_tmp)
+    s_add_u32  s_save_mem_offset, s_save_mem_offset, s_save_tmp
+    s_add_u32 s_save_mem_offset, s_save_mem_offset, get_hwreg_size_bytes()
+
+
+    if (SWIZZLE_EN)
+	s_add_u32	s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0	      //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_save_buf_rsrc2,  0x1000000		      //NUM_RECORDS in bytes
+    end
+
+    s_mov_b32	    m0, 0x0						  //lds_offset initial value = 0
+
+
+var LDS_DMA_ENABLE = 0
+var UNROLL = 0
+if UNROLL==0 && LDS_DMA_ENABLE==1
+	s_mov_b32  s3, 256*2
+	s_nop 0
+	s_nop 0
+	s_nop 0
+  L_SAVE_LDS_LOOP:
+	//TODO: looks the 2 buffer_store/load clause for s/r will hurt performance.???
+    if (SAVE_LDS)     //SPI always alloc LDS space in 128DW granularity
+	    buffer_store_lds_dword s_save_buf_rsrc0, s_save_mem_offset lds:1		// first 64DW
+	    buffer_store_lds_dword s_save_buf_rsrc0, s_save_mem_offset lds:1 offset:256 // second 64DW
+    end
+
+    s_add_u32	    m0, m0, s3						//every buffer_store_lds does 256 bytes
+    s_add_u32	    s_save_mem_offset, s_save_mem_offset, s3				//mem offset increased by 256 bytes
+    s_cmp_lt_u32    m0, s_save_alloc_size						//scc=(m0 < s_save_alloc_size) ? 1 : 0
+    s_cbranch_scc1  L_SAVE_LDS_LOOP							//LDS save is complete?
+
+elsif LDS_DMA_ENABLE==1 && UNROLL==1 // UNROOL	, has ichace miss
+      // store from higest LDS address to lowest
+      s_mov_b32	 s3, 256*2
+      s_sub_u32	 m0, s_save_alloc_size, s3
+      s_add_u32 s_save_mem_offset, s_save_mem_offset, m0
+      s_lshr_b32 s_save_alloc_size, s_save_alloc_size, 9   // how many 128 trunks...
+      s_sub_u32 s_save_alloc_size, 128, s_save_alloc_size   // store from higheset addr to lowest
+      s_mul_i32 s_save_alloc_size, s_save_alloc_size, 6*4   // PC offset increment,  each LDS save block cost 6*4 Bytes instruction
+      s_add_u32 s_save_alloc_size, s_save_alloc_size, 3*4   //2is the below 2 inst...//s_addc and s_setpc
+      s_nop 0
+      s_nop 0
+      s_nop 0	//pad 3 dw to let LDS_DMA align with 64Bytes
+      s_getpc_b64 s[0:1]			      // reuse s[0:1], since s[0:1] already saved
+      s_add_u32	  s0, s0,s_save_alloc_size
+      s_addc_u32  s1, s1, 0
+      s_setpc_b64 s[0:1]
+
+
+       for var i =0; i< 128; i++
+	    // be careful to make here a 64Byte aligned address, which could improve performance...
+	    buffer_store_lds_dword s_save_buf_rsrc0, s_save_mem_offset lds:1 offset:0		// first 64DW
+	    buffer_store_lds_dword s_save_buf_rsrc0, s_save_mem_offset lds:1 offset:256		  // second 64DW
+
+	if i!=127
+	s_sub_u32  m0, m0, s3	   // use a sgpr to shrink 2DW-inst to 1DW inst to improve performance , i.e.  pack more LDS_DMA inst to one Cacheline
+	    s_sub_u32  s_save_mem_offset, s_save_mem_offset,  s3
+	    end
+       end
+
+else   // BUFFER_STORE
+      v_mbcnt_lo_u32_b32 v2, 0xffffffff, 0x0
+      v_mbcnt_hi_u32_b32 v3, 0xffffffff, v2	// tid
+      v_mul_i32_i24 v2, v3, 8	// tid*8
+      v_mov_b32 v3, 256*2
+      s_mov_b32 m0, 0x10000
+      s_mov_b32 s0, s_save_buf_rsrc3
+      s_and_b32 s_save_buf_rsrc3, s_save_buf_rsrc3, 0xFF7FFFFF	  // disable add_tid
+      s_or_b32 s_save_buf_rsrc3, s_save_buf_rsrc3, 0x58000   //DFMT
+
+L_SAVE_LDS_LOOP_VECTOR:
+      ds_read_b64 v[0:1], v2	//x =LDS[a], byte address
+      s_waitcnt lgkmcnt(0)
+      buffer_store_dwordx2  v[0:1], v2, s_save_buf_rsrc0, s_save_mem_offset offen:1  glc:1  slc:1
+//	s_waitcnt vmcnt(0)
+//	v_add_u32 v2, vcc[0:1], v2, v3
+      v_add_u32 v2, v2, v3
+      v_cmp_lt_u32 vcc[0:1], v2, s_save_alloc_size
+      s_cbranch_vccnz L_SAVE_LDS_LOOP_VECTOR
+
+      // restore rsrc3
+      s_mov_b32 s_save_buf_rsrc3, s0
+
+end
+
+L_SAVE_LDS_DONE:
+
+
+    /*		save VGPRs  - set the Rest VGPRs	*/
+    //////////////////////////////////////////////////////////////////////////////////////
+  L_SAVE_VGPR:
+    // VGPR SR memory offset: 0
+    // TODO rearrange the RSRC words to use swizzle for VGPR save...
+
+    s_mov_b32	    s_save_mem_offset, (0+256*4)				    // for the rest VGPRs
+    s_mov_b32	    exec_lo, 0xFFFFFFFF						    //need every thread from now on
+    s_mov_b32	    exec_hi, 0xFFFFFFFF
+
+    s_getreg_b32    s_save_alloc_size, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SIZE)		    //vpgr_size
+    s_add_u32	    s_save_alloc_size, s_save_alloc_size, 1
+    s_lshl_b32	    s_save_alloc_size, s_save_alloc_size, 2			    //Number of VGPRs = (vgpr_size + 1) * 4    (non-zero value)	  //FIXME for GFX, zero is possible
+    s_lshl_b32	    s_save_buf_rsrc2,  s_save_alloc_size, 8			    //NUM_RECORDS in bytes (64 threads*4)
+    if (SWIZZLE_EN)
+	s_add_u32	s_save_buf_rsrc2, s_save_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_save_buf_rsrc2,  0x1000000				    //NUM_RECORDS in bytes
+    end
+
+
+    // VGPR Allocated in 4-GPR granularity
+
+if G8SR_VGPR_SR_IN_DWX4
+	// the const stride for DWx4 is 4*4 bytes
+	s_and_b32 s_save_buf_rsrc1, s_save_buf_rsrc1, 0x0000FFFF   // reset const stride to 0
+	s_or_b32  s_save_buf_rsrc1, s_save_buf_rsrc1, G8SR_SAVE_BUF_RSRC_WORD1_STRIDE_DWx4  // const stride to 4*4 bytes
+
+	s_mov_b32	  m0, 4	    // skip first 4 VGPRs
+	s_cmp_lt_u32	  m0, s_save_alloc_size
+	s_cbranch_scc0	  L_SAVE_VGPR_LOOP_END	    // no more vgprs
+
+	s_set_gpr_idx_on  m0, 0x1   // This will change M0
+	s_add_u32	  s_save_alloc_size, s_save_alloc_size, 0x1000	// because above inst change m0
+L_SAVE_VGPR_LOOP:
+	v_mov_b32	  v0, v0   // v0 = v[0+m0]
+	v_mov_b32	  v1, v1
+	v_mov_b32	  v2, v2
+	v_mov_b32	  v3, v3
+
+
+	buffer_store_dwordx4 v0, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1
+	s_add_u32	  m0, m0, 4
+	s_add_u32	  s_save_mem_offset, s_save_mem_offset, 256*4
+	s_cmp_lt_u32	  m0, s_save_alloc_size
+    s_cbranch_scc1  L_SAVE_VGPR_LOOP						    //VGPR save is complete?
+    s_set_gpr_idx_off
+L_SAVE_VGPR_LOOP_END:
+
+	s_and_b32 s_save_buf_rsrc1, s_save_buf_rsrc1, 0x0000FFFF   // reset const stride to 0
+	s_or_b32  s_save_buf_rsrc1, s_save_buf_rsrc1, S_SAVE_BUF_RSRC_WORD1_STRIDE  // reset const stride to 4 bytes
+else
+    // VGPR store using dw burst
+    s_mov_b32	      m0, 0x4	//VGPR initial index value =0
+    s_cmp_lt_u32      m0, s_save_alloc_size
+    s_cbranch_scc0    L_SAVE_VGPR_END
+
+
+    s_set_gpr_idx_on	m0, 0x1 //M0[7:0] = M0[7:0] and M0[15:12] = 0x1
+    s_add_u32	    s_save_alloc_size, s_save_alloc_size, 0x1000		    //add 0x1000 since we compare m0 against it later
+
+  L_SAVE_VGPR_LOOP:
+    v_mov_b32	    v0, v0		//v0 = v[0+m0]
+    v_mov_b32	    v1, v1		//v0 = v[0+m0]
+    v_mov_b32	    v2, v2		//v0 = v[0+m0]
+    v_mov_b32	    v3, v3		//v0 = v[0+m0]
+
+    if(USE_MTBUF_INSTEAD_OF_MUBUF)
+	tbuffer_store_format_x v0, v0, s_save_buf_rsrc0, s_save_mem_offset format:BUF_NUM_FORMAT_FLOAT format: BUF_DATA_FORMAT_32 slc:1 glc:1
+    else
+	buffer_store_dword v0, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1
+	buffer_store_dword v1, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1  offset:256
+	buffer_store_dword v2, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1  offset:256*2
+	buffer_store_dword v3, v0, s_save_buf_rsrc0, s_save_mem_offset slc:1 glc:1  offset:256*3
+    end
+
+    s_add_u32	    m0, m0, 4							    //next vgpr index
+    s_add_u32	    s_save_mem_offset, s_save_mem_offset, 256*4			    //every buffer_store_dword does 256 bytes
+    s_cmp_lt_u32    m0, s_save_alloc_size					    //scc = (m0 < s_save_alloc_size) ? 1 : 0
+    s_cbranch_scc1  L_SAVE_VGPR_LOOP						    //VGPR save is complete?
+    s_set_gpr_idx_off
+end
+
+L_SAVE_VGPR_END:
+
+
+
+
+
+
+    /*	   S_PGM_END_SAVED  */				    //FIXME  graphics ONLY
+    if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_SAVE_NORMAL_EXIT))
+	s_and_b32 s_save_pc_hi, s_save_pc_hi, 0x0000ffff    //pc[47:32]
+	s_add_u32 s_save_pc_lo, s_save_pc_lo, 4		    //pc[31:0]+4
+	s_addc_u32 s_save_pc_hi, s_save_pc_hi, 0x0	    //carry bit over
+	s_rfe_b64 s_save_pc_lo				    //Return to the main shader program
+    else
+    end
+
+// Save Done timestamp
+if G8SR_DEBUG_TIMESTAMP
+	s_memrealtime	s_g8sr_ts_save_d
+	// SGPR SR memory offset : size(VGPR)
+	get_vgpr_size_bytes(s_save_mem_offset)
+	s_add_u32 s_save_mem_offset, s_save_mem_offset, G8SR_DEBUG_TS_SAVE_D_OFFSET
+	s_waitcnt lgkmcnt(0)	     //FIXME, will cause xnack??
+	// Need reset rsrc2??
+	s_mov_b32 m0, s_save_mem_offset
+	s_mov_b32 s_save_buf_rsrc2,  0x1000000					//NUM_RECORDS in bytes
+	s_buffer_store_dwordx2 s_g8sr_ts_save_d, s_save_buf_rsrc0, m0	    glc:1
+end
+
+
+    s_branch	L_END_PGM
+
+
+
+/**************************************************************************/
+/*			restore routine					  */
+/**************************************************************************/
+
+L_RESTORE:
+    /*	    Setup Resource Contants    */
+    if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_RESTORE_NORMAL))
+	//calculate wd_addr using absolute thread id
+	v_readlane_b32 s_restore_tmp, v9, 0
+	s_lshr_b32 s_restore_tmp, s_restore_tmp, 6
+	s_mul_i32 s_restore_tmp, s_restore_tmp, WAVE_SPACE
+	s_add_i32 s_restore_spi_init_lo, s_restore_tmp, WG_BASE_ADDR_LO
+	s_mov_b32 s_restore_spi_init_hi, WG_BASE_ADDR_HI
+	s_and_b32 s_restore_spi_init_hi, s_restore_spi_init_hi, CTX_RESTORE_CONTROL
+    else
+    end
+
+if G8SR_DEBUG_TIMESTAMP
+	s_memrealtime	s_g8sr_ts_restore_s
+	s_waitcnt lgkmcnt(0)	     //FIXME, will cause xnack??
+	// tma_lo/hi are sgpr 110, 111, which will not used for 112 SGPR allocated case...
+	s_mov_b32 s_restore_pc_lo, s_g8sr_ts_restore_s[0]
+	s_mov_b32 s_restore_pc_hi, s_g8sr_ts_restore_s[1]   //backup ts to ttmp0/1, sicne exec will be finally restored..
+end
+
+
+
+    s_mov_b32	    s_restore_buf_rsrc0,    s_restore_spi_init_lo							    //base_addr_lo
+    s_and_b32	    s_restore_buf_rsrc1,    s_restore_spi_init_hi, 0x0000FFFF						    //base_addr_hi
+    s_or_b32	    s_restore_buf_rsrc1,    s_restore_buf_rsrc1,  S_RESTORE_BUF_RSRC_WORD1_STRIDE
+    s_mov_b32	    s_restore_buf_rsrc2,    0										    //NUM_RECORDS initial value = 0 (in bytes)
+    s_mov_b32	    s_restore_buf_rsrc3,    S_RESTORE_BUF_RSRC_WORD3_MISC
+    s_and_b32	    s_restore_tmp,	    s_restore_spi_init_hi, S_RESTORE_SPI_INIT_ATC_MASK
+    s_lshr_b32	    s_restore_tmp,	    s_restore_tmp, (S_RESTORE_SPI_INIT_ATC_SHIFT-SQ_BUF_RSRC_WORD1_ATC_SHIFT)	    //get ATC bit into position
+    s_or_b32	    s_restore_buf_rsrc3,    s_restore_buf_rsrc3,  s_restore_tmp						    //or ATC
+    s_and_b32	    s_restore_tmp,	    s_restore_spi_init_hi, S_RESTORE_SPI_INIT_MTYPE_MASK
+    s_lshr_b32	    s_restore_tmp,	    s_restore_tmp, (S_RESTORE_SPI_INIT_MTYPE_SHIFT-SQ_BUF_RSRC_WORD3_MTYPE_SHIFT)   //get MTYPE bits into position
+    s_or_b32	    s_restore_buf_rsrc3,    s_restore_buf_rsrc3,  s_restore_tmp						    //or MTYPE
+
+    /*	    global mem offset		*/
+//  s_mov_b32	    s_restore_mem_offset, 0x0				    //mem offset initial value = 0
+
+    /*	    the first wave in the threadgroup	 */
+    s_and_b32	    s_restore_tmp, s_restore_spi_init_hi, S_RESTORE_SPI_INIT_FIRST_WAVE_MASK
+    s_cbranch_scc0  L_RESTORE_VGPR
+
+    /*		restore LDS	*/
+    //////////////////////////////
+  L_RESTORE_LDS:
+
+    s_mov_b32	    exec_lo, 0xFFFFFFFF							    //need every thread from now on   //be consistent with SAVE although can be moved ahead
+    s_mov_b32	    exec_hi, 0xFFFFFFFF
+
+    s_getreg_b32    s_restore_alloc_size, hwreg(HW_REG_LDS_ALLOC,SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT,SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE)		//lds_size
+    s_and_b32	    s_restore_alloc_size, s_restore_alloc_size, 0xFFFFFFFF		    //lds_size is zero?
+    s_cbranch_scc0  L_RESTORE_VGPR							    //no lds used? jump to L_RESTORE_VGPR
+    s_lshl_b32	    s_restore_alloc_size, s_restore_alloc_size, 6			    //LDS size in dwords = lds_size * 64dw
+    s_lshl_b32	    s_restore_alloc_size, s_restore_alloc_size, 2			    //LDS size in bytes
+    s_mov_b32	    s_restore_buf_rsrc2,    s_restore_alloc_size			    //NUM_RECORDS in bytes
+
+    // LDS at offset: size(VGPR)+SIZE(SGPR)+SIZE(HWREG)
+    //
+    get_vgpr_size_bytes(s_restore_mem_offset)
+    get_sgpr_size_bytes(s_restore_tmp)
+    s_add_u32  s_restore_mem_offset, s_restore_mem_offset, s_restore_tmp
+    s_add_u32  s_restore_mem_offset, s_restore_mem_offset, get_hwreg_size_bytes()	     //FIXME, Check if offset overflow???
+
+
+    if (SWIZZLE_EN)
+	s_add_u32	s_restore_buf_rsrc2, s_restore_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_restore_buf_rsrc2,  0x1000000					    //NUM_RECORDS in bytes
+    end
+    s_mov_b32	    m0, 0x0								    //lds_offset initial value = 0
+
+  L_RESTORE_LDS_LOOP:
+    if (SAVE_LDS)
+	buffer_load_dword   v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset lds:1		       // first 64DW
+	buffer_load_dword   v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset lds:1 offset:256	       // second 64DW
+    end
+    s_add_u32	    m0, m0, 256*2						// 128 DW
+    s_add_u32	    s_restore_mem_offset, s_restore_mem_offset, 256*2		//mem offset increased by 128DW
+    s_cmp_lt_u32    m0, s_restore_alloc_size					//scc=(m0 < s_restore_alloc_size) ? 1 : 0
+    s_cbranch_scc1  L_RESTORE_LDS_LOOP							    //LDS restore is complete?
+
+
+    /*		restore VGPRs	    */
+    //////////////////////////////
+  L_RESTORE_VGPR:
+	// VGPR SR memory offset : 0
+    s_mov_b32	    s_restore_mem_offset, 0x0
+    s_mov_b32	    exec_lo, 0xFFFFFFFF							    //need every thread from now on   //be consistent with SAVE although can be moved ahead
+    s_mov_b32	    exec_hi, 0xFFFFFFFF
+
+    s_getreg_b32    s_restore_alloc_size, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SIZE)	//vpgr_size
+    s_add_u32	    s_restore_alloc_size, s_restore_alloc_size, 1
+    s_lshl_b32	    s_restore_alloc_size, s_restore_alloc_size, 2			    //Number of VGPRs = (vgpr_size + 1) * 4    (non-zero value)
+    s_lshl_b32	    s_restore_buf_rsrc2,  s_restore_alloc_size, 8			    //NUM_RECORDS in bytes (64 threads*4)
+    if (SWIZZLE_EN)
+	s_add_u32	s_restore_buf_rsrc2, s_restore_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_restore_buf_rsrc2,  0x1000000					    //NUM_RECORDS in bytes
+    end
+
+if G8SR_VGPR_SR_IN_DWX4
+     get_vgpr_size_bytes(s_restore_mem_offset)
+     s_sub_u32	       s_restore_mem_offset, s_restore_mem_offset, 256*4
+
+     // the const stride for DWx4 is 4*4 bytes
+     s_and_b32 s_restore_buf_rsrc1, s_restore_buf_rsrc1, 0x0000FFFF   // reset const stride to 0
+     s_or_b32  s_restore_buf_rsrc1, s_restore_buf_rsrc1, G8SR_RESTORE_BUF_RSRC_WORD1_STRIDE_DWx4  // const stride to 4*4 bytes
+
+     s_mov_b32	       m0, s_restore_alloc_size
+     s_set_gpr_idx_on  m0, 0x8	  // Note.. This will change m0
+
+L_RESTORE_VGPR_LOOP:
+     buffer_load_dwordx4 v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset slc:1 glc:1
+     s_waitcnt vmcnt(0)
+     s_sub_u32	       m0, m0, 4
+     v_mov_b32	       v0, v0	// v[0+m0] = v0
+     v_mov_b32	       v1, v1
+     v_mov_b32	       v2, v2
+     v_mov_b32	       v3, v3
+     s_sub_u32	       s_restore_mem_offset, s_restore_mem_offset, 256*4
+     s_cmp_eq_u32      m0, 0x8000
+     s_cbranch_scc0    L_RESTORE_VGPR_LOOP
+     s_set_gpr_idx_off
+
+     s_and_b32 s_restore_buf_rsrc1, s_restore_buf_rsrc1, 0x0000FFFF   // reset const stride to 0
+     s_or_b32  s_restore_buf_rsrc1, s_restore_buf_rsrc1, S_RESTORE_BUF_RSRC_WORD1_STRIDE  // const stride to 4*4 bytes
+
+else
+    // VGPR load using dw burst
+    s_mov_b32	    s_restore_mem_offset_save, s_restore_mem_offset	// restore start with v1, v0 will be the last
+    s_add_u32	    s_restore_mem_offset, s_restore_mem_offset, 256*4
+    s_mov_b32	    m0, 4				//VGPR initial index value = 1
+    s_set_gpr_idx_on  m0, 0x8			    //M0[7:0] = M0[7:0] and M0[15:12] = 0x8
+    s_add_u32	    s_restore_alloc_size, s_restore_alloc_size, 0x8000			    //add 0x8000 since we compare m0 against it later
+
+  L_RESTORE_VGPR_LOOP:
+    if(USE_MTBUF_INSTEAD_OF_MUBUF)
+	tbuffer_load_format_x v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset format:BUF_NUM_FORMAT_FLOAT format: BUF_DATA_FORMAT_32 slc:1 glc:1
+    else
+	buffer_load_dword v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset slc:1 glc:1
+	buffer_load_dword v1, v0, s_restore_buf_rsrc0, s_restore_mem_offset slc:1 glc:1 offset:256
+	buffer_load_dword v2, v0, s_restore_buf_rsrc0, s_restore_mem_offset slc:1 glc:1 offset:256*2
+	buffer_load_dword v3, v0, s_restore_buf_rsrc0, s_restore_mem_offset slc:1 glc:1 offset:256*3
+    end
+    s_waitcnt	    vmcnt(0)								    //ensure data ready
+    v_mov_b32	    v0, v0								    //v[0+m0] = v0
+    v_mov_b32	    v1, v1
+    v_mov_b32	    v2, v2
+    v_mov_b32	    v3, v3
+    s_add_u32	    m0, m0, 4								    //next vgpr index
+    s_add_u32	    s_restore_mem_offset, s_restore_mem_offset, 256*4				//every buffer_load_dword does 256 bytes
+    s_cmp_lt_u32    m0, s_restore_alloc_size						    //scc = (m0 < s_restore_alloc_size) ? 1 : 0
+    s_cbranch_scc1  L_RESTORE_VGPR_LOOP							    //VGPR restore (except v0) is complete?
+    s_set_gpr_idx_off
+											    /* VGPR restore on v0 */
+    if(USE_MTBUF_INSTEAD_OF_MUBUF)
+	tbuffer_load_format_x v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save format:BUF_NUM_FORMAT_FLOAT format: BUF_DATA_FORMAT_32 slc:1 glc:1
+    else
+	buffer_load_dword v0, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save    slc:1 glc:1
+	buffer_load_dword v1, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save    slc:1 glc:1 offset:256
+	buffer_load_dword v2, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save    slc:1 glc:1 offset:256*2
+	buffer_load_dword v3, v0, s_restore_buf_rsrc0, s_restore_mem_offset_save    slc:1 glc:1 offset:256*3
+    end
+
+end
+
+    /*		restore SGPRs	    */
+    //////////////////////////////
+
+    // SGPR SR memory offset : size(VGPR)
+    get_vgpr_size_bytes(s_restore_mem_offset)
+    get_sgpr_size_bytes(s_restore_tmp)
+    s_add_u32 s_restore_mem_offset, s_restore_mem_offset, s_restore_tmp
+    s_sub_u32 s_restore_mem_offset, s_restore_mem_offset, 16*4	   // restore SGPR from S[n] to S[0], by 16 sgprs group
+    // TODO, change RSRC word to rearrange memory layout for SGPRS
+
+    s_getreg_b32    s_restore_alloc_size, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SIZE)		    //spgr_size
+    s_add_u32	    s_restore_alloc_size, s_restore_alloc_size, 1
+    s_lshl_b32	    s_restore_alloc_size, s_restore_alloc_size, 4			    //Number of SGPRs = (sgpr_size + 1) * 16   (non-zero value)
+
+    if (SGPR_SAVE_USE_SQC)
+	s_lshl_b32	s_restore_buf_rsrc2,	s_restore_alloc_size, 2			    //NUM_RECORDS in bytes
+    else
+	s_lshl_b32	s_restore_buf_rsrc2,	s_restore_alloc_size, 8			    //NUM_RECORDS in bytes (64 threads)
+    end
+    if (SWIZZLE_EN)
+	s_add_u32	s_restore_buf_rsrc2, s_restore_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_restore_buf_rsrc2,  0x1000000					    //NUM_RECORDS in bytes
+    end
+
+    s_mov_b32 m0, s_restore_alloc_size
+
+ L_RESTORE_SGPR_LOOP:
+    read_16sgpr_from_mem(s0, s_restore_buf_rsrc0, s_restore_mem_offset)	 //PV: further performance improvement can be made
+    s_waitcnt	    lgkmcnt(0)								    //ensure data ready
+
+    s_sub_u32 m0, m0, 16    // Restore from S[n] to S[0]
+    s_nop 0 // hazard SALU M0=> S_MOVREL
+
+    s_movreld_b64   s0, s0	//s[0+m0] = s0
+    s_movreld_b64   s2, s2
+    s_movreld_b64   s4, s4
+    s_movreld_b64   s6, s6
+    s_movreld_b64   s8, s8
+    s_movreld_b64   s10, s10
+    s_movreld_b64   s12, s12
+    s_movreld_b64   s14, s14
+
+    s_cmp_eq_u32    m0, 0		//scc = (m0 < s_restore_alloc_size) ? 1 : 0
+    s_cbranch_scc0  L_RESTORE_SGPR_LOOP		    //SGPR restore (except s0) is complete?
+
+    /*	    restore HW registers    */
+    //////////////////////////////
+  L_RESTORE_HWREG:
+
+
+if G8SR_DEBUG_TIMESTAMP
+      s_mov_b32 s_g8sr_ts_restore_s[0], s_restore_pc_lo
+      s_mov_b32 s_g8sr_ts_restore_s[1], s_restore_pc_hi
+end
+
+    // HWREG SR memory offset : size(VGPR)+size(SGPR)
+    get_vgpr_size_bytes(s_restore_mem_offset)
+    get_sgpr_size_bytes(s_restore_tmp)
+    s_add_u32 s_restore_mem_offset, s_restore_mem_offset, s_restore_tmp
+
+
+    s_mov_b32	    s_restore_buf_rsrc2, 0x4						    //NUM_RECORDS   in bytes
+    if (SWIZZLE_EN)
+	s_add_u32	s_restore_buf_rsrc2, s_restore_buf_rsrc2, 0x0			    //FIXME need to use swizzle to enable bounds checking?
+    else
+	s_mov_b32	s_restore_buf_rsrc2,  0x1000000					    //NUM_RECORDS in bytes
+    end
+
+    read_hwreg_from_mem(s_restore_m0, s_restore_buf_rsrc0, s_restore_mem_offset)		    //M0
+    read_hwreg_from_mem(s_restore_pc_lo, s_restore_buf_rsrc0, s_restore_mem_offset)		//PC
+    read_hwreg_from_mem(s_restore_pc_hi, s_restore_buf_rsrc0, s_restore_mem_offset)
+    read_hwreg_from_mem(s_restore_exec_lo, s_restore_buf_rsrc0, s_restore_mem_offset)		    //EXEC
+    read_hwreg_from_mem(s_restore_exec_hi, s_restore_buf_rsrc0, s_restore_mem_offset)
+    read_hwreg_from_mem(s_restore_status, s_restore_buf_rsrc0, s_restore_mem_offset)		    //STATUS
+    read_hwreg_from_mem(s_restore_trapsts, s_restore_buf_rsrc0, s_restore_mem_offset)		    //TRAPSTS
+    read_hwreg_from_mem(xnack_mask_lo, s_restore_buf_rsrc0, s_restore_mem_offset)		    //XNACK_MASK_LO
+    read_hwreg_from_mem(xnack_mask_hi, s_restore_buf_rsrc0, s_restore_mem_offset)		    //XNACK_MASK_HI
+    read_hwreg_from_mem(s_restore_mode, s_restore_buf_rsrc0, s_restore_mem_offset)		//MODE
+
+    s_waitcnt	    lgkmcnt(0)											    //from now on, it is safe to restore STATUS and IB_STS
+
+    //for normal save & restore, the saved PC points to the next inst to execute, no adjustment needs to be made, otherwise:
+    if ((EMU_RUN_HACK) && (!EMU_RUN_HACK_RESTORE_NORMAL))
+	s_add_u32 s_restore_pc_lo, s_restore_pc_lo, 8		 //pc[31:0]+8	  //two back-to-back s_trap are used (first for save and second for restore)
+	s_addc_u32  s_restore_pc_hi, s_restore_pc_hi, 0x0	 //carry bit over
+    end
+    if ((EMU_RUN_HACK) && (EMU_RUN_HACK_RESTORE_NORMAL))
+	s_add_u32 s_restore_pc_lo, s_restore_pc_lo, 4		 //pc[31:0]+4	  // save is hack through s_trap but restore is normal
+	s_addc_u32  s_restore_pc_hi, s_restore_pc_hi, 0x0	 //carry bit over
+    end
+
+    s_mov_b32	    m0,		s_restore_m0
+    s_mov_b32	    exec_lo,	s_restore_exec_lo
+    s_mov_b32	    exec_hi,	s_restore_exec_hi
+
+    s_and_b32	    s_restore_m0, SQ_WAVE_TRAPSTS_PRE_SAVECTX_MASK, s_restore_trapsts
+    s_setreg_b32    hwreg(HW_REG_TRAPSTS, SQ_WAVE_TRAPSTS_PRE_SAVECTX_SHIFT, SQ_WAVE_TRAPSTS_PRE_SAVECTX_SIZE), s_restore_m0
+    s_and_b32	    s_restore_m0, SQ_WAVE_TRAPSTS_POST_SAVECTX_MASK, s_restore_trapsts
+    s_lshr_b32	    s_restore_m0, s_restore_m0, SQ_WAVE_TRAPSTS_POST_SAVECTX_SHIFT
+    s_setreg_b32    hwreg(HW_REG_TRAPSTS, SQ_WAVE_TRAPSTS_POST_SAVECTX_SHIFT, SQ_WAVE_TRAPSTS_POST_SAVECTX_SIZE), s_restore_m0
+    //s_setreg_b32  hwreg(HW_REG_TRAPSTS),  s_restore_trapsts	   //don't overwrite SAVECTX bit as it may be set through external SAVECTX during restore
+    s_setreg_b32    hwreg(HW_REG_MODE),	    s_restore_mode
+
+    // Restore trap temporaries 6-11, 13-15 initialized by SPI debug dispatch logic
+    // ttmp SR memory offset : size(VGPR)+size(SGPR)+0x40
+    get_vgpr_size_bytes(s_restore_ttmps_lo)
+    get_sgpr_size_bytes(s_restore_ttmps_hi)
+    s_add_u32	    s_restore_ttmps_lo, s_restore_ttmps_lo, s_restore_ttmps_hi
+    s_add_u32	    s_restore_ttmps_lo, s_restore_ttmps_lo, s_restore_buf_rsrc0
+    s_addc_u32	    s_restore_ttmps_hi, s_restore_buf_rsrc1, 0x0
+    s_and_b32	    s_restore_ttmps_hi, s_restore_ttmps_hi, 0xFFFF
+    s_load_dwordx2  [ttmp6, ttmp7], [s_restore_ttmps_lo, s_restore_ttmps_hi], 0x40 glc:1
+    s_load_dwordx4  [ttmp8, ttmp9, ttmp10, ttmp11], [s_restore_ttmps_lo, s_restore_ttmps_hi], 0x48 glc:1
+    s_load_dword    ttmp13, [s_restore_ttmps_lo, s_restore_ttmps_hi], 0x58 glc:1
+    s_load_dwordx2  [ttmp14, ttmp15], [s_restore_ttmps_lo, s_restore_ttmps_hi], 0x5C glc:1
+    s_waitcnt	    lgkmcnt(0)
+
+    //reuse s_restore_m0 as a temp register
+    s_and_b32	    s_restore_m0, s_restore_pc_hi, S_SAVE_PC_HI_RCNT_MASK
+    s_lshr_b32	    s_restore_m0, s_restore_m0, S_SAVE_PC_HI_RCNT_SHIFT
+    s_lshl_b32	    s_restore_m0, s_restore_m0, SQ_WAVE_IB_STS_RCNT_SHIFT
+    s_mov_b32	    s_restore_tmp, 0x0										    //IB_STS is zero
+    s_or_b32	    s_restore_tmp, s_restore_tmp, s_restore_m0
+    s_and_b32	    s_restore_m0, s_restore_pc_hi, S_SAVE_PC_HI_FIRST_REPLAY_MASK
+    s_lshr_b32	    s_restore_m0, s_restore_m0, S_SAVE_PC_HI_FIRST_REPLAY_SHIFT
+    s_lshl_b32	    s_restore_m0, s_restore_m0, SQ_WAVE_IB_STS_FIRST_REPLAY_SHIFT
+    s_or_b32	    s_restore_tmp, s_restore_tmp, s_restore_m0
+    s_and_b32	    s_restore_m0, s_restore_status, SQ_WAVE_STATUS_INST_ATC_MASK
+    s_lshr_b32	    s_restore_m0, s_restore_m0, SQ_WAVE_STATUS_INST_ATC_SHIFT
+    s_setreg_b32    hwreg(HW_REG_IB_STS),   s_restore_tmp
+
+    s_and_b32 s_restore_pc_hi, s_restore_pc_hi, 0x0000ffff	//pc[47:32]	   //Do it here in order not to affect STATUS
+    s_and_b64	 exec, exec, exec  // Restore STATUS.EXECZ, not writable by s_setreg_b32
+    s_and_b64	 vcc, vcc, vcc	// Restore STATUS.VCCZ, not writable by s_setreg_b32
+    s_setreg_b32    hwreg(HW_REG_STATUS),   s_restore_status	 // SCC is included, which is changed by previous salu
+
+    s_barrier							//barrier to ensure the readiness of LDS before access attempts from any other wave in the same TG //FIXME not performance-optimal at this time
+
+if G8SR_DEBUG_TIMESTAMP
+    s_memrealtime s_g8sr_ts_restore_d
+    s_waitcnt lgkmcnt(0)
+end
+
+//  s_rfe_b64 s_restore_pc_lo					//Return to the main shader program and resume execution
+    s_rfe_restore_b64  s_restore_pc_lo, s_restore_m0		// s_restore_m0[0] is used to set STATUS.inst_atc
+
+
+/**************************************************************************/
+/*			the END						  */
+/**************************************************************************/
+L_END_PGM:
+    s_endpgm
+
+end
+
+
+/**************************************************************************/
+/*			the helper functions				  */
+/**************************************************************************/
+
+//Only for save hwreg to mem
+function write_hwreg_to_mem(s, s_rsrc, s_mem_offset)
+	s_mov_b32 exec_lo, m0			//assuming exec_lo is not needed anymore from this point on
+	s_mov_b32 m0, s_mem_offset
+	s_buffer_store_dword s, s_rsrc, m0	glc:1
+	ack_sqc_store_workaround()
+	s_add_u32	s_mem_offset, s_mem_offset, 4
+	s_mov_b32   m0, exec_lo
+end
+
+
+// HWREG are saved before SGPRs, so all HWREG could be use.
+function write_16sgpr_to_mem(s, s_rsrc, s_mem_offset)
+
+	s_buffer_store_dwordx4 s[0], s_rsrc, 0	glc:1
+	ack_sqc_store_workaround()
+	s_buffer_store_dwordx4 s[4], s_rsrc, 16	 glc:1
+	ack_sqc_store_workaround()
+	s_buffer_store_dwordx4 s[8], s_rsrc, 32	 glc:1
+	ack_sqc_store_workaround()
+	s_buffer_store_dwordx4 s[12], s_rsrc, 48 glc:1
+	ack_sqc_store_workaround()
+	s_add_u32	s_rsrc[0], s_rsrc[0], 4*16
+	s_addc_u32	s_rsrc[1], s_rsrc[1], 0x0	      // +scc
+end
+
+
+function read_hwreg_from_mem(s, s_rsrc, s_mem_offset)
+    s_buffer_load_dword s, s_rsrc, s_mem_offset	    glc:1
+    s_add_u32	    s_mem_offset, s_mem_offset, 4
+end
+
+function read_16sgpr_from_mem(s, s_rsrc, s_mem_offset)
+    s_buffer_load_dwordx16 s, s_rsrc, s_mem_offset	glc:1
+    s_sub_u32	    s_mem_offset, s_mem_offset, 4*16
+end
+
+
+
+function get_lds_size_bytes(s_lds_size_byte)
+    // SQ LDS granularity is 64DW, while PGM_RSRC2.lds_size is in granularity 128DW
+    s_getreg_b32   s_lds_size_byte, hwreg(HW_REG_LDS_ALLOC, SQ_WAVE_LDS_ALLOC_LDS_SIZE_SHIFT, SQ_WAVE_LDS_ALLOC_LDS_SIZE_SIZE)		// lds_size
+    s_lshl_b32	   s_lds_size_byte, s_lds_size_byte, 8			    //LDS size in dwords = lds_size * 64 *4Bytes    // granularity 64DW
+end
+
+function get_vgpr_size_bytes(s_vgpr_size_byte)
+    s_getreg_b32   s_vgpr_size_byte, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_VGPR_SIZE_SIZE)	 //vpgr_size
+    s_add_u32	   s_vgpr_size_byte, s_vgpr_size_byte, 1
+    s_lshl_b32	   s_vgpr_size_byte, s_vgpr_size_byte, (2+8) //Number of VGPRs = (vgpr_size + 1) * 4 * 64 * 4	(non-zero value)   //FIXME for GFX, zero is possible
+end
+
+function get_sgpr_size_bytes(s_sgpr_size_byte)
+    s_getreg_b32   s_sgpr_size_byte, hwreg(HW_REG_GPR_ALLOC,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SHIFT,SQ_WAVE_GPR_ALLOC_SGPR_SIZE_SIZE)	 //spgr_size
+    s_add_u32	   s_sgpr_size_byte, s_sgpr_size_byte, 1
+    s_lshl_b32	   s_sgpr_size_byte, s_sgpr_size_byte, 6 //Number of SGPRs = (sgpr_size + 1) * 16 *4   (non-zero value)
+end
+
+function get_hwreg_size_bytes
+    return 128 //HWREG size 128 bytes
+end
+
+function ack_sqc_store_workaround
+    if ACK_SQC_STORE
+        s_waitcnt lgkmcnt(0)
+    end
+end
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
index 59808a39ecf4..f64c5551cdba 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -233,7 +233,7 @@ static int set_queue_properties_from_user(struct queue_properties *q_properties,
 	pr_debug("Queue Size: 0x%llX, %u\n",
 			q_properties->queue_size, args->ring_size);
 
-	pr_debug("Queue r/w Pointers: %p, %p\n",
+	pr_debug("Queue r/w Pointers: %px, %px\n",
 			q_properties->read_ptr,
 			q_properties->write_ptr);
 
@@ -292,8 +292,16 @@ static int kfd_ioctl_create_queue(struct file *filep, struct kfd_process *p,
 
 
 	/* Return gpu_id as doorbell offset for mmap usage */
-	args->doorbell_offset = (KFD_MMAP_DOORBELL_MASK | args->gpu_id);
+	args->doorbell_offset = KFD_MMAP_TYPE_DOORBELL;
+	args->doorbell_offset |= KFD_MMAP_GPU_ID(args->gpu_id);
 	args->doorbell_offset <<= PAGE_SHIFT;
+	if (KFD_IS_SOC15(dev->device_info->asic_family))
+		/* On SOC15 ASICs, doorbell allocation must be
+		 * per-device, and independent from the per-process
+		 * queue_id. Return the doorbell offset within the
+		 * doorbell aperture to user mode.
+		 */
+		args->doorbell_offset |= q_properties.doorbell_off;
 
 	mutex_unlock(&p->mutex);
 
@@ -1296,8 +1304,8 @@ static int kfd_ioctl_map_memory_to_gpu(struct file *filep,
 		return -EINVAL;
 	}
 
-	devices_arr = kmalloc(args->n_devices * sizeof(*devices_arr),
-			      GFP_KERNEL);
+	devices_arr = kmalloc_array(args->n_devices, sizeof(*devices_arr),
+				    GFP_KERNEL);
 	if (!devices_arr)
 		return -ENOMEM;
 
@@ -1405,8 +1413,8 @@ static int kfd_ioctl_unmap_memory_from_gpu(struct file *filep,
 		return -EINVAL;
 	}
 
-	devices_arr = kmalloc(args->n_devices * sizeof(*devices_arr),
-			      GFP_KERNEL);
+	devices_arr = kmalloc_array(args->n_devices, sizeof(*devices_arr),
+				    GFP_KERNEL);
 	if (!devices_arr)
 		return -ENOMEM;
 
@@ -1645,23 +1653,33 @@ err_i1:
 static int kfd_mmap(struct file *filp, struct vm_area_struct *vma)
 {
 	struct kfd_process *process;
+	struct kfd_dev *dev = NULL;
+	unsigned long vm_pgoff;
+	unsigned int gpu_id;
 
 	process = kfd_get_process(current);
 	if (IS_ERR(process))
 		return PTR_ERR(process);
 
-	if ((vma->vm_pgoff & KFD_MMAP_DOORBELL_MASK) ==
-			KFD_MMAP_DOORBELL_MASK) {
-		vma->vm_pgoff = vma->vm_pgoff ^ KFD_MMAP_DOORBELL_MASK;
-		return kfd_doorbell_mmap(process, vma);
-	} else if ((vma->vm_pgoff & KFD_MMAP_EVENTS_MASK) ==
-			KFD_MMAP_EVENTS_MASK) {
-		vma->vm_pgoff = vma->vm_pgoff ^ KFD_MMAP_EVENTS_MASK;
+	vm_pgoff = vma->vm_pgoff;
+	vma->vm_pgoff = KFD_MMAP_OFFSET_VALUE_GET(vm_pgoff);
+	gpu_id = KFD_MMAP_GPU_ID_GET(vm_pgoff);
+	if (gpu_id)
+		dev = kfd_device_by_id(gpu_id);
+
+	switch (vm_pgoff & KFD_MMAP_TYPE_MASK) {
+	case KFD_MMAP_TYPE_DOORBELL:
+		if (!dev)
+			return -ENODEV;
+		return kfd_doorbell_mmap(dev, process, vma);
+
+	case KFD_MMAP_TYPE_EVENTS:
 		return kfd_event_mmap(process, vma);
-	} else if ((vma->vm_pgoff & KFD_MMAP_RESERVED_MEM_MASK) ==
-			KFD_MMAP_RESERVED_MEM_MASK) {
-		vma->vm_pgoff = vma->vm_pgoff ^ KFD_MMAP_RESERVED_MEM_MASK;
-		return kfd_reserved_mem_mmap(process, vma);
+
+	case KFD_MMAP_TYPE_RESERVED_MEM:
+		if (!dev)
+			return -ENODEV;
+		return kfd_reserved_mem_mmap(dev, process, vma);
 	}
 
 	return -EFAULT;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
index 4f126ef6139b..296b3f230280 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
@@ -132,6 +132,9 @@ static struct kfd_gpu_cache_info carrizo_cache_info[] = {
 #define fiji_cache_info  carrizo_cache_info
 #define polaris10_cache_info carrizo_cache_info
 #define polaris11_cache_info carrizo_cache_info
+/* TODO - check & update Vega10 cache details */
+#define vega10_cache_info carrizo_cache_info
+#define raven_cache_info carrizo_cache_info
 
 static void kfd_populated_cu_info_cpu(struct kfd_topology_device *dev,
 		struct crat_subtype_computeunit *cu)
@@ -603,6 +606,14 @@ static int kfd_fill_gpu_cache_info(struct kfd_dev *kdev,
 		pcache_info = polaris11_cache_info;
 		num_of_cache_types = ARRAY_SIZE(polaris11_cache_info);
 		break;
+	case CHIP_VEGA10:
+		pcache_info = vega10_cache_info;
+		num_of_cache_types = ARRAY_SIZE(vega10_cache_info);
+		break;
+	case CHIP_RAVEN:
+		pcache_info = raven_cache_info;
+		num_of_cache_types = ARRAY_SIZE(raven_cache_info);
+		break;
 	default:
 		return -EINVAL;
 	}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index 3346699960dd..7ee6cec2c060 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -20,16 +20,13 @@
  * OTHER DEALINGS IN THE SOFTWARE.
  */
 
-#if defined(CONFIG_AMD_IOMMU_V2_MODULE) || defined(CONFIG_AMD_IOMMU_V2)
-#include <linux/amd-iommu.h>
-#endif
 #include <linux/bsearch.h>
 #include <linux/pci.h>
 #include <linux/slab.h>
 #include "kfd_priv.h"
 #include "kfd_device_queue_manager.h"
 #include "kfd_pm4_headers_vi.h"
-#include "cwsr_trap_handler_gfx8.asm"
+#include "cwsr_trap_handler.h"
 #include "kfd_iommu.h"
 
 #define MQD_SIZE_ALIGNED 768
@@ -41,6 +38,7 @@ static const struct kfd_device_info kaveri_device_info = {
 	.max_pasid_bits = 16,
 	/* max num of queues for KV.TODO should be a dynamic value */
 	.max_no_of_hqd	= 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -55,6 +53,7 @@ static const struct kfd_device_info carrizo_device_info = {
 	.max_pasid_bits = 16,
 	/* max num of queues for CZ.TODO should be a dynamic value */
 	.max_no_of_hqd	= 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -70,6 +69,7 @@ static const struct kfd_device_info hawaii_device_info = {
 	.max_pasid_bits = 16,
 	/* max num of queues for KV.TODO should be a dynamic value */
 	.max_no_of_hqd	= 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -83,6 +83,7 @@ static const struct kfd_device_info tonga_device_info = {
 	.asic_family = CHIP_TONGA,
 	.max_pasid_bits = 16,
 	.max_no_of_hqd  = 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -96,6 +97,7 @@ static const struct kfd_device_info tonga_vf_device_info = {
 	.asic_family = CHIP_TONGA,
 	.max_pasid_bits = 16,
 	.max_no_of_hqd  = 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -109,6 +111,7 @@ static const struct kfd_device_info fiji_device_info = {
 	.asic_family = CHIP_FIJI,
 	.max_pasid_bits = 16,
 	.max_no_of_hqd  = 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -122,6 +125,7 @@ static const struct kfd_device_info fiji_vf_device_info = {
 	.asic_family = CHIP_FIJI,
 	.max_pasid_bits = 16,
 	.max_no_of_hqd  = 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -136,6 +140,7 @@ static const struct kfd_device_info polaris10_device_info = {
 	.asic_family = CHIP_POLARIS10,
 	.max_pasid_bits = 16,
 	.max_no_of_hqd  = 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -149,6 +154,7 @@ static const struct kfd_device_info polaris10_vf_device_info = {
 	.asic_family = CHIP_POLARIS10,
 	.max_pasid_bits = 16,
 	.max_no_of_hqd  = 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -162,6 +168,7 @@ static const struct kfd_device_info polaris11_device_info = {
 	.asic_family = CHIP_POLARIS11,
 	.max_pasid_bits = 16,
 	.max_no_of_hqd  = 24,
+	.doorbell_size  = 4,
 	.ih_ring_entry_size = 4 * sizeof(uint32_t),
 	.event_interrupt_class = &event_interrupt_class_cik,
 	.num_of_watch_points = 4,
@@ -171,6 +178,34 @@ static const struct kfd_device_info polaris11_device_info = {
 	.needs_pci_atomics = true,
 };
 
+static const struct kfd_device_info vega10_device_info = {
+	.asic_family = CHIP_VEGA10,
+	.max_pasid_bits = 16,
+	.max_no_of_hqd  = 24,
+	.doorbell_size  = 8,
+	.ih_ring_entry_size = 8 * sizeof(uint32_t),
+	.event_interrupt_class = &event_interrupt_class_v9,
+	.num_of_watch_points = 4,
+	.mqd_size_aligned = MQD_SIZE_ALIGNED,
+	.supports_cwsr = true,
+	.needs_iommu_device = false,
+	.needs_pci_atomics = false,
+};
+
+static const struct kfd_device_info vega10_vf_device_info = {
+	.asic_family = CHIP_VEGA10,
+	.max_pasid_bits = 16,
+	.max_no_of_hqd  = 24,
+	.doorbell_size  = 8,
+	.ih_ring_entry_size = 8 * sizeof(uint32_t),
+	.event_interrupt_class = &event_interrupt_class_v9,
+	.num_of_watch_points = 4,
+	.mqd_size_aligned = MQD_SIZE_ALIGNED,
+	.supports_cwsr = true,
+	.needs_iommu_device = false,
+	.needs_pci_atomics = false,
+};
+
 
 struct kfd_deviceid {
 	unsigned short did;
@@ -250,6 +285,15 @@ static const struct kfd_deviceid supported_devices[] = {
 	{ 0x67EB, &polaris11_device_info },	/* Polaris11 */
 	{ 0x67EF, &polaris11_device_info },	/* Polaris11 */
 	{ 0x67FF, &polaris11_device_info },	/* Polaris11 */
+	{ 0x6860, &vega10_device_info },	/* Vega10 */
+	{ 0x6861, &vega10_device_info },	/* Vega10 */
+	{ 0x6862, &vega10_device_info },	/* Vega10 */
+	{ 0x6863, &vega10_device_info },	/* Vega10 */
+	{ 0x6864, &vega10_device_info },	/* Vega10 */
+	{ 0x6867, &vega10_device_info },	/* Vega10 */
+	{ 0x6868, &vega10_device_info },	/* Vega10 */
+	{ 0x686C, &vega10_vf_device_info },	/* Vega10  vf*/
+	{ 0x687F, &vega10_device_info },	/* Vega10 */
 };
 
 static int kfd_gtt_sa_init(struct kfd_dev *kfd, unsigned int buf_size,
@@ -279,7 +323,7 @@ struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd,
 	struct pci_dev *pdev, const struct kfd2kgd_calls *f2g)
 {
 	struct kfd_dev *kfd;
-
+	int ret;
 	const struct kfd_device_info *device_info =
 					lookup_device_info(pdev->device);
 
@@ -288,19 +332,18 @@ struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd,
 		return NULL;
 	}
 
-	if (device_info->needs_pci_atomics) {
-		/* Allow BIF to recode atomics to PCIe 3.0
-		 * AtomicOps. 32 and 64-bit requests are possible and
-		 * must be supported.
-		 */
-		if (pci_enable_atomic_ops_to_root(pdev,
-				PCI_EXP_DEVCAP2_ATOMIC_COMP32 |
-				PCI_EXP_DEVCAP2_ATOMIC_COMP64) < 0) {
-			dev_info(kfd_device,
-				"skipped device %x:%x, PCI rejects atomics",
-				 pdev->vendor, pdev->device);
-			return NULL;
-		}
+	/* Allow BIF to recode atomics to PCIe 3.0 AtomicOps.
+	 * 32 and 64-bit requests are possible and must be
+	 * supported.
+	 */
+	ret = pci_enable_atomic_ops_to_root(pdev,
+			PCI_EXP_DEVCAP2_ATOMIC_COMP32 |
+			PCI_EXP_DEVCAP2_ATOMIC_COMP64);
+	if (device_info->needs_pci_atomics && ret < 0) {
+		dev_info(kfd_device,
+			 "skipped device %x:%x, PCI rejects atomics\n",
+			 pdev->vendor, pdev->device);
+		return NULL;
 	}
 
 	kfd = kzalloc(sizeof(*kfd), GFP_KERNEL);
@@ -323,10 +366,16 @@ struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd,
 static void kfd_cwsr_init(struct kfd_dev *kfd)
 {
 	if (cwsr_enable && kfd->device_info->supports_cwsr) {
-		BUILD_BUG_ON(sizeof(cwsr_trap_gfx8_hex) > PAGE_SIZE);
+		if (kfd->device_info->asic_family < CHIP_VEGA10) {
+			BUILD_BUG_ON(sizeof(cwsr_trap_gfx8_hex) > PAGE_SIZE);
+			kfd->cwsr_isa = cwsr_trap_gfx8_hex;
+			kfd->cwsr_isa_size = sizeof(cwsr_trap_gfx8_hex);
+		} else {
+			BUILD_BUG_ON(sizeof(cwsr_trap_gfx9_hex) > PAGE_SIZE);
+			kfd->cwsr_isa = cwsr_trap_gfx9_hex;
+			kfd->cwsr_isa_size = sizeof(cwsr_trap_gfx9_hex);
+		}
 
-		kfd->cwsr_isa = cwsr_trap_gfx8_hex;
-		kfd->cwsr_isa_size = sizeof(cwsr_trap_gfx8_hex);
 		kfd->cwsr_enabled = true;
 	}
 }
@@ -541,6 +590,44 @@ void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry)
 	spin_unlock(&kfd->interrupt_lock);
 }
 
+int kgd2kfd_quiesce_mm(struct mm_struct *mm)
+{
+	struct kfd_process *p;
+	int r;
+
+	/* Because we are called from arbitrary context (workqueue) as opposed
+	 * to process context, kfd_process could attempt to exit while we are
+	 * running so the lookup function increments the process ref count.
+	 */
+	p = kfd_lookup_process_by_mm(mm);
+	if (!p)
+		return -ESRCH;
+
+	r = kfd_process_evict_queues(p);
+
+	kfd_unref_process(p);
+	return r;
+}
+
+int kgd2kfd_resume_mm(struct mm_struct *mm)
+{
+	struct kfd_process *p;
+	int r;
+
+	/* Because we are called from arbitrary context (workqueue) as opposed
+	 * to process context, kfd_process could attempt to exit while we are
+	 * running so the lookup function increments the process ref count.
+	 */
+	p = kfd_lookup_process_by_mm(mm);
+	if (!p)
+		return -ESRCH;
+
+	r = kfd_process_restore_queues(p);
+
+	kfd_unref_process(p);
+	return r;
+}
+
 /** kgd2kfd_schedule_evict_and_restore_process - Schedules work queue that will
  *   prepare for safe eviction of KFD BOs that belong to the specified
  *   process.
@@ -652,7 +739,7 @@ int kfd_gtt_sa_allocate(struct kfd_dev *kfd, unsigned int size,
 	if (size > kfd->gtt_sa_num_of_chunks * kfd->gtt_sa_chunk_size)
 		return -ENOMEM;
 
-	*mem_obj = kmalloc(sizeof(struct kfd_mem_obj), GFP_KERNEL);
+	*mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_NOIO);
 	if ((*mem_obj) == NULL)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
index d55d29d31da4..668ad07ebe1f 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c
@@ -110,6 +110,57 @@ void program_sh_mem_settings(struct device_queue_manager *dqm,
 						qpd->sh_mem_bases);
 }
 
+static int allocate_doorbell(struct qcm_process_device *qpd, struct queue *q)
+{
+	struct kfd_dev *dev = qpd->dqm->dev;
+
+	if (!KFD_IS_SOC15(dev->device_info->asic_family)) {
+		/* On pre-SOC15 chips we need to use the queue ID to
+		 * preserve the user mode ABI.
+		 */
+		q->doorbell_id = q->properties.queue_id;
+	} else if (q->properties.type == KFD_QUEUE_TYPE_SDMA) {
+		/* For SDMA queues on SOC15, use static doorbell
+		 * assignments based on the engine and queue.
+		 */
+		q->doorbell_id = dev->shared_resources.sdma_doorbell
+			[q->properties.sdma_engine_id]
+			[q->properties.sdma_queue_id];
+	} else {
+		/* For CP queues on SOC15 reserve a free doorbell ID */
+		unsigned int found;
+
+		found = find_first_zero_bit(qpd->doorbell_bitmap,
+					    KFD_MAX_NUM_OF_QUEUES_PER_PROCESS);
+		if (found >= KFD_MAX_NUM_OF_QUEUES_PER_PROCESS) {
+			pr_debug("No doorbells available");
+			return -EBUSY;
+		}
+		set_bit(found, qpd->doorbell_bitmap);
+		q->doorbell_id = found;
+	}
+
+	q->properties.doorbell_off =
+		kfd_doorbell_id_to_offset(dev, q->process,
+					  q->doorbell_id);
+
+	return 0;
+}
+
+static void deallocate_doorbell(struct qcm_process_device *qpd,
+				struct queue *q)
+{
+	unsigned int old;
+	struct kfd_dev *dev = qpd->dqm->dev;
+
+	if (!KFD_IS_SOC15(dev->device_info->asic_family) ||
+	    q->properties.type == KFD_QUEUE_TYPE_SDMA)
+		return;
+
+	old = test_and_clear_bit(q->doorbell_id, qpd->doorbell_bitmap);
+	WARN_ON(!old);
+}
+
 static int allocate_vmid(struct device_queue_manager *dqm,
 			struct qcm_process_device *qpd,
 			struct queue *q)
@@ -145,15 +196,19 @@ static int allocate_vmid(struct device_queue_manager *dqm,
 static int flush_texture_cache_nocpsch(struct kfd_dev *kdev,
 				struct qcm_process_device *qpd)
 {
-	uint32_t len;
+	const struct packet_manager_funcs *pmf = qpd->dqm->packets.pmf;
+	int ret;
 
 	if (!qpd->ib_kaddr)
 		return -ENOMEM;
 
-	len = pm_create_release_mem(qpd->ib_base, (uint32_t *)qpd->ib_kaddr);
+	ret = pmf->release_mem(qpd->ib_base, (uint32_t *)qpd->ib_kaddr);
+	if (ret)
+		return ret;
 
 	return kdev->kfd2kgd->submit_ib(kdev->kgd, KGD_ENGINE_MEC1, qpd->vmid,
-				qpd->ib_base, (uint32_t *)qpd->ib_kaddr, len);
+				qpd->ib_base, (uint32_t *)qpd->ib_kaddr,
+				pmf->release_mem_size / sizeof(uint32_t));
 }
 
 static void deallocate_vmid(struct device_queue_manager *dqm,
@@ -301,10 +356,14 @@ static int create_compute_queue_nocpsch(struct device_queue_manager *dqm,
 	if (retval)
 		return retval;
 
+	retval = allocate_doorbell(qpd, q);
+	if (retval)
+		goto out_deallocate_hqd;
+
 	retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj,
 				&q->gart_mqd_addr, &q->properties);
 	if (retval)
-		goto out_deallocate_hqd;
+		goto out_deallocate_doorbell;
 
 	pr_debug("Loading mqd to hqd on pipe %d, queue %d\n",
 			q->pipe, q->queue);
@@ -324,6 +383,8 @@ static int create_compute_queue_nocpsch(struct device_queue_manager *dqm,
 
 out_uninit_mqd:
 	mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj);
+out_deallocate_doorbell:
+	deallocate_doorbell(qpd, q);
 out_deallocate_hqd:
 	deallocate_hqd(dqm, q);
 
@@ -357,6 +418,8 @@ static int destroy_queue_nocpsch_locked(struct device_queue_manager *dqm,
 	}
 	dqm->total_queue_count--;
 
+	deallocate_doorbell(qpd, q);
+
 	retval = mqd->destroy_mqd(mqd, q->mqd,
 				KFD_PREEMPT_TYPE_WAVEFRONT_RESET,
 				KFD_UNMAP_LATENCY_MS,
@@ -861,6 +924,10 @@ static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm,
 	q->properties.sdma_queue_id = q->sdma_id / CIK_SDMA_QUEUES_PER_ENGINE;
 	q->properties.sdma_engine_id = q->sdma_id % CIK_SDMA_QUEUES_PER_ENGINE;
 
+	retval = allocate_doorbell(qpd, q);
+	if (retval)
+		goto out_deallocate_sdma_queue;
+
 	pr_debug("SDMA id is:    %d\n", q->sdma_id);
 	pr_debug("SDMA queue id: %d\n", q->properties.sdma_queue_id);
 	pr_debug("SDMA engine id: %d\n", q->properties.sdma_engine_id);
@@ -869,7 +936,7 @@ static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm,
 	retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj,
 				&q->gart_mqd_addr, &q->properties);
 	if (retval)
-		goto out_deallocate_sdma_queue;
+		goto out_deallocate_doorbell;
 
 	retval = mqd->load_mqd(mqd, q->mqd, 0, 0, &q->properties, NULL);
 	if (retval)
@@ -879,6 +946,8 @@ static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm,
 
 out_uninit_mqd:
 	mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj);
+out_deallocate_doorbell:
+	deallocate_doorbell(qpd, q);
 out_deallocate_sdma_queue:
 	deallocate_sdma_queue(dqm, q->sdma_id);
 
@@ -1070,12 +1139,17 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q,
 		q->properties.sdma_engine_id =
 			q->sdma_id % CIK_SDMA_QUEUES_PER_ENGINE;
 	}
+
+	retval = allocate_doorbell(qpd, q);
+	if (retval)
+		goto out_deallocate_sdma_queue;
+
 	mqd = dqm->ops.get_mqd_manager(dqm,
 			get_mqd_type_from_queue_type(q->properties.type));
 
 	if (!mqd) {
 		retval = -ENOMEM;
-		goto out_deallocate_sdma_queue;
+		goto out_deallocate_doorbell;
 	}
 	/*
 	 * Eviction state logic: we only mark active queues as evicted
@@ -1093,7 +1167,7 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q,
 	retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj,
 				&q->gart_mqd_addr, &q->properties);
 	if (retval)
-		goto out_deallocate_sdma_queue;
+		goto out_deallocate_doorbell;
 
 	list_add(&q->list, &qpd->queues_list);
 	qpd->queue_count++;
@@ -1117,6 +1191,8 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q,
 	mutex_unlock(&dqm->lock);
 	return retval;
 
+out_deallocate_doorbell:
+	deallocate_doorbell(qpd, q);
 out_deallocate_sdma_queue:
 	if (q->properties.type == KFD_QUEUE_TYPE_SDMA)
 		deallocate_sdma_queue(dqm, q->sdma_id);
@@ -1257,6 +1333,8 @@ static int destroy_queue_cpsch(struct device_queue_manager *dqm,
 		goto failed;
 	}
 
+	deallocate_doorbell(qpd, q);
+
 	if (q->properties.type == KFD_QUEUE_TYPE_SDMA) {
 		dqm->sdma_queue_count--;
 		deallocate_sdma_queue(dqm, q->sdma_id);
@@ -1308,7 +1386,10 @@ static bool set_cache_memory_policy(struct device_queue_manager *dqm,
 				   void __user *alternate_aperture_base,
 				   uint64_t alternate_aperture_size)
 {
-	bool retval;
+	bool retval = true;
+
+	if (!dqm->asic_ops.set_cache_memory_policy)
+		return retval;
 
 	mutex_lock(&dqm->lock);
 
@@ -1577,6 +1658,11 @@ struct device_queue_manager *device_queue_manager_init(struct kfd_dev *dev)
 	case CHIP_POLARIS11:
 		device_queue_manager_init_vi_tonga(&dqm->asic_ops);
 		break;
+
+	case CHIP_VEGA10:
+	case CHIP_RAVEN:
+		device_queue_manager_init_v9(&dqm->asic_ops);
+		break;
 	default:
 		WARN(1, "Unexpected ASIC family %u",
 		     dev->device_info->asic_family);
@@ -1627,6 +1713,18 @@ int dqm_debugfs_hqds(struct seq_file *m, void *data)
 	int pipe, queue;
 	int r = 0;
 
+	r = dqm->dev->kfd2kgd->hqd_dump(dqm->dev->kgd,
+		KFD_CIK_HIQ_PIPE, KFD_CIK_HIQ_QUEUE, &dump, &n_regs);
+	if (!r) {
+		seq_printf(m, "  HIQ on MEC %d Pipe %d Queue %d\n",
+				KFD_CIK_HIQ_PIPE/get_pipes_per_mec(dqm)+1,
+				KFD_CIK_HIQ_PIPE%get_pipes_per_mec(dqm),
+				KFD_CIK_HIQ_QUEUE);
+		seq_reg_dump(m, dump, n_regs);
+
+		kfree(dump);
+	}
+
 	for (pipe = 0; pipe < get_pipes_per_mec(dqm); pipe++) {
 		int pipe_offset = pipe * get_queues_per_pipe(dqm);
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
index 412beff3281d..59a6b1956932 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h
@@ -200,6 +200,8 @@ void device_queue_manager_init_vi(
 		struct device_queue_manager_asic_ops *asic_ops);
 void device_queue_manager_init_vi_tonga(
 		struct device_queue_manager_asic_ops *asic_ops);
+void device_queue_manager_init_v9(
+		struct device_queue_manager_asic_ops *asic_ops);
 void program_sh_mem_settings(struct device_queue_manager *dqm,
 					struct qcm_process_device *qpd);
 unsigned int get_queues_num(struct device_queue_manager *dqm);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c
new file mode 100644
index 000000000000..79e5bcf6367c
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c
@@ -0,0 +1,84 @@
+/*
+ * Copyright 2016-2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "kfd_device_queue_manager.h"
+#include "vega10_enum.h"
+#include "gc/gc_9_0_offset.h"
+#include "gc/gc_9_0_sh_mask.h"
+#include "sdma0/sdma0_4_0_sh_mask.h"
+
+static int update_qpd_v9(struct device_queue_manager *dqm,
+			 struct qcm_process_device *qpd);
+static void init_sdma_vm_v9(struct device_queue_manager *dqm, struct queue *q,
+			    struct qcm_process_device *qpd);
+
+void device_queue_manager_init_v9(
+	struct device_queue_manager_asic_ops *asic_ops)
+{
+	asic_ops->update_qpd = update_qpd_v9;
+	asic_ops->init_sdma_vm = init_sdma_vm_v9;
+}
+
+static uint32_t compute_sh_mem_bases_64bit(struct kfd_process_device *pdd)
+{
+	uint32_t shared_base = pdd->lds_base >> 48;
+	uint32_t private_base = pdd->scratch_base >> 48;
+
+	return (shared_base << SH_MEM_BASES__SHARED_BASE__SHIFT) |
+		private_base;
+}
+
+static int update_qpd_v9(struct device_queue_manager *dqm,
+			 struct qcm_process_device *qpd)
+{
+	struct kfd_process_device *pdd;
+
+	pdd = qpd_to_pdd(qpd);
+
+	/* check if sh_mem_config register already configured */
+	if (qpd->sh_mem_config == 0) {
+		qpd->sh_mem_config =
+				SH_MEM_ALIGNMENT_MODE_UNALIGNED <<
+					SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT;
+		if (vega10_noretry &&
+		    !dqm->dev->device_info->needs_iommu_device)
+			qpd->sh_mem_config |=
+				1 << SH_MEM_CONFIG__RETRY_DISABLE__SHIFT;
+
+		qpd->sh_mem_ape1_limit = 0;
+		qpd->sh_mem_ape1_base = 0;
+	}
+
+	qpd->sh_mem_bases = compute_sh_mem_bases_64bit(pdd);
+
+	pr_debug("sh_mem_bases 0x%X\n", qpd->sh_mem_bases);
+
+	return 0;
+}
+
+static void init_sdma_vm_v9(struct device_queue_manager *dqm, struct queue *q,
+			    struct qcm_process_device *qpd)
+{
+	/* Not needed on SDMAv4 any more */
+	q->properties.sdma_vm_addr = 0;
+}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
index ebb4da14e3df..c3744d89352c 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_doorbell.c
@@ -33,7 +33,6 @@
 
 static DEFINE_IDA(doorbell_ida);
 static unsigned int max_doorbell_slices;
-#define KFD_SIZE_OF_DOORBELL_IN_BYTES 4
 
 /*
  * Each device exposes a doorbell aperture, a PCI MMIO aperture that
@@ -50,9 +49,9 @@ static unsigned int max_doorbell_slices;
  */
 
 /* # of doorbell bytes allocated for each process. */
-static inline size_t doorbell_process_allocation(void)
+size_t kfd_doorbell_process_slice(struct kfd_dev *kfd)
 {
-	return roundup(KFD_SIZE_OF_DOORBELL_IN_BYTES *
+	return roundup(kfd->device_info->doorbell_size *
 			KFD_MAX_NUM_OF_QUEUES_PER_PROCESS,
 			PAGE_SIZE);
 }
@@ -72,16 +71,16 @@ int kfd_doorbell_init(struct kfd_dev *kfd)
 
 	doorbell_start_offset =
 			roundup(kfd->shared_resources.doorbell_start_offset,
-					doorbell_process_allocation());
+					kfd_doorbell_process_slice(kfd));
 
 	doorbell_aperture_size =
 			rounddown(kfd->shared_resources.doorbell_aperture_size,
-					doorbell_process_allocation());
+					kfd_doorbell_process_slice(kfd));
 
 	if (doorbell_aperture_size > doorbell_start_offset)
 		doorbell_process_limit =
 			(doorbell_aperture_size - doorbell_start_offset) /
-						doorbell_process_allocation();
+						kfd_doorbell_process_slice(kfd);
 	else
 		return -ENOSPC;
 
@@ -95,7 +94,7 @@ int kfd_doorbell_init(struct kfd_dev *kfd)
 	kfd->doorbell_id_offset = doorbell_start_offset / sizeof(u32);
 
 	kfd->doorbell_kernel_ptr = ioremap(kfd->doorbell_base,
-						doorbell_process_allocation());
+					   kfd_doorbell_process_slice(kfd));
 
 	if (!kfd->doorbell_kernel_ptr)
 		return -ENOMEM;
@@ -127,21 +126,16 @@ void kfd_doorbell_fini(struct kfd_dev *kfd)
 		iounmap(kfd->doorbell_kernel_ptr);
 }
 
-int kfd_doorbell_mmap(struct kfd_process *process, struct vm_area_struct *vma)
+int kfd_doorbell_mmap(struct kfd_dev *dev, struct kfd_process *process,
+		      struct vm_area_struct *vma)
 {
 	phys_addr_t address;
-	struct kfd_dev *dev;
 
 	/*
 	 * For simplicitly we only allow mapping of the entire doorbell
 	 * allocation of a single device & process.
 	 */
-	if (vma->vm_end - vma->vm_start != doorbell_process_allocation())
-		return -EINVAL;
-
-	/* Find kfd device according to gpu id */
-	dev = kfd_device_by_id(vma->vm_pgoff);
-	if (!dev)
+	if (vma->vm_end - vma->vm_start != kfd_doorbell_process_slice(dev))
 		return -EINVAL;
 
 	/* Calculate physical address of doorbell */
@@ -158,19 +152,19 @@ int kfd_doorbell_mmap(struct kfd_process *process, struct vm_area_struct *vma)
 		 "     vm_flags            == 0x%04lX\n"
 		 "     size                == 0x%04lX\n",
 		 (unsigned long long) vma->vm_start, address, vma->vm_flags,
-		 doorbell_process_allocation());
+		 kfd_doorbell_process_slice(dev));
 
 
 	return io_remap_pfn_range(vma,
 				vma->vm_start,
 				address >> PAGE_SHIFT,
-				doorbell_process_allocation(),
+				kfd_doorbell_process_slice(dev),
 				vma->vm_page_prot);
 }
 
 
 /* get kernel iomem pointer for a doorbell */
-u32 __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
+void __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
 					unsigned int *doorbell_off)
 {
 	u32 inx;
@@ -185,6 +179,8 @@ u32 __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
 	if (inx >= KFD_MAX_NUM_OF_QUEUES_PER_PROCESS)
 		return NULL;
 
+	inx *= kfd->device_info->doorbell_size / sizeof(u32);
+
 	/*
 	 * Calculating the kernel doorbell offset using the first
 	 * doorbell page.
@@ -210,7 +206,7 @@ void kfd_release_kernel_doorbell(struct kfd_dev *kfd, u32 __iomem *db_addr)
 	mutex_unlock(&kfd->doorbell_mutex);
 }
 
-inline void write_kernel_doorbell(u32 __iomem *db, u32 value)
+void write_kernel_doorbell(void __iomem *db, u32 value)
 {
 	if (db) {
 		writel(value, db);
@@ -218,30 +214,37 @@ inline void write_kernel_doorbell(u32 __iomem *db, u32 value)
 	}
 }
 
-/*
- * queue_ids are in the range [0,MAX_PROCESS_QUEUES) and are mapped 1:1
- * to doorbells with the process's doorbell page
- */
-unsigned int kfd_queue_id_to_doorbell(struct kfd_dev *kfd,
+void write_kernel_doorbell64(void __iomem *db, u64 value)
+{
+	if (db) {
+		WARN(((unsigned long)db & 7) != 0,
+		     "Unaligned 64-bit doorbell");
+		writeq(value, (u64 __iomem *)db);
+		pr_debug("writing %llu to doorbell address %p\n", value, db);
+	}
+}
+
+unsigned int kfd_doorbell_id_to_offset(struct kfd_dev *kfd,
 					struct kfd_process *process,
-					unsigned int queue_id)
+					unsigned int doorbell_id)
 {
 	/*
 	 * doorbell_id_offset accounts for doorbells taken by KGD.
-	 * index * doorbell_process_allocation/sizeof(u32) adjusts to
-	 * the process's doorbells.
+	 * index * kfd_doorbell_process_slice/sizeof(u32) adjusts to
+	 * the process's doorbells. The offset returned is in dword
+	 * units regardless of the ASIC-dependent doorbell size.
 	 */
 	return kfd->doorbell_id_offset +
 		process->doorbell_index
-		* doorbell_process_allocation() / sizeof(u32) +
-		queue_id;
+		* kfd_doorbell_process_slice(kfd) / sizeof(u32) +
+		doorbell_id * kfd->device_info->doorbell_size / sizeof(u32);
 }
 
 uint64_t kfd_get_number_elems(struct kfd_dev *kfd)
 {
 	uint64_t num_of_elems = (kfd->shared_resources.doorbell_aperture_size -
 				kfd->shared_resources.doorbell_start_offset) /
-					doorbell_process_allocation() + 1;
+					kfd_doorbell_process_slice(kfd) + 1;
 
 	return num_of_elems;
 
@@ -251,7 +254,7 @@ phys_addr_t kfd_get_process_doorbells(struct kfd_dev *dev,
 					struct kfd_process *process)
 {
 	return dev->doorbell_base +
-		process->doorbell_index * doorbell_process_allocation();
+		process->doorbell_index * kfd_doorbell_process_slice(dev);
 }
 
 int kfd_alloc_process_doorbells(struct kfd_process *process)
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
index 4890a90f1e44..5562e94e786a 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
@@ -345,7 +345,7 @@ int kfd_event_create(struct file *devkfd, struct kfd_process *p,
 	case KFD_EVENT_TYPE_DEBUG:
 		ret = create_signal_event(devkfd, p, ev);
 		if (!ret) {
-			*event_page_offset = KFD_MMAP_EVENTS_MASK;
+			*event_page_offset = KFD_MMAP_TYPE_EVENTS;
 			*event_page_offset <<= PAGE_SHIFT;
 			*event_slot_index = ev->event_id;
 		}
@@ -496,7 +496,7 @@ void kfd_signal_event_interrupt(unsigned int pasid, uint32_t partial_id,
 			pr_debug_ratelimited("Partial ID invalid: %u (%u valid bits)\n",
 					     partial_id, valid_id_bits);
 
-		if (p->signal_event_count < KFD_SIGNAL_EVENT_LIMIT/2) {
+		if (p->signal_event_count < KFD_SIGNAL_EVENT_LIMIT / 64) {
 			/* With relatively few events, it's faster to
 			 * iterate over the event IDR
 			 */
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
index 66852de410c8..97d5423c5673 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c
@@ -275,23 +275,35 @@
  * for FLAT_* / S_LOAD operations.
  */
 
-#define MAKE_GPUVM_APP_BASE(gpu_num) \
+#define MAKE_GPUVM_APP_BASE_VI(gpu_num) \
 	(((uint64_t)(gpu_num) << 61) + 0x1000000000000L)
 
 #define MAKE_GPUVM_APP_LIMIT(base, size) \
 	(((uint64_t)(base) & 0xFFFFFF0000000000UL) + (size) - 1)
 
-#define MAKE_SCRATCH_APP_BASE() \
+#define MAKE_SCRATCH_APP_BASE_VI() \
 	(((uint64_t)(0x1UL) << 61) + 0x100000000L)
 
 #define MAKE_SCRATCH_APP_LIMIT(base) \
 	(((uint64_t)base & 0xFFFFFFFF00000000UL) | 0xFFFFFFFF)
 
-#define MAKE_LDS_APP_BASE() \
+#define MAKE_LDS_APP_BASE_VI() \
 	(((uint64_t)(0x1UL) << 61) + 0x0)
 #define MAKE_LDS_APP_LIMIT(base) \
 	(((uint64_t)(base) & 0xFFFFFFFF00000000UL) | 0xFFFFFFFF)
 
+/* On GFXv9 the LDS and scratch apertures are programmed independently
+ * using the high 16 bits of the 64-bit virtual address. They must be
+ * in the hole, which will be the case as long as the high 16 bits are
+ * not 0.
+ *
+ * The aperture sizes are still 4GB implicitly.
+ *
+ * A GPUVM aperture is not applicable on GFXv9.
+ */
+#define MAKE_LDS_APP_BASE_V9() ((uint64_t)(0x1UL) << 48)
+#define MAKE_SCRATCH_APP_BASE_V9() ((uint64_t)(0x2UL) << 48)
+
 /* User mode manages most of the SVM aperture address space. The low
  * 16MB are reserved for kernel use (CWSR trap handler and kernel IB
  * for now).
@@ -300,6 +312,55 @@
 #define SVM_CWSR_BASE (SVM_USER_BASE - KFD_CWSR_TBA_TMA_SIZE)
 #define SVM_IB_BASE   (SVM_CWSR_BASE - PAGE_SIZE)
 
+static void kfd_init_apertures_vi(struct kfd_process_device *pdd, uint8_t id)
+{
+	/*
+	 * node id couldn't be 0 - the three MSB bits of
+	 * aperture shoudn't be 0
+	 */
+	pdd->lds_base = MAKE_LDS_APP_BASE_VI();
+	pdd->lds_limit = MAKE_LDS_APP_LIMIT(pdd->lds_base);
+
+	if (!pdd->dev->device_info->needs_iommu_device) {
+		/* dGPUs: SVM aperture starting at 0
+		 * with small reserved space for kernel.
+		 * Set them to CANONICAL addresses.
+		 */
+		pdd->gpuvm_base = SVM_USER_BASE;
+		pdd->gpuvm_limit =
+			pdd->dev->shared_resources.gpuvm_size - 1;
+	} else {
+		/* set them to non CANONICAL addresses, and no SVM is
+		 * allocated.
+		 */
+		pdd->gpuvm_base = MAKE_GPUVM_APP_BASE_VI(id + 1);
+		pdd->gpuvm_limit = MAKE_GPUVM_APP_LIMIT(pdd->gpuvm_base,
+				pdd->dev->shared_resources.gpuvm_size);
+	}
+
+	pdd->scratch_base = MAKE_SCRATCH_APP_BASE_VI();
+	pdd->scratch_limit = MAKE_SCRATCH_APP_LIMIT(pdd->scratch_base);
+}
+
+static void kfd_init_apertures_v9(struct kfd_process_device *pdd, uint8_t id)
+{
+	pdd->lds_base = MAKE_LDS_APP_BASE_V9();
+	pdd->lds_limit = MAKE_LDS_APP_LIMIT(pdd->lds_base);
+
+	/* Raven needs SVM to support graphic handle, etc. Leave the small
+	 * reserved space before SVM on Raven as well, even though we don't
+	 * have to.
+	 * Set gpuvm_base and gpuvm_limit to CANONICAL addresses so that they
+	 * are used in Thunk to reserve SVM.
+	 */
+	pdd->gpuvm_base = SVM_USER_BASE;
+	pdd->gpuvm_limit =
+		pdd->dev->shared_resources.gpuvm_size - 1;
+
+	pdd->scratch_base = MAKE_SCRATCH_APP_BASE_V9();
+	pdd->scratch_limit = MAKE_SCRATCH_APP_LIMIT(pdd->scratch_base);
+}
+
 int kfd_init_apertures(struct kfd_process *process)
 {
 	uint8_t id  = 0;
@@ -307,9 +368,7 @@ int kfd_init_apertures(struct kfd_process *process)
 	struct kfd_process_device *pdd;
 
 	/*Iterating over all devices*/
-	while (kfd_topology_enum_kfd_devices(id, &dev) == 0 &&
-		id < NUM_OF_SUPPORTED_GPUS) {
-
+	while (kfd_topology_enum_kfd_devices(id, &dev) == 0) {
 		if (!dev) {
 			id++; /* Skip non GPU devices */
 			continue;
@@ -318,7 +377,7 @@ int kfd_init_apertures(struct kfd_process *process)
 		pdd = kfd_create_process_device_data(dev, process);
 		if (!pdd) {
 			pr_err("Failed to create process device data\n");
-			return -1;
+			return -ENOMEM;
 		}
 		/*
 		 * For 64 bit process apertures will be statically reserved in
@@ -330,32 +389,30 @@ int kfd_init_apertures(struct kfd_process *process)
 			pdd->gpuvm_base = pdd->gpuvm_limit = 0;
 			pdd->scratch_base = pdd->scratch_limit = 0;
 		} else {
-			/* Same LDS and scratch apertures can be used
-			 * on all GPUs. This allows using more dGPUs
-			 * than placement options for apertures.
-			 */
-			pdd->lds_base = MAKE_LDS_APP_BASE();
-			pdd->lds_limit = MAKE_LDS_APP_LIMIT(pdd->lds_base);
-
-			pdd->scratch_base = MAKE_SCRATCH_APP_BASE();
-			pdd->scratch_limit =
-				MAKE_SCRATCH_APP_LIMIT(pdd->scratch_base);
+			switch (dev->device_info->asic_family) {
+			case CHIP_KAVERI:
+			case CHIP_HAWAII:
+			case CHIP_CARRIZO:
+			case CHIP_TONGA:
+			case CHIP_FIJI:
+			case CHIP_POLARIS10:
+			case CHIP_POLARIS11:
+				kfd_init_apertures_vi(pdd, id);
+				break;
+			case CHIP_VEGA10:
+			case CHIP_RAVEN:
+				kfd_init_apertures_v9(pdd, id);
+				break;
+			default:
+				WARN(1, "Unexpected ASIC family %u",
+				     dev->device_info->asic_family);
+				return -EINVAL;
+			}
 
-			if (dev->device_info->needs_iommu_device) {
-				/* APUs: GPUVM aperture in
-				 * non-canonical address space
-				 */
-				pdd->gpuvm_base = MAKE_GPUVM_APP_BASE(id + 1);
-				pdd->gpuvm_limit = MAKE_GPUVM_APP_LIMIT(
-					pdd->gpuvm_base,
-					dev->shared_resources.gpuvm_size);
-			} else {
-				/* dGPUs: SVM aperture starting at 0
-				 * with small reserved space for kernel
+			if (!dev->device_info->needs_iommu_device) {
+				/* dGPUs: the reserved space for kernel
+				 * before SVM
 				 */
-				pdd->gpuvm_base = SVM_USER_BASE;
-				pdd->gpuvm_limit =
-					dev->shared_resources.gpuvm_size - 1;
 				pdd->qpd.cwsr_base = SVM_CWSR_BASE;
 				pdd->qpd.ib_base = SVM_IB_BASE;
 			}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
new file mode 100644
index 000000000000..37029baa3346
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_int_process_v9.c
@@ -0,0 +1,92 @@
+/*
+ * Copyright 2016-2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#include "kfd_priv.h"
+#include "kfd_events.h"
+#include "soc15_int.h"
+
+
+static bool event_interrupt_isr_v9(struct kfd_dev *dev,
+					const uint32_t *ih_ring_entry)
+{
+	uint16_t source_id, client_id, pasid, vmid;
+	const uint32_t *data = ih_ring_entry;
+
+	/* Only handle interrupts from KFD VMIDs */
+	vmid = SOC15_VMID_FROM_IH_ENTRY(ih_ring_entry);
+	if (vmid < dev->vm_info.first_vmid_kfd ||
+	    vmid > dev->vm_info.last_vmid_kfd)
+		return 0;
+
+	/* If there is no valid PASID, it's likely a firmware bug */
+	pasid = SOC15_PASID_FROM_IH_ENTRY(ih_ring_entry);
+	if (WARN_ONCE(pasid == 0, "FW bug: No PASID in KFD interrupt"))
+		return 0;
+
+	source_id = SOC15_SOURCE_ID_FROM_IH_ENTRY(ih_ring_entry);
+	client_id = SOC15_CLIENT_ID_FROM_IH_ENTRY(ih_ring_entry);
+
+	pr_debug("client id 0x%x, source id %d, pasid 0x%x. raw data:\n",
+		 client_id, source_id, pasid);
+	pr_debug("%8X, %8X, %8X, %8X, %8X, %8X, %8X, %8X.\n",
+		 data[0], data[1], data[2], data[3],
+		 data[4], data[5], data[6], data[7]);
+
+	/* Interrupt types we care about: various signals and faults.
+	 * They will be forwarded to a work queue (see below).
+	 */
+	return source_id == SOC15_INTSRC_CP_END_OF_PIPE ||
+		source_id == SOC15_INTSRC_SDMA_TRAP ||
+		source_id == SOC15_INTSRC_SQ_INTERRUPT_MSG ||
+		source_id == SOC15_INTSRC_CP_BAD_OPCODE;
+}
+
+static void event_interrupt_wq_v9(struct kfd_dev *dev,
+					const uint32_t *ih_ring_entry)
+{
+	uint16_t source_id, client_id, pasid, vmid;
+	uint32_t context_id;
+
+	source_id = SOC15_SOURCE_ID_FROM_IH_ENTRY(ih_ring_entry);
+	client_id = SOC15_CLIENT_ID_FROM_IH_ENTRY(ih_ring_entry);
+	pasid = SOC15_PASID_FROM_IH_ENTRY(ih_ring_entry);
+	vmid = SOC15_VMID_FROM_IH_ENTRY(ih_ring_entry);
+	context_id = SOC15_CONTEXT_ID0_FROM_IH_ENTRY(ih_ring_entry);
+
+	if (source_id == SOC15_INTSRC_CP_END_OF_PIPE)
+		kfd_signal_event_interrupt(pasid, context_id, 32);
+	else if (source_id == SOC15_INTSRC_SDMA_TRAP)
+		kfd_signal_event_interrupt(pasid, context_id & 0xfffffff, 28);
+	else if (source_id == SOC15_INTSRC_SQ_INTERRUPT_MSG)
+		kfd_signal_event_interrupt(pasid, context_id & 0xffffff, 24);
+	else if (source_id == SOC15_INTSRC_CP_BAD_OPCODE)
+		kfd_signal_hw_exception_event(pasid);
+	else if (client_id == SOC15_IH_CLIENTID_VMC ||
+		 client_id == SOC15_IH_CLIENTID_UTCL2) {
+		/* TODO */
+	}
+}
+
+const struct kfd_event_interrupt_class event_interrupt_class_v9 = {
+	.interrupt_isr = event_interrupt_isr_v9,
+	.interrupt_wq = event_interrupt_wq_v9,
+};
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c
index 035c351f47c5..db6d9336b80d 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_interrupt.c
@@ -139,10 +139,12 @@ static void interrupt_wq(struct work_struct *work)
 {
 	struct kfd_dev *dev = container_of(work, struct kfd_dev,
 						interrupt_work);
+	uint32_t ih_ring_entry[KFD_MAX_RING_ENTRY_SIZE];
 
-	uint32_t ih_ring_entry[DIV_ROUND_UP(
-				dev->device_info->ih_ring_entry_size,
-				sizeof(uint32_t))];
+	if (dev->device_info->ih_ring_entry_size > sizeof(ih_ring_entry)) {
+		dev_err_once(kfd_chardev(), "Ring entry too small\n");
+		return;
+	}
 
 	while (dequeue_ih_ring_entry(dev, ih_ring_entry))
 		dev->device_info->event_interrupt_class->interrupt_wq(dev,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
index 69f496485331..476951d8c91c 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c
@@ -99,7 +99,7 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev,
 	kq->rptr_kernel = kq->rptr_mem->cpu_ptr;
 	kq->rptr_gpu_addr = kq->rptr_mem->gpu_addr;
 
-	retval = kfd_gtt_sa_allocate(dev, sizeof(*kq->wptr_kernel),
+	retval = kfd_gtt_sa_allocate(dev, dev->device_info->doorbell_size,
 					&kq->wptr_mem);
 
 	if (retval != 0)
@@ -208,6 +208,7 @@ static int acquire_packet_buffer(struct kernel_queue *kq,
 	size_t available_size;
 	size_t queue_size_dwords;
 	uint32_t wptr, rptr;
+	uint64_t wptr64;
 	unsigned int *queue_address;
 
 	/* When rptr == wptr, the buffer is empty.
@@ -216,7 +217,8 @@ static int acquire_packet_buffer(struct kernel_queue *kq,
 	 * the opposite. So we can only use up to queue_size_dwords - 1 dwords.
 	 */
 	rptr = *kq->rptr_kernel;
-	wptr = *kq->wptr_kernel;
+	wptr = kq->pending_wptr;
+	wptr64 = kq->pending_wptr64;
 	queue_address = (unsigned int *)kq->pq_kernel_addr;
 	queue_size_dwords = kq->queue->properties.queue_size / 4;
 
@@ -232,29 +234,33 @@ static int acquire_packet_buffer(struct kernel_queue *kq,
 		 * make sure calling functions know
 		 * acquire_packet_buffer() failed
 		 */
-		*buffer_ptr = NULL;
-		return -ENOMEM;
+		goto err_no_space;
 	}
 
 	if (wptr + packet_size_in_dwords >= queue_size_dwords) {
 		/* make sure after rolling back to position 0, there is
 		 * still enough space.
 		 */
-		if (packet_size_in_dwords >= rptr) {
-			*buffer_ptr = NULL;
-			return -ENOMEM;
-		}
+		if (packet_size_in_dwords >= rptr)
+			goto err_no_space;
+
 		/* fill nops, roll back and start at position 0 */
 		while (wptr > 0) {
 			queue_address[wptr] = kq->nop_packet;
 			wptr = (wptr + 1) % queue_size_dwords;
+			wptr64++;
 		}
 	}
 
 	*buffer_ptr = &queue_address[wptr];
 	kq->pending_wptr = wptr + packet_size_in_dwords;
+	kq->pending_wptr64 = wptr64 + packet_size_in_dwords;
 
 	return 0;
+
+err_no_space:
+	*buffer_ptr = NULL;
+	return -ENOMEM;
 }
 
 static void submit_packet(struct kernel_queue *kq)
@@ -270,14 +276,18 @@ static void submit_packet(struct kernel_queue *kq)
 	pr_debug("\n");
 #endif
 
-	*kq->wptr_kernel = kq->pending_wptr;
-	write_kernel_doorbell(kq->queue->properties.doorbell_ptr,
-				kq->pending_wptr);
+	kq->ops_asic_specific.submit_packet(kq);
 }
 
 static void rollback_packet(struct kernel_queue *kq)
 {
-	kq->pending_wptr = *kq->queue->properties.write_ptr;
+	if (kq->dev->device_info->doorbell_size == 8) {
+		kq->pending_wptr64 = *kq->wptr64_kernel;
+		kq->pending_wptr = *kq->wptr_kernel %
+			(kq->queue->properties.queue_size / 4);
+	} else {
+		kq->pending_wptr = *kq->wptr_kernel;
+	}
 }
 
 struct kernel_queue *kernel_queue_init(struct kfd_dev *dev,
@@ -308,6 +318,11 @@ struct kernel_queue *kernel_queue_init(struct kfd_dev *dev,
 	case CHIP_HAWAII:
 		kernel_queue_init_cik(&kq->ops_asic_specific);
 		break;
+
+	case CHIP_VEGA10:
+	case CHIP_RAVEN:
+		kernel_queue_init_v9(&kq->ops_asic_specific);
+		break;
 	default:
 		WARN(1, "Unexpected ASIC family %u",
 		     dev->device_info->asic_family);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h
index 594053136ee4..97aff2041a5d 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.h
@@ -72,6 +72,7 @@ struct kernel_queue {
 	struct kfd_dev		*dev;
 	struct mqd_manager	*mqd;
 	struct queue		*queue;
+	uint64_t		pending_wptr64;
 	uint32_t		pending_wptr;
 	unsigned int		nop_packet;
 
@@ -79,7 +80,10 @@ struct kernel_queue {
 	uint32_t		*rptr_kernel;
 	uint64_t		rptr_gpu_addr;
 	struct kfd_mem_obj	*wptr_mem;
-	uint32_t		*wptr_kernel;
+	union {
+		uint64_t	*wptr64_kernel;
+		uint32_t	*wptr_kernel;
+	};
 	uint64_t		wptr_gpu_addr;
 	struct kfd_mem_obj	*pq;
 	uint64_t		pq_gpu_addr;
@@ -97,5 +101,6 @@ struct kernel_queue {
 
 void kernel_queue_init_cik(struct kernel_queue_ops *ops);
 void kernel_queue_init_vi(struct kernel_queue_ops *ops);
+void kernel_queue_init_v9(struct kernel_queue_ops *ops);
 
 #endif /* KFD_KERNEL_QUEUE_H_ */
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_cik.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_cik.c
index a90eb440b1fb..19e54acb4125 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_cik.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_cik.c
@@ -26,11 +26,13 @@
 static bool initialize_cik(struct kernel_queue *kq, struct kfd_dev *dev,
 			enum kfd_queue_type type, unsigned int queue_size);
 static void uninitialize_cik(struct kernel_queue *kq);
+static void submit_packet_cik(struct kernel_queue *kq);
 
 void kernel_queue_init_cik(struct kernel_queue_ops *ops)
 {
 	ops->initialize = initialize_cik;
 	ops->uninitialize = uninitialize_cik;
+	ops->submit_packet = submit_packet_cik;
 }
 
 static bool initialize_cik(struct kernel_queue *kq, struct kfd_dev *dev,
@@ -42,3 +44,10 @@ static bool initialize_cik(struct kernel_queue *kq, struct kfd_dev *dev,
 static void uninitialize_cik(struct kernel_queue *kq)
 {
 }
+
+static void submit_packet_cik(struct kernel_queue *kq)
+{
+	*kq->wptr_kernel = kq->pending_wptr;
+	write_kernel_doorbell(kq->queue->properties.doorbell_ptr,
+				kq->pending_wptr);
+}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c
new file mode 100644
index 000000000000..684a3bf07efd
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c
@@ -0,0 +1,340 @@
+/*
+ * Copyright 2016-2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "kfd_kernel_queue.h"
+#include "kfd_device_queue_manager.h"
+#include "kfd_pm4_headers_ai.h"
+#include "kfd_pm4_opcodes.h"
+
+static bool initialize_v9(struct kernel_queue *kq, struct kfd_dev *dev,
+			enum kfd_queue_type type, unsigned int queue_size);
+static void uninitialize_v9(struct kernel_queue *kq);
+static void submit_packet_v9(struct kernel_queue *kq);
+
+void kernel_queue_init_v9(struct kernel_queue_ops *ops)
+{
+	ops->initialize = initialize_v9;
+	ops->uninitialize = uninitialize_v9;
+	ops->submit_packet = submit_packet_v9;
+}
+
+static bool initialize_v9(struct kernel_queue *kq, struct kfd_dev *dev,
+			enum kfd_queue_type type, unsigned int queue_size)
+{
+	int retval;
+
+	retval = kfd_gtt_sa_allocate(dev, PAGE_SIZE, &kq->eop_mem);
+	if (retval)
+		return false;
+
+	kq->eop_gpu_addr = kq->eop_mem->gpu_addr;
+	kq->eop_kernel_addr = kq->eop_mem->cpu_ptr;
+
+	memset(kq->eop_kernel_addr, 0, PAGE_SIZE);
+
+	return true;
+}
+
+static void uninitialize_v9(struct kernel_queue *kq)
+{
+	kfd_gtt_sa_free(kq->dev, kq->eop_mem);
+}
+
+static void submit_packet_v9(struct kernel_queue *kq)
+{
+	*kq->wptr64_kernel = kq->pending_wptr64;
+	write_kernel_doorbell64(kq->queue->properties.doorbell_ptr,
+				kq->pending_wptr64);
+}
+
+static int pm_map_process_v9(struct packet_manager *pm,
+		uint32_t *buffer, struct qcm_process_device *qpd)
+{
+	struct pm4_mes_map_process *packet;
+	uint64_t vm_page_table_base_addr =
+		(uint64_t)(qpd->page_table_base) << 12;
+
+	packet = (struct pm4_mes_map_process *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_map_process));
+
+	packet->header.u32All = pm_build_pm4_header(IT_MAP_PROCESS,
+					sizeof(struct pm4_mes_map_process));
+	packet->bitfields2.diq_enable = (qpd->is_debug) ? 1 : 0;
+	packet->bitfields2.process_quantum = 1;
+	packet->bitfields2.pasid = qpd->pqm->process->pasid;
+	packet->bitfields14.gds_size = qpd->gds_size;
+	packet->bitfields14.num_gws = qpd->num_gws;
+	packet->bitfields14.num_oac = qpd->num_oac;
+	packet->bitfields14.sdma_enable = 1;
+	packet->bitfields14.num_queues = (qpd->is_debug) ? 0 : qpd->queue_count;
+
+	packet->sh_mem_config = qpd->sh_mem_config;
+	packet->sh_mem_bases = qpd->sh_mem_bases;
+	packet->sq_shader_tba_lo = lower_32_bits(qpd->tba_addr >> 8);
+	packet->sq_shader_tba_hi = upper_32_bits(qpd->tba_addr >> 8);
+	packet->sq_shader_tma_lo = lower_32_bits(qpd->tma_addr >> 8);
+	packet->sq_shader_tma_hi = upper_32_bits(qpd->tma_addr >> 8);
+
+	packet->gds_addr_lo = lower_32_bits(qpd->gds_context_area);
+	packet->gds_addr_hi = upper_32_bits(qpd->gds_context_area);
+
+	packet->vm_context_page_table_base_addr_lo32 =
+			lower_32_bits(vm_page_table_base_addr);
+	packet->vm_context_page_table_base_addr_hi32 =
+			upper_32_bits(vm_page_table_base_addr);
+
+	return 0;
+}
+
+static int pm_runlist_v9(struct packet_manager *pm, uint32_t *buffer,
+			uint64_t ib, size_t ib_size_in_dwords, bool chain)
+{
+	struct pm4_mes_runlist *packet;
+
+	int concurrent_proc_cnt = 0;
+	struct kfd_dev *kfd = pm->dqm->dev;
+
+	/* Determine the number of processes to map together to HW:
+	 * it can not exceed the number of VMIDs available to the
+	 * scheduler, and it is determined by the smaller of the number
+	 * of processes in the runlist and kfd module parameter
+	 * hws_max_conc_proc.
+	 * Note: the arbitration between the number of VMIDs and
+	 * hws_max_conc_proc has been done in
+	 * kgd2kfd_device_init().
+	 */
+	concurrent_proc_cnt = min(pm->dqm->processes_count,
+			kfd->max_proc_per_quantum);
+
+	packet = (struct pm4_mes_runlist *)buffer;
+
+	memset(buffer, 0, sizeof(struct pm4_mes_runlist));
+	packet->header.u32All = pm_build_pm4_header(IT_RUN_LIST,
+						sizeof(struct pm4_mes_runlist));
+
+	packet->bitfields4.ib_size = ib_size_in_dwords;
+	packet->bitfields4.chain = chain ? 1 : 0;
+	packet->bitfields4.offload_polling = 0;
+	packet->bitfields4.valid = 1;
+	packet->bitfields4.process_cnt = concurrent_proc_cnt;
+	packet->ordinal2 = lower_32_bits(ib);
+	packet->ib_base_hi = upper_32_bits(ib);
+
+	return 0;
+}
+
+static int pm_map_queues_v9(struct packet_manager *pm, uint32_t *buffer,
+		struct queue *q, bool is_static)
+{
+	struct pm4_mes_map_queues *packet;
+	bool use_static = is_static;
+
+	packet = (struct pm4_mes_map_queues *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_map_queues));
+
+	packet->header.u32All = pm_build_pm4_header(IT_MAP_QUEUES,
+					sizeof(struct pm4_mes_map_queues));
+	packet->bitfields2.alloc_format =
+		alloc_format__mes_map_queues__one_per_pipe_vi;
+	packet->bitfields2.num_queues = 1;
+	packet->bitfields2.queue_sel =
+		queue_sel__mes_map_queues__map_to_hws_determined_queue_slots_vi;
+
+	packet->bitfields2.engine_sel =
+		engine_sel__mes_map_queues__compute_vi;
+	packet->bitfields2.queue_type =
+		queue_type__mes_map_queues__normal_compute_vi;
+
+	switch (q->properties.type) {
+	case KFD_QUEUE_TYPE_COMPUTE:
+		if (use_static)
+			packet->bitfields2.queue_type =
+		queue_type__mes_map_queues__normal_latency_static_queue_vi;
+		break;
+	case KFD_QUEUE_TYPE_DIQ:
+		packet->bitfields2.queue_type =
+			queue_type__mes_map_queues__debug_interface_queue_vi;
+		break;
+	case KFD_QUEUE_TYPE_SDMA:
+		packet->bitfields2.engine_sel = q->properties.sdma_engine_id +
+				engine_sel__mes_map_queues__sdma0_vi;
+		use_static = false; /* no static queues under SDMA */
+		break;
+	default:
+		WARN(1, "queue type %d", q->properties.type);
+		return -EINVAL;
+	}
+	packet->bitfields3.doorbell_offset =
+			q->properties.doorbell_off;
+
+	packet->mqd_addr_lo =
+			lower_32_bits(q->gart_mqd_addr);
+
+	packet->mqd_addr_hi =
+			upper_32_bits(q->gart_mqd_addr);
+
+	packet->wptr_addr_lo =
+			lower_32_bits((uint64_t)q->properties.write_ptr);
+
+	packet->wptr_addr_hi =
+			upper_32_bits((uint64_t)q->properties.write_ptr);
+
+	return 0;
+}
+
+static int pm_unmap_queues_v9(struct packet_manager *pm, uint32_t *buffer,
+			enum kfd_queue_type type,
+			enum kfd_unmap_queues_filter filter,
+			uint32_t filter_param, bool reset,
+			unsigned int sdma_engine)
+{
+	struct pm4_mes_unmap_queues *packet;
+
+	packet = (struct pm4_mes_unmap_queues *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_unmap_queues));
+
+	packet->header.u32All = pm_build_pm4_header(IT_UNMAP_QUEUES,
+					sizeof(struct pm4_mes_unmap_queues));
+	switch (type) {
+	case KFD_QUEUE_TYPE_COMPUTE:
+	case KFD_QUEUE_TYPE_DIQ:
+		packet->bitfields2.engine_sel =
+			engine_sel__mes_unmap_queues__compute;
+		break;
+	case KFD_QUEUE_TYPE_SDMA:
+		packet->bitfields2.engine_sel =
+			engine_sel__mes_unmap_queues__sdma0 + sdma_engine;
+		break;
+	default:
+		WARN(1, "queue type %d", type);
+		return -EINVAL;
+	}
+
+	if (reset)
+		packet->bitfields2.action =
+			action__mes_unmap_queues__reset_queues;
+	else
+		packet->bitfields2.action =
+			action__mes_unmap_queues__preempt_queues;
+
+	switch (filter) {
+	case KFD_UNMAP_QUEUES_FILTER_SINGLE_QUEUE:
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__perform_request_on_specified_queues;
+		packet->bitfields2.num_queues = 1;
+		packet->bitfields3b.doorbell_offset0 = filter_param;
+		break;
+	case KFD_UNMAP_QUEUES_FILTER_BY_PASID:
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__perform_request_on_pasid_queues;
+		packet->bitfields3a.pasid = filter_param;
+		break;
+	case KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES:
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__unmap_all_queues;
+		break;
+	case KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES:
+		/* in this case, we do not preempt static queues */
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__unmap_all_non_static_queues;
+		break;
+	default:
+		WARN(1, "filter %d", filter);
+		return -EINVAL;
+	}
+
+	return 0;
+
+}
+
+static int pm_query_status_v9(struct packet_manager *pm, uint32_t *buffer,
+			uint64_t fence_address,	uint32_t fence_value)
+{
+	struct pm4_mes_query_status *packet;
+
+	packet = (struct pm4_mes_query_status *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_query_status));
+
+
+	packet->header.u32All = pm_build_pm4_header(IT_QUERY_STATUS,
+					sizeof(struct pm4_mes_query_status));
+
+	packet->bitfields2.context_id = 0;
+	packet->bitfields2.interrupt_sel =
+			interrupt_sel__mes_query_status__completion_status;
+	packet->bitfields2.command =
+			command__mes_query_status__fence_only_after_write_ack;
+
+	packet->addr_hi = upper_32_bits((uint64_t)fence_address);
+	packet->addr_lo = lower_32_bits((uint64_t)fence_address);
+	packet->data_hi = upper_32_bits((uint64_t)fence_value);
+	packet->data_lo = lower_32_bits((uint64_t)fence_value);
+
+	return 0;
+}
+
+
+static int pm_release_mem_v9(uint64_t gpu_addr, uint32_t *buffer)
+{
+	struct pm4_mec_release_mem *packet;
+
+	packet = (struct pm4_mec_release_mem *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mec_release_mem));
+
+	packet->header.u32All = pm_build_pm4_header(IT_RELEASE_MEM,
+					sizeof(struct pm4_mec_release_mem));
+
+	packet->bitfields2.event_type = CACHE_FLUSH_AND_INV_TS_EVENT;
+	packet->bitfields2.event_index = event_index__mec_release_mem__end_of_pipe;
+	packet->bitfields2.tcl1_action_ena = 1;
+	packet->bitfields2.tc_action_ena = 1;
+	packet->bitfields2.cache_policy = cache_policy__mec_release_mem__lru;
+
+	packet->bitfields3.data_sel = data_sel__mec_release_mem__send_32_bit_low;
+	packet->bitfields3.int_sel =
+		int_sel__mec_release_mem__send_interrupt_after_write_confirm;
+
+	packet->bitfields4.address_lo_32b = (gpu_addr & 0xffffffff) >> 2;
+	packet->address_hi = upper_32_bits(gpu_addr);
+
+	packet->data_lo = 0;
+
+	return 0;
+}
+
+const struct packet_manager_funcs kfd_v9_pm_funcs = {
+	.map_process		= pm_map_process_v9,
+	.runlist		= pm_runlist_v9,
+	.set_resources		= pm_set_resources_vi,
+	.map_queues		= pm_map_queues_v9,
+	.unmap_queues		= pm_unmap_queues_v9,
+	.query_status		= pm_query_status_v9,
+	.release_mem		= pm_release_mem_v9,
+	.map_process_size	= sizeof(struct pm4_mes_map_process),
+	.runlist_size		= sizeof(struct pm4_mes_runlist),
+	.set_resources_size	= sizeof(struct pm4_mes_set_resources),
+	.map_queues_size	= sizeof(struct pm4_mes_map_queues),
+	.unmap_queues_size	= sizeof(struct pm4_mes_unmap_queues),
+	.query_status_size	= sizeof(struct pm4_mes_query_status),
+	.release_mem_size	= sizeof(struct pm4_mec_release_mem)
+};
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c
index f1d48281e322..bf20c6d32ef3 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_vi.c
@@ -22,15 +22,20 @@
  */
 
 #include "kfd_kernel_queue.h"
+#include "kfd_device_queue_manager.h"
+#include "kfd_pm4_headers_vi.h"
+#include "kfd_pm4_opcodes.h"
 
 static bool initialize_vi(struct kernel_queue *kq, struct kfd_dev *dev,
 			enum kfd_queue_type type, unsigned int queue_size);
 static void uninitialize_vi(struct kernel_queue *kq);
+static void submit_packet_vi(struct kernel_queue *kq);
 
 void kernel_queue_init_vi(struct kernel_queue_ops *ops)
 {
 	ops->initialize = initialize_vi;
 	ops->uninitialize = uninitialize_vi;
+	ops->submit_packet = submit_packet_vi;
 }
 
 static bool initialize_vi(struct kernel_queue *kq, struct kfd_dev *dev,
@@ -54,3 +59,317 @@ static void uninitialize_vi(struct kernel_queue *kq)
 {
 	kfd_gtt_sa_free(kq->dev, kq->eop_mem);
 }
+
+static void submit_packet_vi(struct kernel_queue *kq)
+{
+	*kq->wptr_kernel = kq->pending_wptr;
+	write_kernel_doorbell(kq->queue->properties.doorbell_ptr,
+				kq->pending_wptr);
+}
+
+unsigned int pm_build_pm4_header(unsigned int opcode, size_t packet_size)
+{
+	union PM4_MES_TYPE_3_HEADER header;
+
+	header.u32All = 0;
+	header.opcode = opcode;
+	header.count = packet_size / 4 - 2;
+	header.type = PM4_TYPE_3;
+
+	return header.u32All;
+}
+
+static int pm_map_process_vi(struct packet_manager *pm, uint32_t *buffer,
+				struct qcm_process_device *qpd)
+{
+	struct pm4_mes_map_process *packet;
+
+	packet = (struct pm4_mes_map_process *)buffer;
+
+	memset(buffer, 0, sizeof(struct pm4_mes_map_process));
+
+	packet->header.u32All = pm_build_pm4_header(IT_MAP_PROCESS,
+					sizeof(struct pm4_mes_map_process));
+	packet->bitfields2.diq_enable = (qpd->is_debug) ? 1 : 0;
+	packet->bitfields2.process_quantum = 1;
+	packet->bitfields2.pasid = qpd->pqm->process->pasid;
+	packet->bitfields3.page_table_base = qpd->page_table_base;
+	packet->bitfields10.gds_size = qpd->gds_size;
+	packet->bitfields10.num_gws = qpd->num_gws;
+	packet->bitfields10.num_oac = qpd->num_oac;
+	packet->bitfields10.num_queues = (qpd->is_debug) ? 0 : qpd->queue_count;
+
+	packet->sh_mem_config = qpd->sh_mem_config;
+	packet->sh_mem_bases = qpd->sh_mem_bases;
+	packet->sh_mem_ape1_base = qpd->sh_mem_ape1_base;
+	packet->sh_mem_ape1_limit = qpd->sh_mem_ape1_limit;
+
+	packet->sh_hidden_private_base_vmid = qpd->sh_hidden_private_base;
+
+	packet->gds_addr_lo = lower_32_bits(qpd->gds_context_area);
+	packet->gds_addr_hi = upper_32_bits(qpd->gds_context_area);
+
+	return 0;
+}
+
+static int pm_runlist_vi(struct packet_manager *pm, uint32_t *buffer,
+			uint64_t ib, size_t ib_size_in_dwords, bool chain)
+{
+	struct pm4_mes_runlist *packet;
+	int concurrent_proc_cnt = 0;
+	struct kfd_dev *kfd = pm->dqm->dev;
+
+	if (WARN_ON(!ib))
+		return -EFAULT;
+
+	/* Determine the number of processes to map together to HW:
+	 * it can not exceed the number of VMIDs available to the
+	 * scheduler, and it is determined by the smaller of the number
+	 * of processes in the runlist and kfd module parameter
+	 * hws_max_conc_proc.
+	 * Note: the arbitration between the number of VMIDs and
+	 * hws_max_conc_proc has been done in
+	 * kgd2kfd_device_init().
+	 */
+	concurrent_proc_cnt = min(pm->dqm->processes_count,
+			kfd->max_proc_per_quantum);
+
+	packet = (struct pm4_mes_runlist *)buffer;
+
+	memset(buffer, 0, sizeof(struct pm4_mes_runlist));
+	packet->header.u32All = pm_build_pm4_header(IT_RUN_LIST,
+						sizeof(struct pm4_mes_runlist));
+
+	packet->bitfields4.ib_size = ib_size_in_dwords;
+	packet->bitfields4.chain = chain ? 1 : 0;
+	packet->bitfields4.offload_polling = 0;
+	packet->bitfields4.valid = 1;
+	packet->bitfields4.process_cnt = concurrent_proc_cnt;
+	packet->ordinal2 = lower_32_bits(ib);
+	packet->bitfields3.ib_base_hi = upper_32_bits(ib);
+
+	return 0;
+}
+
+int pm_set_resources_vi(struct packet_manager *pm, uint32_t *buffer,
+				struct scheduling_resources *res)
+{
+	struct pm4_mes_set_resources *packet;
+
+	packet = (struct pm4_mes_set_resources *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_set_resources));
+
+	packet->header.u32All = pm_build_pm4_header(IT_SET_RESOURCES,
+					sizeof(struct pm4_mes_set_resources));
+
+	packet->bitfields2.queue_type =
+			queue_type__mes_set_resources__hsa_interface_queue_hiq;
+	packet->bitfields2.vmid_mask = res->vmid_mask;
+	packet->bitfields2.unmap_latency = KFD_UNMAP_LATENCY_MS / 100;
+	packet->bitfields7.oac_mask = res->oac_mask;
+	packet->bitfields8.gds_heap_base = res->gds_heap_base;
+	packet->bitfields8.gds_heap_size = res->gds_heap_size;
+
+	packet->gws_mask_lo = lower_32_bits(res->gws_mask);
+	packet->gws_mask_hi = upper_32_bits(res->gws_mask);
+
+	packet->queue_mask_lo = lower_32_bits(res->queue_mask);
+	packet->queue_mask_hi = upper_32_bits(res->queue_mask);
+
+	return 0;
+}
+
+static int pm_map_queues_vi(struct packet_manager *pm, uint32_t *buffer,
+		struct queue *q, bool is_static)
+{
+	struct pm4_mes_map_queues *packet;
+	bool use_static = is_static;
+
+	packet = (struct pm4_mes_map_queues *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_map_queues));
+
+	packet->header.u32All = pm_build_pm4_header(IT_MAP_QUEUES,
+					sizeof(struct pm4_mes_map_queues));
+	packet->bitfields2.alloc_format =
+		alloc_format__mes_map_queues__one_per_pipe_vi;
+	packet->bitfields2.num_queues = 1;
+	packet->bitfields2.queue_sel =
+		queue_sel__mes_map_queues__map_to_hws_determined_queue_slots_vi;
+
+	packet->bitfields2.engine_sel =
+		engine_sel__mes_map_queues__compute_vi;
+	packet->bitfields2.queue_type =
+		queue_type__mes_map_queues__normal_compute_vi;
+
+	switch (q->properties.type) {
+	case KFD_QUEUE_TYPE_COMPUTE:
+		if (use_static)
+			packet->bitfields2.queue_type =
+		queue_type__mes_map_queues__normal_latency_static_queue_vi;
+		break;
+	case KFD_QUEUE_TYPE_DIQ:
+		packet->bitfields2.queue_type =
+			queue_type__mes_map_queues__debug_interface_queue_vi;
+		break;
+	case KFD_QUEUE_TYPE_SDMA:
+		packet->bitfields2.engine_sel = q->properties.sdma_engine_id +
+				engine_sel__mes_map_queues__sdma0_vi;
+		use_static = false; /* no static queues under SDMA */
+		break;
+	default:
+		WARN(1, "queue type %d", q->properties.type);
+		return -EINVAL;
+	}
+	packet->bitfields3.doorbell_offset =
+			q->properties.doorbell_off;
+
+	packet->mqd_addr_lo =
+			lower_32_bits(q->gart_mqd_addr);
+
+	packet->mqd_addr_hi =
+			upper_32_bits(q->gart_mqd_addr);
+
+	packet->wptr_addr_lo =
+			lower_32_bits((uint64_t)q->properties.write_ptr);
+
+	packet->wptr_addr_hi =
+			upper_32_bits((uint64_t)q->properties.write_ptr);
+
+	return 0;
+}
+
+static int pm_unmap_queues_vi(struct packet_manager *pm, uint32_t *buffer,
+			enum kfd_queue_type type,
+			enum kfd_unmap_queues_filter filter,
+			uint32_t filter_param, bool reset,
+			unsigned int sdma_engine)
+{
+	struct pm4_mes_unmap_queues *packet;
+
+	packet = (struct pm4_mes_unmap_queues *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_unmap_queues));
+
+	packet->header.u32All = pm_build_pm4_header(IT_UNMAP_QUEUES,
+					sizeof(struct pm4_mes_unmap_queues));
+	switch (type) {
+	case KFD_QUEUE_TYPE_COMPUTE:
+	case KFD_QUEUE_TYPE_DIQ:
+		packet->bitfields2.engine_sel =
+			engine_sel__mes_unmap_queues__compute;
+		break;
+	case KFD_QUEUE_TYPE_SDMA:
+		packet->bitfields2.engine_sel =
+			engine_sel__mes_unmap_queues__sdma0 + sdma_engine;
+		break;
+	default:
+		WARN(1, "queue type %d", type);
+		return -EINVAL;
+	}
+
+	if (reset)
+		packet->bitfields2.action =
+			action__mes_unmap_queues__reset_queues;
+	else
+		packet->bitfields2.action =
+			action__mes_unmap_queues__preempt_queues;
+
+	switch (filter) {
+	case KFD_UNMAP_QUEUES_FILTER_SINGLE_QUEUE:
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__perform_request_on_specified_queues;
+		packet->bitfields2.num_queues = 1;
+		packet->bitfields3b.doorbell_offset0 = filter_param;
+		break;
+	case KFD_UNMAP_QUEUES_FILTER_BY_PASID:
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__perform_request_on_pasid_queues;
+		packet->bitfields3a.pasid = filter_param;
+		break;
+	case KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES:
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__unmap_all_queues;
+		break;
+	case KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES:
+		/* in this case, we do not preempt static queues */
+		packet->bitfields2.queue_sel =
+			queue_sel__mes_unmap_queues__unmap_all_non_static_queues;
+		break;
+	default:
+		WARN(1, "filter %d", filter);
+		return -EINVAL;
+	}
+
+	return 0;
+
+}
+
+static int pm_query_status_vi(struct packet_manager *pm, uint32_t *buffer,
+			uint64_t fence_address,	uint32_t fence_value)
+{
+	struct pm4_mes_query_status *packet;
+
+	packet = (struct pm4_mes_query_status *)buffer;
+	memset(buffer, 0, sizeof(struct pm4_mes_query_status));
+
+	packet->header.u32All = pm_build_pm4_header(IT_QUERY_STATUS,
+					sizeof(struct pm4_mes_query_status));
+
+	packet->bitfields2.context_id = 0;
+	packet->bitfields2.interrupt_sel =
+			interrupt_sel__mes_query_status__completion_status;
+	packet->bitfields2.command =
+			command__mes_query_status__fence_only_after_write_ack;
+
+	packet->addr_hi = upper_32_bits((uint64_t)fence_address);
+	packet->addr_lo = lower_32_bits((uint64_t)fence_address);
+	packet->data_hi = upper_32_bits((uint64_t)fence_value);
+	packet->data_lo = lower_32_bits((uint64_t)fence_value);
+
+	return 0;
+}
+
+static int pm_release_mem_vi(uint64_t gpu_addr, uint32_t *buffer)
+{
+	struct pm4_mec_release_mem *packet;
+
+	packet = (struct pm4_mec_release_mem *)buffer;
+	memset(buffer, 0, sizeof(*packet));
+
+	packet->header.u32All = pm_build_pm4_header(IT_RELEASE_MEM,
+						 sizeof(*packet));
+
+	packet->bitfields2.event_type = CACHE_FLUSH_AND_INV_TS_EVENT;
+	packet->bitfields2.event_index = event_index___release_mem__end_of_pipe;
+	packet->bitfields2.tcl1_action_ena = 1;
+	packet->bitfields2.tc_action_ena = 1;
+	packet->bitfields2.cache_policy = cache_policy___release_mem__lru;
+	packet->bitfields2.atc = 0;
+
+	packet->bitfields3.data_sel = data_sel___release_mem__send_32_bit_low;
+	packet->bitfields3.int_sel =
+		int_sel___release_mem__send_interrupt_after_write_confirm;
+
+	packet->bitfields4.address_lo_32b = (gpu_addr & 0xffffffff) >> 2;
+	packet->address_hi = upper_32_bits(gpu_addr);
+
+	packet->data_lo = 0;
+
+	return 0;
+}
+
+const struct packet_manager_funcs kfd_vi_pm_funcs = {
+	.map_process		= pm_map_process_vi,
+	.runlist		= pm_runlist_vi,
+	.set_resources		= pm_set_resources_vi,
+	.map_queues		= pm_map_queues_vi,
+	.unmap_queues		= pm_unmap_queues_vi,
+	.query_status		= pm_query_status_vi,
+	.release_mem		= pm_release_mem_vi,
+	.map_process_size	= sizeof(struct pm4_mes_map_process),
+	.runlist_size		= sizeof(struct pm4_mes_runlist),
+	.set_resources_size	= sizeof(struct pm4_mes_set_resources),
+	.map_queues_size	= sizeof(struct pm4_mes_map_queues),
+	.unmap_queues_size	= sizeof(struct pm4_mes_unmap_queues),
+	.query_status_size	= sizeof(struct pm4_mes_query_status),
+	.release_mem_size	= sizeof(struct pm4_mec_release_mem)
+};
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_module.c b/drivers/gpu/drm/amd/amdkfd/kfd_module.c
index e0c07d24d251..76bf2dc8aec4 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_module.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_module.c
@@ -43,6 +43,8 @@ static const struct kgd2kfd_calls kgd2kfd = {
 	.interrupt	= kgd2kfd_interrupt,
 	.suspend	= kgd2kfd_suspend,
 	.resume		= kgd2kfd_resume,
+	.quiesce_mm	= kgd2kfd_quiesce_mm,
+	.resume_mm	= kgd2kfd_resume_mm,
 	.schedule_evict_and_restore_process =
 			  kgd2kfd_schedule_evict_and_restore_process,
 };
@@ -81,6 +83,11 @@ module_param(ignore_crat, int, 0444);
 MODULE_PARM_DESC(ignore_crat,
 	"Ignore CRAT table during KFD initialization (0 = use CRAT (default), 1 = ignore CRAT)");
 
+int vega10_noretry;
+module_param_named(noretry, vega10_noretry, int, 0644);
+MODULE_PARM_DESC(noretry,
+	"Set sh_mem_config.retry_disable on Vega10 (0 = retry enabled (default), 1 = retry disabled)");
+
 static int amdkfd_init_completed;
 
 int kgd2kfd_init(unsigned int interface_version,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c
index ee7061e1c466..4b8eb506642b 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager.c
@@ -38,6 +38,9 @@ struct mqd_manager *mqd_manager_init(enum KFD_MQD_TYPE type,
 	case CHIP_POLARIS10:
 	case CHIP_POLARIS11:
 		return mqd_manager_init_vi_tonga(type, dev);
+	case CHIP_VEGA10:
+	case CHIP_RAVEN:
+		return mqd_manager_init_v9(type, dev);
 	default:
 		WARN(1, "Unexpected ASIC family %u",
 		     dev->device_info->asic_family);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c
index c00c325ed3c9..06eaa218eba6 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_cik.c
@@ -79,10 +79,6 @@ static int init_mqd(struct mqd_manager *mm, void **mqd,
 	m->cp_mqd_base_addr_lo        = lower_32_bits(addr);
 	m->cp_mqd_base_addr_hi        = upper_32_bits(addr);
 
-	m->cp_hqd_ib_control = DEFAULT_MIN_IB_AVAIL_SIZE | IB_ATC_EN;
-	/* Although WinKFD writes this, I suspect it should not be necessary */
-	m->cp_hqd_ib_control = IB_ATC_EN | DEFAULT_MIN_IB_AVAIL_SIZE;
-
 	m->cp_hqd_quantum = QUANTUM_EN | QUANTUM_SCALE_1MS |
 				QUANTUM_DURATION(10);
 
@@ -412,7 +408,7 @@ struct mqd_manager *mqd_manager_init_cik(enum KFD_MQD_TYPE type,
 	if (WARN_ON(type >= KFD_MQD_TYPE_MAX))
 		return NULL;
 
-	mqd = kzalloc(sizeof(*mqd), GFP_KERNEL);
+	mqd = kzalloc(sizeof(*mqd), GFP_NOIO);
 	if (!mqd)
 		return NULL;
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c
new file mode 100644
index 000000000000..684054ff02cd
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_v9.c
@@ -0,0 +1,443 @@
+/*
+ * Copyright 2016-2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include <linux/printk.h>
+#include <linux/slab.h>
+#include <linux/uaccess.h>
+#include "kfd_priv.h"
+#include "kfd_mqd_manager.h"
+#include "v9_structs.h"
+#include "gc/gc_9_0_offset.h"
+#include "gc/gc_9_0_sh_mask.h"
+#include "sdma0/sdma0_4_0_sh_mask.h"
+
+static inline struct v9_mqd *get_mqd(void *mqd)
+{
+	return (struct v9_mqd *)mqd;
+}
+
+static inline struct v9_sdma_mqd *get_sdma_mqd(void *mqd)
+{
+	return (struct v9_sdma_mqd *)mqd;
+}
+
+static int init_mqd(struct mqd_manager *mm, void **mqd,
+			struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr,
+			struct queue_properties *q)
+{
+	int retval;
+	uint64_t addr;
+	struct v9_mqd *m;
+	struct kfd_dev *kfd = mm->dev;
+
+	/* From V9,  for CWSR, the control stack is located on the next page
+	 * boundary after the mqd, we will use the gtt allocation function
+	 * instead of sub-allocation function.
+	 */
+	if (kfd->cwsr_enabled && (q->type == KFD_QUEUE_TYPE_COMPUTE)) {
+		*mqd_mem_obj = kzalloc(sizeof(struct kfd_mem_obj), GFP_NOIO);
+		if (!*mqd_mem_obj)
+			return -ENOMEM;
+		retval = kfd->kfd2kgd->init_gtt_mem_allocation(kfd->kgd,
+			ALIGN(q->ctl_stack_size, PAGE_SIZE) +
+				ALIGN(sizeof(struct v9_mqd), PAGE_SIZE),
+			&((*mqd_mem_obj)->gtt_mem),
+			&((*mqd_mem_obj)->gpu_addr),
+			(void *)&((*mqd_mem_obj)->cpu_ptr));
+	} else
+		retval = kfd_gtt_sa_allocate(mm->dev, sizeof(struct v9_mqd),
+				mqd_mem_obj);
+	if (retval != 0)
+		return -ENOMEM;
+
+	m = (struct v9_mqd *) (*mqd_mem_obj)->cpu_ptr;
+	addr = (*mqd_mem_obj)->gpu_addr;
+
+	memset(m, 0, sizeof(struct v9_mqd));
+
+	m->header = 0xC0310800;
+	m->compute_pipelinestat_enable = 1;
+	m->compute_static_thread_mgmt_se0 = 0xFFFFFFFF;
+	m->compute_static_thread_mgmt_se1 = 0xFFFFFFFF;
+	m->compute_static_thread_mgmt_se2 = 0xFFFFFFFF;
+	m->compute_static_thread_mgmt_se3 = 0xFFFFFFFF;
+
+	m->cp_hqd_persistent_state = CP_HQD_PERSISTENT_STATE__PRELOAD_REQ_MASK |
+			0x53 << CP_HQD_PERSISTENT_STATE__PRELOAD_SIZE__SHIFT;
+
+	m->cp_mqd_control = 1 << CP_MQD_CONTROL__PRIV_STATE__SHIFT;
+
+	m->cp_mqd_base_addr_lo        = lower_32_bits(addr);
+	m->cp_mqd_base_addr_hi        = upper_32_bits(addr);
+
+	m->cp_hqd_quantum = 1 << CP_HQD_QUANTUM__QUANTUM_EN__SHIFT |
+			1 << CP_HQD_QUANTUM__QUANTUM_SCALE__SHIFT |
+			10 << CP_HQD_QUANTUM__QUANTUM_DURATION__SHIFT;
+
+	m->cp_hqd_pipe_priority = 1;
+	m->cp_hqd_queue_priority = 15;
+
+	if (q->format == KFD_QUEUE_FORMAT_AQL) {
+		m->cp_hqd_aql_control =
+			1 << CP_HQD_AQL_CONTROL__CONTROL0__SHIFT;
+	}
+
+	if (q->tba_addr) {
+		m->compute_pgm_rsrc2 |=
+			(1 << COMPUTE_PGM_RSRC2__TRAP_PRESENT__SHIFT);
+	}
+
+	if (mm->dev->cwsr_enabled && q->ctx_save_restore_area_address) {
+		m->cp_hqd_persistent_state |=
+			(1 << CP_HQD_PERSISTENT_STATE__QSWITCH_MODE__SHIFT);
+		m->cp_hqd_ctx_save_base_addr_lo =
+			lower_32_bits(q->ctx_save_restore_area_address);
+		m->cp_hqd_ctx_save_base_addr_hi =
+			upper_32_bits(q->ctx_save_restore_area_address);
+		m->cp_hqd_ctx_save_size = q->ctx_save_restore_area_size;
+		m->cp_hqd_cntl_stack_size = q->ctl_stack_size;
+		m->cp_hqd_cntl_stack_offset = q->ctl_stack_size;
+		m->cp_hqd_wg_state_offset = q->ctl_stack_size;
+	}
+
+	*mqd = m;
+	if (gart_addr)
+		*gart_addr = addr;
+	retval = mm->update_mqd(mm, m, q);
+
+	return retval;
+}
+
+static int load_mqd(struct mqd_manager *mm, void *mqd,
+			uint32_t pipe_id, uint32_t queue_id,
+			struct queue_properties *p, struct mm_struct *mms)
+{
+	/* AQL write pointer counts in 64B packets, PM4/CP counts in dwords. */
+	uint32_t wptr_shift = (p->format == KFD_QUEUE_FORMAT_AQL ? 4 : 0);
+
+	return mm->dev->kfd2kgd->hqd_load(mm->dev->kgd, mqd, pipe_id, queue_id,
+					  (uint32_t __user *)p->write_ptr,
+					  wptr_shift, 0, mms);
+}
+
+static int update_mqd(struct mqd_manager *mm, void *mqd,
+		      struct queue_properties *q)
+{
+	struct v9_mqd *m;
+
+	m = get_mqd(mqd);
+
+	m->cp_hqd_pq_control = 5 << CP_HQD_PQ_CONTROL__RPTR_BLOCK_SIZE__SHIFT;
+	m->cp_hqd_pq_control |= order_base_2(q->queue_size / 4) - 1;
+	pr_debug("cp_hqd_pq_control 0x%x\n", m->cp_hqd_pq_control);
+
+	m->cp_hqd_pq_base_lo = lower_32_bits((uint64_t)q->queue_address >> 8);
+	m->cp_hqd_pq_base_hi = upper_32_bits((uint64_t)q->queue_address >> 8);
+
+	m->cp_hqd_pq_rptr_report_addr_lo = lower_32_bits((uint64_t)q->read_ptr);
+	m->cp_hqd_pq_rptr_report_addr_hi = upper_32_bits((uint64_t)q->read_ptr);
+	m->cp_hqd_pq_wptr_poll_addr_lo = lower_32_bits((uint64_t)q->write_ptr);
+	m->cp_hqd_pq_wptr_poll_addr_hi = upper_32_bits((uint64_t)q->write_ptr);
+
+	m->cp_hqd_pq_doorbell_control =
+		q->doorbell_off <<
+			CP_HQD_PQ_DOORBELL_CONTROL__DOORBELL_OFFSET__SHIFT;
+	pr_debug("cp_hqd_pq_doorbell_control 0x%x\n",
+			m->cp_hqd_pq_doorbell_control);
+
+	m->cp_hqd_ib_control =
+		3 << CP_HQD_IB_CONTROL__MIN_IB_AVAIL_SIZE__SHIFT |
+		1 << CP_HQD_IB_CONTROL__IB_EXE_DISABLE__SHIFT;
+
+	/*
+	 * HW does not clamp this field correctly. Maximum EOP queue size
+	 * is constrained by per-SE EOP done signal count, which is 8-bit.
+	 * Limit is 0xFF EOP entries (= 0x7F8 dwords). CP will not submit
+	 * more than (EOP entry count - 1) so a queue size of 0x800 dwords
+	 * is safe, giving a maximum field value of 0xA.
+	 */
+	m->cp_hqd_eop_control = min(0xA,
+		order_base_2(q->eop_ring_buffer_size / 4) - 1);
+	m->cp_hqd_eop_base_addr_lo =
+			lower_32_bits(q->eop_ring_buffer_address >> 8);
+	m->cp_hqd_eop_base_addr_hi =
+			upper_32_bits(q->eop_ring_buffer_address >> 8);
+
+	m->cp_hqd_iq_timer = 0;
+
+	m->cp_hqd_vmid = q->vmid;
+
+	if (q->format == KFD_QUEUE_FORMAT_AQL) {
+		m->cp_hqd_pq_control |= CP_HQD_PQ_CONTROL__NO_UPDATE_RPTR_MASK |
+				2 << CP_HQD_PQ_CONTROL__SLOT_BASED_WPTR__SHIFT |
+				1 << CP_HQD_PQ_CONTROL__QUEUE_FULL_EN__SHIFT |
+				1 << CP_HQD_PQ_CONTROL__WPP_CLAMP_EN__SHIFT;
+		m->cp_hqd_pq_doorbell_control |= 1 <<
+			CP_HQD_PQ_DOORBELL_CONTROL__DOORBELL_BIF_DROP__SHIFT;
+	}
+	if (mm->dev->cwsr_enabled && q->ctx_save_restore_area_address)
+		m->cp_hqd_ctx_save_control = 0;
+
+	q->is_active = (q->queue_size > 0 &&
+			q->queue_address != 0 &&
+			q->queue_percent > 0 &&
+			!q->is_evicted);
+
+	return 0;
+}
+
+
+static int destroy_mqd(struct mqd_manager *mm, void *mqd,
+			enum kfd_preempt_type type,
+			unsigned int timeout, uint32_t pipe_id,
+			uint32_t queue_id)
+{
+	return mm->dev->kfd2kgd->hqd_destroy
+		(mm->dev->kgd, mqd, type, timeout,
+		pipe_id, queue_id);
+}
+
+static void uninit_mqd(struct mqd_manager *mm, void *mqd,
+			struct kfd_mem_obj *mqd_mem_obj)
+{
+	struct kfd_dev *kfd = mm->dev;
+
+	if (mqd_mem_obj->gtt_mem) {
+		kfd->kfd2kgd->free_gtt_mem(kfd->kgd, mqd_mem_obj->gtt_mem);
+		kfree(mqd_mem_obj);
+	} else {
+		kfd_gtt_sa_free(mm->dev, mqd_mem_obj);
+	}
+}
+
+static bool is_occupied(struct mqd_manager *mm, void *mqd,
+			uint64_t queue_address,	uint32_t pipe_id,
+			uint32_t queue_id)
+{
+	return mm->dev->kfd2kgd->hqd_is_occupied(
+		mm->dev->kgd, queue_address,
+		pipe_id, queue_id);
+}
+
+static int init_mqd_hiq(struct mqd_manager *mm, void **mqd,
+			struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr,
+			struct queue_properties *q)
+{
+	struct v9_mqd *m;
+	int retval = init_mqd(mm, mqd, mqd_mem_obj, gart_addr, q);
+
+	if (retval != 0)
+		return retval;
+
+	m = get_mqd(*mqd);
+
+	m->cp_hqd_pq_control |= 1 << CP_HQD_PQ_CONTROL__PRIV_STATE__SHIFT |
+			1 << CP_HQD_PQ_CONTROL__KMD_QUEUE__SHIFT;
+
+	return retval;
+}
+
+static int update_mqd_hiq(struct mqd_manager *mm, void *mqd,
+			struct queue_properties *q)
+{
+	struct v9_mqd *m;
+	int retval = update_mqd(mm, mqd, q);
+
+	if (retval != 0)
+		return retval;
+
+	/* TODO: what's the point? update_mqd already does this. */
+	m = get_mqd(mqd);
+	m->cp_hqd_vmid = q->vmid;
+	return retval;
+}
+
+static int init_mqd_sdma(struct mqd_manager *mm, void **mqd,
+		struct kfd_mem_obj **mqd_mem_obj, uint64_t *gart_addr,
+		struct queue_properties *q)
+{
+	int retval;
+	struct v9_sdma_mqd *m;
+
+
+	retval = kfd_gtt_sa_allocate(mm->dev,
+			sizeof(struct v9_sdma_mqd),
+			mqd_mem_obj);
+
+	if (retval != 0)
+		return -ENOMEM;
+
+	m = (struct v9_sdma_mqd *) (*mqd_mem_obj)->cpu_ptr;
+
+	memset(m, 0, sizeof(struct v9_sdma_mqd));
+
+	*mqd = m;
+	if (gart_addr)
+		*gart_addr = (*mqd_mem_obj)->gpu_addr;
+
+	retval = mm->update_mqd(mm, m, q);
+
+	return retval;
+}
+
+static void uninit_mqd_sdma(struct mqd_manager *mm, void *mqd,
+		struct kfd_mem_obj *mqd_mem_obj)
+{
+	kfd_gtt_sa_free(mm->dev, mqd_mem_obj);
+}
+
+static int load_mqd_sdma(struct mqd_manager *mm, void *mqd,
+		uint32_t pipe_id, uint32_t queue_id,
+		struct queue_properties *p, struct mm_struct *mms)
+{
+	return mm->dev->kfd2kgd->hqd_sdma_load(mm->dev->kgd, mqd,
+					       (uint32_t __user *)p->write_ptr,
+					       mms);
+}
+
+#define SDMA_RLC_DUMMY_DEFAULT 0xf
+
+static int update_mqd_sdma(struct mqd_manager *mm, void *mqd,
+		struct queue_properties *q)
+{
+	struct v9_sdma_mqd *m;
+
+	m = get_sdma_mqd(mqd);
+	m->sdmax_rlcx_rb_cntl = order_base_2(q->queue_size / 4)
+		<< SDMA0_RLC0_RB_CNTL__RB_SIZE__SHIFT |
+		q->vmid << SDMA0_RLC0_RB_CNTL__RB_VMID__SHIFT |
+		1 << SDMA0_RLC0_RB_CNTL__RPTR_WRITEBACK_ENABLE__SHIFT |
+		6 << SDMA0_RLC0_RB_CNTL__RPTR_WRITEBACK_TIMER__SHIFT;
+
+	m->sdmax_rlcx_rb_base = lower_32_bits(q->queue_address >> 8);
+	m->sdmax_rlcx_rb_base_hi = upper_32_bits(q->queue_address >> 8);
+	m->sdmax_rlcx_rb_rptr_addr_lo = lower_32_bits((uint64_t)q->read_ptr);
+	m->sdmax_rlcx_rb_rptr_addr_hi = upper_32_bits((uint64_t)q->read_ptr);
+	m->sdmax_rlcx_doorbell_offset =
+		q->doorbell_off << SDMA0_RLC0_DOORBELL_OFFSET__OFFSET__SHIFT;
+
+	m->sdma_engine_id = q->sdma_engine_id;
+	m->sdma_queue_id = q->sdma_queue_id;
+	m->sdmax_rlcx_dummy_reg = SDMA_RLC_DUMMY_DEFAULT;
+
+	q->is_active = (q->queue_size > 0 &&
+			q->queue_address != 0 &&
+			q->queue_percent > 0 &&
+			!q->is_evicted);
+
+	return 0;
+}
+
+/*
+ *  * preempt type here is ignored because there is only one way
+ *  * to preempt sdma queue
+ */
+static int destroy_mqd_sdma(struct mqd_manager *mm, void *mqd,
+		enum kfd_preempt_type type,
+		unsigned int timeout, uint32_t pipe_id,
+		uint32_t queue_id)
+{
+	return mm->dev->kfd2kgd->hqd_sdma_destroy(mm->dev->kgd, mqd, timeout);
+}
+
+static bool is_occupied_sdma(struct mqd_manager *mm, void *mqd,
+		uint64_t queue_address, uint32_t pipe_id,
+		uint32_t queue_id)
+{
+	return mm->dev->kfd2kgd->hqd_sdma_is_occupied(mm->dev->kgd, mqd);
+}
+
+#if defined(CONFIG_DEBUG_FS)
+
+static int debugfs_show_mqd(struct seq_file *m, void *data)
+{
+	seq_hex_dump(m, "    ", DUMP_PREFIX_OFFSET, 32, 4,
+		     data, sizeof(struct v9_mqd), false);
+	return 0;
+}
+
+static int debugfs_show_mqd_sdma(struct seq_file *m, void *data)
+{
+	seq_hex_dump(m, "    ", DUMP_PREFIX_OFFSET, 32, 4,
+		     data, sizeof(struct v9_sdma_mqd), false);
+	return 0;
+}
+
+#endif
+
+struct mqd_manager *mqd_manager_init_v9(enum KFD_MQD_TYPE type,
+		struct kfd_dev *dev)
+{
+	struct mqd_manager *mqd;
+
+	if (WARN_ON(type >= KFD_MQD_TYPE_MAX))
+		return NULL;
+
+	mqd = kzalloc(sizeof(*mqd), GFP_NOIO);
+	if (!mqd)
+		return NULL;
+
+	mqd->dev = dev;
+
+	switch (type) {
+	case KFD_MQD_TYPE_CP:
+	case KFD_MQD_TYPE_COMPUTE:
+		mqd->init_mqd = init_mqd;
+		mqd->uninit_mqd = uninit_mqd;
+		mqd->load_mqd = load_mqd;
+		mqd->update_mqd = update_mqd;
+		mqd->destroy_mqd = destroy_mqd;
+		mqd->is_occupied = is_occupied;
+#if defined(CONFIG_DEBUG_FS)
+		mqd->debugfs_show_mqd = debugfs_show_mqd;
+#endif
+		break;
+	case KFD_MQD_TYPE_HIQ:
+		mqd->init_mqd = init_mqd_hiq;
+		mqd->uninit_mqd = uninit_mqd;
+		mqd->load_mqd = load_mqd;
+		mqd->update_mqd = update_mqd_hiq;
+		mqd->destroy_mqd = destroy_mqd;
+		mqd->is_occupied = is_occupied;
+#if defined(CONFIG_DEBUG_FS)
+		mqd->debugfs_show_mqd = debugfs_show_mqd;
+#endif
+		break;
+	case KFD_MQD_TYPE_SDMA:
+		mqd->init_mqd = init_mqd_sdma;
+		mqd->uninit_mqd = uninit_mqd_sdma;
+		mqd->load_mqd = load_mqd_sdma;
+		mqd->update_mqd = update_mqd_sdma;
+		mqd->destroy_mqd = destroy_mqd_sdma;
+		mqd->is_occupied = is_occupied_sdma;
+#if defined(CONFIG_DEBUG_FS)
+		mqd->debugfs_show_mqd = debugfs_show_mqd_sdma;
+#endif
+		break;
+	default:
+		kfree(mqd);
+		return NULL;
+	}
+
+	return mqd;
+}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c
index 89e4242e43e7..481307b8b4db 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_mqd_manager_vi.c
@@ -394,7 +394,7 @@ struct mqd_manager *mqd_manager_init_vi(enum KFD_MQD_TYPE type,
 	if (WARN_ON(type >= KFD_MQD_TYPE_MAX))
 		return NULL;
 
-	mqd = kzalloc(sizeof(*mqd), GFP_KERNEL);
+	mqd = kzalloc(sizeof(*mqd), GFP_NOIO);
 	if (!mqd)
 		return NULL;
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
index 89ba4c670ec5..c317feb43f69 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c
@@ -26,8 +26,6 @@
 #include "kfd_device_queue_manager.h"
 #include "kfd_kernel_queue.h"
 #include "kfd_priv.h"
-#include "kfd_pm4_headers_vi.h"
-#include "kfd_pm4_opcodes.h"
 
 static inline void inc_wptr(unsigned int *wptr, unsigned int increment_bytes,
 				unsigned int buffer_size_bytes)
@@ -39,18 +37,6 @@ static inline void inc_wptr(unsigned int *wptr, unsigned int increment_bytes,
 	*wptr = temp;
 }
 
-static unsigned int build_pm4_header(unsigned int opcode, size_t packet_size)
-{
-	union PM4_MES_TYPE_3_HEADER header;
-
-	header.u32All = 0;
-	header.opcode = opcode;
-	header.count = packet_size / 4 - 2;
-	header.type = PM4_TYPE_3;
-
-	return header.u32All;
-}
-
 static void pm_calc_rlib_size(struct packet_manager *pm,
 				unsigned int *rlib_size,
 				bool *over_subscription)
@@ -80,9 +66,9 @@ static void pm_calc_rlib_size(struct packet_manager *pm,
 		pr_debug("Over subscribed runlist\n");
 	}
 
-	map_queue_size = sizeof(struct pm4_mes_map_queues);
+	map_queue_size = pm->pmf->map_queues_size;
 	/* calculate run list ib allocation size */
-	*rlib_size = process_count * sizeof(struct pm4_mes_map_process) +
+	*rlib_size = process_count * pm->pmf->map_process_size +
 		     queue_count * map_queue_size;
 
 	/*
@@ -90,7 +76,7 @@ static void pm_calc_rlib_size(struct packet_manager *pm,
 	 * when over subscription
 	 */
 	if (*over_subscription)
-		*rlib_size += sizeof(struct pm4_mes_runlist);
+		*rlib_size += pm->pmf->runlist_size;
 
 	pr_debug("runlist ib size %d\n", *rlib_size);
 }
@@ -108,12 +94,14 @@ static int pm_allocate_runlist_ib(struct packet_manager *pm,
 
 	pm_calc_rlib_size(pm, rl_buffer_size, is_over_subscription);
 
+	mutex_lock(&pm->lock);
+
 	retval = kfd_gtt_sa_allocate(pm->dqm->dev, *rl_buffer_size,
 					&pm->ib_buffer_obj);
 
 	if (retval) {
 		pr_err("Failed to allocate runlist IB\n");
-		return retval;
+		goto out;
 	}
 
 	*(void **)rl_buffer = pm->ib_buffer_obj->cpu_ptr;
@@ -121,138 +109,10 @@ static int pm_allocate_runlist_ib(struct packet_manager *pm,
 
 	memset(*rl_buffer, 0, *rl_buffer_size);
 	pm->allocated = true;
-	return retval;
-}
-
-static int pm_create_runlist(struct packet_manager *pm, uint32_t *buffer,
-			uint64_t ib, size_t ib_size_in_dwords, bool chain)
-{
-	struct pm4_mes_runlist *packet;
-	int concurrent_proc_cnt = 0;
-	struct kfd_dev *kfd = pm->dqm->dev;
-
-	if (WARN_ON(!ib))
-		return -EFAULT;
-
-	/* Determine the number of processes to map together to HW:
-	 * it can not exceed the number of VMIDs available to the
-	 * scheduler, and it is determined by the smaller of the number
-	 * of processes in the runlist and kfd module parameter
-	 * hws_max_conc_proc.
-	 * Note: the arbitration between the number of VMIDs and
-	 * hws_max_conc_proc has been done in
-	 * kgd2kfd_device_init().
-	 */
-	concurrent_proc_cnt = min(pm->dqm->processes_count,
-			kfd->max_proc_per_quantum);
-
-	packet = (struct pm4_mes_runlist *)buffer;
-
-	memset(buffer, 0, sizeof(struct pm4_mes_runlist));
-	packet->header.u32All = build_pm4_header(IT_RUN_LIST,
-						sizeof(struct pm4_mes_runlist));
-
-	packet->bitfields4.ib_size = ib_size_in_dwords;
-	packet->bitfields4.chain = chain ? 1 : 0;
-	packet->bitfields4.offload_polling = 0;
-	packet->bitfields4.valid = 1;
-	packet->bitfields4.process_cnt = concurrent_proc_cnt;
-	packet->ordinal2 = lower_32_bits(ib);
-	packet->bitfields3.ib_base_hi = upper_32_bits(ib);
-
-	return 0;
-}
-
-static int pm_create_map_process(struct packet_manager *pm, uint32_t *buffer,
-				struct qcm_process_device *qpd)
-{
-	struct pm4_mes_map_process *packet;
-
-	packet = (struct pm4_mes_map_process *)buffer;
 
-	memset(buffer, 0, sizeof(struct pm4_mes_map_process));
-
-	packet->header.u32All = build_pm4_header(IT_MAP_PROCESS,
-					sizeof(struct pm4_mes_map_process));
-	packet->bitfields2.diq_enable = (qpd->is_debug) ? 1 : 0;
-	packet->bitfields2.process_quantum = 1;
-	packet->bitfields2.pasid = qpd->pqm->process->pasid;
-	packet->bitfields3.page_table_base = qpd->page_table_base;
-	packet->bitfields10.gds_size = qpd->gds_size;
-	packet->bitfields10.num_gws = qpd->num_gws;
-	packet->bitfields10.num_oac = qpd->num_oac;
-	packet->bitfields10.num_queues = (qpd->is_debug) ? 0 : qpd->queue_count;
-
-	packet->sh_mem_config = qpd->sh_mem_config;
-	packet->sh_mem_bases = qpd->sh_mem_bases;
-	packet->sh_mem_ape1_base = qpd->sh_mem_ape1_base;
-	packet->sh_mem_ape1_limit = qpd->sh_mem_ape1_limit;
-
-	packet->sh_hidden_private_base_vmid = qpd->sh_hidden_private_base;
-
-	packet->gds_addr_lo = lower_32_bits(qpd->gds_context_area);
-	packet->gds_addr_hi = upper_32_bits(qpd->gds_context_area);
-
-	return 0;
-}
-
-static int pm_create_map_queue(struct packet_manager *pm, uint32_t *buffer,
-		struct queue *q, bool is_static)
-{
-	struct pm4_mes_map_queues *packet;
-	bool use_static = is_static;
-
-	packet = (struct pm4_mes_map_queues *)buffer;
-	memset(buffer, 0, sizeof(struct pm4_mes_map_queues));
-
-	packet->header.u32All = build_pm4_header(IT_MAP_QUEUES,
-						sizeof(struct pm4_mes_map_queues));
-	packet->bitfields2.alloc_format =
-		alloc_format__mes_map_queues__one_per_pipe_vi;
-	packet->bitfields2.num_queues = 1;
-	packet->bitfields2.queue_sel =
-		queue_sel__mes_map_queues__map_to_hws_determined_queue_slots_vi;
-
-	packet->bitfields2.engine_sel =
-		engine_sel__mes_map_queues__compute_vi;
-	packet->bitfields2.queue_type =
-		queue_type__mes_map_queues__normal_compute_vi;
-
-	switch (q->properties.type) {
-	case KFD_QUEUE_TYPE_COMPUTE:
-		if (use_static)
-			packet->bitfields2.queue_type =
-		queue_type__mes_map_queues__normal_latency_static_queue_vi;
-		break;
-	case KFD_QUEUE_TYPE_DIQ:
-		packet->bitfields2.queue_type =
-			queue_type__mes_map_queues__debug_interface_queue_vi;
-		break;
-	case KFD_QUEUE_TYPE_SDMA:
-		packet->bitfields2.engine_sel = q->properties.sdma_engine_id +
-				engine_sel__mes_map_queues__sdma0_vi;
-		use_static = false; /* no static queues under SDMA */
-		break;
-	default:
-		WARN(1, "queue type %d", q->properties.type);
-		return -EINVAL;
-	}
-	packet->bitfields3.doorbell_offset =
-			q->properties.doorbell_off;
-
-	packet->mqd_addr_lo =
-			lower_32_bits(q->gart_mqd_addr);
-
-	packet->mqd_addr_hi =
-			upper_32_bits(q->gart_mqd_addr);
-
-	packet->wptr_addr_lo =
-			lower_32_bits((uint64_t)q->properties.write_ptr);
-
-	packet->wptr_addr_hi =
-			upper_32_bits((uint64_t)q->properties.write_ptr);
-
-	return 0;
+out:
+	mutex_unlock(&pm->lock);
+	return retval;
 }
 
 static int pm_create_runlist_ib(struct packet_manager *pm,
@@ -292,12 +152,12 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 			return -ENOMEM;
 		}
 
-		retval = pm_create_map_process(pm, &rl_buffer[rl_wptr], qpd);
+		retval = pm->pmf->map_process(pm, &rl_buffer[rl_wptr], qpd);
 		if (retval)
 			return retval;
 
 		proccesses_mapped++;
-		inc_wptr(&rl_wptr, sizeof(struct pm4_mes_map_process),
+		inc_wptr(&rl_wptr, pm->pmf->map_process_size,
 				alloc_size_bytes);
 
 		list_for_each_entry(kq, &qpd->priv_queue_list, list) {
@@ -307,7 +167,7 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 			pr_debug("static_queue, mapping kernel q %d, is debug status %d\n",
 				kq->queue->queue, qpd->is_debug);
 
-			retval = pm_create_map_queue(pm,
+			retval = pm->pmf->map_queues(pm,
 						&rl_buffer[rl_wptr],
 						kq->queue,
 						qpd->is_debug);
@@ -315,7 +175,7 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 				return retval;
 
 			inc_wptr(&rl_wptr,
-				sizeof(struct pm4_mes_map_queues),
+				pm->pmf->map_queues_size,
 				alloc_size_bytes);
 		}
 
@@ -326,7 +186,7 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 			pr_debug("static_queue, mapping user queue %d, is debug status %d\n",
 				q->queue, qpd->is_debug);
 
-			retval = pm_create_map_queue(pm,
+			retval = pm->pmf->map_queues(pm,
 						&rl_buffer[rl_wptr],
 						q,
 						qpd->is_debug);
@@ -335,7 +195,7 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 				return retval;
 
 			inc_wptr(&rl_wptr,
-				sizeof(struct pm4_mes_map_queues),
+				pm->pmf->map_queues_size,
 				alloc_size_bytes);
 		}
 	}
@@ -343,7 +203,7 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 	pr_debug("Finished map process and queues to runlist\n");
 
 	if (is_over_subscription)
-		retval = pm_create_runlist(pm, &rl_buffer[rl_wptr],
+		retval = pm->pmf->runlist(pm, &rl_buffer[rl_wptr],
 					*rl_gpu_addr,
 					alloc_size_bytes / sizeof(uint32_t),
 					true);
@@ -355,45 +215,29 @@ static int pm_create_runlist_ib(struct packet_manager *pm,
 	return retval;
 }
 
-/* pm_create_release_mem - Create a RELEASE_MEM packet and return the size
- *     of this packet
- *     @gpu_addr - GPU address of the packet. It's a virtual address.
- *     @buffer - buffer to fill up with the packet. It's a CPU kernel pointer
- *     Return - length of the packet
- */
-uint32_t pm_create_release_mem(uint64_t gpu_addr, uint32_t *buffer)
-{
-	struct pm4_mec_release_mem *packet;
-
-	WARN_ON(!buffer);
-
-	packet = (struct pm4_mec_release_mem *)buffer;
-	memset(buffer, 0, sizeof(*packet));
-
-	packet->header.u32All = build_pm4_header(IT_RELEASE_MEM,
-						 sizeof(*packet));
-
-	packet->bitfields2.event_type = CACHE_FLUSH_AND_INV_TS_EVENT;
-	packet->bitfields2.event_index = event_index___release_mem__end_of_pipe;
-	packet->bitfields2.tcl1_action_ena = 1;
-	packet->bitfields2.tc_action_ena = 1;
-	packet->bitfields2.cache_policy = cache_policy___release_mem__lru;
-	packet->bitfields2.atc = 0;
-
-	packet->bitfields3.data_sel = data_sel___release_mem__send_32_bit_low;
-	packet->bitfields3.int_sel =
-		int_sel___release_mem__send_interrupt_after_write_confirm;
-
-	packet->bitfields4.address_lo_32b = (gpu_addr & 0xffffffff) >> 2;
-	packet->address_hi = upper_32_bits(gpu_addr);
-
-	packet->data_lo = 0;
-
-	return sizeof(*packet) / sizeof(unsigned int);
-}
-
 int pm_init(struct packet_manager *pm, struct device_queue_manager *dqm)
 {
+	switch (dqm->dev->device_info->asic_family) {
+	case CHIP_KAVERI:
+	case CHIP_HAWAII:
+		/* PM4 packet structures on CIK are the same as on VI */
+	case CHIP_CARRIZO:
+	case CHIP_TONGA:
+	case CHIP_FIJI:
+	case CHIP_POLARIS10:
+	case CHIP_POLARIS11:
+		pm->pmf = &kfd_vi_pm_funcs;
+		break;
+	case CHIP_VEGA10:
+	case CHIP_RAVEN:
+		pm->pmf = &kfd_v9_pm_funcs;
+		break;
+	default:
+		WARN(1, "Unexpected ASIC family %u",
+		     dqm->dev->device_info->asic_family);
+		return -EINVAL;
+	}
+
 	pm->dqm = dqm;
 	mutex_init(&pm->lock);
 	pm->priv_queue = kernel_queue_init(dqm->dev, KFD_QUEUE_TYPE_HIQ);
@@ -415,38 +259,25 @@ void pm_uninit(struct packet_manager *pm)
 int pm_send_set_resources(struct packet_manager *pm,
 				struct scheduling_resources *res)
 {
-	struct pm4_mes_set_resources *packet;
+	uint32_t *buffer, size;
 	int retval = 0;
 
+	size = pm->pmf->set_resources_size;
 	mutex_lock(&pm->lock);
 	pm->priv_queue->ops.acquire_packet_buffer(pm->priv_queue,
-					sizeof(*packet) / sizeof(uint32_t),
-					(unsigned int **)&packet);
-	if (!packet) {
+					size / sizeof(uint32_t),
+					(unsigned int **)&buffer);
+	if (!buffer) {
 		pr_err("Failed to allocate buffer on kernel queue\n");
 		retval = -ENOMEM;
 		goto out;
 	}
 
-	memset(packet, 0, sizeof(struct pm4_mes_set_resources));
-	packet->header.u32All = build_pm4_header(IT_SET_RESOURCES,
-					sizeof(struct pm4_mes_set_resources));
-
-	packet->bitfields2.queue_type =
-			queue_type__mes_set_resources__hsa_interface_queue_hiq;
-	packet->bitfields2.vmid_mask = res->vmid_mask;
-	packet->bitfields2.unmap_latency = KFD_UNMAP_LATENCY_MS / 100;
-	packet->bitfields7.oac_mask = res->oac_mask;
-	packet->bitfields8.gds_heap_base = res->gds_heap_base;
-	packet->bitfields8.gds_heap_size = res->gds_heap_size;
-
-	packet->gws_mask_lo = lower_32_bits(res->gws_mask);
-	packet->gws_mask_hi = upper_32_bits(res->gws_mask);
-
-	packet->queue_mask_lo = lower_32_bits(res->queue_mask);
-	packet->queue_mask_hi = upper_32_bits(res->queue_mask);
-
-	pm->priv_queue->ops.submit_packet(pm->priv_queue);
+	retval = pm->pmf->set_resources(pm, buffer, res);
+	if (!retval)
+		pm->priv_queue->ops.submit_packet(pm->priv_queue);
+	else
+		pm->priv_queue->ops.rollback_packet(pm->priv_queue);
 
 out:
 	mutex_unlock(&pm->lock);
@@ -468,7 +299,7 @@ int pm_send_runlist(struct packet_manager *pm, struct list_head *dqm_queues)
 
 	pr_debug("runlist IB address: 0x%llX\n", rl_gpu_ib_addr);
 
-	packet_size_dwords = sizeof(struct pm4_mes_runlist) / sizeof(uint32_t);
+	packet_size_dwords = pm->pmf->runlist_size / sizeof(uint32_t);
 	mutex_lock(&pm->lock);
 
 	retval = pm->priv_queue->ops.acquire_packet_buffer(pm->priv_queue,
@@ -476,7 +307,7 @@ int pm_send_runlist(struct packet_manager *pm, struct list_head *dqm_queues)
 	if (retval)
 		goto fail_acquire_packet_buffer;
 
-	retval = pm_create_runlist(pm, rl_buffer, rl_gpu_ib_addr,
+	retval = pm->pmf->runlist(pm, rl_buffer, rl_gpu_ib_addr,
 					rl_ib_size / sizeof(uint32_t), false);
 	if (retval)
 		goto fail_create_runlist;
@@ -499,37 +330,29 @@ fail_create_runlist_ib:
 int pm_send_query_status(struct packet_manager *pm, uint64_t fence_address,
 			uint32_t fence_value)
 {
-	int retval;
-	struct pm4_mes_query_status *packet;
+	uint32_t *buffer, size;
+	int retval = 0;
 
 	if (WARN_ON(!fence_address))
 		return -EFAULT;
 
+	size = pm->pmf->query_status_size;
 	mutex_lock(&pm->lock);
-	retval = pm->priv_queue->ops.acquire_packet_buffer(
-			pm->priv_queue,
-			sizeof(struct pm4_mes_query_status) / sizeof(uint32_t),
-			(unsigned int **)&packet);
-	if (retval)
-		goto fail_acquire_packet_buffer;
-
-	packet->header.u32All = build_pm4_header(IT_QUERY_STATUS,
-					sizeof(struct pm4_mes_query_status));
-
-	packet->bitfields2.context_id = 0;
-	packet->bitfields2.interrupt_sel =
-			interrupt_sel__mes_query_status__completion_status;
-	packet->bitfields2.command =
-			command__mes_query_status__fence_only_after_write_ack;
-
-	packet->addr_hi = upper_32_bits((uint64_t)fence_address);
-	packet->addr_lo = lower_32_bits((uint64_t)fence_address);
-	packet->data_hi = upper_32_bits((uint64_t)fence_value);
-	packet->data_lo = lower_32_bits((uint64_t)fence_value);
+	pm->priv_queue->ops.acquire_packet_buffer(pm->priv_queue,
+			size / sizeof(uint32_t), (unsigned int **)&buffer);
+	if (!buffer) {
+		pr_err("Failed to allocate buffer on kernel queue\n");
+		retval = -ENOMEM;
+		goto out;
+	}
 
-	pm->priv_queue->ops.submit_packet(pm->priv_queue);
+	retval = pm->pmf->query_status(pm, buffer, fence_address, fence_value);
+	if (!retval)
+		pm->priv_queue->ops.submit_packet(pm->priv_queue);
+	else
+		pm->priv_queue->ops.rollback_packet(pm->priv_queue);
 
-fail_acquire_packet_buffer:
+out:
 	mutex_unlock(&pm->lock);
 	return retval;
 }
@@ -539,82 +362,27 @@ int pm_send_unmap_queue(struct packet_manager *pm, enum kfd_queue_type type,
 			uint32_t filter_param, bool reset,
 			unsigned int sdma_engine)
 {
-	int retval;
-	uint32_t *buffer;
-	struct pm4_mes_unmap_queues *packet;
+	uint32_t *buffer, size;
+	int retval = 0;
 
+	size = pm->pmf->unmap_queues_size;
 	mutex_lock(&pm->lock);
-	retval = pm->priv_queue->ops.acquire_packet_buffer(
-			pm->priv_queue,
-			sizeof(struct pm4_mes_unmap_queues) / sizeof(uint32_t),
-			&buffer);
-	if (retval)
-		goto err_acquire_packet_buffer;
-
-	packet = (struct pm4_mes_unmap_queues *)buffer;
-	memset(buffer, 0, sizeof(struct pm4_mes_unmap_queues));
-	pr_debug("static_queue: unmapping queues: filter is %d , reset is %d , type is %d\n",
-		filter, reset, type);
-	packet->header.u32All = build_pm4_header(IT_UNMAP_QUEUES,
-					sizeof(struct pm4_mes_unmap_queues));
-	switch (type) {
-	case KFD_QUEUE_TYPE_COMPUTE:
-	case KFD_QUEUE_TYPE_DIQ:
-		packet->bitfields2.engine_sel =
-			engine_sel__mes_unmap_queues__compute;
-		break;
-	case KFD_QUEUE_TYPE_SDMA:
-		packet->bitfields2.engine_sel =
-			engine_sel__mes_unmap_queues__sdma0 + sdma_engine;
-		break;
-	default:
-		WARN(1, "queue type %d", type);
-		retval = -EINVAL;
-		goto err_invalid;
+	pm->priv_queue->ops.acquire_packet_buffer(pm->priv_queue,
+			size / sizeof(uint32_t), (unsigned int **)&buffer);
+	if (!buffer) {
+		pr_err("Failed to allocate buffer on kernel queue\n");
+		retval = -ENOMEM;
+		goto out;
 	}
 
-	if (reset)
-		packet->bitfields2.action =
-				action__mes_unmap_queues__reset_queues;
+	retval = pm->pmf->unmap_queues(pm, buffer, type, filter, filter_param,
+				       reset, sdma_engine);
+	if (!retval)
+		pm->priv_queue->ops.submit_packet(pm->priv_queue);
 	else
-		packet->bitfields2.action =
-				action__mes_unmap_queues__preempt_queues;
-
-	switch (filter) {
-	case KFD_UNMAP_QUEUES_FILTER_SINGLE_QUEUE:
-		packet->bitfields2.queue_sel =
-				queue_sel__mes_unmap_queues__perform_request_on_specified_queues;
-		packet->bitfields2.num_queues = 1;
-		packet->bitfields3b.doorbell_offset0 = filter_param;
-		break;
-	case KFD_UNMAP_QUEUES_FILTER_BY_PASID:
-		packet->bitfields2.queue_sel =
-				queue_sel__mes_unmap_queues__perform_request_on_pasid_queues;
-		packet->bitfields3a.pasid = filter_param;
-		break;
-	case KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES:
-		packet->bitfields2.queue_sel =
-				queue_sel__mes_unmap_queues__unmap_all_queues;
-		break;
-	case KFD_UNMAP_QUEUES_FILTER_DYNAMIC_QUEUES:
-		/* in this case, we do not preempt static queues */
-		packet->bitfields2.queue_sel =
-				queue_sel__mes_unmap_queues__unmap_all_non_static_queues;
-		break;
-	default:
-		WARN(1, "filter %d", filter);
-		retval = -EINVAL;
-		goto err_invalid;
-	}
+		pm->priv_queue->ops.rollback_packet(pm->priv_queue);
 
-	pm->priv_queue->ops.submit_packet(pm->priv_queue);
-
-	mutex_unlock(&pm->lock);
-	return 0;
-
-err_invalid:
-	pm->priv_queue->ops.rollback_packet(pm->priv_queue);
-err_acquire_packet_buffer:
+out:
 	mutex_unlock(&pm->lock);
 	return retval;
 }
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_ai.h b/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_ai.h
new file mode 100644
index 000000000000..f2bcf5c092ea
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_ai.h
@@ -0,0 +1,583 @@
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef F32_MES_PM4_PACKETS_H
+#define F32_MES_PM4_PACKETS_H
+
+#ifndef PM4_MES_HEADER_DEFINED
+#define PM4_MES_HEADER_DEFINED
+union PM4_MES_TYPE_3_HEADER {
+	struct {
+		uint32_t reserved1 : 8; /* < reserved */
+		uint32_t opcode    : 8; /* < IT opcode */
+		uint32_t count     : 14;/* < number of DWORDs - 1 in the
+					 *   information body.
+					 */
+		uint32_t type      : 2; /* < packet identifier.
+					 *   It should be 3 for type 3 packets
+					 */
+	};
+	uint32_t u32All;
+};
+#endif /* PM4_MES_HEADER_DEFINED */
+
+/*--------------------MES_SET_RESOURCES--------------------*/
+
+#ifndef PM4_MES_SET_RESOURCES_DEFINED
+#define PM4_MES_SET_RESOURCES_DEFINED
+enum mes_set_resources_queue_type_enum {
+	queue_type__mes_set_resources__kernel_interface_queue_kiq = 0,
+	queue_type__mes_set_resources__hsa_interface_queue_hiq = 1,
+	queue_type__mes_set_resources__hsa_debug_interface_queue = 4
+};
+
+
+struct pm4_mes_set_resources {
+	union {
+		union PM4_MES_TYPE_3_HEADER	header;		/* header */
+		uint32_t			ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t vmid_mask:16;
+			uint32_t unmap_latency:8;
+			uint32_t reserved1:5;
+			enum mes_set_resources_queue_type_enum queue_type:3;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	uint32_t queue_mask_lo;
+	uint32_t queue_mask_hi;
+	uint32_t gws_mask_lo;
+	uint32_t gws_mask_hi;
+
+	union {
+		struct {
+			uint32_t oac_mask:16;
+			uint32_t reserved2:16;
+		} bitfields7;
+		uint32_t ordinal7;
+	};
+
+	union {
+		struct {
+		uint32_t gds_heap_base:6;
+		uint32_t reserved3:5;
+		uint32_t gds_heap_size:6;
+		uint32_t reserved4:15;
+		} bitfields8;
+		uint32_t ordinal8;
+	};
+
+};
+#endif
+
+/*--------------------MES_RUN_LIST--------------------*/
+
+#ifndef PM4_MES_RUN_LIST_DEFINED
+#define PM4_MES_RUN_LIST_DEFINED
+
+struct pm4_mes_runlist {
+	union {
+		union PM4_MES_TYPE_3_HEADER header; /* header */
+		uint32_t ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t reserved1:2;
+			uint32_t ib_base_lo:30;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	uint32_t ib_base_hi;
+
+	union {
+		struct {
+			uint32_t ib_size:20;
+			uint32_t chain:1;
+			uint32_t offload_polling:1;
+			uint32_t reserved2:1;
+			uint32_t valid:1;
+			uint32_t process_cnt:4;
+			uint32_t reserved3:4;
+		} bitfields4;
+		uint32_t ordinal4;
+	};
+
+};
+#endif
+
+/*--------------------MES_MAP_PROCESS--------------------*/
+
+#ifndef PM4_MES_MAP_PROCESS_DEFINED
+#define PM4_MES_MAP_PROCESS_DEFINED
+
+struct pm4_mes_map_process {
+	union {
+		union PM4_MES_TYPE_3_HEADER header;	/* header */
+		uint32_t ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t pasid:16;
+			uint32_t reserved1:8;
+			uint32_t diq_enable:1;
+			uint32_t process_quantum:7;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	uint32_t vm_context_page_table_base_addr_lo32;
+
+	uint32_t vm_context_page_table_base_addr_hi32;
+
+	uint32_t sh_mem_bases;
+
+	uint32_t sh_mem_config;
+
+	uint32_t sq_shader_tba_lo;
+
+	uint32_t sq_shader_tba_hi;
+
+	uint32_t sq_shader_tma_lo;
+
+	uint32_t sq_shader_tma_hi;
+
+	uint32_t reserved6;
+
+	uint32_t gds_addr_lo;
+
+	uint32_t gds_addr_hi;
+
+	union {
+		struct {
+			uint32_t num_gws:6;
+			uint32_t reserved7:1;
+			uint32_t sdma_enable:1;
+			uint32_t num_oac:4;
+			uint32_t reserved8:4;
+			uint32_t gds_size:6;
+			uint32_t num_queues:10;
+		} bitfields14;
+		uint32_t ordinal14;
+	};
+
+	uint32_t completion_signal_lo;
+
+	uint32_t completion_signal_hi;
+
+};
+
+#endif
+
+/*--------------------MES_MAP_PROCESS_VM--------------------*/
+
+#ifndef PM4_MES_MAP_PROCESS_VM_DEFINED
+#define PM4_MES_MAP_PROCESS_VM_DEFINED
+
+struct PM4_MES_MAP_PROCESS_VM {
+	union {
+		union PM4_MES_TYPE_3_HEADER header;	/* header */
+		uint32_t ordinal1;
+	};
+
+	uint32_t reserved1;
+
+	uint32_t vm_context_cntl;
+
+	uint32_t reserved2;
+
+	uint32_t vm_context_page_table_end_addr_lo32;
+
+	uint32_t vm_context_page_table_end_addr_hi32;
+
+	uint32_t vm_context_page_table_start_addr_lo32;
+
+	uint32_t vm_context_page_table_start_addr_hi32;
+
+	uint32_t reserved3;
+
+	uint32_t reserved4;
+
+	uint32_t reserved5;
+
+	uint32_t reserved6;
+
+	uint32_t reserved7;
+
+	uint32_t reserved8;
+
+	uint32_t completion_signal_lo32;
+
+	uint32_t completion_signal_hi32;
+
+};
+#endif
+
+/*--------------------MES_MAP_QUEUES--------------------*/
+
+#ifndef PM4_MES_MAP_QUEUES_VI_DEFINED
+#define PM4_MES_MAP_QUEUES_VI_DEFINED
+enum mes_map_queues_queue_sel_enum {
+	queue_sel__mes_map_queues__map_to_specified_queue_slots_vi = 0,
+queue_sel__mes_map_queues__map_to_hws_determined_queue_slots_vi = 1
+};
+
+enum mes_map_queues_queue_type_enum {
+	queue_type__mes_map_queues__normal_compute_vi = 0,
+	queue_type__mes_map_queues__debug_interface_queue_vi = 1,
+	queue_type__mes_map_queues__normal_latency_static_queue_vi = 2,
+queue_type__mes_map_queues__low_latency_static_queue_vi = 3
+};
+
+enum mes_map_queues_alloc_format_enum {
+	alloc_format__mes_map_queues__one_per_pipe_vi = 0,
+alloc_format__mes_map_queues__all_on_one_pipe_vi = 1
+};
+
+enum mes_map_queues_engine_sel_enum {
+	engine_sel__mes_map_queues__compute_vi = 0,
+	engine_sel__mes_map_queues__sdma0_vi = 2,
+	engine_sel__mes_map_queues__sdma1_vi = 3
+};
+
+
+struct pm4_mes_map_queues {
+	union {
+		union PM4_MES_TYPE_3_HEADER   header;            /* header */
+		uint32_t            ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t reserved1:4;
+			enum mes_map_queues_queue_sel_enum queue_sel:2;
+			uint32_t reserved2:15;
+			enum mes_map_queues_queue_type_enum queue_type:3;
+			enum mes_map_queues_alloc_format_enum alloc_format:2;
+			enum mes_map_queues_engine_sel_enum engine_sel:3;
+			uint32_t num_queues:3;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	union {
+		struct {
+			uint32_t reserved3:1;
+			uint32_t check_disable:1;
+			uint32_t doorbell_offset:26;
+			uint32_t reserved4:4;
+		} bitfields3;
+		uint32_t ordinal3;
+	};
+
+	uint32_t mqd_addr_lo;
+	uint32_t mqd_addr_hi;
+	uint32_t wptr_addr_lo;
+	uint32_t wptr_addr_hi;
+};
+#endif
+
+/*--------------------MES_QUERY_STATUS--------------------*/
+
+#ifndef PM4_MES_QUERY_STATUS_DEFINED
+#define PM4_MES_QUERY_STATUS_DEFINED
+enum mes_query_status_interrupt_sel_enum {
+	interrupt_sel__mes_query_status__completion_status = 0,
+	interrupt_sel__mes_query_status__process_status = 1,
+	interrupt_sel__mes_query_status__queue_status = 2
+};
+
+enum mes_query_status_command_enum {
+	command__mes_query_status__interrupt_only = 0,
+	command__mes_query_status__fence_only_immediate = 1,
+	command__mes_query_status__fence_only_after_write_ack = 2,
+	command__mes_query_status__fence_wait_for_write_ack_send_interrupt = 3
+};
+
+enum mes_query_status_engine_sel_enum {
+	engine_sel__mes_query_status__compute = 0,
+	engine_sel__mes_query_status__sdma0_queue = 2,
+	engine_sel__mes_query_status__sdma1_queue = 3
+};
+
+struct pm4_mes_query_status {
+	union {
+		union PM4_MES_TYPE_3_HEADER   header;            /* header */
+		uint32_t            ordinal1;
+	};
+
+	union {
+		struct {
+			uint32_t context_id:28;
+			enum mes_query_status_interrupt_sel_enum	interrupt_sel:2;
+			enum mes_query_status_command_enum command:2;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	union {
+		struct {
+			uint32_t pasid:16;
+			uint32_t reserved1:16;
+		} bitfields3a;
+		struct {
+			uint32_t reserved2:2;
+			uint32_t doorbell_offset:26;
+			enum mes_query_status_engine_sel_enum engine_sel:3;
+			uint32_t reserved3:1;
+		} bitfields3b;
+		uint32_t ordinal3;
+	};
+
+	uint32_t addr_lo;
+	uint32_t addr_hi;
+	uint32_t data_lo;
+	uint32_t data_hi;
+};
+#endif
+
+/*--------------------MES_UNMAP_QUEUES--------------------*/
+
+#ifndef PM4_MES_UNMAP_QUEUES_DEFINED
+#define PM4_MES_UNMAP_QUEUES_DEFINED
+enum mes_unmap_queues_action_enum {
+	action__mes_unmap_queues__preempt_queues = 0,
+	action__mes_unmap_queues__reset_queues = 1,
+	action__mes_unmap_queues__disable_process_queues = 2,
+	action__mes_unmap_queues__reserved = 3
+};
+
+enum mes_unmap_queues_queue_sel_enum {
+	queue_sel__mes_unmap_queues__perform_request_on_specified_queues = 0,
+	queue_sel__mes_unmap_queues__perform_request_on_pasid_queues = 1,
+	queue_sel__mes_unmap_queues__unmap_all_queues = 2,
+	queue_sel__mes_unmap_queues__unmap_all_non_static_queues = 3
+};
+
+enum mes_unmap_queues_engine_sel_enum {
+	engine_sel__mes_unmap_queues__compute = 0,
+	engine_sel__mes_unmap_queues__sdma0 = 2,
+	engine_sel__mes_unmap_queues__sdmal = 3
+};
+
+struct pm4_mes_unmap_queues {
+	union {
+		union PM4_MES_TYPE_3_HEADER   header;            /* header */
+		uint32_t            ordinal1;
+	};
+
+	union {
+		struct {
+			enum mes_unmap_queues_action_enum action:2;
+			uint32_t reserved1:2;
+			enum mes_unmap_queues_queue_sel_enum queue_sel:2;
+			uint32_t reserved2:20;
+			enum mes_unmap_queues_engine_sel_enum engine_sel:3;
+			uint32_t num_queues:3;
+		} bitfields2;
+		uint32_t ordinal2;
+	};
+
+	union {
+		struct {
+			uint32_t pasid:16;
+			uint32_t reserved3:16;
+		} bitfields3a;
+		struct {
+			uint32_t reserved4:2;
+			uint32_t doorbell_offset0:26;
+			int32_t reserved5:4;
+		} bitfields3b;
+		uint32_t ordinal3;
+	};
+
+	union {
+	struct {
+			uint32_t reserved6:2;
+			uint32_t doorbell_offset1:26;
+			uint32_t reserved7:4;
+		} bitfields4;
+		uint32_t ordinal4;
+	};
+
+	union {
+		struct {
+			uint32_t reserved8:2;
+			uint32_t doorbell_offset2:26;
+			uint32_t reserved9:4;
+		} bitfields5;
+		uint32_t ordinal5;
+	};
+
+	union {
+		struct {
+			uint32_t reserved10:2;
+			uint32_t doorbell_offset3:26;
+			uint32_t reserved11:4;
+		} bitfields6;
+		uint32_t ordinal6;
+	};
+};
+#endif
+
+#ifndef PM4_MEC_RELEASE_MEM_DEFINED
+#define PM4_MEC_RELEASE_MEM_DEFINED
+
+enum mec_release_mem_event_index_enum {
+	event_index__mec_release_mem__end_of_pipe = 5,
+	event_index__mec_release_mem__shader_done = 6
+};
+
+enum mec_release_mem_cache_policy_enum {
+	cache_policy__mec_release_mem__lru = 0,
+	cache_policy__mec_release_mem__stream = 1
+};
+
+enum mec_release_mem_pq_exe_status_enum {
+	pq_exe_status__mec_release_mem__default = 0,
+	pq_exe_status__mec_release_mem__phase_update = 1
+};
+
+enum mec_release_mem_dst_sel_enum {
+	dst_sel__mec_release_mem__memory_controller = 0,
+	dst_sel__mec_release_mem__tc_l2 = 1,
+	dst_sel__mec_release_mem__queue_write_pointer_register = 2,
+	dst_sel__mec_release_mem__queue_write_pointer_poll_mask_bit = 3
+};
+
+enum mec_release_mem_int_sel_enum {
+	int_sel__mec_release_mem__none = 0,
+	int_sel__mec_release_mem__send_interrupt_only = 1,
+	int_sel__mec_release_mem__send_interrupt_after_write_confirm = 2,
+	int_sel__mec_release_mem__send_data_after_write_confirm = 3,
+	int_sel__mec_release_mem__unconditionally_send_int_ctxid = 4,
+	int_sel__mec_release_mem__conditionally_send_int_ctxid_based_on_32_bit_compare = 5,
+	int_sel__mec_release_mem__conditionally_send_int_ctxid_based_on_64_bit_compare = 6
+};
+
+enum mec_release_mem_data_sel_enum {
+	data_sel__mec_release_mem__none = 0,
+	data_sel__mec_release_mem__send_32_bit_low = 1,
+	data_sel__mec_release_mem__send_64_bit_data = 2,
+	data_sel__mec_release_mem__send_gpu_clock_counter = 3,
+	data_sel__mec_release_mem__send_cp_perfcounter_hi_lo = 4,
+	data_sel__mec_release_mem__store_gds_data_to_memory = 5
+};
+
+struct pm4_mec_release_mem {
+	union {
+		union PM4_MES_TYPE_3_HEADER header;     /*header */
+		unsigned int ordinal1;
+	};
+
+	union {
+		struct {
+			unsigned int event_type:6;
+			unsigned int reserved1:2;
+			enum mec_release_mem_event_index_enum event_index:4;
+			unsigned int tcl1_vol_action_ena:1;
+			unsigned int tc_vol_action_ena:1;
+			unsigned int reserved2:1;
+			unsigned int tc_wb_action_ena:1;
+			unsigned int tcl1_action_ena:1;
+			unsigned int tc_action_ena:1;
+			uint32_t reserved3:1;
+			uint32_t tc_nc_action_ena:1;
+			uint32_t tc_wc_action_ena:1;
+			uint32_t tc_md_action_ena:1;
+			uint32_t reserved4:3;
+			enum mec_release_mem_cache_policy_enum cache_policy:2;
+			uint32_t reserved5:2;
+			enum mec_release_mem_pq_exe_status_enum pq_exe_status:1;
+			uint32_t reserved6:2;
+		} bitfields2;
+		unsigned int ordinal2;
+	};
+
+	union {
+		struct {
+			uint32_t reserved7:16;
+			enum mec_release_mem_dst_sel_enum dst_sel:2;
+			uint32_t reserved8:6;
+			enum mec_release_mem_int_sel_enum int_sel:3;
+			uint32_t reserved9:2;
+			enum mec_release_mem_data_sel_enum data_sel:3;
+		} bitfields3;
+		unsigned int ordinal3;
+	};
+
+	union {
+		struct {
+			uint32_t reserved10:2;
+			unsigned int address_lo_32b:30;
+		} bitfields4;
+		struct {
+			uint32_t reserved11:3;
+			uint32_t address_lo_64b:29;
+		} bitfields4b;
+		uint32_t reserved12;
+		unsigned int ordinal4;
+	};
+
+	union {
+		uint32_t address_hi;
+		uint32_t reserved13;
+		uint32_t ordinal5;
+	};
+
+	union {
+		uint32_t data_lo;
+		uint32_t cmp_data_lo;
+		struct {
+			uint32_t dw_offset:16;
+			uint32_t num_dwords:16;
+		} bitfields6c;
+		uint32_t reserved14;
+		uint32_t ordinal6;
+	};
+
+	union {
+		uint32_t data_hi;
+		uint32_t cmp_data_hi;
+		uint32_t reserved15;
+		uint32_t reserved16;
+		uint32_t ordinal7;
+	};
+
+	uint32_t int_ctxid;
+
+};
+
+#endif
+
+enum {
+	CACHE_FLUSH_AND_INV_TS_EVENT = 0x00000014
+};
+#endif
+
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index 96a9cc0f02c9..5e3990bb4c4b 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -39,11 +39,37 @@
 
 #include "amd_shared.h"
 
+#define KFD_MAX_RING_ENTRY_SIZE	8
+
 #define KFD_SYSFS_FILE_MODE 0444
 
-#define KFD_MMAP_DOORBELL_MASK 0x8000000000000ull
-#define KFD_MMAP_EVENTS_MASK 0x4000000000000ull
-#define KFD_MMAP_RESERVED_MEM_MASK 0x2000000000000ull
+/* GPU ID hash width in bits */
+#define KFD_GPU_ID_HASH_WIDTH 16
+
+/* Use upper bits of mmap offset to store KFD driver specific information.
+ * BITS[63:62] - Encode MMAP type
+ * BITS[61:46] - Encode gpu_id. To identify to which GPU the offset belongs to
+ * BITS[45:0]  - MMAP offset value
+ *
+ * NOTE: struct vm_area_struct.vm_pgoff uses offset in pages. Hence, these
+ *  defines are w.r.t to PAGE_SIZE
+ */
+#define KFD_MMAP_TYPE_SHIFT	(62 - PAGE_SHIFT)
+#define KFD_MMAP_TYPE_MASK	(0x3ULL << KFD_MMAP_TYPE_SHIFT)
+#define KFD_MMAP_TYPE_DOORBELL	(0x3ULL << KFD_MMAP_TYPE_SHIFT)
+#define KFD_MMAP_TYPE_EVENTS	(0x2ULL << KFD_MMAP_TYPE_SHIFT)
+#define KFD_MMAP_TYPE_RESERVED_MEM	(0x1ULL << KFD_MMAP_TYPE_SHIFT)
+
+#define KFD_MMAP_GPU_ID_SHIFT (46 - PAGE_SHIFT)
+#define KFD_MMAP_GPU_ID_MASK (((1ULL << KFD_GPU_ID_HASH_WIDTH) - 1) \
+				<< KFD_MMAP_GPU_ID_SHIFT)
+#define KFD_MMAP_GPU_ID(gpu_id) ((((uint64_t)gpu_id) << KFD_MMAP_GPU_ID_SHIFT)\
+				& KFD_MMAP_GPU_ID_MASK)
+#define KFD_MMAP_GPU_ID_GET(offset)    ((offset & KFD_MMAP_GPU_ID_MASK) \
+				>> KFD_MMAP_GPU_ID_SHIFT)
+
+#define KFD_MMAP_OFFSET_VALUE_MASK	(0x3FFFFFFFFFFFULL >> PAGE_SHIFT)
+#define KFD_MMAP_OFFSET_VALUE_GET(offset) (offset & KFD_MMAP_OFFSET_VALUE_MASK)
 
 /*
  * When working with cp scheduler we should assign the HIQ manually or via
@@ -55,9 +81,6 @@
 #define KFD_CIK_HIQ_PIPE 4
 #define KFD_CIK_HIQ_QUEUE 0
 
-/* GPU ID hash width in bits */
-#define KFD_GPU_ID_HASH_WIDTH 16
-
 /* Macro for allocating structures */
 #define kfd_alloc_struct(ptr_to_struct)	\
 	((typeof(ptr_to_struct)) kzalloc(sizeof(*ptr_to_struct), GFP_KERNEL))
@@ -116,6 +139,11 @@ extern int debug_largebar;
  */
 extern int ignore_crat;
 
+/*
+ * Set sh_mem_config.retry_disable on Vega10
+ */
+extern int vega10_noretry;
+
 /**
  * enum kfd_sched_policy
  *
@@ -148,6 +176,8 @@ enum cache_policy {
 	cache_policy_noncoherent
 };
 
+#define KFD_IS_SOC15(chip) ((chip) >= CHIP_VEGA10)
+
 struct kfd_event_interrupt_class {
 	bool (*interrupt_isr)(struct kfd_dev *dev,
 				const uint32_t *ih_ring_entry);
@@ -160,6 +190,7 @@ struct kfd_device_info {
 	const struct kfd_event_interrupt_class *event_interrupt_class;
 	unsigned int max_pasid_bits;
 	unsigned int max_no_of_hqd;
+	unsigned int doorbell_size;
 	size_t ih_ring_entry_size;
 	uint8_t num_of_watch_points;
 	uint16_t mqd_size_aligned;
@@ -173,6 +204,7 @@ struct kfd_mem_obj {
 	uint32_t range_end;
 	uint64_t gpu_addr;
 	uint32_t *cpu_ptr;
+	void *gtt_mem;
 };
 
 struct kfd_vmid_info {
@@ -364,7 +396,7 @@ struct queue_properties {
 	uint32_t queue_percent;
 	uint32_t *read_ptr;
 	uint32_t *write_ptr;
-	uint32_t __iomem *doorbell_ptr;
+	void __iomem *doorbell_ptr;
 	uint32_t doorbell_off;
 	bool is_interop;
 	bool is_evicted;
@@ -427,6 +459,7 @@ struct queue {
 	uint32_t queue;
 
 	unsigned int sdma_id;
+	unsigned int doorbell_id;
 
 	struct kfd_process	*process;
 	struct kfd_dev		*device;
@@ -501,6 +534,9 @@ struct qcm_process_device {
 	/* IB memory */
 	uint64_t ib_base;
 	void *ib_kaddr;
+
+	/* doorbell resources per process per device */
+	unsigned long *doorbell_bitmap;
 };
 
 /* KFD Memory Eviction */
@@ -512,6 +548,8 @@ struct qcm_process_device {
 /* Approx. time before evicting the process again */
 #define PROCESS_ACTIVE_TIME_MS 10
 
+int kgd2kfd_quiesce_mm(struct mm_struct *mm);
+int kgd2kfd_resume_mm(struct mm_struct *mm);
 int kgd2kfd_schedule_evict_and_restore_process(struct mm_struct *mm,
 					       struct dma_fence *fence);
 
@@ -681,6 +719,8 @@ struct kfd_process *kfd_get_process(const struct task_struct *);
 struct kfd_process *kfd_lookup_process_by_pasid(unsigned int pasid);
 struct kfd_process *kfd_lookup_process_by_mm(const struct mm_struct *mm);
 void kfd_unref_process(struct kfd_process *p);
+int kfd_process_evict_queues(struct kfd_process *p);
+int kfd_process_restore_queues(struct kfd_process *p);
 void kfd_suspend_all_processes(void);
 int kfd_resume_all_processes(void);
 
@@ -693,7 +733,7 @@ struct kfd_process_device *kfd_get_process_device_data(struct kfd_dev *dev,
 struct kfd_process_device *kfd_create_process_device_data(struct kfd_dev *dev,
 							struct kfd_process *p);
 
-int kfd_reserved_mem_mmap(struct kfd_process *process,
+int kfd_reserved_mem_mmap(struct kfd_dev *dev, struct kfd_process *process,
 			  struct vm_area_struct *vma);
 
 /* KFD process API for creating and translating handles */
@@ -721,17 +761,20 @@ unsigned int kfd_pasid_alloc(void);
 void kfd_pasid_free(unsigned int pasid);
 
 /* Doorbells */
+size_t kfd_doorbell_process_slice(struct kfd_dev *kfd);
 int kfd_doorbell_init(struct kfd_dev *kfd);
 void kfd_doorbell_fini(struct kfd_dev *kfd);
-int kfd_doorbell_mmap(struct kfd_process *process, struct vm_area_struct *vma);
-u32 __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
+int kfd_doorbell_mmap(struct kfd_dev *dev, struct kfd_process *process,
+		      struct vm_area_struct *vma);
+void __iomem *kfd_get_kernel_doorbell(struct kfd_dev *kfd,
 					unsigned int *doorbell_off);
 void kfd_release_kernel_doorbell(struct kfd_dev *kfd, u32 __iomem *db_addr);
 u32 read_kernel_doorbell(u32 __iomem *db);
-void write_kernel_doorbell(u32 __iomem *db, u32 value);
-unsigned int kfd_queue_id_to_doorbell(struct kfd_dev *kfd,
+void write_kernel_doorbell(void __iomem *db, u32 value);
+void write_kernel_doorbell64(void __iomem *db, u64 value);
+unsigned int kfd_doorbell_id_to_offset(struct kfd_dev *kfd,
 					struct kfd_process *process,
-					unsigned int queue_id);
+					unsigned int doorbell_id);
 phys_addr_t kfd_get_process_doorbells(struct kfd_dev *dev,
 					struct kfd_process *process);
 int kfd_alloc_process_doorbells(struct kfd_process *process);
@@ -788,6 +831,8 @@ struct mqd_manager *mqd_manager_init_vi(enum KFD_MQD_TYPE type,
 		struct kfd_dev *dev);
 struct mqd_manager *mqd_manager_init_vi_tonga(enum KFD_MQD_TYPE type,
 		struct kfd_dev *dev);
+struct mqd_manager *mqd_manager_init_v9(enum KFD_MQD_TYPE type,
+		struct kfd_dev *dev);
 struct device_queue_manager *device_queue_manager_init(struct kfd_dev *dev);
 void device_queue_manager_uninit(struct device_queue_manager *dqm);
 struct kernel_queue *kernel_queue_init(struct kfd_dev *dev,
@@ -832,8 +877,42 @@ struct packet_manager {
 	bool allocated;
 	struct kfd_mem_obj *ib_buffer_obj;
 	unsigned int ib_size_bytes;
+
+	const struct packet_manager_funcs *pmf;
+};
+
+struct packet_manager_funcs {
+	/* Support ASIC-specific packet formats for PM4 packets */
+	int (*map_process)(struct packet_manager *pm, uint32_t *buffer,
+			struct qcm_process_device *qpd);
+	int (*runlist)(struct packet_manager *pm, uint32_t *buffer,
+			uint64_t ib, size_t ib_size_in_dwords, bool chain);
+	int (*set_resources)(struct packet_manager *pm, uint32_t *buffer,
+			struct scheduling_resources *res);
+	int (*map_queues)(struct packet_manager *pm, uint32_t *buffer,
+			struct queue *q, bool is_static);
+	int (*unmap_queues)(struct packet_manager *pm, uint32_t *buffer,
+			enum kfd_queue_type type,
+			enum kfd_unmap_queues_filter mode,
+			uint32_t filter_param, bool reset,
+			unsigned int sdma_engine);
+	int (*query_status)(struct packet_manager *pm, uint32_t *buffer,
+			uint64_t fence_address,	uint32_t fence_value);
+	int (*release_mem)(uint64_t gpu_addr, uint32_t *buffer);
+
+	/* Packet sizes */
+	int map_process_size;
+	int runlist_size;
+	int set_resources_size;
+	int map_queues_size;
+	int unmap_queues_size;
+	int query_status_size;
+	int release_mem_size;
 };
 
+extern const struct packet_manager_funcs kfd_vi_pm_funcs;
+extern const struct packet_manager_funcs kfd_v9_pm_funcs;
+
 int pm_init(struct packet_manager *pm, struct device_queue_manager *dqm);
 void pm_uninit(struct packet_manager *pm);
 int pm_send_set_resources(struct packet_manager *pm,
@@ -849,12 +928,17 @@ int pm_send_unmap_queue(struct packet_manager *pm, enum kfd_queue_type type,
 
 void pm_release_ib(struct packet_manager *pm);
 
-uint32_t pm_create_release_mem(uint64_t gpu_addr, uint32_t *buffer);
+/* Following PM funcs can be shared among VI and AI */
+unsigned int pm_build_pm4_header(unsigned int opcode, size_t packet_size);
+int pm_set_resources_vi(struct packet_manager *pm, uint32_t *buffer,
+				struct scheduling_resources *res);
 
 uint64_t kfd_get_number_elems(struct kfd_dev *kfd);
 
 /* Events */
 extern const struct kfd_event_interrupt_class event_interrupt_class_cik;
+extern const struct kfd_event_interrupt_class event_interrupt_class_v9;
+
 extern const struct kfd_device_global_init_class device_global_init_class_cik;
 
 void kfd_event_init_process(struct kfd_process *p);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
index 1711ad0642f7..1d80b4f7c681 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
@@ -332,6 +332,7 @@ static void kfd_process_destroy_pdds(struct kfd_process *p)
 			free_pages((unsigned long)pdd->qpd.cwsr_kaddr,
 				get_order(KFD_CWSR_TBA_TMA_SIZE));
 
+		kfree(pdd->qpd.doorbell_bitmap);
 		idr_destroy(&pdd->alloc_idr);
 
 		kfree(pdd);
@@ -451,7 +452,8 @@ static int kfd_process_init_cwsr_apu(struct kfd_process *p, struct file *filep)
 		if (!dev->cwsr_enabled || qpd->cwsr_kaddr || qpd->cwsr_base)
 			continue;
 
-		offset = (dev->id | KFD_MMAP_RESERVED_MEM_MASK) << PAGE_SHIFT;
+		offset = (KFD_MMAP_TYPE_RESERVED_MEM | KFD_MMAP_GPU_ID(dev->id))
+			<< PAGE_SHIFT;
 		qpd->tba_addr = (int64_t)vm_mmap(filep, 0,
 			KFD_CWSR_TBA_TMA_SIZE, PROT_READ | PROT_EXEC,
 			MAP_SHARED, offset);
@@ -585,6 +587,31 @@ err_alloc_process:
 	return ERR_PTR(err);
 }
 
+static int init_doorbell_bitmap(struct qcm_process_device *qpd,
+			struct kfd_dev *dev)
+{
+	unsigned int i;
+
+	if (!KFD_IS_SOC15(dev->device_info->asic_family))
+		return 0;
+
+	qpd->doorbell_bitmap =
+		kzalloc(DIV_ROUND_UP(KFD_MAX_NUM_OF_QUEUES_PER_PROCESS,
+				     BITS_PER_BYTE), GFP_KERNEL);
+	if (!qpd->doorbell_bitmap)
+		return -ENOMEM;
+
+	/* Mask out any reserved doorbells */
+	for (i = 0; i < KFD_MAX_NUM_OF_QUEUES_PER_PROCESS; i++)
+		if ((dev->shared_resources.reserved_doorbell_mask & i) ==
+		    dev->shared_resources.reserved_doorbell_val) {
+			set_bit(i, qpd->doorbell_bitmap);
+			pr_debug("reserved doorbell 0x%03x\n", i);
+		}
+
+	return 0;
+}
+
 struct kfd_process_device *kfd_get_process_device_data(struct kfd_dev *dev,
 							struct kfd_process *p)
 {
@@ -606,6 +633,12 @@ struct kfd_process_device *kfd_create_process_device_data(struct kfd_dev *dev,
 	if (!pdd)
 		return NULL;
 
+	if (init_doorbell_bitmap(&pdd->qpd, dev)) {
+		pr_err("Failed to init doorbell for process\n");
+		kfree(pdd);
+		return NULL;
+	}
+
 	pdd->dev = dev;
 	INIT_LIST_HEAD(&pdd->qpd.queues_list);
 	INIT_LIST_HEAD(&pdd->qpd.priv_queue_list);
@@ -808,7 +841,7 @@ struct kfd_process *kfd_lookup_process_by_mm(const struct mm_struct *mm)
  * Eviction is reference-counted per process-device. This means multiple
  * evictions from different sources can be nested safely.
  */
-static int process_evict_queues(struct kfd_process *p)
+int kfd_process_evict_queues(struct kfd_process *p)
 {
 	struct kfd_process_device *pdd;
 	int r = 0;
@@ -844,7 +877,7 @@ fail:
 }
 
 /* process_restore_queues - Restore all user queues of a process */
-static  int process_restore_queues(struct kfd_process *p)
+int kfd_process_restore_queues(struct kfd_process *p)
 {
 	struct kfd_process_device *pdd;
 	int r, ret = 0;
@@ -886,7 +919,7 @@ static void evict_process_worker(struct work_struct *work)
 	flush_delayed_work(&p->restore_work);
 
 	pr_debug("Started evicting pasid %d\n", p->pasid);
-	ret = process_evict_queues(p);
+	ret = kfd_process_evict_queues(p);
 	if (!ret) {
 		dma_fence_signal(p->ef);
 		dma_fence_put(p->ef);
@@ -946,7 +979,7 @@ static void restore_process_worker(struct work_struct *work)
 		return;
 	}
 
-	ret = process_restore_queues(p);
+	ret = kfd_process_restore_queues(p);
 	if (!ret)
 		pr_debug("Finished restoring pasid %d\n", p->pasid);
 	else
@@ -963,7 +996,7 @@ void kfd_suspend_all_processes(void)
 		cancel_delayed_work_sync(&p->eviction_work);
 		cancel_delayed_work_sync(&p->restore_work);
 
-		if (process_evict_queues(p))
+		if (kfd_process_evict_queues(p))
 			pr_err("Failed to suspend process %d\n", p->pasid);
 		dma_fence_signal(p->ef);
 		dma_fence_put(p->ef);
@@ -989,15 +1022,12 @@ int kfd_resume_all_processes(void)
 	return ret;
 }
 
-int kfd_reserved_mem_mmap(struct kfd_process *process,
+int kfd_reserved_mem_mmap(struct kfd_dev *dev, struct kfd_process *process,
 			  struct vm_area_struct *vma)
 {
-	struct kfd_dev *dev = kfd_device_by_id(vma->vm_pgoff);
 	struct kfd_process_device *pdd;
 	struct qcm_process_device *qpd;
 
-	if (!dev)
-		return -EINVAL;
 	if ((vma->vm_end - vma->vm_start) != KFD_CWSR_TBA_TMA_SIZE) {
 		pr_err("Incorrect CWSR mapping size.\n");
 		return -EINVAL;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
index 7817e327ea6d..d65ce0436b31 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c
@@ -119,9 +119,6 @@ static int create_cp_queue(struct process_queue_manager *pqm,
 	/* Doorbell initialized in user space*/
 	q_properties->doorbell_ptr = NULL;
 
-	q_properties->doorbell_off =
-			kfd_queue_id_to_doorbell(dev, pqm->process, qid);
-
 	/* let DQM handle it*/
 	q_properties->vmid = 0;
 	q_properties->queue_id = qid;
@@ -244,10 +241,20 @@ int pqm_create_queue(struct process_queue_manager *pqm,
 	}
 
 	if (retval != 0) {
-		pr_err("DQM create queue failed\n");
+		pr_err("Pasid %d DQM create queue %d failed. ret %d\n",
+			pqm->process->pasid, type, retval);
 		goto err_create_queue;
 	}
 
+	if (q)
+		/* Return the doorbell offset within the doorbell page
+		 * to the caller so it can be passed up to user mode
+		 * (in bytes).
+		 */
+		properties->doorbell_off =
+			(q->properties.doorbell_off * sizeof(uint32_t)) &
+			(kfd_doorbell_process_slice(dev) - 1);
+
 	pr_debug("PQM After DQM create queue\n");
 
 	list_add(&pqn->process_queue_list, &pqm->queues);
@@ -313,8 +320,11 @@ int pqm_destroy_queue(struct process_queue_manager *pqm, unsigned int qid)
 		dqm = pqn->q->device->dqm;
 		retval = dqm->ops.destroy_queue(dqm, &pdd->qpd, pqn->q);
 		if (retval) {
-			pr_debug("Destroy queue failed, returned %d\n", retval);
-			goto err_destroy_queue;
+			pr_err("Pasid %d destroy queue %d failed, ret %d\n",
+				pqm->process->pasid,
+				pqn->q->properties.queue_id, retval);
+			if (retval != -ETIME)
+				goto err_destroy_queue;
 		}
 		uninit_queue(pqn->q);
 	}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_queue.c
index a5315d4f1c95..6dcd621e5b71 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_queue.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_queue.c
@@ -36,8 +36,8 @@ void print_queue_properties(struct queue_properties *q)
 	pr_debug("Queue Address: 0x%llX\n", q->queue_address);
 	pr_debug("Queue Id: %u\n", q->queue_id);
 	pr_debug("Queue Process Vmid: %u\n", q->vmid);
-	pr_debug("Queue Read Pointer: 0x%p\n", q->read_ptr);
-	pr_debug("Queue Write Pointer: 0x%p\n", q->write_ptr);
+	pr_debug("Queue Read Pointer: 0x%px\n", q->read_ptr);
+	pr_debug("Queue Write Pointer: 0x%px\n", q->write_ptr);
 	pr_debug("Queue Doorbell Pointer: 0x%p\n", q->doorbell_ptr);
 	pr_debug("Queue Doorbell Offset: %u\n", q->doorbell_off);
 }
@@ -53,8 +53,8 @@ void print_queue(struct queue *q)
 	pr_debug("Queue Address: 0x%llX\n", q->properties.queue_address);
 	pr_debug("Queue Id: %u\n", q->properties.queue_id);
 	pr_debug("Queue Process Vmid: %u\n", q->properties.vmid);
-	pr_debug("Queue Read Pointer: 0x%p\n", q->properties.read_ptr);
-	pr_debug("Queue Write Pointer: 0x%p\n", q->properties.write_ptr);
+	pr_debug("Queue Read Pointer: 0x%px\n", q->properties.read_ptr);
+	pr_debug("Queue Write Pointer: 0x%px\n", q->properties.write_ptr);
 	pr_debug("Queue Doorbell Pointer: 0x%p\n", q->properties.doorbell_ptr);
 	pr_debug("Queue Doorbell Offset: %u\n", q->properties.doorbell_off);
 	pr_debug("Queue MQD Address: 0x%p\n", q->mqd);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
index ac28abc94e57..bc95d4dfee2e 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
@@ -1239,6 +1239,12 @@ int kfd_topology_add_device(struct kfd_dev *gpu)
 			HSA_CAP_DOORBELL_TYPE_TOTALBITS_SHIFT) &
 			HSA_CAP_DOORBELL_TYPE_TOTALBITS_MASK);
 		break;
+	case CHIP_VEGA10:
+	case CHIP_RAVEN:
+		dev->node_props.capability |= ((HSA_CAP_DOORBELL_TYPE_2_0 <<
+			HSA_CAP_DOORBELL_TYPE_TOTALBITS_SHIFT) &
+			HSA_CAP_DOORBELL_TYPE_TOTALBITS_MASK);
+		break;
 	default:
 		WARN(1, "Unexpected ASIC family %u",
 		     dev->gpu->device_info->asic_family);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
index eb54cfcaf039..7d9c3f948dff 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
@@ -45,6 +45,7 @@
 
 #define HSA_CAP_DOORBELL_TYPE_PRE_1_0		0x0
 #define HSA_CAP_DOORBELL_TYPE_1_0		0x1
+#define HSA_CAP_DOORBELL_TYPE_2_0		0x2
 #define HSA_CAP_AQL_QUEUE_DOUBLE_MAP		0x00004000
 
 struct kfd_node_properties {
diff --git a/drivers/gpu/drm/amd/amdkfd/soc15_int.h b/drivers/gpu/drm/amd/amdkfd/soc15_int.h
new file mode 100644
index 000000000000..0bc0b25cb410
--- /dev/null
+++ b/drivers/gpu/drm/amd/amdkfd/soc15_int.h
@@ -0,0 +1,47 @@
+/*
+ * Copyright 2016-2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+
+#ifndef HSA_SOC15_INT_H_INCLUDED
+#define HSA_SOC15_INT_H_INCLUDED
+
+#include "soc15_ih_clientid.h"
+
+#define SOC15_INTSRC_CP_END_OF_PIPE	181
+#define SOC15_INTSRC_CP_BAD_OPCODE	183
+#define SOC15_INTSRC_SQ_INTERRUPT_MSG	239
+#define SOC15_INTSRC_VMC_FAULT		0
+#define SOC15_INTSRC_SDMA_TRAP		224
+
+
+#define SOC15_CLIENT_ID_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[0]) & 0xff)
+#define SOC15_SOURCE_ID_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[0]) >> 8 & 0xff)
+#define SOC15_RING_ID_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[0]) >> 16 & 0xff)
+#define SOC15_VMID_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[0]) >> 24 & 0xf)
+#define SOC15_VMID_TYPE_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[0]) >> 31 & 0x1)
+#define SOC15_PASID_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[3]) & 0xffff)
+#define SOC15_CONTEXT_ID0_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[4]))
+#define SOC15_CONTEXT_ID1_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[5]))
+#define SOC15_CONTEXT_ID2_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[6]))
+#define SOC15_CONTEXT_ID3_FROM_IH_ENTRY(entry) (le32_to_cpu(entry[7]))
+
+#endif
+
diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig
index 5b124a67404c..325083b0297e 100644
--- a/drivers/gpu/drm/amd/display/Kconfig
+++ b/drivers/gpu/drm/amd/display/Kconfig
@@ -9,31 +9,6 @@ config DRM_AMD_DC
 	  support for AMDGPU. This adds required support for Vega and
 	  Raven ASICs.
 
-config DRM_AMD_DC_PRE_VEGA
-	bool "DC support for Polaris and older ASICs"
-	default y
-	help
-	  Choose this option to enable the new DC support for older asics
-	  by default. This includes Polaris, Carrizo, Tonga, Bonaire,
-	  and Hawaii.
-
-config DRM_AMD_DC_FBC
-	bool "AMD FBC - Enable Frame Buffer Compression"
-	depends on DRM_AMD_DC
-	help
-	  Choose this option if you want to use frame buffer compression
-	  support.
-	  This is a power optimisation feature, check its availability
-	  on your hardware before enabling this option.
-
-
-config DRM_AMD_DC_DCN1_0
-	bool "DCN 1.0 Raven family"
-	depends on DRM_AMD_DC && X86
-	help
-	  Choose this option if you want to have
-	  RV family for display engine
-
 config DEBUG_KERNEL_DC
 	bool "Enable kgdb break in DC"
 	depends on DRM_AMD_DC
diff --git a/drivers/gpu/drm/amd/display/TODO b/drivers/gpu/drm/amd/display/TODO
index 357d59648401..a8a6c106e8c7 100644
--- a/drivers/gpu/drm/amd/display/TODO
+++ b/drivers/gpu/drm/amd/display/TODO
@@ -97,10 +97,10 @@ share it with drivers. But that's a very long term goal, and by far not just an
 issue with DC - other drivers, especially around DP sink handling, are equally
 guilty.
 
-19. The DC logger is still a rather sore thing, but I know that the DRM_DEBUG
-stuff just isn't up to the challenges either. We need to figure out something
-that integrates better with DRM and linux debug printing, while not being
-useless with filtering output. dynamic debug printing might be an option.
+19. DONE - The DC logger is still a rather sore thing, but I know that the
+DRM_DEBUG stuff just isn't up to the challenges either. We need to figure out
+something that integrates better with DRM and linux debug printing, while not
+being useless with filtering output. dynamic debug printing might be an option.
 
 20. Use kernel i2c device to program HDMI retimer. Some boards have an HDMI
 retimer that we need to program to pass PHY compliance. Currently that's
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
index af16973f2c41..94911871eb9b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
@@ -28,11 +28,11 @@
 AMDGPUDM = amdgpu_dm.o amdgpu_dm_irq.o amdgpu_dm_mst_types.o amdgpu_dm_color.o
 
 ifneq ($(CONFIG_DRM_AMD_DC),)
-AMDGPUDM += amdgpu_dm_services.o amdgpu_dm_helpers.o
+AMDGPUDM += amdgpu_dm_services.o amdgpu_dm_helpers.o amdgpu_dm_pp_smu.o
 endif
 
 ifneq ($(CONFIG_DEBUG_FS),)
-AMDGPUDM += amdgpu_dm_crc.o
+AMDGPUDM += amdgpu_dm_crc.o amdgpu_dm_debugfs.o
 endif
 
 subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/dc
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index 1dd1142246c2..5fc13e71a3b5 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -39,6 +39,9 @@
 #include "dm_helpers.h"
 #include "dm_services_types.h"
 #include "amdgpu_dm_mst_types.h"
+#if defined(CONFIG_DEBUG_FS)
+#include "amdgpu_dm_debugfs.h"
+#endif
 
 #include "ivsrcid/ivsrcid_vislands30.h"
 
@@ -46,6 +49,7 @@
 #include <linux/moduleparam.h>
 #include <linux/version.h>
 #include <linux/types.h>
+#include <linux/pm_runtime.h>
 
 #include <drm/drmP.h>
 #include <drm/drm_atomic.h>
@@ -56,7 +60,7 @@
 
 #include "modules/inc/mod_freesync.h"
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 #include "ivsrcid/irqsrcs_dcn_1_0.h"
 
 #include "dcn/dcn_1_0_offset.h"
@@ -346,7 +350,6 @@ static void hotplug_notify_work_func(struct work_struct *work)
 	drm_kms_helper_hotplug_event(dev);
 }
 
-#if defined(CONFIG_DRM_AMD_DC_FBC)
 /* Allocate memory for FBC compressed data  */
 static void amdgpu_dm_fbc_init(struct drm_connector *connector)
 {
@@ -387,7 +390,6 @@ static void amdgpu_dm_fbc_init(struct drm_connector *connector)
 	}
 
 }
-#endif
 
 
 /* Init display KMS
@@ -433,11 +435,6 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
 
 	init_data.dce_environment = DCE_ENV_PRODUCTION_DRV;
 
-	if (amdgpu_dc_log)
-		init_data.log_mask = DC_DEFAULT_LOG_MASK;
-	else
-		init_data.log_mask = DC_MIN_LOG_MASK;
-
 	/*
 	 * TODO debug why this doesn't work on Raven
 	 */
@@ -649,18 +646,6 @@ amdgpu_dm_find_first_crtc_matching_connector(struct drm_atomic_state *state,
 static int dm_resume(void *handle)
 {
 	struct amdgpu_device *adev = handle;
-	struct amdgpu_display_manager *dm = &adev->dm;
-	int ret = 0;
-
-	/* power on hardware */
-	dc_set_power_state(dm->dc, DC_ACPI_CM_POWER_STATE_D0);
-
-	ret = amdgpu_dm_display_resume(adev);
-	return ret;
-}
-
-int amdgpu_dm_display_resume(struct amdgpu_device *adev)
-{
 	struct drm_device *ddev = adev->ddev;
 	struct amdgpu_display_manager *dm = &adev->dm;
 	struct amdgpu_dm_connector *aconnector;
@@ -671,10 +656,12 @@ int amdgpu_dm_display_resume(struct amdgpu_device *adev)
 	struct drm_plane *plane;
 	struct drm_plane_state *new_plane_state;
 	struct dm_plane_state *dm_new_plane_state;
-
-	int ret = 0;
+	int ret;
 	int i;
 
+	/* power on hardware */
+	dc_set_power_state(dm->dc, DC_ACPI_CM_POWER_STATE_D0);
+
 	/* program HPD filter */
 	dc_resume(dm->dc);
 
@@ -688,8 +675,7 @@ int amdgpu_dm_display_resume(struct amdgpu_device *adev)
 	amdgpu_dm_irq_resume_early(adev);
 
 	/* Do detection*/
-	list_for_each_entry(connector,
-			&ddev->mode_config.connector_list, head) {
+	list_for_each_entry(connector, &ddev->mode_config.connector_list, head) {
 		aconnector = to_amdgpu_dm_connector(connector);
 
 		/*
@@ -711,7 +697,7 @@ int amdgpu_dm_display_resume(struct amdgpu_device *adev)
 	}
 
 	/* Force mode set in atomic comit */
-	for_each_new_crtc_in_state(adev->dm.cached_state, crtc, new_crtc_state, i)
+	for_each_new_crtc_in_state(dm->cached_state, crtc, new_crtc_state, i)
 		new_crtc_state->active_changed = true;
 
 	/*
@@ -719,7 +705,7 @@ int amdgpu_dm_display_resume(struct amdgpu_device *adev)
 	 * them here, since they were duplicated as part of the suspend
 	 * procedure.
 	 */
-	for_each_new_crtc_in_state(adev->dm.cached_state, crtc, new_crtc_state, i) {
+	for_each_new_crtc_in_state(dm->cached_state, crtc, new_crtc_state, i) {
 		dm_new_crtc_state = to_dm_crtc_state(new_crtc_state);
 		if (dm_new_crtc_state->stream) {
 			WARN_ON(kref_read(&dm_new_crtc_state->stream->refcount) > 1);
@@ -728,7 +714,7 @@ int amdgpu_dm_display_resume(struct amdgpu_device *adev)
 		}
 	}
 
-	for_each_new_plane_in_state(adev->dm.cached_state, plane, new_plane_state, i) {
+	for_each_new_plane_in_state(dm->cached_state, plane, new_plane_state, i) {
 		dm_new_plane_state = to_dm_plane_state(new_plane_state);
 		if (dm_new_plane_state->dc_state) {
 			WARN_ON(kref_read(&dm_new_plane_state->dc_state->refcount) > 1);
@@ -737,9 +723,9 @@ int amdgpu_dm_display_resume(struct amdgpu_device *adev)
 		}
 	}
 
-	ret = drm_atomic_helper_resume(ddev, adev->dm.cached_state);
+	ret = drm_atomic_helper_resume(ddev, dm->cached_state);
 
-	adev->dm.cached_state = NULL;
+	dm->cached_state = NULL;
 
 	amdgpu_dm_irq_resume_late(adev);
 
@@ -917,16 +903,17 @@ amdgpu_dm_update_connector_after_detect(struct amdgpu_dm_connector *aconnector)
 				(struct edid *) sink->dc_edid.raw_edid;
 
 
-			drm_mode_connector_update_edid_property(connector,
+			drm_connector_update_edid_property(connector,
 					aconnector->edid);
 		}
 		amdgpu_dm_add_sink_to_freesync_module(connector, aconnector->edid);
 
 	} else {
 		amdgpu_dm_remove_sink_from_freesync_module(connector);
-		drm_mode_connector_update_edid_property(connector, NULL);
+		drm_connector_update_edid_property(connector, NULL);
 		aconnector->num_modes = 0;
 		aconnector->dc_sink = NULL;
+		aconnector->edid = NULL;
 	}
 
 	mutex_unlock(&dev->mode_config.mutex);
@@ -1054,7 +1041,7 @@ static void handle_hpd_rx_irq(void *param)
 	if (dc_link->type != dc_connection_mst_branch)
 		mutex_lock(&aconnector->hpd_lock);
 
-	if (dc_link_handle_hpd_rx_irq(dc_link, NULL) &&
+	if (dc_link_handle_hpd_rx_irq(dc_link, NULL, NULL) &&
 			!is_mst_root_connector) {
 		/* Downstream Port status changed. */
 		if (dc_link_detect(dc_link, DETECT_REASON_HPDRX)) {
@@ -1131,6 +1118,7 @@ static int dce110_register_irq_handlers(struct amdgpu_device *adev)
 
 	if (adev->asic_type == CHIP_VEGA10 ||
 	    adev->asic_type == CHIP_VEGA12 ||
+	    adev->asic_type == CHIP_VEGA20 ||
 	    adev->asic_type == CHIP_RAVEN)
 		client_id = SOC15_IH_CLIENTID_DCE;
 
@@ -1204,7 +1192,7 @@ static int dce110_register_irq_handlers(struct amdgpu_device *adev)
 	return 0;
 }
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 /* Register IRQ sources and initialize IRQ callbacks */
 static int dcn10_register_irq_handlers(struct amdgpu_device *adev)
 {
@@ -1529,14 +1517,16 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
 	case CHIP_POLARIS11:
 	case CHIP_POLARIS10:
 	case CHIP_POLARIS12:
+	case CHIP_VEGAM:
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		if (dce110_register_irq_handlers(dm->adev)) {
 			DRM_ERROR("DM: Failed to initialize IRQ\n");
 			goto fail;
 		}
 		break;
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	case CHIP_RAVEN:
 		if (dcn10_register_irq_handlers(dm->adev)) {
 			DRM_ERROR("DM: Failed to initialize IRQ\n");
@@ -1545,11 +1535,11 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
 		/*
 		 * Temporary disable until pplib/smu interaction is implemented
 		 */
-		dm->dc->debug.disable_stutter = true;
+		dm->dc->debug.disable_stutter = amdgpu_pp_feature_mask & PP_STUTTER_MODE ? false : true;
 		break;
 #endif
 	default:
-		DRM_ERROR("Usupported ASIC type: 0x%X\n", adev->asic_type);
+		DRM_ERROR("Unsupported ASIC type: 0x%X\n", adev->asic_type);
 		goto fail;
 	}
 
@@ -1657,7 +1647,6 @@ static ssize_t s3_debug_store(struct device *device,
 	if (ret == 0) {
 		if (s3_state) {
 			dm_resume(adev);
-			amdgpu_dm_display_resume(adev);
 			drm_kms_helper_hotplug_event(adev->ddev);
 		} else
 			dm_suspend(adev);
@@ -1722,6 +1711,7 @@ static int dm_early_init(void *handle)
 		adev->mode_info.plane_type = dm_plane_type_default;
 		break;
 	case CHIP_POLARIS10:
+	case CHIP_VEGAM:
 		adev->mode_info.num_crtc = 6;
 		adev->mode_info.num_hpd = 6;
 		adev->mode_info.num_dig = 6;
@@ -1729,12 +1719,13 @@ static int dm_early_init(void *handle)
 		break;
 	case CHIP_VEGA10:
 	case CHIP_VEGA12:
+	case CHIP_VEGA20:
 		adev->mode_info.num_crtc = 6;
 		adev->mode_info.num_hpd = 6;
 		adev->mode_info.num_dig = 6;
 		adev->mode_info.plane_type = dm_plane_type_default;
 		break;
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	case CHIP_RAVEN:
 		adev->mode_info.num_crtc = 4;
 		adev->mode_info.num_hpd = 4;
@@ -1743,7 +1734,7 @@ static int dm_early_init(void *handle)
 		break;
 #endif
 	default:
-		DRM_ERROR("Usupported ASIC type: 0x%X\n", adev->asic_type);
+		DRM_ERROR("Unsupported ASIC type: 0x%X\n", adev->asic_type);
 		return -EINVAL;
 	}
 
@@ -1848,7 +1839,7 @@ static bool fill_rects_from_plane_state(const struct drm_plane_state *state,
 static int get_fb_info(const struct amdgpu_framebuffer *amdgpu_fb,
 		       uint64_t *tiling_flags)
 {
-	struct amdgpu_bo *rbo = gem_to_amdgpu_bo(amdgpu_fb->obj);
+	struct amdgpu_bo *rbo = gem_to_amdgpu_bo(amdgpu_fb->base.obj[0]);
 	int r = amdgpu_bo_reserve(rbo, false);
 
 	if (unlikely(r)) {
@@ -1977,6 +1968,7 @@ static int fill_plane_attributes_from_fb(struct amdgpu_device *adev,
 
 	if (adev->asic_type == CHIP_VEGA10 ||
 	    adev->asic_type == CHIP_VEGA12 ||
+	    adev->asic_type == CHIP_VEGA20 ||
 	    adev->asic_type == CHIP_RAVEN) {
 		/* Fill GFX9 params */
 		plane_state->tiling_info.gfx9.num_pipes =
@@ -2017,7 +2009,6 @@ static int fill_plane_attributes(struct amdgpu_device *adev,
 	const struct amdgpu_framebuffer *amdgpu_fb =
 		to_amdgpu_framebuffer(plane_state->fb);
 	const struct drm_crtc *crtc = plane_state->crtc;
-	struct dc_transfer_func *input_tf;
 	int ret = 0;
 
 	if (!fill_rects_from_plane_state(plane_state, dc_plane_state))
@@ -2031,13 +2022,6 @@ static int fill_plane_attributes(struct amdgpu_device *adev,
 	if (ret)
 		return ret;
 
-	input_tf = dc_create_transfer_func();
-
-	if (input_tf == NULL)
-		return -ENOMEM;
-
-	dc_plane_state->in_transfer_func = input_tf;
-
 	/*
 	 * Always set input transfer function, since plane state is refreshed
 	 * every time.
@@ -2113,12 +2097,6 @@ convert_color_depth_from_display_info(const struct drm_connector *connector)
 {
 	uint32_t bpc = connector->display_info.bpc;
 
-	/* Limited color depth to 8bit
-	 * TODO: Still need to handle deep color
-	 */
-	if (bpc > 8)
-		bpc = 8;
-
 	switch (bpc) {
 	case 0:
 		/* Temporary Work around, DRM don't parse color depth for
@@ -2198,6 +2176,46 @@ get_output_color_space(const struct dc_crtc_timing *dc_crtc_timing)
 	return color_space;
 }
 
+static void reduce_mode_colour_depth(struct dc_crtc_timing *timing_out)
+{
+	if (timing_out->display_color_depth <= COLOR_DEPTH_888)
+		return;
+
+	timing_out->display_color_depth--;
+}
+
+static void adjust_colour_depth_from_display_info(struct dc_crtc_timing *timing_out,
+						const struct drm_display_info *info)
+{
+	int normalized_clk;
+	if (timing_out->display_color_depth <= COLOR_DEPTH_888)
+		return;
+	do {
+		normalized_clk = timing_out->pix_clk_khz;
+		/* YCbCr 4:2:0 requires additional adjustment of 1/2 */
+		if (timing_out->pixel_encoding == PIXEL_ENCODING_YCBCR420)
+			normalized_clk /= 2;
+		/* Adjusting pix clock following on HDMI spec based on colour depth */
+		switch (timing_out->display_color_depth) {
+		case COLOR_DEPTH_101010:
+			normalized_clk = (normalized_clk * 30) / 24;
+			break;
+		case COLOR_DEPTH_121212:
+			normalized_clk = (normalized_clk * 36) / 24;
+			break;
+		case COLOR_DEPTH_161616:
+			normalized_clk = (normalized_clk * 48) / 24;
+			break;
+		default:
+			return;
+		}
+		if (normalized_clk <= info->max_tmds_clock)
+			return;
+		reduce_mode_colour_depth(timing_out);
+
+	} while (timing_out->display_color_depth > COLOR_DEPTH_888);
+
+}
 /*****************************************************************************/
 
 static void
@@ -2206,7 +2224,7 @@ fill_stream_properties_from_drm_display_mode(struct dc_stream_state *stream,
 					     const struct drm_connector *connector)
 {
 	struct dc_crtc_timing *timing_out = &stream->timing;
-	struct dc_transfer_func *tf = dc_create_transfer_func();
+	const struct drm_display_info *info = &connector->display_info;
 
 	memset(timing_out, 0, sizeof(struct dc_crtc_timing));
 
@@ -2215,8 +2233,10 @@ fill_stream_properties_from_drm_display_mode(struct dc_stream_state *stream,
 	timing_out->v_border_top = 0;
 	timing_out->v_border_bottom = 0;
 	/* TODO: un-hardcode */
-
-	if ((connector->display_info.color_formats & DRM_COLOR_FORMAT_YCRCB444)
+	if (drm_mode_is_420_only(info, mode_in)
+			&& stream->sink->sink_signal == SIGNAL_TYPE_HDMI_TYPE_A)
+		timing_out->pixel_encoding = PIXEL_ENCODING_YCBCR420;
+	else if ((connector->display_info.color_formats & DRM_COLOR_FORMAT_YCRCB444)
 			&& stream->sink->sink_signal == SIGNAL_TYPE_HDMI_TYPE_A)
 		timing_out->pixel_encoding = PIXEL_ENCODING_YCBCR444;
 	else
@@ -2250,9 +2270,10 @@ fill_stream_properties_from_drm_display_mode(struct dc_stream_state *stream,
 
 	stream->output_color_space = get_output_color_space(timing_out);
 
-	tf->type = TF_TYPE_PREDEFINED;
-	tf->tf = TRANSFER_FUNCTION_SRGB;
-	stream->out_transfer_func = tf;
+	stream->out_transfer_func->type = TF_TYPE_PREDEFINED;
+	stream->out_transfer_func->tf = TRANSFER_FUNCTION_SRGB;
+	if (stream->sink->sink_signal == SIGNAL_TYPE_HDMI_TYPE_A)
+		adjust_colour_depth_from_display_info(timing_out, info);
 }
 
 static void fill_audio_info(struct audio_info *audio_info,
@@ -2336,27 +2357,22 @@ decide_crtc_timing_for_drm_display_mode(struct drm_display_mode *drm_mode,
 	}
 }
 
-static int create_fake_sink(struct amdgpu_dm_connector *aconnector)
+static struct dc_sink *
+create_fake_sink(struct amdgpu_dm_connector *aconnector)
 {
-	struct dc_sink *sink = NULL;
 	struct dc_sink_init_data sink_init_data = { 0 };
-
+	struct dc_sink *sink = NULL;
 	sink_init_data.link = aconnector->dc_link;
 	sink_init_data.sink_signal = aconnector->dc_link->connector_signal;
 
 	sink = dc_sink_create(&sink_init_data);
 	if (!sink) {
 		DRM_ERROR("Failed to create sink!\n");
-		return -ENOMEM;
+		return NULL;
 	}
-
 	sink->sink_signal = SIGNAL_TYPE_VIRTUAL;
-	aconnector->fake_enable = true;
-
-	aconnector->dc_sink = sink;
-	aconnector->dc_link->local_sink = sink;
 
-	return 0;
+	return sink;
 }
 
 static void set_multisync_trigger_params(
@@ -2419,7 +2435,7 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 	struct dc_stream_state *stream = NULL;
 	struct drm_display_mode mode = *drm_mode;
 	bool native_mode_found = false;
-
+	struct dc_sink *sink = NULL;
 	if (aconnector == NULL) {
 		DRM_ERROR("aconnector is NULL!\n");
 		return stream;
@@ -2437,15 +2453,18 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 			return stream;
 		}
 
-		if (create_fake_sink(aconnector))
+		sink = create_fake_sink(aconnector);
+		if (!sink)
 			return stream;
+	} else {
+		sink = aconnector->dc_sink;
 	}
 
-	stream = dc_create_stream_for_sink(aconnector->dc_sink);
+	stream = dc_create_stream_for_sink(sink);
 
 	if (stream == NULL) {
 		DRM_ERROR("Failed to create stream for sink!\n");
-		return stream;
+		goto finish;
 	}
 
 	list_for_each_entry(preferred_mode, &aconnector->base.modes, head) {
@@ -2484,10 +2503,16 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 	fill_audio_info(
 		&stream->audio_info,
 		drm_connector,
-		aconnector->dc_sink);
+		sink);
 
 	update_stream_signal(stream);
 
+	if (dm_state && dm_state->freesync_capable)
+		stream->ignore_msa_timing_param = true;
+finish:
+	if (sink && sink->sink_signal == SIGNAL_TYPE_VIRTUAL)
+		dc_sink_release(sink);
+
 	return stream;
 }
 
@@ -2710,18 +2735,15 @@ static void amdgpu_dm_connector_destroy(struct drm_connector *connector)
 	const struct dc_link *link = aconnector->dc_link;
 	struct amdgpu_device *adev = connector->dev->dev_private;
 	struct amdgpu_display_manager *dm = &adev->dm;
+
 #if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) ||\
 	defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE)
 
 	if ((link->connector_signal & (SIGNAL_TYPE_EDP | SIGNAL_TYPE_LVDS)) &&
-	    link->type != dc_connection_none) {
-		amdgpu_dm_register_backlight_device(dm);
-
-		if (dm->backlight_dev) {
-			backlight_device_unregister(dm->backlight_dev);
-			dm->backlight_dev = NULL;
-		}
-
+	    link->type != dc_connection_none &&
+	    dm->backlight_dev) {
+		backlight_device_unregister(dm->backlight_dev);
+		dm->backlight_dev = NULL;
 	}
 #endif
 	drm_connector_unregister(connector);
@@ -2734,6 +2756,9 @@ void amdgpu_dm_connector_funcs_reset(struct drm_connector *connector)
 	struct dm_connector_state *state =
 		to_dm_connector_state(connector->state);
 
+	if (connector->state)
+		__drm_atomic_helper_connector_destroy_state(connector->state);
+
 	kfree(state);
 
 	state = kzalloc(sizeof(*state), GFP_KERNEL);
@@ -2744,8 +2769,7 @@ void amdgpu_dm_connector_funcs_reset(struct drm_connector *connector)
 		state->underscan_hborder = 0;
 		state->underscan_vborder = 0;
 
-		connector->state = &state->base;
-		connector->state->connector = connector;
+		__drm_atomic_helper_connector_reset(connector, &state->base);
 	}
 }
 
@@ -2855,7 +2879,7 @@ static void handle_edid_mgmt(struct amdgpu_dm_connector *aconnector)
 	create_eml_sink(aconnector);
 }
 
-int amdgpu_dm_connector_mode_valid(struct drm_connector *connector,
+enum drm_mode_status amdgpu_dm_connector_mode_valid(struct drm_connector *connector,
 				   struct drm_display_mode *mode)
 {
 	int result = MODE_ERROR;
@@ -3058,8 +3082,7 @@ static int dm_plane_helper_prepare_fb(struct drm_plane *plane,
 	}
 
 	afb = to_amdgpu_framebuffer(new_state->fb);
-
-	obj = afb->obj;
+	obj = new_state->fb->obj[0];
 	rbo = gem_to_amdgpu_bo(obj);
 	adev = amdgpu_ttm_adev(rbo->tbo.bdev);
 	r = amdgpu_bo_reserve(rbo, false);
@@ -3067,20 +3090,29 @@ static int dm_plane_helper_prepare_fb(struct drm_plane *plane,
 		return r;
 
 	if (plane->type != DRM_PLANE_TYPE_CURSOR)
-		domain = amdgpu_display_framebuffer_domains(adev);
+		domain = amdgpu_display_supported_domains(adev);
 	else
 		domain = AMDGPU_GEM_DOMAIN_VRAM;
 
-	r = amdgpu_bo_pin(rbo, domain, &afb->address);
-
-	amdgpu_bo_unreserve(rbo);
-
+	r = amdgpu_bo_pin(rbo, domain);
 	if (unlikely(r != 0)) {
 		if (r != -ERESTARTSYS)
 			DRM_ERROR("Failed to pin framebuffer with error %d\n", r);
+		amdgpu_bo_unreserve(rbo);
 		return r;
 	}
 
+	r = amdgpu_ttm_alloc_gart(&rbo->tbo);
+	if (unlikely(r != 0)) {
+		amdgpu_bo_unpin(rbo);
+		amdgpu_bo_unreserve(rbo);
+		DRM_ERROR("%p bind failed\n", rbo);
+		return r;
+	}
+	amdgpu_bo_unreserve(rbo);
+
+	afb->address = amdgpu_bo_gpu_offset(rbo);
+
 	amdgpu_bo_ref(rbo);
 
 	if (dm_plane_state_new->dc_state &&
@@ -3105,17 +3137,6 @@ static int dm_plane_helper_prepare_fb(struct drm_plane *plane,
 		}
 	}
 
-	/* It's a hack for s3 since in 4.9 kernel filter out cursor buffer
-	 * prepare and cleanup in drm_atomic_helper_prepare_planes
-	 * and drm_atomic_helper_cleanup_planes because fb doens't in s3.
-	 * IN 4.10 kernel this code should be removed and amdgpu_device_suspend
-	 * code touching fram buffers should be avoided for DC.
-	 */
-	if (plane->type == DRM_PLANE_TYPE_CURSOR) {
-		struct amdgpu_crtc *acrtc = to_amdgpu_crtc(new_state->crtc);
-
-		acrtc->cursor_bo = obj;
-	}
 	return 0;
 }
 
@@ -3123,14 +3144,12 @@ static void dm_plane_helper_cleanup_fb(struct drm_plane *plane,
 				       struct drm_plane_state *old_state)
 {
 	struct amdgpu_bo *rbo;
-	struct amdgpu_framebuffer *afb;
 	int r;
 
 	if (!old_state->fb)
 		return;
 
-	afb = to_amdgpu_framebuffer(old_state->fb);
-	rbo = gem_to_amdgpu_bo(afb->obj);
+	rbo = gem_to_amdgpu_bo(old_state->fb->obj[0]);
 	r = amdgpu_bo_reserve(rbo, false);
 	if (unlikely(r)) {
 		DRM_ERROR("failed to reserve rbo before unpin\n");
@@ -3463,12 +3482,15 @@ static int amdgpu_dm_connector_get_modes(struct drm_connector *connector)
 	struct edid *edid = amdgpu_dm_connector->edid;
 
 	encoder = helper->best_encoder(connector);
-	amdgpu_dm_connector_ddc_get_modes(connector, edid);
-	amdgpu_dm_connector_add_common_modes(encoder, connector);
 
-#if defined(CONFIG_DRM_AMD_DC_FBC)
+	if (!edid || !drm_edid_is_valid(edid)) {
+		drm_add_modes_noedid(connector, 640, 480);
+	} else {
+		amdgpu_dm_connector_ddc_get_modes(connector, edid);
+		amdgpu_dm_connector_add_common_modes(encoder, connector);
+	}
 	amdgpu_dm_fbc_init(connector);
-#endif
+
 	return amdgpu_dm_connector->num_modes;
 }
 
@@ -3487,7 +3509,6 @@ void amdgpu_dm_connector_init_helper(struct amdgpu_display_manager *dm,
 	aconnector->base.stereo_allowed = false;
 	aconnector->base.dpms = DRM_MODE_DPMS_OFF;
 	aconnector->hpd.hpd = AMDGPU_HPD_NONE; /* not used */
-
 	mutex_init(&aconnector->hpd_lock);
 
 	/* configure support HPD hot plug connector_>polled default value is 0
@@ -3496,9 +3517,13 @@ void amdgpu_dm_connector_init_helper(struct amdgpu_display_manager *dm,
 	switch (connector_type) {
 	case DRM_MODE_CONNECTOR_HDMIA:
 		aconnector->base.polled = DRM_CONNECTOR_POLL_HPD;
+		aconnector->base.ycbcr_420_allowed =
+			link->link_enc->features.ycbcr420_supported ? true : false;
 		break;
 	case DRM_MODE_CONNECTOR_DisplayPort:
 		aconnector->base.polled = DRM_CONNECTOR_POLL_HPD;
+		aconnector->base.ycbcr_420_allowed =
+			link->link_enc->features.ycbcr420_supported ? true : false;
 		break;
 	case DRM_MODE_CONNECTOR_DVID:
 		aconnector->base.polled = DRM_CONNECTOR_POLL_HPD;
@@ -3651,10 +3676,17 @@ static int amdgpu_dm_connector_init(struct amdgpu_display_manager *dm,
 		link,
 		link_index);
 
-	drm_mode_connector_attach_encoder(
+	drm_connector_attach_encoder(
 		&aconnector->base, &aencoder->base);
 
 	drm_connector_register(&aconnector->base);
+#if defined(CONFIG_DEBUG_FS)
+	res = connector_debugfs_init(aconnector);
+	if (res) {
+		DRM_ERROR("Failed to create debugfs for connector");
+		goto out_free;
+	}
+#endif
 
 	if (connector_type == DRM_MODE_CONNECTOR_DisplayPort
 		|| connector_type == DRM_MODE_CONNECTOR_eDP)
@@ -3773,7 +3805,7 @@ static void remove_stream(struct amdgpu_device *adev,
 static int get_cursor_position(struct drm_plane *plane, struct drm_crtc *crtc,
 			       struct dc_cursor_position *position)
 {
-	struct amdgpu_crtc *amdgpu_crtc = amdgpu_crtc = to_amdgpu_crtc(crtc);
+	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
 	int x, y;
 	int xorigin = 0, yorigin = 0;
 
@@ -3905,7 +3937,7 @@ static void amdgpu_dm_do_flip(struct drm_crtc *crtc,
 	int r, vpos, hpos;
 	struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc);
 	struct amdgpu_framebuffer *afb = to_amdgpu_framebuffer(fb);
-	struct amdgpu_bo *abo = gem_to_amdgpu_bo(afb->obj);
+	struct amdgpu_bo *abo = gem_to_amdgpu_bo(fb->obj[0]);
 	struct amdgpu_device *adev = crtc->dev->dev_private;
 	bool async_flip = (crtc->state->pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC) != 0;
 	struct dc_flip_addrs addr = { {0} };
@@ -3951,8 +3983,6 @@ static void amdgpu_dm_do_flip(struct drm_crtc *crtc,
 
 	/* Flip */
 	spin_lock_irqsave(&crtc->dev->event_lock, flags);
-	/* update crtc fb */
-	crtc->primary->fb = fb;
 
 	WARN_ON(acrtc->pflip_status != AMDGPU_FLIP_NONE);
 	WARN_ON(!acrtc_state->stream);
@@ -3965,10 +3995,11 @@ static void amdgpu_dm_do_flip(struct drm_crtc *crtc,
 	if (acrtc->base.state->event)
 		prepare_flip_isr(acrtc);
 
+	spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
+
 	surface_updates->surface = dc_stream_get_status(acrtc_state->stream)->plane_states[0];
 	surface_updates->flip_addr = &addr;
 
-
 	dc_commit_updates_for_stream(adev->dm.dc,
 					     surface_updates,
 					     1,
@@ -3981,9 +4012,96 @@ static void amdgpu_dm_do_flip(struct drm_crtc *crtc,
 			 __func__,
 			 addr.address.grph.addr.high_part,
 			 addr.address.grph.addr.low_part);
+}
 
+/*
+ * TODO this whole function needs to go
+ *
+ * dc_surface_update is needlessly complex. See if we can just replace this
+ * with a dc_plane_state and follow the atomic model a bit more closely here.
+ */
+static bool commit_planes_to_stream(
+		struct dc *dc,
+		struct dc_plane_state **plane_states,
+		uint8_t new_plane_count,
+		struct dm_crtc_state *dm_new_crtc_state,
+		struct dm_crtc_state *dm_old_crtc_state,
+		struct dc_state *state)
+{
+	/* no need to dynamically allocate this. it's pretty small */
+	struct dc_surface_update updates[MAX_SURFACES];
+	struct dc_flip_addrs *flip_addr;
+	struct dc_plane_info *plane_info;
+	struct dc_scaling_info *scaling_info;
+	int i;
+	struct dc_stream_state *dc_stream = dm_new_crtc_state->stream;
+	struct dc_stream_update *stream_update =
+			kzalloc(sizeof(struct dc_stream_update), GFP_KERNEL);
 
-	spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
+	if (!stream_update) {
+		BREAK_TO_DEBUGGER();
+		return false;
+	}
+
+	flip_addr = kcalloc(MAX_SURFACES, sizeof(struct dc_flip_addrs),
+			    GFP_KERNEL);
+	plane_info = kcalloc(MAX_SURFACES, sizeof(struct dc_plane_info),
+			     GFP_KERNEL);
+	scaling_info = kcalloc(MAX_SURFACES, sizeof(struct dc_scaling_info),
+			       GFP_KERNEL);
+
+	if (!flip_addr || !plane_info || !scaling_info) {
+		kfree(flip_addr);
+		kfree(plane_info);
+		kfree(scaling_info);
+		kfree(stream_update);
+		return false;
+	}
+
+	memset(updates, 0, sizeof(updates));
+
+	stream_update->src = dc_stream->src;
+	stream_update->dst = dc_stream->dst;
+	stream_update->out_transfer_func = dc_stream->out_transfer_func;
+
+	for (i = 0; i < new_plane_count; i++) {
+		updates[i].surface = plane_states[i];
+		updates[i].gamma =
+			(struct dc_gamma *)plane_states[i]->gamma_correction;
+		updates[i].in_transfer_func = plane_states[i]->in_transfer_func;
+		flip_addr[i].address = plane_states[i]->address;
+		flip_addr[i].flip_immediate = plane_states[i]->flip_immediate;
+		plane_info[i].color_space = plane_states[i]->color_space;
+		plane_info[i].format = plane_states[i]->format;
+		plane_info[i].plane_size = plane_states[i]->plane_size;
+		plane_info[i].rotation = plane_states[i]->rotation;
+		plane_info[i].horizontal_mirror = plane_states[i]->horizontal_mirror;
+		plane_info[i].stereo_format = plane_states[i]->stereo_format;
+		plane_info[i].tiling_info = plane_states[i]->tiling_info;
+		plane_info[i].visible = plane_states[i]->visible;
+		plane_info[i].per_pixel_alpha = plane_states[i]->per_pixel_alpha;
+		plane_info[i].dcc = plane_states[i]->dcc;
+		scaling_info[i].scaling_quality = plane_states[i]->scaling_quality;
+		scaling_info[i].src_rect = plane_states[i]->src_rect;
+		scaling_info[i].dst_rect = plane_states[i]->dst_rect;
+		scaling_info[i].clip_rect = plane_states[i]->clip_rect;
+
+		updates[i].flip_addr = &flip_addr[i];
+		updates[i].plane_info = &plane_info[i];
+		updates[i].scaling_info = &scaling_info[i];
+	}
+
+	dc_commit_updates_for_stream(
+			dc,
+			updates,
+			new_plane_count,
+			dc_stream, stream_update, plane_states, state);
+
+	kfree(flip_addr);
+	kfree(plane_info);
+	kfree(scaling_info);
+	kfree(stream_update);
+	return true;
 }
 
 static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
@@ -4001,6 +4119,8 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 	struct drm_crtc_state *new_pcrtc_state =
 			drm_atomic_get_new_crtc_state(state, pcrtc);
 	struct dm_crtc_state *acrtc_state = to_dm_crtc_state(new_pcrtc_state);
+	struct dm_crtc_state *dm_old_crtc_state =
+			to_dm_crtc_state(drm_atomic_get_old_crtc_state(state, pcrtc));
 	struct dm_atomic_state *dm_state = to_dm_atomic_state(state);
 	int planes_count = 0;
 	unsigned long flags;
@@ -4037,7 +4157,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 		}
 		spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
 
-		if (!pflip_needed) {
+		if (!pflip_needed || plane->type == DRM_PLANE_TYPE_OVERLAY) {
 			WARN_ON(!dm_new_plane_state->dc_state);
 
 			plane_states_constructed[planes_count] = dm_new_plane_state->dc_state;
@@ -4079,10 +4199,12 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 			spin_unlock_irqrestore(&pcrtc->dev->event_lock, flags);
 		}
 
-		if (false == dc_commit_planes_to_stream(dm->dc,
+
+		if (false == commit_planes_to_stream(dm->dc,
 							plane_states_constructed,
 							planes_count,
-							dc_stream_attach,
+							acrtc_state,
+							dm_old_crtc_state,
 							dm_state->context))
 			dm_error("%s: Failed to attach plane!\n", __func__);
 	} else {
@@ -4149,6 +4271,7 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
 	struct drm_connector *connector;
 	struct drm_connector_state *old_con_state, *new_con_state;
 	struct dm_crtc_state *dm_old_crtc_state, *dm_new_crtc_state;
+	int crtc_disable_count = 0;
 
 	drm_atomic_helper_update_legacy_modeset_state(dev, state);
 
@@ -4211,6 +4334,8 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
 			if (dm_old_crtc_state->stream)
 				remove_stream(adev, acrtc, dm_old_crtc_state->stream);
 
+			pm_runtime_get_noresume(dev->dev);
+
 			acrtc->enabled = true;
 			acrtc->hw_mode = new_crtc_state->mode;
 			crtc->hwmode = new_crtc_state->mode;
@@ -4307,8 +4432,10 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
 		struct amdgpu_crtc *acrtc = to_amdgpu_crtc(dm_new_con_state->base.crtc);
 		struct dc_stream_status *status = NULL;
 
-		if (acrtc)
+		if (acrtc) {
 			new_crtc_state = drm_atomic_get_new_crtc_state(state, &acrtc->base);
+			old_crtc_state = drm_atomic_get_old_crtc_state(state, &acrtc->base);
+		}
 
 		/* Skip any modesets/resets */
 		if (!acrtc || drm_atomic_crtc_needs_modeset(new_crtc_state))
@@ -4331,11 +4458,12 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
 		WARN_ON(!status->plane_count);
 
 		/*TODO How it works with MPO ?*/
-		if (!dc_commit_planes_to_stream(
+		if (!commit_planes_to_stream(
 				dm->dc,
 				status->plane_states,
 				status->plane_count,
-				dm_new_crtc_state->stream,
+				dm_new_crtc_state,
+				to_dm_crtc_state(old_crtc_state),
 				dm_state->context))
 			dm_error("%s: Failed to update stream scaling!\n", __func__);
 	}
@@ -4348,6 +4476,9 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
 		struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc);
 		bool modeset_needed;
 
+		if (old_crtc_state->active && !new_crtc_state->active)
+			crtc_disable_count++;
+
 		dm_new_crtc_state = to_dm_crtc_state(new_crtc_state);
 		dm_old_crtc_state = to_dm_crtc_state(old_crtc_state);
 		modeset_needed = modeset_required(
@@ -4396,6 +4527,14 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
 		drm_atomic_helper_wait_for_flip_done(dev, state);
 
 	drm_atomic_helper_cleanup_planes(dev, state);
+
+	/* Finally, drop a runtime PM reference for each newly disabled CRTC,
+	 * so we can put the GPU into runtime suspend if we're not driving any
+	 * displays anymore
+	 */
+	for (i = 0; i < crtc_disable_count; i++)
+		pm_runtime_put_autosuspend(dev->dev);
+	pm_runtime_mark_last_busy(dev->dev);
 }
 
 
@@ -4555,8 +4694,8 @@ static int dm_update_crtcs_state(struct dc *dc,
 	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
 		struct amdgpu_crtc *acrtc = NULL;
 		struct amdgpu_dm_connector *aconnector = NULL;
-		struct drm_connector_state *new_con_state = NULL;
-		struct dm_connector_state *dm_conn_state = NULL;
+		struct drm_connector_state *drm_new_conn_state = NULL, *drm_old_conn_state = NULL;
+		struct dm_connector_state *dm_new_conn_state = NULL, *dm_old_conn_state = NULL;
 		struct drm_plane_state *new_plane_state = NULL;
 
 		new_stream = NULL;
@@ -4577,19 +4716,22 @@ static int dm_update_crtcs_state(struct dc *dc,
 		/* TODO This hack should go away */
 		if (aconnector && enable) {
 			// Make sure fake sink is created in plug-in scenario
-			new_con_state = drm_atomic_get_connector_state(state,
+			drm_new_conn_state = drm_atomic_get_new_connector_state(state,
  								    &aconnector->base);
+			drm_old_conn_state = drm_atomic_get_old_connector_state(state,
+								    &aconnector->base);
 
-			if (IS_ERR(new_con_state)) {
-				ret = PTR_ERR_OR_ZERO(new_con_state);
+			if (IS_ERR(drm_new_conn_state)) {
+				ret = PTR_ERR_OR_ZERO(drm_new_conn_state);
 				break;
 			}
 
-			dm_conn_state = to_dm_connector_state(new_con_state);
+			dm_new_conn_state = to_dm_connector_state(drm_new_conn_state);
+			dm_old_conn_state = to_dm_connector_state(drm_old_conn_state);
 
 			new_stream = create_stream_for_sink(aconnector,
 							     &new_crtc_state->mode,
-							    dm_conn_state);
+							    dm_new_conn_state);
 
 			/*
 			 * we can have no stream on ACTION_SET if a display
@@ -4695,20 +4837,30 @@ next_crtc:
 		 * We want to do dc stream updates that do not require a
 		 * full modeset below.
 		 */
-		if (!enable || !aconnector || modereset_required(new_crtc_state))
+		if (!(enable && aconnector && new_crtc_state->enable &&
+		      new_crtc_state->active))
 			continue;
 		/*
 		 * Given above conditions, the dc state cannot be NULL because:
-		 * 1. We're attempting to enable a CRTC. Which has a...
-		 * 2. Valid connector attached, and
-		 * 3. User does not want to reset it (disable or mark inactive,
-		 *    which can happen on a CRTC that's already disabled).
-		 * => It currently exists.
+		 * 1. We're in the process of enabling CRTCs (just been added
+		 *    to the dc context, or already is on the context)
+		 * 2. Has a valid connector attached, and
+		 * 3. Is currently active and enabled.
+		 * => The dc stream state currently exists.
 		 */
 		BUG_ON(dm_new_crtc_state->stream == NULL);
 
-		/* Color managment settings */
-		if (dm_new_crtc_state->base.color_mgmt_changed) {
+		/* Scaling or underscan settings */
+		if (is_scaling_state_different(dm_old_conn_state, dm_new_conn_state))
+			update_stream_scaling_settings(
+				&new_crtc_state->mode, dm_new_conn_state, dm_new_crtc_state->stream);
+
+		/*
+		 * Color management settings. We also update color properties
+		 * when a modeset is needed, to ensure it gets reprogrammed.
+		 */
+		if (dm_new_crtc_state->base.color_mgmt_changed ||
+		    drm_atomic_crtc_needs_modeset(new_crtc_state)) {
 			ret = amdgpu_dm_set_regamma_lut(dm_new_crtc_state);
 			if (ret)
 				goto fail;
@@ -4755,7 +4907,8 @@ static int dm_update_planes_state(struct dc *dc,
 
 		/* Remove any changed/removed planes */
 		if (!enable) {
-			if (pflip_needed)
+			if (pflip_needed &&
+			    plane->type != DRM_PLANE_TYPE_OVERLAY)
 				continue;
 
 			if (!old_plane_crtc)
@@ -4802,7 +4955,8 @@ static int dm_update_planes_state(struct dc *dc,
 			if (!dm_new_crtc_state->stream)
 				continue;
 
-			if (pflip_needed)
+			if (pflip_needed &&
+			    plane->type != DRM_PLANE_TYPE_OVERLAY)
 				continue;
 
 			WARN_ON(dm_new_plane_state->dc_state);
@@ -5009,17 +5163,24 @@ void amdgpu_dm_add_sink_to_freesync_module(struct drm_connector *connector,
 					   struct edid *edid)
 {
 	int i;
-	uint64_t val_capable;
 	bool edid_check_required;
 	struct detailed_timing *timing;
 	struct detailed_non_pixel *data;
 	struct detailed_data_monitor_range *range;
 	struct amdgpu_dm_connector *amdgpu_dm_connector =
 			to_amdgpu_dm_connector(connector);
+	struct dm_connector_state *dm_con_state;
 
 	struct drm_device *dev = connector->dev;
 	struct amdgpu_device *adev = dev->dev_private;
 
+	if (!connector->state) {
+		DRM_ERROR("%s - Connector has no state", __func__);
+		return;
+	}
+
+	dm_con_state = to_dm_connector_state(connector->state);
+
 	edid_check_required = false;
 	if (!amdgpu_dm_connector->dc_sink) {
 		DRM_ERROR("dc_sink NULL, could not add free_sync module.\n");
@@ -5038,7 +5199,7 @@ void amdgpu_dm_add_sink_to_freesync_module(struct drm_connector *connector,
 						amdgpu_dm_connector);
 		}
 	}
-	val_capable = 0;
+	dm_con_state->freesync_capable = false;
 	if (edid_check_required == true && (edid->version > 1 ||
 	   (edid->version == 1 && edid->revision > 1))) {
 		for (i = 0; i < 4; i++) {
@@ -5074,7 +5235,7 @@ void amdgpu_dm_add_sink_to_freesync_module(struct drm_connector *connector,
 					amdgpu_dm_connector->min_vfreq * 1000000;
 			amdgpu_dm_connector->caps.max_refresh_in_micro_hz =
 					amdgpu_dm_connector->max_vfreq * 1000000;
-				val_capable = 1;
+			dm_con_state->freesync_capable = true;
 		}
 	}
 
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
index b68400c1154b..a29dc35954c9 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
@@ -28,7 +28,6 @@
 
 #include <drm/drmP.h>
 #include <drm/drm_atomic.h>
-#include "dc.h"
 
 /*
  * This file contains the definition for amdgpu_display_manager
@@ -53,6 +52,7 @@
 struct amdgpu_device;
 struct drm_device;
 struct amdgpu_dm_irq_handler_data;
+struct dc;
 
 struct amdgpu_dm_prev_state {
 	struct drm_framebuffer *fb;
@@ -72,13 +72,11 @@ struct irq_list_head {
 	struct work_struct work;
 };
 
-#if defined(CONFIG_DRM_AMD_DC_FBC)
 struct dm_comressor_info {
 	void *cpu_addr;
 	struct amdgpu_bo *bo_ptr;
 	uint64_t gpu_addr;
 };
-#endif
 
 
 struct amdgpu_display_manager {
@@ -129,9 +127,8 @@ struct amdgpu_display_manager {
 	 * Caches device atomic state for suspend/resume
 	 */
 	struct drm_atomic_state *cached_state;
-#if defined(CONFIG_DRM_AMD_DC_FBC)
+
 	struct dm_comressor_info compressor;
-#endif
 };
 
 struct amdgpu_dm_connector {
@@ -220,6 +217,7 @@ struct dm_connector_state {
 	uint8_t underscan_hborder;
 	bool underscan_enable;
 	struct mod_freesync_user_enable user_enable;
+	bool freesync_capable;
 };
 
 #define to_dm_connector_state(x)\
@@ -246,7 +244,7 @@ void amdgpu_dm_connector_init_helper(struct amdgpu_display_manager *dm,
 				     struct dc_link *link,
 				     int link_index);
 
-int amdgpu_dm_connector_mode_valid(struct drm_connector *connector,
+enum drm_mode_status amdgpu_dm_connector_mode_valid(struct drm_connector *connector,
 				   struct drm_display_mode *mode);
 
 void dm_restore_drm_connector_state(struct drm_device *dev,
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
index 25f064c01038..b329393307e5 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_color.c
@@ -25,6 +25,7 @@
 
 #include "amdgpu_mode.h"
 #include "amdgpu_dm.h"
+#include "dc.h"
 #include "modules/color/color_gamma.h"
 
 #define MAX_DRM_LUT_VALUE 0xFFFF
@@ -87,9 +88,9 @@ static void __drm_lut_to_dc_gamma(struct drm_color_lut *lut,
 			g = drm_color_lut_extract(lut[i].green, 16);
 			b = drm_color_lut_extract(lut[i].blue, 16);
 
-			gamma->entries.red[i] = dal_fixed31_32_from_int(r);
-			gamma->entries.green[i] = dal_fixed31_32_from_int(g);
-			gamma->entries.blue[i] = dal_fixed31_32_from_int(b);
+			gamma->entries.red[i] = dc_fixpt_from_int(r);
+			gamma->entries.green[i] = dc_fixpt_from_int(g);
+			gamma->entries.blue[i] = dc_fixpt_from_int(b);
 		}
 		return;
 	}
@@ -100,9 +101,9 @@ static void __drm_lut_to_dc_gamma(struct drm_color_lut *lut,
 		g = drm_color_lut_extract(lut[i].green, 16);
 		b = drm_color_lut_extract(lut[i].blue, 16);
 
-		gamma->entries.red[i] = dal_fixed31_32_from_fraction(r, MAX_DRM_LUT_VALUE);
-		gamma->entries.green[i] = dal_fixed31_32_from_fraction(g, MAX_DRM_LUT_VALUE);
-		gamma->entries.blue[i] = dal_fixed31_32_from_fraction(b, MAX_DRM_LUT_VALUE);
+		gamma->entries.red[i] = dc_fixpt_from_fraction(r, MAX_DRM_LUT_VALUE);
+		gamma->entries.green[i] = dc_fixpt_from_fraction(g, MAX_DRM_LUT_VALUE);
+		gamma->entries.blue[i] = dc_fixpt_from_fraction(b, MAX_DRM_LUT_VALUE);
 	}
 }
 
@@ -207,7 +208,7 @@ void amdgpu_dm_set_ctm(struct dm_crtc_state *crtc)
 	for (i = 0; i < 12; i++) {
 		/* Skip 4th element */
 		if (i % 4 == 3) {
-			stream->gamut_remap_matrix.matrix[i] = dal_fixed31_32_zero;
+			stream->gamut_remap_matrix.matrix[i] = dc_fixpt_zero;
 			continue;
 		}
 
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
new file mode 100644
index 000000000000..0d9e410ca01e
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
@@ -0,0 +1,722 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include <linux/debugfs.h>
+
+#include "dc.h"
+#include "amdgpu.h"
+#include "amdgpu_dm.h"
+#include "amdgpu_dm_debugfs.h"
+
+/* function description
+ * get/ set DP configuration: lane_count, link_rate, spread_spectrum
+ *
+ * valid lane count value: 1, 2, 4
+ * valid link rate value:
+ * 06h = 1.62Gbps per lane
+ * 0Ah = 2.7Gbps per lane
+ * 0Ch = 3.24Gbps per lane
+ * 14h = 5.4Gbps per lane
+ * 1Eh = 8.1Gbps per lane
+ *
+ * debugfs is located at /sys/kernel/debug/dri/0/DP-x/link_settings
+ *
+ * --- to get dp configuration
+ *
+ * cat link_settings
+ *
+ * It will list current, verified, reported, preferred dp configuration.
+ * current -- for current video mode
+ * verified --- maximum configuration which pass link training
+ * reported --- DP rx report caps (DPCD register offset 0, 1 2)
+ * preferred --- user force settings
+ *
+ * --- set (or force) dp configuration
+ *
+ * echo <lane_count>  <link_rate> > link_settings
+ *
+ * for example, to force to  2 lane, 2.7GHz,
+ * echo 4 0xa > link_settings
+ *
+ * spread_spectrum could not be changed dynamically.
+ *
+ * in case invalid lane count, link rate are force, no hw programming will be
+ * done. please check link settings after force operation to see if HW get
+ * programming.
+ *
+ * cat link_settings
+ *
+ * check current and preferred settings.
+ *
+ */
+static ssize_t dp_link_settings_read(struct file *f, char __user *buf,
+				 size_t size, loff_t *pos)
+{
+	struct amdgpu_dm_connector *connector = file_inode(f)->i_private;
+	struct dc_link *link = connector->dc_link;
+	char *rd_buf = NULL;
+	char *rd_buf_ptr = NULL;
+	const uint32_t rd_buf_size = 100;
+	uint32_t result = 0;
+	uint8_t str_len = 0;
+	int r;
+
+	if (*pos & 3 || size & 3)
+		return -EINVAL;
+
+	rd_buf = kcalloc(rd_buf_size, sizeof(char), GFP_KERNEL);
+	if (!rd_buf)
+		return 0;
+
+	rd_buf_ptr = rd_buf;
+
+	str_len = strlen("Current:  %d  %d  %d  ");
+	snprintf(rd_buf_ptr, str_len, "Current:  %d  %d  %d  ",
+			link->cur_link_settings.lane_count,
+			link->cur_link_settings.link_rate,
+			link->cur_link_settings.link_spread);
+	rd_buf_ptr += str_len;
+
+	str_len = strlen("Verified:  %d  %d  %d  ");
+	snprintf(rd_buf_ptr, str_len, "Verified:  %d  %d  %d  ",
+			link->verified_link_cap.lane_count,
+			link->verified_link_cap.link_rate,
+			link->verified_link_cap.link_spread);
+	rd_buf_ptr += str_len;
+
+	str_len = strlen("Reported:  %d  %d  %d  ");
+	snprintf(rd_buf_ptr, str_len, "Reported:  %d  %d  %d  ",
+			link->reported_link_cap.lane_count,
+			link->reported_link_cap.link_rate,
+			link->reported_link_cap.link_spread);
+	rd_buf_ptr += str_len;
+
+	str_len = strlen("Preferred:  %d  %d  %d  ");
+	snprintf(rd_buf_ptr, str_len, "Preferred:  %d  %d  %d\n",
+			link->preferred_link_setting.lane_count,
+			link->preferred_link_setting.link_rate,
+			link->preferred_link_setting.link_spread);
+
+	while (size) {
+		if (*pos >= rd_buf_size)
+			break;
+
+		r = put_user(*(rd_buf + result), buf);
+		if (r)
+			return r; /* r = -EFAULT */
+
+		buf += 1;
+		size -= 1;
+		*pos += 1;
+		result += 1;
+	}
+
+	kfree(rd_buf);
+	return result;
+}
+
+static ssize_t dp_link_settings_write(struct file *f, const char __user *buf,
+				 size_t size, loff_t *pos)
+{
+	struct amdgpu_dm_connector *connector = file_inode(f)->i_private;
+	struct dc_link *link = connector->dc_link;
+	struct dc *dc = (struct dc *)link->dc;
+	struct dc_link_settings prefer_link_settings;
+	char *wr_buf = NULL;
+	char *wr_buf_ptr = NULL;
+	const uint32_t wr_buf_size = 40;
+	int r;
+	int bytes_from_user;
+	char *sub_str;
+	/* 0: lane_count; 1: link_rate */
+	uint8_t param_index = 0;
+	long param[2];
+	const char delimiter[3] = {' ', '\n', '\0'};
+	bool valid_input = false;
+
+	if (size == 0)
+		return -EINVAL;
+
+	wr_buf = kcalloc(wr_buf_size, sizeof(char), GFP_KERNEL);
+	if (!wr_buf)
+		return -EINVAL;
+	wr_buf_ptr = wr_buf;
+
+	r = copy_from_user(wr_buf_ptr, buf, wr_buf_size);
+
+	/* r is bytes not be copied */
+	if (r >= wr_buf_size) {
+		kfree(wr_buf);
+		DRM_DEBUG_DRIVER("user data not read\n");
+		return -EINVAL;
+	}
+
+	bytes_from_user = wr_buf_size - r;
+
+	while (isspace(*wr_buf_ptr))
+		wr_buf_ptr++;
+
+	while ((*wr_buf_ptr != '\0') && (param_index < 2)) {
+
+		sub_str = strsep(&wr_buf_ptr, delimiter);
+
+		r = kstrtol(sub_str, 16, &param[param_index]);
+
+		if (r)
+			DRM_DEBUG_DRIVER("string to int convert error code: %d\n", r);
+
+		param_index++;
+		while (isspace(*wr_buf_ptr))
+			wr_buf_ptr++;
+	}
+
+	switch (param[0]) {
+	case LANE_COUNT_ONE:
+	case LANE_COUNT_TWO:
+	case LANE_COUNT_FOUR:
+		valid_input = true;
+		break;
+	default:
+		break;
+	}
+
+	switch (param[1]) {
+	case LINK_RATE_LOW:
+	case LINK_RATE_HIGH:
+	case LINK_RATE_RBR2:
+	case LINK_RATE_HIGH2:
+	case LINK_RATE_HIGH3:
+		valid_input = true;
+		break;
+	default:
+		break;
+	}
+
+	if (!valid_input) {
+		kfree(wr_buf);
+		DRM_DEBUG_DRIVER("Invalid Input value No HW will be programmed\n");
+		return bytes_from_user;
+	}
+
+	/* save user force lane_count, link_rate to preferred settings
+	 * spread spectrum will not be changed
+	 */
+	prefer_link_settings.link_spread = link->cur_link_settings.link_spread;
+	prefer_link_settings.lane_count = param[0];
+	prefer_link_settings.link_rate = param[1];
+
+	dc_link_set_preferred_link_settings(dc, &prefer_link_settings, link);
+
+	kfree(wr_buf);
+	return bytes_from_user;
+}
+
+/* function: get current DP PHY settings: voltage swing, pre-emphasis,
+ * post-cursor2 (defined by VESA DP specification)
+ *
+ * valid values
+ * voltage swing: 0,1,2,3
+ * pre-emphasis : 0,1,2,3
+ * post cursor2 : 0,1,2,3
+ *
+ *
+ * how to use this debugfs
+ *
+ * debugfs is located at /sys/kernel/debug/dri/0/DP-x
+ *
+ * there will be directories, like DP-1, DP-2,DP-3, etc. for DP display
+ *
+ * To figure out which DP-x is the display for DP to be check,
+ * cd DP-x
+ * ls -ll
+ * There should be debugfs file, like link_settings, phy_settings.
+ * cat link_settings
+ * from lane_count, link_rate to figure which DP-x is for display to be worked
+ * on
+ *
+ * To get current DP PHY settings,
+ * cat phy_settings
+ *
+ * To change DP PHY settings,
+ * echo <voltage_swing> <pre-emphasis> <post_cursor2> > phy_settings
+ * for examle, to change voltage swing to 2, pre-emphasis to 3, post_cursor2 to
+ * 0,
+ * echo 2 3 0 > phy_settings
+ *
+ * To check if change be applied, get current phy settings by
+ * cat phy_settings
+ *
+ * In case invalid values are set by user, like
+ * echo 1 4 0 > phy_settings
+ *
+ * HW will NOT be programmed by these settings.
+ * cat phy_settings will show the previous valid settings.
+ */
+static ssize_t dp_phy_settings_read(struct file *f, char __user *buf,
+				 size_t size, loff_t *pos)
+{
+	struct amdgpu_dm_connector *connector = file_inode(f)->i_private;
+	struct dc_link *link = connector->dc_link;
+	char *rd_buf = NULL;
+	const uint32_t rd_buf_size = 20;
+	uint32_t result = 0;
+	int r;
+
+	if (*pos & 3 || size & 3)
+		return -EINVAL;
+
+	rd_buf = kcalloc(rd_buf_size, sizeof(char), GFP_KERNEL);
+	if (!rd_buf)
+		return -EINVAL;
+
+	snprintf(rd_buf, rd_buf_size, "  %d  %d  %d  ",
+			link->cur_lane_setting.VOLTAGE_SWING,
+			link->cur_lane_setting.PRE_EMPHASIS,
+			link->cur_lane_setting.POST_CURSOR2);
+
+	while (size) {
+		if (*pos >= rd_buf_size)
+			break;
+
+		r = put_user((*(rd_buf + result)), buf);
+		if (r)
+			return r; /* r = -EFAULT */
+
+		buf += 1;
+		size -= 1;
+		*pos += 1;
+		result += 1;
+	}
+
+	kfree(rd_buf);
+	return result;
+}
+
+static ssize_t dp_phy_settings_write(struct file *f, const char __user *buf,
+				 size_t size, loff_t *pos)
+{
+	struct amdgpu_dm_connector *connector = file_inode(f)->i_private;
+	struct dc_link *link = connector->dc_link;
+	struct dc *dc = (struct dc *)link->dc;
+	char *wr_buf = NULL;
+	char *wr_buf_ptr = NULL;
+	uint32_t wr_buf_size = 40;
+	int r;
+	int bytes_from_user;
+	char *sub_str;
+	uint8_t param_index = 0;
+	long param[3];
+	const char delimiter[3] = {' ', '\n', '\0'};
+	bool use_prefer_link_setting;
+	struct link_training_settings link_lane_settings;
+
+	if (size == 0)
+		return 0;
+
+	wr_buf = kcalloc(wr_buf_size, sizeof(char), GFP_KERNEL);
+	if (!wr_buf)
+		return 0;
+	wr_buf_ptr = wr_buf;
+
+	r = copy_from_user(wr_buf_ptr, buf, wr_buf_size);
+
+	/* r is bytes not be copied */
+	if (r >= wr_buf_size) {
+		kfree(wr_buf);
+		DRM_DEBUG_DRIVER("user data not be read\n");
+		return 0;
+	}
+
+	bytes_from_user = wr_buf_size - r;
+
+	while (isspace(*wr_buf_ptr))
+		wr_buf_ptr++;
+
+	while ((*wr_buf_ptr != '\0') && (param_index < 3)) {
+
+		sub_str = strsep(&wr_buf_ptr, delimiter);
+
+		r = kstrtol(sub_str, 16, &param[param_index]);
+
+		if (r)
+			DRM_DEBUG_DRIVER("string to int convert error code: %d\n", r);
+
+		param_index++;
+		while (isspace(*wr_buf_ptr))
+			wr_buf_ptr++;
+	}
+
+	if ((param[0] > VOLTAGE_SWING_MAX_LEVEL) ||
+			(param[1] > PRE_EMPHASIS_MAX_LEVEL) ||
+			(param[2] > POST_CURSOR2_MAX_LEVEL)) {
+		kfree(wr_buf);
+		DRM_DEBUG_DRIVER("Invalid Input No HW will be programmed\n");
+		return bytes_from_user;
+	}
+
+	/* get link settings: lane count, link rate */
+	use_prefer_link_setting =
+		((link->preferred_link_setting.link_rate != LINK_RATE_UNKNOWN) &&
+		(link->test_pattern_enabled));
+
+	memset(&link_lane_settings, 0, sizeof(link_lane_settings));
+
+	if (use_prefer_link_setting) {
+		link_lane_settings.link_settings.lane_count =
+				link->preferred_link_setting.lane_count;
+		link_lane_settings.link_settings.link_rate =
+				link->preferred_link_setting.link_rate;
+		link_lane_settings.link_settings.link_spread =
+				link->preferred_link_setting.link_spread;
+	} else {
+		link_lane_settings.link_settings.lane_count =
+				link->cur_link_settings.lane_count;
+		link_lane_settings.link_settings.link_rate =
+				link->cur_link_settings.link_rate;
+		link_lane_settings.link_settings.link_spread =
+				link->cur_link_settings.link_spread;
+	}
+
+	/* apply phy settings from user */
+	for (r = 0; r < link_lane_settings.link_settings.lane_count; r++) {
+		link_lane_settings.lane_settings[r].VOLTAGE_SWING =
+				(enum dc_voltage_swing) (param[0]);
+		link_lane_settings.lane_settings[r].PRE_EMPHASIS =
+				(enum dc_pre_emphasis) (param[1]);
+		link_lane_settings.lane_settings[r].POST_CURSOR2 =
+				(enum dc_post_cursor2) (param[2]);
+	}
+
+	/* program ASIC registers and DPCD registers */
+	dc_link_set_drive_settings(dc, &link_lane_settings, link);
+
+	kfree(wr_buf);
+	return bytes_from_user;
+}
+
+/* function description
+ *
+ * set PHY layer or Link layer test pattern
+ * PHY test pattern is used for PHY SI check.
+ * Link layer test will not affect PHY SI.
+ *
+ * Reset Test Pattern:
+ * 0 = DP_TEST_PATTERN_VIDEO_MODE
+ *
+ * PHY test pattern supported:
+ * 1 = DP_TEST_PATTERN_D102
+ * 2 = DP_TEST_PATTERN_SYMBOL_ERROR
+ * 3 = DP_TEST_PATTERN_PRBS7
+ * 4 = DP_TEST_PATTERN_80BIT_CUSTOM
+ * 5 = DP_TEST_PATTERN_CP2520_1
+ * 6 = DP_TEST_PATTERN_CP2520_2 = DP_TEST_PATTERN_HBR2_COMPLIANCE_EYE
+ * 7 = DP_TEST_PATTERN_CP2520_3
+ *
+ * DP PHY Link Training Patterns
+ * 8 = DP_TEST_PATTERN_TRAINING_PATTERN1
+ * 9 = DP_TEST_PATTERN_TRAINING_PATTERN2
+ * a = DP_TEST_PATTERN_TRAINING_PATTERN3
+ * b = DP_TEST_PATTERN_TRAINING_PATTERN4
+ *
+ * DP Link Layer Test pattern
+ * c = DP_TEST_PATTERN_COLOR_SQUARES
+ * d = DP_TEST_PATTERN_COLOR_SQUARES_CEA
+ * e = DP_TEST_PATTERN_VERTICAL_BARS
+ * f = DP_TEST_PATTERN_HORIZONTAL_BARS
+ * 10= DP_TEST_PATTERN_COLOR_RAMP
+ *
+ * debugfs phy_test_pattern is located at /syskernel/debug/dri/0/DP-x
+ *
+ * --- set test pattern
+ * echo <test pattern #> > test_pattern
+ *
+ * If test pattern # is not supported, NO HW programming will be done.
+ * for DP_TEST_PATTERN_80BIT_CUSTOM, it needs extra 10 bytes of data
+ * for the user pattern. input 10 bytes data are separated by space
+ *
+ * echo 0x4 0x11 0x22 0x33 0x44 0x55 0x66 0x77 0x88 0x99 0xaa > test_pattern
+ *
+ * --- reset test pattern
+ * echo 0 > test_pattern
+ *
+ * --- HPD detection is disabled when set PHY test pattern
+ *
+ * when PHY test pattern (pattern # within [1,7]) is set, HPD pin of HW ASIC
+ * is disable. User could unplug DP display from DP connected and plug scope to
+ * check test pattern PHY SI.
+ * If there is need unplug scope and plug DP display back, do steps below:
+ * echo 0 > phy_test_pattern
+ * unplug scope
+ * plug DP display.
+ *
+ * "echo 0 > phy_test_pattern" will re-enable HPD pin again so that video sw
+ * driver could detect "unplug scope" and "plug DP display"
+ */
+static ssize_t dp_phy_test_pattern_debugfs_write(struct file *f, const char __user *buf,
+				 size_t size, loff_t *pos)
+{
+	struct amdgpu_dm_connector *connector = file_inode(f)->i_private;
+	struct dc_link *link = connector->dc_link;
+	char *wr_buf = NULL;
+	char *wr_buf_ptr = NULL;
+	uint32_t wr_buf_size = 100;
+	uint32_t wr_buf_count = 0;
+	int r;
+	int bytes_from_user;
+	char *sub_str = NULL;
+	uint8_t param_index = 0;
+	uint8_t param_nums = 0;
+	long param[11] = {0x0};
+	const char delimiter[3] = {' ', '\n', '\0'};
+	enum dp_test_pattern test_pattern = DP_TEST_PATTERN_UNSUPPORTED;
+	bool disable_hpd = false;
+	bool valid_test_pattern = false;
+	/* init with defalut 80bit custom pattern */
+	uint8_t custom_pattern[10] = {
+			0x1f, 0x7c, 0xf0, 0xc1, 0x07,
+			0x1f, 0x7c, 0xf0, 0xc1, 0x07
+			};
+	struct dc_link_settings prefer_link_settings = {LANE_COUNT_UNKNOWN,
+			LINK_RATE_UNKNOWN, LINK_SPREAD_DISABLED};
+	struct dc_link_settings cur_link_settings = {LANE_COUNT_UNKNOWN,
+			LINK_RATE_UNKNOWN, LINK_SPREAD_DISABLED};
+	struct link_training_settings link_training_settings;
+	int i;
+
+	if (size == 0)
+		return 0;
+
+	wr_buf = kcalloc(wr_buf_size, sizeof(char), GFP_KERNEL);
+	if (!wr_buf)
+		return 0;
+	wr_buf_ptr = wr_buf;
+
+	r = copy_from_user(wr_buf_ptr, buf, wr_buf_size);
+
+	/* r is bytes not be copied */
+	if (r >= wr_buf_size) {
+		kfree(wr_buf);
+		DRM_DEBUG_DRIVER("user data not be read\n");
+		return 0;
+	}
+
+	bytes_from_user = wr_buf_size - r;
+
+	/* check number of parameters. isspace could not differ space and \n */
+	while ((*wr_buf_ptr != 0xa) && (wr_buf_count < wr_buf_size)) {
+		/* skip space*/
+		while (isspace(*wr_buf_ptr) && (wr_buf_count < wr_buf_size)) {
+			wr_buf_ptr++;
+			wr_buf_count++;
+			}
+
+		if (wr_buf_count == wr_buf_size)
+			break;
+
+		/* skip non-space*/
+		while ((!isspace(*wr_buf_ptr)) && (wr_buf_count < wr_buf_size)) {
+			wr_buf_ptr++;
+			wr_buf_count++;
+			}
+
+		param_nums++;
+
+		if (wr_buf_count == wr_buf_size)
+			break;
+	}
+
+	/* max 11 parameters */
+	if (param_nums > 11)
+		param_nums = 11;
+
+	wr_buf_ptr = wr_buf; /* reset buf pinter */
+	wr_buf_count = 0; /* number of char already checked */
+
+	while (isspace(*wr_buf_ptr) && (wr_buf_count < wr_buf_size)) {
+		wr_buf_ptr++;
+		wr_buf_count++;
+	}
+
+	while (param_index < param_nums) {
+		/* after strsep, wr_buf_ptr will be moved to after space */
+		sub_str = strsep(&wr_buf_ptr, delimiter);
+
+		r = kstrtol(sub_str, 16, &param[param_index]);
+
+		if (r)
+			DRM_DEBUG_DRIVER("string to int convert error code: %d\n", r);
+
+		param_index++;
+	}
+
+	test_pattern = param[0];
+
+	switch (test_pattern) {
+	case DP_TEST_PATTERN_VIDEO_MODE:
+	case DP_TEST_PATTERN_COLOR_SQUARES:
+	case DP_TEST_PATTERN_COLOR_SQUARES_CEA:
+	case DP_TEST_PATTERN_VERTICAL_BARS:
+	case DP_TEST_PATTERN_HORIZONTAL_BARS:
+	case DP_TEST_PATTERN_COLOR_RAMP:
+		valid_test_pattern = true;
+		break;
+
+	case DP_TEST_PATTERN_D102:
+	case DP_TEST_PATTERN_SYMBOL_ERROR:
+	case DP_TEST_PATTERN_PRBS7:
+	case DP_TEST_PATTERN_80BIT_CUSTOM:
+	case DP_TEST_PATTERN_HBR2_COMPLIANCE_EYE:
+	case DP_TEST_PATTERN_TRAINING_PATTERN4:
+		disable_hpd = true;
+		valid_test_pattern = true;
+		break;
+
+	default:
+		valid_test_pattern = false;
+		test_pattern = DP_TEST_PATTERN_UNSUPPORTED;
+		break;
+	}
+
+	if (!valid_test_pattern) {
+		kfree(wr_buf);
+		DRM_DEBUG_DRIVER("Invalid Test Pattern Parameters\n");
+		return bytes_from_user;
+	}
+
+	if (test_pattern == DP_TEST_PATTERN_80BIT_CUSTOM) {
+		for (i = 0; i < 10; i++) {
+			if ((uint8_t) param[i + 1] != 0x0)
+				break;
+		}
+
+		if (i < 10) {
+			/* not use default value */
+			for (i = 0; i < 10; i++)
+				custom_pattern[i] = (uint8_t) param[i + 1];
+		}
+	}
+
+	/* Usage: set DP physical test pattern using debugfs with normal DP
+	 * panel. Then plug out DP panel and connect a scope to measure
+	 * For normal video mode and test pattern generated from CRCT,
+	 * they are visibile to user. So do not disable HPD.
+	 * Video Mode is also set to clear the test pattern, so enable HPD
+	 * because it might have been disabled after a test pattern was set.
+	 * AUX depends on HPD * sequence dependent, do not move!
+	 */
+	if (!disable_hpd)
+		dc_link_enable_hpd(link);
+
+	prefer_link_settings.lane_count = link->verified_link_cap.lane_count;
+	prefer_link_settings.link_rate = link->verified_link_cap.link_rate;
+	prefer_link_settings.link_spread = link->verified_link_cap.link_spread;
+
+	cur_link_settings.lane_count = link->cur_link_settings.lane_count;
+	cur_link_settings.link_rate = link->cur_link_settings.link_rate;
+	cur_link_settings.link_spread = link->cur_link_settings.link_spread;
+
+	link_training_settings.link_settings = cur_link_settings;
+
+
+	if (test_pattern != DP_TEST_PATTERN_VIDEO_MODE) {
+		if (prefer_link_settings.lane_count != LANE_COUNT_UNKNOWN &&
+			prefer_link_settings.link_rate !=  LINK_RATE_UNKNOWN &&
+			(prefer_link_settings.lane_count != cur_link_settings.lane_count ||
+			prefer_link_settings.link_rate != cur_link_settings.link_rate))
+			link_training_settings.link_settings = prefer_link_settings;
+	}
+
+	for (i = 0; i < (unsigned int)(link_training_settings.link_settings.lane_count); i++)
+		link_training_settings.lane_settings[i] = link->cur_lane_setting;
+
+	dc_link_set_test_pattern(
+		link,
+		test_pattern,
+		&link_training_settings,
+		custom_pattern,
+		10);
+
+	/* Usage: Set DP physical test pattern using AMDDP with normal DP panel
+	 * Then plug out DP panel and connect a scope to measure DP PHY signal.
+	 * Need disable interrupt to avoid SW driver disable DP output. This is
+	 * done after the test pattern is set.
+	 */
+	if (valid_test_pattern && disable_hpd)
+		dc_link_disable_hpd(link);
+
+	kfree(wr_buf);
+
+	return bytes_from_user;
+}
+
+static const struct file_operations dp_link_settings_debugfs_fops = {
+	.owner = THIS_MODULE,
+	.read = dp_link_settings_read,
+	.write = dp_link_settings_write,
+	.llseek = default_llseek
+};
+
+static const struct file_operations dp_phy_settings_debugfs_fop = {
+	.owner = THIS_MODULE,
+	.read = dp_phy_settings_read,
+	.write = dp_phy_settings_write,
+	.llseek = default_llseek
+};
+
+static const struct file_operations dp_phy_test_pattern_fops = {
+	.owner = THIS_MODULE,
+	.write = dp_phy_test_pattern_debugfs_write,
+	.llseek = default_llseek
+};
+
+static const struct {
+	char *name;
+	const struct file_operations *fops;
+} dp_debugfs_entries[] = {
+		{"link_settings", &dp_link_settings_debugfs_fops},
+		{"phy_settings", &dp_phy_settings_debugfs_fop},
+		{"test_pattern", &dp_phy_test_pattern_fops}
+};
+
+int connector_debugfs_init(struct amdgpu_dm_connector *connector)
+{
+	int i;
+	struct dentry *ent, *dir = connector->base.debugfs_entry;
+
+	if (connector->base.connector_type == DRM_MODE_CONNECTOR_DisplayPort) {
+		for (i = 0; i < ARRAY_SIZE(dp_debugfs_entries); i++) {
+			ent = debugfs_create_file(dp_debugfs_entries[i].name,
+						  0644,
+						  dir,
+						  connector,
+						  dp_debugfs_entries[i].fops);
+			if (IS_ERR(ent))
+				return PTR_ERR(ent);
+		}
+	}
+
+	return 0;
+}
+
diff --git a/drivers/gpu/drm/amd/powerplay/inc/pp_power_source.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.h
index b43315cc5d58..d9ed1b2aa811 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/pp_power_source.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.h
@@ -1,5 +1,5 @@
 /*
- * Copyright 2015 Advanced Micro Devices, Inc.
+ * Copyright 2018 Advanced Micro Devices, Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -19,18 +19,16 @@
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
  *
+ * Authors: AMD
+ *
  */
 
-#ifndef PP_POWERSOURCE_H
-#define PP_POWERSOURCE_H
+#ifndef __AMDGPU_DM_DEBUGFS_H__
+#define __AMDGPU_DM_DEBUGFS_H__
 
-enum pp_power_source {
-	PP_PowerSource_AC = 0,
-	PP_PowerSource_DC,
-	PP_PowerSource_LimitedPower,
-	PP_PowerSource_LimitedPower_2,
-	PP_PowerSource_Max
-};
+#include "amdgpu.h"
+#include "amdgpu_dm.h"
 
+int connector_debugfs_init(struct amdgpu_dm_connector *connector);
 
 #endif
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
index ca0b08bfa2cf..8403b6a9a77b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
@@ -169,6 +169,11 @@ static void get_payload_table(
 	mutex_unlock(&mst_mgr->payload_lock);
 }
 
+void dm_helpers_dp_update_branch_info(
+	struct dc_context *ctx,
+	const struct dc_link *link)
+{}
+
 /*
  * Writes payload allocation table in immediate downstream device.
  */
@@ -330,11 +335,6 @@ bool dm_helpers_dp_mst_send_payload_allocation(
 	return true;
 }
 
-bool dm_helpers_dc_conn_log(struct dc_context *ctx, struct log_entry *entry, enum dc_log_type event)
-{
-	return true;
-}
-
 void dm_dtn_log_begin(struct dc_context *ctx)
 {}
 
@@ -440,7 +440,7 @@ bool dm_helpers_submit_i2c(
 		return false;
 	}
 
-	msgs = kzalloc(num * sizeof(struct i2c_msg), GFP_KERNEL);
+	msgs = kcalloc(num, sizeof(struct i2c_msg), GFP_KERNEL);
 
 	if (!msgs)
 		return false;
@@ -459,6 +459,22 @@ bool dm_helpers_submit_i2c(
 	return result;
 }
 
+bool dm_helpers_is_dp_sink_present(struct dc_link *link)
+{
+	bool dp_sink_present;
+	struct amdgpu_dm_connector *aconnector = link->priv;
+
+	if (!aconnector) {
+		BUG_ON("Failed to found connector for link!");
+		return true;
+	}
+
+	mutex_lock(&aconnector->dm_dp_aux.aux.hw_mutex);
+	dp_sink_present = dc_link_is_dp_sink_present(link);
+	mutex_unlock(&aconnector->dm_dp_aux.aux.hw_mutex);
+	return dp_sink_present;
+}
+
 enum dc_edid_status dm_helpers_read_local_edid(
 		struct dc_context *ctx,
 		struct dc_link *link,
@@ -502,6 +518,34 @@ enum dc_edid_status dm_helpers_read_local_edid(
 		DRM_ERROR("EDID err: %d, on connector: %s",
 				edid_status,
 				aconnector->base.name);
+	if (link->aux_mode) {
+		union test_request test_request = { {0} };
+		union test_response test_response = { {0} };
+
+		dm_helpers_dp_read_dpcd(ctx,
+					link,
+					DP_TEST_REQUEST,
+					&test_request.raw,
+					sizeof(union test_request));
+
+		if (!test_request.bits.EDID_READ)
+			return edid_status;
+
+		test_response.bits.EDID_CHECKSUM_WRITE = 1;
+
+		dm_helpers_dp_write_dpcd(ctx,
+					link,
+					DP_TEST_EDID_CHECKSUM,
+					&sink->dc_edid.raw_edid[sink->dc_edid.length-1],
+					1);
+
+		dm_helpers_dp_write_dpcd(ctx,
+					link,
+					DP_TEST_RESPONSE,
+					&test_response.raw,
+					sizeof(test_response));
+
+	}
 
 	return edid_status;
 }
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
index 4be21bf54749..a910f01838ab 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
@@ -555,6 +555,9 @@ static inline int dm_irq_state(struct amdgpu_device *adev,
 		return 0;
 	}
 
+	if (acrtc->otg_inst == -1)
+		return 0;
+
 	irq_source = dal_irq_type + acrtc->otg_inst;
 
 	st = (state == AMDGPU_IRQ_STATE_ENABLE);
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
index ace9ad578ca0..9a300732ba37 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
@@ -80,53 +80,72 @@ static void log_dpcd(uint8_t type,
 static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
 				  struct drm_dp_aux_msg *msg)
 {
-	enum i2c_mot_mode mot = (msg->request & DP_AUX_I2C_MOT) ?
-		I2C_MOT_TRUE : I2C_MOT_FALSE;
-	enum ddc_result res;
-	ssize_t read_bytes;
+	ssize_t result = 0;
+	enum i2caux_transaction_action action;
+	enum aux_transaction_type type;
 
 	if (WARN_ON(msg->size > 16))
 		return -E2BIG;
 
 	switch (msg->request & ~DP_AUX_I2C_MOT) {
 	case DP_AUX_NATIVE_READ:
-		read_bytes = dal_ddc_service_read_dpcd_data(
-				TO_DM_AUX(aux)->ddc_service,
-				false,
-				I2C_MOT_UNDEF,
-				msg->address,
-				msg->buffer,
-				msg->size);
-		return read_bytes;
+		type = AUX_TRANSACTION_TYPE_DP;
+		action = I2CAUX_TRANSACTION_ACTION_DP_READ;
+
+		result = dc_link_aux_transfer(TO_DM_AUX(aux)->ddc_service,
+					      msg->address,
+					      &msg->reply,
+					      msg->buffer,
+					      msg->size,
+					      type,
+					      action);
+		break;
 	case DP_AUX_NATIVE_WRITE:
-		res = dal_ddc_service_write_dpcd_data(
-				TO_DM_AUX(aux)->ddc_service,
-				false,
-				I2C_MOT_UNDEF,
-				msg->address,
-				msg->buffer,
-				msg->size);
+		type = AUX_TRANSACTION_TYPE_DP;
+		action = I2CAUX_TRANSACTION_ACTION_DP_WRITE;
+
+		dc_link_aux_transfer(TO_DM_AUX(aux)->ddc_service,
+				     msg->address,
+				     &msg->reply,
+				     msg->buffer,
+				     msg->size,
+				     type,
+				     action);
+		result = msg->size;
 		break;
 	case DP_AUX_I2C_READ:
-		read_bytes = dal_ddc_service_read_dpcd_data(
-				TO_DM_AUX(aux)->ddc_service,
-				true,
-				mot,
-				msg->address,
-				msg->buffer,
-				msg->size);
-		return read_bytes;
+		type = AUX_TRANSACTION_TYPE_I2C;
+		if (msg->request & DP_AUX_I2C_MOT)
+			action = I2CAUX_TRANSACTION_ACTION_I2C_READ_MOT;
+		else
+			action = I2CAUX_TRANSACTION_ACTION_I2C_READ;
+
+		result = dc_link_aux_transfer(TO_DM_AUX(aux)->ddc_service,
+					      msg->address,
+					      &msg->reply,
+					      msg->buffer,
+					      msg->size,
+					      type,
+					      action);
+		break;
 	case DP_AUX_I2C_WRITE:
-		res = dal_ddc_service_write_dpcd_data(
-				TO_DM_AUX(aux)->ddc_service,
-				true,
-				mot,
-				msg->address,
-				msg->buffer,
-				msg->size);
+		type = AUX_TRANSACTION_TYPE_I2C;
+		if (msg->request & DP_AUX_I2C_MOT)
+			action = I2CAUX_TRANSACTION_ACTION_I2C_WRITE_MOT;
+		else
+			action = I2CAUX_TRANSACTION_ACTION_I2C_WRITE;
+
+		dc_link_aux_transfer(TO_DM_AUX(aux)->ddc_service,
+				     msg->address,
+				     &msg->reply,
+				     msg->buffer,
+				     msg->size,
+				     type,
+				     action);
+		result = msg->size;
 		break;
 	default:
-		return 0;
+		return -EINVAL;
 	}
 
 #ifdef TRACE_DPCD
@@ -137,7 +156,10 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
 		 r == DDC_RESULT_SUCESSFULL);
 #endif
 
-	return msg->size;
+	if (result < 0) /* DC doesn't know about kernel error codes */
+		result = -EIO;
+
+	return result;
 }
 
 static enum drm_connector_status
@@ -229,7 +251,7 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector)
 		edid = drm_dp_mst_get_edid(connector, &aconnector->mst_port->mst_mgr, aconnector->port);
 
 		if (!edid) {
-			drm_mode_connector_update_edid_property(
+			drm_connector_update_edid_property(
 				&aconnector->base,
 				NULL);
 			return ret;
@@ -257,7 +279,7 @@ static int dm_dp_mst_get_modes(struct drm_connector *connector)
 					connector, aconnector->edid);
 	}
 
-	drm_mode_connector_update_edid_property(
+	drm_connector_update_edid_property(
 					&aconnector->base, aconnector->edid);
 
 	ret = drm_add_edid_modes(connector, aconnector->edid);
@@ -341,7 +363,7 @@ dm_dp_add_mst_connector(struct drm_dp_mst_topology_mgr *mgr,
 						aconnector, connector->base.id, aconnector->mst_port);
 
 			aconnector->port = port;
-			drm_mode_connector_set_path_property(connector, pathprop);
+			drm_connector_set_path_property(connector, pathprop);
 
 			drm_connector_list_iter_end(&conn_iter);
 			aconnector->mst_connected = true;
@@ -389,7 +411,7 @@ dm_dp_add_mst_connector(struct drm_dp_mst_topology_mgr *mgr,
 		dev->mode_config.tile_property,
 		0);
 
-	drm_mode_connector_set_path_property(connector, pathprop);
+	drm_connector_set_path_property(connector, pathprop);
 
 	/*
 	 * Initialize connector state before adding the connectror to drm and
@@ -437,7 +459,7 @@ static void dm_dp_mst_hotplug(struct drm_dp_mst_topology_mgr *mgr)
 static void dm_dp_mst_link_status_reset(struct drm_connector *connector)
 {
 	mutex_lock(&connector->dev->mode_config.mutex);
-	drm_mode_connector_set_link_status_property(connector, DRM_MODE_LINK_STATUS_BAD);
+	drm_connector_set_link_status_property(connector, DRM_MODE_LINK_STATUS_BAD);
 	mutex_unlock(&connector->dev->mode_config.mutex);
 }
 
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c
new file mode 100644
index 000000000000..c69ae78d82b2
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c
@@ -0,0 +1,562 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ */
+#include <linux/string.h>
+#include <linux/acpi.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/amdgpu_drm.h>
+#include "dm_services.h"
+#include "amdgpu.h"
+#include "amdgpu_dm.h"
+#include "amdgpu_dm_irq.h"
+#include "amdgpu_pm.h"
+#include "dm_pp_smu.h"
+
+
+bool dm_pp_apply_display_requirements(
+		const struct dc_context *ctx,
+		const struct dm_pp_display_configuration *pp_display_cfg)
+{
+	struct amdgpu_device *adev = ctx->driver_context;
+	int i;
+
+	if (adev->pm.dpm_enabled) {
+
+		memset(&adev->pm.pm_display_cfg, 0,
+				sizeof(adev->pm.pm_display_cfg));
+
+		adev->pm.pm_display_cfg.cpu_cc6_disable =
+			pp_display_cfg->cpu_cc6_disable;
+
+		adev->pm.pm_display_cfg.cpu_pstate_disable =
+			pp_display_cfg->cpu_pstate_disable;
+
+		adev->pm.pm_display_cfg.cpu_pstate_separation_time =
+			pp_display_cfg->cpu_pstate_separation_time;
+
+		adev->pm.pm_display_cfg.nb_pstate_switch_disable =
+			pp_display_cfg->nb_pstate_switch_disable;
+
+		adev->pm.pm_display_cfg.num_display =
+				pp_display_cfg->display_count;
+		adev->pm.pm_display_cfg.num_path_including_non_display =
+				pp_display_cfg->display_count;
+
+		adev->pm.pm_display_cfg.min_core_set_clock =
+				pp_display_cfg->min_engine_clock_khz/10;
+		adev->pm.pm_display_cfg.min_core_set_clock_in_sr =
+				pp_display_cfg->min_engine_clock_deep_sleep_khz/10;
+		adev->pm.pm_display_cfg.min_mem_set_clock =
+				pp_display_cfg->min_memory_clock_khz/10;
+
+		adev->pm.pm_display_cfg.min_dcef_deep_sleep_set_clk =
+				pp_display_cfg->min_engine_clock_deep_sleep_khz/10;
+		adev->pm.pm_display_cfg.min_dcef_set_clk =
+				pp_display_cfg->min_dcfclock_khz/10;
+
+		adev->pm.pm_display_cfg.multi_monitor_in_sync =
+				pp_display_cfg->all_displays_in_sync;
+		adev->pm.pm_display_cfg.min_vblank_time =
+				pp_display_cfg->avail_mclk_switch_time_us;
+
+		adev->pm.pm_display_cfg.display_clk =
+				pp_display_cfg->disp_clk_khz/10;
+
+		adev->pm.pm_display_cfg.dce_tolerable_mclk_in_active_latency =
+				pp_display_cfg->avail_mclk_switch_time_in_disp_active_us;
+
+		adev->pm.pm_display_cfg.crtc_index = pp_display_cfg->crtc_index;
+		adev->pm.pm_display_cfg.line_time_in_us =
+				pp_display_cfg->line_time_in_us;
+
+		adev->pm.pm_display_cfg.vrefresh = pp_display_cfg->disp_configs[0].v_refresh;
+		adev->pm.pm_display_cfg.crossfire_display_index = -1;
+		adev->pm.pm_display_cfg.min_bus_bandwidth = 0;
+
+		for (i = 0; i < pp_display_cfg->display_count; i++) {
+			const struct dm_pp_single_disp_config *dc_cfg =
+						&pp_display_cfg->disp_configs[i];
+			adev->pm.pm_display_cfg.displays[i].controller_id = dc_cfg->pipe_idx + 1;
+		}
+
+		/* TODO: complete implementation of
+		 * pp_display_configuration_change().
+		 * Follow example of:
+		 * PHM_StoreDALConfigurationData - powerplay\hwmgr\hardwaremanager.c
+		 * PP_IRI_DisplayConfigurationChange - powerplay\eventmgr\iri.c */
+		if (adev->powerplay.pp_funcs->display_configuration_change)
+			adev->powerplay.pp_funcs->display_configuration_change(
+				adev->powerplay.pp_handle,
+				&adev->pm.pm_display_cfg);
+
+		/* TODO: replace by a separate call to 'apply display cfg'? */
+		amdgpu_pm_compute_clocks(adev);
+	}
+
+	return true;
+}
+
+static void get_default_clock_levels(
+		enum dm_pp_clock_type clk_type,
+		struct dm_pp_clock_levels *clks)
+{
+	uint32_t disp_clks_in_khz[6] = {
+			300000, 400000, 496560, 626090, 685720, 757900 };
+	uint32_t sclks_in_khz[6] = {
+			300000, 360000, 423530, 514290, 626090, 720000 };
+	uint32_t mclks_in_khz[2] = { 333000, 800000 };
+
+	switch (clk_type) {
+	case DM_PP_CLOCK_TYPE_DISPLAY_CLK:
+		clks->num_levels = 6;
+		memmove(clks->clocks_in_khz, disp_clks_in_khz,
+				sizeof(disp_clks_in_khz));
+		break;
+	case DM_PP_CLOCK_TYPE_ENGINE_CLK:
+		clks->num_levels = 6;
+		memmove(clks->clocks_in_khz, sclks_in_khz,
+				sizeof(sclks_in_khz));
+		break;
+	case DM_PP_CLOCK_TYPE_MEMORY_CLK:
+		clks->num_levels = 2;
+		memmove(clks->clocks_in_khz, mclks_in_khz,
+				sizeof(mclks_in_khz));
+		break;
+	default:
+		clks->num_levels = 0;
+		break;
+	}
+}
+
+static enum amd_pp_clock_type dc_to_pp_clock_type(
+		enum dm_pp_clock_type dm_pp_clk_type)
+{
+	enum amd_pp_clock_type amd_pp_clk_type = 0;
+
+	switch (dm_pp_clk_type) {
+	case DM_PP_CLOCK_TYPE_DISPLAY_CLK:
+		amd_pp_clk_type = amd_pp_disp_clock;
+		break;
+	case DM_PP_CLOCK_TYPE_ENGINE_CLK:
+		amd_pp_clk_type = amd_pp_sys_clock;
+		break;
+	case DM_PP_CLOCK_TYPE_MEMORY_CLK:
+		amd_pp_clk_type = amd_pp_mem_clock;
+		break;
+	case DM_PP_CLOCK_TYPE_DCEFCLK:
+		amd_pp_clk_type  = amd_pp_dcef_clock;
+		break;
+	case DM_PP_CLOCK_TYPE_DCFCLK:
+		amd_pp_clk_type = amd_pp_dcf_clock;
+		break;
+	case DM_PP_CLOCK_TYPE_PIXELCLK:
+		amd_pp_clk_type = amd_pp_pixel_clock;
+		break;
+	case DM_PP_CLOCK_TYPE_FCLK:
+		amd_pp_clk_type = amd_pp_f_clock;
+		break;
+	case DM_PP_CLOCK_TYPE_DISPLAYPHYCLK:
+		amd_pp_clk_type = amd_pp_phy_clock;
+		break;
+	case DM_PP_CLOCK_TYPE_DPPCLK:
+		amd_pp_clk_type = amd_pp_dpp_clock;
+		break;
+	default:
+		DRM_ERROR("DM_PPLIB: invalid clock type: %d!\n",
+				dm_pp_clk_type);
+		break;
+	}
+
+	return amd_pp_clk_type;
+}
+
+static enum dm_pp_clocks_state pp_to_dc_powerlevel_state(
+			enum PP_DAL_POWERLEVEL max_clocks_state)
+{
+	switch (max_clocks_state) {
+	case PP_DAL_POWERLEVEL_0:
+		return DM_PP_CLOCKS_DPM_STATE_LEVEL_0;
+	case PP_DAL_POWERLEVEL_1:
+		return DM_PP_CLOCKS_DPM_STATE_LEVEL_1;
+	case PP_DAL_POWERLEVEL_2:
+		return DM_PP_CLOCKS_DPM_STATE_LEVEL_2;
+	case PP_DAL_POWERLEVEL_3:
+		return DM_PP_CLOCKS_DPM_STATE_LEVEL_3;
+	case PP_DAL_POWERLEVEL_4:
+		return DM_PP_CLOCKS_DPM_STATE_LEVEL_4;
+	case PP_DAL_POWERLEVEL_5:
+		return DM_PP_CLOCKS_DPM_STATE_LEVEL_5;
+	case PP_DAL_POWERLEVEL_6:
+		return DM_PP_CLOCKS_DPM_STATE_LEVEL_6;
+	case PP_DAL_POWERLEVEL_7:
+		return DM_PP_CLOCKS_DPM_STATE_LEVEL_7;
+	default:
+		DRM_ERROR("DM_PPLIB: invalid powerlevel state: %d!\n",
+				max_clocks_state);
+		return DM_PP_CLOCKS_STATE_INVALID;
+	}
+}
+
+static void pp_to_dc_clock_levels(
+		const struct amd_pp_clocks *pp_clks,
+		struct dm_pp_clock_levels *dc_clks,
+		enum dm_pp_clock_type dc_clk_type)
+{
+	uint32_t i;
+
+	if (pp_clks->count > DM_PP_MAX_CLOCK_LEVELS) {
+		DRM_INFO("DM_PPLIB: Warning: %s clock: number of levels %d exceeds maximum of %d!\n",
+				DC_DECODE_PP_CLOCK_TYPE(dc_clk_type),
+				pp_clks->count,
+				DM_PP_MAX_CLOCK_LEVELS);
+
+		dc_clks->num_levels = DM_PP_MAX_CLOCK_LEVELS;
+	} else
+		dc_clks->num_levels = pp_clks->count;
+
+	DRM_INFO("DM_PPLIB: values for %s clock\n",
+			DC_DECODE_PP_CLOCK_TYPE(dc_clk_type));
+
+	for (i = 0; i < dc_clks->num_levels; i++) {
+		DRM_INFO("DM_PPLIB:\t %d\n", pp_clks->clock[i]);
+		dc_clks->clocks_in_khz[i] = pp_clks->clock[i];
+	}
+}
+
+static void pp_to_dc_clock_levels_with_latency(
+		const struct pp_clock_levels_with_latency *pp_clks,
+		struct dm_pp_clock_levels_with_latency *clk_level_info,
+		enum dm_pp_clock_type dc_clk_type)
+{
+	uint32_t i;
+
+	if (pp_clks->num_levels > DM_PP_MAX_CLOCK_LEVELS) {
+		DRM_INFO("DM_PPLIB: Warning: %s clock: number of levels %d exceeds maximum of %d!\n",
+				DC_DECODE_PP_CLOCK_TYPE(dc_clk_type),
+				pp_clks->num_levels,
+				DM_PP_MAX_CLOCK_LEVELS);
+
+		clk_level_info->num_levels = DM_PP_MAX_CLOCK_LEVELS;
+	} else
+		clk_level_info->num_levels = pp_clks->num_levels;
+
+	DRM_DEBUG("DM_PPLIB: values for %s clock\n",
+			DC_DECODE_PP_CLOCK_TYPE(dc_clk_type));
+
+	for (i = 0; i < clk_level_info->num_levels; i++) {
+		DRM_DEBUG("DM_PPLIB:\t %d in kHz\n", pp_clks->data[i].clocks_in_khz);
+		clk_level_info->data[i].clocks_in_khz = pp_clks->data[i].clocks_in_khz;
+		clk_level_info->data[i].latency_in_us = pp_clks->data[i].latency_in_us;
+	}
+}
+
+static void pp_to_dc_clock_levels_with_voltage(
+		const struct pp_clock_levels_with_voltage *pp_clks,
+		struct dm_pp_clock_levels_with_voltage *clk_level_info,
+		enum dm_pp_clock_type dc_clk_type)
+{
+	uint32_t i;
+
+	if (pp_clks->num_levels > DM_PP_MAX_CLOCK_LEVELS) {
+		DRM_INFO("DM_PPLIB: Warning: %s clock: number of levels %d exceeds maximum of %d!\n",
+				DC_DECODE_PP_CLOCK_TYPE(dc_clk_type),
+				pp_clks->num_levels,
+				DM_PP_MAX_CLOCK_LEVELS);
+
+		clk_level_info->num_levels = DM_PP_MAX_CLOCK_LEVELS;
+	} else
+		clk_level_info->num_levels = pp_clks->num_levels;
+
+	DRM_INFO("DM_PPLIB: values for %s clock\n",
+			DC_DECODE_PP_CLOCK_TYPE(dc_clk_type));
+
+	for (i = 0; i < clk_level_info->num_levels; i++) {
+		DRM_INFO("DM_PPLIB:\t %d in kHz\n", pp_clks->data[i].clocks_in_khz);
+		clk_level_info->data[i].clocks_in_khz = pp_clks->data[i].clocks_in_khz;
+		clk_level_info->data[i].voltage_in_mv = pp_clks->data[i].voltage_in_mv;
+	}
+}
+
+bool dm_pp_get_clock_levels_by_type(
+		const struct dc_context *ctx,
+		enum dm_pp_clock_type clk_type,
+		struct dm_pp_clock_levels *dc_clks)
+{
+	struct amdgpu_device *adev = ctx->driver_context;
+	void *pp_handle = adev->powerplay.pp_handle;
+	struct amd_pp_clocks pp_clks = { 0 };
+	struct amd_pp_simple_clock_info validation_clks = { 0 };
+	uint32_t i;
+
+	if (adev->powerplay.pp_funcs->get_clock_by_type) {
+		if (adev->powerplay.pp_funcs->get_clock_by_type(pp_handle,
+			dc_to_pp_clock_type(clk_type), &pp_clks)) {
+		/* Error in pplib. Provide default values. */
+			get_default_clock_levels(clk_type, dc_clks);
+			return true;
+		}
+	}
+
+	pp_to_dc_clock_levels(&pp_clks, dc_clks, clk_type);
+
+	if (adev->powerplay.pp_funcs->get_display_mode_validation_clocks) {
+		if (adev->powerplay.pp_funcs->get_display_mode_validation_clocks(
+						pp_handle, &validation_clks)) {
+			/* Error in pplib. Provide default values. */
+			DRM_INFO("DM_PPLIB: Warning: using default validation clocks!\n");
+			validation_clks.engine_max_clock = 72000;
+			validation_clks.memory_max_clock = 80000;
+			validation_clks.level = 0;
+		}
+	}
+
+	DRM_INFO("DM_PPLIB: Validation clocks:\n");
+	DRM_INFO("DM_PPLIB:    engine_max_clock: %d\n",
+			validation_clks.engine_max_clock);
+	DRM_INFO("DM_PPLIB:    memory_max_clock: %d\n",
+			validation_clks.memory_max_clock);
+	DRM_INFO("DM_PPLIB:    level           : %d\n",
+			validation_clks.level);
+
+	/* Translate 10 kHz to kHz. */
+	validation_clks.engine_max_clock *= 10;
+	validation_clks.memory_max_clock *= 10;
+
+	/* Determine the highest non-boosted level from the Validation Clocks */
+	if (clk_type == DM_PP_CLOCK_TYPE_ENGINE_CLK) {
+		for (i = 0; i < dc_clks->num_levels; i++) {
+			if (dc_clks->clocks_in_khz[i] > validation_clks.engine_max_clock) {
+				/* This clock is higher the validation clock.
+				 * Than means the previous one is the highest
+				 * non-boosted one. */
+				DRM_INFO("DM_PPLIB: reducing engine clock level from %d to %d\n",
+						dc_clks->num_levels, i);
+				dc_clks->num_levels = i > 0 ? i : 1;
+				break;
+			}
+		}
+	} else if (clk_type == DM_PP_CLOCK_TYPE_MEMORY_CLK) {
+		for (i = 0; i < dc_clks->num_levels; i++) {
+			if (dc_clks->clocks_in_khz[i] > validation_clks.memory_max_clock) {
+				DRM_INFO("DM_PPLIB: reducing memory clock level from %d to %d\n",
+						dc_clks->num_levels, i);
+				dc_clks->num_levels = i > 0 ? i : 1;
+				break;
+			}
+		}
+	}
+
+	return true;
+}
+
+bool dm_pp_get_clock_levels_by_type_with_latency(
+	const struct dc_context *ctx,
+	enum dm_pp_clock_type clk_type,
+	struct dm_pp_clock_levels_with_latency *clk_level_info)
+{
+	struct amdgpu_device *adev = ctx->driver_context;
+	void *pp_handle = adev->powerplay.pp_handle;
+	struct pp_clock_levels_with_latency pp_clks = { 0 };
+	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
+
+	if (!pp_funcs || !pp_funcs->get_clock_by_type_with_latency)
+		return false;
+
+	if (pp_funcs->get_clock_by_type_with_latency(pp_handle,
+						     dc_to_pp_clock_type(clk_type),
+						     &pp_clks))
+		return false;
+
+	pp_to_dc_clock_levels_with_latency(&pp_clks, clk_level_info, clk_type);
+
+	return true;
+}
+
+bool dm_pp_get_clock_levels_by_type_with_voltage(
+	const struct dc_context *ctx,
+	enum dm_pp_clock_type clk_type,
+	struct dm_pp_clock_levels_with_voltage *clk_level_info)
+{
+	struct amdgpu_device *adev = ctx->driver_context;
+	void *pp_handle = adev->powerplay.pp_handle;
+	struct pp_clock_levels_with_voltage pp_clk_info = {0};
+	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
+
+	if (pp_funcs->get_clock_by_type_with_voltage(pp_handle,
+						     dc_to_pp_clock_type(clk_type),
+						     &pp_clk_info))
+		return false;
+
+	pp_to_dc_clock_levels_with_voltage(&pp_clk_info, clk_level_info, clk_type);
+
+	return true;
+}
+
+bool dm_pp_notify_wm_clock_changes(
+	const struct dc_context *ctx,
+	struct dm_pp_wm_sets_with_clock_ranges *wm_with_clock_ranges)
+{
+	/* TODO: to be implemented */
+	return false;
+}
+
+bool dm_pp_apply_power_level_change_request(
+	const struct dc_context *ctx,
+	struct dm_pp_power_level_change_request *level_change_req)
+{
+	/* TODO: to be implemented */
+	return false;
+}
+
+bool dm_pp_apply_clock_for_voltage_request(
+	const struct dc_context *ctx,
+	struct dm_pp_clock_for_voltage_req *clock_for_voltage_req)
+{
+	struct amdgpu_device *adev = ctx->driver_context;
+	struct pp_display_clock_request pp_clock_request = {0};
+	int ret = 0;
+
+	pp_clock_request.clock_type = dc_to_pp_clock_type(clock_for_voltage_req->clk_type);
+	pp_clock_request.clock_freq_in_khz = clock_for_voltage_req->clocks_in_khz;
+
+	if (!pp_clock_request.clock_type)
+		return false;
+
+	if (adev->powerplay.pp_funcs->display_clock_voltage_request)
+		ret = adev->powerplay.pp_funcs->display_clock_voltage_request(
+			adev->powerplay.pp_handle,
+			&pp_clock_request);
+	if (ret)
+		return false;
+	return true;
+}
+
+bool dm_pp_get_static_clocks(
+	const struct dc_context *ctx,
+	struct dm_pp_static_clock_info *static_clk_info)
+{
+	struct amdgpu_device *adev = ctx->driver_context;
+	struct amd_pp_clock_info pp_clk_info = {0};
+	int ret = 0;
+
+	if (adev->powerplay.pp_funcs->get_current_clocks)
+		ret = adev->powerplay.pp_funcs->get_current_clocks(
+			adev->powerplay.pp_handle,
+			&pp_clk_info);
+	if (ret)
+		return false;
+
+	static_clk_info->max_clocks_state = pp_to_dc_powerlevel_state(pp_clk_info.max_clocks_state);
+	static_clk_info->max_mclk_khz = pp_clk_info.max_memory_clock;
+	static_clk_info->max_sclk_khz = pp_clk_info.max_engine_clock;
+
+	return true;
+}
+
+void pp_rv_set_display_requirement(struct pp_smu *pp,
+		struct pp_smu_display_requirement_rv *req)
+{
+	struct dc_context *ctx = pp->ctx;
+	struct amdgpu_device *adev = ctx->driver_context;
+	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
+
+	if (!pp_funcs || !pp_funcs->display_configuration_changed)
+		return;
+
+	amdgpu_dpm_display_configuration_changed(adev);
+}
+
+void pp_rv_set_wm_ranges(struct pp_smu *pp,
+		struct pp_smu_wm_range_sets *ranges)
+{
+	struct dc_context *ctx = pp->ctx;
+	struct amdgpu_device *adev = ctx->driver_context;
+	void *pp_handle = adev->powerplay.pp_handle;
+	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
+	struct dm_pp_wm_sets_with_clock_ranges_soc15 wm_with_clock_ranges;
+	struct dm_pp_clock_range_for_dmif_wm_set_soc15 *wm_dce_clocks = wm_with_clock_ranges.wm_dmif_clocks_ranges;
+	struct dm_pp_clock_range_for_mcif_wm_set_soc15 *wm_soc_clocks = wm_with_clock_ranges.wm_mcif_clocks_ranges;
+	int32_t i;
+
+	wm_with_clock_ranges.num_wm_dmif_sets = ranges->num_reader_wm_sets;
+	wm_with_clock_ranges.num_wm_mcif_sets = ranges->num_writer_wm_sets;
+
+	for (i = 0; i < wm_with_clock_ranges.num_wm_dmif_sets; i++) {
+		if (ranges->reader_wm_sets[i].wm_inst > 3)
+			wm_dce_clocks[i].wm_set_id = WM_SET_A;
+		else
+			wm_dce_clocks[i].wm_set_id =
+					ranges->reader_wm_sets[i].wm_inst;
+		wm_dce_clocks[i].wm_max_dcfclk_clk_in_khz =
+				ranges->reader_wm_sets[i].max_drain_clk_khz;
+		wm_dce_clocks[i].wm_min_dcfclk_clk_in_khz =
+				ranges->reader_wm_sets[i].min_drain_clk_khz;
+		wm_dce_clocks[i].wm_max_mem_clk_in_khz =
+				ranges->reader_wm_sets[i].max_fill_clk_khz;
+		wm_dce_clocks[i].wm_min_mem_clk_in_khz =
+				ranges->reader_wm_sets[i].min_fill_clk_khz;
+	}
+
+	for (i = 0; i < wm_with_clock_ranges.num_wm_mcif_sets; i++) {
+		if (ranges->writer_wm_sets[i].wm_inst > 3)
+			wm_soc_clocks[i].wm_set_id = WM_SET_A;
+		else
+			wm_soc_clocks[i].wm_set_id =
+					ranges->writer_wm_sets[i].wm_inst;
+		wm_soc_clocks[i].wm_max_socclk_clk_in_khz =
+				ranges->writer_wm_sets[i].max_fill_clk_khz;
+		wm_soc_clocks[i].wm_min_socclk_clk_in_khz =
+				ranges->writer_wm_sets[i].min_fill_clk_khz;
+		wm_soc_clocks[i].wm_max_mem_clk_in_khz =
+				ranges->writer_wm_sets[i].max_drain_clk_khz;
+		wm_soc_clocks[i].wm_min_mem_clk_in_khz =
+				ranges->writer_wm_sets[i].min_drain_clk_khz;
+	}
+
+	pp_funcs->set_watermarks_for_clocks_ranges(pp_handle, &wm_with_clock_ranges);
+}
+
+void pp_rv_set_pme_wa_enable(struct pp_smu *pp)
+{
+	struct dc_context *ctx = pp->ctx;
+	struct amdgpu_device *adev = ctx->driver_context;
+	void *pp_handle = adev->powerplay.pp_handle;
+	const struct amd_pm_funcs *pp_funcs = adev->powerplay.pp_funcs;
+
+	if (!pp_funcs || !pp_funcs->notify_smu_enable_pwe)
+		return;
+
+	pp_funcs->notify_smu_enable_pwe(pp_handle);
+}
+
+void dm_pp_get_funcs_rv(
+		struct dc_context *ctx,
+		struct pp_smu_funcs_rv *funcs)
+{
+	funcs->pp_smu.ctx = ctx;
+	funcs->set_display_requirement = pp_rv_set_display_requirement;
+	funcs->set_wm_ranges = pp_rv_set_wm_ranges;
+	funcs->set_pme_wa_enable = pp_rv_set_pme_wa_enable;
+}
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
index 89342b48be6b..9f0a217603ad 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_services.c
@@ -35,10 +35,13 @@
 #include "amdgpu_dm_irq.h"
 #include "amdgpu_pm.h"
 
-unsigned long long dm_get_timestamp(struct dc_context *ctx)
+
+
+unsigned long long dm_get_elapse_time_in_ns(struct dc_context *ctx,
+		unsigned long long current_time_stamp,
+		unsigned long long last_time_stamp)
 {
-	/* TODO: return actual timestamp */
-	return 0;
+	return current_time_stamp - last_time_stamp;
 }
 
 void dm_perf_trace_timestamp(const char *func_name, unsigned int line)
@@ -71,285 +74,4 @@ bool dm_read_persistent_data(struct dc_context *ctx,
 
 /**** power component interfaces ****/
 
-bool dm_pp_apply_display_requirements(
-		const struct dc_context *ctx,
-		const struct dm_pp_display_configuration *pp_display_cfg)
-{
-	struct amdgpu_device *adev = ctx->driver_context;
-
-	if (adev->pm.dpm_enabled) {
-
-		memset(&adev->pm.pm_display_cfg, 0,
-				sizeof(adev->pm.pm_display_cfg));
-
-		adev->pm.pm_display_cfg.cpu_cc6_disable =
-			pp_display_cfg->cpu_cc6_disable;
-
-		adev->pm.pm_display_cfg.cpu_pstate_disable =
-			pp_display_cfg->cpu_pstate_disable;
-
-		adev->pm.pm_display_cfg.cpu_pstate_separation_time =
-			pp_display_cfg->cpu_pstate_separation_time;
-
-		adev->pm.pm_display_cfg.nb_pstate_switch_disable =
-			pp_display_cfg->nb_pstate_switch_disable;
-
-		adev->pm.pm_display_cfg.num_display =
-				pp_display_cfg->display_count;
-		adev->pm.pm_display_cfg.num_path_including_non_display =
-				pp_display_cfg->display_count;
-
-		adev->pm.pm_display_cfg.min_core_set_clock =
-				pp_display_cfg->min_engine_clock_khz/10;
-		adev->pm.pm_display_cfg.min_core_set_clock_in_sr =
-				pp_display_cfg->min_engine_clock_deep_sleep_khz/10;
-		adev->pm.pm_display_cfg.min_mem_set_clock =
-				pp_display_cfg->min_memory_clock_khz/10;
-
-		adev->pm.pm_display_cfg.multi_monitor_in_sync =
-				pp_display_cfg->all_displays_in_sync;
-		adev->pm.pm_display_cfg.min_vblank_time =
-				pp_display_cfg->avail_mclk_switch_time_us;
-
-		adev->pm.pm_display_cfg.display_clk =
-				pp_display_cfg->disp_clk_khz/10;
-
-		adev->pm.pm_display_cfg.dce_tolerable_mclk_in_active_latency =
-				pp_display_cfg->avail_mclk_switch_time_in_disp_active_us;
-
-		adev->pm.pm_display_cfg.crtc_index = pp_display_cfg->crtc_index;
-		adev->pm.pm_display_cfg.line_time_in_us =
-				pp_display_cfg->line_time_in_us;
-
-		adev->pm.pm_display_cfg.vrefresh = pp_display_cfg->disp_configs[0].v_refresh;
-		adev->pm.pm_display_cfg.crossfire_display_index = -1;
-		adev->pm.pm_display_cfg.min_bus_bandwidth = 0;
-
-		/* TODO: complete implementation of
-		 * pp_display_configuration_change().
-		 * Follow example of:
-		 * PHM_StoreDALConfigurationData - powerplay\hwmgr\hardwaremanager.c
-		 * PP_IRI_DisplayConfigurationChange - powerplay\eventmgr\iri.c */
-		if (adev->powerplay.pp_funcs->display_configuration_change)
-			adev->powerplay.pp_funcs->display_configuration_change(
-				adev->powerplay.pp_handle,
-				&adev->pm.pm_display_cfg);
-
-		/* TODO: replace by a separate call to 'apply display cfg'? */
-		amdgpu_pm_compute_clocks(adev);
-	}
-
-	return true;
-}
-
-static void get_default_clock_levels(
-		enum dm_pp_clock_type clk_type,
-		struct dm_pp_clock_levels *clks)
-{
-	uint32_t disp_clks_in_khz[6] = {
-			300000, 400000, 496560, 626090, 685720, 757900 };
-	uint32_t sclks_in_khz[6] = {
-			300000, 360000, 423530, 514290, 626090, 720000 };
-	uint32_t mclks_in_khz[2] = { 333000, 800000 };
-
-	switch (clk_type) {
-	case DM_PP_CLOCK_TYPE_DISPLAY_CLK:
-		clks->num_levels = 6;
-		memmove(clks->clocks_in_khz, disp_clks_in_khz,
-				sizeof(disp_clks_in_khz));
-		break;
-	case DM_PP_CLOCK_TYPE_ENGINE_CLK:
-		clks->num_levels = 6;
-		memmove(clks->clocks_in_khz, sclks_in_khz,
-				sizeof(sclks_in_khz));
-		break;
-	case DM_PP_CLOCK_TYPE_MEMORY_CLK:
-		clks->num_levels = 2;
-		memmove(clks->clocks_in_khz, mclks_in_khz,
-				sizeof(mclks_in_khz));
-		break;
-	default:
-		clks->num_levels = 0;
-		break;
-	}
-}
-
-static enum amd_pp_clock_type dc_to_pp_clock_type(
-		enum dm_pp_clock_type dm_pp_clk_type)
-{
-	enum amd_pp_clock_type amd_pp_clk_type = 0;
-
-	switch (dm_pp_clk_type) {
-	case DM_PP_CLOCK_TYPE_DISPLAY_CLK:
-		amd_pp_clk_type = amd_pp_disp_clock;
-		break;
-	case DM_PP_CLOCK_TYPE_ENGINE_CLK:
-		amd_pp_clk_type = amd_pp_sys_clock;
-		break;
-	case DM_PP_CLOCK_TYPE_MEMORY_CLK:
-		amd_pp_clk_type = amd_pp_mem_clock;
-		break;
-	default:
-		DRM_ERROR("DM_PPLIB: invalid clock type: %d!\n",
-				dm_pp_clk_type);
-		break;
-	}
-
-	return amd_pp_clk_type;
-}
-
-static void pp_to_dc_clock_levels(
-		const struct amd_pp_clocks *pp_clks,
-		struct dm_pp_clock_levels *dc_clks,
-		enum dm_pp_clock_type dc_clk_type)
-{
-	uint32_t i;
-
-	if (pp_clks->count > DM_PP_MAX_CLOCK_LEVELS) {
-		DRM_INFO("DM_PPLIB: Warning: %s clock: number of levels %d exceeds maximum of %d!\n",
-				DC_DECODE_PP_CLOCK_TYPE(dc_clk_type),
-				pp_clks->count,
-				DM_PP_MAX_CLOCK_LEVELS);
-
-		dc_clks->num_levels = DM_PP_MAX_CLOCK_LEVELS;
-	} else
-		dc_clks->num_levels = pp_clks->count;
-
-	DRM_INFO("DM_PPLIB: values for %s clock\n",
-			DC_DECODE_PP_CLOCK_TYPE(dc_clk_type));
-
-	for (i = 0; i < dc_clks->num_levels; i++) {
-		DRM_INFO("DM_PPLIB:\t %d\n", pp_clks->clock[i]);
-		/* translate 10kHz to kHz */
-		dc_clks->clocks_in_khz[i] = pp_clks->clock[i] * 10;
-	}
-}
-
-bool dm_pp_get_clock_levels_by_type(
-		const struct dc_context *ctx,
-		enum dm_pp_clock_type clk_type,
-		struct dm_pp_clock_levels *dc_clks)
-{
-	struct amdgpu_device *adev = ctx->driver_context;
-	void *pp_handle = adev->powerplay.pp_handle;
-	struct amd_pp_clocks pp_clks = { 0 };
-	struct amd_pp_simple_clock_info validation_clks = { 0 };
-	uint32_t i;
-
-	if (adev->powerplay.pp_funcs->get_clock_by_type) {
-		if (adev->powerplay.pp_funcs->get_clock_by_type(pp_handle,
-			dc_to_pp_clock_type(clk_type), &pp_clks)) {
-		/* Error in pplib. Provide default values. */
-			get_default_clock_levels(clk_type, dc_clks);
-			return true;
-		}
-	}
-
-	pp_to_dc_clock_levels(&pp_clks, dc_clks, clk_type);
-
-	if (adev->powerplay.pp_funcs->get_display_mode_validation_clocks) {
-		if (adev->powerplay.pp_funcs->get_display_mode_validation_clocks(
-						pp_handle, &validation_clks)) {
-			/* Error in pplib. Provide default values. */
-			DRM_INFO("DM_PPLIB: Warning: using default validation clocks!\n");
-			validation_clks.engine_max_clock = 72000;
-			validation_clks.memory_max_clock = 80000;
-			validation_clks.level = 0;
-		}
-	}
-
-	DRM_INFO("DM_PPLIB: Validation clocks:\n");
-	DRM_INFO("DM_PPLIB:    engine_max_clock: %d\n",
-			validation_clks.engine_max_clock);
-	DRM_INFO("DM_PPLIB:    memory_max_clock: %d\n",
-			validation_clks.memory_max_clock);
-	DRM_INFO("DM_PPLIB:    level           : %d\n",
-			validation_clks.level);
-
-	/* Translate 10 kHz to kHz. */
-	validation_clks.engine_max_clock *= 10;
-	validation_clks.memory_max_clock *= 10;
-
-	/* Determine the highest non-boosted level from the Validation Clocks */
-	if (clk_type == DM_PP_CLOCK_TYPE_ENGINE_CLK) {
-		for (i = 0; i < dc_clks->num_levels; i++) {
-			if (dc_clks->clocks_in_khz[i] > validation_clks.engine_max_clock) {
-				/* This clock is higher the validation clock.
-				 * Than means the previous one is the highest
-				 * non-boosted one. */
-				DRM_INFO("DM_PPLIB: reducing engine clock level from %d to %d\n",
-						dc_clks->num_levels, i);
-				dc_clks->num_levels = i > 0 ? i : 1;
-				break;
-			}
-		}
-	} else if (clk_type == DM_PP_CLOCK_TYPE_MEMORY_CLK) {
-		for (i = 0; i < dc_clks->num_levels; i++) {
-			if (dc_clks->clocks_in_khz[i] > validation_clks.memory_max_clock) {
-				DRM_INFO("DM_PPLIB: reducing memory clock level from %d to %d\n",
-						dc_clks->num_levels, i);
-				dc_clks->num_levels = i > 0 ? i : 1;
-				break;
-			}
-		}
-	}
-
-	return true;
-}
-
-bool dm_pp_get_clock_levels_by_type_with_latency(
-	const struct dc_context *ctx,
-	enum dm_pp_clock_type clk_type,
-	struct dm_pp_clock_levels_with_latency *clk_level_info)
-{
-	/* TODO: to be implemented */
-	return false;
-}
-
-bool dm_pp_get_clock_levels_by_type_with_voltage(
-	const struct dc_context *ctx,
-	enum dm_pp_clock_type clk_type,
-	struct dm_pp_clock_levels_with_voltage *clk_level_info)
-{
-	/* TODO: to be implemented */
-	return false;
-}
-
-bool dm_pp_notify_wm_clock_changes(
-	const struct dc_context *ctx,
-	struct dm_pp_wm_sets_with_clock_ranges *wm_with_clock_ranges)
-{
-	/* TODO: to be implemented */
-	return false;
-}
-
-bool dm_pp_apply_power_level_change_request(
-	const struct dc_context *ctx,
-	struct dm_pp_power_level_change_request *level_change_req)
-{
-	/* TODO: to be implemented */
-	return false;
-}
-
-bool dm_pp_apply_clock_for_voltage_request(
-	const struct dc_context *ctx,
-	struct dm_pp_clock_for_voltage_req *clock_for_voltage_req)
-{
-	/* TODO: to be implemented */
-	return false;
-}
-
-bool dm_pp_get_static_clocks(
-	const struct dc_context *ctx,
-	struct dm_pp_static_clock_info *static_clk_info)
-{
-	/* TODO: to be implemented */
-	return false;
-}
-
-void dm_pp_get_funcs_rv(
-		struct dc_context *ctx,
-		struct pp_smu_funcs_rv *funcs)
-{}
 
-/**** end of power component interfaces ****/
diff --git a/drivers/gpu/drm/amd/display/dc/Makefile b/drivers/gpu/drm/amd/display/dc/Makefile
index aed538a4d1ba..532a515fda9a 100644
--- a/drivers/gpu/drm/amd/display/dc/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/Makefile
@@ -25,7 +25,7 @@
 
 DC_LIBS = basics bios calcs dce gpio i2caux irq virtual
 
-ifdef CONFIG_DRM_AMD_DC_DCN1_0
+ifdef CONFIG_X86
 DC_LIBS += dcn10 dml
 endif
 
diff --git a/drivers/gpu/drm/amd/display/dc/basics/Makefile b/drivers/gpu/drm/amd/display/dc/basics/Makefile
index bca33bd9a0d2..a50a76471107 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/basics/Makefile
@@ -24,8 +24,8 @@
 # It provides the general basic services required by other DAL
 # subcomponents.
 
-BASICS = conversion.o fixpt31_32.o fixpt32_32.o \
-	logger.o log_helpers.o vector.o
+BASICS = conversion.o fixpt31_32.o \
+	log_helpers.o vector.o
 
 AMD_DAL_BASICS = $(addprefix $(AMDDALPATH)/dc/basics/,$(BASICS))
 
diff --git a/drivers/gpu/drm/amd/display/dc/basics/conversion.c b/drivers/gpu/drm/amd/display/dc/basics/conversion.c
index 310964915a83..50b47f11875c 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/conversion.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/conversion.c
@@ -41,22 +41,22 @@ uint16_t fixed_point_to_int_frac(
 
 	uint16_t result;
 
-	uint16_t d = (uint16_t)dal_fixed31_32_floor(
-		dal_fixed31_32_abs(
+	uint16_t d = (uint16_t)dc_fixpt_floor(
+		dc_fixpt_abs(
 			arg));
 
 	if (d <= (uint16_t)(1 << integer_bits) - (1 / (uint16_t)divisor))
-		numerator = (uint16_t)dal_fixed31_32_round(
-			dal_fixed31_32_mul_int(
+		numerator = (uint16_t)dc_fixpt_round(
+			dc_fixpt_mul_int(
 				arg,
 				divisor));
 	else {
-		numerator = dal_fixed31_32_floor(
-			dal_fixed31_32_sub(
-				dal_fixed31_32_from_int(
+		numerator = dc_fixpt_floor(
+			dc_fixpt_sub(
+				dc_fixpt_from_int(
 					1LL << integer_bits),
-				dal_fixed31_32_recip(
-					dal_fixed31_32_from_int(
+				dc_fixpt_recip(
+					dc_fixpt_from_int(
 						divisor))));
 	}
 
@@ -66,8 +66,8 @@ uint16_t fixed_point_to_int_frac(
 		result = (uint16_t)(
 		(1 << (integer_bits + fractional_bits + 1)) + numerator);
 
-	if ((result != 0) && dal_fixed31_32_lt(
-		arg, dal_fixed31_32_zero))
+	if ((result != 0) && dc_fixpt_lt(
+		arg, dc_fixpt_zero))
 		result |= 1 << (integer_bits + fractional_bits);
 
 	return result;
@@ -84,15 +84,15 @@ void convert_float_matrix(
 	uint32_t buffer_size)
 {
 	const struct fixed31_32 min_2_13 =
-		dal_fixed31_32_from_fraction(S2D13_MIN, DIVIDER);
+		dc_fixpt_from_fraction(S2D13_MIN, DIVIDER);
 	const struct fixed31_32 max_2_13 =
-		dal_fixed31_32_from_fraction(S2D13_MAX, DIVIDER);
+		dc_fixpt_from_fraction(S2D13_MAX, DIVIDER);
 	uint32_t i;
 
 	for (i = 0; i < buffer_size; ++i) {
 		uint32_t reg_value =
 				fixed_point_to_int_frac(
-					dal_fixed31_32_clamp(
+					dc_fixpt_clamp(
 						flt[i],
 						min_2_13,
 						max_2_13),
diff --git a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
index 8a9bba879207..f28989860fd8 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
@@ -26,13 +26,13 @@
 #include "dm_services.h"
 #include "include/fixed31_32.h"
 
-static inline uint64_t abs_i64(
-	int64_t arg)
+static inline unsigned long long abs_i64(
+	long long arg)
 {
 	if (arg > 0)
-		return (uint64_t)arg;
+		return (unsigned long long)arg;
 	else
-		return (uint64_t)(-arg);
+		return (unsigned long long)(-arg);
 }
 
 /*
@@ -40,12 +40,12 @@ static inline uint64_t abs_i64(
  * result = dividend / divisor
  * *remainder = dividend % divisor
  */
-static inline uint64_t complete_integer_division_u64(
-	uint64_t dividend,
-	uint64_t divisor,
-	uint64_t *remainder)
+static inline unsigned long long complete_integer_division_u64(
+	unsigned long long dividend,
+	unsigned long long divisor,
+	unsigned long long *remainder)
 {
-	uint64_t result;
+	unsigned long long result;
 
 	ASSERT(divisor);
 
@@ -64,30 +64,28 @@ static inline uint64_t complete_integer_division_u64(
 #define GET_FRACTIONAL_PART(x) \
 	(FRACTIONAL_PART_MASK & (x))
 
-struct fixed31_32 dal_fixed31_32_from_fraction(
-	int64_t numerator,
-	int64_t denominator)
+struct fixed31_32 dc_fixpt_from_fraction(long long numerator, long long denominator)
 {
 	struct fixed31_32 res;
 
 	bool arg1_negative = numerator < 0;
 	bool arg2_negative = denominator < 0;
 
-	uint64_t arg1_value = arg1_negative ? -numerator : numerator;
-	uint64_t arg2_value = arg2_negative ? -denominator : denominator;
+	unsigned long long arg1_value = arg1_negative ? -numerator : numerator;
+	unsigned long long arg2_value = arg2_negative ? -denominator : denominator;
 
-	uint64_t remainder;
+	unsigned long long remainder;
 
 	/* determine integer part */
 
-	uint64_t res_value = complete_integer_division_u64(
+	unsigned long long res_value = complete_integer_division_u64(
 		arg1_value, arg2_value, &remainder);
 
 	ASSERT(res_value <= LONG_MAX);
 
 	/* determine fractional part */
 	{
-		uint32_t i = FIXED31_32_BITS_PER_FRACTIONAL_PART;
+		unsigned int i = FIXED31_32_BITS_PER_FRACTIONAL_PART;
 
 		do {
 			remainder <<= 1;
@@ -103,14 +101,14 @@ struct fixed31_32 dal_fixed31_32_from_fraction(
 
 	/* round up LSB */
 	{
-		uint64_t summand = (remainder << 1) >= arg2_value;
+		unsigned long long summand = (remainder << 1) >= arg2_value;
 
 		ASSERT(res_value <= LLONG_MAX - summand);
 
 		res_value += summand;
 	}
 
-	res.value = (int64_t)res_value;
+	res.value = (long long)res_value;
 
 	if (arg1_negative ^ arg2_negative)
 		res.value = -res.value;
@@ -118,79 +116,23 @@ struct fixed31_32 dal_fixed31_32_from_fraction(
 	return res;
 }
 
-struct fixed31_32 dal_fixed31_32_from_int_nonconst(
-	int64_t arg)
-{
-	struct fixed31_32 res;
-
-	ASSERT((LONG_MIN <= arg) && (arg <= LONG_MAX));
-
-	res.value = arg << FIXED31_32_BITS_PER_FRACTIONAL_PART;
-
-	return res;
-}
-
-struct fixed31_32 dal_fixed31_32_shl(
-	struct fixed31_32 arg,
-	uint8_t shift)
-{
-	struct fixed31_32 res;
-
-	ASSERT(((arg.value >= 0) && (arg.value <= LLONG_MAX >> shift)) ||
-		((arg.value < 0) && (arg.value >= LLONG_MIN >> shift)));
-
-	res.value = arg.value << shift;
-
-	return res;
-}
-
-struct fixed31_32 dal_fixed31_32_add(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2)
-{
-	struct fixed31_32 res;
-
-	ASSERT(((arg1.value >= 0) && (LLONG_MAX - arg1.value >= arg2.value)) ||
-		((arg1.value < 0) && (LLONG_MIN - arg1.value <= arg2.value)));
-
-	res.value = arg1.value + arg2.value;
-
-	return res;
-}
-
-struct fixed31_32 dal_fixed31_32_sub(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2)
-{
-	struct fixed31_32 res;
-
-	ASSERT(((arg2.value >= 0) && (LLONG_MIN + arg2.value <= arg1.value)) ||
-		((arg2.value < 0) && (LLONG_MAX + arg2.value >= arg1.value)));
-
-	res.value = arg1.value - arg2.value;
-
-	return res;
-}
-
-struct fixed31_32 dal_fixed31_32_mul(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2)
+struct fixed31_32 dc_fixpt_mul(struct fixed31_32 arg1, struct fixed31_32 arg2)
 {
 	struct fixed31_32 res;
 
 	bool arg1_negative = arg1.value < 0;
 	bool arg2_negative = arg2.value < 0;
 
-	uint64_t arg1_value = arg1_negative ? -arg1.value : arg1.value;
-	uint64_t arg2_value = arg2_negative ? -arg2.value : arg2.value;
+	unsigned long long arg1_value = arg1_negative ? -arg1.value : arg1.value;
+	unsigned long long arg2_value = arg2_negative ? -arg2.value : arg2.value;
 
-	uint64_t arg1_int = GET_INTEGER_PART(arg1_value);
-	uint64_t arg2_int = GET_INTEGER_PART(arg2_value);
+	unsigned long long arg1_int = GET_INTEGER_PART(arg1_value);
+	unsigned long long arg2_int = GET_INTEGER_PART(arg2_value);
 
-	uint64_t arg1_fra = GET_FRACTIONAL_PART(arg1_value);
-	uint64_t arg2_fra = GET_FRACTIONAL_PART(arg2_value);
+	unsigned long long arg1_fra = GET_FRACTIONAL_PART(arg1_value);
+	unsigned long long arg2_fra = GET_FRACTIONAL_PART(arg2_value);
 
-	uint64_t tmp;
+	unsigned long long tmp;
 
 	res.value = arg1_int * arg2_int;
 
@@ -200,22 +142,22 @@ struct fixed31_32 dal_fixed31_32_mul(
 
 	tmp = arg1_int * arg2_fra;
 
-	ASSERT(tmp <= (uint64_t)(LLONG_MAX - res.value));
+	ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value));
 
 	res.value += tmp;
 
 	tmp = arg2_int * arg1_fra;
 
-	ASSERT(tmp <= (uint64_t)(LLONG_MAX - res.value));
+	ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value));
 
 	res.value += tmp;
 
 	tmp = arg1_fra * arg2_fra;
 
 	tmp = (tmp >> FIXED31_32_BITS_PER_FRACTIONAL_PART) +
-		(tmp >= (uint64_t)dal_fixed31_32_half.value);
+		(tmp >= (unsigned long long)dc_fixpt_half.value);
 
-	ASSERT(tmp <= (uint64_t)(LLONG_MAX - res.value));
+	ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value));
 
 	res.value += tmp;
 
@@ -225,18 +167,17 @@ struct fixed31_32 dal_fixed31_32_mul(
 	return res;
 }
 
-struct fixed31_32 dal_fixed31_32_sqr(
-	struct fixed31_32 arg)
+struct fixed31_32 dc_fixpt_sqr(struct fixed31_32 arg)
 {
 	struct fixed31_32 res;
 
-	uint64_t arg_value = abs_i64(arg.value);
+	unsigned long long arg_value = abs_i64(arg.value);
 
-	uint64_t arg_int = GET_INTEGER_PART(arg_value);
+	unsigned long long arg_int = GET_INTEGER_PART(arg_value);
 
-	uint64_t arg_fra = GET_FRACTIONAL_PART(arg_value);
+	unsigned long long arg_fra = GET_FRACTIONAL_PART(arg_value);
 
-	uint64_t tmp;
+	unsigned long long tmp;
 
 	res.value = arg_int * arg_int;
 
@@ -246,28 +187,27 @@ struct fixed31_32 dal_fixed31_32_sqr(
 
 	tmp = arg_int * arg_fra;
 
-	ASSERT(tmp <= (uint64_t)(LLONG_MAX - res.value));
+	ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value));
 
 	res.value += tmp;
 
-	ASSERT(tmp <= (uint64_t)(LLONG_MAX - res.value));
+	ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value));
 
 	res.value += tmp;
 
 	tmp = arg_fra * arg_fra;
 
 	tmp = (tmp >> FIXED31_32_BITS_PER_FRACTIONAL_PART) +
-		(tmp >= (uint64_t)dal_fixed31_32_half.value);
+		(tmp >= (unsigned long long)dc_fixpt_half.value);
 
-	ASSERT(tmp <= (uint64_t)(LLONG_MAX - res.value));
+	ASSERT(tmp <= (unsigned long long)(LLONG_MAX - res.value));
 
 	res.value += tmp;
 
 	return res;
 }
 
-struct fixed31_32 dal_fixed31_32_recip(
-	struct fixed31_32 arg)
+struct fixed31_32 dc_fixpt_recip(struct fixed31_32 arg)
 {
 	/*
 	 * @note
@@ -276,41 +216,40 @@ struct fixed31_32 dal_fixed31_32_recip(
 
 	ASSERT(arg.value);
 
-	return dal_fixed31_32_from_fraction(
-		dal_fixed31_32_one.value,
+	return dc_fixpt_from_fraction(
+		dc_fixpt_one.value,
 		arg.value);
 }
 
-struct fixed31_32 dal_fixed31_32_sinc(
-	struct fixed31_32 arg)
+struct fixed31_32 dc_fixpt_sinc(struct fixed31_32 arg)
 {
 	struct fixed31_32 square;
 
-	struct fixed31_32 res = dal_fixed31_32_one;
+	struct fixed31_32 res = dc_fixpt_one;
 
-	int32_t n = 27;
+	int n = 27;
 
 	struct fixed31_32 arg_norm = arg;
 
-	if (dal_fixed31_32_le(
-		dal_fixed31_32_two_pi,
-		dal_fixed31_32_abs(arg))) {
-		arg_norm = dal_fixed31_32_sub(
+	if (dc_fixpt_le(
+		dc_fixpt_two_pi,
+		dc_fixpt_abs(arg))) {
+		arg_norm = dc_fixpt_sub(
 			arg_norm,
-			dal_fixed31_32_mul_int(
-				dal_fixed31_32_two_pi,
-				(int32_t)div64_s64(
+			dc_fixpt_mul_int(
+				dc_fixpt_two_pi,
+				(int)div64_s64(
 					arg_norm.value,
-					dal_fixed31_32_two_pi.value)));
+					dc_fixpt_two_pi.value)));
 	}
 
-	square = dal_fixed31_32_sqr(arg_norm);
+	square = dc_fixpt_sqr(arg_norm);
 
 	do {
-		res = dal_fixed31_32_sub(
-			dal_fixed31_32_one,
-			dal_fixed31_32_div_int(
-				dal_fixed31_32_mul(
+		res = dc_fixpt_sub(
+			dc_fixpt_one,
+			dc_fixpt_div_int(
+				dc_fixpt_mul(
 					square,
 					res),
 				n * (n - 1)));
@@ -319,37 +258,35 @@ struct fixed31_32 dal_fixed31_32_sinc(
 	} while (n > 2);
 
 	if (arg.value != arg_norm.value)
-		res = dal_fixed31_32_div(
-			dal_fixed31_32_mul(res, arg_norm),
+		res = dc_fixpt_div(
+			dc_fixpt_mul(res, arg_norm),
 			arg);
 
 	return res;
 }
 
-struct fixed31_32 dal_fixed31_32_sin(
-	struct fixed31_32 arg)
+struct fixed31_32 dc_fixpt_sin(struct fixed31_32 arg)
 {
-	return dal_fixed31_32_mul(
+	return dc_fixpt_mul(
 		arg,
-		dal_fixed31_32_sinc(arg));
+		dc_fixpt_sinc(arg));
 }
 
-struct fixed31_32 dal_fixed31_32_cos(
-	struct fixed31_32 arg)
+struct fixed31_32 dc_fixpt_cos(struct fixed31_32 arg)
 {
 	/* TODO implement argument normalization */
 
-	const struct fixed31_32 square = dal_fixed31_32_sqr(arg);
+	const struct fixed31_32 square = dc_fixpt_sqr(arg);
 
-	struct fixed31_32 res = dal_fixed31_32_one;
+	struct fixed31_32 res = dc_fixpt_one;
 
-	int32_t n = 26;
+	int n = 26;
 
 	do {
-		res = dal_fixed31_32_sub(
-			dal_fixed31_32_one,
-			dal_fixed31_32_div_int(
-				dal_fixed31_32_mul(
+		res = dc_fixpt_sub(
+			dc_fixpt_one,
+			dc_fixpt_div_int(
+				dc_fixpt_mul(
 					square,
 					res),
 				n * (n - 1)));
@@ -367,37 +304,35 @@ struct fixed31_32 dal_fixed31_32_cos(
  *
  * Calculated as Taylor series.
  */
-static struct fixed31_32 fixed31_32_exp_from_taylor_series(
-	struct fixed31_32 arg)
+static struct fixed31_32 fixed31_32_exp_from_taylor_series(struct fixed31_32 arg)
 {
-	uint32_t n = 9;
+	unsigned int n = 9;
 
-	struct fixed31_32 res = dal_fixed31_32_from_fraction(
+	struct fixed31_32 res = dc_fixpt_from_fraction(
 		n + 2,
 		n + 1);
 	/* TODO find correct res */
 
-	ASSERT(dal_fixed31_32_lt(arg, dal_fixed31_32_one));
+	ASSERT(dc_fixpt_lt(arg, dc_fixpt_one));
 
 	do
-		res = dal_fixed31_32_add(
-			dal_fixed31_32_one,
-			dal_fixed31_32_div_int(
-				dal_fixed31_32_mul(
+		res = dc_fixpt_add(
+			dc_fixpt_one,
+			dc_fixpt_div_int(
+				dc_fixpt_mul(
 					arg,
 					res),
 				n));
 	while (--n != 1);
 
-	return dal_fixed31_32_add(
-		dal_fixed31_32_one,
-		dal_fixed31_32_mul(
+	return dc_fixpt_add(
+		dc_fixpt_one,
+		dc_fixpt_mul(
 			arg,
 			res));
 }
 
-struct fixed31_32 dal_fixed31_32_exp(
-	struct fixed31_32 arg)
+struct fixed31_32 dc_fixpt_exp(struct fixed31_32 arg)
 {
 	/*
 	 * @brief
@@ -406,44 +341,43 @@ struct fixed31_32 dal_fixed31_32_exp(
 	 * where m = round(x / ln(2)), r = x - m * ln(2)
 	 */
 
-	if (dal_fixed31_32_le(
-		dal_fixed31_32_ln2_div_2,
-		dal_fixed31_32_abs(arg))) {
-		int32_t m = dal_fixed31_32_round(
-			dal_fixed31_32_div(
+	if (dc_fixpt_le(
+		dc_fixpt_ln2_div_2,
+		dc_fixpt_abs(arg))) {
+		int m = dc_fixpt_round(
+			dc_fixpt_div(
 				arg,
-				dal_fixed31_32_ln2));
+				dc_fixpt_ln2));
 
-		struct fixed31_32 r = dal_fixed31_32_sub(
+		struct fixed31_32 r = dc_fixpt_sub(
 			arg,
-			dal_fixed31_32_mul_int(
-				dal_fixed31_32_ln2,
+			dc_fixpt_mul_int(
+				dc_fixpt_ln2,
 				m));
 
 		ASSERT(m != 0);
 
-		ASSERT(dal_fixed31_32_lt(
-			dal_fixed31_32_abs(r),
-			dal_fixed31_32_one));
+		ASSERT(dc_fixpt_lt(
+			dc_fixpt_abs(r),
+			dc_fixpt_one));
 
 		if (m > 0)
-			return dal_fixed31_32_shl(
+			return dc_fixpt_shl(
 				fixed31_32_exp_from_taylor_series(r),
-				(uint8_t)m);
+				(unsigned char)m);
 		else
-			return dal_fixed31_32_div_int(
+			return dc_fixpt_div_int(
 				fixed31_32_exp_from_taylor_series(r),
 				1LL << -m);
 	} else if (arg.value != 0)
 		return fixed31_32_exp_from_taylor_series(arg);
 	else
-		return dal_fixed31_32_one;
+		return dc_fixpt_one;
 }
 
-struct fixed31_32 dal_fixed31_32_log(
-	struct fixed31_32 arg)
+struct fixed31_32 dc_fixpt_log(struct fixed31_32 arg)
 {
-	struct fixed31_32 res = dal_fixed31_32_neg(dal_fixed31_32_one);
+	struct fixed31_32 res = dc_fixpt_neg(dc_fixpt_one);
 	/* TODO improve 1st estimation */
 
 	struct fixed31_32 error;
@@ -453,15 +387,15 @@ struct fixed31_32 dal_fixed31_32_log(
 	/* TODO if arg is zero, return -INF */
 
 	do {
-		struct fixed31_32 res1 = dal_fixed31_32_add(
-			dal_fixed31_32_sub(
+		struct fixed31_32 res1 = dc_fixpt_add(
+			dc_fixpt_sub(
 				res,
-				dal_fixed31_32_one),
-			dal_fixed31_32_div(
+				dc_fixpt_one),
+			dc_fixpt_div(
 				arg,
-				dal_fixed31_32_exp(res)));
+				dc_fixpt_exp(res)));
 
-		error = dal_fixed31_32_sub(
+		error = dc_fixpt_sub(
 			res,
 			res1);
 
@@ -472,78 +406,23 @@ struct fixed31_32 dal_fixed31_32_log(
 	return res;
 }
 
-struct fixed31_32 dal_fixed31_32_pow(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2)
-{
-	return dal_fixed31_32_exp(
-		dal_fixed31_32_mul(
-			dal_fixed31_32_log(arg1),
-			arg2));
-}
-
-int32_t dal_fixed31_32_floor(
-	struct fixed31_32 arg)
-{
-	uint64_t arg_value = abs_i64(arg.value);
-
-	if (arg.value >= 0)
-		return (int32_t)GET_INTEGER_PART(arg_value);
-	else
-		return -(int32_t)GET_INTEGER_PART(arg_value);
-}
-
-int32_t dal_fixed31_32_round(
-	struct fixed31_32 arg)
-{
-	uint64_t arg_value = abs_i64(arg.value);
-
-	const int64_t summand = dal_fixed31_32_half.value;
-
-	ASSERT(LLONG_MAX - (int64_t)arg_value >= summand);
-
-	arg_value += summand;
-
-	if (arg.value >= 0)
-		return (int32_t)GET_INTEGER_PART(arg_value);
-	else
-		return -(int32_t)GET_INTEGER_PART(arg_value);
-}
-
-int32_t dal_fixed31_32_ceil(
-	struct fixed31_32 arg)
-{
-	uint64_t arg_value = abs_i64(arg.value);
-
-	const int64_t summand = dal_fixed31_32_one.value -
-		dal_fixed31_32_epsilon.value;
-
-	ASSERT(LLONG_MAX - (int64_t)arg_value >= summand);
-
-	arg_value += summand;
-
-	if (arg.value >= 0)
-		return (int32_t)GET_INTEGER_PART(arg_value);
-	else
-		return -(int32_t)GET_INTEGER_PART(arg_value);
-}
 
 /* this function is a generic helper to translate fixed point value to
  * specified integer format that will consist of integer_bits integer part and
  * fractional_bits fractional part. For example it is used in
- * dal_fixed31_32_u2d19 to receive 2 bits integer part and 19 bits fractional
+ * dc_fixpt_u2d19 to receive 2 bits integer part and 19 bits fractional
  * part in 32 bits. It is used in hw programming (scaler)
  */
 
-static inline uint32_t ux_dy(
-	int64_t value,
-	uint32_t integer_bits,
-	uint32_t fractional_bits)
+static inline unsigned int ux_dy(
+	long long value,
+	unsigned int integer_bits,
+	unsigned int fractional_bits)
 {
 	/* 1. create mask of integer part */
-	uint32_t result = (1 << integer_bits) - 1;
+	unsigned int result = (1 << integer_bits) - 1;
 	/* 2. mask out fractional part */
-	uint32_t fractional_part = FRACTIONAL_PART_MASK & value;
+	unsigned int fractional_part = FRACTIONAL_PART_MASK & value;
 	/* 3. shrink fixed point integer part to be of integer_bits width*/
 	result &= GET_INTEGER_PART(value);
 	/* 4. make space for fractional part to be filled in after integer */
@@ -554,13 +433,13 @@ static inline uint32_t ux_dy(
 	return result | fractional_part;
 }
 
-static inline uint32_t clamp_ux_dy(
-	int64_t value,
-	uint32_t integer_bits,
-	uint32_t fractional_bits,
-	uint32_t min_clamp)
+static inline unsigned int clamp_ux_dy(
+	long long value,
+	unsigned int integer_bits,
+	unsigned int fractional_bits,
+	unsigned int min_clamp)
 {
-	uint32_t truncated_val = ux_dy(value, integer_bits, fractional_bits);
+	unsigned int truncated_val = ux_dy(value, integer_bits, fractional_bits);
 
 	if (value >= (1LL << (integer_bits + FIXED31_32_BITS_PER_FRACTIONAL_PART)))
 		return (1 << (integer_bits + fractional_bits)) - 1;
@@ -570,35 +449,35 @@ static inline uint32_t clamp_ux_dy(
 		return min_clamp;
 }
 
-uint32_t dal_fixed31_32_u2d19(
-	struct fixed31_32 arg)
+unsigned int dc_fixpt_u3d19(struct fixed31_32 arg)
+{
+	return ux_dy(arg.value, 3, 19);
+}
+
+unsigned int dc_fixpt_u2d19(struct fixed31_32 arg)
 {
 	return ux_dy(arg.value, 2, 19);
 }
 
-uint32_t dal_fixed31_32_u0d19(
-	struct fixed31_32 arg)
+unsigned int dc_fixpt_u0d19(struct fixed31_32 arg)
 {
 	return ux_dy(arg.value, 0, 19);
 }
 
-uint32_t dal_fixed31_32_clamp_u0d14(
-	struct fixed31_32 arg)
+unsigned int dc_fixpt_clamp_u0d14(struct fixed31_32 arg)
 {
 	return clamp_ux_dy(arg.value, 0, 14, 1);
 }
 
-uint32_t dal_fixed31_32_clamp_u0d10(
-	struct fixed31_32 arg)
+unsigned int dc_fixpt_clamp_u0d10(struct fixed31_32 arg)
 {
 	return clamp_ux_dy(arg.value, 0, 10, 1);
 }
 
-int32_t dal_fixed31_32_s4d19(
-	struct fixed31_32 arg)
+int dc_fixpt_s4d19(struct fixed31_32 arg)
 {
 	if (arg.value < 0)
-		return -(int32_t)ux_dy(dal_fixed31_32_abs(arg).value, 4, 19);
+		return -(int)ux_dy(dc_fixpt_abs(arg).value, 4, 19);
 	else
 		return ux_dy(arg.value, 4, 19);
 }
diff --git a/drivers/gpu/drm/amd/display/dc/basics/fixpt32_32.c b/drivers/gpu/drm/amd/display/dc/basics/fixpt32_32.c
deleted file mode 100644
index 4d3aaa82a07b..000000000000
--- a/drivers/gpu/drm/amd/display/dc/basics/fixpt32_32.c
+++ /dev/null
@@ -1,161 +0,0 @@
-/*
- * Copyright 2012-15 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: AMD
- *
- */
-
-#include "dm_services.h"
-#include "include/fixed32_32.h"
-
-static uint64_t u64_div(uint64_t n, uint64_t d)
-{
-	uint32_t i = 0;
-	uint64_t r;
-	uint64_t q = div64_u64_rem(n, d, &r);
-
-	for (i = 0; i < 32; ++i) {
-		uint64_t sbit = q & (1ULL<<63);
-
-		r <<= 1;
-		r |= sbit ? 1 : 0;
-		q <<= 1;
-		if (r >= d) {
-			r -= d;
-			q |= 1;
-		}
-	}
-
-	if (2*r >= d)
-		q += 1;
-	return q;
-}
-
-struct fixed32_32 dal_fixed32_32_from_fraction(uint32_t n, uint32_t d)
-{
-	struct fixed32_32 fx;
-
-	fx.value = u64_div((uint64_t)n << 32, (uint64_t)d << 32);
-	return fx;
-}
-
-struct fixed32_32 dal_fixed32_32_add(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs)
-{
-	struct fixed32_32 fx = {lhs.value + rhs.value};
-
-	ASSERT(fx.value >= rhs.value);
-	return fx;
-}
-
-struct fixed32_32 dal_fixed32_32_add_int(struct fixed32_32 lhs, uint32_t rhs)
-{
-	struct fixed32_32 fx = {lhs.value + ((uint64_t)rhs << 32)};
-
-	ASSERT(fx.value >= (uint64_t)rhs << 32);
-	return fx;
-
-}
-struct fixed32_32 dal_fixed32_32_sub(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs)
-{
-	struct fixed32_32 fx;
-
-	ASSERT(lhs.value >= rhs.value);
-	fx.value = lhs.value - rhs.value;
-	return fx;
-}
-
-struct fixed32_32 dal_fixed32_32_sub_int(struct fixed32_32 lhs, uint32_t rhs)
-{
-	struct fixed32_32 fx;
-
-	ASSERT(lhs.value >= ((uint64_t)rhs<<32));
-	fx.value = lhs.value - ((uint64_t)rhs<<32);
-	return fx;
-}
-
-struct fixed32_32 dal_fixed32_32_mul(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs)
-{
-	struct fixed32_32 fx;
-	uint64_t lhs_int = lhs.value>>32;
-	uint64_t lhs_frac = (uint32_t)lhs.value;
-	uint64_t rhs_int = rhs.value>>32;
-	uint64_t rhs_frac = (uint32_t)rhs.value;
-	uint64_t ahbh = lhs_int * rhs_int;
-	uint64_t ahbl = lhs_int * rhs_frac;
-	uint64_t albh = lhs_frac * rhs_int;
-	uint64_t albl = lhs_frac * rhs_frac;
-
-	ASSERT((ahbh>>32) == 0);
-
-	fx.value = (ahbh<<32) + ahbl + albh + (albl>>32);
-	return fx;
-
-}
-
-struct fixed32_32 dal_fixed32_32_mul_int(struct fixed32_32 lhs, uint32_t rhs)
-{
-	struct fixed32_32 fx;
-	uint64_t lhsi = (lhs.value>>32) * (uint64_t)rhs;
-	uint64_t lhsf;
-
-	ASSERT((lhsi>>32) == 0);
-	lhsf = ((uint32_t)lhs.value) * (uint64_t)rhs;
-	ASSERT((lhsi<<32) + lhsf >= lhsf);
-	fx.value = (lhsi<<32) + lhsf;
-	return fx;
-}
-
-struct fixed32_32 dal_fixed32_32_div(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs)
-{
-	struct fixed32_32 fx;
-
-	fx.value = u64_div(lhs.value, rhs.value);
-	return fx;
-}
-
-struct fixed32_32 dal_fixed32_32_div_int(struct fixed32_32 lhs, uint32_t rhs)
-{
-	struct fixed32_32 fx;
-
-	fx.value = u64_div(lhs.value, (uint64_t)rhs << 32);
-	return fx;
-}
-
-uint32_t dal_fixed32_32_ceil(struct fixed32_32 v)
-{
-	ASSERT((uint32_t)v.value ? (v.value >> 32) + 1 >= 1 : true);
-	return (v.value>>32) + ((uint32_t)v.value ? 1 : 0);
-}
-
-uint32_t dal_fixed32_32_round(struct fixed32_32 v)
-{
-	ASSERT(v.value + (1ULL<<31) >= (1ULL<<31));
-	return (v.value + (1ULL<<31))>>32;
-}
-
diff --git a/drivers/gpu/drm/amd/display/dc/basics/log_helpers.c b/drivers/gpu/drm/amd/display/dc/basics/log_helpers.c
index 854678a0c54b..26583f346c39 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/log_helpers.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/log_helpers.c
@@ -28,76 +28,12 @@
 #include "include/logger_interface.h"
 #include "dm_helpers.h"
 
-#define NUM_ELEMENTS(a) (sizeof(a) / sizeof((a)[0]))
-
-struct dc_signal_type_info {
-	enum signal_type type;
-	char name[MAX_NAME_LEN];
-};
-
-static const struct dc_signal_type_info signal_type_info_tbl[] = {
-		{SIGNAL_TYPE_NONE,             "NC"},
-		{SIGNAL_TYPE_DVI_SINGLE_LINK,  "DVI"},
-		{SIGNAL_TYPE_DVI_DUAL_LINK,    "DDVI"},
-		{SIGNAL_TYPE_HDMI_TYPE_A,      "HDMIA"},
-		{SIGNAL_TYPE_LVDS,             "LVDS"},
-		{SIGNAL_TYPE_RGB,              "VGA"},
-		{SIGNAL_TYPE_DISPLAY_PORT,     "DP"},
-		{SIGNAL_TYPE_DISPLAY_PORT_MST, "MST"},
-		{SIGNAL_TYPE_EDP,              "eDP"},
-		{SIGNAL_TYPE_VIRTUAL,          "Virtual"}
-};
-
-void dc_conn_log(struct dc_context *ctx,
-		const struct dc_link *link,
-		uint8_t *hex_data,
-		int hex_data_count,
-		enum dc_log_type event,
-		const char *msg,
-		...)
+void dc_conn_log_hex_linux(const uint8_t *hex_data, int hex_data_count)
 {
 	int i;
-	va_list args;
-	struct log_entry entry = { 0 };
-	enum signal_type signal;
-
-	if (link->local_sink)
-		signal = link->local_sink->sink_signal;
-	else
-		signal = link->connector_signal;
-
-	if (link->type == dc_connection_mst_branch)
-		signal = SIGNAL_TYPE_DISPLAY_PORT_MST;
-
-	dm_logger_open(ctx->logger, &entry, event);
-
-	for (i = 0; i < NUM_ELEMENTS(signal_type_info_tbl); i++)
-		if (signal == signal_type_info_tbl[i].type)
-			break;
-
-	if (i == NUM_ELEMENTS(signal_type_info_tbl))
-		goto fail;
-
-	dm_logger_append(&entry, "[%s][ConnIdx:%d] ",
-			signal_type_info_tbl[i].name,
-			link->link_index);
-
-	va_start(args, msg);
-	dm_logger_append_va(&entry, msg, args);
-
-	if (entry.buf_offset > 0 &&
-	    entry.buf[entry.buf_offset - 1] == '\n')
-		entry.buf_offset--;
 
 	if (hex_data)
 		for (i = 0; i < hex_data_count; i++)
-			dm_logger_append(&entry, "%2.2X ", hex_data[i]);
-
-	dm_logger_append(&entry, "^\n");
-	dm_helpers_dc_conn_log(ctx, &entry, event);
-
-fail:
-	dm_logger_close(&entry);
-
-	va_end(args);
+			DC_LOG_DEBUG("%2.2X ", hex_data[i]);
 }
+
diff --git a/drivers/gpu/drm/amd/display/dc/basics/logger.c b/drivers/gpu/drm/amd/display/dc/basics/logger.c
deleted file mode 100644
index 31bee054f43a..000000000000
--- a/drivers/gpu/drm/amd/display/dc/basics/logger.c
+++ /dev/null
@@ -1,404 +0,0 @@
-/*
- * Copyright 2012-15 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: AMD
- *
- */
-#include "dm_services.h"
-#include "include/logger_interface.h"
-#include "logger.h"
-
-
-#define NUM_ELEMENTS(a) (sizeof(a) / sizeof((a)[0]))
-
-static const struct dc_log_type_info log_type_info_tbl[] = {
-		{LOG_ERROR,                 "Error"},
-		{LOG_WARNING,               "Warning"},
-		{LOG_DEBUG,		    "Debug"},
-		{LOG_DC,                    "DC_Interface"},
-		{LOG_SURFACE,               "Surface"},
-		{LOG_HW_HOTPLUG,            "HW_Hotplug"},
-		{LOG_HW_LINK_TRAINING,      "HW_LKTN"},
-		{LOG_HW_SET_MODE,           "HW_Mode"},
-		{LOG_HW_RESUME_S3,          "HW_Resume"},
-		{LOG_HW_AUDIO,              "HW_Audio"},
-		{LOG_HW_HPD_IRQ,            "HW_HPDIRQ"},
-		{LOG_MST,                   "MST"},
-		{LOG_SCALER,                "Scaler"},
-		{LOG_BIOS,                  "BIOS"},
-		{LOG_BANDWIDTH_CALCS,       "BWCalcs"},
-		{LOG_BANDWIDTH_VALIDATION,  "BWValidation"},
-		{LOG_I2C_AUX,               "I2C_AUX"},
-		{LOG_SYNC,                  "Sync"},
-		{LOG_BACKLIGHT,             "Backlight"},
-		{LOG_FEATURE_OVERRIDE,      "Override"},
-		{LOG_DETECTION_EDID_PARSER, "Edid"},
-		{LOG_DETECTION_DP_CAPS,     "DP_Caps"},
-		{LOG_RESOURCE,              "Resource"},
-		{LOG_DML,                   "DML"},
-		{LOG_EVENT_MODE_SET,        "Mode"},
-		{LOG_EVENT_DETECTION,       "Detect"},
-		{LOG_EVENT_LINK_TRAINING,   "LKTN"},
-		{LOG_EVENT_LINK_LOSS,       "LinkLoss"},
-		{LOG_EVENT_UNDERFLOW,       "Underflow"},
-		{LOG_IF_TRACE,              "InterfaceTrace"},
-		{LOG_DTN,                   "DTN"},
-		{LOG_PROFILING,             "Profiling"}
-};
-
-
-/* ----------- Object init and destruction ----------- */
-static bool construct(struct dc_context *ctx, struct dal_logger *logger,
-		      uint32_t log_mask)
-{
-	/* malloc buffer and init offsets */
-	logger->log_buffer_size = DAL_LOGGER_BUFFER_MAX_SIZE;
-	logger->log_buffer = kcalloc(logger->log_buffer_size, sizeof(char),
-				     GFP_KERNEL);
-	if (!logger->log_buffer)
-		return false;
-
-	/* Initialize both offsets to start of buffer (empty) */
-	logger->buffer_read_offset = 0;
-	logger->buffer_write_offset = 0;
-
-	logger->open_count = 0;
-
-	logger->flags.bits.ENABLE_CONSOLE = 1;
-	logger->flags.bits.ENABLE_BUFFER = 0;
-
-	logger->ctx = ctx;
-
-	logger->mask = log_mask;
-
-	return true;
-}
-
-static void destruct(struct dal_logger *logger)
-{
-	if (logger->log_buffer) {
-		kfree(logger->log_buffer);
-		logger->log_buffer = NULL;
-	}
-}
-
-struct dal_logger *dal_logger_create(struct dc_context *ctx, uint32_t log_mask)
-{
-	/* malloc struct */
-	struct dal_logger *logger = kzalloc(sizeof(struct dal_logger),
-					    GFP_KERNEL);
-
-	if (!logger)
-		return NULL;
-	if (!construct(ctx, logger, log_mask)) {
-		kfree(logger);
-		return NULL;
-	}
-
-	return logger;
-}
-
-uint32_t dal_logger_destroy(struct dal_logger **logger)
-{
-	if (logger == NULL || *logger == NULL)
-		return 1;
-	destruct(*logger);
-	kfree(*logger);
-	*logger = NULL;
-
-	return 0;
-}
-
-/* ------------------------------------------------------------------------ */
-
-
-static bool dal_logger_should_log(
-	struct dal_logger *logger,
-	enum dc_log_type log_type)
-{
-	if (logger->mask & (1 << log_type))
-		return true;
-
-	return false;
-}
-
-static void log_to_debug_console(struct log_entry *entry)
-{
-	struct dal_logger *logger = entry->logger;
-
-	if (logger->flags.bits.ENABLE_CONSOLE == 0)
-		return;
-
-	if (entry->buf_offset) {
-		switch (entry->type) {
-		case LOG_ERROR:
-			dm_error("%s", entry->buf);
-			break;
-		default:
-			dm_output_to_console("%s", entry->buf);
-			break;
-		}
-	}
-}
-
-/* Print everything unread existing in log_buffer to debug console*/
-void dm_logger_flush_buffer(struct dal_logger *logger, bool should_warn)
-{
-	char *string_start = &logger->log_buffer[logger->buffer_read_offset];
-
-	if (should_warn)
-		dm_output_to_console(
-			"---------------- FLUSHING LOG BUFFER ----------------\n");
-	while (logger->buffer_read_offset < logger->buffer_write_offset) {
-
-		if (logger->log_buffer[logger->buffer_read_offset] == '\0') {
-			dm_output_to_console("%s", string_start);
-			string_start = logger->log_buffer + logger->buffer_read_offset + 1;
-		}
-		logger->buffer_read_offset++;
-	}
-	if (should_warn)
-		dm_output_to_console(
-			"-------------- END FLUSHING LOG BUFFER --------------\n\n");
-}
-
-static void log_to_internal_buffer(struct log_entry *entry)
-{
-
-	uint32_t size = entry->buf_offset;
-	struct dal_logger *logger = entry->logger;
-
-	if (logger->flags.bits.ENABLE_BUFFER == 0)
-		return;
-
-	if (logger->log_buffer == NULL)
-		return;
-
-	if (size > 0 && size < logger->log_buffer_size) {
-
-		int buffer_space = logger->log_buffer_size -
-				logger->buffer_write_offset;
-
-		if (logger->buffer_write_offset == logger->buffer_read_offset) {
-			/* Buffer is empty, start writing at beginning */
-			buffer_space = logger->log_buffer_size;
-			logger->buffer_write_offset = 0;
-			logger->buffer_read_offset = 0;
-		}
-
-		if (buffer_space > size) {
-			/* No wrap around, copy 'size' bytes
-			 * from 'entry->buf' to 'log_buffer'
-			 */
-			memmove(logger->log_buffer +
-					logger->buffer_write_offset,
-					entry->buf, size);
-			logger->buffer_write_offset += size;
-
-		} else {
-			/* Not enough room remaining, we should flush
-			 * existing logs */
-
-			/* Flush existing unread logs to console */
-			dm_logger_flush_buffer(logger, true);
-
-			/* Start writing to beginning of buffer */
-			memmove(logger->log_buffer, entry->buf, size);
-			logger->buffer_write_offset = size;
-			logger->buffer_read_offset = 0;
-		}
-
-	}
-}
-
-static void log_heading(struct log_entry *entry)
-{
-	int j;
-
-	for (j = 0; j < NUM_ELEMENTS(log_type_info_tbl); j++) {
-
-		const struct dc_log_type_info *info = &log_type_info_tbl[j];
-
-		if (info->type == entry->type)
-			dm_logger_append(entry, "[%s]\t", info->name);
-	}
-}
-
-static void append_entry(
-		struct log_entry *entry,
-		char *buffer,
-		uint32_t buf_size)
-{
-	if (!entry->buf ||
-		entry->buf_offset + buf_size > entry->max_buf_bytes
-	) {
-		BREAK_TO_DEBUGGER();
-		return;
-	}
-
-	/* Todo: check if off by 1 byte due to \0 anywhere */
-	memmove(entry->buf + entry->buf_offset, buffer, buf_size);
-	entry->buf_offset += buf_size;
-}
-
-/* ------------------------------------------------------------------------ */
-
-/* Warning: Be careful that 'msg' is null terminated and the total size is
- * less than DAL_LOGGER_BUFFER_MAX_LOG_LINE_SIZE (256) including '\0'
- */
-void dm_logger_write(
-	struct dal_logger *logger,
-	enum dc_log_type log_type,
-	const char *msg,
-	...)
-{
-	if (logger && dal_logger_should_log(logger, log_type)) {
-		uint32_t size;
-		va_list args;
-		char buffer[LOG_MAX_LINE_SIZE];
-		struct log_entry entry;
-
-		va_start(args, msg);
-
-		entry.logger = logger;
-
-		entry.buf = buffer;
-
-		entry.buf_offset = 0;
-		entry.max_buf_bytes = DAL_LOGGER_BUFFER_MAX_SIZE * sizeof(char);
-
-		entry.type = log_type;
-
-		log_heading(&entry);
-
-		size = dm_log_to_buffer(
-			buffer, LOG_MAX_LINE_SIZE - 1, msg, args);
-
-		buffer[entry.buf_offset + size] = '\0';
-		entry.buf_offset += size + 1;
-
-		/* --Flush log_entry buffer-- */
-		/* print to kernel console */
-		log_to_debug_console(&entry);
-		/* log internally for dsat */
-		log_to_internal_buffer(&entry);
-
-		va_end(args);
-	}
-}
-
-/* Same as dm_logger_write, except without open() and close(), which must
- * be done separately.
- */
-void dm_logger_append(
-	struct log_entry *entry,
-	const char *msg,
-	...)
-{
-	va_list args;
-
-	va_start(args, msg);
-	dm_logger_append_va(entry, msg, args);
-	va_end(args);
-}
-
-void dm_logger_append_va(
-	struct log_entry *entry,
-	const char *msg,
-	va_list args)
-{
-	struct dal_logger *logger;
-
-	if (!entry) {
-		BREAK_TO_DEBUGGER();
-		return;
-	}
-
-	logger = entry->logger;
-
-	if (logger && logger->open_count > 0 &&
-		dal_logger_should_log(logger, entry->type)) {
-
-		uint32_t size;
-		char buffer[LOG_MAX_LINE_SIZE];
-
-		size = dm_log_to_buffer(
-			buffer, LOG_MAX_LINE_SIZE, msg, args);
-
-		if (size < LOG_MAX_LINE_SIZE - 1) {
-			append_entry(entry, buffer, size);
-		} else {
-			append_entry(entry, "LOG_ERROR, line too long\n", 27);
-		}
-	}
-}
-
-void dm_logger_open(
-		struct dal_logger *logger,
-		struct log_entry *entry, /* out */
-		enum dc_log_type log_type)
-{
-	if (!entry) {
-		BREAK_TO_DEBUGGER();
-		return;
-	}
-
-	entry->type = log_type;
-	entry->logger = logger;
-
-	entry->buf = kzalloc(DAL_LOGGER_BUFFER_MAX_SIZE * sizeof(char),
-			     GFP_KERNEL);
-
-	entry->buf_offset = 0;
-	entry->max_buf_bytes = DAL_LOGGER_BUFFER_MAX_SIZE * sizeof(char);
-
-	logger->open_count++;
-
-	log_heading(entry);
-}
-
-void dm_logger_close(struct log_entry *entry)
-{
-	struct dal_logger *logger = entry->logger;
-
-	if (logger && logger->open_count > 0) {
-		logger->open_count--;
-	} else {
-		BREAK_TO_DEBUGGER();
-		goto cleanup;
-	}
-
-	/* --Flush log_entry buffer-- */
-	/* print to kernel console */
-	log_to_debug_console(entry);
-	/* log internally for dsat */
-	log_to_internal_buffer(entry);
-
-	/* TODO: Write end heading */
-
-cleanup:
-	if (entry->buf) {
-		kfree(entry->buf);
-		entry->buf = NULL;
-		entry->buf_offset = 0;
-		entry->max_buf_bytes = 0;
-	}
-}
diff --git a/drivers/gpu/drm/amd/display/dc/basics/vector.c b/drivers/gpu/drm/amd/display/dc/basics/vector.c
index 217b8f1f7bf6..d28e9cf0e961 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/vector.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/vector.c
@@ -40,7 +40,7 @@ bool dal_vector_construct(
 		return false;
 	}
 
-	vector->container = kzalloc(struct_size * capacity, GFP_KERNEL);
+	vector->container = kcalloc(capacity, struct_size, GFP_KERNEL);
 	if (vector->container == NULL)
 		return false;
 	vector->capacity = capacity;
@@ -67,7 +67,7 @@ bool dal_vector_presized_costruct(
 		return false;
 	}
 
-	vector->container = kzalloc(struct_size * count, GFP_KERNEL);
+	vector->container = kcalloc(count, struct_size, GFP_KERNEL);
 
 	if (vector->container == NULL)
 		return false;
diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c
index c7f0b27e457e..be8a2494355a 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c
@@ -3762,6 +3762,200 @@ static struct integrated_info *bios_parser_create_integrated_info(
 	return NULL;
 }
 
+enum bp_result update_slot_layout_info(
+	struct dc_bios *dcb,
+	unsigned int i,
+	struct slot_layout_info *slot_layout_info,
+	unsigned int record_offset)
+{
+	unsigned int j;
+	struct bios_parser *bp;
+	ATOM_BRACKET_LAYOUT_RECORD *record;
+	ATOM_COMMON_RECORD_HEADER *record_header;
+	enum bp_result result = BP_RESULT_NORECORD;
+
+	bp = BP_FROM_DCB(dcb);
+	record = NULL;
+	record_header = NULL;
+
+	for (;;) {
+
+		record_header = (ATOM_COMMON_RECORD_HEADER *)
+			GET_IMAGE(ATOM_COMMON_RECORD_HEADER, record_offset);
+		if (record_header == NULL) {
+			result = BP_RESULT_BADBIOSTABLE;
+			break;
+		}
+
+		/* the end of the list */
+		if (record_header->ucRecordType == 0xff ||
+			record_header->ucRecordSize == 0)	{
+			break;
+		}
+
+		if (record_header->ucRecordType ==
+			ATOM_BRACKET_LAYOUT_RECORD_TYPE &&
+			sizeof(ATOM_BRACKET_LAYOUT_RECORD)
+			<= record_header->ucRecordSize) {
+			record = (ATOM_BRACKET_LAYOUT_RECORD *)
+				(record_header);
+			result = BP_RESULT_OK;
+			break;
+		}
+
+		record_offset += record_header->ucRecordSize;
+	}
+
+	/* return if the record not found */
+	if (result != BP_RESULT_OK)
+		return result;
+
+	/* get slot sizes */
+	slot_layout_info->length = record->ucLength;
+	slot_layout_info->width = record->ucWidth;
+
+	/* get info for each connector in the slot */
+	slot_layout_info->num_of_connectors = record->ucConnNum;
+	for (j = 0; j < slot_layout_info->num_of_connectors; ++j) {
+		slot_layout_info->connectors[j].connector_type =
+			(enum connector_layout_type)
+			(record->asConnInfo[j].ucConnectorType);
+		switch (record->asConnInfo[j].ucConnectorType) {
+		case CONNECTOR_TYPE_DVI_D:
+			slot_layout_info->connectors[j].connector_type =
+				CONNECTOR_LAYOUT_TYPE_DVI_D;
+			slot_layout_info->connectors[j].length =
+				CONNECTOR_SIZE_DVI;
+			break;
+
+		case CONNECTOR_TYPE_HDMI:
+			slot_layout_info->connectors[j].connector_type =
+				CONNECTOR_LAYOUT_TYPE_HDMI;
+			slot_layout_info->connectors[j].length =
+				CONNECTOR_SIZE_HDMI;
+			break;
+
+		case CONNECTOR_TYPE_DISPLAY_PORT:
+			slot_layout_info->connectors[j].connector_type =
+				CONNECTOR_LAYOUT_TYPE_DP;
+			slot_layout_info->connectors[j].length =
+				CONNECTOR_SIZE_DP;
+			break;
+
+		case CONNECTOR_TYPE_MINI_DISPLAY_PORT:
+			slot_layout_info->connectors[j].connector_type =
+				CONNECTOR_LAYOUT_TYPE_MINI_DP;
+			slot_layout_info->connectors[j].length =
+				CONNECTOR_SIZE_MINI_DP;
+			break;
+
+		default:
+			slot_layout_info->connectors[j].connector_type =
+				CONNECTOR_LAYOUT_TYPE_UNKNOWN;
+			slot_layout_info->connectors[j].length =
+				CONNECTOR_SIZE_UNKNOWN;
+		}
+
+		slot_layout_info->connectors[j].position =
+			record->asConnInfo[j].ucPosition;
+		slot_layout_info->connectors[j].connector_id =
+			object_id_from_bios_object_id(
+				record->asConnInfo[j].usConnectorObjectId);
+	}
+	return result;
+}
+
+
+enum bp_result get_bracket_layout_record(
+	struct dc_bios *dcb,
+	unsigned int bracket_layout_id,
+	struct slot_layout_info *slot_layout_info)
+{
+	unsigned int i;
+	unsigned int record_offset;
+	struct bios_parser *bp;
+	enum bp_result result;
+	ATOM_OBJECT *object;
+	ATOM_OBJECT_TABLE *object_table;
+	unsigned int genericTableOffset;
+
+	bp = BP_FROM_DCB(dcb);
+	object = NULL;
+	if (slot_layout_info == NULL) {
+		DC_LOG_DETECTION_EDID_PARSER("Invalid slot_layout_info\n");
+		return BP_RESULT_BADINPUT;
+	}
+
+
+	genericTableOffset = bp->object_info_tbl_offset +
+		bp->object_info_tbl.v1_3->usMiscObjectTableOffset;
+	object_table = (ATOM_OBJECT_TABLE *)
+		GET_IMAGE(ATOM_OBJECT_TABLE, genericTableOffset);
+	if (!object_table)
+		return BP_RESULT_FAILURE;
+
+	result = BP_RESULT_NORECORD;
+	for (i = 0; i < object_table->ucNumberOfObjects; ++i) {
+
+		if (bracket_layout_id ==
+			object_table->asObjects[i].usObjectID) {
+
+			object = &object_table->asObjects[i];
+			record_offset = object->usRecordOffset +
+				bp->object_info_tbl_offset;
+
+			result = update_slot_layout_info(dcb, i,
+				slot_layout_info, record_offset);
+			break;
+		}
+	}
+	return result;
+}
+
+static enum bp_result bios_get_board_layout_info(
+	struct dc_bios *dcb,
+	struct board_layout_info *board_layout_info)
+{
+	unsigned int i;
+	struct bios_parser *bp;
+	enum bp_result record_result;
+
+	const unsigned int slot_index_to_vbios_id[MAX_BOARD_SLOTS] = {
+		GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID1,
+		GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID2,
+		0, 0
+	};
+
+	bp = BP_FROM_DCB(dcb);
+	if (board_layout_info == NULL) {
+		DC_LOG_DETECTION_EDID_PARSER("Invalid board_layout_info\n");
+		return BP_RESULT_BADINPUT;
+	}
+
+	board_layout_info->num_of_slots = 0;
+
+	for (i = 0; i < MAX_BOARD_SLOTS; ++i) {
+		record_result = get_bracket_layout_record(dcb,
+			slot_index_to_vbios_id[i],
+			&board_layout_info->slots[i]);
+
+		if (record_result == BP_RESULT_NORECORD && i > 0)
+			break; /* no more slots present in bios */
+		else if (record_result != BP_RESULT_OK)
+			return record_result;  /* fail */
+
+		++board_layout_info->num_of_slots;
+	}
+
+	/* all data is valid */
+	board_layout_info->is_number_of_slots_valid = 1;
+	board_layout_info->is_slots_size_valid = 1;
+	board_layout_info->is_connector_offsets_valid = 1;
+	board_layout_info->is_connector_lengths_valid = 1;
+
+	return BP_RESULT_OK;
+}
+
 /******************************************************************************/
 
 static const struct dc_vbios_funcs vbios_funcs = {
@@ -3836,6 +4030,8 @@ static const struct dc_vbios_funcs vbios_funcs = {
 	.post_init = bios_parser_post_init,  /* patch vbios table for mxm module by reading i2c */
 
 	.bios_parser_destroy = bios_parser_destroy,
+
+	.get_board_layout_info = bios_get_board_layout_info,
 };
 
 static bool bios_parser_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
index 985fe8c22875..eab007e1793c 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
@@ -43,6 +43,29 @@
 #include "bios_parser_interface.h"
 
 #include "bios_parser_common.h"
+
+/* Temporarily add in defines until ObjectID.h patch is updated in a few days */
+#ifndef GENERIC_OBJECT_ID_BRACKET_LAYOUT
+#define GENERIC_OBJECT_ID_BRACKET_LAYOUT          0x05
+#endif /* GENERIC_OBJECT_ID_BRACKET_LAYOUT */
+
+#ifndef GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID1
+#define GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID1	\
+	(GRAPH_OBJECT_TYPE_GENERIC << OBJECT_TYPE_SHIFT |\
+	GRAPH_OBJECT_ENUM_ID1 << ENUM_ID_SHIFT |\
+	GENERIC_OBJECT_ID_BRACKET_LAYOUT << OBJECT_ID_SHIFT)
+#endif /* GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID1 */
+
+#ifndef GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID2
+#define GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID2	\
+	(GRAPH_OBJECT_TYPE_GENERIC << OBJECT_TYPE_SHIFT |\
+	GRAPH_OBJECT_ENUM_ID2 << ENUM_ID_SHIFT |\
+	GENERIC_OBJECT_ID_BRACKET_LAYOUT << OBJECT_ID_SHIFT)
+#endif /* GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID2 */
+
+#define DC_LOGGER \
+	bp->base.ctx->logger
+
 #define LAST_RECORD_TYPE 0xff
 #define SMU9_SYSPLL0_ID  0
 
@@ -70,6 +93,10 @@ static enum bp_result get_firmware_info_v3_1(
 	struct bios_parser *bp,
 	struct dc_firmware_info *info);
 
+static enum bp_result get_firmware_info_v3_2(
+	struct bios_parser *bp,
+	struct dc_firmware_info *info);
+
 static struct atom_hpd_int_record *get_hpd_record(struct bios_parser *bp,
 		struct atom_display_object_path_v2 *object);
 
@@ -82,7 +109,6 @@ static struct atom_encoder_caps_record *get_encoder_cap_record(
 
 #define DATA_TABLES(table) (bp->master_data_tbl->listOfdatatables.table)
 
-
 static void destruct(struct bios_parser *bp)
 {
 	kfree(bp->base.bios_local_image);
@@ -652,7 +678,7 @@ static enum bp_result bios_parser_get_gpio_pin_info(
 		return BP_RESULT_BADBIOSTABLE;
 
 	if (sizeof(struct atom_common_table_header) +
-			sizeof(struct atom_gpio_pin_lut_v2_1)
+			sizeof(struct atom_gpio_pin_assignment)
 			> le16_to_cpu(header->table_header.structuresize))
 		return BP_RESULT_BADBIOSTABLE;
 
@@ -1321,9 +1347,14 @@ static enum bp_result bios_parser_get_firmware_info(
 		case 3:
 			switch (revision.minor) {
 			case 1:
-			case 2:
 				result = get_firmware_info_v3_1(bp, info);
 				break;
+			case 2:
+				result = get_firmware_info_v3_2(bp, info);
+				break;
+			case 3:
+				result = get_firmware_info_v3_2(bp, info);
+				break;
 			default:
 				break;
 			}
@@ -1383,6 +1414,84 @@ static enum bp_result get_firmware_info_v3_1(
 	return BP_RESULT_OK;
 }
 
+static enum bp_result get_firmware_info_v3_2(
+	struct bios_parser *bp,
+	struct dc_firmware_info *info)
+{
+	struct atom_firmware_info_v3_2 *firmware_info;
+	struct atom_display_controller_info_v4_1 *dce_info = NULL;
+	struct atom_common_table_header *header;
+	struct atom_data_revision revision;
+	struct atom_smu_info_v3_2 *smu_info_v3_2 = NULL;
+	struct atom_smu_info_v3_3 *smu_info_v3_3 = NULL;
+
+	if (!info)
+		return BP_RESULT_BADINPUT;
+
+	firmware_info = GET_IMAGE(struct atom_firmware_info_v3_2,
+			DATA_TABLES(firmwareinfo));
+
+	dce_info = GET_IMAGE(struct atom_display_controller_info_v4_1,
+			DATA_TABLES(dce_info));
+
+	if (!firmware_info || !dce_info)
+		return BP_RESULT_BADBIOSTABLE;
+
+	memset(info, 0, sizeof(*info));
+
+	header = GET_IMAGE(struct atom_common_table_header,
+					DATA_TABLES(smu_info));
+	get_atom_data_table_revision(header, &revision);
+
+	if (revision.minor == 2) {
+		/* Vega12 */
+		smu_info_v3_2 = GET_IMAGE(struct atom_smu_info_v3_2,
+							DATA_TABLES(smu_info));
+
+		if (!smu_info_v3_2)
+			return BP_RESULT_BADBIOSTABLE;
+
+		info->default_engine_clk = smu_info_v3_2->bootup_dcefclk_10khz * 10;
+	} else if (revision.minor == 3) {
+		/* Vega20 */
+		smu_info_v3_3 = GET_IMAGE(struct atom_smu_info_v3_3,
+							DATA_TABLES(smu_info));
+
+		if (!smu_info_v3_3)
+			return BP_RESULT_BADBIOSTABLE;
+
+		info->default_engine_clk = smu_info_v3_3->bootup_dcefclk_10khz * 10;
+	}
+
+	 // We need to convert from 10KHz units into KHz units.
+	info->default_memory_clk = firmware_info->bootup_mclk_in10khz * 10;
+
+	 /* 27MHz for Vega10 & Vega12; 100MHz for Vega20 */
+	info->pll_info.crystal_frequency = dce_info->dce_refclk_10khz * 10;
+	/* Hardcode frequency if BIOS gives no DCE Ref Clk */
+	if (info->pll_info.crystal_frequency == 0) {
+		if (revision.minor == 2)
+			info->pll_info.crystal_frequency = 27000;
+		else if (revision.minor == 3)
+			info->pll_info.crystal_frequency = 100000;
+	}
+	/*dp_phy_ref_clk is not correct for atom_display_controller_info_v4_2, but we don't use it*/
+	info->dp_phy_ref_clk     = dce_info->dpphy_refclk_10khz * 10;
+	info->i2c_engine_ref_clk = dce_info->i2c_engine_refclk_10khz * 10;
+
+	/* Get GPU PLL VCO Clock */
+	if (bp->cmd_tbl.get_smu_clock_info != NULL) {
+		if (revision.minor == 2)
+			info->smu_gpu_pll_output_freq =
+					bp->cmd_tbl.get_smu_clock_info(bp, SMU9_SYSPLL0_ID) * 10;
+		else if (revision.minor == 3)
+			info->smu_gpu_pll_output_freq =
+					bp->cmd_tbl.get_smu_clock_info(bp, SMU11_SYSPLL3_0_ID) * 10;
+	}
+
+	return BP_RESULT_OK;
+}
+
 static enum bp_result bios_parser_get_encoder_cap_info(
 	struct dc_bios *dcb,
 	struct graphics_object_id object_id,
@@ -1767,6 +1876,198 @@ static struct integrated_info *bios_parser_create_integrated_info(
 	return NULL;
 }
 
+static enum bp_result update_slot_layout_info(
+	struct dc_bios *dcb,
+	unsigned int i,
+	struct slot_layout_info *slot_layout_info)
+{
+	unsigned int record_offset;
+	unsigned int j;
+	struct atom_display_object_path_v2 *object;
+	struct atom_bracket_layout_record *record;
+	struct atom_common_record_header *record_header;
+	enum bp_result result;
+	struct bios_parser *bp;
+	struct object_info_table *tbl;
+	struct display_object_info_table_v1_4 *v1_4;
+
+	record = NULL;
+	record_header = NULL;
+	result = BP_RESULT_NORECORD;
+
+	bp = BP_FROM_DCB(dcb);
+	tbl = &bp->object_info_tbl;
+	v1_4 = tbl->v1_4;
+
+	object = &v1_4->display_path[i];
+	record_offset = (unsigned int)
+		(object->disp_recordoffset) +
+		(unsigned int)(bp->object_info_tbl_offset);
+
+	for (;;) {
+
+		record_header = (struct atom_common_record_header *)
+			GET_IMAGE(struct atom_common_record_header,
+			record_offset);
+		if (record_header == NULL) {
+			result = BP_RESULT_BADBIOSTABLE;
+			break;
+		}
+
+		/* the end of the list */
+		if (record_header->record_type == 0xff ||
+			record_header->record_size == 0)	{
+			break;
+		}
+
+		if (record_header->record_type ==
+			ATOM_BRACKET_LAYOUT_RECORD_TYPE &&
+			sizeof(struct atom_bracket_layout_record)
+			<= record_header->record_size) {
+			record = (struct atom_bracket_layout_record *)
+				(record_header);
+			result = BP_RESULT_OK;
+			break;
+		}
+
+		record_offset += record_header->record_size;
+	}
+
+	/* return if the record not found */
+	if (result != BP_RESULT_OK)
+		return result;
+
+	/* get slot sizes */
+	slot_layout_info->length = record->bracketlen;
+	slot_layout_info->width = record->bracketwidth;
+
+	/* get info for each connector in the slot */
+	slot_layout_info->num_of_connectors = record->conn_num;
+	for (j = 0; j < slot_layout_info->num_of_connectors; ++j) {
+		slot_layout_info->connectors[j].connector_type =
+			(enum connector_layout_type)
+			(record->conn_info[j].connector_type);
+		switch (record->conn_info[j].connector_type) {
+		case CONNECTOR_TYPE_DVI_D:
+			slot_layout_info->connectors[j].connector_type =
+				CONNECTOR_LAYOUT_TYPE_DVI_D;
+			slot_layout_info->connectors[j].length =
+				CONNECTOR_SIZE_DVI;
+			break;
+
+		case CONNECTOR_TYPE_HDMI:
+			slot_layout_info->connectors[j].connector_type =
+				CONNECTOR_LAYOUT_TYPE_HDMI;
+			slot_layout_info->connectors[j].length =
+				CONNECTOR_SIZE_HDMI;
+			break;
+
+		case CONNECTOR_TYPE_DISPLAY_PORT:
+			slot_layout_info->connectors[j].connector_type =
+				CONNECTOR_LAYOUT_TYPE_DP;
+			slot_layout_info->connectors[j].length =
+				CONNECTOR_SIZE_DP;
+			break;
+
+		case CONNECTOR_TYPE_MINI_DISPLAY_PORT:
+			slot_layout_info->connectors[j].connector_type =
+				CONNECTOR_LAYOUT_TYPE_MINI_DP;
+			slot_layout_info->connectors[j].length =
+				CONNECTOR_SIZE_MINI_DP;
+			break;
+
+		default:
+			slot_layout_info->connectors[j].connector_type =
+				CONNECTOR_LAYOUT_TYPE_UNKNOWN;
+			slot_layout_info->connectors[j].length =
+				CONNECTOR_SIZE_UNKNOWN;
+		}
+
+		slot_layout_info->connectors[j].position =
+			record->conn_info[j].position;
+		slot_layout_info->connectors[j].connector_id =
+			object_id_from_bios_object_id(
+				record->conn_info[j].connectorobjid);
+	}
+	return result;
+}
+
+
+static enum bp_result get_bracket_layout_record(
+	struct dc_bios *dcb,
+	unsigned int bracket_layout_id,
+	struct slot_layout_info *slot_layout_info)
+{
+	unsigned int i;
+	struct bios_parser *bp = BP_FROM_DCB(dcb);
+	enum bp_result result;
+	struct object_info_table *tbl;
+	struct display_object_info_table_v1_4 *v1_4;
+
+	if (slot_layout_info == NULL) {
+		DC_LOG_DETECTION_EDID_PARSER("Invalid slot_layout_info\n");
+		return BP_RESULT_BADINPUT;
+	}
+	tbl = &bp->object_info_tbl;
+	v1_4 = tbl->v1_4;
+
+	result = BP_RESULT_NORECORD;
+	for (i = 0; i < v1_4->number_of_path; ++i)	{
+
+		if (bracket_layout_id ==
+			v1_4->display_path[i].display_objid) {
+			result = update_slot_layout_info(dcb, i,
+				slot_layout_info);
+			break;
+		}
+	}
+	return result;
+}
+
+static enum bp_result bios_get_board_layout_info(
+	struct dc_bios *dcb,
+	struct board_layout_info *board_layout_info)
+{
+	unsigned int i;
+	struct bios_parser *bp;
+	enum bp_result record_result;
+
+	const unsigned int slot_index_to_vbios_id[MAX_BOARD_SLOTS] = {
+		GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID1,
+		GENERICOBJECT_BRACKET_LAYOUT_ENUM_ID2,
+		0, 0
+	};
+
+	bp = BP_FROM_DCB(dcb);
+	if (board_layout_info == NULL) {
+		DC_LOG_DETECTION_EDID_PARSER("Invalid board_layout_info\n");
+		return BP_RESULT_BADINPUT;
+	}
+
+	board_layout_info->num_of_slots = 0;
+
+	for (i = 0; i < MAX_BOARD_SLOTS; ++i) {
+		record_result = get_bracket_layout_record(dcb,
+			slot_index_to_vbios_id[i],
+			&board_layout_info->slots[i]);
+
+		if (record_result == BP_RESULT_NORECORD && i > 0)
+			break; /* no more slots present in bios */
+		else if (record_result != BP_RESULT_OK)
+			return record_result;  /* fail */
+
+		++board_layout_info->num_of_slots;
+	}
+
+	/* all data is valid */
+	board_layout_info->is_number_of_slots_valid = 1;
+	board_layout_info->is_slots_size_valid = 1;
+	board_layout_info->is_connector_offsets_valid = 1;
+	board_layout_info->is_connector_lengths_valid = 1;
+
+	return BP_RESULT_OK;
+}
+
 static const struct dc_vbios_funcs vbios_funcs = {
 	.get_connectors_number = bios_parser_get_connectors_number,
 
@@ -1838,6 +2139,8 @@ static const struct dc_vbios_funcs vbios_funcs = {
 	.bios_parser_destroy = firmware_parser_destroy,
 
 	.get_smu_clock_info = bios_parser_get_smu_clock_info,
+
+	.get_board_layout_info = bios_get_board_layout_info,
 };
 
 static bool bios_parser_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table.c b/drivers/gpu/drm/amd/display/dc/bios/command_table.c
index 4b5fdd577848..651e1fd4622f 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/command_table.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/command_table.c
@@ -24,7 +24,7 @@
  */
 
 #include "dm_services.h"
-
+#include "amdgpu.h"
 #include "atom.h"
 
 #include "include/bios_parser_interface.h"
@@ -35,16 +35,16 @@
 #include "bios_parser_types_internal.h"
 
 #define EXEC_BIOS_CMD_TABLE(command, params)\
-	(cgs_atom_exec_cmd_table(bp->base.ctx->cgs_device, \
+	(amdgpu_atom_execute_table(((struct amdgpu_device *)bp->base.ctx->driver_context)->mode_info.atom_context, \
 		GetIndexIntoMasterTable(COMMAND, command), \
-		&params) == 0)
+		(uint32_t *)&params) == 0)
 
 #define BIOS_CMD_TABLE_REVISION(command, frev, crev)\
-	cgs_atom_get_cmd_table_revs(bp->base.ctx->cgs_device, \
+	amdgpu_atom_parse_cmd_header(((struct amdgpu_device *)bp->base.ctx->driver_context)->mode_info.atom_context, \
 		GetIndexIntoMasterTable(COMMAND, command), &frev, &crev)
 
 #define BIOS_CMD_TABLE_PARA_REVISION(command)\
-	bios_cmd_table_para_revision(bp->base.ctx->cgs_device, \
+	bios_cmd_table_para_revision(bp->base.ctx->driver_context, \
 		GetIndexIntoMasterTable(COMMAND, command))
 
 static void init_dig_encoder_control(struct bios_parser *bp);
@@ -82,16 +82,18 @@ void dal_bios_parser_init_cmd_tbl(struct bios_parser *bp)
 	init_set_dce_clock(bp);
 }
 
-static uint32_t bios_cmd_table_para_revision(void *cgs_device,
+static uint32_t bios_cmd_table_para_revision(void *dev,
 					     uint32_t index)
 {
+	struct amdgpu_device *adev = dev;
 	uint8_t frev, crev;
 
-	if (cgs_atom_get_cmd_table_revs(cgs_device,
+	if (amdgpu_atom_parse_cmd_header(adev->mode_info.atom_context,
 					index,
-					&frev, &crev) != 0)
+					&frev, &crev))
+		return crev;
+	else
 		return 0;
-	return crev;
 }
 
 /*******************************************************************************
@@ -368,7 +370,7 @@ static void init_transmitter_control(struct bios_parser *bp)
 	uint8_t crev;
 
 	if (BIOS_CMD_TABLE_REVISION(UNIPHYTransmitterControl,
-			frev, crev) != 0)
+			frev, crev) == false)
 		BREAK_TO_DEBUGGER();
 	switch (crev) {
 	case 2:
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table2.c b/drivers/gpu/drm/amd/display/dc/bios/command_table2.c
index 3f63f712c8a4..2b5dc499a35e 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/command_table2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/command_table2.c
@@ -26,14 +26,18 @@
 #include "dm_services.h"
 
 #include "ObjectID.h"
-#include "atomfirmware.h"
 
+#include "atomfirmware.h"
+#include "atom.h"
 #include "include/bios_parser_interface.h"
 
 #include "command_table2.h"
 #include "command_table_helper2.h"
 #include "bios_parser_helper.h"
 #include "bios_parser_types_internal2.h"
+#include "amdgpu.h"
+
+
 #define DC_LOGGER \
 	bp->base.ctx->logger
 
@@ -43,59 +47,32 @@
 		->FieldName)-(char *)0)/sizeof(uint16_t))
 
 #define EXEC_BIOS_CMD_TABLE(fname, params)\
-	(cgs_atom_exec_cmd_table(bp->base.ctx->cgs_device, \
+	(amdgpu_atom_execute_table(((struct amdgpu_device *)bp->base.ctx->driver_context)->mode_info.atom_context, \
 		GET_INDEX_INTO_MASTER_TABLE(command, fname), \
-		&params) == 0)
+		(uint32_t *)&params) == 0)
 
 #define BIOS_CMD_TABLE_REVISION(fname, frev, crev)\
-	cgs_atom_get_cmd_table_revs(bp->base.ctx->cgs_device, \
+	amdgpu_atom_parse_cmd_header(((struct amdgpu_device *)bp->base.ctx->driver_context)->mode_info.atom_context, \
 		GET_INDEX_INTO_MASTER_TABLE(command, fname), &frev, &crev)
 
 #define BIOS_CMD_TABLE_PARA_REVISION(fname)\
-	bios_cmd_table_para_revision(bp->base.ctx->cgs_device, \
+	bios_cmd_table_para_revision(bp->base.ctx->driver_context, \
 			GET_INDEX_INTO_MASTER_TABLE(command, fname))
 
-static void init_dig_encoder_control(struct bios_parser *bp);
-static void init_transmitter_control(struct bios_parser *bp);
-static void init_set_pixel_clock(struct bios_parser *bp);
-
-static void init_set_crtc_timing(struct bios_parser *bp);
-
-static void init_select_crtc_source(struct bios_parser *bp);
-static void init_enable_crtc(struct bios_parser *bp);
 
-static void init_external_encoder_control(struct bios_parser *bp);
-static void init_enable_disp_power_gating(struct bios_parser *bp);
-static void init_set_dce_clock(struct bios_parser *bp);
-static void init_get_smu_clock_info(struct bios_parser *bp);
 
-void dal_firmware_parser_init_cmd_tbl(struct bios_parser *bp)
-{
-	init_dig_encoder_control(bp);
-	init_transmitter_control(bp);
-	init_set_pixel_clock(bp);
-
-	init_set_crtc_timing(bp);
-
-	init_select_crtc_source(bp);
-	init_enable_crtc(bp);
-
-	init_external_encoder_control(bp);
-	init_enable_disp_power_gating(bp);
-	init_set_dce_clock(bp);
-	init_get_smu_clock_info(bp);
-}
-
-static uint32_t bios_cmd_table_para_revision(void *cgs_device,
+static uint32_t bios_cmd_table_para_revision(void *dev,
 					     uint32_t index)
 {
+	struct amdgpu_device *adev = dev;
 	uint8_t frev, crev;
 
-	if (cgs_atom_get_cmd_table_revs(cgs_device,
+	if (amdgpu_atom_parse_cmd_header(adev->mode_info.atom_context,
 					index,
-					&frev, &crev) != 0)
+					&frev, &crev))
+		return crev;
+	else
 		return 0;
-	return crev;
 }
 
 /******************************************************************************
@@ -201,7 +178,7 @@ static void init_transmitter_control(struct bios_parser *bp)
 	uint8_t frev;
 	uint8_t crev;
 
-	if (BIOS_CMD_TABLE_REVISION(dig1transmittercontrol, frev, crev) != 0)
+	if (BIOS_CMD_TABLE_REVISION(dig1transmittercontrol, frev, crev) == false)
 		BREAK_TO_DEBUGGER();
 	switch (crev) {
 	case 6:
@@ -823,3 +800,20 @@ static unsigned int get_smu_clock_info_v3_1(struct bios_parser *bp, uint8_t id)
 	return 0;
 }
 
+void dal_firmware_parser_init_cmd_tbl(struct bios_parser *bp)
+{
+	init_dig_encoder_control(bp);
+	init_transmitter_control(bp);
+	init_set_pixel_clock(bp);
+
+	init_set_crtc_timing(bp);
+
+	init_select_crtc_source(bp);
+	init_enable_crtc(bp);
+
+	init_external_encoder_control(bp);
+	init_enable_disp_power_gating(bp);
+	init_set_dce_clock(bp);
+	init_get_smu_clock_info(bp);
+
+}
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c
index 2979358c6a55..253bbb1eea60 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c
@@ -51,6 +51,7 @@ bool dal_bios_parser_init_cmd_tbl_helper(
 		return true;
 
 	case DCE_VERSION_11_2:
+	case DCE_VERSION_11_22:
 		*h = dal_cmd_tbl_helper_dce112_get_table();
 		return true;
 
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c
index 9a4d30dd4969..770ff89ba7e1 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c
@@ -52,9 +52,10 @@ bool dal_bios_parser_init_cmd_tbl_helper2(
 		return true;
 
 	case DCE_VERSION_11_2:
+	case DCE_VERSION_11_22:
 		*h = dal_cmd_tbl_helper_dce112_get_table2();
 		return true;
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	case DCN_VERSION_1_0:
 		*h = dal_cmd_tbl_helper_dce112_get_table2();
 		return true;
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/Makefile b/drivers/gpu/drm/amd/display/dc/calcs/Makefile
index 95f332ee3e7e..416500e51b8d 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/calcs/Makefile
@@ -38,7 +38,7 @@ CFLAGS_dcn_calc_math.o := $(calcs_ccflags) -Wno-tautological-compare
 
 BW_CALCS = dce_calcs.o bw_fixed.o custom_float.o
 
-ifdef CONFIG_DRM_AMD_DC_DCN1_0
+ifdef CONFIG_X86
 BW_CALCS += dcn_calcs.o dcn_calc_math.o dcn_calc_auto.o
 endif
 
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/calcs_logger.h b/drivers/gpu/drm/amd/display/dc/calcs/calcs_logger.h
new file mode 100644
index 000000000000..62435bfc274d
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/calcs/calcs_logger.h
@@ -0,0 +1,578 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef _CALCS_CALCS_LOGGER_H_
+#define _CALCS_CALCS_LOGGER_H_
+#define DC_LOGGER ctx->logger
+
+static void print_bw_calcs_dceip(struct dc_context *ctx, const struct bw_calcs_dceip *dceip)
+{
+
+	DC_LOG_BANDWIDTH_CALCS("#####################################################################");
+	DC_LOG_BANDWIDTH_CALCS("struct bw_calcs_dceip");
+	DC_LOG_BANDWIDTH_CALCS("#####################################################################");
+	DC_LOG_BANDWIDTH_CALCS("	[enum]   bw_calcs_version version %d", dceip->version);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] large_cursor: %d", dceip->large_cursor);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] dmif_pipe_en_fbc_chunk_tracker: %d", dceip->dmif_pipe_en_fbc_chunk_tracker);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] display_write_back_supported: %d", dceip->display_write_back_supported);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] argb_compression_support: %d", dceip->argb_compression_support);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] pre_downscaler_enabled: %d", dceip->pre_downscaler_enabled);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] underlay_downscale_prefetch_enabled: %d",
+				dceip->underlay_downscale_prefetch_enabled);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] graphics_lb_nodownscaling_multi_line_prefetching: %d",
+				dceip->graphics_lb_nodownscaling_multi_line_prefetching);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] limit_excessive_outstanding_dmif_requests: %d",
+				dceip->limit_excessive_outstanding_dmif_requests);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] cursor_max_outstanding_group_num: %d",
+				dceip->cursor_max_outstanding_group_num);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] lines_interleaved_into_lb: %d", dceip->lines_interleaved_into_lb);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] low_power_tiling_mode: %d", dceip->low_power_tiling_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] chunk_width: %d", dceip->chunk_width);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_graphics_pipes: %d", dceip->number_of_graphics_pipes);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_underlay_pipes: %d", dceip->number_of_underlay_pipes);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] max_dmif_buffer_allocated: %d", dceip->max_dmif_buffer_allocated);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] graphics_dmif_size: %d", dceip->graphics_dmif_size);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] underlay_luma_dmif_size: %d", dceip->underlay_luma_dmif_size);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] underlay_chroma_dmif_size: %d", dceip->underlay_chroma_dmif_size);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] scatter_gather_lines_of_pte_prefetching_in_linear_mode: %d",
+				dceip->scatter_gather_lines_of_pte_prefetching_in_linear_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] display_write_back420_luma_mcifwr_buffer_size: %d",
+				dceip->display_write_back420_luma_mcifwr_buffer_size);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] display_write_back420_chroma_mcifwr_buffer_size: %d",
+				dceip->display_write_back420_chroma_mcifwr_buffer_size);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] scatter_gather_pte_request_rows_in_tiling_mode: %d",
+				dceip->scatter_gather_pte_request_rows_in_tiling_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay_vscaler_efficiency10_bit_per_component: %d",
+				bw_fixed_to_int(dceip->underlay_vscaler_efficiency10_bit_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay_vscaler_efficiency12_bit_per_component: %d",
+				bw_fixed_to_int(dceip->underlay_vscaler_efficiency12_bit_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] graphics_vscaler_efficiency6_bit_per_component: %d",
+				bw_fixed_to_int(dceip->graphics_vscaler_efficiency6_bit_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] graphics_vscaler_efficiency8_bit_per_component: %d",
+				bw_fixed_to_int(dceip->graphics_vscaler_efficiency8_bit_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] graphics_vscaler_efficiency10_bit_per_component: %d",
+				bw_fixed_to_int(dceip->graphics_vscaler_efficiency10_bit_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] graphics_vscaler_efficiency12_bit_per_component: %d",
+				bw_fixed_to_int(dceip->graphics_vscaler_efficiency12_bit_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] alpha_vscaler_efficiency: %d",
+				bw_fixed_to_int(dceip->alpha_vscaler_efficiency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_write_pixels_per_dispclk: %d",
+				bw_fixed_to_int(dceip->lb_write_pixels_per_dispclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_size_per_component444: %d",
+				bw_fixed_to_int(dceip->lb_size_per_component444));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_and_dram_clock_state_change_gated_before_cursor: %d",
+				bw_fixed_to_int(dceip->stutter_and_dram_clock_state_change_gated_before_cursor));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay420_luma_lb_size_per_component: %d",
+				bw_fixed_to_int(dceip->underlay420_luma_lb_size_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay420_chroma_lb_size_per_component: %d",
+				bw_fixed_to_int(dceip->underlay420_chroma_lb_size_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay422_lb_size_per_component: %d",
+				bw_fixed_to_int(dceip->underlay422_lb_size_per_component));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] cursor_chunk_width: %d", bw_fixed_to_int(dceip->cursor_chunk_width));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] cursor_dcp_buffer_lines: %d",
+				bw_fixed_to_int(dceip->cursor_dcp_buffer_lines));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay_maximum_width_efficient_for_tiling: %d",
+				bw_fixed_to_int(dceip->underlay_maximum_width_efficient_for_tiling));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay_maximum_height_efficient_for_tiling: %d",
+				bw_fixed_to_int(dceip->underlay_maximum_height_efficient_for_tiling));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] peak_pte_request_to_eviction_ratio_limiting_multiple_displays_or_single_rotated_display: %d",
+				bw_fixed_to_int(dceip->peak_pte_request_to_eviction_ratio_limiting_multiple_displays_or_single_rotated_display));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] peak_pte_request_to_eviction_ratio_limiting_single_display_no_rotation: %d",
+				bw_fixed_to_int(dceip->peak_pte_request_to_eviction_ratio_limiting_single_display_no_rotation));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] minimum_outstanding_pte_request_limit: %d",
+				bw_fixed_to_int(dceip->minimum_outstanding_pte_request_limit));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] maximum_total_outstanding_pte_requests_allowed_by_saw: %d",
+				bw_fixed_to_int(dceip->maximum_total_outstanding_pte_requests_allowed_by_saw));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] linear_mode_line_request_alternation_slice: %d",
+				bw_fixed_to_int(dceip->linear_mode_line_request_alternation_slice));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] request_efficiency: %d", bw_fixed_to_int(dceip->request_efficiency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_per_request: %d", bw_fixed_to_int(dceip->dispclk_per_request));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_ramping_factor: %d",
+				bw_fixed_to_int(dceip->dispclk_ramping_factor));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_pipe_throughput_factor: %d",
+				bw_fixed_to_int(dceip->display_pipe_throughput_factor));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwr_all_surfaces_burst_time: %d",
+				bw_fixed_to_int(dceip->mcifwr_all_surfaces_burst_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_request_buffer_size: %d",
+				bw_fixed_to_int(dceip->dmif_request_buffer_size));
+
+
+}
+
+static void print_bw_calcs_vbios(struct dc_context *ctx, const struct bw_calcs_vbios *vbios)
+{
+
+	DC_LOG_BANDWIDTH_CALCS("#####################################################################");
+	DC_LOG_BANDWIDTH_CALCS("struct bw_calcs_vbios vbios");
+	DC_LOG_BANDWIDTH_CALCS("#####################################################################");
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines memory_type: %d", vbios->memory_type);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines memory_type: %d", vbios->memory_type);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] dram_channel_width_in_bits: %d", vbios->dram_channel_width_in_bits);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_dram_channels: %d", vbios->number_of_dram_channels);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_dram_banks: %d", vbios->number_of_dram_banks);
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] low_yclk: %d", bw_fixed_to_int(vbios->low_yclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid_yclk: %d", bw_fixed_to_int(vbios->mid_yclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] high_yclk: %d", bw_fixed_to_int(vbios->high_yclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] low_sclk: %d", bw_fixed_to_int(vbios->low_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid1_sclk: %d", bw_fixed_to_int(vbios->mid1_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid2_sclk: %d", bw_fixed_to_int(vbios->mid2_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid3_sclk: %d", bw_fixed_to_int(vbios->mid3_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid4_sclk: %d", bw_fixed_to_int(vbios->mid4_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid5_sclk: %d", bw_fixed_to_int(vbios->mid5_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid6_sclk: %d", bw_fixed_to_int(vbios->mid6_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] high_sclk: %d", bw_fixed_to_int(vbios->high_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] low_voltage_max_dispclk: %d",
+				bw_fixed_to_int(vbios->low_voltage_max_dispclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid_voltage_max_dispclk;: %d",
+				bw_fixed_to_int(vbios->mid_voltage_max_dispclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] high_voltage_max_dispclk;: %d",
+				bw_fixed_to_int(vbios->high_voltage_max_dispclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] low_voltage_max_phyclk: %d",
+				bw_fixed_to_int(vbios->low_voltage_max_phyclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mid_voltage_max_phyclk: %d",
+				bw_fixed_to_int(vbios->mid_voltage_max_phyclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] high_voltage_max_phyclk: %d",
+				bw_fixed_to_int(vbios->high_voltage_max_phyclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] data_return_bus_width: %d", bw_fixed_to_int(vbios->data_return_bus_width));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] trc: %d", bw_fixed_to_int(vbios->trc));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmifmc_urgent_latency: %d", bw_fixed_to_int(vbios->dmifmc_urgent_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_self_refresh_exit_latency: %d",
+				bw_fixed_to_int(vbios->stutter_self_refresh_exit_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_self_refresh_entry_latency: %d",
+				bw_fixed_to_int(vbios->stutter_self_refresh_entry_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] nbp_state_change_latency: %d",
+				bw_fixed_to_int(vbios->nbp_state_change_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwrmc_urgent_latency: %d",
+				bw_fixed_to_int(vbios->mcifwrmc_urgent_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bool] scatter_gather_enable: %d", vbios->scatter_gather_enable);
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] down_spread_percentage: %d",
+				bw_fixed_to_int(vbios->down_spread_percentage));
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] cursor_width: %d", vbios->cursor_width);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] average_compression_rate: %d", vbios->average_compression_rate);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_request_slots_gmc_reserves_for_dmif_per_channel: %d",
+				vbios->number_of_request_slots_gmc_reserves_for_dmif_per_channel);
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] blackout_duration: %d", bw_fixed_to_int(vbios->blackout_duration));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] maximum_blackout_recovery_time: %d",
+				bw_fixed_to_int(vbios->maximum_blackout_recovery_time));
+
+
+}
+
+static void print_bw_calcs_data(struct dc_context *ctx, struct bw_calcs_data *data)
+{
+
+	int i, j, k;
+
+	DC_LOG_BANDWIDTH_CALCS("#####################################################################");
+	DC_LOG_BANDWIDTH_CALCS("struct bw_calcs_data data");
+	DC_LOG_BANDWIDTH_CALCS("#####################################################################");
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_displays: %d", data->number_of_displays);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines underlay_surface_type: %d", data->underlay_surface_type);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines panning_and_bezel_adjustment: %d",
+				data->panning_and_bezel_adjustment);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines graphics_tiling_mode: %d", data->graphics_tiling_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] graphics_lb_bpc: %d", data->graphics_lb_bpc);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] underlay_lb_bpc: %d", data->underlay_lb_bpc);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines underlay_tiling_mode: %d", data->underlay_tiling_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines d0_underlay_mode: %d", data->d0_underlay_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] d1_display_write_back_dwb_enable: %d", data->d1_display_write_back_dwb_enable);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines d1_underlay_mode: %d", data->d1_underlay_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] cpup_state_change_enable: %d", data->cpup_state_change_enable);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] cpuc_state_change_enable: %d", data->cpuc_state_change_enable);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] nbp_state_change_enable: %d", data->nbp_state_change_enable);
+	DC_LOG_BANDWIDTH_CALCS("	[bool] stutter_mode_enable: %d", data->stutter_mode_enable);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] y_clk_level: %d", data->y_clk_level);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] sclk_level: %d", data->sclk_level);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_underlay_surfaces: %d", data->number_of_underlay_surfaces);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_dram_wrchannels: %d", data->number_of_dram_wrchannels);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] chunk_request_delay: %d", data->chunk_request_delay);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] number_of_dram_channels: %d", data->number_of_dram_channels);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines underlay_micro_tile_mode: %d", data->underlay_micro_tile_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines graphics_micro_tile_mode: %d", data->graphics_micro_tile_mode);
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] max_phyclk: %d", bw_fixed_to_int(data->max_phyclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dram_efficiency: %d", bw_fixed_to_int(data->dram_efficiency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_width_after_surface_type: %d",
+				bw_fixed_to_int(data->src_width_after_surface_type));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_height_after_surface_type: %d",
+				bw_fixed_to_int(data->src_height_after_surface_type));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] hsr_after_surface_type: %d",
+				bw_fixed_to_int(data->hsr_after_surface_type));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] vsr_after_surface_type: %d", bw_fixed_to_int(data->vsr_after_surface_type));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_width_after_rotation: %d",
+				bw_fixed_to_int(data->src_width_after_rotation));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_height_after_rotation: %d",
+				bw_fixed_to_int(data->src_height_after_rotation));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] hsr_after_rotation: %d", bw_fixed_to_int(data->hsr_after_rotation));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] vsr_after_rotation: %d", bw_fixed_to_int(data->vsr_after_rotation));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] source_height_pixels: %d", bw_fixed_to_int(data->source_height_pixels));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] hsr_after_stereo: %d", bw_fixed_to_int(data->hsr_after_stereo));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] vsr_after_stereo: %d", bw_fixed_to_int(data->vsr_after_stereo));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] source_width_in_lb: %d", bw_fixed_to_int(data->source_width_in_lb));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_line_pitch: %d", bw_fixed_to_int(data->lb_line_pitch));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] underlay_maximum_source_efficient_for_tiling: %d",
+				bw_fixed_to_int(data->underlay_maximum_source_efficient_for_tiling));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] num_lines_at_frame_start: %d",
+				bw_fixed_to_int(data->num_lines_at_frame_start));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_dmif_size_in_time: %d", bw_fixed_to_int(data->min_dmif_size_in_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_mcifwr_size_in_time: %d",
+				bw_fixed_to_int(data->min_mcifwr_size_in_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_requests_for_dmif_size: %d",
+				bw_fixed_to_int(data->total_requests_for_dmif_size));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] peak_pte_request_to_eviction_ratio_limiting: %d",
+				bw_fixed_to_int(data->peak_pte_request_to_eviction_ratio_limiting));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] useful_pte_per_pte_request: %d",
+				bw_fixed_to_int(data->useful_pte_per_pte_request));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_pte_request_rows: %d",
+				bw_fixed_to_int(data->scatter_gather_pte_request_rows));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_row_height: %d",
+				bw_fixed_to_int(data->scatter_gather_row_height));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_pte_requests_in_vblank: %d",
+				bw_fixed_to_int(data->scatter_gather_pte_requests_in_vblank));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] inefficient_linear_pitch_in_bytes: %d",
+				bw_fixed_to_int(data->inefficient_linear_pitch_in_bytes));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] cursor_total_data: %d", bw_fixed_to_int(data->cursor_total_data));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] cursor_total_request_groups: %d",
+				bw_fixed_to_int(data->cursor_total_request_groups));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_total_pte_requests: %d",
+				bw_fixed_to_int(data->scatter_gather_total_pte_requests));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_total_pte_request_groups: %d",
+				bw_fixed_to_int(data->scatter_gather_total_pte_request_groups));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] tile_width_in_pixels: %d", bw_fixed_to_int(data->tile_width_in_pixels));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_total_number_of_data_request_page_close_open: %d",
+				bw_fixed_to_int(data->dmif_total_number_of_data_request_page_close_open));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwr_total_number_of_data_request_page_close_open: %d",
+				bw_fixed_to_int(data->mcifwr_total_number_of_data_request_page_close_open));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] bytes_per_page_close_open: %d",
+				bw_fixed_to_int(data->bytes_per_page_close_open));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwr_total_page_close_open_time: %d",
+				bw_fixed_to_int(data->mcifwr_total_page_close_open_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_requests_for_adjusted_dmif_size: %d",
+				bw_fixed_to_int(data->total_requests_for_adjusted_dmif_size));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_dmifmc_urgent_trips: %d",
+				bw_fixed_to_int(data->total_dmifmc_urgent_trips));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_dmifmc_urgent_latency: %d",
+				bw_fixed_to_int(data->total_dmifmc_urgent_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_display_reads_required_data: %d",
+				bw_fixed_to_int(data->total_display_reads_required_data));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_display_reads_required_dram_access_data: %d",
+				bw_fixed_to_int(data->total_display_reads_required_dram_access_data));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_display_writes_required_data: %d",
+				bw_fixed_to_int(data->total_display_writes_required_data));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_display_writes_required_dram_access_data: %d",
+				bw_fixed_to_int(data->total_display_writes_required_dram_access_data));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_reads_required_data: %d",
+				bw_fixed_to_int(data->display_reads_required_data));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_reads_required_dram_access_data: %d",
+				bw_fixed_to_int(data->display_reads_required_dram_access_data));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_total_page_close_open_time: %d",
+				bw_fixed_to_int(data->dmif_total_page_close_open_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_cursor_memory_interface_buffer_size_in_time: %d",
+				bw_fixed_to_int(data->min_cursor_memory_interface_buffer_size_in_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_read_buffer_size_in_time: %d",
+				bw_fixed_to_int(data->min_read_buffer_size_in_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_reads_time_for_data_transfer: %d",
+				bw_fixed_to_int(data->display_reads_time_for_data_transfer));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_writes_time_for_data_transfer: %d",
+				bw_fixed_to_int(data->display_writes_time_for_data_transfer));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_required_dram_bandwidth: %d",
+				bw_fixed_to_int(data->dmif_required_dram_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwr_required_dram_bandwidth: %d",
+				bw_fixed_to_int(data->mcifwr_required_dram_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] required_dmifmc_urgent_latency_for_page_close_open: %d",
+				bw_fixed_to_int(data->required_dmifmc_urgent_latency_for_page_close_open));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] required_mcifmcwr_urgent_latency: %d",
+				bw_fixed_to_int(data->required_mcifmcwr_urgent_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] required_dram_bandwidth_gbyte_per_second: %d",
+				bw_fixed_to_int(data->required_dram_bandwidth_gbyte_per_second));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dram_bandwidth: %d", bw_fixed_to_int(data->dram_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_required_sclk: %d", bw_fixed_to_int(data->dmif_required_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwr_required_sclk: %d", bw_fixed_to_int(data->mcifwr_required_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] required_sclk: %d", bw_fixed_to_int(data->required_sclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] downspread_factor: %d", bw_fixed_to_int(data->downspread_factor));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] v_scaler_efficiency: %d", bw_fixed_to_int(data->v_scaler_efficiency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scaler_limits_factor: %d", bw_fixed_to_int(data->scaler_limits_factor));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_pipe_pixel_throughput: %d",
+				bw_fixed_to_int(data->display_pipe_pixel_throughput));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_dispclk_required_with_ramping: %d",
+				bw_fixed_to_int(data->total_dispclk_required_with_ramping));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_dispclk_required_without_ramping: %d",
+				bw_fixed_to_int(data->total_dispclk_required_without_ramping));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_read_request_bandwidth: %d",
+				bw_fixed_to_int(data->total_read_request_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_write_request_bandwidth: %d",
+				bw_fixed_to_int(data->total_write_request_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_required_for_total_read_request_bandwidth: %d",
+				bw_fixed_to_int(data->dispclk_required_for_total_read_request_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_dispclk_required_with_ramping_with_request_bandwidth: %d",
+				bw_fixed_to_int(data->total_dispclk_required_with_ramping_with_request_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_dispclk_required_without_ramping_with_request_bandwidth: %d",
+				bw_fixed_to_int(data->total_dispclk_required_without_ramping_with_request_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk: %d", bw_fixed_to_int(data->dispclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] blackout_recovery_time: %d", bw_fixed_to_int(data->blackout_recovery_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_pixels_per_data_fifo_entry: %d",
+				bw_fixed_to_int(data->min_pixels_per_data_fifo_entry));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] sclk_deep_sleep: %d", bw_fixed_to_int(data->sclk_deep_sleep));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] chunk_request_time: %d", bw_fixed_to_int(data->chunk_request_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] cursor_request_time: %d", bw_fixed_to_int(data->cursor_request_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] line_source_pixels_transfer_time: %d",
+				bw_fixed_to_int(data->line_source_pixels_transfer_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmifdram_access_efficiency: %d",
+				bw_fixed_to_int(data->dmifdram_access_efficiency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwrdram_access_efficiency: %d",
+				bw_fixed_to_int(data->mcifwrdram_access_efficiency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_average_bandwidth_no_compression: %d",
+				bw_fixed_to_int(data->total_average_bandwidth_no_compression));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_average_bandwidth: %d",
+				bw_fixed_to_int(data->total_average_bandwidth));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] total_stutter_cycle_duration: %d",
+				bw_fixed_to_int(data->total_stutter_cycle_duration));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_burst_time: %d", bw_fixed_to_int(data->stutter_burst_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] time_in_self_refresh: %d", bw_fixed_to_int(data->time_in_self_refresh));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_efficiency: %d", bw_fixed_to_int(data->stutter_efficiency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] worst_number_of_trips_to_memory: %d",
+				bw_fixed_to_int(data->worst_number_of_trips_to_memory));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] immediate_flip_time: %d", bw_fixed_to_int(data->immediate_flip_time));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] latency_for_non_dmif_clients: %d",
+				bw_fixed_to_int(data->latency_for_non_dmif_clients));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] latency_for_non_mcifwr_clients: %d",
+				bw_fixed_to_int(data->latency_for_non_mcifwr_clients));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmifmc_urgent_latency_supported_in_high_sclk_and_yclk: %d",
+				bw_fixed_to_int(data->dmifmc_urgent_latency_supported_in_high_sclk_and_yclk));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] nbp_state_dram_speed_change_margin: %d",
+				bw_fixed_to_int(data->nbp_state_dram_speed_change_margin));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_reads_time_for_data_transfer_and_urgent_latency: %d",
+				bw_fixed_to_int(data->display_reads_time_for_data_transfer_and_urgent_latency));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dram_speed_change_margin: %d",
+				bw_fixed_to_int(data->dram_speed_change_margin));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_vblank_dram_speed_change_margin: %d",
+				bw_fixed_to_int(data->min_vblank_dram_speed_change_margin));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_stutter_refresh_duration: %d",
+				bw_fixed_to_int(data->min_stutter_refresh_duration));
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] total_stutter_dmif_buffer_size: %d", data->total_stutter_dmif_buffer_size);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] total_bytes_requested: %d", data->total_bytes_requested);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] min_stutter_dmif_buffer_size: %d", data->min_stutter_dmif_buffer_size);
+	DC_LOG_BANDWIDTH_CALCS("	[uint32_t] num_stutter_bursts: %d", data->num_stutter_bursts);
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] v_blank_nbp_state_dram_speed_change_latency_supported: %d",
+				bw_fixed_to_int(data->v_blank_nbp_state_dram_speed_change_latency_supported));
+	DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] nbp_state_dram_speed_change_latency_supported: %d",
+				bw_fixed_to_int(data->nbp_state_dram_speed_change_latency_supported));
+
+	for (i = 0; i < maximum_number_of_surfaces; i++) {
+		DC_LOG_BANDWIDTH_CALCS("	[bool] fbc_en[%d]:%d\n", i, data->fbc_en[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] lpt_en[%d]:%d", i, data->lpt_en[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] displays_match_flag[%d]:%d", i, data->displays_match_flag[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] use_alpha[%d]:%d", i, data->use_alpha[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] orthogonal_rotation[%d]:%d", i, data->orthogonal_rotation[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] enable[%d]:%d", i, data->enable[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] access_one_channel_only[%d]:%d", i, data->access_one_channel_only[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] scatter_gather_enable_for_pipe[%d]:%d",
+					i, data->scatter_gather_enable_for_pipe[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] interlace_mode[%d]:%d",
+					i, data->interlace_mode[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] display_pstate_change_enable[%d]:%d",
+					i, data->display_pstate_change_enable[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bool] line_buffer_prefetch[%d]:%d", i, data->line_buffer_prefetch[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[uint32_t] bytes_per_pixel[%d]:%d", i, data->bytes_per_pixel[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[uint32_t] max_chunks_non_fbc_mode[%d]:%d",
+					i, data->max_chunks_non_fbc_mode[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[uint32_t] lb_bpc[%d]:%d", i, data->lb_bpc[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[uint32_t] output_bpphdmi[%d]:%d", i, data->output_bpphdmi[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[uint32_t] output_bppdp4_lane_hbr[%d]:%d", i, data->output_bppdp4_lane_hbr[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[uint32_t] output_bppdp4_lane_hbr2[%d]:%d",
+					i, data->output_bppdp4_lane_hbr2[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[uint32_t] output_bppdp4_lane_hbr3[%d]:%d",
+					i, data->output_bppdp4_lane_hbr3[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[enum] bw_defines stereo_mode[%d]:%d", i, data->stereo_mode[i]);
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_buffer_transfer_time[%d]:%d",
+					i, bw_fixed_to_int(data->dmif_buffer_transfer_time[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] displays_with_same_mode[%d]:%d",
+					i, bw_fixed_to_int(data->displays_with_same_mode[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_dmif_buffer_size[%d]:%d",
+					i, bw_fixed_to_int(data->stutter_dmif_buffer_size[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_refresh_duration[%d]:%d",
+					i, bw_fixed_to_int(data->stutter_refresh_duration[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_exit_watermark[%d]:%d",
+					i, bw_fixed_to_int(data->stutter_exit_watermark[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_entry_watermark[%d]:%d",
+					i, bw_fixed_to_int(data->stutter_entry_watermark[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] h_total[%d]:%d", i, bw_fixed_to_int(data->h_total[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] v_total[%d]:%d", i, bw_fixed_to_int(data->v_total[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] pixel_rate[%d]:%d", i, bw_fixed_to_int(data->pixel_rate[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_width[%d]:%d", i, bw_fixed_to_int(data->src_width[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] pitch_in_pixels[%d]:%d",
+					i, bw_fixed_to_int(data->pitch_in_pixels[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] pitch_in_pixels_after_surface_type[%d]:%d",
+					i, bw_fixed_to_int(data->pitch_in_pixels_after_surface_type[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_height[%d]:%d", i, bw_fixed_to_int(data->src_height[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scale_ratio[%d]:%d", i, bw_fixed_to_int(data->scale_ratio[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] h_taps[%d]:%d", i, bw_fixed_to_int(data->h_taps[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] v_taps[%d]:%d", i, bw_fixed_to_int(data->v_taps[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] h_scale_ratio[%d]:%d", i, bw_fixed_to_int(data->h_scale_ratio[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] v_scale_ratio[%d]:%d", i, bw_fixed_to_int(data->v_scale_ratio[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] rotation_angle[%d]:%d",
+					i, bw_fixed_to_int(data->rotation_angle[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] compression_rate[%d]:%d",
+					i, bw_fixed_to_int(data->compression_rate[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] hsr[%d]:%d", i, bw_fixed_to_int(data->hsr[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] vsr[%d]:%d", i, bw_fixed_to_int(data->vsr[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] source_width_rounded_up_to_chunks[%d]:%d",
+					i, bw_fixed_to_int(data->source_width_rounded_up_to_chunks[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] source_width_pixels[%d]:%d",
+					i, bw_fixed_to_int(data->source_width_pixels[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] source_height_rounded_up_to_chunks[%d]:%d",
+					i, bw_fixed_to_int(data->source_height_rounded_up_to_chunks[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] display_bandwidth[%d]:%d",
+					i, bw_fixed_to_int(data->display_bandwidth[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] request_bandwidth[%d]:%d",
+					i, bw_fixed_to_int(data->request_bandwidth[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] bytes_per_request[%d]:%d",
+					i, bw_fixed_to_int(data->bytes_per_request[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] useful_bytes_per_request[%d]:%d",
+					i, bw_fixed_to_int(data->useful_bytes_per_request[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lines_interleaved_in_mem_access[%d]:%d",
+					i, bw_fixed_to_int(data->lines_interleaved_in_mem_access[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] latency_hiding_lines[%d]:%d",
+					i, bw_fixed_to_int(data->latency_hiding_lines[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_partitions[%d]:%d",
+					i, bw_fixed_to_int(data->lb_partitions[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_partitions_max[%d]:%d",
+					i, bw_fixed_to_int(data->lb_partitions_max[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_required_with_ramping[%d]:%d",
+					i, bw_fixed_to_int(data->dispclk_required_with_ramping[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_required_without_ramping[%d]:%d",
+					i, bw_fixed_to_int(data->dispclk_required_without_ramping[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] data_buffer_size[%d]:%d",
+					i, bw_fixed_to_int(data->data_buffer_size[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] outstanding_chunk_request_limit[%d]:%d",
+					i, bw_fixed_to_int(data->outstanding_chunk_request_limit[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] urgent_watermark[%d]:%d",
+					i, bw_fixed_to_int(data->urgent_watermark[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] nbp_state_change_watermark[%d]:%d",
+					i, bw_fixed_to_int(data->nbp_state_change_watermark[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] v_filter_init[%d]:%d", i, bw_fixed_to_int(data->v_filter_init[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] stutter_cycle_duration[%d]:%d",
+					i, bw_fixed_to_int(data->stutter_cycle_duration[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] average_bandwidth[%d]:%d",
+					i, bw_fixed_to_int(data->average_bandwidth[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] average_bandwidth_no_compression[%d]:%d",
+					i, bw_fixed_to_int(data->average_bandwidth_no_compression[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_pte_request_limit[%d]:%d",
+					i, bw_fixed_to_int(data->scatter_gather_pte_request_limit[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_size_per_component[%d]:%d",
+					i, bw_fixed_to_int(data->lb_size_per_component[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] memory_chunk_size_in_bytes[%d]:%d",
+					i, bw_fixed_to_int(data->memory_chunk_size_in_bytes[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] pipe_chunk_size_in_bytes[%d]:%d",
+					i, bw_fixed_to_int(data->pipe_chunk_size_in_bytes[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] number_of_trips_to_memory_for_getting_apte_row[%d]:%d",
+					i, bw_fixed_to_int(data->number_of_trips_to_memory_for_getting_apte_row[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] adjusted_data_buffer_size[%d]:%d",
+					i, bw_fixed_to_int(data->adjusted_data_buffer_size[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] adjusted_data_buffer_size_in_memory[%d]:%d",
+					i, bw_fixed_to_int(data->adjusted_data_buffer_size_in_memory[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] pixels_per_data_fifo_entry[%d]:%d",
+					i, bw_fixed_to_int(data->pixels_per_data_fifo_entry[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_pte_requests_in_row[%d]:%d",
+					i, bw_fixed_to_int(data->scatter_gather_pte_requests_in_row[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] pte_request_per_chunk[%d]:%d",
+					i, bw_fixed_to_int(data->pte_request_per_chunk[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_page_width[%d]:%d",
+					i, bw_fixed_to_int(data->scatter_gather_page_width[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] scatter_gather_page_height[%d]:%d",
+					i, bw_fixed_to_int(data->scatter_gather_page_height[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_lines_in_per_line_out_in_beginning_of_frame[%d]:%d",
+					i, bw_fixed_to_int(data->lb_lines_in_per_line_out_in_beginning_of_frame[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] lb_lines_in_per_line_out_in_middle_of_frame[%d]:%d",
+					i, bw_fixed_to_int(data->lb_lines_in_per_line_out_in_middle_of_frame[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] cursor_width_pixels[%d]:%d",
+					i, bw_fixed_to_int(data->cursor_width_pixels[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] minimum_latency_hiding[%d]:%d",
+					i, bw_fixed_to_int(data->minimum_latency_hiding[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] maximum_latency_hiding[%d]:%d",
+					i, bw_fixed_to_int(data->maximum_latency_hiding[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] minimum_latency_hiding_with_cursor[%d]:%d",
+					i, bw_fixed_to_int(data->minimum_latency_hiding_with_cursor[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] maximum_latency_hiding_with_cursor[%d]:%d",
+					i, bw_fixed_to_int(data->maximum_latency_hiding_with_cursor[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_pixels_for_first_output_pixel[%d]:%d",
+					i, bw_fixed_to_int(data->src_pixels_for_first_output_pixel[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_pixels_for_last_output_pixel[%d]:%d",
+					i, bw_fixed_to_int(data->src_pixels_for_last_output_pixel[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_data_for_first_output_pixel[%d]:%d",
+					i, bw_fixed_to_int(data->src_data_for_first_output_pixel[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] src_data_for_last_output_pixel[%d]:%d",
+					i, bw_fixed_to_int(data->src_data_for_last_output_pixel[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] active_time[%d]:%d", i, bw_fixed_to_int(data->active_time[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] horizontal_blank_and_chunk_granularity_factor[%d]:%d",
+					i, bw_fixed_to_int(data->horizontal_blank_and_chunk_granularity_factor[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] cursor_latency_hiding[%d]:%d",
+					i, bw_fixed_to_int(data->cursor_latency_hiding[i]));
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] v_blank_dram_speed_change_margin[%d]:%d",
+					i, bw_fixed_to_int(data->v_blank_dram_speed_change_margin[i]));
+		}
+
+	for (i = 0; i < maximum_number_of_surfaces; i++) {
+		for (j = 0; j < 3; j++) {
+			for (k = 0; k < 8; k++) {
+
+				DC_LOG_BANDWIDTH_CALCS("\n	[bw_fixed] line_source_transfer_time[%d][%d][%d]:%d",
+					i, j, k, bw_fixed_to_int(data->line_source_transfer_time[i][j][k]));
+				DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dram_speed_change_line_source_transfer_time[%d][%d][%d]:%d",
+					i, j, k,
+					bw_fixed_to_int(data->dram_speed_change_line_source_transfer_time[i][j][k]));
+			}
+		}
+	}
+
+	for (i = 0; i < 3; i++) {
+		for (j = 0; j < 8; j++) {
+
+			DC_LOG_BANDWIDTH_CALCS("\n	[uint32_t] num_displays_with_margin[%d][%d]:%d",
+					i, j, data->num_displays_with_margin[i][j]);
+			DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_burst_time[%d][%d]:%d",
+					i, j, bw_fixed_to_int(data->dmif_burst_time[i][j]));
+			DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] mcifwr_burst_time[%d][%d]:%d",
+					i, j, bw_fixed_to_int(data->mcifwr_burst_time[i][j]));
+			DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] min_dram_speed_change_margin[%d][%d]:%d",
+					i, j, bw_fixed_to_int(data->min_dram_speed_change_margin[i][j]));
+			DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_required_for_dram_speed_change[%d][%d]:%d",
+					i, j, bw_fixed_to_int(data->dispclk_required_for_dram_speed_change[i][j]));
+			DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] blackout_duration_margin[%d][%d]:%d",
+					i, j, bw_fixed_to_int(data->blackout_duration_margin[i][j]));
+			DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_required_for_blackout_duration[%d][%d]:%d",
+					i, j, bw_fixed_to_int(data->dispclk_required_for_blackout_duration[i][j]));
+			DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dispclk_required_for_blackout_recovery[%d][%d]:%d",
+					i, j, bw_fixed_to_int(data->dispclk_required_for_blackout_recovery[i][j]));
+		}
+	}
+
+	for (i = 0; i < 6; i++) {
+		DC_LOG_BANDWIDTH_CALCS("	[bw_fixed] dmif_required_sclk_for_urgent_latency[%d]:%d",
+					i, bw_fixed_to_int(data->dmif_required_sclk_for_urgent_latency[i]));
+	}
+}
+;
+
+#endif /* _CALCS_CALCS_LOGGER_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/custom_float.c b/drivers/gpu/drm/amd/display/dc/calcs/custom_float.c
index 7243c37f569e..31d167bc548f 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/custom_float.c
+++ b/drivers/gpu/drm/amd/display/dc/calcs/custom_float.c
@@ -36,41 +36,41 @@ static bool build_custom_float(
 	uint32_t exp_offset = (1 << (format->exponenta_bits - 1)) - 1;
 
 	const struct fixed31_32 mantissa_constant_plus_max_fraction =
-		dal_fixed31_32_from_fraction(
+		dc_fixpt_from_fraction(
 			(1LL << (format->mantissa_bits + 1)) - 1,
 			1LL << format->mantissa_bits);
 
 	struct fixed31_32 mantiss;
 
-	if (dal_fixed31_32_eq(
+	if (dc_fixpt_eq(
 		value,
-		dal_fixed31_32_zero)) {
+		dc_fixpt_zero)) {
 		*negative = false;
 		*mantissa = 0;
 		*exponenta = 0;
 		return true;
 	}
 
-	if (dal_fixed31_32_lt(
+	if (dc_fixpt_lt(
 		value,
-		dal_fixed31_32_zero)) {
+		dc_fixpt_zero)) {
 		*negative = format->sign;
-		value = dal_fixed31_32_neg(value);
+		value = dc_fixpt_neg(value);
 	} else {
 		*negative = false;
 	}
 
-	if (dal_fixed31_32_lt(
+	if (dc_fixpt_lt(
 		value,
-		dal_fixed31_32_one)) {
+		dc_fixpt_one)) {
 		uint32_t i = 1;
 
 		do {
-			value = dal_fixed31_32_shl(value, 1);
+			value = dc_fixpt_shl(value, 1);
 			++i;
-		} while (dal_fixed31_32_lt(
+		} while (dc_fixpt_lt(
 			value,
-			dal_fixed31_32_one));
+			dc_fixpt_one));
 
 		--i;
 
@@ -81,15 +81,15 @@ static bool build_custom_float(
 		}
 
 		*exponenta = exp_offset - i;
-	} else if (dal_fixed31_32_le(
+	} else if (dc_fixpt_le(
 		mantissa_constant_plus_max_fraction,
 		value)) {
 		uint32_t i = 1;
 
 		do {
-			value = dal_fixed31_32_shr(value, 1);
+			value = dc_fixpt_shr(value, 1);
 			++i;
-		} while (dal_fixed31_32_lt(
+		} while (dc_fixpt_lt(
 			mantissa_constant_plus_max_fraction,
 			value));
 
@@ -98,23 +98,23 @@ static bool build_custom_float(
 		*exponenta = exp_offset;
 	}
 
-	mantiss = dal_fixed31_32_sub(
+	mantiss = dc_fixpt_sub(
 		value,
-		dal_fixed31_32_one);
+		dc_fixpt_one);
 
-	if (dal_fixed31_32_lt(
+	if (dc_fixpt_lt(
 			mantiss,
-			dal_fixed31_32_zero) ||
-		dal_fixed31_32_lt(
-			dal_fixed31_32_one,
+			dc_fixpt_zero) ||
+		dc_fixpt_lt(
+			dc_fixpt_one,
 			mantiss))
-		mantiss = dal_fixed31_32_zero;
+		mantiss = dc_fixpt_zero;
 	else
-		mantiss = dal_fixed31_32_shl(
+		mantiss = dc_fixpt_shl(
 			mantiss,
 			format->mantissa_bits);
 
-	*mantissa = dal_fixed31_32_floor(mantiss);
+	*mantissa = dc_fixpt_floor(mantiss);
 
 	return true;
 }
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
index 0cbab81ab304..160d11a15eac 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
+++ b/drivers/gpu/drm/amd/display/dc/calcs/dce_calcs.c
@@ -28,6 +28,7 @@
 #include "dc.h"
 #include "core_types.h"
 #include "dal_asic_id.h"
+#include "calcs_logger.h"
 
 /*
  * NOTE:
@@ -52,11 +53,14 @@ static enum bw_calcs_version bw_calcs_version_from_asic_id(struct hw_asic_id asi
 		return BW_CALCS_VERSION_CARRIZO;
 
 	case FAMILY_VI:
+		if (ASIC_REV_IS_POLARIS12_V(asic_id.hw_internal_rev))
+			return BW_CALCS_VERSION_POLARIS12;
 		if (ASIC_REV_IS_POLARIS10_P(asic_id.hw_internal_rev))
 			return BW_CALCS_VERSION_POLARIS10;
-		if (ASIC_REV_IS_POLARIS11_M(asic_id.hw_internal_rev) ||
-				ASIC_REV_IS_POLARIS12_V(asic_id.hw_internal_rev))
+		if (ASIC_REV_IS_POLARIS11_M(asic_id.hw_internal_rev))
 			return BW_CALCS_VERSION_POLARIS11;
+		if (ASIC_REV_IS_VEGAM(asic_id.hw_internal_rev))
+			return BW_CALCS_VERSION_VEGAM;
 		return BW_CALCS_VERSION_INVALID;
 
 	case FAMILY_AI:
@@ -2145,6 +2149,9 @@ void bw_calcs_init(struct bw_calcs_dceip *bw_dceip,
 		dceip.mcifwr_all_surfaces_burst_time = bw_int_to_fixed(0); /* todo: this is a bug*/
 		break;
 	case BW_CALCS_VERSION_POLARIS10:
+		/* TODO: Treat VEGAM the same as P10 for now
+		 * Need to tune the para for VEGAM if needed */
+	case BW_CALCS_VERSION_VEGAM:
 		vbios.memory_type = bw_def_gddr5;
 		vbios.dram_channel_width_in_bits = 32;
 		vbios.number_of_dram_channels = asic_id.vram_width / vbios.dram_channel_width_in_bits;
@@ -2373,6 +2380,122 @@ void bw_calcs_init(struct bw_calcs_dceip *bw_dceip,
 		dceip.scatter_gather_pte_request_rows_in_tiling_mode = 2;
 		dceip.mcifwr_all_surfaces_burst_time = bw_int_to_fixed(0);
 		break;
+	case BW_CALCS_VERSION_POLARIS12:
+		vbios.memory_type = bw_def_gddr5;
+		vbios.dram_channel_width_in_bits = 32;
+		vbios.number_of_dram_channels = asic_id.vram_width / vbios.dram_channel_width_in_bits;
+		vbios.number_of_dram_banks = 8;
+		vbios.high_yclk = bw_int_to_fixed(6000);
+		vbios.mid_yclk = bw_int_to_fixed(3200);
+		vbios.low_yclk = bw_int_to_fixed(1000);
+		vbios.low_sclk = bw_int_to_fixed(678);
+		vbios.mid1_sclk = bw_int_to_fixed(864);
+		vbios.mid2_sclk = bw_int_to_fixed(900);
+		vbios.mid3_sclk = bw_int_to_fixed(920);
+		vbios.mid4_sclk = bw_int_to_fixed(940);
+		vbios.mid5_sclk = bw_int_to_fixed(960);
+		vbios.mid6_sclk = bw_int_to_fixed(980);
+		vbios.high_sclk = bw_int_to_fixed(1049);
+		vbios.low_voltage_max_dispclk = bw_int_to_fixed(459);
+		vbios.mid_voltage_max_dispclk = bw_int_to_fixed(654);
+		vbios.high_voltage_max_dispclk = bw_int_to_fixed(1108);
+		vbios.low_voltage_max_phyclk = bw_int_to_fixed(540);
+		vbios.mid_voltage_max_phyclk = bw_int_to_fixed(810);
+		vbios.high_voltage_max_phyclk = bw_int_to_fixed(810);
+		vbios.data_return_bus_width = bw_int_to_fixed(32);
+		vbios.trc = bw_int_to_fixed(48);
+		if (vbios.number_of_dram_channels == 2) // 64-bit
+			vbios.dmifmc_urgent_latency = bw_int_to_fixed(4);
+		else
+			vbios.dmifmc_urgent_latency = bw_int_to_fixed(3);
+		vbios.stutter_self_refresh_exit_latency = bw_int_to_fixed(5);
+		vbios.stutter_self_refresh_entry_latency = bw_int_to_fixed(0);
+		vbios.nbp_state_change_latency = bw_int_to_fixed(250);
+		vbios.mcifwrmc_urgent_latency = bw_int_to_fixed(10);
+		vbios.scatter_gather_enable = false;
+		vbios.down_spread_percentage = bw_frc_to_fixed(5, 10);
+		vbios.cursor_width = 32;
+		vbios.average_compression_rate = 4;
+		vbios.number_of_request_slots_gmc_reserves_for_dmif_per_channel = 256;
+		vbios.blackout_duration = bw_int_to_fixed(0); /* us */
+		vbios.maximum_blackout_recovery_time = bw_int_to_fixed(0);
+
+		dceip.max_average_percent_of_ideal_port_bw_display_can_use_in_normal_system_operation = 100;
+		dceip.max_average_percent_of_ideal_drambw_display_can_use_in_normal_system_operation = 100;
+		dceip.percent_of_ideal_port_bw_received_after_urgent_latency = 100;
+		dceip.large_cursor = false;
+		dceip.dmif_request_buffer_size = bw_int_to_fixed(768);
+		dceip.dmif_pipe_en_fbc_chunk_tracker = false;
+		dceip.cursor_max_outstanding_group_num = 1;
+		dceip.lines_interleaved_into_lb = 2;
+		dceip.chunk_width = 256;
+		dceip.number_of_graphics_pipes = 5;
+		dceip.number_of_underlay_pipes = 0;
+		dceip.low_power_tiling_mode = 0;
+		dceip.display_write_back_supported = true;
+		dceip.argb_compression_support = true;
+		dceip.underlay_vscaler_efficiency6_bit_per_component =
+			bw_frc_to_fixed(35556, 10000);
+		dceip.underlay_vscaler_efficiency8_bit_per_component =
+			bw_frc_to_fixed(34286, 10000);
+		dceip.underlay_vscaler_efficiency10_bit_per_component =
+			bw_frc_to_fixed(32, 10);
+		dceip.underlay_vscaler_efficiency12_bit_per_component =
+			bw_int_to_fixed(3);
+		dceip.graphics_vscaler_efficiency6_bit_per_component =
+			bw_frc_to_fixed(35, 10);
+		dceip.graphics_vscaler_efficiency8_bit_per_component =
+			bw_frc_to_fixed(34286, 10000);
+		dceip.graphics_vscaler_efficiency10_bit_per_component =
+			bw_frc_to_fixed(32, 10);
+		dceip.graphics_vscaler_efficiency12_bit_per_component =
+			bw_int_to_fixed(3);
+		dceip.alpha_vscaler_efficiency = bw_int_to_fixed(3);
+		dceip.max_dmif_buffer_allocated = 4;
+		dceip.graphics_dmif_size = 12288;
+		dceip.underlay_luma_dmif_size = 19456;
+		dceip.underlay_chroma_dmif_size = 23552;
+		dceip.pre_downscaler_enabled = true;
+		dceip.underlay_downscale_prefetch_enabled = true;
+		dceip.lb_write_pixels_per_dispclk = bw_int_to_fixed(1);
+		dceip.lb_size_per_component444 = bw_int_to_fixed(245952);
+		dceip.graphics_lb_nodownscaling_multi_line_prefetching = true;
+		dceip.stutter_and_dram_clock_state_change_gated_before_cursor =
+			bw_int_to_fixed(1);
+		dceip.underlay420_luma_lb_size_per_component = bw_int_to_fixed(
+			82176);
+		dceip.underlay420_chroma_lb_size_per_component =
+			bw_int_to_fixed(164352);
+		dceip.underlay422_lb_size_per_component = bw_int_to_fixed(
+			82176);
+		dceip.cursor_chunk_width = bw_int_to_fixed(64);
+		dceip.cursor_dcp_buffer_lines = bw_int_to_fixed(4);
+		dceip.underlay_maximum_width_efficient_for_tiling =
+			bw_int_to_fixed(1920);
+		dceip.underlay_maximum_height_efficient_for_tiling =
+			bw_int_to_fixed(1080);
+		dceip.peak_pte_request_to_eviction_ratio_limiting_multiple_displays_or_single_rotated_display =
+			bw_frc_to_fixed(3, 10);
+		dceip.peak_pte_request_to_eviction_ratio_limiting_single_display_no_rotation =
+			bw_int_to_fixed(25);
+		dceip.minimum_outstanding_pte_request_limit = bw_int_to_fixed(
+			2);
+		dceip.maximum_total_outstanding_pte_requests_allowed_by_saw =
+			bw_int_to_fixed(128);
+		dceip.limit_excessive_outstanding_dmif_requests = true;
+		dceip.linear_mode_line_request_alternation_slice =
+			bw_int_to_fixed(64);
+		dceip.scatter_gather_lines_of_pte_prefetching_in_linear_mode =
+			32;
+		dceip.display_write_back420_luma_mcifwr_buffer_size = 12288;
+		dceip.display_write_back420_chroma_mcifwr_buffer_size = 8192;
+		dceip.request_efficiency = bw_frc_to_fixed(8, 10);
+		dceip.dispclk_per_request = bw_int_to_fixed(2);
+		dceip.dispclk_ramping_factor = bw_frc_to_fixed(105, 100);
+		dceip.display_pipe_throughput_factor = bw_frc_to_fixed(105, 100);
+		dceip.scatter_gather_pte_request_rows_in_tiling_mode = 2;
+		dceip.mcifwr_all_surfaces_burst_time = bw_int_to_fixed(0);
+		break;
 	case BW_CALCS_VERSION_STONEY:
 		vbios.memory_type = bw_def_gddr5;
 		vbios.dram_channel_width_in_bits = 64;
@@ -2815,6 +2938,19 @@ static void populate_initial_data(
 				data->bytes_per_pixel[num_displays + 4] = 4;
 				break;
 			}
+		} else if (pipe[i].stream->dst.width != 0 &&
+					pipe[i].stream->dst.height != 0 &&
+					pipe[i].stream->src.width != 0 &&
+					pipe[i].stream->src.height != 0) {
+			data->src_width[num_displays + 4] = bw_int_to_fixed(pipe[i].stream->src.width);
+			data->pitch_in_pixels[num_displays + 4] = data->src_width[num_displays + 4];
+			data->src_height[num_displays + 4] = bw_int_to_fixed(pipe[i].stream->src.height);
+			data->h_taps[num_displays + 4] = pipe[i].stream->src.width == pipe[i].stream->dst.width ? bw_int_to_fixed(1) : bw_int_to_fixed(2);
+			data->v_taps[num_displays + 4] = pipe[i].stream->src.height == pipe[i].stream->dst.height ? bw_int_to_fixed(1) : bw_int_to_fixed(2);
+			data->h_scale_ratio[num_displays + 4] = bw_frc_to_fixed(pipe[i].stream->src.width, pipe[i].stream->dst.width);
+			data->v_scale_ratio[num_displays + 4] = bw_frc_to_fixed(pipe[i].stream->src.height, pipe[i].stream->dst.height);
+			data->rotation_angle[num_displays + 4] = bw_int_to_fixed(0);
+			data->bytes_per_pixel[num_displays + 4] = 4;
 		} else {
 			data->src_width[num_displays + 4] = bw_int_to_fixed(pipe[i].stream->timing.h_addressable);
 			data->pitch_in_pixels[num_displays + 4] = data->src_width[num_displays + 4];
@@ -2873,6 +3009,11 @@ bool bw_calcs(struct dc_context *ctx,
 		struct bw_fixed mid_yclk = vbios->mid_yclk;
 		struct bw_fixed low_yclk = vbios->low_yclk;
 
+		if (ctx->dc->debug.bandwidth_calcs_trace) {
+			print_bw_calcs_dceip(ctx, dceip);
+			print_bw_calcs_vbios(ctx, vbios);
+			print_bw_calcs_data(ctx, data);
+		}
 		calculate_bandwidth(dceip, vbios, data);
 
 		yclk_lvl = data->y_clk_level;
@@ -2968,7 +3109,33 @@ bool bw_calcs(struct dc_context *ctx,
 			bw_fixed_to_int(bw_mul(data->
 				stutter_exit_watermark[9], bw_int_to_fixed(1000)));
 
-
+		calcs_output->stutter_entry_wm_ns[0].a_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[4], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[1].a_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[5], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[2].a_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[6], bw_int_to_fixed(1000)));
+		if (ctx->dc->caps.max_slave_planes) {
+			calcs_output->stutter_entry_wm_ns[3].a_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[0], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].a_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[1], bw_int_to_fixed(1000)));
+		} else {
+			calcs_output->stutter_entry_wm_ns[3].a_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[7], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].a_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[8], bw_int_to_fixed(1000)));
+		}
+		calcs_output->stutter_entry_wm_ns[5].a_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[9], bw_int_to_fixed(1000)));
 
 		calcs_output->urgent_wm_ns[0].a_mark =
 			bw_fixed_to_int(bw_mul(data->
@@ -3063,7 +3230,33 @@ bool bw_calcs(struct dc_context *ctx,
 				bw_fixed_to_int(bw_mul(data->
 					stutter_exit_watermark[9], bw_int_to_fixed(1000)));
 
-
+		calcs_output->stutter_entry_wm_ns[0].b_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[4], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[1].b_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[5], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[2].b_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[6], bw_int_to_fixed(1000)));
+		if (ctx->dc->caps.max_slave_planes) {
+			calcs_output->stutter_entry_wm_ns[3].b_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[0], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].b_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[1], bw_int_to_fixed(1000)));
+		} else {
+			calcs_output->stutter_entry_wm_ns[3].b_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[7], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].b_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[8], bw_int_to_fixed(1000)));
+		}
+		calcs_output->stutter_entry_wm_ns[5].b_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[9], bw_int_to_fixed(1000)));
 
 			calcs_output->urgent_wm_ns[0].b_mark =
 				bw_fixed_to_int(bw_mul(data->
@@ -3156,6 +3349,34 @@ bool bw_calcs(struct dc_context *ctx,
 				bw_fixed_to_int(bw_mul(data->
 					stutter_exit_watermark[9], bw_int_to_fixed(1000)));
 
+		calcs_output->stutter_entry_wm_ns[0].c_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[4], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[1].c_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[5], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[2].c_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[6], bw_int_to_fixed(1000)));
+		if (ctx->dc->caps.max_slave_planes) {
+			calcs_output->stutter_entry_wm_ns[3].c_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[0], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].c_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[1], bw_int_to_fixed(1000)));
+		} else {
+			calcs_output->stutter_entry_wm_ns[3].c_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[7], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].c_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[8], bw_int_to_fixed(1000)));
+		}
+		calcs_output->stutter_entry_wm_ns[5].c_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[9], bw_int_to_fixed(1000)));
+
 			calcs_output->urgent_wm_ns[0].c_mark =
 				bw_fixed_to_int(bw_mul(data->
 					urgent_watermark[4], bw_int_to_fixed(1000)));
@@ -3260,6 +3481,33 @@ bool bw_calcs(struct dc_context *ctx,
 			bw_fixed_to_int(bw_mul(data->
 				stutter_exit_watermark[9], bw_int_to_fixed(1000)));
 
+		calcs_output->stutter_entry_wm_ns[0].d_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[4], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[1].d_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[5], bw_int_to_fixed(1000)));
+		calcs_output->stutter_entry_wm_ns[2].d_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[6], bw_int_to_fixed(1000)));
+		if (ctx->dc->caps.max_slave_planes) {
+			calcs_output->stutter_entry_wm_ns[3].d_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[0], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].d_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[1], bw_int_to_fixed(1000)));
+		} else {
+			calcs_output->stutter_entry_wm_ns[3].d_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[7], bw_int_to_fixed(1000)));
+			calcs_output->stutter_entry_wm_ns[4].d_mark =
+				bw_fixed_to_int(bw_mul(data->
+					stutter_entry_watermark[8], bw_int_to_fixed(1000)));
+		}
+		calcs_output->stutter_entry_wm_ns[5].d_mark =
+			bw_fixed_to_int(bw_mul(data->
+				stutter_entry_watermark[9], bw_int_to_fixed(1000)));
 
 		calcs_output->urgent_wm_ns[0].d_mark =
 			bw_fixed_to_int(bw_mul(data->
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c b/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c
index 4bb43a371292..080f777d705e 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c
+++ b/drivers/gpu/drm/amd/display/dc/calcs/dcn_calcs.c
@@ -31,6 +31,8 @@
 
 #include "resource.h"
 #include "dcn10/dcn10_resource.h"
+#include "dcn10/dcn10_hubbub.h"
+
 #include "dcn_calc_math.h"
 
 #define DC_LOGGER \
@@ -248,7 +250,24 @@ static void pipe_ctx_to_e2e_pipe_params (
 	else if (pipe->bottom_pipe != NULL && pipe->bottom_pipe->plane_state == pipe->plane_state)
 		input->src.is_hsplit = true;
 
-	input->src.dcc                 = pipe->plane_state->dcc.enable;
+	if (pipe->plane_res.dpp->ctx->dc->debug.optimized_watermark) {
+		/*
+		 * this method requires us to always re-calculate watermark when dcc change
+		 * between flip.
+		 */
+		input->src.dcc = pipe->plane_state->dcc.enable ? 1 : 0;
+	} else {
+		/*
+		 * allow us to disable dcc on the fly without re-calculating WM
+		 *
+		 * extra overhead for DCC is quite small.  for 1080p WM without
+		 * DCC is only 0.417us lower (urgent goes from 6.979us to 6.562us)
+		 */
+		unsigned int bpe;
+
+		input->src.dcc = pipe->plane_res.dpp->ctx->dc->res_pool->hubbub->funcs->
+			dcc_support_pixel_format(pipe->plane_state->format, &bpe) ? 1 : 0;
+	}
 	input->src.dcc_rate            = 1;
 	input->src.meta_pitch          = pipe->plane_state->dcc.grph.meta_pitch;
 	input->src.source_scan         = dm_horz;
@@ -423,6 +442,10 @@ static void dcn_bw_calc_rq_dlg_ttu(
 	int total_flip_bytes = 0;
 	int i;
 
+	memset(dlg_regs, 0, sizeof(*dlg_regs));
+	memset(ttu_regs, 0, sizeof(*ttu_regs));
+	memset(rq_regs, 0, sizeof(*rq_regs));
+
 	for (i = 0; i < number_of_planes; i++) {
 		total_active_bw += v->read_bandwidth[i];
 		total_prefetch_bw += v->prefetch_bandwidth[i];
@@ -501,6 +524,7 @@ static void split_stream_across_pipes(
 	resource_build_scaling_params(secondary_pipe);
 }
 
+#if 0
 static void calc_wm_sets_and_perf_params(
 		struct dc_state *context,
 		struct dcn_bw_internal_vars *v)
@@ -582,6 +606,7 @@ static void calc_wm_sets_and_perf_params(
 	if (v->voltage_level >= 3)
 		context->bw.dcn.watermarks.d = context->bw.dcn.watermarks.a;
 }
+#endif
 
 static bool dcn_bw_apply_registry_override(struct dc *dc)
 {
@@ -873,17 +898,36 @@ bool dcn_validate_bandwidth(
 			}
 
 			if (pipe->plane_state->rotation % 2 == 0) {
-				ASSERT(pipe->plane_res.scl_data.ratios.horz.value != dal_fixed31_32_one.value
+				ASSERT(pipe->plane_res.scl_data.ratios.horz.value != dc_fixpt_one.value
 					|| v->scaler_rec_out_width[input_idx] == v->viewport_width[input_idx]);
-				ASSERT(pipe->plane_res.scl_data.ratios.vert.value != dal_fixed31_32_one.value
+				ASSERT(pipe->plane_res.scl_data.ratios.vert.value != dc_fixpt_one.value
 					|| v->scaler_recout_height[input_idx] == v->viewport_height[input_idx]);
 			} else {
-				ASSERT(pipe->plane_res.scl_data.ratios.horz.value != dal_fixed31_32_one.value
+				ASSERT(pipe->plane_res.scl_data.ratios.horz.value != dc_fixpt_one.value
 					|| v->scaler_recout_height[input_idx] == v->viewport_width[input_idx]);
-				ASSERT(pipe->plane_res.scl_data.ratios.vert.value != dal_fixed31_32_one.value
+				ASSERT(pipe->plane_res.scl_data.ratios.vert.value != dc_fixpt_one.value
 					|| v->scaler_rec_out_width[input_idx] == v->viewport_height[input_idx]);
 			}
-			v->dcc_enable[input_idx] = pipe->plane_state->dcc.enable ? dcn_bw_yes : dcn_bw_no;
+
+			if (dc->debug.optimized_watermark) {
+				/*
+				 * this method requires us to always re-calculate watermark when dcc change
+				 * between flip.
+				 */
+				v->dcc_enable[input_idx] = pipe->plane_state->dcc.enable ? dcn_bw_yes : dcn_bw_no;
+			} else {
+				/*
+				 * allow us to disable dcc on the fly without re-calculating WM
+				 *
+				 * extra overhead for DCC is quite small.  for 1080p WM without
+				 * DCC is only 0.417us lower (urgent goes from 6.979us to 6.562us)
+				 */
+				unsigned int bpe;
+
+				v->dcc_enable[input_idx] = dc->res_pool->hubbub->funcs->dcc_support_pixel_format(
+						pipe->plane_state->format, &bpe) ? dcn_bw_yes : dcn_bw_no;
+			}
+
 			v->source_pixel_format[input_idx] = tl_pixel_format_to_bw_defs(
 					pipe->plane_state->format);
 			v->source_surface_mode[input_idx] = tl_sw_mode_to_bw_defs(
@@ -976,43 +1020,60 @@ bool dcn_validate_bandwidth(
 				bw_consumed = v->fabric_and_dram_bandwidth;
 
 		display_pipe_configuration(v);
-		calc_wm_sets_and_perf_params(context, v);
-		context->bw.dcn.calc_clk.fclk_khz = (int)(bw_consumed * 1000000 /
+		/*calc_wm_sets_and_perf_params(context, v);*/
+		/* Only 1 set is used by dcn since no noticeable
+		 * performance improvement was measured and due to hw bug DEGVIDCN10-254
+		 */
+		dispclkdppclkdcfclk_deep_sleep_prefetch_parameters_watermarks_and_performance_calculation(v);
+
+		context->bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns =
+			v->stutter_exit_watermark * 1000;
+		context->bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns =
+				v->stutter_enter_plus_exit_watermark * 1000;
+		context->bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns =
+				v->dram_clock_change_watermark * 1000;
+		context->bw.dcn.watermarks.a.pte_meta_urgent_ns = v->ptemeta_urgent_watermark * 1000;
+		context->bw.dcn.watermarks.a.urgent_ns = v->urgent_watermark * 1000;
+		context->bw.dcn.watermarks.b = context->bw.dcn.watermarks.a;
+		context->bw.dcn.watermarks.c = context->bw.dcn.watermarks.a;
+		context->bw.dcn.watermarks.d = context->bw.dcn.watermarks.a;
+
+		context->bw.dcn.clk.fclk_khz = (int)(bw_consumed * 1000000 /
 				(ddr4_dram_factor_single_Channel * v->number_of_channels));
 		if (bw_consumed == v->fabric_and_dram_bandwidth_vmin0p65) {
-			context->bw.dcn.calc_clk.fclk_khz = (int)(bw_consumed * 1000000 / 32);
+			context->bw.dcn.clk.fclk_khz = (int)(bw_consumed * 1000000 / 32);
 		}
 
-		context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz = (int)(v->dcf_clk_deep_sleep * 1000);
-		context->bw.dcn.calc_clk.dcfclk_khz = (int)(v->dcfclk * 1000);
+		context->bw.dcn.clk.dcfclk_deep_sleep_khz = (int)(v->dcf_clk_deep_sleep * 1000);
+		context->bw.dcn.clk.dcfclk_khz = (int)(v->dcfclk * 1000);
 
-		context->bw.dcn.calc_clk.dispclk_khz = (int)(v->dispclk * 1000);
+		context->bw.dcn.clk.dispclk_khz = (int)(v->dispclk * 1000);
 		if (dc->debug.max_disp_clk == true)
-			context->bw.dcn.calc_clk.dispclk_khz = (int)(dc->dcn_soc->max_dispclk_vmax0p9 * 1000);
+			context->bw.dcn.clk.dispclk_khz = (int)(dc->dcn_soc->max_dispclk_vmax0p9 * 1000);
 
-		if (context->bw.dcn.calc_clk.dispclk_khz <
+		if (context->bw.dcn.clk.dispclk_khz <
 				dc->debug.min_disp_clk_khz) {
-			context->bw.dcn.calc_clk.dispclk_khz =
+			context->bw.dcn.clk.dispclk_khz =
 					dc->debug.min_disp_clk_khz;
 		}
 
-		context->bw.dcn.calc_clk.dppclk_khz = context->bw.dcn.calc_clk.dispclk_khz / v->dispclk_dppclk_ratio;
-
+		context->bw.dcn.clk.dppclk_khz = context->bw.dcn.clk.dispclk_khz / v->dispclk_dppclk_ratio;
+		context->bw.dcn.clk.phyclk_khz = v->phyclk_per_state[v->voltage_level];
 		switch (v->voltage_level) {
 		case 0:
-			context->bw.dcn.calc_clk.max_supported_dppclk_khz =
+			context->bw.dcn.clk.max_supported_dppclk_khz =
 					(int)(dc->dcn_soc->max_dppclk_vmin0p65 * 1000);
 			break;
 		case 1:
-			context->bw.dcn.calc_clk.max_supported_dppclk_khz =
+			context->bw.dcn.clk.max_supported_dppclk_khz =
 					(int)(dc->dcn_soc->max_dppclk_vmid0p72 * 1000);
 			break;
 		case 2:
-			context->bw.dcn.calc_clk.max_supported_dppclk_khz =
+			context->bw.dcn.clk.max_supported_dppclk_khz =
 					(int)(dc->dcn_soc->max_dppclk_vnom0p8 * 1000);
 			break;
 		default:
-			context->bw.dcn.calc_clk.max_supported_dppclk_khz =
+			context->bw.dcn.clk.max_supported_dppclk_khz =
 					(int)(dc->dcn_soc->max_dppclk_vmax0p9 * 1000);
 			break;
 		}
@@ -1225,27 +1286,27 @@ static unsigned int dcn_find_normalized_clock_vdd_Level(
 
 unsigned int dcn_find_dcfclk_suits_all(
 	const struct dc *dc,
-	struct clocks_value *clocks)
+	struct dc_clocks *clocks)
 {
 	unsigned vdd_level, vdd_level_temp;
 	unsigned dcf_clk;
 
 	/*find a common supported voltage level*/
 	vdd_level = dcn_find_normalized_clock_vdd_Level(
-		dc, DM_PP_CLOCK_TYPE_DISPLAY_CLK, clocks->dispclk_in_khz);
+		dc, DM_PP_CLOCK_TYPE_DISPLAY_CLK, clocks->dispclk_khz);
 	vdd_level_temp = dcn_find_normalized_clock_vdd_Level(
-		dc, DM_PP_CLOCK_TYPE_DISPLAYPHYCLK, clocks->phyclk_in_khz);
+		dc, DM_PP_CLOCK_TYPE_DISPLAYPHYCLK, clocks->phyclk_khz);
 
 	vdd_level = dcn_bw_max(vdd_level, vdd_level_temp);
 	vdd_level_temp = dcn_find_normalized_clock_vdd_Level(
-		dc, DM_PP_CLOCK_TYPE_DPPCLK, clocks->dppclk_in_khz);
+		dc, DM_PP_CLOCK_TYPE_DPPCLK, clocks->dppclk_khz);
 	vdd_level = dcn_bw_max(vdd_level, vdd_level_temp);
 
 	vdd_level_temp = dcn_find_normalized_clock_vdd_Level(
-		dc, DM_PP_CLOCK_TYPE_MEMORY_CLK, clocks->dcfclock_in_khz);
+		dc, DM_PP_CLOCK_TYPE_MEMORY_CLK, clocks->fclk_khz);
 	vdd_level = dcn_bw_max(vdd_level, vdd_level_temp);
 	vdd_level_temp = dcn_find_normalized_clock_vdd_Level(
-		dc, DM_PP_CLOCK_TYPE_DCFCLK, clocks->dcfclock_in_khz);
+		dc, DM_PP_CLOCK_TYPE_DCFCLK, clocks->dcfclk_khz);
 
 	/*find that level conresponding dcfclk*/
 	vdd_level = dcn_bw_max(vdd_level, vdd_level_temp);
@@ -1331,21 +1392,14 @@ void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc)
 {
 	struct pp_smu_funcs_rv *pp = dc->res_pool->pp_smu;
 	struct pp_smu_wm_range_sets ranges = {0};
-	int max_fclk_khz, nom_fclk_khz, mid_fclk_khz, min_fclk_khz;
-	int max_dcfclk_khz, min_dcfclk_khz;
-	int socclk_khz;
+	int min_fclk_khz, min_dcfclk_khz, socclk_khz;
 	const int overdrive = 5000000; /* 5 GHz to cover Overdrive */
-	unsigned factor = (ddr4_dram_factor_single_Channel * dc->dcn_soc->number_of_channels);
 
 	if (!pp->set_wm_ranges)
 		return;
 
 	kernel_fpu_begin();
-	max_fclk_khz = dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9 * 1000000 / factor;
-	nom_fclk_khz = dc->dcn_soc->fabric_and_dram_bandwidth_vnom0p8 * 1000000 / factor;
-	mid_fclk_khz = dc->dcn_soc->fabric_and_dram_bandwidth_vmid0p72 * 1000000 / factor;
 	min_fclk_khz = dc->dcn_soc->fabric_and_dram_bandwidth_vmin0p65 * 1000000 / 32;
-	max_dcfclk_khz = dc->dcn_soc->dcfclkv_max0p9 * 1000;
 	min_dcfclk_khz = dc->dcn_soc->dcfclkv_min0p65 * 1000;
 	socclk_khz = dc->dcn_soc->socclk * 1000;
 	kernel_fpu_end();
@@ -1353,105 +1407,46 @@ void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc)
 	/* Now notify PPLib/SMU about which Watermarks sets they should select
 	 * depending on DPM state they are in. And update BW MGR GFX Engine and
 	 * Memory clock member variables for Watermarks calculations for each
-	 * Watermark Set
+	 * Watermark Set. Only one watermark set for dcn1 due to hw bug DEGVIDCN10-254.
 	 */
 	/* SOCCLK does not affect anytihng but writeback for DCN so for now we dont
 	 * care what the value is, hence min to overdrive level
 	 */
-	ranges.num_reader_wm_sets = WM_COUNT;
-	ranges.num_writer_wm_sets = WM_COUNT;
+	ranges.num_reader_wm_sets = WM_SET_COUNT;
+	ranges.num_writer_wm_sets = WM_SET_COUNT;
 	ranges.reader_wm_sets[0].wm_inst = WM_A;
 	ranges.reader_wm_sets[0].min_drain_clk_khz = min_dcfclk_khz;
-	ranges.reader_wm_sets[0].max_drain_clk_khz = max_dcfclk_khz;
+	ranges.reader_wm_sets[0].max_drain_clk_khz = overdrive;
 	ranges.reader_wm_sets[0].min_fill_clk_khz = min_fclk_khz;
-	ranges.reader_wm_sets[0].max_fill_clk_khz = min_fclk_khz;
+	ranges.reader_wm_sets[0].max_fill_clk_khz = overdrive;
 	ranges.writer_wm_sets[0].wm_inst = WM_A;
 	ranges.writer_wm_sets[0].min_fill_clk_khz = socclk_khz;
 	ranges.writer_wm_sets[0].max_fill_clk_khz = overdrive;
 	ranges.writer_wm_sets[0].min_drain_clk_khz = min_fclk_khz;
-	ranges.writer_wm_sets[0].max_drain_clk_khz = min_fclk_khz;
-
-	ranges.reader_wm_sets[1].wm_inst = WM_B;
-	ranges.reader_wm_sets[1].min_drain_clk_khz = min_fclk_khz;
-	ranges.reader_wm_sets[1].max_drain_clk_khz = max_dcfclk_khz;
-	ranges.reader_wm_sets[1].min_fill_clk_khz = mid_fclk_khz;
-	ranges.reader_wm_sets[1].max_fill_clk_khz = mid_fclk_khz;
-	ranges.writer_wm_sets[1].wm_inst = WM_B;
-	ranges.writer_wm_sets[1].min_fill_clk_khz = socclk_khz;
-	ranges.writer_wm_sets[1].max_fill_clk_khz = overdrive;
-	ranges.writer_wm_sets[1].min_drain_clk_khz = mid_fclk_khz;
-	ranges.writer_wm_sets[1].max_drain_clk_khz = mid_fclk_khz;
-
-
-	ranges.reader_wm_sets[2].wm_inst = WM_C;
-	ranges.reader_wm_sets[2].min_drain_clk_khz = min_fclk_khz;
-	ranges.reader_wm_sets[2].max_drain_clk_khz = max_dcfclk_khz;
-	ranges.reader_wm_sets[2].min_fill_clk_khz = nom_fclk_khz;
-	ranges.reader_wm_sets[2].max_fill_clk_khz = nom_fclk_khz;
-	ranges.writer_wm_sets[2].wm_inst = WM_C;
-	ranges.writer_wm_sets[2].min_fill_clk_khz = socclk_khz;
-	ranges.writer_wm_sets[2].max_fill_clk_khz = overdrive;
-	ranges.writer_wm_sets[2].min_drain_clk_khz = nom_fclk_khz;
-	ranges.writer_wm_sets[2].max_drain_clk_khz = nom_fclk_khz;
-
-	ranges.reader_wm_sets[3].wm_inst = WM_D;
-	ranges.reader_wm_sets[3].min_drain_clk_khz = min_fclk_khz;
-	ranges.reader_wm_sets[3].max_drain_clk_khz = max_dcfclk_khz;
-	ranges.reader_wm_sets[3].min_fill_clk_khz = max_fclk_khz;
-	ranges.reader_wm_sets[3].max_fill_clk_khz = max_fclk_khz;
-	ranges.writer_wm_sets[3].wm_inst = WM_D;
-	ranges.writer_wm_sets[3].min_fill_clk_khz = socclk_khz;
-	ranges.writer_wm_sets[3].max_fill_clk_khz = overdrive;
-	ranges.writer_wm_sets[3].min_drain_clk_khz = max_fclk_khz;
-	ranges.writer_wm_sets[3].max_drain_clk_khz = max_fclk_khz;
+	ranges.writer_wm_sets[0].max_drain_clk_khz = overdrive;
 
 	if (dc->debug.pplib_wm_report_mode == WM_REPORT_OVERRIDE) {
 		ranges.reader_wm_sets[0].wm_inst = WM_A;
 		ranges.reader_wm_sets[0].min_drain_clk_khz = 300000;
-		ranges.reader_wm_sets[0].max_drain_clk_khz = 654000;
+		ranges.reader_wm_sets[0].max_drain_clk_khz = 5000000;
 		ranges.reader_wm_sets[0].min_fill_clk_khz = 800000;
-		ranges.reader_wm_sets[0].max_fill_clk_khz = 800000;
+		ranges.reader_wm_sets[0].max_fill_clk_khz = 5000000;
 		ranges.writer_wm_sets[0].wm_inst = WM_A;
 		ranges.writer_wm_sets[0].min_fill_clk_khz = 200000;
-		ranges.writer_wm_sets[0].max_fill_clk_khz = 757000;
+		ranges.writer_wm_sets[0].max_fill_clk_khz = 5000000;
 		ranges.writer_wm_sets[0].min_drain_clk_khz = 800000;
-		ranges.writer_wm_sets[0].max_drain_clk_khz = 800000;
-
-		ranges.reader_wm_sets[1].wm_inst = WM_B;
-		ranges.reader_wm_sets[1].min_drain_clk_khz = 300000;
-		ranges.reader_wm_sets[1].max_drain_clk_khz = 654000;
-		ranges.reader_wm_sets[1].min_fill_clk_khz = 933000;
-		ranges.reader_wm_sets[1].max_fill_clk_khz = 933000;
-		ranges.writer_wm_sets[1].wm_inst = WM_B;
-		ranges.writer_wm_sets[1].min_fill_clk_khz = 200000;
-		ranges.writer_wm_sets[1].max_fill_clk_khz = 757000;
-		ranges.writer_wm_sets[1].min_drain_clk_khz = 933000;
-		ranges.writer_wm_sets[1].max_drain_clk_khz = 933000;
-
-
-		ranges.reader_wm_sets[2].wm_inst = WM_C;
-		ranges.reader_wm_sets[2].min_drain_clk_khz = 300000;
-		ranges.reader_wm_sets[2].max_drain_clk_khz = 654000;
-		ranges.reader_wm_sets[2].min_fill_clk_khz = 1067000;
-		ranges.reader_wm_sets[2].max_fill_clk_khz = 1067000;
-		ranges.writer_wm_sets[2].wm_inst = WM_C;
-		ranges.writer_wm_sets[2].min_fill_clk_khz = 200000;
-		ranges.writer_wm_sets[2].max_fill_clk_khz = 757000;
-		ranges.writer_wm_sets[2].min_drain_clk_khz = 1067000;
-		ranges.writer_wm_sets[2].max_drain_clk_khz = 1067000;
-
-		ranges.reader_wm_sets[3].wm_inst = WM_D;
-		ranges.reader_wm_sets[3].min_drain_clk_khz = 300000;
-		ranges.reader_wm_sets[3].max_drain_clk_khz = 654000;
-		ranges.reader_wm_sets[3].min_fill_clk_khz = 1200000;
-		ranges.reader_wm_sets[3].max_fill_clk_khz = 1200000;
-		ranges.writer_wm_sets[3].wm_inst = WM_D;
-		ranges.writer_wm_sets[3].min_fill_clk_khz = 200000;
-		ranges.writer_wm_sets[3].max_fill_clk_khz = 757000;
-		ranges.writer_wm_sets[3].min_drain_clk_khz = 1200000;
-		ranges.writer_wm_sets[3].max_drain_clk_khz = 1200000;
+		ranges.writer_wm_sets[0].max_drain_clk_khz = 5000000;
 	}
 
+	ranges.reader_wm_sets[1] = ranges.writer_wm_sets[0];
+	ranges.reader_wm_sets[1].wm_inst = WM_B;
+
+	ranges.reader_wm_sets[2] = ranges.writer_wm_sets[0];
+	ranges.reader_wm_sets[2].wm_inst = WM_C;
+
+	ranges.reader_wm_sets[3] = ranges.writer_wm_sets[0];
+	ranges.reader_wm_sets[3].wm_inst = WM_D;
+
 	/* Notify PP Lib/SMU which Watermarks to use for which clock ranges */
 	pp->set_wm_ranges(&pp->pp_smu, &ranges);
 }
@@ -1459,39 +1454,39 @@ void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc)
 void dcn_bw_sync_calcs_and_dml(struct dc *dc)
 {
 	kernel_fpu_begin();
-	DC_LOG_BANDWIDTH_CALCS("sr_exit_time: %d ns\n"
-			"sr_enter_plus_exit_time: %d ns\n"
-			"urgent_latency: %d ns\n"
-			"write_back_latency: %d ns\n"
-			"percent_of_ideal_drambw_received_after_urg_latency: %d %\n"
+	DC_LOG_BANDWIDTH_CALCS("sr_exit_time: %f ns\n"
+			"sr_enter_plus_exit_time: %f ns\n"
+			"urgent_latency: %f ns\n"
+			"write_back_latency: %f ns\n"
+			"percent_of_ideal_drambw_received_after_urg_latency: %f %%\n"
 			"max_request_size: %d bytes\n"
-			"dcfclkv_max0p9: %d kHz\n"
-			"dcfclkv_nom0p8: %d kHz\n"
-			"dcfclkv_mid0p72: %d kHz\n"
-			"dcfclkv_min0p65: %d kHz\n"
-			"max_dispclk_vmax0p9: %d kHz\n"
-			"max_dispclk_vnom0p8: %d kHz\n"
-			"max_dispclk_vmid0p72: %d kHz\n"
-			"max_dispclk_vmin0p65: %d kHz\n"
-			"max_dppclk_vmax0p9: %d kHz\n"
-			"max_dppclk_vnom0p8: %d kHz\n"
-			"max_dppclk_vmid0p72: %d kHz\n"
-			"max_dppclk_vmin0p65: %d kHz\n"
-			"socclk: %d kHz\n"
-			"fabric_and_dram_bandwidth_vmax0p9: %d MB/s\n"
-			"fabric_and_dram_bandwidth_vnom0p8: %d MB/s\n"
-			"fabric_and_dram_bandwidth_vmid0p72: %d MB/s\n"
-			"fabric_and_dram_bandwidth_vmin0p65: %d MB/s\n"
-			"phyclkv_max0p9: %d kHz\n"
-			"phyclkv_nom0p8: %d kHz\n"
-			"phyclkv_mid0p72: %d kHz\n"
-			"phyclkv_min0p65: %d kHz\n"
-			"downspreading: %d %\n"
+			"dcfclkv_max0p9: %f kHz\n"
+			"dcfclkv_nom0p8: %f kHz\n"
+			"dcfclkv_mid0p72: %f kHz\n"
+			"dcfclkv_min0p65: %f kHz\n"
+			"max_dispclk_vmax0p9: %f kHz\n"
+			"max_dispclk_vnom0p8: %f kHz\n"
+			"max_dispclk_vmid0p72: %f kHz\n"
+			"max_dispclk_vmin0p65: %f kHz\n"
+			"max_dppclk_vmax0p9: %f kHz\n"
+			"max_dppclk_vnom0p8: %f kHz\n"
+			"max_dppclk_vmid0p72: %f kHz\n"
+			"max_dppclk_vmin0p65: %f kHz\n"
+			"socclk: %f kHz\n"
+			"fabric_and_dram_bandwidth_vmax0p9: %f MB/s\n"
+			"fabric_and_dram_bandwidth_vnom0p8: %f MB/s\n"
+			"fabric_and_dram_bandwidth_vmid0p72: %f MB/s\n"
+			"fabric_and_dram_bandwidth_vmin0p65: %f MB/s\n"
+			"phyclkv_max0p9: %f kHz\n"
+			"phyclkv_nom0p8: %f kHz\n"
+			"phyclkv_mid0p72: %f kHz\n"
+			"phyclkv_min0p65: %f kHz\n"
+			"downspreading: %f %%\n"
 			"round_trip_ping_latency_cycles: %d DCFCLK Cycles\n"
 			"urgent_out_of_order_return_per_channel: %d Bytes\n"
 			"number_of_channels: %d\n"
 			"vmm_page_size: %d Bytes\n"
-			"dram_clock_change_latency: %d ns\n"
+			"dram_clock_change_latency: %f ns\n"
 			"return_bus_width: %d Bytes\n",
 			dc->dcn_soc->sr_exit_time * 1000,
 			dc->dcn_soc->sr_enter_plus_exit_time * 1000,
@@ -1527,11 +1522,11 @@ void dcn_bw_sync_calcs_and_dml(struct dc *dc)
 			dc->dcn_soc->vmm_page_size,
 			dc->dcn_soc->dram_clock_change_latency * 1000,
 			dc->dcn_soc->return_bus_width);
-	DC_LOG_BANDWIDTH_CALCS("rob_buffer_size_in_kbyte: %d\n"
-			"det_buffer_size_in_kbyte: %d\n"
-			"dpp_output_buffer_pixels: %d\n"
-			"opp_output_buffer_lines: %d\n"
-			"pixel_chunk_size_in_kbyte: %d\n"
+	DC_LOG_BANDWIDTH_CALCS("rob_buffer_size_in_kbyte: %f\n"
+			"det_buffer_size_in_kbyte: %f\n"
+			"dpp_output_buffer_pixels: %f\n"
+			"opp_output_buffer_lines: %f\n"
+			"pixel_chunk_size_in_kbyte: %f\n"
 			"pte_enable: %d\n"
 			"pte_chunk_size: %d kbytes\n"
 			"meta_chunk_size: %d kbytes\n"
@@ -1550,13 +1545,13 @@ void dcn_bw_sync_calcs_and_dml(struct dc *dc)
 			"max_pscl_tolb_throughput: %d pixels/dppclk\n"
 			"max_lb_tovscl_throughput: %d pixels/dppclk\n"
 			"max_vscl_tohscl_throughput: %d pixels/dppclk\n"
-			"max_hscl_ratio: %d\n"
-			"max_vscl_ratio: %d\n"
+			"max_hscl_ratio: %f\n"
+			"max_vscl_ratio: %f\n"
 			"max_hscl_taps: %d\n"
 			"max_vscl_taps: %d\n"
 			"pte_buffer_size_in_requests: %d\n"
-			"dispclk_ramping_margin: %d %\n"
-			"under_scan_factor: %d %\n"
+			"dispclk_ramping_margin: %f %%\n"
+			"under_scan_factor: %f %%\n"
 			"max_inter_dcn_tile_repeaters: %d\n"
 			"can_vstartup_lines_exceed_vsync_plus_back_porch_lines_minus_one: %d\n"
 			"bug_forcing_luma_and_chroma_request_to_same_size_fixed: %d\n"
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 9cd3566def8d..733ac224e7fd 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -169,6 +169,22 @@ failed_alloc:
 	return false;
 }
 
+/**
+ *****************************************************************************
+ *  Function: dc_stream_adjust_vmin_vmax
+ *
+ *  @brief
+ *     Looks up the pipe context of dc_stream_state and updates the
+ *     vertical_total_min and vertical_total_max of the DRR, Dynamic Refresh
+ *     Rate, which is a power-saving feature that targets reducing panel
+ *     refresh rate while the screen is static
+ *
+ *  @param [in] dc: dc reference
+ *  @param [in] stream: Initial dc stream state
+ *  @param [in] adjust: Updated parameters for vertical_total_min and
+ *  vertical_total_max
+ *****************************************************************************
+ */
 bool dc_stream_adjust_vmin_vmax(struct dc *dc,
 		struct dc_stream_state **streams, int num_streams,
 		int vmin, int vmax)
@@ -368,6 +384,71 @@ void dc_stream_set_static_screen_events(struct dc *dc,
 	dc->hwss.set_static_screen_control(pipes_affected, num_pipes_affected, events);
 }
 
+void dc_link_set_drive_settings(struct dc *dc,
+				struct link_training_settings *lt_settings,
+				const struct dc_link *link)
+{
+
+	int i;
+
+	for (i = 0; i < dc->link_count; i++) {
+		if (dc->links[i] == link)
+			break;
+	}
+
+	if (i >= dc->link_count)
+		ASSERT_CRITICAL(false);
+
+	dc_link_dp_set_drive_settings(dc->links[i], lt_settings);
+}
+
+void dc_link_perform_link_training(struct dc *dc,
+				   struct dc_link_settings *link_setting,
+				   bool skip_video_pattern)
+{
+	int i;
+
+	for (i = 0; i < dc->link_count; i++)
+		dc_link_dp_perform_link_training(
+			dc->links[i],
+			link_setting,
+			skip_video_pattern);
+}
+
+void dc_link_set_preferred_link_settings(struct dc *dc,
+					 struct dc_link_settings *link_setting,
+					 struct dc_link *link)
+{
+	link->preferred_link_setting = *link_setting;
+	dp_retrain_link_dp_test(link, link_setting, false);
+}
+
+void dc_link_enable_hpd(const struct dc_link *link)
+{
+	dc_link_dp_enable_hpd(link);
+}
+
+void dc_link_disable_hpd(const struct dc_link *link)
+{
+	dc_link_dp_disable_hpd(link);
+}
+
+
+void dc_link_set_test_pattern(struct dc_link *link,
+			      enum dp_test_pattern test_pattern,
+			      const struct link_training_settings *p_link_settings,
+			      const unsigned char *p_custom_pattern,
+			      unsigned int cust_pattern_size)
+{
+	if (link != NULL)
+		dc_link_dp_set_test_pattern(
+			link,
+			test_pattern,
+			p_link_settings,
+			p_custom_pattern,
+			cust_pattern_size);
+}
+
 static void destruct(struct dc *dc)
 {
 	dc_release_state(dc->current_state);
@@ -386,9 +467,6 @@ static void destruct(struct dc *dc)
 	if (dc->ctx->created_bios)
 		dal_bios_parser_destroy(&dc->ctx->dc_bios);
 
-	if (dc->ctx->logger)
-		dal_logger_destroy(&dc->ctx->logger);
-
 	kfree(dc->ctx);
 	dc->ctx = NULL;
 
@@ -398,7 +476,7 @@ static void destruct(struct dc *dc)
 	kfree(dc->bw_dceip);
 	dc->bw_dceip = NULL;
 
-#ifdef CONFIG_DRM_AMD_DC_DCN1_0
+#ifdef CONFIG_X86
 	kfree(dc->dcn_soc);
 	dc->dcn_soc = NULL;
 
@@ -411,11 +489,10 @@ static void destruct(struct dc *dc)
 static bool construct(struct dc *dc,
 		const struct dc_init_data *init_params)
 {
-	struct dal_logger *logger;
 	struct dc_context *dc_ctx;
 	struct bw_calcs_dceip *dc_dceip;
 	struct bw_calcs_vbios *dc_vbios;
-#ifdef CONFIG_DRM_AMD_DC_DCN1_0
+#ifdef CONFIG_X86
 	struct dcn_soc_bounding_box *dcn_soc;
 	struct dcn_ip_params *dcn_ip;
 #endif
@@ -437,7 +514,7 @@ static bool construct(struct dc *dc,
 	}
 
 	dc->bw_vbios = dc_vbios;
-#ifdef CONFIG_DRM_AMD_DC_DCN1_0
+#ifdef CONFIG_X86
 	dcn_soc = kzalloc(sizeof(*dcn_soc), GFP_KERNEL);
 	if (!dcn_soc) {
 		dm_error("%s: failed to create dcn_soc\n", __func__);
@@ -465,6 +542,7 @@ static bool construct(struct dc *dc,
 	dc_ctx->driver_context = init_params->driver;
 	dc_ctx->dc = dc;
 	dc_ctx->asic_id = init_params->asic_id;
+	dc_ctx->dc_sink_id_count = 0;
 	dc->ctx = dc_ctx;
 
 	dc->current_state = dc_create_state();
@@ -475,14 +553,7 @@ static bool construct(struct dc *dc,
 	}
 
 	/* Create logger */
-	logger = dal_logger_create(dc_ctx, init_params->log_mask);
 
-	if (!logger) {
-		/* can *not* call logger. call base driver 'print error' */
-		dm_error("%s: failed to create Logger!\n", __func__);
-		goto fail;
-	}
-	dc_ctx->logger = logger;
 	dc_ctx->dce_environment = init_params->dce_environment;
 
 	dc_version = resource_parse_asic_id(init_params->asic_id);
@@ -901,9 +972,7 @@ bool dc_commit_state(struct dc *dc, struct dc_state *context)
 	for (i = 0; i < context->stream_count; i++) {
 		struct dc_stream_state *stream = context->streams[i];
 
-		dc_stream_log(stream,
-				dc->ctx->logger,
-				LOG_DC);
+		dc_stream_log(dc, stream);
 	}
 
 	result = dc_commit_state_no_check(dc, context);
@@ -927,101 +996,7 @@ bool dc_post_update_surfaces_to_stream(struct dc *dc)
 
 	dc->optimized_required = false;
 
-	/* 3rd param should be true, temp w/a for RV*/
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
-	dc->hwss.set_bandwidth(dc, context, dc->ctx->dce_version < DCN_VERSION_1_0);
-#else
 	dc->hwss.set_bandwidth(dc, context, true);
-#endif
-	return true;
-}
-
-/*
- * TODO this whole function needs to go
- *
- * dc_surface_update is needlessly complex. See if we can just replace this
- * with a dc_plane_state and follow the atomic model a bit more closely here.
- */
-bool dc_commit_planes_to_stream(
-		struct dc *dc,
-		struct dc_plane_state **plane_states,
-		uint8_t new_plane_count,
-		struct dc_stream_state *dc_stream,
-		struct dc_state *state)
-{
-	/* no need to dynamically allocate this. it's pretty small */
-	struct dc_surface_update updates[MAX_SURFACES];
-	struct dc_flip_addrs *flip_addr;
-	struct dc_plane_info *plane_info;
-	struct dc_scaling_info *scaling_info;
-	int i;
-	struct dc_stream_update *stream_update =
-			kzalloc(sizeof(struct dc_stream_update), GFP_KERNEL);
-
-	if (!stream_update) {
-		BREAK_TO_DEBUGGER();
-		return false;
-	}
-
-	flip_addr = kcalloc(MAX_SURFACES, sizeof(struct dc_flip_addrs),
-			    GFP_KERNEL);
-	plane_info = kcalloc(MAX_SURFACES, sizeof(struct dc_plane_info),
-			     GFP_KERNEL);
-	scaling_info = kcalloc(MAX_SURFACES, sizeof(struct dc_scaling_info),
-			       GFP_KERNEL);
-
-	if (!flip_addr || !plane_info || !scaling_info) {
-		kfree(flip_addr);
-		kfree(plane_info);
-		kfree(scaling_info);
-		kfree(stream_update);
-		return false;
-	}
-
-	memset(updates, 0, sizeof(updates));
-
-	stream_update->src = dc_stream->src;
-	stream_update->dst = dc_stream->dst;
-	stream_update->out_transfer_func = dc_stream->out_transfer_func;
-
-	for (i = 0; i < new_plane_count; i++) {
-		updates[i].surface = plane_states[i];
-		updates[i].gamma =
-			(struct dc_gamma *)plane_states[i]->gamma_correction;
-		updates[i].in_transfer_func = plane_states[i]->in_transfer_func;
-		flip_addr[i].address = plane_states[i]->address;
-		flip_addr[i].flip_immediate = plane_states[i]->flip_immediate;
-		plane_info[i].color_space = plane_states[i]->color_space;
-		plane_info[i].input_tf = plane_states[i]->input_tf;
-		plane_info[i].format = plane_states[i]->format;
-		plane_info[i].plane_size = plane_states[i]->plane_size;
-		plane_info[i].rotation = plane_states[i]->rotation;
-		plane_info[i].horizontal_mirror = plane_states[i]->horizontal_mirror;
-		plane_info[i].stereo_format = plane_states[i]->stereo_format;
-		plane_info[i].tiling_info = plane_states[i]->tiling_info;
-		plane_info[i].visible = plane_states[i]->visible;
-		plane_info[i].per_pixel_alpha = plane_states[i]->per_pixel_alpha;
-		plane_info[i].dcc = plane_states[i]->dcc;
-		scaling_info[i].scaling_quality = plane_states[i]->scaling_quality;
-		scaling_info[i].src_rect = plane_states[i]->src_rect;
-		scaling_info[i].dst_rect = plane_states[i]->dst_rect;
-		scaling_info[i].clip_rect = plane_states[i]->clip_rect;
-
-		updates[i].flip_addr = &flip_addr[i];
-		updates[i].plane_info = &plane_info[i];
-		updates[i].scaling_info = &scaling_info[i];
-	}
-
-	dc_commit_updates_for_stream(
-			dc,
-			updates,
-			new_plane_count,
-			dc_stream, stream_update, plane_states, state);
-
-	kfree(flip_addr);
-	kfree(plane_info);
-	kfree(scaling_info);
-	kfree(stream_update);
 	return true;
 }
 
@@ -1107,9 +1082,6 @@ static enum surface_update_type get_plane_info_update_type(const struct dc_surfa
 	if (u->plane_info->color_space != u->surface->color_space)
 		update_flags->bits.color_space_change = 1;
 
-	if (u->plane_info->input_tf != u->surface->input_tf)
-		update_flags->bits.input_tf_change = 1;
-
 	if (u->plane_info->horizontal_mirror != u->surface->horizontal_mirror)
 		update_flags->bits.horizontal_mirror_change = 1;
 
@@ -1243,12 +1215,20 @@ static enum surface_update_type det_surface_update(const struct dc *dc,
 	if (u->input_csc_color_matrix)
 		update_flags->bits.input_csc_change = 1;
 
-	if (update_flags->bits.in_transfer_func_change
-			|| update_flags->bits.input_csc_change) {
+	if (u->coeff_reduction_factor)
+		update_flags->bits.coeff_reduction_change = 1;
+
+	if (update_flags->bits.in_transfer_func_change) {
 		type = UPDATE_TYPE_MED;
 		elevate_update_type(&overall_type, type);
 	}
 
+	if (update_flags->bits.input_csc_change
+			|| update_flags->bits.coeff_reduction_change) {
+		type = UPDATE_TYPE_FULL;
+		elevate_update_type(&overall_type, type);
+	}
+
 	return overall_type;
 }
 
@@ -1297,7 +1277,7 @@ enum surface_update_type dc_check_update_surfaces_for_stream(
 	type = check_update_surfaces_for_stream(dc, updates, surface_count, stream_update, stream_status);
 	if (type == UPDATE_TYPE_FULL)
 		for (i = 0; i < surface_count; i++)
-			updates[i].surface->update_flags.bits.full_update = 1;
+			updates[i].surface->update_flags.raw = 0xFFFFFFFF;
 
 	return type;
 }
@@ -1375,6 +1355,12 @@ static void commit_planes_for_stream(struct dc *dc,
 					pipe_ctx->stream_res.abm->funcs->set_abm_level(
 							pipe_ctx->stream_res.abm, stream->abm_level);
 			}
+
+			if (stream_update && stream_update->periodic_fn_vsync_delta &&
+					pipe_ctx->stream_res.tg->funcs->program_vline_interrupt)
+				pipe_ctx->stream_res.tg->funcs->program_vline_interrupt(
+						pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing,
+						pipe_ctx->stream->periodic_fn_vsync_delta);
 		}
 	}
 
@@ -1626,7 +1612,7 @@ struct dc_sink *dc_link_add_remote_sink(
 	struct dc_sink *dc_sink;
 	enum dc_edid_status edid_status;
 
-	if (len > MAX_EDID_BUFFER_SIZE) {
+	if (len > DC_MAX_EDID_BUFFER_SIZE) {
 		dm_error("Max EDID buffer size breached!\n");
 		return NULL;
 	}
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
index 5a552cb3f8a7..caece7c13bc6 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
@@ -36,8 +36,9 @@
 #include "hw_sequencer.h"
 
 #include "resource.h"
-#define DC_LOGGER \
-	logger
+
+#define DC_LOGGER_INIT(logger)
+
 
 #define SURFACE_TRACE(...) do {\
 		if (dc->debug.surface_trace) \
@@ -60,8 +61,7 @@ void pre_surface_trace(
 		int surface_count)
 {
 	int i;
-	struct dc  *core_dc = dc;
-	struct dal_logger *logger =  core_dc->ctx->logger;
+	DC_LOGGER_INIT(dc->ctx->logger);
 
 	for (i = 0; i < surface_count; i++) {
 		const struct dc_plane_state *plane_state = plane_states[i];
@@ -72,8 +72,8 @@ void pre_surface_trace(
 				"plane_state->visible = %d;\n"
 				"plane_state->flip_immediate = %d;\n"
 				"plane_state->address.type = %d;\n"
-				"plane_state->address.grph.addr.quad_part = 0x%X;\n"
-				"plane_state->address.grph.meta_addr.quad_part = 0x%X;\n"
+				"plane_state->address.grph.addr.quad_part = 0x%llX;\n"
+				"plane_state->address.grph.meta_addr.quad_part = 0x%llX;\n"
 				"plane_state->scaling_quality.h_taps = %d;\n"
 				"plane_state->scaling_quality.v_taps = %d;\n"
 				"plane_state->scaling_quality.h_taps_c = %d;\n"
@@ -155,7 +155,6 @@ void pre_surface_trace(
 				"plane_state->tiling_info.gfx8.pipe_config = %d;\n"
 				"plane_state->tiling_info.gfx8.array_mode = %d;\n"
 				"plane_state->color_space = %d;\n"
-				"plane_state->input_tf = %d;\n"
 				"plane_state->dcc.enable = %d;\n"
 				"plane_state->format = %d;\n"
 				"plane_state->rotation = %d;\n"
@@ -163,7 +162,6 @@ void pre_surface_trace(
 				plane_state->tiling_info.gfx8.pipe_config,
 				plane_state->tiling_info.gfx8.array_mode,
 				plane_state->color_space,
-				plane_state->input_tf,
 				plane_state->dcc.enable,
 				plane_state->format,
 				plane_state->rotation,
@@ -183,8 +181,7 @@ void update_surface_trace(
 		int surface_count)
 {
 	int i;
-	struct dc  *core_dc = dc;
-	struct dal_logger *logger =  core_dc->ctx->logger;
+	DC_LOGGER_INIT(dc->ctx->logger);
 
 	for (i = 0; i < surface_count; i++) {
 		const struct dc_surface_update *update = &updates[i];
@@ -192,8 +189,8 @@ void update_surface_trace(
 		SURFACE_TRACE("Update %d\n", i);
 		if (update->flip_addr) {
 			SURFACE_TRACE("flip_addr->address.type = %d;\n"
-					"flip_addr->address.grph.addr.quad_part = 0x%X;\n"
-					"flip_addr->address.grph.meta_addr.quad_part = 0x%X;\n"
+					"flip_addr->address.grph.addr.quad_part = 0x%llX;\n"
+					"flip_addr->address.grph.meta_addr.quad_part = 0x%llX;\n"
 					"flip_addr->flip_immediate = %d;\n",
 					update->flip_addr->address.type,
 					update->flip_addr->address.grph.addr.quad_part,
@@ -204,16 +201,15 @@ void update_surface_trace(
 		if (update->plane_info) {
 			SURFACE_TRACE(
 					"plane_info->color_space = %d;\n"
-					"plane_info->input_tf = %d;\n"
 					"plane_info->format = %d;\n"
 					"plane_info->plane_size.grph.surface_pitch = %d;\n"
 					"plane_info->plane_size.grph.surface_size.height = %d;\n"
 					"plane_info->plane_size.grph.surface_size.width = %d;\n"
 					"plane_info->plane_size.grph.surface_size.x = %d;\n"
 					"plane_info->plane_size.grph.surface_size.y = %d;\n"
-					"plane_info->rotation = %d;\n",
+					"plane_info->rotation = %d;\n"
+					"plane_info->stereo_format = %d;\n",
 					update->plane_info->color_space,
-					update->plane_info->input_tf,
 					update->plane_info->format,
 					update->plane_info->plane_size.grph.surface_pitch,
 					update->plane_info->plane_size.grph.surface_size.height,
@@ -303,8 +299,7 @@ void update_surface_trace(
 
 void post_surface_trace(struct dc *dc)
 {
-	struct dc  *core_dc = dc;
-	struct dal_logger *logger =  core_dc->ctx->logger;
+	DC_LOGGER_INIT(dc->ctx->logger);
 
 	SURFACE_TRACE("post surface process.\n");
 
@@ -316,10 +311,10 @@ void context_timing_trace(
 {
 	int i;
 	struct dc  *core_dc = dc;
-	struct dal_logger *logger =  core_dc->ctx->logger;
 	int h_pos[MAX_PIPES], v_pos[MAX_PIPES];
 	struct crtc_position position;
 	unsigned int underlay_idx = core_dc->res_pool->underlay_pipe_index;
+	DC_LOGGER_INIT(dc->ctx->logger);
 
 
 	for (i = 0; i < core_dc->res_pool->pipe_count; i++) {
@@ -353,24 +348,23 @@ void context_clock_trace(
 		struct dc *dc,
 		struct dc_state *context)
 {
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
-	struct dc  *core_dc = dc;
-	struct dal_logger *logger =  core_dc->ctx->logger;
-
+#ifdef CONFIG_X86
+	DC_LOGGER_INIT(dc->ctx->logger);
 	CLOCK_TRACE("Current: dispclk_khz:%d  max_dppclk_khz:%d  dcfclk_khz:%d\n"
 			"dcfclk_deep_sleep_khz:%d  fclk_khz:%d  socclk_khz:%d\n",
-			context->bw.dcn.calc_clk.dispclk_khz,
-			context->bw.dcn.calc_clk.dppclk_khz,
-			context->bw.dcn.calc_clk.dcfclk_khz,
-			context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz,
-			context->bw.dcn.calc_clk.fclk_khz,
-			context->bw.dcn.calc_clk.socclk_khz);
+			context->bw.dcn.clk.dispclk_khz,
+			context->bw.dcn.clk.dppclk_khz,
+			context->bw.dcn.clk.dcfclk_khz,
+			context->bw.dcn.clk.dcfclk_deep_sleep_khz,
+			context->bw.dcn.clk.fclk_khz,
+			context->bw.dcn.clk.socclk_khz);
 	CLOCK_TRACE("Calculated: dispclk_khz:%d  max_dppclk_khz:%d  dcfclk_khz:%d\n"
 			"dcfclk_deep_sleep_khz:%d  fclk_khz:%d  socclk_khz:%d\n",
-			context->bw.dcn.calc_clk.dispclk_khz,
-			context->bw.dcn.calc_clk.dppclk_khz,
-			context->bw.dcn.calc_clk.dcfclk_khz,
-			context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz,
-			context->bw.dcn.calc_clk.fclk_khz);
+			context->bw.dcn.clk.dispclk_khz,
+			context->bw.dcn.clk.dppclk_khz,
+			context->bw.dcn.clk.dcfclk_khz,
+			context->bw.dcn.clk.dcfclk_deep_sleep_khz,
+			context->bw.dcn.clk.fclk_khz,
+			context->bw.dcn.clk.socclk_khz);
 #endif
 }
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
index ebc96b720083..83d121510ef5 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
@@ -208,6 +208,7 @@ void color_space_to_black_color(
 	case COLOR_SPACE_YCBCR709:
 	case COLOR_SPACE_YCBCR601_LIMITED:
 	case COLOR_SPACE_YCBCR709_LIMITED:
+	case COLOR_SPACE_2020_YCBCR:
 		*black_color = black_color_format[BLACK_COLOR_FORMAT_YUV_CV];
 		break;
 
@@ -216,7 +217,25 @@ void color_space_to_black_color(
 			black_color_format[BLACK_COLOR_FORMAT_RGB_LIMITED];
 		break;
 
-	default:
+	/**
+	 * Remove default and add case for all color space
+	 * so when we forget to add new color space
+	 * compiler will give a warning
+	 */
+	case COLOR_SPACE_UNKNOWN:
+	case COLOR_SPACE_SRGB:
+	case COLOR_SPACE_XR_RGB:
+	case COLOR_SPACE_MSREF_SCRGB:
+	case COLOR_SPACE_XV_YCC_709:
+	case COLOR_SPACE_XV_YCC_601:
+	case COLOR_SPACE_2020_RGB_FULLRANGE:
+	case COLOR_SPACE_2020_RGB_LIMITEDRANGE:
+	case COLOR_SPACE_ADOBERGB:
+	case COLOR_SPACE_DCIP3:
+	case COLOR_SPACE_DISPLAYNATIVE:
+	case COLOR_SPACE_DOLBYVISION:
+	case COLOR_SPACE_APPCTRL:
+	case COLOR_SPACE_CUSTOMPOINTS:
 		/* fefault is sRGB black (full range). */
 		*black_color =
 			black_color_format[BLACK_COLOR_FORMAT_RGB_FULLRANGE];
@@ -230,6 +249,9 @@ bool hwss_wait_for_blank_complete(
 {
 	int counter;
 
+	/* Not applicable if the pipe is not primary, save 300ms of boot time */
+	if (!tg->funcs->is_blanked)
+		return true;
 	for (counter = 0; counter < 100; counter++) {
 		if (tg->funcs->is_blanked(tg))
 			break;
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
index 6d1c4981a185..a4429c90c60c 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
@@ -33,6 +33,7 @@
 #include "dc_link_dp.h"
 #include "dc_link_ddc.h"
 #include "link_hwss.h"
+#include "opp.h"
 
 #include "link_encoder.h"
 #include "hw_sequencer.h"
@@ -45,8 +46,9 @@
 #include "dce/dce_11_0_d.h"
 #include "dce/dce_11_0_enum.h"
 #include "dce/dce_11_0_sh_mask.h"
-#define DC_LOGGER \
-	dc_ctx->logger
+
+#define DC_LOGGER_INIT(logger)
+
 
 #define LINK_INFO(...) \
 	DC_LOG_HW_HOTPLUG(  \
@@ -311,7 +313,7 @@ static enum signal_type get_basic_signal_type(
  * @brief
  * Check whether there is a dongle on DP connector
  */
-static bool is_dp_sink_present(struct dc_link *link)
+bool dc_link_is_dp_sink_present(struct dc_link *link)
 {
 	enum gpio_result gpio_result;
 	uint32_t clock_pin = 0;
@@ -404,7 +406,7 @@ static enum signal_type link_detect_sink(
 			 * we assume signal is DVI; it could be corrected
 			 * to HDMI after dongle detection
 			 */
-			if (!is_dp_sink_present(link))
+			if (!dm_helpers_is_dp_sink_present(link))
 				result = SIGNAL_TYPE_DVI_SINGLE_LINK;
 		}
 	}
@@ -468,6 +470,13 @@ static void link_disconnect_sink(struct dc_link *link)
 	link->dpcd_sink_count = 0;
 }
 
+static void link_disconnect_remap(struct dc_sink *prev_sink, struct dc_link *link)
+{
+	dc_sink_release(link->local_sink);
+	link->local_sink = prev_sink;
+}
+
+
 static bool detect_dp(
 	struct dc_link *link,
 	struct display_sink_capability *sink_caps,
@@ -489,6 +498,10 @@ static bool detect_dp(
 			sink_caps->signal = SIGNAL_TYPE_DISPLAY_PORT_MST;
 			link->type = dc_connection_mst_branch;
 
+			dal_ddc_service_set_transaction_type(
+							link->ddc,
+							sink_caps->transaction_type);
+
 			/*
 			 * This call will initiate MST topology discovery. Which
 			 * will detect MST ports and add new DRM connector DRM
@@ -516,6 +529,10 @@ static bool detect_dp(
 			if (reason == DETECT_REASON_BOOT)
 				boot = true;
 
+			dm_helpers_dp_update_branch_info(
+				link->ctx,
+				link);
+
 			if (!dm_helpers_dp_mst_start_top_mgr(
 				link->ctx,
 				link, boot)) {
@@ -550,6 +567,17 @@ static bool detect_dp(
 	return true;
 }
 
+static bool is_same_edid(struct dc_edid *old_edid, struct dc_edid *new_edid)
+{
+	if (old_edid->length != new_edid->length)
+		return false;
+
+	if (new_edid->length == 0)
+		return false;
+
+	return (memcmp(old_edid->raw_edid, new_edid->raw_edid, new_edid->length) == 0);
+}
+
 bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 {
 	struct dc_sink_init_data sink_init_data = { 0 };
@@ -557,11 +585,15 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 	uint8_t i;
 	bool converter_disable_audio = false;
 	struct audio_support *aud_support = &link->dc->res_pool->audio_support;
+	bool same_edid = false;
 	enum dc_edid_status edid_status;
 	struct dc_context *dc_ctx = link->ctx;
 	struct dc_sink *sink = NULL;
+	struct dc_sink *prev_sink = NULL;
+	struct dpcd_caps prev_dpcd_caps;
+	bool same_dpcd = true;
 	enum dc_connection_type new_connection_type = dc_connection_none;
-
+	DC_LOGGER_INIT(link->ctx->logger);
 	if (link->connector_signal == SIGNAL_TYPE_VIRTUAL)
 		return false;
 
@@ -574,6 +606,11 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 			link->local_sink)
 		return true;
 
+	prev_sink = link->local_sink;
+	if (prev_sink != NULL) {
+		dc_sink_retain(prev_sink);
+		memcpy(&prev_dpcd_caps, &link->dpcd_caps, sizeof(struct dpcd_caps));
+	}
 	link_disconnect_sink(link);
 
 	if (new_connection_type != dc_connection_none) {
@@ -615,14 +652,25 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 				link,
 				&sink_caps,
 				&converter_disable_audio,
-				aud_support, reason))
+				aud_support, reason)) {
+				if (prev_sink != NULL)
+					dc_sink_release(prev_sink);
 				return false;
+			}
 
+			// Check if dpcp block is the same
+			if (prev_sink != NULL) {
+				if (memcmp(&link->dpcd_caps, &prev_dpcd_caps, sizeof(struct dpcd_caps)))
+					same_dpcd = false;
+			}
 			/* Active dongle downstream unplug */
 			if (link->type == dc_connection_active_dongle
 					&& link->dpcd_caps.sink_count.
-					bits.SINK_COUNT == 0)
+					bits.SINK_COUNT == 0) {
+				if (prev_sink != NULL)
+					dc_sink_release(prev_sink);
 				return true;
+			}
 
 			if (link->type == dc_connection_mst_branch) {
 				LINK_INFO("link=%d, mst branch is now Connected\n",
@@ -630,9 +678,11 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 				/* Need to setup mst link_cap struct here
 				 * otherwise dc_link_detect() will leave mst link_cap
 				 * empty which leads to allocate_mst_payload() has "0"
-				 * pbn_per_slot value leading to exception on dal_fixed31_32_div()
+				 * pbn_per_slot value leading to exception on dc_fixpt_div()
 				 */
 				link->verified_link_cap = link->reported_link_cap;
+				if (prev_sink != NULL)
+					dc_sink_release(prev_sink);
 				return false;
 			}
 
@@ -642,6 +692,8 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 		default:
 			DC_ERROR("Invalid connector type! signal:%d\n",
 				link->connector_signal);
+			if (prev_sink != NULL)
+				dc_sink_release(prev_sink);
 			return false;
 		} /* switch() */
 
@@ -664,6 +716,8 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 		sink = dc_sink_create(&sink_init_data);
 		if (!sink) {
 			DC_ERROR("Failed to create sink!\n");
+			if (prev_sink != NULL)
+				dc_sink_release(prev_sink);
 			return false;
 		}
 
@@ -687,22 +741,33 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 			break;
 		}
 
-		if (link->connector_signal == SIGNAL_TYPE_DISPLAY_PORT &&
-			sink_caps.transaction_type ==
-			DDC_TRANSACTION_TYPE_I2C_OVER_AUX) {
-			/*
-			 * TODO debug why Dell 2413 doesn't like
-			 *  two link trainings
-			 */
+		// Check if edid is the same
+		if ((prev_sink != NULL) && ((edid_status == EDID_THE_SAME) || (edid_status == EDID_OK)))
+			same_edid = is_same_edid(&prev_sink->dc_edid, &sink->dc_edid);
 
-			/* deal with non-mst cases */
-			dp_hbr_verify_link_cap(link, &link->reported_link_cap);
-		}
+		// If both edid and dpcd are the same, then discard new sink and revert back to original sink
+		if ((same_edid) && (same_dpcd)) {
+			link_disconnect_remap(prev_sink, link);
+			sink = prev_sink;
+			prev_sink = NULL;
+		} else {
+			if (link->connector_signal == SIGNAL_TYPE_DISPLAY_PORT &&
+					sink_caps.transaction_type ==
+						DDC_TRANSACTION_TYPE_I2C_OVER_AUX) {
+				/*
+				 * TODO debug why Dell 2413 doesn't like
+				 *  two link trainings
+				 */
 
-		/* HDMI-DVI Dongle */
-		if (sink->sink_signal == SIGNAL_TYPE_HDMI_TYPE_A &&
-				!sink->edid_caps.edid_hdmi)
-			sink->sink_signal = SIGNAL_TYPE_DVI_SINGLE_LINK;
+				/* deal with non-mst cases */
+				dp_hbr_verify_link_cap(link, &link->reported_link_cap);
+			}
+
+			/* HDMI-DVI Dongle */
+			if (sink->sink_signal == SIGNAL_TYPE_HDMI_TYPE_A &&
+					!sink->edid_caps.edid_hdmi)
+				sink->sink_signal = SIGNAL_TYPE_DVI_SINGLE_LINK;
+		}
 
 		/* Connectivity log: detection */
 		for (i = 0; i < sink->dc_edid.length / EDID_BLOCK_SIZE; i++) {
@@ -761,10 +826,14 @@ bool dc_link_detect(struct dc_link *link, enum dc_detect_reason reason)
 		sink_caps.signal = SIGNAL_TYPE_NONE;
 	}
 
-	LINK_INFO("link=%d, dc_sink_in=%p is now %s\n",
+	LINK_INFO("link=%d, dc_sink_in=%p is now %s prev_sink=%p dpcd same=%d edid same=%d\n",
 		link->link_index, sink,
 		(sink_caps.signal == SIGNAL_TYPE_NONE ?
-			"Disconnected":"Connected"));
+			"Disconnected":"Connected"), prev_sink,
+			same_dpcd, same_edid);
+
+	if (prev_sink != NULL)
+		dc_sink_release(prev_sink);
 
 	return true;
 }
@@ -927,6 +996,7 @@ static bool construct(
 	struct integrated_info info = {{{ 0 }}};
 	struct dc_bios *bios = init_params->dc->ctx->dc_bios;
 	const struct dc_vbios_funcs *bp_funcs = bios->funcs;
+	DC_LOGGER_INIT(dc_ctx->logger);
 
 	link->irq_source_hpd = DC_IRQ_SOURCE_INVALID;
 	link->irq_source_hpd_rx = DC_IRQ_SOURCE_INVALID;
@@ -1135,7 +1205,8 @@ static void dpcd_configure_panel_mode(
 {
 	union dpcd_edp_config edp_config_set;
 	bool panel_mode_edp = false;
-	struct dc_context *dc_ctx = link->ctx;
+	DC_LOGGER_INIT(link->ctx->logger);
+
 	memset(&edp_config_set, '\0', sizeof(union dpcd_edp_config));
 
 	if (DP_PANEL_MODE_DEFAULT != panel_mode) {
@@ -1183,16 +1254,21 @@ static void enable_stream_features(struct pipe_ctx *pipe_ctx)
 {
 	struct dc_stream_state *stream = pipe_ctx->stream;
 	struct dc_link *link = stream->sink->link;
-	union down_spread_ctrl downspread;
+	union down_spread_ctrl old_downspread;
+	union down_spread_ctrl new_downspread;
 
 	core_link_read_dpcd(link, DP_DOWNSPREAD_CTRL,
-			&downspread.raw, sizeof(downspread));
+			&old_downspread.raw, sizeof(old_downspread));
 
-	downspread.bits.IGNORE_MSA_TIMING_PARAM =
+	new_downspread.raw = old_downspread.raw;
+
+	new_downspread.bits.IGNORE_MSA_TIMING_PARAM =
 			(stream->ignore_msa_timing_param) ? 1 : 0;
 
-	core_link_write_dpcd(link, DP_DOWNSPREAD_CTRL,
-			&downspread.raw, sizeof(downspread));
+	if (new_downspread.raw != old_downspread.raw) {
+		core_link_write_dpcd(link, DP_DOWNSPREAD_CTRL,
+			&new_downspread.raw, sizeof(new_downspread));
+	}
 }
 
 static enum dc_status enable_link_dp(
@@ -1217,29 +1293,15 @@ static enum dc_status enable_link_dp(
 		max_link_rate = LINK_RATE_HIGH3;
 
 	if (link_settings.link_rate == max_link_rate) {
-		if (state->dis_clk->funcs->set_min_clocks_state) {
-			if (state->dis_clk->cur_min_clks_state < DM_PP_CLOCKS_STATE_NOMINAL)
-				state->dis_clk->funcs->set_min_clocks_state(
-					state->dis_clk, DM_PP_CLOCKS_STATE_NOMINAL);
-		} else {
-			uint32_t dp_phyclk_in_khz;
-			const struct clocks_value clocks_value =
-					state->dis_clk->cur_clocks_value;
-
-			/* 27mhz = 27000000hz= 27000khz */
-			dp_phyclk_in_khz = link_settings.link_rate * 27000;
-
-			if (((clocks_value.max_non_dp_phyclk_in_khz != 0) &&
-				(dp_phyclk_in_khz > clocks_value.max_non_dp_phyclk_in_khz)) ||
-				(dp_phyclk_in_khz > clocks_value.max_dp_phyclk_in_khz)) {
-				state->dis_clk->funcs->apply_clock_voltage_request(
-						state->dis_clk,
-						DM_PP_CLOCK_TYPE_DISPLAYPHYCLK,
-						dp_phyclk_in_khz,
-						false,
-						true);
-			}
-		}
+		struct dc_clocks clocks = state->bw.dcn.clk;
+
+		/* dce/dcn compat, do not update dispclk */
+		clocks.dispclk_khz = 0;
+		/* 27mhz = 27000000hz= 27000khz */
+		clocks.phyclk_khz = link_settings.link_rate * 27000;
+
+		state->dis_clk->funcs->update_clocks(
+				state->dis_clk, &clocks, false);
 	}
 
 	dp_enable_link_phy(
@@ -1794,28 +1856,6 @@ static enum dc_status enable_link(
 		break;
 	}
 
-	if (pipe_ctx->stream_res.audio && status == DC_OK) {
-		struct dc *core_dc = pipe_ctx->stream->ctx->dc;
-		/* notify audio driver for audio modes of monitor */
-		struct pp_smu_funcs_rv *pp_smu = core_dc->res_pool->pp_smu;
-		unsigned int i, num_audio = 1;
-		for (i = 0; i < MAX_PIPES; i++) {
-			/*current_state not updated yet*/
-			if (core_dc->current_state->res_ctx.pipe_ctx[i].stream_res.audio != NULL)
-				num_audio++;
-		}
-
-		pipe_ctx->stream_res.audio->funcs->az_enable(pipe_ctx->stream_res.audio);
-
-		if (num_audio == 1 && pp_smu != NULL && pp_smu->set_pme_wa_enable != NULL)
-			/*this is the first audio. apply the PME w/a in order to wake AZ from D3*/
-			pp_smu->set_pme_wa_enable(&pp_smu->pp_smu);
-		/* un-mute audio */
-		/* TODO: audio should be per stream rather than per link */
-		pipe_ctx->stream_res.stream_enc->funcs->audio_mute_control(
-			pipe_ctx->stream_res.stream_enc, false);
-	}
-
 	return status;
 }
 
@@ -1843,9 +1883,22 @@ static void disable_link(struct dc_link *link, enum signal_type signal)
 
 static bool dp_active_dongle_validate_timing(
 		const struct dc_crtc_timing *timing,
-		const struct dc_dongle_caps *dongle_caps)
+		const struct dpcd_caps *dpcd_caps)
 {
 	unsigned int required_pix_clk = timing->pix_clk_khz;
+	const struct dc_dongle_caps *dongle_caps = &dpcd_caps->dongle_caps;
+
+	switch (dpcd_caps->dongle_type) {
+	case DISPLAY_DONGLE_DP_VGA_CONVERTER:
+	case DISPLAY_DONGLE_DP_DVI_CONVERTER:
+	case DISPLAY_DONGLE_DP_DVI_DONGLE:
+		if (timing->pixel_encoding == PIXEL_ENCODING_RGB)
+			return true;
+		else
+			return false;
+	default:
+		break;
+	}
 
 	if (dongle_caps->dongle_type != DISPLAY_DONGLE_DP_HDMI_CONVERTER ||
 		dongle_caps->extendedCapValid == false)
@@ -1911,7 +1964,7 @@ enum dc_status dc_link_validate_mode_timing(
 		const struct dc_crtc_timing *timing)
 {
 	uint32_t max_pix_clk = stream->sink->dongle_max_pix_clk;
-	struct dc_dongle_caps *dongle_caps = &link->dpcd_caps.dongle_caps;
+	struct dpcd_caps *dpcd_caps = &link->dpcd_caps;
 
 	/* A hack to avoid failing any modes for EDID override feature on
 	 * topology change such as lower quality cable for DP or different dongle
@@ -1924,7 +1977,7 @@ enum dc_status dc_link_validate_mode_timing(
 		return DC_EXCEED_DONGLE_CAP;
 
 	/* Active Dongle*/
-	if (!dp_active_dongle_validate_timing(timing, dongle_caps))
+	if (!dp_active_dongle_validate_timing(timing, dpcd_caps))
 		return DC_EXCEED_DONGLE_CAP;
 
 	switch (stream->signal) {
@@ -1950,10 +2003,10 @@ bool dc_link_set_backlight_level(const struct dc_link *link, uint32_t level,
 	struct dc  *core_dc = link->ctx->dc;
 	struct abm *abm = core_dc->res_pool->abm;
 	struct dmcu *dmcu = core_dc->res_pool->dmcu;
-	struct dc_context *dc_ctx = link->ctx;
 	unsigned int controller_id = 0;
 	bool use_smooth_brightness = true;
 	int i;
+	DC_LOGGER_INIT(link->ctx->logger);
 
 	if ((dmcu == NULL) ||
 		(abm == NULL) ||
@@ -1961,7 +2014,7 @@ bool dc_link_set_backlight_level(const struct dc_link *link, uint32_t level,
 		return false;
 
 	if (stream) {
-		if (stream->bl_pwm_level == 0)
+		if (stream->bl_pwm_level == EDP_BACKLIGHT_RAMP_DISABLE_LEVEL)
 			frame_ramp = 0;
 
 		((struct dc_stream_state *)stream)->bl_pwm_level = level;
@@ -2038,10 +2091,10 @@ static struct fixed31_32 get_pbn_per_slot(struct dc_stream_state *stream)
 			&stream->sink->link->cur_link_settings;
 	uint32_t link_rate_in_mbps =
 			link_settings->link_rate * LINK_RATE_REF_FREQ_IN_MHZ;
-	struct fixed31_32 mbps = dal_fixed31_32_from_int(
+	struct fixed31_32 mbps = dc_fixpt_from_int(
 			link_rate_in_mbps * link_settings->lane_count);
 
-	return dal_fixed31_32_div_int(mbps, 54);
+	return dc_fixpt_div_int(mbps, 54);
 }
 
 static int get_color_depth(enum dc_color_depth color_depth)
@@ -2082,7 +2135,7 @@ static struct fixed31_32 get_pbn_from_timing(struct pipe_ctx *pipe_ctx)
 	numerator = 64 * PEAK_FACTOR_X1000;
 	denominator = 54 * 8 * 1000 * 1000;
 	kbps *= numerator;
-	peak_kbps = dal_fixed31_32_from_fraction(kbps, denominator);
+	peak_kbps = dc_fixpt_from_fraction(kbps, denominator);
 
 	return peak_kbps;
 }
@@ -2149,8 +2202,8 @@ static enum dc_status allocate_mst_payload(struct pipe_ctx *pipe_ctx)
 	struct fixed31_32 avg_time_slots_per_mtp;
 	struct fixed31_32 pbn;
 	struct fixed31_32 pbn_per_slot;
-	struct dc_context *dc_ctx = link->ctx;
 	uint8_t i;
+	DC_LOGGER_INIT(link->ctx->logger);
 
 	/* enable_link_dp_mst already check link->enabled_stream_count
 	 * and stream is in link->stream[]. This is called during set mode,
@@ -2178,11 +2231,11 @@ static enum dc_status allocate_mst_payload(struct pipe_ctx *pipe_ctx)
 			link->mst_stream_alloc_table.stream_count);
 
 	for (i = 0; i < MAX_CONTROLLER_NUM; i++) {
-		DC_LOG_MST("stream_enc[%d]: 0x%x      "
+		DC_LOG_MST("stream_enc[%d]: %p      "
 		"stream[%d].vcp_id: %d      "
 		"stream[%d].slot_count: %d\n",
 		i,
-		link->mst_stream_alloc_table.stream_allocations[i].stream_enc,
+		(void *) link->mst_stream_alloc_table.stream_allocations[i].stream_enc,
 		i,
 		link->mst_stream_alloc_table.stream_allocations[i].vcp_id,
 		i,
@@ -2209,7 +2262,7 @@ static enum dc_status allocate_mst_payload(struct pipe_ctx *pipe_ctx)
 	/* slot X.Y for only current stream */
 	pbn_per_slot = get_pbn_per_slot(stream);
 	pbn = get_pbn_from_timing(pipe_ctx);
-	avg_time_slots_per_mtp = dal_fixed31_32_div(pbn, pbn_per_slot);
+	avg_time_slots_per_mtp = dc_fixpt_div(pbn, pbn_per_slot);
 
 	stream_encoder->funcs->set_mst_bandwidth(
 		stream_encoder,
@@ -2226,10 +2279,10 @@ static enum dc_status deallocate_mst_payload(struct pipe_ctx *pipe_ctx)
 	struct link_encoder *link_encoder = link->link_enc;
 	struct stream_encoder *stream_encoder = pipe_ctx->stream_res.stream_enc;
 	struct dp_mst_stream_allocation_table proposed_table = {0};
-	struct fixed31_32 avg_time_slots_per_mtp = dal_fixed31_32_from_int(0);
+	struct fixed31_32 avg_time_slots_per_mtp = dc_fixpt_from_int(0);
 	uint8_t i;
 	bool mst_mode = (link->type == dc_connection_mst_branch);
-	struct dc_context *dc_ctx = link->ctx;
+	DC_LOGGER_INIT(link->ctx->logger);
 
 	/* deallocate_mst_payload is called before disable link. When mode or
 	 * disable/enable monitor, new stream is created which is not in link
@@ -2268,11 +2321,11 @@ static enum dc_status deallocate_mst_payload(struct pipe_ctx *pipe_ctx)
 			link->mst_stream_alloc_table.stream_count);
 
 	for (i = 0; i < MAX_CONTROLLER_NUM; i++) {
-		DC_LOG_MST("stream_enc[%d]: 0x%x      "
+		DC_LOG_MST("stream_enc[%d]: %p      "
 		"stream[%d].vcp_id: %d      "
 		"stream[%d].slot_count: %d\n",
 		i,
-		link->mst_stream_alloc_table.stream_allocations[i].stream_enc,
+		(void *) link->mst_stream_alloc_table.stream_allocations[i].stream_enc,
 		i,
 		link->mst_stream_alloc_table.stream_allocations[i].vcp_id,
 		i,
@@ -2302,8 +2355,8 @@ void core_link_enable_stream(
 		struct pipe_ctx *pipe_ctx)
 {
 	struct dc  *core_dc = pipe_ctx->stream->ctx->dc;
-	struct dc_context *dc_ctx = pipe_ctx->stream->ctx;
 	enum dc_status status;
+	DC_LOGGER_INIT(pipe_ctx->stream->ctx->logger);
 
 	/* eDP lit up by bios already, no need to enable again. */
 	if (pipe_ctx->stream->signal == SIGNAL_TYPE_EDP &&
@@ -2335,10 +2388,13 @@ void core_link_enable_stream(
 			}
 	}
 
+	core_dc->hwss.enable_audio_stream(pipe_ctx);
+
 	/* turn off otg test pattern if enable */
-	pipe_ctx->stream_res.tg->funcs->set_test_pattern(pipe_ctx->stream_res.tg,
-			CONTROLLER_DP_TEST_PATTERN_VIDEOMODE,
-			COLOR_DEPTH_UNDEFINED);
+	if (pipe_ctx->stream_res.tg->funcs->set_test_pattern)
+		pipe_ctx->stream_res.tg->funcs->set_test_pattern(pipe_ctx->stream_res.tg,
+				CONTROLLER_DP_TEST_PATTERN_VIDEOMODE,
+				COLOR_DEPTH_UNDEFINED);
 
 	core_dc->hwss.enable_stream(pipe_ctx);
 
@@ -2373,6 +2429,22 @@ void core_link_set_avmute(struct pipe_ctx *pipe_ctx, bool enable)
 	core_dc->hwss.set_avmute(pipe_ctx, enable);
 }
 
+/**
+ *****************************************************************************
+ *  Function: dc_link_enable_hpd_filter
+ *
+ *  @brief
+ *     If enable is true, programs HPD filter on associated HPD line using
+ *     delay_on_disconnect/delay_on_connect values dependent on
+ *     link->connector_signal
+ *
+ *     If enable is false, programs HPD filter on associated HPD line with no
+ *     delays on connect or disconnect
+ *
+ *  @param [in] link: pointer to the dc link
+ *  @param [in] enable: boolean specifying whether to enable hbd
+ *****************************************************************************
+ */
 void dc_link_enable_hpd_filter(struct dc_link *link, bool enable)
 {
 	struct gpio *hpd;
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
index 49c2face1e7a..08c9d73b9ab7 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_ddc.c
@@ -33,6 +33,10 @@
 #include "include/vector.h"
 #include "core_types.h"
 #include "dc_link_ddc.h"
+#include "i2caux/engine.h"
+#include "i2caux/i2c_engine.h"
+#include "i2caux/aux_engine.h"
+#include "i2caux/i2caux.h"
 
 #define AUX_POWER_UP_WA_DELAY 500
 #define I2C_OVER_AUX_DEFER_WA_DELAY 70
@@ -629,79 +633,62 @@ bool dal_ddc_service_query_ddc_data(
 	return ret;
 }
 
-ssize_t dal_ddc_service_read_dpcd_data(
-	struct ddc_service *ddc,
-	bool i2c,
-	enum i2c_mot_mode mot,
-	uint32_t address,
-	uint8_t *data,
-	uint32_t len)
+int dc_link_aux_transfer(struct ddc_service *ddc,
+			     unsigned int address,
+			     uint8_t *reply,
+			     void *buffer,
+			     unsigned int size,
+			     enum aux_transaction_type type,
+			     enum i2caux_transaction_action action)
 {
-	struct aux_payload read_payload = {
-		.i2c_over_aux = i2c,
-		.write = false,
-		.address = address,
-		.length = len,
-		.data = data,
-	};
-	struct aux_command command = {
-		.payloads = &read_payload,
-		.number_of_payloads = 1,
-		.defer_delay = 0,
-		.max_defer_write_retry = 0,
-		.mot = mot
-	};
-
-	if (len > DEFAULT_AUX_MAX_DATA_SIZE) {
-		BREAK_TO_DEBUGGER();
-		return DDC_RESULT_FAILED_INVALID_OPERATION;
-	}
+	struct i2caux *i2caux = ddc->ctx->i2caux;
+	struct ddc *ddc_pin = ddc->ddc_pin;
+	struct aux_engine *engine;
+	enum aux_channel_operation_result operation_result;
+	struct aux_request_transaction_data aux_req;
+	struct aux_reply_transaction_data aux_rep;
+	uint8_t returned_bytes = 0;
+	int res = -1;
+	uint32_t status;
 
-	if (dal_i2caux_submit_aux_command(
-		ddc->ctx->i2caux,
-		ddc->ddc_pin,
-		&command)) {
-		return (ssize_t)command.payloads->length;
-	}
+	memset(&aux_req, 0, sizeof(aux_req));
+	memset(&aux_rep, 0, sizeof(aux_rep));
 
-	return DDC_RESULT_FAILED_OPERATION;
-}
+	engine = i2caux->funcs->acquire_aux_engine(i2caux, ddc_pin);
 
-enum ddc_result dal_ddc_service_write_dpcd_data(
-	struct ddc_service *ddc,
-	bool i2c,
-	enum i2c_mot_mode mot,
-	uint32_t address,
-	const uint8_t *data,
-	uint32_t len)
-{
-	struct aux_payload write_payload = {
-		.i2c_over_aux = i2c,
-		.write = true,
-		.address = address,
-		.length = len,
-		.data = (uint8_t *)data,
-	};
-	struct aux_command command = {
-		.payloads = &write_payload,
-		.number_of_payloads = 1,
-		.defer_delay = 0,
-		.max_defer_write_retry = 0,
-		.mot = mot
-	};
-
-	if (len > DEFAULT_AUX_MAX_DATA_SIZE) {
-		BREAK_TO_DEBUGGER();
-		return DDC_RESULT_FAILED_INVALID_OPERATION;
-	}
+	aux_req.type = type;
+	aux_req.action = action;
+
+	aux_req.address = address;
+	aux_req.delay = 0;
+	aux_req.length = size;
+	aux_req.data = buffer;
 
-	if (dal_i2caux_submit_aux_command(
-		ddc->ctx->i2caux,
-		ddc->ddc_pin,
-		&command))
-		return DDC_RESULT_SUCESSFULL;
+	engine->funcs->submit_channel_request(engine, &aux_req);
+	operation_result = engine->funcs->get_channel_status(engine, &returned_bytes);
+
+	switch (operation_result) {
+	case AUX_CHANNEL_OPERATION_SUCCEEDED:
+		res = returned_bytes;
+
+		if (res <= size && res >= 0)
+			res = engine->funcs->read_channel_reply(engine, size,
+								buffer, reply,
+								&status);
+
+		break;
+	case AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON:
+		res = 0;
+		break;
+	case AUX_CHANNEL_OPERATION_FAILED_REASON_UNKNOWN:
+	case AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY:
+	case AUX_CHANNEL_OPERATION_FAILED_TIMEOUT:
+		res = -1;
+		break;
+	}
 
-	return DDC_RESULT_FAILED_OPERATION;
+	i2caux->funcs->release_engine(i2caux, &engine->base);
+	return res;
 }
 
 /*test only function*/
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
index 3b5053570229..474cd3e01752 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
@@ -3,6 +3,7 @@
 #include "dc.h"
 #include "dc_link_dp.h"
 #include "dm_helpers.h"
+#include "opp.h"
 
 #include "inc/core_types.h"
 #include "link_hwss.h"
@@ -38,7 +39,7 @@ static bool decide_fallback_link_setting(
 		struct dc_link_settings initial_link_settings,
 		struct dc_link_settings *current_link_setting,
 		enum link_training_result training_result);
-static struct dc_link_settings get_common_supported_link_settings (
+static struct dc_link_settings get_common_supported_link_settings(
 		struct dc_link_settings link_setting_a,
 		struct dc_link_settings link_setting_b);
 
@@ -93,8 +94,8 @@ static void dpcd_set_link_settings(
 	uint8_t rate = (uint8_t)
 	(lt_settings->link_settings.link_rate);
 
-	union down_spread_ctrl downspread = {{0}};
-	union lane_count_set lane_count_set = {{0}};
+	union down_spread_ctrl downspread = { {0} };
+	union lane_count_set lane_count_set = { {0} };
 	uint8_t link_set_buffer[2];
 
 	downspread.raw = (uint8_t)
@@ -164,11 +165,11 @@ static void dpcd_set_lt_pattern_and_lane_settings(
 	const struct link_training_settings *lt_settings,
 	enum hw_dp_training_pattern pattern)
 {
-	union dpcd_training_lane dpcd_lane[LANE_COUNT_DP_MAX] = {{{0}}};
+	union dpcd_training_lane dpcd_lane[LANE_COUNT_DP_MAX] = { { {0} } };
 	const uint32_t dpcd_base_lt_offset =
 	DP_TRAINING_PATTERN_SET;
 	uint8_t dpcd_lt_buffer[5] = {0};
-	union dpcd_training_pattern dpcd_pattern = {{0}};
+	union dpcd_training_pattern dpcd_pattern = { {0} };
 	uint32_t lane;
 	uint32_t size_in_bytes;
 	bool edp_workaround = false; /* TODO link_prop.INTERNAL */
@@ -232,7 +233,7 @@ static void dpcd_set_lt_pattern_and_lane_settings(
 			link,
 			DP_TRAINING_PATTERN_SET,
 			&dpcd_pattern.raw,
-			sizeof(dpcd_pattern.raw) );
+			sizeof(dpcd_pattern.raw));
 
 		core_link_write_dpcd(
 			link,
@@ -246,7 +247,7 @@ static void dpcd_set_lt_pattern_and_lane_settings(
 				link,
 				dpcd_base_lt_offset,
 				dpcd_lt_buffer,
-				size_in_bytes + sizeof(dpcd_pattern.raw) );
+				size_in_bytes + sizeof(dpcd_pattern.raw));
 
 	link->cur_lane_setting = lt_settings->lane_settings[0];
 }
@@ -428,8 +429,8 @@ static void get_lane_status_and_drive_settings(
 	struct link_training_settings *req_settings)
 {
 	uint8_t dpcd_buf[6] = {0};
-	union lane_adjust dpcd_lane_adjust[LANE_COUNT_DP_MAX] = {{{0}}};
-	struct link_training_settings request_settings = {{0}};
+	union lane_adjust dpcd_lane_adjust[LANE_COUNT_DP_MAX] = { { {0} } };
+	struct link_training_settings request_settings = { {0} };
 	uint32_t lane;
 
 	memset(req_settings, '\0', sizeof(struct link_training_settings));
@@ -651,7 +652,7 @@ static bool perform_post_lt_adj_req_sequence(
 
 			if (req_drv_setting_changed) {
 				update_drive_settings(
-					lt_settings,req_settings);
+					lt_settings, req_settings);
 
 				dc_link_dp_set_drive_settings(link,
 						lt_settings);
@@ -724,8 +725,8 @@ static enum link_training_result perform_channel_equalization_sequence(
 	enum hw_dp_training_pattern hw_tr_pattern;
 	uint32_t retries_ch_eq;
 	enum dc_lane_count lane_count = lt_settings->link_settings.lane_count;
-	union lane_align_status_updated dpcd_lane_status_updated = {{0}};
-	union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX] = {{{0}}};
+	union lane_align_status_updated dpcd_lane_status_updated = { {0} };
+	union lane_status dpcd_lane_status[LANE_COUNT_DP_MAX] = { { {0} } };
 
 	hw_tr_pattern = get_supported_tp(link);
 
@@ -1027,6 +1028,9 @@ enum link_training_result dc_link_dp_perform_link_training(
 			lt_settings.lane_settings[0].VOLTAGE_SWING,
 			lt_settings.lane_settings[0].PRE_EMPHASIS);
 
+	if (status != LINK_TRAINING_SUCCESS)
+		link->ctx->dc->debug.debug_data.ltFailCount++;
+
 	return status;
 }
 
@@ -1182,7 +1186,7 @@ bool dp_hbr_verify_link_cap(
 	return success;
 }
 
-static struct dc_link_settings get_common_supported_link_settings (
+static struct dc_link_settings get_common_supported_link_settings(
 		struct dc_link_settings link_setting_a,
 		struct dc_link_settings link_setting_b)
 {
@@ -1378,8 +1382,8 @@ static uint32_t bandwidth_in_kbps_from_timing(
 {
 	uint32_t bits_per_channel = 0;
 	uint32_t kbps;
-	switch (timing->display_color_depth) {
 
+	switch (timing->display_color_depth) {
 	case COLOR_DEPTH_666:
 		bits_per_channel = 6;
 		break;
@@ -1401,14 +1405,20 @@ static uint32_t bandwidth_in_kbps_from_timing(
 	default:
 		break;
 	}
+
 	ASSERT(bits_per_channel != 0);
 
 	kbps = timing->pix_clk_khz;
 	kbps *= bits_per_channel;
 
-	if (timing->flags.Y_ONLY != 1)
+	if (timing->flags.Y_ONLY != 1) {
 		/*Only YOnly make reduce bandwidth by 1/3 compares to RGB*/
 		kbps *= 3;
+		if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR420)
+			kbps /= 2;
+		else if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR422)
+			kbps = kbps * 2 / 3;
+	}
 
 	return kbps;
 
@@ -1422,6 +1432,7 @@ static uint32_t bandwidth_in_kbps_from_link_settings(
 
 	uint32_t lane_count  = link_setting->lane_count;
 	uint32_t kbps = link_rate_in_kbps;
+
 	kbps *= lane_count;
 	kbps *= 8;   /* 8 bits per byte*/
 
@@ -1439,9 +1450,9 @@ bool dp_validate_mode_timing(
 	const struct dc_link_settings *link_setting;
 
 	/*always DP fail safe mode*/
-	if (timing->pix_clk_khz == (uint32_t)25175 &&
-		timing->h_addressable == (uint32_t)640 &&
-		timing->v_addressable == (uint32_t)480)
+	if (timing->pix_clk_khz == (uint32_t) 25175 &&
+		timing->h_addressable == (uint32_t) 640 &&
+		timing->v_addressable == (uint32_t) 480)
 		return true;
 
 	/* We always use verified link settings */
@@ -1624,17 +1635,46 @@ static enum dc_status read_hpd_rx_irq_data(
 	struct dc_link *link,
 	union hpd_irq_data *irq_data)
 {
+	static enum dc_status retval;
+
 	/* The HW reads 16 bytes from 200h on HPD,
 	 * but if we get an AUX_DEFER, the HW cannot retry
 	 * and this causes the CTS tests 4.3.2.1 - 3.2.4 to
 	 * fail, so we now explicitly read 6 bytes which is
 	 * the req from the above mentioned test cases.
+	 *
+	 * For DP 1.4 we need to read those from 2002h range.
 	 */
-	return core_link_read_dpcd(
-	link,
-	DP_SINK_COUNT,
-	irq_data->raw,
-	sizeof(union hpd_irq_data));
+	if (link->dpcd_caps.dpcd_rev.raw < DPCD_REV_14)
+		retval = core_link_read_dpcd(
+			link,
+			DP_SINK_COUNT,
+			irq_data->raw,
+			sizeof(union hpd_irq_data));
+	else {
+		/* Read 14 bytes in a single read and then copy only the required fields.
+		 * This is more efficient than doing it in two separate AUX reads. */
+
+		uint8_t tmp[DP_SINK_STATUS_ESI - DP_SINK_COUNT_ESI + 1];
+
+		retval = core_link_read_dpcd(
+			link,
+			DP_SINK_COUNT_ESI,
+			tmp,
+			sizeof(tmp));
+
+		if (retval != DC_OK)
+			return retval;
+
+		irq_data->bytes.sink_cnt.raw = tmp[DP_SINK_COUNT_ESI - DP_SINK_COUNT_ESI];
+		irq_data->bytes.device_service_irq.raw = tmp[DP_DEVICE_SERVICE_IRQ_VECTOR_ESI0 - DP_SINK_COUNT_ESI];
+		irq_data->bytes.lane01_status.raw = tmp[DP_LANE0_1_STATUS_ESI - DP_SINK_COUNT_ESI];
+		irq_data->bytes.lane23_status.raw = tmp[DP_LANE2_3_STATUS_ESI - DP_SINK_COUNT_ESI];
+		irq_data->bytes.lane_status_updated.raw = tmp[DP_LANE_ALIGN_STATUS_UPDATED_ESI - DP_SINK_COUNT_ESI];
+		irq_data->bytes.sink_status.raw = tmp[DP_SINK_STATUS_ESI - DP_SINK_COUNT_ESI];
+	}
+
+	return retval;
 }
 
 static bool allow_hpd_rx_irq(const struct dc_link *link)
@@ -1960,12 +2000,16 @@ static void handle_automated_test(struct dc_link *link)
 			sizeof(test_response));
 }
 
-bool dc_link_handle_hpd_rx_irq(struct dc_link *link, union hpd_irq_data *out_hpd_irq_dpcd_data)
+bool dc_link_handle_hpd_rx_irq(struct dc_link *link, union hpd_irq_data *out_hpd_irq_dpcd_data, bool *out_link_loss)
 {
-	union hpd_irq_data hpd_irq_dpcd_data = {{{{0}}}};
+	union hpd_irq_data hpd_irq_dpcd_data = { { { {0} } } };
 	union device_service_irq device_service_clear = { { 0 } };
-	enum dc_status result = DDC_RESULT_UNKNOWN;
+	enum dc_status result;
+
 	bool status = false;
+
+	if (out_link_loss)
+		*out_link_loss = false;
 	/* For use cases related to down stream connection status change,
 	 * PSR and device auto test, refer to function handle_sst_hpd_irq
 	 * in DAL2.1*/
@@ -2040,6 +2084,8 @@ bool dc_link_handle_hpd_rx_irq(struct dc_link *link, union hpd_irq_data *out_hpd
 			true, LINK_TRAINING_ATTEMPTS);
 
 		status = false;
+		if (out_link_loss)
+			*out_link_loss = true;
 	}
 
 	if (link->type == dc_connection_active_dongle &&
@@ -2226,6 +2272,11 @@ static void get_active_converter_info(
 
 		link->dpcd_caps.branch_hw_revision =
 			dp_hw_fw_revision.ieee_hw_rev;
+
+		memmove(
+			link->dpcd_caps.branch_fw_revision,
+			dp_hw_fw_revision.ieee_fw_rev,
+			sizeof(dp_hw_fw_revision.ieee_fw_rev));
 	}
 }
 
@@ -2272,12 +2323,16 @@ static void dp_wa_power_up_0010FA(struct dc_link *link, uint8_t *dpcd_data,
 
 static bool retrieve_link_cap(struct dc_link *link)
 {
-	uint8_t dpcd_data[DP_TRAINING_AUX_RD_INTERVAL - DP_DPCD_REV + 1];
+	uint8_t dpcd_data[DP_ADAPTER_CAP - DP_DPCD_REV + 1];
 
+	struct dp_device_vendor_id sink_id;
 	union down_stream_port_count down_strm_port_count;
 	union edp_configuration_cap edp_config_cap;
 	union dp_downstream_port_present ds_port = { 0 };
 	enum dc_status status = DC_ERROR_UNEXPECTED;
+	uint32_t read_dpcd_retry_cnt = 3;
+	int i;
+	struct dp_sink_hw_fw_revision dp_hw_fw_revision;
 
 	memset(dpcd_data, '\0', sizeof(dpcd_data));
 	memset(&down_strm_port_count,
@@ -2285,11 +2340,15 @@ static bool retrieve_link_cap(struct dc_link *link)
 	memset(&edp_config_cap, '\0',
 		sizeof(union edp_configuration_cap));
 
-	status = core_link_read_dpcd(
-			link,
-			DP_DPCD_REV,
-			dpcd_data,
-			sizeof(dpcd_data));
+	for (i = 0; i < read_dpcd_retry_cnt; i++) {
+		status = core_link_read_dpcd(
+				link,
+				DP_DPCD_REV,
+				dpcd_data,
+				sizeof(dpcd_data));
+		if (status == DC_OK)
+			break;
+	}
 
 	if (status != DC_OK) {
 		dm_error("%s: Read dpcd data failed.\n", __func__);
@@ -2354,6 +2413,36 @@ static bool retrieve_link_cap(struct dc_link *link)
 			&link->dpcd_caps.sink_count.raw,
 			sizeof(link->dpcd_caps.sink_count.raw));
 
+	/* read sink ieee oui */
+	core_link_read_dpcd(link,
+			DP_SINK_OUI,
+			(uint8_t *)(&sink_id),
+			sizeof(sink_id));
+
+	link->dpcd_caps.sink_dev_id =
+			(sink_id.ieee_oui[0] << 16) +
+			(sink_id.ieee_oui[1] << 8) +
+			(sink_id.ieee_oui[2]);
+
+	memmove(
+		link->dpcd_caps.sink_dev_id_str,
+		sink_id.ieee_device_id,
+		sizeof(sink_id.ieee_device_id));
+
+	core_link_read_dpcd(
+		link,
+		DP_SINK_HW_REVISION_START,
+		(uint8_t *)&dp_hw_fw_revision,
+		sizeof(dp_hw_fw_revision));
+
+	link->dpcd_caps.sink_hw_revision =
+		dp_hw_fw_revision.ieee_hw_rev;
+
+	memmove(
+		link->dpcd_caps.sink_fw_revision,
+		dp_hw_fw_revision.ieee_fw_rev,
+		sizeof(dp_hw_fw_revision.ieee_fw_rev));
+
 	/* Connectivity log: detection */
 	CONN_DATA_DETECT(link, dpcd_data, sizeof(dpcd_data), "Rx Caps: ");
 
@@ -2376,6 +2465,10 @@ bool detect_dp_sink_caps(struct dc_link *link)
 void detect_edp_sink_caps(struct dc_link *link)
 {
 	retrieve_link_cap(link);
+
+	if (link->reported_link_cap.link_rate == LINK_RATE_UNKNOWN)
+		link->reported_link_cap.link_rate = LINK_RATE_HIGH2;
+
 	link->verified_link_cap = link->reported_link_cap;
 }
 
@@ -2454,8 +2547,8 @@ static void set_crtc_test_pattern(struct dc_link *link,
 		pipe_ctx->stream->bit_depth_params = params;
 		pipe_ctx->stream_res.opp->funcs->
 			opp_program_bit_depth_reduction(pipe_ctx->stream_res.opp, &params);
-
-		pipe_ctx->stream_res.tg->funcs->set_test_pattern(pipe_ctx->stream_res.tg,
+		if (pipe_ctx->stream_res.tg->funcs->set_test_pattern)
+			pipe_ctx->stream_res.tg->funcs->set_test_pattern(pipe_ctx->stream_res.tg,
 				controller_test_pattern, color_depth);
 	}
 	break;
@@ -2467,8 +2560,8 @@ static void set_crtc_test_pattern(struct dc_link *link,
 		pipe_ctx->stream->bit_depth_params = params;
 		pipe_ctx->stream_res.opp->funcs->
 			opp_program_bit_depth_reduction(pipe_ctx->stream_res.opp, &params);
-
-		pipe_ctx->stream_res.tg->funcs->set_test_pattern(pipe_ctx->stream_res.tg,
+		if (pipe_ctx->stream_res.tg->funcs->set_test_pattern)
+			pipe_ctx->stream_res.tg->funcs->set_test_pattern(pipe_ctx->stream_res.tg,
 				CONTROLLER_DP_TEST_PATTERN_VIDEOMODE,
 				color_depth);
 	}
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c
index 7c866a7d5e77..82cd1d6e6e59 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_hwss.c
@@ -11,8 +11,6 @@
 #include "dc_link_dp.h"
 #include "dc_link_ddc.h"
 #include "dm_helpers.h"
-#include "dce/dce_link_encoder.h"
-#include "dce/dce_stream_encoder.h"
 #include "dpcd_defs.h"
 
 enum dc_status core_link_read_dpcd(
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index ba3487e97361..2e65715f76a1 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -41,12 +41,13 @@
 #include "dce100/dce100_resource.h"
 #include "dce110/dce110_resource.h"
 #include "dce112/dce112_resource.h"
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 #include "dcn10/dcn10_resource.h"
 #endif
 #include "dce120/dce120_resource.h"
-#define DC_LOGGER \
-	ctx->logger
+
+#define DC_LOGGER_INIT(logger)
+
 enum dce_version resource_parse_asic_id(struct hw_asic_id asic_id)
 {
 	enum dce_version dc_version = DCE_VERSION_UNKNOWN;
@@ -78,11 +79,13 @@ enum dce_version resource_parse_asic_id(struct hw_asic_id asic_id)
 				ASIC_REV_IS_POLARIS12_V(asic_id.hw_internal_rev)) {
 			dc_version = DCE_VERSION_11_2;
 		}
+		if (ASIC_REV_IS_VEGAM(asic_id.hw_internal_rev))
+			dc_version = DCE_VERSION_11_22;
 		break;
 	case FAMILY_AI:
 		dc_version = DCE_VERSION_12_0;
 		break;
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	case FAMILY_RV:
 		dc_version = DCN_VERSION_1_0;
 		break;
@@ -124,6 +127,7 @@ struct resource_pool *dc_create_resource_pool(
 			num_virtual_links, dc, asic_id);
 		break;
 	case DCE_VERSION_11_2:
+	case DCE_VERSION_11_22:
 		res_pool = dce112_create_resource_pool(
 			num_virtual_links, dc);
 		break;
@@ -132,7 +136,7 @@ struct resource_pool *dc_create_resource_pool(
 			num_virtual_links, dc);
 		break;
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	case DCN_VERSION_1_0:
 		res_pool = dcn10_create_resource_pool(
 				num_virtual_links, dc);
@@ -492,9 +496,9 @@ static void calculate_viewport(struct pipe_ctx *pipe_ctx)
 	data->viewport_c.x = data->viewport.x / vpc_div;
 	data->viewport_c.y = data->viewport.y / vpc_div;
 	data->inits.h_c = (data->viewport.x % vpc_div) != 0 ?
-			dal_fixed31_32_half : dal_fixed31_32_zero;
+			dc_fixpt_half : dc_fixpt_zero;
 	data->inits.v_c = (data->viewport.y % vpc_div) != 0 ?
-			dal_fixed31_32_half : dal_fixed31_32_zero;
+			dc_fixpt_half : dc_fixpt_zero;
 	/* Round up, assume original video size always even dimensions */
 	data->viewport_c.width = (data->viewport.width + vpc_div - 1) / vpc_div;
 	data->viewport_c.height = (data->viewport.height + vpc_div - 1) / vpc_div;
@@ -518,13 +522,12 @@ static void calculate_viewport(struct pipe_ctx *pipe_ctx)
 	}
 }
 
-static void calculate_recout(struct pipe_ctx *pipe_ctx, struct view *recout_skip)
+static void calculate_recout(struct pipe_ctx *pipe_ctx, struct rect *recout_full)
 {
 	const struct dc_plane_state *plane_state = pipe_ctx->plane_state;
 	const struct dc_stream_state *stream = pipe_ctx->stream;
 	struct rect surf_src = plane_state->src_rect;
 	struct rect surf_clip = plane_state->clip_rect;
-	int recout_full_x, recout_full_y;
 	bool pri_split = pipe_ctx->bottom_pipe &&
 			pipe_ctx->bottom_pipe->plane_state == pipe_ctx->plane_state;
 	bool sec_split = pipe_ctx->top_pipe &&
@@ -593,20 +596,22 @@ static void calculate_recout(struct pipe_ctx *pipe_ctx, struct view *recout_skip
 		}
 	}
 	/* Unclipped recout offset = stream dst offset + ((surf dst offset - stream surf_src offset)
-	 * 				* 1/ stream scaling ratio) - (surf surf_src offset * 1/ full scl
-	 * 				ratio)
+	 *			* 1/ stream scaling ratio) - (surf surf_src offset * 1/ full scl
+	 *			ratio)
 	 */
-	recout_full_x = stream->dst.x + (plane_state->dst_rect.x - stream->src.x)
+	recout_full->x = stream->dst.x + (plane_state->dst_rect.x - stream->src.x)
 					* stream->dst.width / stream->src.width -
 			surf_src.x * plane_state->dst_rect.width / surf_src.width
 					* stream->dst.width / stream->src.width;
-	recout_full_y = stream->dst.y + (plane_state->dst_rect.y - stream->src.y)
+	recout_full->y = stream->dst.y + (plane_state->dst_rect.y - stream->src.y)
 					* stream->dst.height / stream->src.height -
 			surf_src.y * plane_state->dst_rect.height / surf_src.height
 					* stream->dst.height / stream->src.height;
 
-	recout_skip->width = pipe_ctx->plane_res.scl_data.recout.x - recout_full_x;
-	recout_skip->height = pipe_ctx->plane_res.scl_data.recout.y - recout_full_y;
+	recout_full->width = plane_state->dst_rect.width
+					* stream->dst.width / stream->src.width;
+	recout_full->height = plane_state->dst_rect.height
+					* stream->dst.height / stream->src.height;
 }
 
 static void calculate_scaling_ratios(struct pipe_ctx *pipe_ctx)
@@ -623,10 +628,10 @@ static void calculate_scaling_ratios(struct pipe_ctx *pipe_ctx)
 			pipe_ctx->plane_state->rotation == ROTATION_ANGLE_270)
 		rect_swap_helper(&surf_src);
 
-	pipe_ctx->plane_res.scl_data.ratios.horz = dal_fixed31_32_from_fraction(
+	pipe_ctx->plane_res.scl_data.ratios.horz = dc_fixpt_from_fraction(
 					surf_src.width,
 					plane_state->dst_rect.width);
-	pipe_ctx->plane_res.scl_data.ratios.vert = dal_fixed31_32_from_fraction(
+	pipe_ctx->plane_res.scl_data.ratios.vert = dc_fixpt_from_fraction(
 					surf_src.height,
 					plane_state->dst_rect.height);
 
@@ -648,9 +653,17 @@ static void calculate_scaling_ratios(struct pipe_ctx *pipe_ctx)
 		pipe_ctx->plane_res.scl_data.ratios.horz_c.value /= 2;
 		pipe_ctx->plane_res.scl_data.ratios.vert_c.value /= 2;
 	}
+	pipe_ctx->plane_res.scl_data.ratios.horz = dc_fixpt_truncate(
+			pipe_ctx->plane_res.scl_data.ratios.horz, 19);
+	pipe_ctx->plane_res.scl_data.ratios.vert = dc_fixpt_truncate(
+			pipe_ctx->plane_res.scl_data.ratios.vert, 19);
+	pipe_ctx->plane_res.scl_data.ratios.horz_c = dc_fixpt_truncate(
+			pipe_ctx->plane_res.scl_data.ratios.horz_c, 19);
+	pipe_ctx->plane_res.scl_data.ratios.vert_c = dc_fixpt_truncate(
+			pipe_ctx->plane_res.scl_data.ratios.vert_c, 19);
 }
 
-static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *recout_skip)
+static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct rect *recout_full)
 {
 	struct scaler_data *data = &pipe_ctx->plane_res.scl_data;
 	struct rect src = pipe_ctx->plane_state->src_rect;
@@ -668,15 +681,14 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 		flip_vert_scan_dir = true;
 	else if (pipe_ctx->plane_state->rotation == ROTATION_ANGLE_270)
 		flip_horz_scan_dir = true;
-	if (pipe_ctx->plane_state->horizontal_mirror)
-		flip_horz_scan_dir = !flip_horz_scan_dir;
 
 	if (pipe_ctx->plane_state->rotation == ROTATION_ANGLE_90 ||
 			pipe_ctx->plane_state->rotation == ROTATION_ANGLE_270) {
 		rect_swap_helper(&src);
 		rect_swap_helper(&data->viewport_c);
 		rect_swap_helper(&data->viewport);
-	}
+	} else if (pipe_ctx->plane_state->horizontal_mirror)
+			flip_horz_scan_dir = !flip_horz_scan_dir;
 
 	/*
 	 * Init calculated according to formula:
@@ -684,143 +696,303 @@ static void calculate_inits_and_adj_vp(struct pipe_ctx *pipe_ctx, struct view *r
 	 * 	init_bot = init + scaling_ratio
 	 * 	init_c = init + truncated_vp_c_offset(from calculate viewport)
 	 */
-	data->inits.h = dal_fixed31_32_div_int(
-			dal_fixed31_32_add_int(data->ratios.horz, data->taps.h_taps + 1), 2);
-
-	data->inits.h_c = dal_fixed31_32_add(data->inits.h_c, dal_fixed31_32_div_int(
-			dal_fixed31_32_add_int(data->ratios.horz_c, data->taps.h_taps_c + 1), 2));
-
-	data->inits.v = dal_fixed31_32_div_int(
-			dal_fixed31_32_add_int(data->ratios.vert, data->taps.v_taps + 1), 2);
-
-	data->inits.v_c = dal_fixed31_32_add(data->inits.v_c, dal_fixed31_32_div_int(
-			dal_fixed31_32_add_int(data->ratios.vert_c, data->taps.v_taps_c + 1), 2));
-
-
-	/* Adjust for viewport end clip-off */
-	if ((data->viewport.x + data->viewport.width) < (src.x + src.width) && !flip_horz_scan_dir) {
-		int vp_clip = src.x + src.width - data->viewport.width - data->viewport.x;
-		int int_part = dal_fixed31_32_floor(
-				dal_fixed31_32_sub(data->inits.h, data->ratios.horz));
-
-		int_part = int_part > 0 ? int_part : 0;
-		data->viewport.width += int_part < vp_clip ? int_part : vp_clip;
-	}
-	if ((data->viewport.y + data->viewport.height) < (src.y + src.height) && !flip_vert_scan_dir) {
-		int vp_clip = src.y + src.height - data->viewport.height - data->viewport.y;
-		int int_part = dal_fixed31_32_floor(
-				dal_fixed31_32_sub(data->inits.v, data->ratios.vert));
-
-		int_part = int_part > 0 ? int_part : 0;
-		data->viewport.height += int_part < vp_clip ? int_part : vp_clip;
-	}
-	if ((data->viewport_c.x + data->viewport_c.width) < (src.x + src.width) / vpc_div && !flip_horz_scan_dir) {
-		int vp_clip = (src.x + src.width) / vpc_div -
-				data->viewport_c.width - data->viewport_c.x;
-		int int_part = dal_fixed31_32_floor(
-				dal_fixed31_32_sub(data->inits.h_c, data->ratios.horz_c));
-
-		int_part = int_part > 0 ? int_part : 0;
-		data->viewport_c.width += int_part < vp_clip ? int_part : vp_clip;
-	}
-	if ((data->viewport_c.y + data->viewport_c.height) < (src.y + src.height) / vpc_div && !flip_vert_scan_dir) {
-		int vp_clip = (src.y + src.height) / vpc_div -
-				data->viewport_c.height - data->viewport_c.y;
-		int int_part = dal_fixed31_32_floor(
-				dal_fixed31_32_sub(data->inits.v_c, data->ratios.vert_c));
-
-		int_part = int_part > 0 ? int_part : 0;
-		data->viewport_c.height += int_part < vp_clip ? int_part : vp_clip;
-	}
-
-	/* Adjust for non-0 viewport offset */
-	if (data->viewport.x && !flip_horz_scan_dir) {
-		int int_part;
-
-		data->inits.h = dal_fixed31_32_add(data->inits.h, dal_fixed31_32_mul_int(
-				data->ratios.horz, recout_skip->width));
-		int_part = dal_fixed31_32_floor(data->inits.h) - data->viewport.x;
-		if (int_part < data->taps.h_taps) {
-			int int_adj = data->viewport.x >= (data->taps.h_taps - int_part) ?
-						(data->taps.h_taps - int_part) : data->viewport.x;
-			data->viewport.x -= int_adj;
-			data->viewport.width += int_adj;
-			int_part += int_adj;
-		} else if (int_part > data->taps.h_taps) {
-			data->viewport.x += int_part - data->taps.h_taps;
-			data->viewport.width -= int_part - data->taps.h_taps;
-			int_part = data->taps.h_taps;
+	data->inits.h = dc_fixpt_truncate(dc_fixpt_div_int(
+			dc_fixpt_add_int(data->ratios.horz, data->taps.h_taps + 1), 2), 19);
+
+	data->inits.h_c = dc_fixpt_truncate(dc_fixpt_add(data->inits.h_c, dc_fixpt_div_int(
+			dc_fixpt_add_int(data->ratios.horz_c, data->taps.h_taps_c + 1), 2)), 19);
+
+	data->inits.v = dc_fixpt_truncate(dc_fixpt_div_int(
+			dc_fixpt_add_int(data->ratios.vert, data->taps.v_taps + 1), 2), 19);
+
+	data->inits.v_c = dc_fixpt_truncate(dc_fixpt_add(data->inits.v_c, dc_fixpt_div_int(
+			dc_fixpt_add_int(data->ratios.vert_c, data->taps.v_taps_c + 1), 2)), 19);
+
+	if (!flip_horz_scan_dir) {
+		/* Adjust for viewport end clip-off */
+		if ((data->viewport.x + data->viewport.width) < (src.x + src.width)) {
+			int vp_clip = src.x + src.width - data->viewport.width - data->viewport.x;
+			int int_part = dc_fixpt_floor(
+					dc_fixpt_sub(data->inits.h, data->ratios.horz));
+
+			int_part = int_part > 0 ? int_part : 0;
+			data->viewport.width += int_part < vp_clip ? int_part : vp_clip;
 		}
-		data->inits.h.value &= 0xffffffff;
-		data->inits.h = dal_fixed31_32_add_int(data->inits.h, int_part);
-	}
-
-	if (data->viewport_c.x && !flip_horz_scan_dir) {
-		int int_part;
-
-		data->inits.h_c = dal_fixed31_32_add(data->inits.h_c, dal_fixed31_32_mul_int(
-				data->ratios.horz_c, recout_skip->width));
-		int_part = dal_fixed31_32_floor(data->inits.h_c) - data->viewport_c.x;
-		if (int_part < data->taps.h_taps_c) {
-			int int_adj = data->viewport_c.x >= (data->taps.h_taps_c - int_part) ?
-					(data->taps.h_taps_c - int_part) : data->viewport_c.x;
-			data->viewport_c.x -= int_adj;
-			data->viewport_c.width += int_adj;
-			int_part += int_adj;
-		} else if (int_part > data->taps.h_taps_c) {
-			data->viewport_c.x += int_part - data->taps.h_taps_c;
-			data->viewport_c.width -= int_part - data->taps.h_taps_c;
-			int_part = data->taps.h_taps_c;
+		if ((data->viewport_c.x + data->viewport_c.width) < (src.x + src.width) / vpc_div) {
+			int vp_clip = (src.x + src.width) / vpc_div -
+					data->viewport_c.width - data->viewport_c.x;
+			int int_part = dc_fixpt_floor(
+					dc_fixpt_sub(data->inits.h_c, data->ratios.horz_c));
+
+			int_part = int_part > 0 ? int_part : 0;
+			data->viewport_c.width += int_part < vp_clip ? int_part : vp_clip;
+		}
+
+		/* Adjust for non-0 viewport offset */
+		if (data->viewport.x) {
+			int int_part;
+
+			data->inits.h = dc_fixpt_add(data->inits.h, dc_fixpt_mul_int(
+					data->ratios.horz, data->recout.x - recout_full->x));
+			int_part = dc_fixpt_floor(data->inits.h) - data->viewport.x;
+			if (int_part < data->taps.h_taps) {
+				int int_adj = data->viewport.x >= (data->taps.h_taps - int_part) ?
+							(data->taps.h_taps - int_part) : data->viewport.x;
+				data->viewport.x -= int_adj;
+				data->viewport.width += int_adj;
+				int_part += int_adj;
+			} else if (int_part > data->taps.h_taps) {
+				data->viewport.x += int_part - data->taps.h_taps;
+				data->viewport.width -= int_part - data->taps.h_taps;
+				int_part = data->taps.h_taps;
+			}
+			data->inits.h.value &= 0xffffffff;
+			data->inits.h = dc_fixpt_add_int(data->inits.h, int_part);
 		}
-		data->inits.h_c.value &= 0xffffffff;
-		data->inits.h_c = dal_fixed31_32_add_int(data->inits.h_c, int_part);
-	}
-
-	if (data->viewport.y && !flip_vert_scan_dir) {
-		int int_part;
-
-		data->inits.v = dal_fixed31_32_add(data->inits.v, dal_fixed31_32_mul_int(
-				data->ratios.vert, recout_skip->height));
-		int_part = dal_fixed31_32_floor(data->inits.v) - data->viewport.y;
-		if (int_part < data->taps.v_taps) {
-			int int_adj = data->viewport.y >= (data->taps.v_taps - int_part) ?
-						(data->taps.v_taps - int_part) : data->viewport.y;
-			data->viewport.y -= int_adj;
-			data->viewport.height += int_adj;
-			int_part += int_adj;
-		} else if (int_part > data->taps.v_taps) {
-			data->viewport.y += int_part - data->taps.v_taps;
-			data->viewport.height -= int_part - data->taps.v_taps;
-			int_part = data->taps.v_taps;
+
+		if (data->viewport_c.x) {
+			int int_part;
+
+			data->inits.h_c = dc_fixpt_add(data->inits.h_c, dc_fixpt_mul_int(
+					data->ratios.horz_c, data->recout.x - recout_full->x));
+			int_part = dc_fixpt_floor(data->inits.h_c) - data->viewport_c.x;
+			if (int_part < data->taps.h_taps_c) {
+				int int_adj = data->viewport_c.x >= (data->taps.h_taps_c - int_part) ?
+						(data->taps.h_taps_c - int_part) : data->viewport_c.x;
+				data->viewport_c.x -= int_adj;
+				data->viewport_c.width += int_adj;
+				int_part += int_adj;
+			} else if (int_part > data->taps.h_taps_c) {
+				data->viewport_c.x += int_part - data->taps.h_taps_c;
+				data->viewport_c.width -= int_part - data->taps.h_taps_c;
+				int_part = data->taps.h_taps_c;
+			}
+			data->inits.h_c.value &= 0xffffffff;
+			data->inits.h_c = dc_fixpt_add_int(data->inits.h_c, int_part);
 		}
-		data->inits.v.value &= 0xffffffff;
-		data->inits.v = dal_fixed31_32_add_int(data->inits.v, int_part);
-	}
-
-	if (data->viewport_c.y && !flip_vert_scan_dir) {
-		int int_part;
-
-		data->inits.v_c = dal_fixed31_32_add(data->inits.v_c, dal_fixed31_32_mul_int(
-				data->ratios.vert_c, recout_skip->height));
-		int_part = dal_fixed31_32_floor(data->inits.v_c) - data->viewport_c.y;
-		if (int_part < data->taps.v_taps_c) {
-			int int_adj = data->viewport_c.y >= (data->taps.v_taps_c - int_part) ?
-					(data->taps.v_taps_c - int_part) : data->viewport_c.y;
-			data->viewport_c.y -= int_adj;
-			data->viewport_c.height += int_adj;
-			int_part += int_adj;
-		} else if (int_part > data->taps.v_taps_c) {
-			data->viewport_c.y += int_part - data->taps.v_taps_c;
-			data->viewport_c.height -= int_part - data->taps.v_taps_c;
-			int_part = data->taps.v_taps_c;
+	} else {
+		/* Adjust for non-0 viewport offset */
+		if (data->viewport.x) {
+			int int_part = dc_fixpt_floor(
+					dc_fixpt_sub(data->inits.h, data->ratios.horz));
+
+			int_part = int_part > 0 ? int_part : 0;
+			data->viewport.width += int_part < data->viewport.x ? int_part : data->viewport.x;
+			data->viewport.x -= int_part < data->viewport.x ? int_part : data->viewport.x;
+		}
+		if (data->viewport_c.x) {
+			int int_part = dc_fixpt_floor(
+					dc_fixpt_sub(data->inits.h_c, data->ratios.horz_c));
+
+			int_part = int_part > 0 ? int_part : 0;
+			data->viewport_c.width += int_part < data->viewport_c.x ? int_part : data->viewport_c.x;
+			data->viewport_c.x -= int_part < data->viewport_c.x ? int_part : data->viewport_c.x;
+		}
+
+		/* Adjust for viewport end clip-off */
+		if ((data->viewport.x + data->viewport.width) < (src.x + src.width)) {
+			int int_part;
+			int end_offset = src.x + src.width
+					- data->viewport.x - data->viewport.width;
+
+			/*
+			 * this is init if vp had no offset, keep in mind this is from the
+			 * right side of vp due to scan direction
+			 */
+			data->inits.h = dc_fixpt_add(data->inits.h, dc_fixpt_mul_int(
+					data->ratios.horz, data->recout.x - recout_full->x));
+			/*
+			 * this is the difference between first pixel of viewport available to read
+			 * and init position, takning into account scan direction
+			 */
+			int_part = dc_fixpt_floor(data->inits.h) - end_offset;
+			if (int_part < data->taps.h_taps) {
+				int int_adj = end_offset >= (data->taps.h_taps - int_part) ?
+							(data->taps.h_taps - int_part) : end_offset;
+				data->viewport.width += int_adj;
+				int_part += int_adj;
+			} else if (int_part > data->taps.h_taps) {
+				data->viewport.width += int_part - data->taps.h_taps;
+				int_part = data->taps.h_taps;
+			}
+			data->inits.h.value &= 0xffffffff;
+			data->inits.h = dc_fixpt_add_int(data->inits.h, int_part);
+		}
+
+		if ((data->viewport_c.x + data->viewport_c.width) < (src.x + src.width) / vpc_div) {
+			int int_part;
+			int end_offset = (src.x + src.width) / vpc_div
+					- data->viewport_c.x - data->viewport_c.width;
+
+			/*
+			 * this is init if vp had no offset, keep in mind this is from the
+			 * right side of vp due to scan direction
+			 */
+			data->inits.h_c = dc_fixpt_add(data->inits.h_c, dc_fixpt_mul_int(
+					data->ratios.horz_c, data->recout.x - recout_full->x));
+			/*
+			 * this is the difference between first pixel of viewport available to read
+			 * and init position, takning into account scan direction
+			 */
+			int_part = dc_fixpt_floor(data->inits.h_c) - end_offset;
+			if (int_part < data->taps.h_taps_c) {
+				int int_adj = end_offset >= (data->taps.h_taps_c - int_part) ?
+							(data->taps.h_taps_c - int_part) : end_offset;
+				data->viewport_c.width += int_adj;
+				int_part += int_adj;
+			} else if (int_part > data->taps.h_taps_c) {
+				data->viewport_c.width += int_part - data->taps.h_taps_c;
+				int_part = data->taps.h_taps_c;
+			}
+			data->inits.h_c.value &= 0xffffffff;
+			data->inits.h_c = dc_fixpt_add_int(data->inits.h_c, int_part);
+		}
+
+	}
+	if (!flip_vert_scan_dir) {
+		/* Adjust for viewport end clip-off */
+		if ((data->viewport.y + data->viewport.height) < (src.y + src.height)) {
+			int vp_clip = src.y + src.height - data->viewport.height - data->viewport.y;
+			int int_part = dc_fixpt_floor(
+					dc_fixpt_sub(data->inits.v, data->ratios.vert));
+
+			int_part = int_part > 0 ? int_part : 0;
+			data->viewport.height += int_part < vp_clip ? int_part : vp_clip;
+		}
+		if ((data->viewport_c.y + data->viewport_c.height) < (src.y + src.height) / vpc_div) {
+			int vp_clip = (src.y + src.height) / vpc_div -
+					data->viewport_c.height - data->viewport_c.y;
+			int int_part = dc_fixpt_floor(
+					dc_fixpt_sub(data->inits.v_c, data->ratios.vert_c));
+
+			int_part = int_part > 0 ? int_part : 0;
+			data->viewport_c.height += int_part < vp_clip ? int_part : vp_clip;
+		}
+
+		/* Adjust for non-0 viewport offset */
+		if (data->viewport.y) {
+			int int_part;
+
+			data->inits.v = dc_fixpt_add(data->inits.v, dc_fixpt_mul_int(
+					data->ratios.vert, data->recout.y - recout_full->y));
+			int_part = dc_fixpt_floor(data->inits.v) - data->viewport.y;
+			if (int_part < data->taps.v_taps) {
+				int int_adj = data->viewport.y >= (data->taps.v_taps - int_part) ?
+							(data->taps.v_taps - int_part) : data->viewport.y;
+				data->viewport.y -= int_adj;
+				data->viewport.height += int_adj;
+				int_part += int_adj;
+			} else if (int_part > data->taps.v_taps) {
+				data->viewport.y += int_part - data->taps.v_taps;
+				data->viewport.height -= int_part - data->taps.v_taps;
+				int_part = data->taps.v_taps;
+			}
+			data->inits.v.value &= 0xffffffff;
+			data->inits.v = dc_fixpt_add_int(data->inits.v, int_part);
+		}
+
+		if (data->viewport_c.y) {
+			int int_part;
+
+			data->inits.v_c = dc_fixpt_add(data->inits.v_c, dc_fixpt_mul_int(
+					data->ratios.vert_c, data->recout.y - recout_full->y));
+			int_part = dc_fixpt_floor(data->inits.v_c) - data->viewport_c.y;
+			if (int_part < data->taps.v_taps_c) {
+				int int_adj = data->viewport_c.y >= (data->taps.v_taps_c - int_part) ?
+						(data->taps.v_taps_c - int_part) : data->viewport_c.y;
+				data->viewport_c.y -= int_adj;
+				data->viewport_c.height += int_adj;
+				int_part += int_adj;
+			} else if (int_part > data->taps.v_taps_c) {
+				data->viewport_c.y += int_part - data->taps.v_taps_c;
+				data->viewport_c.height -= int_part - data->taps.v_taps_c;
+				int_part = data->taps.v_taps_c;
+			}
+			data->inits.v_c.value &= 0xffffffff;
+			data->inits.v_c = dc_fixpt_add_int(data->inits.v_c, int_part);
+		}
+	} else {
+		/* Adjust for non-0 viewport offset */
+		if (data->viewport.y) {
+			int int_part = dc_fixpt_floor(
+					dc_fixpt_sub(data->inits.v, data->ratios.vert));
+
+			int_part = int_part > 0 ? int_part : 0;
+			data->viewport.height += int_part < data->viewport.y ? int_part : data->viewport.y;
+			data->viewport.y -= int_part < data->viewport.y ? int_part : data->viewport.y;
+		}
+		if (data->viewport_c.y) {
+			int int_part = dc_fixpt_floor(
+					dc_fixpt_sub(data->inits.v_c, data->ratios.vert_c));
+
+			int_part = int_part > 0 ? int_part : 0;
+			data->viewport_c.height += int_part < data->viewport_c.y ? int_part : data->viewport_c.y;
+			data->viewport_c.y -= int_part < data->viewport_c.y ? int_part : data->viewport_c.y;
+		}
+
+		/* Adjust for viewport end clip-off */
+		if ((data->viewport.y + data->viewport.height) < (src.y + src.height)) {
+			int int_part;
+			int end_offset = src.y + src.height
+					- data->viewport.y - data->viewport.height;
+
+			/*
+			 * this is init if vp had no offset, keep in mind this is from the
+			 * right side of vp due to scan direction
+			 */
+			data->inits.v = dc_fixpt_add(data->inits.v, dc_fixpt_mul_int(
+					data->ratios.vert, data->recout.y - recout_full->y));
+			/*
+			 * this is the difference between first pixel of viewport available to read
+			 * and init position, taking into account scan direction
+			 */
+			int_part = dc_fixpt_floor(data->inits.v) - end_offset;
+			if (int_part < data->taps.v_taps) {
+				int int_adj = end_offset >= (data->taps.v_taps - int_part) ?
+							(data->taps.v_taps - int_part) : end_offset;
+				data->viewport.height += int_adj;
+				int_part += int_adj;
+			} else if (int_part > data->taps.v_taps) {
+				data->viewport.height += int_part - data->taps.v_taps;
+				int_part = data->taps.v_taps;
+			}
+			data->inits.v.value &= 0xffffffff;
+			data->inits.v = dc_fixpt_add_int(data->inits.v, int_part);
+		}
+
+		if ((data->viewport_c.y + data->viewport_c.height) < (src.y + src.height) / vpc_div) {
+			int int_part;
+			int end_offset = (src.y + src.height) / vpc_div
+					- data->viewport_c.y - data->viewport_c.height;
+
+			/*
+			 * this is init if vp had no offset, keep in mind this is from the
+			 * right side of vp due to scan direction
+			 */
+			data->inits.v_c = dc_fixpt_add(data->inits.v_c, dc_fixpt_mul_int(
+					data->ratios.vert_c, data->recout.y - recout_full->y));
+			/*
+			 * this is the difference between first pixel of viewport available to read
+			 * and init position, taking into account scan direction
+			 */
+			int_part = dc_fixpt_floor(data->inits.v_c) - end_offset;
+			if (int_part < data->taps.v_taps_c) {
+				int int_adj = end_offset >= (data->taps.v_taps_c - int_part) ?
+							(data->taps.v_taps_c - int_part) : end_offset;
+				data->viewport_c.height += int_adj;
+				int_part += int_adj;
+			} else if (int_part > data->taps.v_taps_c) {
+				data->viewport_c.height += int_part - data->taps.v_taps_c;
+				int_part = data->taps.v_taps_c;
+			}
+			data->inits.v_c.value &= 0xffffffff;
+			data->inits.v_c = dc_fixpt_add_int(data->inits.v_c, int_part);
 		}
-		data->inits.v_c.value &= 0xffffffff;
-		data->inits.v_c = dal_fixed31_32_add_int(data->inits.v_c, int_part);
 	}
 
 	/* Interlaced inits based on final vert inits */
-	data->inits.v_bot = dal_fixed31_32_add(data->inits.v, data->ratios.vert);
-	data->inits.v_c_bot = dal_fixed31_32_add(data->inits.v_c, data->ratios.vert_c);
+	data->inits.v_bot = dc_fixpt_add(data->inits.v, data->ratios.vert);
+	data->inits.v_c_bot = dc_fixpt_add(data->inits.v_c, data->ratios.vert_c);
 
 	if (pipe_ctx->plane_state->rotation == ROTATION_ANGLE_90 ||
 			pipe_ctx->plane_state->rotation == ROTATION_ANGLE_270) {
@@ -833,9 +1005,9 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 {
 	const struct dc_plane_state *plane_state = pipe_ctx->plane_state;
 	struct dc_crtc_timing *timing = &pipe_ctx->stream->timing;
-	struct view recout_skip = { 0 };
+	struct rect recout_full = { 0 };
 	bool res = false;
-	struct dc_context *ctx = pipe_ctx->stream->ctx;
+	DC_LOGGER_INIT(pipe_ctx->stream->ctx->logger);
 	/* Important: scaling ratio calculation requires pixel format,
 	 * lb depth calculation requires recout and taps require scaling ratios.
 	 * Inits require viewport, taps, ratios and recout of split pipe
@@ -843,6 +1015,9 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 	pipe_ctx->plane_res.scl_data.format = convert_pixel_format_to_dalsurface(
 			pipe_ctx->plane_state->format);
 
+	if (pipe_ctx->stream->timing.flags.INTERLACE)
+		pipe_ctx->stream->dst.height *= 2;
+
 	calculate_scaling_ratios(pipe_ctx);
 
 	calculate_viewport(pipe_ctx);
@@ -850,7 +1025,7 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 	if (pipe_ctx->plane_res.scl_data.viewport.height < 16 || pipe_ctx->plane_res.scl_data.viewport.width < 16)
 		return false;
 
-	calculate_recout(pipe_ctx, &recout_skip);
+	calculate_recout(pipe_ctx, &recout_full);
 
 	/**
 	 * Setting line buffer pixel depth to 24bpp yields banding
@@ -863,6 +1038,8 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 
 	pipe_ctx->plane_res.scl_data.h_active = timing->h_addressable + timing->h_border_left + timing->h_border_right;
 	pipe_ctx->plane_res.scl_data.v_active = timing->v_addressable + timing->v_border_top + timing->v_border_bottom;
+	if (pipe_ctx->stream->timing.flags.INTERLACE)
+		pipe_ctx->plane_res.scl_data.v_active *= 2;
 
 
 	/* Taps calculations */
@@ -892,7 +1069,7 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 
 	if (res)
 		/* May need to re-check lb size after this in some obscure scenario */
-		calculate_inits_and_adj_vp(pipe_ctx, &recout_skip);
+		calculate_inits_and_adj_vp(pipe_ctx, &recout_full);
 
 	DC_LOG_SCALER(
 				"%s: Viewport:\nheight:%d width:%d x:%d "
@@ -908,6 +1085,9 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
 				plane_state->dst_rect.x,
 				plane_state->dst_rect.y);
 
+	if (pipe_ctx->stream->timing.flags.INTERLACE)
+		pipe_ctx->stream->dst.height /= 2;
+
 	return res;
 }
 
@@ -1033,7 +1213,7 @@ static struct pipe_ctx *acquire_free_pipe_for_stream(
 
 }
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 static int acquire_first_split_pipe(
 		struct resource_context *res_ctx,
 		const struct resource_pool *pool,
@@ -1104,7 +1284,7 @@ bool dc_add_plane_to_context(
 
 	free_pipe = acquire_free_pipe_for_stream(context, pool, stream);
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	if (!free_pipe) {
 		int pipe_idx = acquire_first_split_pipe(&context->res_ctx, pool, stream);
 		if (pipe_idx >= 0)
@@ -1294,6 +1474,19 @@ bool dc_add_all_planes_for_stream(
 }
 
 
+static bool is_hdr_static_meta_changed(struct dc_stream_state *cur_stream,
+	struct dc_stream_state *new_stream)
+{
+	if (cur_stream == NULL)
+		return true;
+
+	if (memcmp(&cur_stream->hdr_static_metadata,
+			&new_stream->hdr_static_metadata,
+			sizeof(struct dc_info_packet)) != 0)
+		return true;
+
+	return false;
+}
 
 static bool is_timing_changed(struct dc_stream_state *cur_stream,
 		struct dc_stream_state *new_stream)
@@ -1329,6 +1522,9 @@ static bool are_stream_backends_same(
 	if (is_timing_changed(stream_a, stream_b))
 		return false;
 
+	if (is_hdr_static_meta_changed(stream_a, stream_b))
+		return false;
+
 	return true;
 }
 
@@ -1509,8 +1705,8 @@ enum dc_status dc_add_stream_to_ctx(
 	struct dc_context *dc_ctx = dc->ctx;
 	enum dc_status res;
 
-	if (new_ctx->stream_count >= dc->res_pool->pipe_count) {
-		DC_ERROR("Max streams reached, can add stream %p !\n", stream);
+	if (new_ctx->stream_count >= dc->res_pool->timing_generator_count) {
+		DC_ERROR("Max streams reached, can't add stream %p !\n", stream);
 		return DC_ERROR_UNEXPECTED;
 	}
 
@@ -1599,18 +1795,6 @@ enum dc_status dc_remove_stream_from_ctx(
 	return DC_OK;
 }
 
-static void copy_pipe_ctx(
-	const struct pipe_ctx *from_pipe_ctx, struct pipe_ctx *to_pipe_ctx)
-{
-	struct dc_plane_state *plane_state = to_pipe_ctx->plane_state;
-	struct dc_stream_state *stream = to_pipe_ctx->stream;
-
-	*to_pipe_ctx = *from_pipe_ctx;
-	to_pipe_ctx->stream = stream;
-	if (plane_state != NULL)
-		to_pipe_ctx->plane_state = plane_state;
-}
-
 static struct dc_stream_state *find_pll_sharable_stream(
 		struct dc_stream_state *stream_needs_pll,
 		struct dc_state *context)
@@ -1698,12 +1882,12 @@ enum dc_status resource_map_pool_resources(
 	/* acquire new resources */
 	pipe_idx = acquire_first_free_pipe(&context->res_ctx, pool, stream);
 
-#ifdef CONFIG_DRM_AMD_DC_DCN1_0
+#ifdef CONFIG_X86
 	if (pipe_idx < 0)
 		pipe_idx = acquire_first_split_pipe(&context->res_ctx, pool, stream);
 #endif
 
-	if (pipe_idx < 0)
+	if (pipe_idx < 0 || context->res_ctx.pipe_ctx[pipe_idx].stream_res.tg == NULL)
 		return DC_NO_CONTROLLER_RESOURCE;
 
 	pipe_ctx = &context->res_ctx.pipe_ctx[pipe_idx];
@@ -1752,26 +1936,6 @@ enum dc_status resource_map_pool_resources(
 	return DC_ERROR_UNEXPECTED;
 }
 
-/* first stream in the context is used to populate the rest */
-void validate_guaranteed_copy_streams(
-		struct dc_state *context,
-		int max_streams)
-{
-	int i;
-
-	for (i = 1; i < max_streams; i++) {
-		context->streams[i] = context->streams[0];
-
-		copy_pipe_ctx(&context->res_ctx.pipe_ctx[0],
-			      &context->res_ctx.pipe_ctx[i]);
-		context->res_ctx.pipe_ctx[i].stream =
-				context->res_ctx.pipe_ctx[0].stream;
-
-		dc_stream_retain(context->streams[i]);
-		context->stream_count++;
-	}
-}
-
 void dc_resource_state_copy_construct_current(
 		const struct dc *dc,
 		struct dc_state *dst_ctx)
@@ -1784,7 +1948,7 @@ void dc_resource_state_construct(
 		const struct dc *dc,
 		struct dc_state *dst_ctx)
 {
-	dst_ctx->dis_clk = dc->res_pool->display_clock;
+	dst_ctx->dis_clk = dc->res_pool->dccg;
 }
 
 enum dc_status dc_validate_global_state(
@@ -1798,9 +1962,9 @@ enum dc_status dc_validate_global_state(
 		return DC_ERROR_UNEXPECTED;
 
 	if (dc->res_pool->funcs->validate_global) {
-			result = dc->res_pool->funcs->validate_global(dc, new_ctx);
-			if (result != DC_OK)
-				return result;
+		result = dc->res_pool->funcs->validate_global(dc, new_ctx);
+		if (result != DC_OK)
+			return result;
 	}
 
 	for (i = 0; i < new_ctx->stream_count; i++) {
@@ -1843,7 +2007,7 @@ enum dc_status dc_validate_global_state(
 }
 
 static void patch_gamut_packet_checksum(
-		struct encoder_info_packet *gamut_packet)
+		struct dc_info_packet *gamut_packet)
 {
 	/* For gamut we recalc checksum */
 	if (gamut_packet->valid) {
@@ -1862,12 +2026,11 @@ static void patch_gamut_packet_checksum(
 }
 
 static void set_avi_info_frame(
-		struct encoder_info_packet *info_packet,
+		struct dc_info_packet *info_packet,
 		struct pipe_ctx *pipe_ctx)
 {
 	struct dc_stream_state *stream = pipe_ctx->stream;
 	enum dc_color_space color_space = COLOR_SPACE_UNKNOWN;
-	struct info_frame info_frame = { {0} };
 	uint32_t pixel_encoding = 0;
 	enum scanning_type scan_type = SCANNING_TYPE_NODATA;
 	enum dc_aspect_ratio aspect = ASPECT_RATIO_NO_DATA;
@@ -1877,22 +2040,24 @@ static void set_avi_info_frame(
 	unsigned int cn0_cn1_value = 0;
 	uint8_t *check_sum = NULL;
 	uint8_t byte_index = 0;
-	union hdmi_info_packet *hdmi_info = &info_frame.avi_info_packet.info_packet_hdmi;
+	union hdmi_info_packet hdmi_info;
 	union display_content_support support = {0};
 	unsigned int vic = pipe_ctx->stream->timing.vic;
 	enum dc_timing_3d_format format;
 
+	memset(&hdmi_info, 0, sizeof(union hdmi_info_packet));
+
 	color_space = pipe_ctx->stream->output_color_space;
 	if (color_space == COLOR_SPACE_UNKNOWN)
 		color_space = (stream->timing.pixel_encoding == PIXEL_ENCODING_RGB) ?
 			COLOR_SPACE_SRGB:COLOR_SPACE_YCBCR709;
 
 	/* Initialize header */
-	hdmi_info->bits.header.info_frame_type = HDMI_INFOFRAME_TYPE_AVI;
+	hdmi_info.bits.header.info_frame_type = HDMI_INFOFRAME_TYPE_AVI;
 	/* InfoFrameVersion_3 is defined by CEA861F (Section 6.4), but shall
 	* not be used in HDMI 2.0 (Section 10.1) */
-	hdmi_info->bits.header.version = 2;
-	hdmi_info->bits.header.length = HDMI_AVI_INFOFRAME_SIZE;
+	hdmi_info.bits.header.version = 2;
+	hdmi_info.bits.header.length = HDMI_AVI_INFOFRAME_SIZE;
 
 	/*
 	 * IDO-defined (Y2,Y1,Y0 = 1,1,1) shall not be used by devices built
@@ -1918,39 +2083,39 @@ static void set_avi_info_frame(
 
 	/* Y0_Y1_Y2 : The pixel encoding */
 	/* H14b AVI InfoFrame has extension on Y-field from 2 bits to 3 bits */
-	hdmi_info->bits.Y0_Y1_Y2 = pixel_encoding;
+	hdmi_info.bits.Y0_Y1_Y2 = pixel_encoding;
 
 	/* A0 = 1 Active Format Information valid */
-	hdmi_info->bits.A0 = ACTIVE_FORMAT_VALID;
+	hdmi_info.bits.A0 = ACTIVE_FORMAT_VALID;
 
 	/* B0, B1 = 3; Bar info data is valid */
-	hdmi_info->bits.B0_B1 = BAR_INFO_BOTH_VALID;
+	hdmi_info.bits.B0_B1 = BAR_INFO_BOTH_VALID;
 
-	hdmi_info->bits.SC0_SC1 = PICTURE_SCALING_UNIFORM;
+	hdmi_info.bits.SC0_SC1 = PICTURE_SCALING_UNIFORM;
 
 	/* S0, S1 : Underscan / Overscan */
 	/* TODO: un-hardcode scan type */
 	scan_type = SCANNING_TYPE_UNDERSCAN;
-	hdmi_info->bits.S0_S1 = scan_type;
+	hdmi_info.bits.S0_S1 = scan_type;
 
 	/* C0, C1 : Colorimetry */
 	if (color_space == COLOR_SPACE_YCBCR709 ||
 			color_space == COLOR_SPACE_YCBCR709_LIMITED)
-		hdmi_info->bits.C0_C1 = COLORIMETRY_ITU709;
+		hdmi_info.bits.C0_C1 = COLORIMETRY_ITU709;
 	else if (color_space == COLOR_SPACE_YCBCR601 ||
 			color_space == COLOR_SPACE_YCBCR601_LIMITED)
-		hdmi_info->bits.C0_C1 = COLORIMETRY_ITU601;
+		hdmi_info.bits.C0_C1 = COLORIMETRY_ITU601;
 	else {
-		hdmi_info->bits.C0_C1 = COLORIMETRY_NO_DATA;
+		hdmi_info.bits.C0_C1 = COLORIMETRY_NO_DATA;
 	}
 	if (color_space == COLOR_SPACE_2020_RGB_FULLRANGE ||
 			color_space == COLOR_SPACE_2020_RGB_LIMITEDRANGE ||
 			color_space == COLOR_SPACE_2020_YCBCR) {
-		hdmi_info->bits.EC0_EC2 = COLORIMETRYEX_BT2020RGBYCBCR;
-		hdmi_info->bits.C0_C1   = COLORIMETRY_EXTENDED;
+		hdmi_info.bits.EC0_EC2 = COLORIMETRYEX_BT2020RGBYCBCR;
+		hdmi_info.bits.C0_C1   = COLORIMETRY_EXTENDED;
 	} else if (color_space == COLOR_SPACE_ADOBERGB) {
-		hdmi_info->bits.EC0_EC2 = COLORIMETRYEX_ADOBERGB;
-		hdmi_info->bits.C0_C1   = COLORIMETRY_EXTENDED;
+		hdmi_info.bits.EC0_EC2 = COLORIMETRYEX_ADOBERGB;
+		hdmi_info.bits.C0_C1   = COLORIMETRY_EXTENDED;
 	}
 
 	/* TODO: un-hardcode aspect ratio */
@@ -1959,18 +2124,18 @@ static void set_avi_info_frame(
 	switch (aspect) {
 	case ASPECT_RATIO_4_3:
 	case ASPECT_RATIO_16_9:
-		hdmi_info->bits.M0_M1 = aspect;
+		hdmi_info.bits.M0_M1 = aspect;
 		break;
 
 	case ASPECT_RATIO_NO_DATA:
 	case ASPECT_RATIO_64_27:
 	case ASPECT_RATIO_256_135:
 	default:
-		hdmi_info->bits.M0_M1 = 0;
+		hdmi_info.bits.M0_M1 = 0;
 	}
 
 	/* Active Format Aspect ratio - same as Picture Aspect Ratio. */
-	hdmi_info->bits.R0_R3 = ACTIVE_FORMAT_ASPECT_RATIO_SAME_AS_PICTURE;
+	hdmi_info.bits.R0_R3 = ACTIVE_FORMAT_ASPECT_RATIO_SAME_AS_PICTURE;
 
 	/* TODO: un-hardcode cn0_cn1 and itc */
 
@@ -2013,8 +2178,8 @@ static void set_avi_info_frame(
 				}
 			}
 		}
-		hdmi_info->bits.CN0_CN1 = cn0_cn1_value;
-		hdmi_info->bits.ITC = itc_value;
+		hdmi_info.bits.CN0_CN1 = cn0_cn1_value;
+		hdmi_info.bits.ITC = itc_value;
 	}
 
 	/* TODO : We should handle YCC quantization */
@@ -2023,19 +2188,19 @@ static void set_avi_info_frame(
 			stream->sink->edid_caps.qy_bit == 1) {
 		if (color_space == COLOR_SPACE_SRGB ||
 			color_space == COLOR_SPACE_2020_RGB_FULLRANGE) {
-			hdmi_info->bits.Q0_Q1   = RGB_QUANTIZATION_FULL_RANGE;
-			hdmi_info->bits.YQ0_YQ1 = YYC_QUANTIZATION_FULL_RANGE;
+			hdmi_info.bits.Q0_Q1   = RGB_QUANTIZATION_FULL_RANGE;
+			hdmi_info.bits.YQ0_YQ1 = YYC_QUANTIZATION_FULL_RANGE;
 		} else if (color_space == COLOR_SPACE_SRGB_LIMITED ||
 					color_space == COLOR_SPACE_2020_RGB_LIMITEDRANGE) {
-			hdmi_info->bits.Q0_Q1   = RGB_QUANTIZATION_LIMITED_RANGE;
-			hdmi_info->bits.YQ0_YQ1 = YYC_QUANTIZATION_LIMITED_RANGE;
+			hdmi_info.bits.Q0_Q1   = RGB_QUANTIZATION_LIMITED_RANGE;
+			hdmi_info.bits.YQ0_YQ1 = YYC_QUANTIZATION_LIMITED_RANGE;
 		} else {
-			hdmi_info->bits.Q0_Q1   = RGB_QUANTIZATION_DEFAULT_RANGE;
-			hdmi_info->bits.YQ0_YQ1 = YYC_QUANTIZATION_LIMITED_RANGE;
+			hdmi_info.bits.Q0_Q1   = RGB_QUANTIZATION_DEFAULT_RANGE;
+			hdmi_info.bits.YQ0_YQ1 = YYC_QUANTIZATION_LIMITED_RANGE;
 		}
 	} else {
-		hdmi_info->bits.Q0_Q1   = RGB_QUANTIZATION_DEFAULT_RANGE;
-		hdmi_info->bits.YQ0_YQ1   = YYC_QUANTIZATION_LIMITED_RANGE;
+		hdmi_info.bits.Q0_Q1   = RGB_QUANTIZATION_DEFAULT_RANGE;
+		hdmi_info.bits.YQ0_YQ1   = YYC_QUANTIZATION_LIMITED_RANGE;
 	}
 
 	///VIC
@@ -2060,51 +2225,49 @@ static void set_avi_info_frame(
 			break;
 		}
 	}
-	hdmi_info->bits.VIC0_VIC7 = vic;
+	hdmi_info.bits.VIC0_VIC7 = vic;
 
 	/* pixel repetition
 	 * PR0 - PR3 start from 0 whereas pHwPathMode->mode.timing.flags.pixel
 	 * repetition start from 1 */
-	hdmi_info->bits.PR0_PR3 = 0;
+	hdmi_info.bits.PR0_PR3 = 0;
 
 	/* Bar Info
 	 * barTop:    Line Number of End of Top Bar.
 	 * barBottom: Line Number of Start of Bottom Bar.
 	 * barLeft:   Pixel Number of End of Left Bar.
 	 * barRight:  Pixel Number of Start of Right Bar. */
-	hdmi_info->bits.bar_top = stream->timing.v_border_top;
-	hdmi_info->bits.bar_bottom = (stream->timing.v_total
+	hdmi_info.bits.bar_top = stream->timing.v_border_top;
+	hdmi_info.bits.bar_bottom = (stream->timing.v_total
 			- stream->timing.v_border_bottom + 1);
-	hdmi_info->bits.bar_left  = stream->timing.h_border_left;
-	hdmi_info->bits.bar_right = (stream->timing.h_total
+	hdmi_info.bits.bar_left  = stream->timing.h_border_left;
+	hdmi_info.bits.bar_right = (stream->timing.h_total
 			- stream->timing.h_border_right + 1);
 
 	/* check_sum - Calculate AFMT_AVI_INFO0 ~ AFMT_AVI_INFO3 */
-	check_sum = &info_frame.avi_info_packet.info_packet_hdmi.packet_raw_data.sb[0];
+	check_sum = &hdmi_info.packet_raw_data.sb[0];
 
 	*check_sum = HDMI_INFOFRAME_TYPE_AVI + HDMI_AVI_INFOFRAME_SIZE + 2;
 
 	for (byte_index = 1; byte_index <= HDMI_AVI_INFOFRAME_SIZE; byte_index++)
-		*check_sum += hdmi_info->packet_raw_data.sb[byte_index];
+		*check_sum += hdmi_info.packet_raw_data.sb[byte_index];
 
 	/* one byte complement */
 	*check_sum = (uint8_t) (0x100 - *check_sum);
 
 	/* Store in hw_path_mode */
-	info_packet->hb0 = hdmi_info->packet_raw_data.hb0;
-	info_packet->hb1 = hdmi_info->packet_raw_data.hb1;
-	info_packet->hb2 = hdmi_info->packet_raw_data.hb2;
+	info_packet->hb0 = hdmi_info.packet_raw_data.hb0;
+	info_packet->hb1 = hdmi_info.packet_raw_data.hb1;
+	info_packet->hb2 = hdmi_info.packet_raw_data.hb2;
 
-	for (byte_index = 0; byte_index < sizeof(info_frame.avi_info_packet.
-				info_packet_hdmi.packet_raw_data.sb); byte_index++)
-		info_packet->sb[byte_index] = info_frame.avi_info_packet.
-				info_packet_hdmi.packet_raw_data.sb[byte_index];
+	for (byte_index = 0; byte_index < sizeof(hdmi_info.packet_raw_data.sb); byte_index++)
+		info_packet->sb[byte_index] = hdmi_info.packet_raw_data.sb[byte_index];
 
 	info_packet->valid = true;
 }
 
 static void set_vendor_info_packet(
-		struct encoder_info_packet *info_packet,
+		struct dc_info_packet *info_packet,
 		struct dc_stream_state *stream)
 {
 	uint32_t length = 0;
@@ -2217,7 +2380,7 @@ static void set_vendor_info_packet(
 }
 
 static void set_spd_info_packet(
-		struct encoder_info_packet *info_packet,
+		struct dc_info_packet *info_packet,
 		struct dc_stream_state *stream)
 {
 	/* SPD info packet for FreeSync */
@@ -2338,104 +2501,20 @@ static void set_spd_info_packet(
 }
 
 static void set_hdr_static_info_packet(
-		struct encoder_info_packet *info_packet,
+		struct dc_info_packet *info_packet,
 		struct dc_stream_state *stream)
 {
-	uint16_t i = 0;
-	enum signal_type signal = stream->signal;
-	uint32_t data;
+	/* HDR Static Metadata info packet for HDR10 */
 
-	if (!stream->hdr_static_metadata.hdr_supported)
+	if (!stream->hdr_static_metadata.valid ||
+			stream->use_dynamic_meta)
 		return;
 
-	if (dc_is_hdmi_signal(signal)) {
-		info_packet->valid = true;
-
-		info_packet->hb0 = 0x87;
-		info_packet->hb1 = 0x01;
-		info_packet->hb2 = 0x1A;
-		i = 1;
-	} else if (dc_is_dp_signal(signal)) {
-		info_packet->valid = true;
-
-		info_packet->hb0 = 0x00;
-		info_packet->hb1 = 0x87;
-		info_packet->hb2 = 0x1D;
-		info_packet->hb3 = (0x13 << 2);
-		i = 2;
-	}
-
-	data = stream->hdr_static_metadata.is_hdr;
-	info_packet->sb[i++] = data ? 0x02 : 0x00;
-	info_packet->sb[i++] = 0x00;
-
-	data = stream->hdr_static_metadata.chromaticity_green_x / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.chromaticity_green_y / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.chromaticity_blue_x / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.chromaticity_blue_y / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.chromaticity_red_x / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.chromaticity_red_y / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.chromaticity_white_point_x / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.chromaticity_white_point_y / 2;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.max_luminance;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.min_luminance;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.maximum_content_light_level;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	data = stream->hdr_static_metadata.maximum_frame_average_light_level;
-	info_packet->sb[i++] = data & 0xFF;
-	info_packet->sb[i++] = (data & 0xFF00) >> 8;
-
-	if (dc_is_hdmi_signal(signal)) {
-		uint32_t checksum = 0;
-
-		checksum += info_packet->hb0;
-		checksum += info_packet->hb1;
-		checksum += info_packet->hb2;
-
-		for (i = 1; i <= info_packet->hb2; i++)
-			checksum += info_packet->sb[i];
-
-		info_packet->sb[0] = 0x100 - checksum;
-	} else if (dc_is_dp_signal(signal)) {
-		info_packet->sb[0] = 0x01;
-		info_packet->sb[1] = 0x1A;
-	}
+	*info_packet = stream->hdr_static_metadata;
 }
 
 static void set_vsc_info_packet(
-		struct encoder_info_packet *info_packet,
+		struct dc_info_packet *info_packet,
 		struct dc_stream_state *stream)
 {
 	unsigned int vscPacketRevision = 0;
@@ -2650,6 +2729,8 @@ bool pipe_need_reprogram(
 	if (is_timing_changed(pipe_ctx_old->stream, pipe_ctx->stream))
 		return true;
 
+	if (is_hdr_static_meta_changed(pipe_ctx_old->stream, pipe_ctx->stream))
+		return true;
 
 	return false;
 }
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_sink.c b/drivers/gpu/drm/amd/display/dc/core/dc_sink.c
index 25fae38409ab..9971b515c3eb 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_sink.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_sink.c
@@ -53,6 +53,10 @@ static bool construct(struct dc_sink *sink, const struct dc_sink_init_data *init
 	sink->dongle_max_pix_clk = init_params->dongle_max_pix_clk;
 	sink->converter_disable_audio = init_params->converter_disable_audio;
 	sink->dc_container_id = NULL;
+	sink->sink_id = init_params->link->ctx->dc_sink_id_count;
+	// increment dc_sink_id_count because we don't want two sinks with same ID
+	// unless they are actually the same
+	init_params->link->ctx->dc_sink_id_count++;
 
 	return true;
 }
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
index ce0747ed0f00..fdcc8ab19bf3 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
@@ -30,6 +30,8 @@
 #include "ipp.h"
 #include "timing_generator.h"
 
+#define DC_LOGGER dc->ctx->logger
+
 /*******************************************************************************
  * Private functions
  ******************************************************************************/
@@ -101,14 +103,16 @@ static void construct(struct dc_stream_state *stream,
 	stream->status.link = stream->sink->link;
 
 	update_stream_signal(stream);
+
+	stream->out_transfer_func = dc_create_transfer_func();
+	stream->out_transfer_func->type = TF_TYPE_BYPASS;
 }
 
 static void destruct(struct dc_stream_state *stream)
 {
 	dc_sink_release(stream->sink);
 	if (stream->out_transfer_func != NULL) {
-		dc_transfer_func_release(
-				stream->out_transfer_func);
+		dc_transfer_func_release(stream->out_transfer_func);
 		stream->out_transfer_func = NULL;
 	}
 }
@@ -176,6 +180,7 @@ bool dc_stream_set_cursor_attributes(
 	int i;
 	struct dc  *core_dc;
 	struct resource_context *res_ctx;
+	struct pipe_ctx *pipe_to_program = NULL;
 
 	if (NULL == stream) {
 		dm_error("DC: dc_stream is NULL!\n");
@@ -203,9 +208,19 @@ bool dc_stream_set_cursor_attributes(
 		if (pipe_ctx->top_pipe && pipe_ctx->plane_state != pipe_ctx->top_pipe->plane_state)
 			continue;
 
+		if (!pipe_to_program) {
+			pipe_to_program = pipe_ctx;
+			core_dc->hwss.pipe_control_lock(core_dc, pipe_to_program, true);
+		}
 
 		core_dc->hwss.set_cursor_attribute(pipe_ctx);
+		if (core_dc->hwss.set_cursor_sdr_white_level)
+			core_dc->hwss.set_cursor_sdr_white_level(pipe_ctx);
 	}
+
+	if (pipe_to_program)
+		core_dc->hwss.pipe_control_lock(core_dc, pipe_to_program, false);
+
 	return true;
 }
 
@@ -216,6 +231,7 @@ bool dc_stream_set_cursor_position(
 	int i;
 	struct dc  *core_dc;
 	struct resource_context *res_ctx;
+	struct pipe_ctx *pipe_to_program = NULL;
 
 	if (NULL == stream) {
 		dm_error("DC: dc_stream is NULL!\n");
@@ -241,9 +257,17 @@ bool dc_stream_set_cursor_position(
 				!pipe_ctx->plane_res.ipp)
 			continue;
 
+		if (!pipe_to_program) {
+			pipe_to_program = pipe_ctx;
+			core_dc->hwss.pipe_control_lock(core_dc, pipe_to_program, true);
+		}
+
 		core_dc->hwss.set_cursor_position(pipe_ctx);
 	}
 
+	if (pipe_to_program)
+		core_dc->hwss.pipe_control_lock(core_dc, pipe_to_program, false);
+
 	return true;
 }
 
@@ -297,16 +321,10 @@ bool dc_stream_get_scanoutpos(const struct dc_stream_state *stream,
 	return ret;
 }
 
-
-void dc_stream_log(
-	const struct dc_stream_state *stream,
-	struct dal_logger *dm_logger,
-	enum dc_log_type log_type)
+void dc_stream_log(const struct dc *dc, const struct dc_stream_state *stream)
 {
-
-	dm_logger_write(dm_logger,
-			log_type,
-			"core_stream 0x%x: src: %d, %d, %d, %d; dst: %d, %d, %d, %d, colorSpace:%d\n",
+	DC_LOG_DC(
+			"core_stream 0x%p: src: %d, %d, %d, %d; dst: %d, %d, %d, %d, colorSpace:%d\n",
 			stream,
 			stream->src.x,
 			stream->src.y,
@@ -317,21 +335,18 @@ void dc_stream_log(
 			stream->dst.width,
 			stream->dst.height,
 			stream->output_color_space);
-	dm_logger_write(dm_logger,
-			log_type,
+	DC_LOG_DC(
 			"\tpix_clk_khz: %d, h_total: %d, v_total: %d, pixelencoder:%d, displaycolorDepth:%d\n",
 			stream->timing.pix_clk_khz,
 			stream->timing.h_total,
 			stream->timing.v_total,
 			stream->timing.pixel_encoding,
 			stream->timing.display_color_depth);
-	dm_logger_write(dm_logger,
-			log_type,
+	DC_LOG_DC(
 			"\tsink name: %s, serial: %d\n",
 			stream->sink->edid_caps.display_name,
 			stream->sink->edid_caps.serial_number);
-	dm_logger_write(dm_logger,
-			log_type,
+	DC_LOG_DC(
 			"\tlink: %d\n",
 			stream->sink->link->link_index);
 }
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
index ade5b8ee9c3c..815dfb50089b 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
@@ -38,6 +38,12 @@
 static void construct(struct dc_context *ctx, struct dc_plane_state *plane_state)
 {
 	plane_state->ctx = ctx;
+
+	plane_state->gamma_correction = dc_create_gamma();
+	plane_state->gamma_correction->is_identity = true;
+
+	plane_state->in_transfer_func = dc_create_transfer_func();
+	plane_state->in_transfer_func->type = TF_TYPE_BYPASS;
 }
 
 static void destruct(struct dc_plane_state *plane_state)
@@ -66,8 +72,8 @@ struct dc_plane_state *dc_create_plane_state(struct dc *dc)
 {
 	struct dc *core_dc = dc;
 
-	struct dc_plane_state *plane_state = kzalloc(sizeof(*plane_state),
-						     GFP_KERNEL);
+	struct dc_plane_state *plane_state = kvzalloc(sizeof(*plane_state),
+						      GFP_KERNEL);
 
 	if (NULL == plane_state)
 		return NULL;
@@ -78,6 +84,17 @@ struct dc_plane_state *dc_create_plane_state(struct dc *dc)
 	return plane_state;
 }
 
+/**
+ *****************************************************************************
+ *  Function: dc_plane_get_status
+ *
+ *  @brief
+ *     Looks up the pipe context of plane_state and updates the pending status
+ *     of the pipe context. Then returns plane_state->status
+ *
+ *  @param [in] plane_state: pointer to the plane_state to get the status of
+ *****************************************************************************
+ */
 const struct dc_plane_status *dc_plane_get_status(
 		const struct dc_plane_state *plane_state)
 {
@@ -120,7 +137,7 @@ static void dc_plane_state_free(struct kref *kref)
 {
 	struct dc_plane_state *plane_state = container_of(kref, struct dc_plane_state, refcount);
 	destruct(plane_state);
-	kfree(plane_state);
+	kvfree(plane_state);
 }
 
 void dc_plane_state_release(struct dc_plane_state *plane_state)
@@ -136,7 +153,7 @@ void dc_gamma_retain(struct dc_gamma *gamma)
 static void dc_gamma_free(struct kref *kref)
 {
 	struct dc_gamma *gamma = container_of(kref, struct dc_gamma, refcount);
-	kfree(gamma);
+	kvfree(gamma);
 }
 
 void dc_gamma_release(struct dc_gamma **gamma)
@@ -147,7 +164,7 @@ void dc_gamma_release(struct dc_gamma **gamma)
 
 struct dc_gamma *dc_create_gamma(void)
 {
-	struct dc_gamma *gamma = kzalloc(sizeof(*gamma), GFP_KERNEL);
+	struct dc_gamma *gamma = kvzalloc(sizeof(*gamma), GFP_KERNEL);
 
 	if (gamma == NULL)
 		goto alloc_fail;
@@ -167,7 +184,7 @@ void dc_transfer_func_retain(struct dc_transfer_func *tf)
 static void dc_transfer_func_free(struct kref *kref)
 {
 	struct dc_transfer_func *tf = container_of(kref, struct dc_transfer_func, refcount);
-	kfree(tf);
+	kvfree(tf);
 }
 
 void dc_transfer_func_release(struct dc_transfer_func *tf)
@@ -175,9 +192,9 @@ void dc_transfer_func_release(struct dc_transfer_func *tf)
 	kref_put(&tf->refcount, dc_transfer_func_free);
 }
 
-struct dc_transfer_func *dc_create_transfer_func(void)
+struct dc_transfer_func *dc_create_transfer_func()
 {
-	struct dc_transfer_func *tf = kzalloc(sizeof(*tf), GFP_KERNEL);
+	struct dc_transfer_func *tf = kvzalloc(sizeof(*tf), GFP_KERNEL);
 
 	if (tf == NULL)
 		goto alloc_fail;
diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index fa4b3c8b3bb7..ceb4c3725893 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -38,7 +38,7 @@
 #include "inc/compressor.h"
 #include "dml/display_mode_lib.h"
 
-#define DC_VER "3.1.38"
+#define DC_VER "3.1.56"
 
 #define MAX_SURFACES 3
 #define MAX_STREAMS 6
@@ -68,6 +68,7 @@ struct dc_caps {
 	uint32_t max_planes;
 	uint32_t max_downscale_ratio;
 	uint32_t i2c_speed_in_khz;
+	uint32_t dmdata_alloc_size;
 	unsigned int max_cursor_size;
 	unsigned int max_video_width;
 	int linear_pitch_alignment;
@@ -75,6 +76,7 @@ struct dc_caps {
 	bool dynamic_audio;
 	bool is_apu;
 	bool dual_link_dvi;
+	bool post_blend_color_processing;
 };
 
 struct dc_dcc_surface_param {
@@ -167,6 +169,12 @@ struct dc_config {
 	bool disable_disp_pll_sharing;
 };
 
+enum visual_confirm {
+	VISUAL_CONFIRM_DISABLE = 0,
+	VISUAL_CONFIRM_SURFACE = 1,
+	VISUAL_CONFIRM_HDR = 2,
+};
+
 enum dcc_option {
 	DCC_ENABLE = 0,
 	DCC_DISABLE = 1,
@@ -184,6 +192,10 @@ enum wm_report_mode {
 	WM_REPORT_OVERRIDE = 1,
 };
 
+/*
+ * For any clocks that may differ per pipe
+ * only the max is stored in this structure
+ */
 struct dc_clocks {
 	int dispclk_khz;
 	int max_supported_dppclk_khz;
@@ -192,16 +204,19 @@ struct dc_clocks {
 	int socclk_khz;
 	int dcfclk_deep_sleep_khz;
 	int fclk_khz;
+	int phyclk_khz;
 };
 
 struct dc_debug {
-	bool surface_visual_confirm;
+	enum visual_confirm visual_confirm;
 	bool sanity_checks;
 	bool max_disp_clk;
 	bool surface_trace;
 	bool timing_trace;
 	bool clock_trace;
 	bool validation_trace;
+	bool bandwidth_calcs_trace;
+	int max_downscale_src_width;
 
 	/* stutter efficiency related */
 	bool disable_stutter;
@@ -224,6 +239,7 @@ struct dc_debug {
 	int urgent_latency_ns;
 	int percent_of_ideal_drambw;
 	int dram_clock_change_latency_ns;
+	bool optimized_watermark;
 	int always_scale;
 	bool disable_pplib_clock_request;
 	bool disable_clock_gate;
@@ -238,6 +254,16 @@ struct dc_debug {
 	bool az_endpoint_mute_only;
 	bool always_use_regamma;
 	bool p010_mpo_support;
+	bool recovery_enabled;
+	bool avoid_vbios_exec_table;
+	bool scl_reset_length10;
+	bool hdmi20_disable;
+
+	struct {
+		uint32_t ltFailCount;
+		uint32_t i2cErrorCount;
+		uint32_t auxErrorCount;
+	} debug_data;
 };
 struct dc_state;
 struct resource_pool;
@@ -263,7 +289,7 @@ struct dc {
 	/* Inputs into BW and WM calculations. */
 	struct bw_calcs_dceip *bw_dceip;
 	struct bw_calcs_vbios *bw_vbios;
-#ifdef CONFIG_DRM_AMD_DC_DCN1_0
+#ifdef CONFIG_X86
 	struct dcn_soc_bounding_box *dcn_soc;
 	struct dcn_ip_params *dcn_ip;
 	struct display_mode_lib dml;
@@ -283,9 +309,7 @@ struct dc {
 	bool apply_edp_fast_boot_optimization;
 
 	/* FBC compressor */
-#if defined(CONFIG_DRM_AMD_DC_FBC)
 	struct compressor *fbc_compressor;
-#endif
 };
 
 enum frame_buffer_mode {
@@ -332,20 +356,6 @@ enum {
 	TRANSFER_FUNC_POINTS = 1025
 };
 
-// Moved here from color module for linux
-enum color_transfer_func {
-	transfer_func_unknown,
-	transfer_func_srgb,
-	transfer_func_bt709,
-	transfer_func_pq2084,
-	transfer_func_pq2084_interim,
-	transfer_func_linear_0_1,
-	transfer_func_linear_0_125,
-	transfer_func_dolbyvision,
-	transfer_func_gamma_22,
-	transfer_func_gamma_26
-};
-
 struct dc_hdr_static_metadata {
 	/* display chromaticities and white point in units of 0.00001 */
 	unsigned int chromaticity_green_x;
@@ -361,15 +371,13 @@ struct dc_hdr_static_metadata {
 	uint32_t max_luminance;
 	uint32_t maximum_content_light_level;
 	uint32_t maximum_frame_average_light_level;
-
-	bool hdr_supported;
-	bool is_hdr;
 };
 
 enum dc_transfer_func_type {
 	TF_TYPE_PREDEFINED,
 	TF_TYPE_DISTRIBUTED_POINTS,
 	TF_TYPE_BYPASS,
+	TF_TYPE_HWPWL
 };
 
 struct dc_transfer_func_distributed_points {
@@ -389,16 +397,22 @@ enum dc_transfer_func_predefined {
 	TRANSFER_FUNCTION_PQ,
 	TRANSFER_FUNCTION_LINEAR,
 	TRANSFER_FUNCTION_UNITY,
+	TRANSFER_FUNCTION_HLG,
+	TRANSFER_FUNCTION_HLG12,
+	TRANSFER_FUNCTION_GAMMA22
 };
 
 struct dc_transfer_func {
 	struct kref refcount;
-	struct dc_transfer_func_distributed_points tf_pts;
 	enum dc_transfer_func_type type;
 	enum dc_transfer_func_predefined tf;
 	/* FP16 1.0 reference level in nits, default is 80 nits, only for PQ*/
 	uint32_t sdr_ref_white_level;
 	struct dc_context *ctx;
+	union {
+		struct pwl_params pwl;
+		struct dc_transfer_func_distributed_points tf_pts;
+	};
 };
 
 /*
@@ -419,7 +433,6 @@ union surface_update_flags {
 		/* Medium updates */
 		uint32_t dcc_change:1;
 		uint32_t color_space_change:1;
-		uint32_t input_tf_change:1;
 		uint32_t horizontal_mirror_change:1;
 		uint32_t per_pixel_alpha_change:1;
 		uint32_t rotation_change:1;
@@ -428,6 +441,7 @@ union surface_update_flags {
 		uint32_t position_change:1;
 		uint32_t in_transfer_func_change:1;
 		uint32_t input_csc_change:1;
+		uint32_t coeff_reduction_change:1;
 		uint32_t output_tf_change:1;
 		uint32_t pixel_format_change:1;
 
@@ -460,7 +474,7 @@ struct dc_plane_state {
 	struct dc_gamma *gamma_correction;
 	struct dc_transfer_func *in_transfer_func;
 	struct dc_bias_and_scale *bias_and_scale;
-	struct csc_transform input_csc_color_matrix;
+	struct dc_csc_transform input_csc_color_matrix;
 	struct fixed31_32 coeff_reduction_factor;
 	uint32_t sdr_white_level;
 
@@ -468,7 +482,6 @@ struct dc_plane_state {
 	struct dc_hdr_static_metadata hdr_static_ctx;
 
 	enum dc_color_space color_space;
-	enum color_transfer_func input_tf;
 
 	enum surface_pixel_format format;
 	enum dc_rotation_angle rotation;
@@ -498,7 +511,6 @@ struct dc_plane_info {
 	enum dc_rotation_angle rotation;
 	enum plane_stereo_format stereo_format;
 	enum dc_color_space color_space;
-	enum color_transfer_func input_tf;
 	unsigned int sdr_white_level;
 	bool horizontal_mirror;
 	bool visible;
@@ -517,19 +529,18 @@ struct dc_surface_update {
 	struct dc_plane_state *surface;
 
 	/* isr safe update parameters.  null means no updates */
-	struct dc_flip_addrs *flip_addr;
-	struct dc_plane_info *plane_info;
-	struct dc_scaling_info *scaling_info;
+	const struct dc_flip_addrs *flip_addr;
+	const struct dc_plane_info *plane_info;
+	const struct dc_scaling_info *scaling_info;
 
 	/* following updates require alloc/sleep/spin that is not isr safe,
 	 * null means no updates
 	 */
-	struct dc_gamma *gamma;
-	enum color_transfer_func color_input_tf;
-	struct dc_transfer_func *in_transfer_func;
+	const struct dc_gamma *gamma;
+	const struct dc_transfer_func *in_transfer_func;
 
-	struct csc_transform *input_csc_color_matrix;
-	struct fixed31_32 *coeff_reduction_factor;
+	const struct dc_csc_transform *input_csc_color_matrix;
+	const struct fixed31_32 *coeff_reduction_factor;
 };
 
 /*
@@ -631,9 +642,14 @@ struct dpcd_caps {
 	struct dc_dongle_caps dongle_caps;
 
 	uint32_t sink_dev_id;
+	int8_t sink_dev_id_str[6];
+	int8_t sink_hw_revision;
+	int8_t sink_fw_revision[2];
+
 	uint32_t branch_dev_id;
 	int8_t branch_dev_name[6];
 	int8_t branch_hw_revision;
+	int8_t branch_fw_revision[2];
 
 	bool allow_invalid_MSA_timing_param;
 	bool panel_mode_edp;
@@ -676,9 +692,13 @@ struct dc_sink {
 	struct dc_link *link;
 	struct dc_context *ctx;
 
+	uint32_t sink_id;
+
 	/* private to dc_sink.c */
+	// refcount must be the last member in dc_sink, since we want the
+	// sink structure to be logically cloneable up to (but not including)
+	// refcount
 	struct kref refcount;
-
 };
 
 void dc_sink_retain(struct dc_sink *sink);
@@ -699,6 +719,7 @@ struct dc_cursor {
 	struct dc_cursor_attributes attributes;
 };
 
+
 /*******************************************************************************
  * Interrupt interfaces
  ******************************************************************************/
diff --git a/drivers/gpu/drm/amd/display/dc/dc_bios_types.h b/drivers/gpu/drm/amd/display/dc/dc_bios_types.h
index d9b84ec7954c..90082bab71f0 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_bios_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_bios_types.h
@@ -198,6 +198,10 @@ struct dc_vbios_funcs {
 	void (*post_init)(struct dc_bios *bios);
 
 	void (*bios_parser_destroy)(struct dc_bios **dcb);
+
+	enum bp_result (*get_board_layout_info)(
+		struct dc_bios *dcb,
+		struct board_layout_info *board_layout_info);
 };
 
 struct bios_registers {
diff --git a/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h b/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h
index e1affeb5cc51..05c8c31d8b31 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_ddc_types.h
@@ -25,6 +25,65 @@
 #ifndef DC_DDC_TYPES_H_
 #define DC_DDC_TYPES_H_
 
+enum aux_transaction_type {
+	AUX_TRANSACTION_TYPE_DP,
+	AUX_TRANSACTION_TYPE_I2C
+};
+
+
+enum i2caux_transaction_action {
+	I2CAUX_TRANSACTION_ACTION_I2C_WRITE = 0x00,
+	I2CAUX_TRANSACTION_ACTION_I2C_READ = 0x10,
+	I2CAUX_TRANSACTION_ACTION_I2C_STATUS_REQUEST = 0x20,
+
+	I2CAUX_TRANSACTION_ACTION_I2C_WRITE_MOT = 0x40,
+	I2CAUX_TRANSACTION_ACTION_I2C_READ_MOT = 0x50,
+	I2CAUX_TRANSACTION_ACTION_I2C_STATUS_REQUEST_MOT = 0x60,
+
+	I2CAUX_TRANSACTION_ACTION_DP_WRITE = 0x80,
+	I2CAUX_TRANSACTION_ACTION_DP_READ = 0x90
+};
+
+enum aux_channel_operation_result {
+	AUX_CHANNEL_OPERATION_SUCCEEDED,
+	AUX_CHANNEL_OPERATION_FAILED_REASON_UNKNOWN,
+	AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY,
+	AUX_CHANNEL_OPERATION_FAILED_TIMEOUT,
+	AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON
+};
+
+
+struct aux_request_transaction_data {
+	enum aux_transaction_type type;
+	enum i2caux_transaction_action action;
+	/* 20-bit AUX channel transaction address */
+	uint32_t address;
+	/* delay, in 100-microsecond units */
+	uint8_t delay;
+	uint32_t length;
+	uint8_t *data;
+};
+
+enum aux_transaction_reply {
+	AUX_TRANSACTION_REPLY_AUX_ACK = 0x00,
+	AUX_TRANSACTION_REPLY_AUX_NACK = 0x01,
+	AUX_TRANSACTION_REPLY_AUX_DEFER = 0x02,
+
+	AUX_TRANSACTION_REPLY_I2C_ACK = 0x00,
+	AUX_TRANSACTION_REPLY_I2C_NACK = 0x10,
+	AUX_TRANSACTION_REPLY_I2C_DEFER = 0x20,
+
+	AUX_TRANSACTION_REPLY_HPD_DISCON = 0x40,
+
+	AUX_TRANSACTION_REPLY_INVALID = 0xFF
+};
+
+struct aux_reply_transaction_data {
+	enum aux_transaction_reply status;
+	uint32_t length;
+	uint8_t *data;
+};
+
 struct i2c_payload {
 	bool write;
 	uint8_t address;
@@ -109,7 +168,7 @@ struct ddc_service {
 
 	uint32_t address;
 	uint32_t edid_buf_len;
-	uint8_t edid_buf[MAX_EDID_BUFFER_SIZE];
+	uint8_t edid_buf[DC_MAX_EDID_BUFFER_SIZE];
 };
 
 #endif /* DC_DDC_TYPES_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h
index 2726b02e006b..da93ab43f2d8 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h
@@ -26,6 +26,8 @@
 #ifndef DC_DP_TYPES_H
 #define DC_DP_TYPES_H
 
+#include "os_types.h"
+
 enum dc_lane_count {
 	LANE_COUNT_UNKNOWN = 0,
 	LANE_COUNT_ONE = 1,
@@ -428,7 +430,7 @@ union test_request {
 	struct {
 	uint8_t LINK_TRAINING         :1;
 	uint8_t LINK_TEST_PATTRN      :1;
-	uint8_t EDID_REAT             :1;
+	uint8_t EDID_READ             :1;
 	uint8_t PHY_TEST_PATTERN      :1;
 	uint8_t AUDIO_TEST_PATTERN    :1;
 	uint8_t RESERVED              :1;
@@ -441,7 +443,8 @@ union test_response {
 	struct {
 		uint8_t ACK         :1;
 		uint8_t NO_ACK      :1;
-		uint8_t RESERVED    :6;
+		uint8_t EDID_CHECKSUM_WRITE:1;
+		uint8_t RESERVED    :5;
 	} bits;
 	uint8_t raw;
 };
diff --git a/drivers/gpu/drm/amd/display/dc/dc_helper.c b/drivers/gpu/drm/amd/display/dc/dc_helper.c
index 48e1fcf53d43..e68077e65565 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_helper.c
+++ b/drivers/gpu/drm/amd/display/dc/dc_helper.c
@@ -117,6 +117,65 @@ uint32_t generic_reg_get5(const struct dc_context *ctx, uint32_t addr,
 	return reg_val;
 }
 
+uint32_t generic_reg_get6(const struct dc_context *ctx, uint32_t addr,
+		uint8_t shift1, uint32_t mask1, uint32_t *field_value1,
+		uint8_t shift2, uint32_t mask2, uint32_t *field_value2,
+		uint8_t shift3, uint32_t mask3, uint32_t *field_value3,
+		uint8_t shift4, uint32_t mask4, uint32_t *field_value4,
+		uint8_t shift5, uint32_t mask5, uint32_t *field_value5,
+		uint8_t shift6, uint32_t mask6, uint32_t *field_value6)
+{
+	uint32_t reg_val = dm_read_reg(ctx, addr);
+	*field_value1 = get_reg_field_value_ex(reg_val, mask1, shift1);
+	*field_value2 = get_reg_field_value_ex(reg_val, mask2, shift2);
+	*field_value3 = get_reg_field_value_ex(reg_val, mask3, shift3);
+	*field_value4 = get_reg_field_value_ex(reg_val, mask4, shift4);
+	*field_value5 = get_reg_field_value_ex(reg_val, mask5, shift5);
+	*field_value6 = get_reg_field_value_ex(reg_val, mask6, shift6);
+	return reg_val;
+}
+
+uint32_t generic_reg_get7(const struct dc_context *ctx, uint32_t addr,
+		uint8_t shift1, uint32_t mask1, uint32_t *field_value1,
+		uint8_t shift2, uint32_t mask2, uint32_t *field_value2,
+		uint8_t shift3, uint32_t mask3, uint32_t *field_value3,
+		uint8_t shift4, uint32_t mask4, uint32_t *field_value4,
+		uint8_t shift5, uint32_t mask5, uint32_t *field_value5,
+		uint8_t shift6, uint32_t mask6, uint32_t *field_value6,
+		uint8_t shift7, uint32_t mask7, uint32_t *field_value7)
+{
+	uint32_t reg_val = dm_read_reg(ctx, addr);
+	*field_value1 = get_reg_field_value_ex(reg_val, mask1, shift1);
+	*field_value2 = get_reg_field_value_ex(reg_val, mask2, shift2);
+	*field_value3 = get_reg_field_value_ex(reg_val, mask3, shift3);
+	*field_value4 = get_reg_field_value_ex(reg_val, mask4, shift4);
+	*field_value5 = get_reg_field_value_ex(reg_val, mask5, shift5);
+	*field_value6 = get_reg_field_value_ex(reg_val, mask6, shift6);
+	*field_value7 = get_reg_field_value_ex(reg_val, mask7, shift7);
+	return reg_val;
+}
+
+uint32_t generic_reg_get8(const struct dc_context *ctx, uint32_t addr,
+		uint8_t shift1, uint32_t mask1, uint32_t *field_value1,
+		uint8_t shift2, uint32_t mask2, uint32_t *field_value2,
+		uint8_t shift3, uint32_t mask3, uint32_t *field_value3,
+		uint8_t shift4, uint32_t mask4, uint32_t *field_value4,
+		uint8_t shift5, uint32_t mask5, uint32_t *field_value5,
+		uint8_t shift6, uint32_t mask6, uint32_t *field_value6,
+		uint8_t shift7, uint32_t mask7, uint32_t *field_value7,
+		uint8_t shift8, uint32_t mask8, uint32_t *field_value8)
+{
+	uint32_t reg_val = dm_read_reg(ctx, addr);
+	*field_value1 = get_reg_field_value_ex(reg_val, mask1, shift1);
+	*field_value2 = get_reg_field_value_ex(reg_val, mask2, shift2);
+	*field_value3 = get_reg_field_value_ex(reg_val, mask3, shift3);
+	*field_value4 = get_reg_field_value_ex(reg_val, mask4, shift4);
+	*field_value5 = get_reg_field_value_ex(reg_val, mask5, shift5);
+	*field_value6 = get_reg_field_value_ex(reg_val, mask6, shift6);
+	*field_value7 = get_reg_field_value_ex(reg_val, mask7, shift7);
+	*field_value8 = get_reg_field_value_ex(reg_val, mask8, shift8);
+	return reg_val;
+}
 /* note:  va version of this is pretty bad idea, since there is a output parameter pass by pointer
  * compiler won't be able to check for size match and is prone to stack corruption type of bugs
 
@@ -196,3 +255,54 @@ uint32_t generic_reg_wait(const struct dc_context *ctx,
 
 	return reg_val;
 }
+
+void generic_write_indirect_reg(const struct dc_context *ctx,
+		uint32_t addr_index, uint32_t addr_data,
+		uint32_t index, uint32_t data)
+{
+	dm_write_reg(ctx, addr_index, index);
+	dm_write_reg(ctx, addr_data, data);
+}
+
+uint32_t generic_read_indirect_reg(const struct dc_context *ctx,
+		uint32_t addr_index, uint32_t addr_data,
+		uint32_t index)
+{
+	uint32_t value = 0;
+
+	dm_write_reg(ctx, addr_index, index);
+	value = dm_read_reg(ctx, addr_data);
+
+	return value;
+}
+
+
+uint32_t generic_indirect_reg_update_ex(const struct dc_context *ctx,
+		uint32_t addr_index, uint32_t addr_data,
+		uint32_t index, uint32_t reg_val, int n,
+		uint8_t shift1, uint32_t mask1, uint32_t field_value1,
+		...)
+{
+	uint32_t shift, mask, field_value;
+	int i = 1;
+
+	va_list ap;
+
+	va_start(ap, field_value1);
+
+	reg_val = set_reg_field_value_ex(reg_val, field_value1, mask1, shift1);
+
+	while (i < n) {
+		shift = va_arg(ap, uint32_t);
+		mask = va_arg(ap, uint32_t);
+		field_value = va_arg(ap, uint32_t);
+
+		reg_val = set_reg_field_value_ex(reg_val, field_value, mask, shift);
+		i++;
+	}
+
+	generic_write_indirect_reg(ctx, addr_index, addr_data, index, reg_val);
+	va_end(ap);
+
+	return reg_val;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/dc_hw_types.h b/drivers/gpu/drm/amd/display/dc/dc_hw_types.h
index b83a7dc2f5a9..9cfd7ea845e3 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_hw_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_hw_types.h
@@ -199,6 +199,7 @@ enum surface_pixel_format {
 	SURFACE_PIXEL_FORMAT_VIDEO_420_YCrCb,
 	SURFACE_PIXEL_FORMAT_VIDEO_420_10bpc_YCbCr,
 	SURFACE_PIXEL_FORMAT_VIDEO_420_10bpc_YCrCb,
+		SURFACE_PIXEL_FORMAT_SUBSAMPLE_END,
 	SURFACE_PIXEL_FORMAT_INVALID
 
 	/*grow 444 video here if necessary */
@@ -403,9 +404,11 @@ struct dc_cursor_position {
 struct dc_cursor_mi_param {
 	unsigned int pixel_clk_khz;
 	unsigned int ref_clk_khz;
-	unsigned int viewport_x_start;
-	unsigned int viewport_width;
+	struct rect viewport;
 	struct fixed31_32 h_scale_ratio;
+	struct fixed31_32 v_scale_ratio;
+	enum dc_rotation_angle rotation;
+	bool mirror;
 };
 
 /* IPP related types */
@@ -423,6 +426,11 @@ enum dc_gamma_type {
 	GAMMA_CS_TFM_1D = 3,
 };
 
+struct dc_csc_transform {
+	uint16_t matrix[12];
+	bool enable_adjustment;
+};
+
 struct dc_gamma {
 	struct kref refcount;
 	enum dc_gamma_type type;
@@ -484,6 +492,7 @@ struct dc_cursor_attributes {
 	uint32_t height;
 
 	enum dc_cursor_color_format color_format;
+	uint32_t sdr_white_level; // for boosting (SDR) cursor in HDR mode
 
 	/* In case we support HW Cursor rotation in the future */
 	enum dc_rotation_angle rotation_angle;
@@ -491,6 +500,11 @@ struct dc_cursor_attributes {
 	union dc_cursor_attribute_flags attribute_flags;
 };
 
+struct dpp_cursor_attributes {
+	int bias;
+	int scale;
+};
+
 /* OPP */
 
 enum dc_color_space {
@@ -562,25 +576,25 @@ struct scaling_taps {
 };
 
 enum dc_timing_standard {
-	TIMING_STANDARD_UNDEFINED,
-	TIMING_STANDARD_DMT,
-	TIMING_STANDARD_GTF,
-	TIMING_STANDARD_CVT,
-	TIMING_STANDARD_CVT_RB,
-	TIMING_STANDARD_CEA770,
-	TIMING_STANDARD_CEA861,
-	TIMING_STANDARD_HDMI,
-	TIMING_STANDARD_TV_NTSC,
-	TIMING_STANDARD_TV_NTSC_J,
-	TIMING_STANDARD_TV_PAL,
-	TIMING_STANDARD_TV_PAL_M,
-	TIMING_STANDARD_TV_PAL_CN,
-	TIMING_STANDARD_TV_SECAM,
-	TIMING_STANDARD_EXPLICIT,
+	DC_TIMING_STANDARD_UNDEFINED,
+	DC_TIMING_STANDARD_DMT,
+	DC_TIMING_STANDARD_GTF,
+	DC_TIMING_STANDARD_CVT,
+	DC_TIMING_STANDARD_CVT_RB,
+	DC_TIMING_STANDARD_CEA770,
+	DC_TIMING_STANDARD_CEA861,
+	DC_TIMING_STANDARD_HDMI,
+	DC_TIMING_STANDARD_TV_NTSC,
+	DC_TIMING_STANDARD_TV_NTSC_J,
+	DC_TIMING_STANDARD_TV_PAL,
+	DC_TIMING_STANDARD_TV_PAL_M,
+	DC_TIMING_STANDARD_TV_PAL_CN,
+	DC_TIMING_STANDARD_TV_SECAM,
+	DC_TIMING_STANDARD_EXPLICIT,
 	/*!< For explicit timings from EDID, VBIOS, etc.*/
-	TIMING_STANDARD_USER_OVERRIDE,
+	DC_TIMING_STANDARD_USER_OVERRIDE,
 	/*!< For mode timing override by user*/
-	TIMING_STANDARD_MAX
+	DC_TIMING_STANDARD_MAX
 };
 
 enum dc_color_depth {
diff --git a/drivers/gpu/drm/amd/display/dc/dc_link.h b/drivers/gpu/drm/amd/display/dc/dc_link.h
index dc34515ef01f..070a56926308 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_link.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_link.h
@@ -51,6 +51,14 @@ struct link_mst_stream_allocation_table {
 	struct link_mst_stream_allocation stream_allocations[MAX_CONTROLLER_NUM];
 };
 
+struct time_stamp {
+	uint64_t edp_poweroff;
+	uint64_t edp_poweron;
+};
+
+struct link_trace {
+	struct time_stamp time_stamp;
+};
 /*
  * A link contains one or more sinks and their connected status.
  * The currently active signal type (HDMI, DP-SST, DP-MST) is also reported.
@@ -114,6 +122,7 @@ struct dc_link {
 
 	struct dc_link_status link_status;
 
+	struct link_trace link_trace;
 };
 
 const struct dc_link_status *dc_link_get_status(const struct dc_link *dc_link);
@@ -163,7 +172,7 @@ bool dc_link_detect(struct dc_link *dc_link, enum dc_detect_reason reason);
  * false - no change in Downstream port status. No further action required
  * from DM. */
 bool dc_link_handle_hpd_rx_irq(struct dc_link *dc_link,
-		union hpd_irq_data *hpd_irq_dpcd_data);
+		union hpd_irq_data *hpd_irq_dpcd_data, bool *out_link_loss);
 
 struct dc_sink_init_data;
 
@@ -201,10 +210,29 @@ bool dc_link_dp_set_test_pattern(
 
 void dc_link_enable_hpd_filter(struct dc_link *link, bool enable);
 
+bool dc_link_is_dp_sink_present(struct dc_link *link);
+
 /*
  * DPCD access interfaces
  */
 
+void dc_link_set_drive_settings(struct dc *dc,
+				struct link_training_settings *lt_settings,
+				const struct dc_link *link);
+void dc_link_perform_link_training(struct dc *dc,
+				   struct dc_link_settings *link_setting,
+				   bool skip_video_pattern);
+void dc_link_set_preferred_link_settings(struct dc *dc,
+					 struct dc_link_settings *link_setting,
+					 struct dc_link *link);
+void dc_link_enable_hpd(const struct dc_link *link);
+void dc_link_disable_hpd(const struct dc_link *link);
+void dc_link_set_test_pattern(struct dc_link *link,
+			enum dp_test_pattern test_pattern,
+			const struct link_training_settings *p_link_settings,
+			const unsigned char *p_custom_pattern,
+			unsigned int cust_pattern_size);
+
 bool dc_submit_i2c(
 		struct dc *dc,
 		uint32_t link_index,
diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h
index d017df56b2ba..cbfe418006cb 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_stream.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -58,18 +58,23 @@ struct dc_stream_state {
 
 	struct freesync_context freesync_ctx;
 
-	struct dc_hdr_static_metadata hdr_static_metadata;
+	struct dc_info_packet hdr_static_metadata;
+	PHYSICAL_ADDRESS_LOC dmdata_address;
+	bool   use_dynamic_meta;
+
 	struct dc_transfer_func *out_transfer_func;
 	struct colorspace_transform gamut_remap_matrix;
-	struct csc_transform csc_color_matrix;
+	struct dc_csc_transform csc_color_matrix;
 
 	enum dc_color_space output_color_space;
 	enum dc_dither_option dither_option;
 
 	enum view_3d_format view_format;
-	enum color_transfer_func output_tf;
 
 	bool ignore_msa_timing_param;
+
+	unsigned long long periodic_fn_vsync_delta;
+
 	/* TODO: custom INFO packets */
 	/* TODO: ABM info (DMCU) */
 	/* PSR info */
@@ -95,6 +100,7 @@ struct dc_stream_state {
 
 	struct dc_cursor_attributes cursor_attributes;
 	struct dc_cursor_position cursor_position;
+	uint32_t sdr_white_level; // for boosting (SDR) cursor in HDR mode
 
 	/* from stream struct */
 	struct kref refcount;
@@ -110,9 +116,10 @@ struct dc_stream_update {
 	struct rect src;
 	struct rect dst;
 	struct dc_transfer_func *out_transfer_func;
-	struct dc_hdr_static_metadata *hdr_static_metadata;
-	enum color_transfer_func color_output_tf;
+	struct dc_info_packet *hdr_static_metadata;
 	unsigned int *abm_level;
+
+	unsigned long long *periodic_fn_vsync_delta;
 };
 
 bool dc_is_stream_unchanged(
@@ -131,13 +138,6 @@ bool dc_is_stream_scaling_unchanged(
  *   This does not trigger a flip.  No surface address is programmed.
  */
 
-bool dc_commit_planes_to_stream(
-		struct dc *dc,
-		struct dc_plane_state **plane_states,
-		uint8_t new_plane_count,
-		struct dc_stream_state *dc_stream,
-		struct dc_state *state);
-
 void dc_commit_updates_for_stream(struct dc *dc,
 		struct dc_surface_update *srf_updates,
 		int surface_count,
@@ -148,10 +148,7 @@ void dc_commit_updates_for_stream(struct dc *dc,
 /*
  * Log the current stream state.
  */
-void dc_stream_log(
-	const struct dc_stream_state *stream,
-	struct dal_logger *dc_logger,
-	enum dc_log_type log_type);
+void dc_stream_log(const struct dc *dc, const struct dc_stream_state *stream);
 
 uint8_t dc_get_current_stream_count(struct dc *dc);
 struct dc_stream_state *dc_get_stream_at_index(struct dc *dc, uint8_t i);
@@ -209,14 +206,6 @@ bool dc_add_all_planes_for_stream(
 enum dc_status dc_validate_stream(struct dc *dc, struct dc_stream_state *stream);
 
 /*
- * This function takes a stream and checks if it is guaranteed to be supported.
- * Guaranteed means that MAX_COFUNC similar streams are supported.
- *
- * After this call:
- *   No hardware is programmed for call.  Only validation is done.
- */
-
-/*
  * Set up streams and links associated to drive sinks
  * The streams parameter is an absolute set of all active streams.
  *
@@ -267,6 +256,7 @@ bool dc_stream_set_cursor_position(
 	struct dc_stream_state *stream,
 	const struct dc_cursor_position *position);
 
+
 bool dc_stream_adjust_vmin_vmax(struct dc *dc,
 				struct dc_stream_state **stream,
 				int num_streams,
@@ -311,9 +301,4 @@ bool dc_stream_get_crtc_position(struct dc *dc,
 				 unsigned int *v_pos,
 				 unsigned int *nom_v_pos);
 
-void dc_stream_set_static_screen_events(struct dc *dc,
-					struct dc_stream_state **stream,
-					int num_streams,
-					const struct dc_static_screen_events *events);
-
 #endif /* DC_STREAM_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/dc_types.h b/drivers/gpu/drm/amd/display/dc/dc_types.h
index 9441305d3ab5..8c6eb78b0c3b 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_types.h
@@ -25,7 +25,7 @@
 #ifndef DC_TYPES_H_
 #define DC_TYPES_H_
 
-#include "fixed32_32.h"
+#include "os_types.h"
 #include "fixed31_32.h"
 #include "irq_types.h"
 #include "dc_dp_types.h"
@@ -77,8 +77,6 @@ struct dc_context {
 	struct dc *dc;
 
 	void *driver_context; /* e.g. amdgpu_device */
-
-	struct dal_logger *logger;
 	void *cgs_device;
 
 	enum dce_environment dce_environment;
@@ -92,13 +90,12 @@ struct dc_context {
 	bool created_bios;
 	struct gpio_service *gpio_service;
 	struct i2caux *i2caux;
-#if defined(CONFIG_DRM_AMD_DC_FBC)
+	uint32_t dc_sink_id_count;
 	uint64_t fbc_gpu_addr;
-#endif
 };
 
 
-#define MAX_EDID_BUFFER_SIZE 512
+#define DC_MAX_EDID_BUFFER_SIZE 512
 #define EDID_BLOCK_SIZE 128
 #define MAX_SURFACE_NUM 4
 #define NUM_PIXEL_FORMATS 10
@@ -137,13 +134,13 @@ enum plane_stereo_format {
  */
 
 enum dc_edid_connector_type {
-	EDID_CONNECTOR_UNKNOWN = 0,
-	EDID_CONNECTOR_ANALOG = 1,
-	EDID_CONNECTOR_DIGITAL = 10,
-	EDID_CONNECTOR_DVI = 11,
-	EDID_CONNECTOR_HDMIA = 12,
-	EDID_CONNECTOR_MDDI = 14,
-	EDID_CONNECTOR_DISPLAYPORT = 15
+	DC_EDID_CONNECTOR_UNKNOWN = 0,
+	DC_EDID_CONNECTOR_ANALOG = 1,
+	DC_EDID_CONNECTOR_DIGITAL = 10,
+	DC_EDID_CONNECTOR_DVI = 11,
+	DC_EDID_CONNECTOR_HDMIA = 12,
+	DC_EDID_CONNECTOR_MDDI = 14,
+	DC_EDID_CONNECTOR_DISPLAYPORT = 15
 };
 
 enum dc_edid_status {
@@ -169,7 +166,7 @@ struct dc_cea_audio_mode {
 
 struct dc_edid {
 	uint32_t length;
-	uint8_t raw_edid[MAX_EDID_BUFFER_SIZE];
+	uint8_t raw_edid[DC_MAX_EDID_BUFFER_SIZE];
 };
 
 /* When speaker location data block is not available, DEFAULT_SPEAKER_LOCATION
@@ -195,6 +192,7 @@ union display_content_support {
 
 struct dc_panel_patch {
 	unsigned int dppowerup_delay;
+	unsigned int extra_t12_ms;
 };
 
 struct dc_edid_caps {
@@ -370,12 +368,6 @@ struct dc_csc_adjustments {
 	struct fixed31_32 hue;
 };
 
-enum {
-	MAX_LANES = 2,
-	MAX_COFUNC_PATH = 6,
-	LAYER_INDEX_PRIMARY = -1,
-};
-
 enum dpcd_downstream_port_max_bpc {
 	DOWN_STREAM_MAX_8BPC = 0,
 	DOWN_STREAM_MAX_10BPC,
@@ -530,6 +522,15 @@ struct vrr_params {
 	uint32_t frame_counter;
 };
 
+struct dc_info_packet {
+	bool valid;
+	uint8_t hb0;
+	uint8_t hb1;
+	uint8_t hb2;
+	uint8_t hb3;
+	uint8_t sb[32];
+};
+
 #define DC_PLANE_UPDATE_TIMES_MAX 10
 
 struct dc_plane_flip_time {
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c b/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c
index fe92a1222803..29294db1a96b 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_abm.c
@@ -26,7 +26,7 @@
 #include "dce_abm.h"
 #include "dm_services.h"
 #include "reg_helper.h"
-#include "fixed32_32.h"
+#include "fixed31_32.h"
 #include "dc.h"
 
 #include "atom.h"
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c b/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
index 6d5cdcdc8ec9..7f6d724686f1 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
@@ -33,8 +33,9 @@
 
 #define CTX \
 	aud->base.ctx
-#define DC_LOGGER \
-	aud->base.ctx->logger
+
+#define DC_LOGGER_INIT()
+
 #define REG(reg)\
 	(aud->regs->reg)
 
@@ -348,8 +349,8 @@ static void set_audio_latency(
 
 void dce_aud_az_enable(struct audio *audio)
 {
-	struct dce_audio *aud = DCE_AUD(audio);
 	uint32_t value = AZ_REG_READ(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL);
+	DC_LOGGER_INIT();
 
 	set_reg_field_value(value, 1,
 			    AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL,
@@ -371,7 +372,7 @@ void dce_aud_az_enable(struct audio *audio)
 void dce_aud_az_disable(struct audio *audio)
 {
 	uint32_t value;
-	struct dce_audio *aud = DCE_AUD(audio);
+	DC_LOGGER_INIT();
 
 	value = AZ_REG_READ(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL);
 	set_reg_field_value(value, 1,
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
index 0aa2cda60890..439dcf3b596c 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
@@ -41,8 +41,9 @@
 
 #define CTX \
 	clk_src->base.ctx
-#define DC_LOGGER \
-	calc_pll_cs->ctx->logger
+
+#define DC_LOGGER_INIT()
+
 #undef FN
 #define FN(reg_name, field_name) \
 	clk_src->cs_shift->field_name, clk_src->cs_mask->field_name
@@ -132,7 +133,7 @@ static bool calculate_fb_and_fractional_fb_divider(
 	uint64_t feedback_divider;
 
 	feedback_divider =
-		(uint64_t)(target_pix_clk_khz * ref_divider * post_divider);
+		(uint64_t)target_pix_clk_khz * ref_divider * post_divider;
 	feedback_divider *= 10;
 	/* additional factor, since we divide by 10 afterwards */
 	feedback_divider *= (uint64_t)(calc_pll_cs->fract_fb_divider_factor);
@@ -144,8 +145,8 @@ static bool calculate_fb_and_fractional_fb_divider(
  * of fractional feedback decimal point and the fractional FB Divider precision
  * is 2 then the equation becomes (ullfeedbackDivider + 5*100) / (10*100))*/
 
-	feedback_divider += (uint64_t)
-			(5 * calc_pll_cs->fract_fb_divider_precision_factor);
+	feedback_divider += 5ULL *
+			    calc_pll_cs->fract_fb_divider_precision_factor;
 	feedback_divider =
 		div_u64(feedback_divider,
 			calc_pll_cs->fract_fb_divider_precision_factor * 10);
@@ -202,8 +203,8 @@ static bool calc_fb_divider_checking_tolerance(
 			&fract_feedback_divider);
 
 	/*Actual calculated value*/
-	actual_calc_clk_khz = (uint64_t)(feedback_divider *
-					calc_pll_cs->fract_fb_divider_factor) +
+	actual_calc_clk_khz = (uint64_t)feedback_divider *
+					calc_pll_cs->fract_fb_divider_factor +
 							fract_feedback_divider;
 	actual_calc_clk_khz *= calc_pll_cs->ref_freq_khz;
 	actual_calc_clk_khz =
@@ -467,7 +468,7 @@ static uint32_t dce110_get_pix_clk_dividers_helper (
 {
 	uint32_t field = 0;
 	uint32_t pll_calc_error = MAX_PLL_CALC_ERROR;
-	struct calc_pll_clock_source *calc_pll_cs = &clk_src->calc_pll;
+	DC_LOGGER_INIT();
 	/* Check if reference clock is external (not pcie/xtalin)
 	* HW Dce80 spec:
 	* 00 - PCIE_REFCLK, 01 - XTALIN,    02 - GENERICA,    03 - GENERICB
@@ -557,8 +558,8 @@ static uint32_t dce110_get_pix_clk_dividers(
 		struct pll_settings *pll_settings)
 {
 	struct dce110_clk_src *clk_src = TO_DCE110_CLK_SRC(cs);
-	struct calc_pll_clock_source *calc_pll_cs = &clk_src->calc_pll;
 	uint32_t pll_calc_error = MAX_PLL_CALC_ERROR;
+	DC_LOGGER_INIT();
 
 	if (pix_clk_params == NULL || pll_settings == NULL
 			|| pix_clk_params->requested_pix_clk == 0) {
@@ -589,8 +590,9 @@ static uint32_t dce110_get_pix_clk_dividers(
 			pll_settings, pix_clk_params);
 		break;
 	case DCE_VERSION_11_2:
+	case DCE_VERSION_11_22:
 	case DCE_VERSION_12_0:
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	case DCN_VERSION_1_0:
 #endif
 
@@ -655,12 +657,12 @@ static uint32_t dce110_get_d_to_pixel_rate_in_hz(
 			return 0;
 		}
 
-		pix_rate = dal_fixed31_32_from_int(clk_src->ref_freq_khz);
-		pix_rate = dal_fixed31_32_mul_int(pix_rate, 1000);
-		pix_rate = dal_fixed31_32_mul_int(pix_rate, phase);
-		pix_rate = dal_fixed31_32_div_int(pix_rate, modulo);
+		pix_rate = dc_fixpt_from_int(clk_src->ref_freq_khz);
+		pix_rate = dc_fixpt_mul_int(pix_rate, 1000);
+		pix_rate = dc_fixpt_mul_int(pix_rate, phase);
+		pix_rate = dc_fixpt_div_int(pix_rate, modulo);
 
-		return dal_fixed31_32_round(pix_rate);
+		return dc_fixpt_round(pix_rate);
 	} else {
 		return dce110_get_dp_pixel_rate_from_combo_phy_pll(cs, pix_clk_params, pll_settings);
 	}
@@ -709,12 +711,12 @@ static bool calculate_ss(
 		const struct spread_spectrum_data *ss_data,
 		struct delta_sigma_data *ds_data)
 {
-	struct fixed32_32 fb_div;
-	struct fixed32_32 ss_amount;
-	struct fixed32_32 ss_nslip_amount;
-	struct fixed32_32 ss_ds_frac_amount;
-	struct fixed32_32 ss_step_size;
-	struct fixed32_32 modulation_time;
+	struct fixed31_32 fb_div;
+	struct fixed31_32 ss_amount;
+	struct fixed31_32 ss_nslip_amount;
+	struct fixed31_32 ss_ds_frac_amount;
+	struct fixed31_32 ss_step_size;
+	struct fixed31_32 modulation_time;
 
 	if (ds_data == NULL)
 		return false;
@@ -729,42 +731,42 @@ static bool calculate_ss(
 
 	/* compute SS_AMOUNT_FBDIV & SS_AMOUNT_NFRAC_SLIP & SS_AMOUNT_DSFRAC*/
 	/* 6 decimal point support in fractional feedback divider */
-	fb_div  = dal_fixed32_32_from_fraction(
+	fb_div  = dc_fixpt_from_fraction(
 		pll_settings->fract_feedback_divider, 1000000);
-	fb_div = dal_fixed32_32_add_int(fb_div, pll_settings->feedback_divider);
+	fb_div = dc_fixpt_add_int(fb_div, pll_settings->feedback_divider);
 
 	ds_data->ds_frac_amount = 0;
 	/*spreadSpectrumPercentage is in the unit of .01%,
 	 * so have to divided by 100 * 100*/
-	ss_amount = dal_fixed32_32_mul(
-		fb_div, dal_fixed32_32_from_fraction(ss_data->percentage,
+	ss_amount = dc_fixpt_mul(
+		fb_div, dc_fixpt_from_fraction(ss_data->percentage,
 					100 * ss_data->percentage_divider));
-	ds_data->feedback_amount = dal_fixed32_32_floor(ss_amount);
+	ds_data->feedback_amount = dc_fixpt_floor(ss_amount);
 
-	ss_nslip_amount = dal_fixed32_32_sub(ss_amount,
-		dal_fixed32_32_from_int(ds_data->feedback_amount));
-	ss_nslip_amount = dal_fixed32_32_mul_int(ss_nslip_amount, 10);
-	ds_data->nfrac_amount = dal_fixed32_32_floor(ss_nslip_amount);
+	ss_nslip_amount = dc_fixpt_sub(ss_amount,
+		dc_fixpt_from_int(ds_data->feedback_amount));
+	ss_nslip_amount = dc_fixpt_mul_int(ss_nslip_amount, 10);
+	ds_data->nfrac_amount = dc_fixpt_floor(ss_nslip_amount);
 
-	ss_ds_frac_amount = dal_fixed32_32_sub(ss_nslip_amount,
-		dal_fixed32_32_from_int(ds_data->nfrac_amount));
-	ss_ds_frac_amount = dal_fixed32_32_mul_int(ss_ds_frac_amount, 65536);
-	ds_data->ds_frac_amount = dal_fixed32_32_floor(ss_ds_frac_amount);
+	ss_ds_frac_amount = dc_fixpt_sub(ss_nslip_amount,
+		dc_fixpt_from_int(ds_data->nfrac_amount));
+	ss_ds_frac_amount = dc_fixpt_mul_int(ss_ds_frac_amount, 65536);
+	ds_data->ds_frac_amount = dc_fixpt_floor(ss_ds_frac_amount);
 
 	/* compute SS_STEP_SIZE_DSFRAC */
-	modulation_time = dal_fixed32_32_from_fraction(
+	modulation_time = dc_fixpt_from_fraction(
 		pll_settings->reference_freq * 1000,
 		pll_settings->reference_divider * ss_data->modulation_freq_hz);
 
 	if (ss_data->flags.CENTER_SPREAD)
-		modulation_time = dal_fixed32_32_div_int(modulation_time, 4);
+		modulation_time = dc_fixpt_div_int(modulation_time, 4);
 	else
-		modulation_time = dal_fixed32_32_div_int(modulation_time, 2);
+		modulation_time = dc_fixpt_div_int(modulation_time, 2);
 
-	ss_step_size = dal_fixed32_32_div(ss_amount, modulation_time);
+	ss_step_size = dc_fixpt_div(ss_amount, modulation_time);
 	/* SS_STEP_SIZE_DSFRAC_DEC = Int(SS_STEP_SIZE * 2 ^ 16 * 10)*/
-	ss_step_size = dal_fixed32_32_mul_int(ss_step_size, 65536 * 10);
-	ds_data->ds_frac_size =  dal_fixed32_32_floor(ss_step_size);
+	ss_step_size = dc_fixpt_mul_int(ss_step_size, 65536 * 10);
+	ds_data->ds_frac_size =  dc_fixpt_floor(ss_step_size);
 
 	return true;
 }
@@ -907,7 +909,7 @@ static bool dce110_program_pix_clk(
 	struct dce110_clk_src *clk_src = TO_DCE110_CLK_SRC(clock_source);
 	struct bp_pixel_clock_parameters bp_pc_params = {0};
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	if (IS_FPGA_MAXIMUS_DC(clock_source->ctx->dce_environment)) {
 		unsigned int inst = pix_clk_params->controller_id - CONTROLLER_ID_D0;
 		unsigned dp_dto_ref_kHz = 700000;
@@ -978,8 +980,9 @@ static bool dce110_program_pix_clk(
 
 		break;
 	case DCE_VERSION_11_2:
+	case DCE_VERSION_11_22:
 	case DCE_VERSION_12_0:
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	case DCN_VERSION_1_0:
 #endif
 
@@ -1054,7 +1057,7 @@ static void get_ss_info_from_atombios(
 	struct spread_spectrum_info *ss_info_cur;
 	struct spread_spectrum_data *ss_data_cur;
 	uint32_t i;
-	struct calc_pll_clock_source *calc_pll_cs = &clk_src->calc_pll;
+	DC_LOGGER_INIT();
 	if (ss_entries_num == NULL) {
 		DC_LOG_SYNC(
 			"Invalid entry !!!\n");
@@ -1076,13 +1079,15 @@ static void get_ss_info_from_atombios(
 	if (*ss_entries_num == 0)
 		return;
 
-	ss_info = kzalloc(sizeof(struct spread_spectrum_info) * (*ss_entries_num),
+	ss_info = kcalloc(*ss_entries_num,
+			  sizeof(struct spread_spectrum_info),
 			  GFP_KERNEL);
 	ss_info_cur = ss_info;
 	if (ss_info == NULL)
 		return;
 
-	ss_data = kzalloc(sizeof(struct spread_spectrum_data) * (*ss_entries_num),
+	ss_data = kcalloc(*ss_entries_num,
+			  sizeof(struct spread_spectrum_data),
 			  GFP_KERNEL);
 	if (ss_data == NULL)
 		goto out_free_info;
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.h b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.h
index c45e2f76189e..801bb65707b3 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.h
@@ -55,7 +55,7 @@
 	CS_SF(PHYPLLA_PIXCLK_RESYNC_CNTL, PHYPLLA_DCCG_DEEP_COLOR_CNTL, mask_sh),\
 	CS_SF(PHYPLLA_PIXCLK_RESYNC_CNTL, PHYPLLA_PIXCLK_DOUBLE_RATE_ENABLE, mask_sh)
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 
 #define CS_COMMON_REG_LIST_DCN1_0(index, pllid) \
 		SRI(PIXCLK_RESYNC_CNTL, PHYPLL, pllid),\
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c
index 78e6beb6cf26..8f8a2abac3f3 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.c
@@ -26,19 +26,19 @@
 #include "dce_clocks.h"
 #include "dm_services.h"
 #include "reg_helper.h"
-#include "fixed32_32.h"
+#include "fixed31_32.h"
 #include "bios_parser_interface.h"
 #include "dc.h"
 #include "dmcu.h"
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 #include "dcn_calcs.h"
 #endif
 #include "core_types.h"
 #include "dc_types.h"
-
+#include "dal_asic_id.h"
 
 #define TO_DCE_CLOCKS(clocks)\
-	container_of(clocks, struct dce_disp_clk, base)
+	container_of(clocks, struct dce_dccg, base)
 
 #define REG(reg) \
 	(clk_dce->regs->reg)
@@ -101,99 +101,78 @@ static const struct state_dependent_clocks dce120_max_clks_by_state[] = {
 /*ClocksStatePerformance*/
 { .display_clk_khz = 1133000, .pixel_clk_khz = 600000 } };
 
-/* Starting point for each divider range.*/
-enum dce_divider_range_start {
-	DIVIDER_RANGE_01_START = 200, /* 2.00*/
-	DIVIDER_RANGE_02_START = 1600, /* 16.00*/
-	DIVIDER_RANGE_03_START = 3200, /* 32.00*/
-	DIVIDER_RANGE_SCALE_FACTOR = 100 /* Results are scaled up by 100.*/
-};
-
-/* Ranges for divider identifiers (Divider ID or DID)
- mmDENTIST_DISPCLK_CNTL.DENTIST_DISPCLK_WDIVIDER*/
-enum dce_divider_id_register_setting {
-	DIVIDER_RANGE_01_BASE_DIVIDER_ID = 0X08,
-	DIVIDER_RANGE_02_BASE_DIVIDER_ID = 0X40,
-	DIVIDER_RANGE_03_BASE_DIVIDER_ID = 0X60,
-	DIVIDER_RANGE_MAX_DIVIDER_ID = 0X80
+/* Starting DID for each range */
+enum dentist_base_divider_id {
+	DENTIST_BASE_DID_1 = 0x08,
+	DENTIST_BASE_DID_2 = 0x40,
+	DENTIST_BASE_DID_3 = 0x60,
+	DENTIST_MAX_DID    = 0x80
 };
 
-/* Step size between each divider within a range.
- Incrementing the DENTIST_DISPCLK_WDIVIDER by one
- will increment the divider by this much.*/
-enum dce_divider_range_step_size {
-	DIVIDER_RANGE_01_STEP_SIZE = 25, /* 0.25*/
-	DIVIDER_RANGE_02_STEP_SIZE = 50, /* 0.50*/
-	DIVIDER_RANGE_03_STEP_SIZE = 100 /* 1.00 */
+/* Starting point and step size for each divider range.*/
+enum dentist_divider_range {
+	DENTIST_DIVIDER_RANGE_1_START = 8,   /* 2.00  */
+	DENTIST_DIVIDER_RANGE_1_STEP  = 1,   /* 0.25  */
+	DENTIST_DIVIDER_RANGE_2_START = 64,  /* 16.00 */
+	DENTIST_DIVIDER_RANGE_2_STEP  = 2,   /* 0.50  */
+	DENTIST_DIVIDER_RANGE_3_START = 128, /* 32.00 */
+	DENTIST_DIVIDER_RANGE_3_STEP  = 4,   /* 1.00  */
+	DENTIST_DIVIDER_RANGE_SCALE_FACTOR = 4
 };
 
-static bool dce_divider_range_construct(
-	struct dce_divider_range *div_range,
-	int range_start,
-	int range_step,
-	int did_min,
-	int did_max)
+static int dentist_get_divider_from_did(int did)
 {
-	div_range->div_range_start = range_start;
-	div_range->div_range_step = range_step;
-	div_range->did_min = did_min;
-	div_range->did_max = did_max;
-
-	if (div_range->div_range_step == 0) {
-		div_range->div_range_step = 1;
-		/*div_range_step cannot be zero*/
-		BREAK_TO_DEBUGGER();
+	if (did < DENTIST_BASE_DID_1)
+		did = DENTIST_BASE_DID_1;
+	if (did > DENTIST_MAX_DID)
+		did = DENTIST_MAX_DID;
+
+	if (did < DENTIST_BASE_DID_2) {
+		return DENTIST_DIVIDER_RANGE_1_START + DENTIST_DIVIDER_RANGE_1_STEP
+							* (did - DENTIST_BASE_DID_1);
+	} else if (did < DENTIST_BASE_DID_3) {
+		return DENTIST_DIVIDER_RANGE_2_START + DENTIST_DIVIDER_RANGE_2_STEP
+							* (did - DENTIST_BASE_DID_2);
+	} else {
+		return DENTIST_DIVIDER_RANGE_3_START + DENTIST_DIVIDER_RANGE_3_STEP
+							* (did - DENTIST_BASE_DID_3);
 	}
-	/* Calculate this based on the other inputs.*/
-	/* See DividerRange.h for explanation of */
-	/* the relationship between divider id (DID) and a divider.*/
-	/* Number of Divider IDs = (Maximum Divider ID - Minimum Divider ID)*/
-	/* Maximum divider identified in this range =
-	 * (Number of Divider IDs)*Step size between dividers
-	 *  + The start of this range.*/
-	div_range->div_range_end = (did_max - did_min) * range_step
-		+ range_start;
-	return true;
-}
-
-static int dce_divider_range_calc_divider(
-	struct dce_divider_range *div_range,
-	int did)
-{
-	/* Is this DID within our range?*/
-	if ((did < div_range->did_min) || (did >= div_range->did_max))
-		return INVALID_DIVIDER;
-
-	return ((did - div_range->did_min) * div_range->div_range_step)
-			+ div_range->div_range_start;
-
 }
 
-static int dce_divider_range_get_divider(
-	struct dce_divider_range *div_range,
-	int ranges_num,
-	int did)
+/* SW will adjust DP REF Clock average value for all purposes
+ * (DP DTO / DP Audio DTO and DP GTC)
+ if clock is spread for all cases:
+ -if SS enabled on DP Ref clock and HW de-spreading enabled with SW
+ calculations for DS_INCR/DS_MODULO (this is planned to be default case)
+ -if SS enabled on DP Ref clock and HW de-spreading enabled with HW
+ calculations (not planned to be used, but average clock should still
+ be valid)
+ -if SS enabled on DP Ref clock and HW de-spreading disabled
+ (should not be case with CIK) then SW should program all rates
+ generated according to average value (case as with previous ASICs)
+  */
+static int dccg_adjust_dp_ref_freq_for_ss(struct dce_dccg *clk_dce, int dp_ref_clk_khz)
 {
-	int div = INVALID_DIVIDER;
-	int i;
-
-	for (i = 0; i < ranges_num; i++) {
-		/* Calculate divider with given divider ID*/
-		div = dce_divider_range_calc_divider(&div_range[i], did);
-		/* Found a valid return divider*/
-		if (div != INVALID_DIVIDER)
-			break;
+	if (clk_dce->ss_on_dprefclk && clk_dce->dprefclk_ss_divider != 0) {
+		struct fixed31_32 ss_percentage = dc_fixpt_div_int(
+				dc_fixpt_from_fraction(clk_dce->dprefclk_ss_percentage,
+							clk_dce->dprefclk_ss_divider), 200);
+		struct fixed31_32 adj_dp_ref_clk_khz;
+
+		ss_percentage = dc_fixpt_sub(dc_fixpt_one, ss_percentage);
+		adj_dp_ref_clk_khz = dc_fixpt_mul_int(ss_percentage, dp_ref_clk_khz);
+		dp_ref_clk_khz = dc_fixpt_floor(adj_dp_ref_clk_khz);
 	}
-	return div;
+	return dp_ref_clk_khz;
 }
 
-static int dce_clocks_get_dp_ref_freq(struct display_clock *clk)
+static int dce_get_dp_ref_freq_khz(struct dccg *clk)
 {
-	struct dce_disp_clk *clk_dce = TO_DCE_CLOCKS(clk);
+	struct dce_dccg *clk_dce = TO_DCE_CLOCKS(clk);
 	int dprefclk_wdivider;
 	int dprefclk_src_sel;
 	int dp_ref_clk_khz = 600000;
-	int target_div = INVALID_DIVIDER;
+	int target_div;
 
 	/* ASSERT DP Reference Clock source is from DFS*/
 	REG_GET(DPREFCLK_CNTL, DPREFCLK_SRC_SEL, &dprefclk_src_sel);
@@ -204,80 +183,27 @@ static int dce_clocks_get_dp_ref_freq(struct display_clock *clk)
 	REG_GET(DENTIST_DISPCLK_CNTL, DENTIST_DPREFCLK_WDIVIDER, &dprefclk_wdivider);
 
 	/* Convert DENTIST_DPREFCLK_WDIVIDERto actual divider*/
-	target_div = dce_divider_range_get_divider(
-			clk_dce->divider_ranges,
-			DIVIDER_RANGE_MAX,
-			dprefclk_wdivider);
-
-	if (target_div != INVALID_DIVIDER) {
-		/* Calculate the current DFS clock, in kHz.*/
-		dp_ref_clk_khz = (DIVIDER_RANGE_SCALE_FACTOR
-			* clk_dce->dentist_vco_freq_khz) / target_div;
-	}
+	target_div = dentist_get_divider_from_did(dprefclk_wdivider);
 
-	/* SW will adjust DP REF Clock average value for all purposes
-	 * (DP DTO / DP Audio DTO and DP GTC)
-	 if clock is spread for all cases:
-	 -if SS enabled on DP Ref clock and HW de-spreading enabled with SW
-	 calculations for DS_INCR/DS_MODULO (this is planned to be default case)
-	 -if SS enabled on DP Ref clock and HW de-spreading enabled with HW
-	 calculations (not planned to be used, but average clock should still
-	 be valid)
-	 -if SS enabled on DP Ref clock and HW de-spreading disabled
-	 (should not be case with CIK) then SW should program all rates
-	 generated according to average value (case as with previous ASICs)
-	  */
-	if (clk_dce->ss_on_dprefclk && clk_dce->dprefclk_ss_divider != 0) {
-		struct fixed32_32 ss_percentage = dal_fixed32_32_div_int(
-				dal_fixed32_32_from_fraction(
-						clk_dce->dprefclk_ss_percentage,
-						clk_dce->dprefclk_ss_divider), 200);
-		struct fixed32_32 adj_dp_ref_clk_khz;
-
-		ss_percentage = dal_fixed32_32_sub(dal_fixed32_32_one,
-								ss_percentage);
-		adj_dp_ref_clk_khz =
-			dal_fixed32_32_mul_int(
-				ss_percentage,
-				dp_ref_clk_khz);
-		dp_ref_clk_khz = dal_fixed32_32_floor(adj_dp_ref_clk_khz);
-	}
+	/* Calculate the current DFS clock, in kHz.*/
+	dp_ref_clk_khz = (DENTIST_DIVIDER_RANGE_SCALE_FACTOR
+		* clk_dce->dentist_vco_freq_khz) / target_div;
 
-	return dp_ref_clk_khz;
+	return dccg_adjust_dp_ref_freq_for_ss(clk_dce, dp_ref_clk_khz);
 }
 
-/* TODO: This is DCN DPREFCLK: it could be program by DENTIST by VBIOS
- * or CLK0_CLK11 by SMU. For DCE120, it is wlays 600Mhz. Will re-visit
- * clock implementation
- */
-static int dce_clocks_get_dp_ref_freq_wrkaround(struct display_clock *clk)
+static int dce12_get_dp_ref_freq_khz(struct dccg *clk)
 {
-	struct dce_disp_clk *clk_dce = TO_DCE_CLOCKS(clk);
-	int dp_ref_clk_khz = 600000;
-
-	if (clk_dce->ss_on_dprefclk && clk_dce->dprefclk_ss_divider != 0) {
-		struct fixed32_32 ss_percentage = dal_fixed32_32_div_int(
-				dal_fixed32_32_from_fraction(
-						clk_dce->dprefclk_ss_percentage,
-						clk_dce->dprefclk_ss_divider), 200);
-		struct fixed32_32 adj_dp_ref_clk_khz;
-
-		ss_percentage = dal_fixed32_32_sub(dal_fixed32_32_one,
-								ss_percentage);
-		adj_dp_ref_clk_khz =
-			dal_fixed32_32_mul_int(
-				ss_percentage,
-				dp_ref_clk_khz);
-		dp_ref_clk_khz = dal_fixed32_32_floor(adj_dp_ref_clk_khz);
-	}
+	struct dce_dccg *clk_dce = TO_DCE_CLOCKS(clk);
 
-	return dp_ref_clk_khz;
+	return dccg_adjust_dp_ref_freq_for_ss(clk_dce, 600000);
 }
+
 static enum dm_pp_clocks_state dce_get_required_clocks_state(
-	struct display_clock *clk,
-	struct state_dependent_clocks *req_clocks)
+	struct dccg *clk,
+	struct dc_clocks *req_clocks)
 {
-	struct dce_disp_clk *clk_dce = TO_DCE_CLOCKS(clk);
+	struct dce_dccg *clk_dce = TO_DCE_CLOCKS(clk);
 	int i;
 	enum dm_pp_clocks_state low_req_clk;
 
@@ -286,53 +212,30 @@ static enum dm_pp_clocks_state dce_get_required_clocks_state(
 	 * all required clocks
 	 */
 	for (i = clk->max_clks_state; i >= DM_PP_CLOCKS_STATE_ULTRA_LOW; i--)
-		if (req_clocks->display_clk_khz >
+		if (req_clocks->dispclk_khz >
 				clk_dce->max_clks_by_state[i].display_clk_khz
-			|| req_clocks->pixel_clk_khz >
+			|| req_clocks->phyclk_khz >
 				clk_dce->max_clks_by_state[i].pixel_clk_khz)
 			break;
 
 	low_req_clk = i + 1;
 	if (low_req_clk > clk->max_clks_state) {
-		DC_LOG_WARNING("%s: clocks unsupported disp_clk %d pix_clk %d",
-				__func__,
-				req_clocks->display_clk_khz,
-				req_clocks->pixel_clk_khz);
-		low_req_clk = DM_PP_CLOCKS_STATE_INVALID;
+		/* set max clock state for high phyclock, invalid on exceeding display clock */
+		if (clk_dce->max_clks_by_state[clk->max_clks_state].display_clk_khz
+				< req_clocks->dispclk_khz)
+			low_req_clk = DM_PP_CLOCKS_STATE_INVALID;
+		else
+			low_req_clk = clk->max_clks_state;
 	}
 
 	return low_req_clk;
 }
 
-static bool dce_clock_set_min_clocks_state(
-	struct display_clock *clk,
-	enum dm_pp_clocks_state clocks_state)
-{
-	struct dm_pp_power_level_change_request level_change_req = {
-			clocks_state };
-
-	if (clocks_state > clk->max_clks_state) {
-		/*Requested state exceeds max supported state.*/
-		DC_LOG_WARNING("Requested state exceeds max supported state");
-		return false;
-	} else if (clocks_state == clk->cur_min_clks_state) {
-		/*if we're trying to set the same state, we can just return
-		 * since nothing needs to be done*/
-		return true;
-	}
-
-	/* get max clock state from PPLIB */
-	if (dm_pp_apply_power_level_change_request(clk->ctx, &level_change_req))
-		clk->cur_min_clks_state = clocks_state;
-
-	return true;
-}
-
 static int dce_set_clock(
-	struct display_clock *clk,
+	struct dccg *clk,
 	int requested_clk_khz)
 {
-	struct dce_disp_clk *clk_dce = TO_DCE_CLOCKS(clk);
+	struct dce_dccg *clk_dce = TO_DCE_CLOCKS(clk);
 	struct bp_pixel_clock_parameters pxl_clk_params = { 0 };
 	struct dc_bios *bp = clk->ctx->dc_bios;
 	int actual_clock = requested_clk_khz;
@@ -364,10 +267,10 @@ static int dce_set_clock(
 }
 
 static int dce_psr_set_clock(
-	struct display_clock *clk,
+	struct dccg *clk,
 	int requested_clk_khz)
 {
-	struct dce_disp_clk *clk_dce = TO_DCE_CLOCKS(clk);
+	struct dce_dccg *clk_dce = TO_DCE_CLOCKS(clk);
 	struct dc_context *ctx = clk_dce->base.ctx;
 	struct dc *core_dc = ctx->dc;
 	struct dmcu *dmcu = core_dc->res_pool->dmcu;
@@ -380,10 +283,10 @@ static int dce_psr_set_clock(
 }
 
 static int dce112_set_clock(
-	struct display_clock *clk,
+	struct dccg *clk,
 	int requested_clk_khz)
 {
-	struct dce_disp_clk *clk_dce = TO_DCE_CLOCKS(clk);
+	struct dce_dccg *clk_dce = TO_DCE_CLOCKS(clk);
 	struct bp_set_dce_clock_parameters dce_clk_params;
 	struct dc_bios *bp = clk->ctx->dc_bios;
 	struct dc *core_dc = clk->ctx->dc;
@@ -413,9 +316,12 @@ static int dce112_set_clock(
 	/*VBIOS will determine DPREFCLK frequency, so we don't set it*/
 	dce_clk_params.target_clock_frequency = 0;
 	dce_clk_params.clock_type = DCECLOCK_TYPE_DPREFCLK;
-	dce_clk_params.flags.USE_GENLOCK_AS_SOURCE_FOR_DPREFCLK =
+	if (!ASICREV_IS_VEGA20_P(clk->ctx->asic_id.hw_internal_rev))
+		dce_clk_params.flags.USE_GENLOCK_AS_SOURCE_FOR_DPREFCLK =
 			(dce_clk_params.pll_id ==
 					CLOCK_SOURCE_COMBO_DISPLAY_PLL0);
+	else
+		dce_clk_params.flags.USE_GENLOCK_AS_SOURCE_FOR_DPREFCLK = false;
 
 	bp->funcs->set_dce_clock(bp, &dce_clk_params);
 
@@ -429,7 +335,7 @@ static int dce112_set_clock(
 	return actual_clock;
 }
 
-static void dce_clock_read_integrated_info(struct dce_disp_clk *clk_dce)
+static void dce_clock_read_integrated_info(struct dce_dccg *clk_dce)
 {
 	struct dc_debug *debug = &clk_dce->base.ctx->dc->debug;
 	struct dc_bios *bp = clk_dce->base.ctx->dc_bios;
@@ -485,11 +391,9 @@ static void dce_clock_read_integrated_info(struct dce_disp_clk *clk_dce)
 	if (!debug->disable_dfs_bypass && bp->integrated_info)
 		if (bp->integrated_info->gpu_cap_info & DFS_BYPASS_ENABLE)
 			clk_dce->dfs_bypass_enabled = true;
-
-	clk_dce->use_max_disp_clk = debug->max_disp_clk;
 }
 
-static void dce_clock_read_ss_info(struct dce_disp_clk *clk_dce)
+static void dce_clock_read_ss_info(struct dce_dccg *clk_dce)
 {
 	struct dc_bios *bp = clk_dce->base.ctx->dc_bios;
 	int ss_info_num = bp->funcs->get_ss_entry_number(
@@ -545,139 +449,263 @@ static void dce_clock_read_ss_info(struct dce_disp_clk *clk_dce)
 	}
 }
 
-static bool dce_apply_clock_voltage_request(
-	struct display_clock *clk,
-	enum dm_pp_clock_type clocks_type,
-	int clocks_in_khz,
-	bool pre_mode_set,
-	bool update_dp_phyclk)
+static inline bool should_set_clock(bool safe_to_lower, int calc_clk, int cur_clk)
+{
+	return ((safe_to_lower && calc_clk < cur_clk) || calc_clk > cur_clk);
+}
+
+static void dce12_update_clocks(struct dccg *dccg,
+			struct dc_clocks *new_clocks,
+			bool safe_to_lower)
 {
-	bool send_request = false;
 	struct dm_pp_clock_for_voltage_req clock_voltage_req = {0};
 
-	switch (clocks_type) {
-	case DM_PP_CLOCK_TYPE_DISPLAY_CLK:
-	case DM_PP_CLOCK_TYPE_PIXELCLK:
-	case DM_PP_CLOCK_TYPE_DISPLAYPHYCLK:
-		break;
-	default:
-		BREAK_TO_DEBUGGER();
-		return false;
+	if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, dccg->clks.dispclk_khz)) {
+		clock_voltage_req.clk_type = DM_PP_CLOCK_TYPE_DISPLAY_CLK;
+		clock_voltage_req.clocks_in_khz = new_clocks->dispclk_khz;
+		dccg->funcs->set_dispclk(dccg, new_clocks->dispclk_khz);
+		dccg->clks.dispclk_khz = new_clocks->dispclk_khz;
+
+		dm_pp_apply_clock_for_voltage_request(dccg->ctx, &clock_voltage_req);
 	}
 
-	clock_voltage_req.clk_type = clocks_type;
-	clock_voltage_req.clocks_in_khz = clocks_in_khz;
-
-	/* to pplib */
-	if (pre_mode_set) {
-		switch (clocks_type) {
-		case DM_PP_CLOCK_TYPE_DISPLAY_CLK:
-			if (clocks_in_khz > clk->cur_clocks_value.dispclk_in_khz) {
-				clk->cur_clocks_value.dispclk_notify_pplib_done = true;
-				send_request = true;
-			} else
-				clk->cur_clocks_value.dispclk_notify_pplib_done = false;
-			/* no matter incrase or decrase clock, update current clock value */
-			clk->cur_clocks_value.dispclk_in_khz = clocks_in_khz;
-			break;
-		case DM_PP_CLOCK_TYPE_PIXELCLK:
-			if (clocks_in_khz > clk->cur_clocks_value.max_pixelclk_in_khz) {
-				clk->cur_clocks_value.pixelclk_notify_pplib_done = true;
-				send_request = true;
-			} else
-				clk->cur_clocks_value.pixelclk_notify_pplib_done = false;
-			/* no matter incrase or decrase clock, update current clock value */
-			clk->cur_clocks_value.max_pixelclk_in_khz = clocks_in_khz;
-			break;
-		case DM_PP_CLOCK_TYPE_DISPLAYPHYCLK:
-			if (clocks_in_khz > clk->cur_clocks_value.max_non_dp_phyclk_in_khz) {
-				clk->cur_clocks_value.phyclk_notigy_pplib_done = true;
-				send_request = true;
-			} else
-				clk->cur_clocks_value.phyclk_notigy_pplib_done = false;
-			/* no matter incrase or decrase clock, update current clock value */
-			clk->cur_clocks_value.max_non_dp_phyclk_in_khz = clocks_in_khz;
-			break;
-		default:
-			ASSERT(0);
-			break;
-		}
+	if (should_set_clock(safe_to_lower, new_clocks->phyclk_khz, dccg->clks.phyclk_khz)) {
+		clock_voltage_req.clk_type = DM_PP_CLOCK_TYPE_DISPLAYPHYCLK;
+		clock_voltage_req.clocks_in_khz = new_clocks->phyclk_khz;
+		dccg->clks.phyclk_khz = new_clocks->phyclk_khz;
+
+		dm_pp_apply_clock_for_voltage_request(dccg->ctx, &clock_voltage_req);
+	}
+}
+
+#ifdef CONFIG_X86
+static int dcn1_determine_dppclk_threshold(struct dccg *dccg, struct dc_clocks *new_clocks)
+{
+	bool request_dpp_div = new_clocks->dispclk_khz > new_clocks->dppclk_khz;
+	bool dispclk_increase = new_clocks->dispclk_khz > dccg->clks.dispclk_khz;
+	int disp_clk_threshold = new_clocks->max_supported_dppclk_khz;
+	bool cur_dpp_div = dccg->clks.dispclk_khz > dccg->clks.dppclk_khz;
+
+	/* increase clock, looking for div is 0 for current, request div is 1*/
+	if (dispclk_increase) {
+		/* already divided by 2, no need to reach target clk with 2 steps*/
+		if (cur_dpp_div)
+			return new_clocks->dispclk_khz;
+
+		/* request disp clk is lower than maximum supported dpp clk,
+		 * no need to reach target clk with two steps.
+		 */
+		if (new_clocks->dispclk_khz <= disp_clk_threshold)
+			return new_clocks->dispclk_khz;
+
+		/* target dpp clk not request divided by 2, still within threshold */
+		if (!request_dpp_div)
+			return new_clocks->dispclk_khz;
 
 	} else {
-		switch (clocks_type) {
-		case DM_PP_CLOCK_TYPE_DISPLAY_CLK:
-			if (!clk->cur_clocks_value.dispclk_notify_pplib_done)
-				send_request = true;
-			break;
-		case DM_PP_CLOCK_TYPE_PIXELCLK:
-			if (!clk->cur_clocks_value.pixelclk_notify_pplib_done)
-				send_request = true;
-			break;
-		case DM_PP_CLOCK_TYPE_DISPLAYPHYCLK:
-			if (!clk->cur_clocks_value.phyclk_notigy_pplib_done)
-				send_request = true;
-			break;
-		default:
-			ASSERT(0);
-			break;
-		}
+		/* decrease clock, looking for current dppclk divided by 2,
+		 * request dppclk not divided by 2.
+		 */
+
+		/* current dpp clk not divided by 2, no need to ramp*/
+		if (!cur_dpp_div)
+			return new_clocks->dispclk_khz;
+
+		/* current disp clk is lower than current maximum dpp clk,
+		 * no need to ramp
+		 */
+		if (dccg->clks.dispclk_khz <= disp_clk_threshold)
+			return new_clocks->dispclk_khz;
+
+		/* request dpp clk need to be divided by 2 */
+		if (request_dpp_div)
+			return new_clocks->dispclk_khz;
 	}
-	if (send_request) {
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
-		if (clk->ctx->dce_version >= DCN_VERSION_1_0) {
-			struct dc *core_dc = clk->ctx->dc;
-			/*use dcfclk request voltage*/
-			clock_voltage_req.clk_type = DM_PP_CLOCK_TYPE_DCFCLK;
-			clock_voltage_req.clocks_in_khz =
-				dcn_find_dcfclk_suits_all(core_dc, &clk->cur_clocks_value);
-		}
+
+	return disp_clk_threshold;
+}
+
+static void dcn1_ramp_up_dispclk_with_dpp(struct dccg *dccg, struct dc_clocks *new_clocks)
+{
+	struct dc *dc = dccg->ctx->dc;
+	int dispclk_to_dpp_threshold = dcn1_determine_dppclk_threshold(dccg, new_clocks);
+	bool request_dpp_div = new_clocks->dispclk_khz > new_clocks->dppclk_khz;
+	int i;
+
+	/* set disp clk to dpp clk threshold */
+	dccg->funcs->set_dispclk(dccg, dispclk_to_dpp_threshold);
+
+	/* update request dpp clk division option */
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i];
+
+		if (!pipe_ctx->plane_state)
+			continue;
+
+		pipe_ctx->plane_res.dpp->funcs->dpp_dppclk_control(
+				pipe_ctx->plane_res.dpp,
+				request_dpp_div,
+				true);
+	}
+
+	/* If target clk not same as dppclk threshold, set to target clock */
+	if (dispclk_to_dpp_threshold != new_clocks->dispclk_khz)
+		dccg->funcs->set_dispclk(dccg, new_clocks->dispclk_khz);
+
+	dccg->clks.dispclk_khz = new_clocks->dispclk_khz;
+	dccg->clks.dppclk_khz = new_clocks->dppclk_khz;
+	dccg->clks.max_supported_dppclk_khz = new_clocks->max_supported_dppclk_khz;
+}
+
+static void dcn1_update_clocks(struct dccg *dccg,
+			struct dc_clocks *new_clocks,
+			bool safe_to_lower)
+{
+	struct dc *dc = dccg->ctx->dc;
+	struct pp_smu_display_requirement_rv *smu_req_cur =
+			&dc->res_pool->pp_smu_req;
+	struct pp_smu_display_requirement_rv smu_req = *smu_req_cur;
+	struct pp_smu_funcs_rv *pp_smu = dc->res_pool->pp_smu;
+	struct dm_pp_clock_for_voltage_req clock_voltage_req = {0};
+	bool send_request_to_increase = false;
+	bool send_request_to_lower = false;
+
+	if (new_clocks->phyclk_khz)
+		smu_req.display_count = 1;
+	else
+		smu_req.display_count = 0;
+
+	if (new_clocks->dispclk_khz > dccg->clks.dispclk_khz
+			|| new_clocks->phyclk_khz > dccg->clks.phyclk_khz
+			|| new_clocks->fclk_khz > dccg->clks.fclk_khz
+			|| new_clocks->dcfclk_khz > dccg->clks.dcfclk_khz)
+		send_request_to_increase = true;
+
+	if (should_set_clock(safe_to_lower, new_clocks->phyclk_khz, dccg->clks.phyclk_khz)) {
+		dccg->clks.phyclk_khz = new_clocks->phyclk_khz;
+
+		send_request_to_lower = true;
+	}
+
+	if (should_set_clock(safe_to_lower, new_clocks->fclk_khz, dccg->clks.fclk_khz)) {
+		dccg->clks.fclk_khz = new_clocks->fclk_khz;
+		clock_voltage_req.clk_type = DM_PP_CLOCK_TYPE_FCLK;
+		clock_voltage_req.clocks_in_khz = new_clocks->fclk_khz;
+		smu_req.hard_min_fclk_khz = new_clocks->fclk_khz;
+
+		dm_pp_apply_clock_for_voltage_request(dccg->ctx, &clock_voltage_req);
+		send_request_to_lower = true;
+	}
+
+	if (should_set_clock(safe_to_lower, new_clocks->dcfclk_khz, dccg->clks.dcfclk_khz)) {
+		dccg->clks.dcfclk_khz = new_clocks->dcfclk_khz;
+		smu_req.hard_min_dcefclk_khz = new_clocks->dcfclk_khz;
+
+		send_request_to_lower = true;
+	}
+
+	if (should_set_clock(safe_to_lower,
+			new_clocks->dcfclk_deep_sleep_khz, dccg->clks.dcfclk_deep_sleep_khz)) {
+		dccg->clks.dcfclk_deep_sleep_khz = new_clocks->dcfclk_deep_sleep_khz;
+		smu_req.min_deep_sleep_dcefclk_mhz = new_clocks->dcfclk_deep_sleep_khz;
+
+		send_request_to_lower = true;
+	}
+
+	/* make sure dcf clk is before dpp clk to
+	 * make sure we have enough voltage to run dpp clk
+	 */
+	if (send_request_to_increase) {
+		/*use dcfclk to request voltage*/
+		clock_voltage_req.clk_type = DM_PP_CLOCK_TYPE_DCFCLK;
+		clock_voltage_req.clocks_in_khz = dcn_find_dcfclk_suits_all(dc, new_clocks);
+		dm_pp_apply_clock_for_voltage_request(dccg->ctx, &clock_voltage_req);
+		if (pp_smu->set_display_requirement)
+			pp_smu->set_display_requirement(&pp_smu->pp_smu, &smu_req);
+	}
+
+	/* dcn1 dppclk is tied to dispclk */
+	if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, dccg->clks.dispclk_khz)) {
+		dcn1_ramp_up_dispclk_with_dpp(dccg, new_clocks);
+		dccg->clks.dispclk_khz = new_clocks->dispclk_khz;
+
+		send_request_to_lower = true;
+	}
+
+	if (!send_request_to_increase && send_request_to_lower) {
+		/*use dcfclk to request voltage*/
+		clock_voltage_req.clk_type = DM_PP_CLOCK_TYPE_DCFCLK;
+		clock_voltage_req.clocks_in_khz = dcn_find_dcfclk_suits_all(dc, new_clocks);
+		dm_pp_apply_clock_for_voltage_request(dccg->ctx, &clock_voltage_req);
+		if (pp_smu->set_display_requirement)
+			pp_smu->set_display_requirement(&pp_smu->pp_smu, &smu_req);
+	}
+
+
+	*smu_req_cur = smu_req;
+}
 #endif
-		dm_pp_apply_clock_for_voltage_request(
-			clk->ctx, &clock_voltage_req);
+
+static void dce_update_clocks(struct dccg *dccg,
+			struct dc_clocks *new_clocks,
+			bool safe_to_lower)
+{
+	struct dm_pp_power_level_change_request level_change_req;
+
+	level_change_req.power_level = dce_get_required_clocks_state(dccg, new_clocks);
+	/* get max clock state from PPLIB */
+	if ((level_change_req.power_level < dccg->cur_min_clks_state && safe_to_lower)
+			|| level_change_req.power_level > dccg->cur_min_clks_state) {
+		if (dm_pp_apply_power_level_change_request(dccg->ctx, &level_change_req))
+			dccg->cur_min_clks_state = level_change_req.power_level;
 	}
-	if (update_dp_phyclk && (clocks_in_khz >
-	clk->cur_clocks_value.max_dp_phyclk_in_khz))
-		clk->cur_clocks_value.max_dp_phyclk_in_khz = clocks_in_khz;
 
-	return true;
+	if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, dccg->clks.dispclk_khz)) {
+		dccg->funcs->set_dispclk(dccg, new_clocks->dispclk_khz);
+		dccg->clks.dispclk_khz = new_clocks->dispclk_khz;
+	}
 }
 
+#ifdef CONFIG_X86
+static const struct display_clock_funcs dcn1_funcs = {
+	.get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz,
+	.set_dispclk = dce112_set_clock,
+	.update_clocks = dcn1_update_clocks
+};
+#endif
 
 static const struct display_clock_funcs dce120_funcs = {
-	.get_dp_ref_clk_frequency = dce_clocks_get_dp_ref_freq_wrkaround,
-	.apply_clock_voltage_request = dce_apply_clock_voltage_request,
-	.set_clock = dce112_set_clock
+	.get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz,
+	.set_dispclk = dce112_set_clock,
+	.update_clocks = dce12_update_clocks
 };
 
 static const struct display_clock_funcs dce112_funcs = {
-	.get_dp_ref_clk_frequency = dce_clocks_get_dp_ref_freq,
-	.get_required_clocks_state = dce_get_required_clocks_state,
-	.set_min_clocks_state = dce_clock_set_min_clocks_state,
-	.set_clock = dce112_set_clock
+	.get_dp_ref_clk_frequency = dce_get_dp_ref_freq_khz,
+	.set_dispclk = dce112_set_clock,
+	.update_clocks = dce_update_clocks
 };
 
 static const struct display_clock_funcs dce110_funcs = {
-	.get_dp_ref_clk_frequency = dce_clocks_get_dp_ref_freq,
-	.get_required_clocks_state = dce_get_required_clocks_state,
-	.set_min_clocks_state = dce_clock_set_min_clocks_state,
-	.set_clock = dce_psr_set_clock
+	.get_dp_ref_clk_frequency = dce_get_dp_ref_freq_khz,
+	.set_dispclk = dce_psr_set_clock,
+	.update_clocks = dce_update_clocks
 };
 
 static const struct display_clock_funcs dce_funcs = {
-	.get_dp_ref_clk_frequency = dce_clocks_get_dp_ref_freq,
-	.get_required_clocks_state = dce_get_required_clocks_state,
-	.set_min_clocks_state = dce_clock_set_min_clocks_state,
-	.set_clock = dce_set_clock
+	.get_dp_ref_clk_frequency = dce_get_dp_ref_freq_khz,
+	.set_dispclk = dce_set_clock,
+	.update_clocks = dce_update_clocks
 };
 
-static void dce_disp_clk_construct(
-	struct dce_disp_clk *clk_dce,
+static void dce_dccg_construct(
+	struct dce_dccg *clk_dce,
 	struct dc_context *ctx,
-	const struct dce_disp_clk_registers *regs,
-	const struct dce_disp_clk_shift *clk_shift,
-	const struct dce_disp_clk_mask *clk_mask)
+	const struct dccg_registers *regs,
+	const struct dccg_shift *clk_shift,
+	const struct dccg_mask *clk_mask)
 {
-	struct display_clock *base = &clk_dce->base;
+	struct dccg *base = &clk_dce->base;
 
 	base->ctx = ctx;
 	base->funcs = &dce_funcs;
@@ -697,34 +725,15 @@ static void dce_disp_clk_construct(
 
 	dce_clock_read_integrated_info(clk_dce);
 	dce_clock_read_ss_info(clk_dce);
-
-	dce_divider_range_construct(
-		&clk_dce->divider_ranges[DIVIDER_RANGE_01],
-		DIVIDER_RANGE_01_START,
-		DIVIDER_RANGE_01_STEP_SIZE,
-		DIVIDER_RANGE_01_BASE_DIVIDER_ID,
-		DIVIDER_RANGE_02_BASE_DIVIDER_ID);
-	dce_divider_range_construct(
-		&clk_dce->divider_ranges[DIVIDER_RANGE_02],
-		DIVIDER_RANGE_02_START,
-		DIVIDER_RANGE_02_STEP_SIZE,
-		DIVIDER_RANGE_02_BASE_DIVIDER_ID,
-		DIVIDER_RANGE_03_BASE_DIVIDER_ID);
-	dce_divider_range_construct(
-		&clk_dce->divider_ranges[DIVIDER_RANGE_03],
-		DIVIDER_RANGE_03_START,
-		DIVIDER_RANGE_03_STEP_SIZE,
-		DIVIDER_RANGE_03_BASE_DIVIDER_ID,
-		DIVIDER_RANGE_MAX_DIVIDER_ID);
 }
 
-struct display_clock *dce_disp_clk_create(
+struct dccg *dce_dccg_create(
 	struct dc_context *ctx,
-	const struct dce_disp_clk_registers *regs,
-	const struct dce_disp_clk_shift *clk_shift,
-	const struct dce_disp_clk_mask *clk_mask)
+	const struct dccg_registers *regs,
+	const struct dccg_shift *clk_shift,
+	const struct dccg_mask *clk_mask)
 {
-	struct dce_disp_clk *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL);
+	struct dce_dccg *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL);
 
 	if (clk_dce == NULL) {
 		BREAK_TO_DEBUGGER();
@@ -735,19 +744,19 @@ struct display_clock *dce_disp_clk_create(
 		dce80_max_clks_by_state,
 		sizeof(dce80_max_clks_by_state));
 
-	dce_disp_clk_construct(
+	dce_dccg_construct(
 		clk_dce, ctx, regs, clk_shift, clk_mask);
 
 	return &clk_dce->base;
 }
 
-struct display_clock *dce110_disp_clk_create(
+struct dccg *dce110_dccg_create(
 	struct dc_context *ctx,
-	const struct dce_disp_clk_registers *regs,
-	const struct dce_disp_clk_shift *clk_shift,
-	const struct dce_disp_clk_mask *clk_mask)
+	const struct dccg_registers *regs,
+	const struct dccg_shift *clk_shift,
+	const struct dccg_mask *clk_mask)
 {
-	struct dce_disp_clk *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL);
+	struct dce_dccg *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL);
 
 	if (clk_dce == NULL) {
 		BREAK_TO_DEBUGGER();
@@ -758,7 +767,7 @@ struct display_clock *dce110_disp_clk_create(
 		dce110_max_clks_by_state,
 		sizeof(dce110_max_clks_by_state));
 
-	dce_disp_clk_construct(
+	dce_dccg_construct(
 		clk_dce, ctx, regs, clk_shift, clk_mask);
 
 	clk_dce->base.funcs = &dce110_funcs;
@@ -766,13 +775,13 @@ struct display_clock *dce110_disp_clk_create(
 	return &clk_dce->base;
 }
 
-struct display_clock *dce112_disp_clk_create(
+struct dccg *dce112_dccg_create(
 	struct dc_context *ctx,
-	const struct dce_disp_clk_registers *regs,
-	const struct dce_disp_clk_shift *clk_shift,
-	const struct dce_disp_clk_mask *clk_mask)
+	const struct dccg_registers *regs,
+	const struct dccg_shift *clk_shift,
+	const struct dccg_mask *clk_mask)
 {
-	struct dce_disp_clk *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL);
+	struct dce_dccg *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL);
 
 	if (clk_dce == NULL) {
 		BREAK_TO_DEBUGGER();
@@ -783,7 +792,7 @@ struct display_clock *dce112_disp_clk_create(
 		dce112_max_clks_by_state,
 		sizeof(dce112_max_clks_by_state));
 
-	dce_disp_clk_construct(
+	dce_dccg_construct(
 		clk_dce, ctx, regs, clk_shift, clk_mask);
 
 	clk_dce->base.funcs = &dce112_funcs;
@@ -791,10 +800,9 @@ struct display_clock *dce112_disp_clk_create(
 	return &clk_dce->base;
 }
 
-struct display_clock *dce120_disp_clk_create(struct dc_context *ctx)
+struct dccg *dce120_dccg_create(struct dc_context *ctx)
 {
-	struct dce_disp_clk *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL);
-	struct dm_pp_clock_levels_with_voltage clk_level_info = {0};
+	struct dce_dccg *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL);
 
 	if (clk_dce == NULL) {
 		BREAK_TO_DEBUGGER();
@@ -805,28 +813,59 @@ struct display_clock *dce120_disp_clk_create(struct dc_context *ctx)
 		dce120_max_clks_by_state,
 		sizeof(dce120_max_clks_by_state));
 
-	dce_disp_clk_construct(
+	dce_dccg_construct(
 		clk_dce, ctx, NULL, NULL, NULL);
 
 	clk_dce->base.funcs = &dce120_funcs;
 
-	/* new in dce120 */
-	if (!ctx->dc->debug.disable_pplib_clock_request  &&
-			dm_pp_get_clock_levels_by_type_with_voltage(
-			ctx, DM_PP_CLOCK_TYPE_DISPLAY_CLK, &clk_level_info)
-						&& clk_level_info.num_levels)
-		clk_dce->max_displ_clk_in_khz =
-			clk_level_info.data[clk_level_info.num_levels - 1].clocks_in_khz;
-	else
-		clk_dce->max_displ_clk_in_khz = 1133000;
+	return &clk_dce->base;
+}
+
+#ifdef CONFIG_X86
+struct dccg *dcn1_dccg_create(struct dc_context *ctx)
+{
+	struct dc_debug *debug = &ctx->dc->debug;
+	struct dc_bios *bp = ctx->dc_bios;
+	struct dc_firmware_info fw_info = { { 0 } };
+	struct dce_dccg *clk_dce = kzalloc(sizeof(*clk_dce), GFP_KERNEL);
+
+	if (clk_dce == NULL) {
+		BREAK_TO_DEBUGGER();
+		return NULL;
+	}
+
+	clk_dce->base.ctx = ctx;
+	clk_dce->base.funcs = &dcn1_funcs;
+
+	clk_dce->dfs_bypass_disp_clk = 0;
+
+	clk_dce->dprefclk_ss_percentage = 0;
+	clk_dce->dprefclk_ss_divider = 1000;
+	clk_dce->ss_on_dprefclk = false;
+
+	if (bp->integrated_info)
+		clk_dce->dentist_vco_freq_khz = bp->integrated_info->dentist_vco_freq;
+	if (clk_dce->dentist_vco_freq_khz == 0) {
+		bp->funcs->get_firmware_info(bp, &fw_info);
+		clk_dce->dentist_vco_freq_khz = fw_info.smu_gpu_pll_output_freq;
+		if (clk_dce->dentist_vco_freq_khz == 0)
+			clk_dce->dentist_vco_freq_khz = 3600000;
+	}
+
+	if (!debug->disable_dfs_bypass && bp->integrated_info)
+		if (bp->integrated_info->gpu_cap_info & DFS_BYPASS_ENABLE)
+			clk_dce->dfs_bypass_enabled = true;
+
+	dce_clock_read_ss_info(clk_dce);
 
 	return &clk_dce->base;
 }
+#endif
 
-void dce_disp_clk_destroy(struct display_clock **disp_clk)
+void dce_dccg_destroy(struct dccg **dccg)
 {
-	struct dce_disp_clk *clk_dce = TO_DCE_CLOCKS(*disp_clk);
+	struct dce_dccg *clk_dce = TO_DCE_CLOCKS(*dccg);
 
 	kfree(clk_dce);
-	*disp_clk = NULL;
+	*dccg = NULL;
 }
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.h b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.h
index 0e717e0dc8f0..e5e44adc6c27 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clocks.h
@@ -33,6 +33,9 @@
 	.DPREFCLK_CNTL = mmDPREFCLK_CNTL, \
 	.DENTIST_DISPCLK_CNTL = mmDENTIST_DISPCLK_CNTL
 
+#define CLK_COMMON_REG_LIST_DCN_BASE() \
+	SR(DENTIST_DISPCLK_CNTL)
+
 #define CLK_SF(reg_name, field_name, post_fix)\
 	.field_name = reg_name ## __ ## field_name ## post_fix
 
@@ -40,58 +43,37 @@
 	CLK_SF(DPREFCLK_CNTL, DPREFCLK_SRC_SEL, mask_sh), \
 	CLK_SF(DENTIST_DISPCLK_CNTL, DENTIST_DPREFCLK_WDIVIDER, mask_sh)
 
+#define CLK_COMMON_MASK_SH_LIST_DCN_COMMON_BASE(mask_sh) \
+	CLK_SF(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_WDIVIDER, mask_sh),\
+	CLK_SF(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_CHG_DONE, mask_sh)
+
 #define CLK_REG_FIELD_LIST(type) \
 	type DPREFCLK_SRC_SEL; \
-	type DENTIST_DPREFCLK_WDIVIDER;
+	type DENTIST_DPREFCLK_WDIVIDER; \
+	type DENTIST_DISPCLK_WDIVIDER; \
+	type DENTIST_DISPCLK_CHG_DONE;
 
-struct dce_disp_clk_shift {
+struct dccg_shift {
 	CLK_REG_FIELD_LIST(uint8_t)
 };
 
-struct dce_disp_clk_mask {
+struct dccg_mask {
 	CLK_REG_FIELD_LIST(uint32_t)
 };
 
-struct dce_disp_clk_registers {
+struct dccg_registers {
 	uint32_t DPREFCLK_CNTL;
 	uint32_t DENTIST_DISPCLK_CNTL;
 };
 
-/* Array identifiers and count for the divider ranges.*/
-enum dce_divider_range_count {
-	DIVIDER_RANGE_01 = 0,
-	DIVIDER_RANGE_02,
-	DIVIDER_RANGE_03,
-	DIVIDER_RANGE_MAX /* == 3*/
-};
-
-enum dce_divider_error_types {
-	INVALID_DID = 0,
-	INVALID_DIVIDER = 1
-};
-
-struct dce_divider_range {
-	int div_range_start;
-	/* The end of this range of dividers.*/
-	int div_range_end;
-	/* The distance between each divider in this range.*/
-	int div_range_step;
-	/* The divider id for the lowest divider.*/
-	int did_min;
-	/* The divider id for the highest divider.*/
-	int did_max;
-};
-
-struct dce_disp_clk {
-	struct display_clock base;
-	const struct dce_disp_clk_registers *regs;
-	const struct dce_disp_clk_shift *clk_shift;
-	const struct dce_disp_clk_mask *clk_mask;
+struct dce_dccg {
+	struct dccg base;
+	const struct dccg_registers *regs;
+	const struct dccg_shift *clk_shift;
+	const struct dccg_mask *clk_mask;
 
 	struct state_dependent_clocks max_clks_by_state[DM_PP_CLOCKS_MAX_STATES];
-	struct dce_divider_range divider_ranges[DIVIDER_RANGE_MAX];
 
-	bool use_max_disp_clk;
 	int dentist_vco_freq_khz;
 
 	/* Cache the status of DFS-bypass feature*/
@@ -106,32 +88,33 @@ struct dce_disp_clk {
 	int dprefclk_ss_percentage;
 	/* DPREFCLK SS percentage Divider (100 or 1000) */
 	int dprefclk_ss_divider;
-
-	/* max disp_clk from PPLIB for max validation display clock*/
-	int max_displ_clk_in_khz;
 };
 
 
-struct display_clock *dce_disp_clk_create(
+struct dccg *dce_dccg_create(
 	struct dc_context *ctx,
-	const struct dce_disp_clk_registers *regs,
-	const struct dce_disp_clk_shift *clk_shift,
-	const struct dce_disp_clk_mask *clk_mask);
+	const struct dccg_registers *regs,
+	const struct dccg_shift *clk_shift,
+	const struct dccg_mask *clk_mask);
 
-struct display_clock *dce110_disp_clk_create(
+struct dccg *dce110_dccg_create(
 	struct dc_context *ctx,
-	const struct dce_disp_clk_registers *regs,
-	const struct dce_disp_clk_shift *clk_shift,
-	const struct dce_disp_clk_mask *clk_mask);
+	const struct dccg_registers *regs,
+	const struct dccg_shift *clk_shift,
+	const struct dccg_mask *clk_mask);
 
-struct display_clock *dce112_disp_clk_create(
+struct dccg *dce112_dccg_create(
 	struct dc_context *ctx,
-	const struct dce_disp_clk_registers *regs,
-	const struct dce_disp_clk_shift *clk_shift,
-	const struct dce_disp_clk_mask *clk_mask);
+	const struct dccg_registers *regs,
+	const struct dccg_shift *clk_shift,
+	const struct dccg_mask *clk_mask);
+
+struct dccg *dce120_dccg_create(struct dc_context *ctx);
 
-struct display_clock *dce120_disp_clk_create(struct dc_context *ctx);
+#ifdef CONFIG_X86
+struct dccg *dcn1_dccg_create(struct dc_context *ctx);
+#endif
 
-void dce_disp_clk_destroy(struct display_clock **disp_clk);
+void dce_dccg_destroy(struct dccg **dccg);
 
 #endif /* _DCE_CLOCKS_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c
index 2ee3d9bf1062..062a46543887 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_dmcu.c
@@ -28,7 +28,7 @@
 #include "dce_dmcu.h"
 #include "dm_services.h"
 #include "reg_helper.h"
-#include "fixed32_32.h"
+#include "fixed31_32.h"
 #include "dc.h"
 
 #define TO_DCE_DMCU(dmcu)\
@@ -314,7 +314,7 @@ static void dce_get_psr_wait_loop(
 	return;
 }
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 static void dcn10_get_dmcu_state(struct dmcu *dmcu)
 {
 	struct dce_dmcu *dmcu_dce = TO_DCE_DMCU(dmcu);
@@ -735,7 +735,7 @@ static const struct dmcu_funcs dce_funcs = {
 	.is_dmcu_initialized = dce_is_dmcu_initialized
 };
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 static const struct dmcu_funcs dcn10_funcs = {
 	.dmcu_init = dcn10_dmcu_init,
 	.load_iram = dcn10_dmcu_load_iram,
@@ -787,7 +787,7 @@ struct dmcu *dce_dmcu_create(
 	return &dmcu_dce->base;
 }
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 struct dmcu *dcn10_dmcu_create(
 	struct dc_context *ctx,
 	const struct dce_dmcu_registers *regs,
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.c b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.c
index 487724345d9d..0275d6d60da4 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.c
@@ -53,7 +53,8 @@ void dce_pipe_control_lock(struct dc *dc,
 	struct dce_hwseq *hws = dc->hwseq;
 
 	/* Not lock pipe when blank */
-	if (lock && pipe->stream_res.tg->funcs->is_blanked(pipe->stream_res.tg))
+	if (lock && pipe->stream_res.tg->funcs->is_blanked &&
+	    pipe->stream_res.tg->funcs->is_blanked(pipe->stream_res.tg))
 		return;
 
 	val = REG_GET_4(BLND_V_UPDATE_LOCK[pipe->stream_res.tg->inst],
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h
index 057b8afd74bc..64dc75378541 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_hwseq.h
@@ -147,6 +147,7 @@
 	SR(DCCG_GATE_DISABLE_CNTL2), \
 	SR(DCFCLK_CNTL),\
 	SR(DCFCLK_CNTL), \
+	SR(DC_MEM_GLOBAL_PWR_REQ_CNTL), \
 	/* todo:  get these from GVM instead of reading registers ourselves */\
 	MMHUB_SR(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_HI32),\
 	MMHUB_SR(VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LO32),\
@@ -249,7 +250,6 @@ struct dce_hwseq_registers {
 	uint32_t DISPCLK_FREQ_CHANGE_CNTL;
 	uint32_t RBBMIF_TIMEOUT_DIS;
 	uint32_t RBBMIF_TIMEOUT_DIS_2;
-	uint32_t DENTIST_DISPCLK_CNTL;
 	uint32_t DCHUBBUB_CRC_CTRL;
 	uint32_t DPP_TOP0_DPP_CRC_CTRL;
 	uint32_t DPP_TOP0_DPP_CRC_VAL_R_G;
@@ -276,6 +276,8 @@ struct dce_hwseq_registers {
 	uint32_t MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR_LSB;
 	uint32_t MC_VM_SYSTEM_APERTURE_LOW_ADDR;
 	uint32_t MC_VM_SYSTEM_APERTURE_HIGH_ADDR;
+	uint32_t AZALIA_AUDIO_DTO;
+	uint32_t AZALIA_CONTROLLER_CLOCK_GATING;
 };
  /* set field name */
 #define HWS_SF(blk_name, reg_name, field_name, post_fix)\
@@ -362,7 +364,8 @@ struct dce_hwseq_registers {
 	HWSEQ_PIXEL_RATE_MASK_SH_LIST(mask_sh, OTG0_),\
 	HWS_SF1(OTG0_, PHYPLL_PIXEL_RATE_CNTL, PHYPLL_PIXEL_RATE_SOURCE, mask_sh), \
 	HWS_SF(, DCHUBBUB_GLOBAL_TIMER_CNTL, DCHUBBUB_GLOBAL_TIMER_ENABLE, mask_sh), \
-	HWS_SF(, DCFCLK_CNTL, DCFCLK_GATE_DIS, mask_sh)
+	HWS_SF(, DCFCLK_CNTL, DCFCLK_GATE_DIS, mask_sh), \
+	HWS_SF(, DC_MEM_GLOBAL_PWR_REQ_CNTL, DC_MEM_GLOBAL_PWR_REQ_DIS, mask_sh)
 
 #define HWSEQ_DCN1_MASK_SH_LIST(mask_sh)\
 	HWSEQ_DCN_MASK_SH_LIST(mask_sh), \
@@ -496,14 +499,13 @@ struct dce_hwseq_registers {
 	type DOMAIN7_PGFSM_PWR_STATUS; \
 	type DCFCLK_GATE_DIS; \
 	type DCHUBBUB_GLOBAL_TIMER_REFDIV; \
-	type DENTIST_DPPCLK_WDIVIDER; \
-	type DENTIST_DISPCLK_WDIVIDER; \
 	type VGA_TEST_ENABLE; \
 	type VGA_TEST_RENDER_START; \
 	type D1VGA_MODE_ENABLE; \
 	type D2VGA_MODE_ENABLE; \
 	type D3VGA_MODE_ENABLE; \
-	type D4VGA_MODE_ENABLE;
+	type D4VGA_MODE_ENABLE; \
+	type AZALIA_AUDIO_DTO_MODULE;
 
 struct dce_hwseq_shift {
 	HWSEQ_REG_FIELD_LIST(uint8_t)
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_ipp.c b/drivers/gpu/drm/amd/display/dc/dce/dce_ipp.c
index d737e911971b..5d9506b3d46b 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_ipp.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_ipp.c
@@ -195,13 +195,13 @@ static void dce_ipp_program_input_lut(
 
 	for (i = 0; i < gamma->num_entries; i++) {
 		REG_SET(DC_LUT_SEQ_COLOR, 0, DC_LUT_SEQ_COLOR,
-				dal_fixed31_32_round(
+				dc_fixpt_round(
 					gamma->entries.red[i]));
 		REG_SET(DC_LUT_SEQ_COLOR, 0, DC_LUT_SEQ_COLOR,
-				dal_fixed31_32_round(
+				dc_fixpt_round(
 					gamma->entries.green[i]));
 		REG_SET(DC_LUT_SEQ_COLOR, 0, DC_LUT_SEQ_COLOR,
-				dal_fixed31_32_round(
+				dc_fixpt_round(
 					gamma->entries.blue[i]));
 	}
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c
index 8167cad7bcf7..60e3c6a73d37 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.c
@@ -113,6 +113,7 @@ static const struct link_encoder_funcs dce110_lnk_enc_funcs = {
 	.connect_dig_be_to_fe = dce110_link_encoder_connect_dig_be_to_fe,
 	.enable_hpd = dce110_link_encoder_enable_hpd,
 	.disable_hpd = dce110_link_encoder_disable_hpd,
+	.is_dig_enabled = dce110_is_dig_enabled,
 	.destroy = dce110_link_encoder_destroy
 };
 
@@ -535,8 +536,9 @@ void dce110_psr_program_secondary_packet(struct link_encoder *enc,
 		DP_SEC_GSP0_PRIORITY, 1);
 }
 
-static bool is_dig_enabled(const struct dce110_link_encoder *enc110)
+bool dce110_is_dig_enabled(struct link_encoder *enc)
 {
+	struct dce110_link_encoder *enc110 = TO_DCE110_LINK_ENC(enc);
 	uint32_t value;
 
 	REG_GET(DIG_BE_EN_CNTL, DIG_ENABLE, &value);
@@ -644,6 +646,9 @@ static bool dce110_link_encoder_validate_hdmi_output(
 	if (!enc110->base.features.flags.bits.HDMI_6GB_EN &&
 		adjusted_pix_clk_khz >= 300000)
 		return false;
+	if (enc110->base.ctx->dc->debug.hdmi20_disable &&
+		crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR420)
+		return false;
 	return true;
 }
 
@@ -771,6 +776,9 @@ void dce110_link_encoder_construct(
 				__func__,
 				result);
 	}
+	if (enc110->base.ctx->dc->debug.hdmi20_disable) {
+		enc110->base.features.flags.bits.HDMI_6GB_EN = 0;
+	}
 }
 
 bool dce110_link_encoder_validate_output_with_stream(
@@ -1031,7 +1039,7 @@ void dce110_link_encoder_disable_output(
 	struct bp_transmitter_control cntl = { 0 };
 	enum bp_result result;
 
-	if (!is_dig_enabled(enc110)) {
+	if (!dce110_is_dig_enabled(enc)) {
 		/* OF_SKIP_POWER_DOWN_INACTIVE_ENCODER */
 		return;
 	}
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.h
index 0ec3433d34b6..347069461a22 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_link_encoder.h
@@ -263,4 +263,6 @@ void dce110_psr_program_dp_dphy_fast_training(struct link_encoder *enc,
 void dce110_psr_program_secondary_packet(struct link_encoder *enc,
 			unsigned int sdp_transmit_line_num_deadline);
 
+bool dce110_is_dig_enabled(struct link_encoder *enc);
+
 #endif /* __DC_LINK_ENCODER__DCE110_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
index 0790f25c7b3b..85686d917636 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
@@ -174,6 +174,25 @@ static void program_urgency_watermark(
 		URGENCY_HIGH_WATERMARK, urgency_high_wm);
 }
 
+static void dce120_program_urgency_watermark(
+	struct dce_mem_input *dce_mi,
+	uint32_t wm_select,
+	uint32_t urgency_low_wm,
+	uint32_t urgency_high_wm)
+{
+	REG_UPDATE(DPG_WATERMARK_MASK_CONTROL,
+		URGENCY_WATERMARK_MASK, wm_select);
+
+	REG_SET_2(DPG_PIPE_URGENCY_CONTROL, 0,
+		URGENCY_LOW_WATERMARK, urgency_low_wm,
+		URGENCY_HIGH_WATERMARK, urgency_high_wm);
+
+	REG_SET_2(DPG_PIPE_URGENT_LEVEL_CONTROL, 0,
+		URGENT_LEVEL_LOW_WATERMARK, urgency_low_wm,
+		URGENT_LEVEL_HIGH_WATERMARK, urgency_high_wm);
+
+}
+
 static void program_nbp_watermark(
 	struct dce_mem_input *dce_mi,
 	uint32_t wm_select,
@@ -206,6 +225,25 @@ static void program_nbp_watermark(
 	}
 }
 
+static void dce120_program_stutter_watermark(
+	struct dce_mem_input *dce_mi,
+	uint32_t wm_select,
+	uint32_t stutter_mark,
+	uint32_t stutter_entry)
+{
+	REG_UPDATE(DPG_WATERMARK_MASK_CONTROL,
+		STUTTER_EXIT_SELF_REFRESH_WATERMARK_MASK, wm_select);
+
+	if (REG(DPG_PIPE_STUTTER_CONTROL2))
+		REG_UPDATE_2(DPG_PIPE_STUTTER_CONTROL2,
+				STUTTER_EXIT_SELF_REFRESH_WATERMARK, stutter_mark,
+				STUTTER_ENTER_SELF_REFRESH_WATERMARK, stutter_entry);
+	else
+		REG_UPDATE_2(DPG_PIPE_STUTTER_CONTROL,
+				STUTTER_EXIT_SELF_REFRESH_WATERMARK, stutter_mark,
+				STUTTER_ENTER_SELF_REFRESH_WATERMARK, stutter_entry);
+}
+
 static void program_stutter_watermark(
 	struct dce_mem_input *dce_mi,
 	uint32_t wm_select,
@@ -225,7 +263,8 @@ static void program_stutter_watermark(
 static void dce_mi_program_display_marks(
 	struct mem_input *mi,
 	struct dce_watermarks nbp,
-	struct dce_watermarks stutter,
+	struct dce_watermarks stutter_exit,
+	struct dce_watermarks stutter_enter,
 	struct dce_watermarks urgent,
 	uint32_t total_dest_line_time_ns)
 {
@@ -243,13 +282,14 @@ static void dce_mi_program_display_marks(
 	program_nbp_watermark(dce_mi, 2, nbp.a_mark); /* set a */
 	program_nbp_watermark(dce_mi, 1, nbp.d_mark); /* set d */
 
-	program_stutter_watermark(dce_mi, 2, stutter.a_mark); /* set a */
-	program_stutter_watermark(dce_mi, 1, stutter.d_mark); /* set d */
+	program_stutter_watermark(dce_mi, 2, stutter_exit.a_mark); /* set a */
+	program_stutter_watermark(dce_mi, 1, stutter_exit.d_mark); /* set d */
 }
 
-static void dce120_mi_program_display_marks(struct mem_input *mi,
+static void dce112_mi_program_display_marks(struct mem_input *mi,
 	struct dce_watermarks nbp,
-	struct dce_watermarks stutter,
+	struct dce_watermarks stutter_exit,
+	struct dce_watermarks stutter_entry,
 	struct dce_watermarks urgent,
 	uint32_t total_dest_line_time_ns)
 {
@@ -273,10 +313,43 @@ static void dce120_mi_program_display_marks(struct mem_input *mi,
 	program_nbp_watermark(dce_mi, 2, nbp.c_mark); /* set c */
 	program_nbp_watermark(dce_mi, 3, nbp.d_mark); /* set d */
 
-	program_stutter_watermark(dce_mi, 0, stutter.a_mark); /* set a */
-	program_stutter_watermark(dce_mi, 1, stutter.b_mark); /* set b */
-	program_stutter_watermark(dce_mi, 2, stutter.c_mark); /* set c */
-	program_stutter_watermark(dce_mi, 3, stutter.d_mark); /* set d */
+	program_stutter_watermark(dce_mi, 0, stutter_exit.a_mark); /* set a */
+	program_stutter_watermark(dce_mi, 1, stutter_exit.b_mark); /* set b */
+	program_stutter_watermark(dce_mi, 2, stutter_exit.c_mark); /* set c */
+	program_stutter_watermark(dce_mi, 3, stutter_exit.d_mark); /* set d */
+}
+
+static void dce120_mi_program_display_marks(struct mem_input *mi,
+	struct dce_watermarks nbp,
+	struct dce_watermarks stutter_exit,
+	struct dce_watermarks stutter_entry,
+	struct dce_watermarks urgent,
+	uint32_t total_dest_line_time_ns)
+{
+	struct dce_mem_input *dce_mi = TO_DCE_MEM_INPUT(mi);
+	uint32_t stutter_en = mi->ctx->dc->debug.disable_stutter ? 0 : 1;
+
+	dce120_program_urgency_watermark(dce_mi, 0, /* set a */
+			urgent.a_mark, total_dest_line_time_ns);
+	dce120_program_urgency_watermark(dce_mi, 1, /* set b */
+			urgent.b_mark, total_dest_line_time_ns);
+	dce120_program_urgency_watermark(dce_mi, 2, /* set c */
+			urgent.c_mark, total_dest_line_time_ns);
+	dce120_program_urgency_watermark(dce_mi, 3, /* set d */
+			urgent.d_mark, total_dest_line_time_ns);
+
+	REG_UPDATE_2(DPG_PIPE_STUTTER_CONTROL,
+		STUTTER_ENABLE, stutter_en,
+		STUTTER_IGNORE_FBC, 1);
+	program_nbp_watermark(dce_mi, 0, nbp.a_mark); /* set a */
+	program_nbp_watermark(dce_mi, 1, nbp.b_mark); /* set b */
+	program_nbp_watermark(dce_mi, 2, nbp.c_mark); /* set c */
+	program_nbp_watermark(dce_mi, 3, nbp.d_mark); /* set d */
+
+	dce120_program_stutter_watermark(dce_mi, 0, stutter_exit.a_mark, stutter_entry.a_mark); /* set a */
+	dce120_program_stutter_watermark(dce_mi, 1, stutter_exit.b_mark, stutter_entry.b_mark); /* set b */
+	dce120_program_stutter_watermark(dce_mi, 2, stutter_exit.c_mark, stutter_entry.c_mark); /* set c */
+	dce120_program_stutter_watermark(dce_mi, 3, stutter_exit.d_mark, stutter_entry.d_mark); /* set d */
 }
 
 static void program_tiling(
@@ -656,7 +729,7 @@ static bool dce_mi_program_surface_flip_and_addr(
 	return true;
 }
 
-static struct mem_input_funcs dce_mi_funcs = {
+static const struct mem_input_funcs dce_mi_funcs = {
 	.mem_input_program_display_marks = dce_mi_program_display_marks,
 	.allocate_mem_input = dce_mi_allocate_dmif,
 	.free_mem_input = dce_mi_free_dmif,
@@ -668,6 +741,29 @@ static struct mem_input_funcs dce_mi_funcs = {
 	.mem_input_is_flip_pending = dce_mi_is_flip_pending
 };
 
+static const struct mem_input_funcs dce112_mi_funcs = {
+	.mem_input_program_display_marks = dce112_mi_program_display_marks,
+	.allocate_mem_input = dce_mi_allocate_dmif,
+	.free_mem_input = dce_mi_free_dmif,
+	.mem_input_program_surface_flip_and_addr =
+			dce_mi_program_surface_flip_and_addr,
+	.mem_input_program_pte_vm = dce_mi_program_pte_vm,
+	.mem_input_program_surface_config =
+			dce_mi_program_surface_config,
+	.mem_input_is_flip_pending = dce_mi_is_flip_pending
+};
+
+static const struct mem_input_funcs dce120_mi_funcs = {
+	.mem_input_program_display_marks = dce120_mi_program_display_marks,
+	.allocate_mem_input = dce_mi_allocate_dmif,
+	.free_mem_input = dce_mi_free_dmif,
+	.mem_input_program_surface_flip_and_addr =
+			dce_mi_program_surface_flip_and_addr,
+	.mem_input_program_pte_vm = dce_mi_program_pte_vm,
+	.mem_input_program_surface_config =
+			dce_mi_program_surface_config,
+	.mem_input_is_flip_pending = dce_mi_is_flip_pending
+};
 
 void dce_mem_input_construct(
 	struct dce_mem_input *dce_mi,
@@ -696,5 +792,17 @@ void dce112_mem_input_construct(
 	const struct dce_mem_input_mask *mi_mask)
 {
 	dce_mem_input_construct(dce_mi, ctx, inst, regs, mi_shift, mi_mask);
-	dce_mi->base.funcs->mem_input_program_display_marks = dce120_mi_program_display_marks;
+	dce_mi->base.funcs = &dce112_mi_funcs;
+}
+
+void dce120_mem_input_construct(
+	struct dce_mem_input *dce_mi,
+	struct dc_context *ctx,
+	int inst,
+	const struct dce_mem_input_registers *regs,
+	const struct dce_mem_input_shift *mi_shift,
+	const struct dce_mem_input_mask *mi_mask)
+{
+	dce_mem_input_construct(dce_mi, ctx, inst, regs, mi_shift, mi_mask);
+	dce_mi->base.funcs = &dce120_mi_funcs;
 }
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h
index 05d39c0cbe87..d15b0d7f47fc 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.h
@@ -106,6 +106,7 @@ struct dce_mem_input_registers {
 	uint32_t DPG_PIPE_ARBITRATION_CONTROL1;
 	uint32_t DPG_WATERMARK_MASK_CONTROL;
 	uint32_t DPG_PIPE_URGENCY_CONTROL;
+	uint32_t DPG_PIPE_URGENT_LEVEL_CONTROL;
 	uint32_t DPG_PIPE_NB_PSTATE_CHANGE_CONTROL;
 	uint32_t DPG_PIPE_LOW_POWER_CONTROL;
 	uint32_t DPG_PIPE_STUTTER_CONTROL;
@@ -213,6 +214,11 @@ struct dce_mem_input_registers {
 
 #define MI_DCE12_DMIF_PG_MASK_SH_LIST(mask_sh, blk)\
 	SFB(blk, DPG_PIPE_STUTTER_CONTROL2, STUTTER_EXIT_SELF_REFRESH_WATERMARK, mask_sh),\
+	SFB(blk, DPG_PIPE_STUTTER_CONTROL2, STUTTER_ENTER_SELF_REFRESH_WATERMARK, mask_sh),\
+	SFB(blk, DPG_PIPE_URGENT_LEVEL_CONTROL, URGENT_LEVEL_LOW_WATERMARK, mask_sh),\
+	SFB(blk, DPG_PIPE_URGENT_LEVEL_CONTROL, URGENT_LEVEL_HIGH_WATERMARK, mask_sh),\
+	SFB(blk, DPG_PIPE_URGENCY_CONTROL, URGENCY_LOW_WATERMARK, mask_sh),\
+	SFB(blk, DPG_PIPE_URGENCY_CONTROL, URGENCY_HIGH_WATERMARK, mask_sh),\
 	SFB(blk, DPG_WATERMARK_MASK_CONTROL, PSTATE_CHANGE_WATERMARK_MASK, mask_sh),\
 	SFB(blk, DPG_PIPE_LOW_POWER_CONTROL, PSTATE_CHANGE_ENABLE, mask_sh),\
 	SFB(blk, DPG_PIPE_LOW_POWER_CONTROL, PSTATE_CHANGE_URGENT_DURING_REQUEST, mask_sh),\
@@ -286,6 +292,8 @@ struct dce_mem_input_registers {
 	type STUTTER_EXIT_SELF_REFRESH_WATERMARK_MASK; \
 	type URGENCY_LOW_WATERMARK; \
 	type URGENCY_HIGH_WATERMARK; \
+	type URGENT_LEVEL_LOW_WATERMARK;\
+	type URGENT_LEVEL_HIGH_WATERMARK;\
 	type NB_PSTATE_CHANGE_ENABLE; \
 	type NB_PSTATE_CHANGE_URGENT_DURING_REQUEST; \
 	type NB_PSTATE_CHANGE_NOT_SELF_REFRESH_DURING_REQUEST; \
@@ -297,6 +305,7 @@ struct dce_mem_input_registers {
 	type STUTTER_ENABLE; \
 	type STUTTER_IGNORE_FBC; \
 	type STUTTER_EXIT_SELF_REFRESH_WATERMARK; \
+	type STUTTER_ENTER_SELF_REFRESH_WATERMARK; \
 	type DMIF_BUFFERS_ALLOCATED; \
 	type DMIF_BUFFERS_ALLOCATION_COMPLETED; \
 	type ENABLE; /* MC_HUB_RDREQ_DMIF_LIMIT */\
@@ -344,4 +353,12 @@ void dce112_mem_input_construct(
 	const struct dce_mem_input_shift *mi_shift,
 	const struct dce_mem_input_mask *mi_mask);
 
+void dce120_mem_input_construct(
+	struct dce_mem_input *dce_mi,
+	struct dc_context *ctx,
+	int inst,
+	const struct dce_mem_input_registers *regs,
+	const struct dce_mem_input_shift *mi_shift,
+	const struct dce_mem_input_mask *mi_mask);
+
 #endif /*__DCE_MEM_INPUT_H__*/
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_scl_filters.c b/drivers/gpu/drm/amd/display/dc/dce/dce_scl_filters.c
index 6243450b41b7..48862bebf29e 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_scl_filters.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_scl_filters.c
@@ -1014,11 +1014,11 @@ static const uint16_t filter_8tap_64p_183[264] = {
 
 const uint16_t *get_filter_3tap_16p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_3tap_16p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_3tap_16p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_3tap_16p_150;
 	else
 		return filter_3tap_16p_183;
@@ -1026,11 +1026,11 @@ const uint16_t *get_filter_3tap_16p(struct fixed31_32 ratio)
 
 const uint16_t *get_filter_3tap_64p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_3tap_64p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_3tap_64p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_3tap_64p_150;
 	else
 		return filter_3tap_64p_183;
@@ -1038,11 +1038,11 @@ const uint16_t *get_filter_3tap_64p(struct fixed31_32 ratio)
 
 const uint16_t *get_filter_4tap_16p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_4tap_16p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_4tap_16p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_4tap_16p_150;
 	else
 		return filter_4tap_16p_183;
@@ -1050,11 +1050,11 @@ const uint16_t *get_filter_4tap_16p(struct fixed31_32 ratio)
 
 const uint16_t *get_filter_4tap_64p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_4tap_64p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_4tap_64p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_4tap_64p_150;
 	else
 		return filter_4tap_64p_183;
@@ -1062,11 +1062,11 @@ const uint16_t *get_filter_4tap_64p(struct fixed31_32 ratio)
 
 const uint16_t *get_filter_5tap_64p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_5tap_64p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_5tap_64p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_5tap_64p_150;
 	else
 		return filter_5tap_64p_183;
@@ -1074,11 +1074,11 @@ const uint16_t *get_filter_5tap_64p(struct fixed31_32 ratio)
 
 const uint16_t *get_filter_6tap_64p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_6tap_64p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_6tap_64p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_6tap_64p_150;
 	else
 		return filter_6tap_64p_183;
@@ -1086,11 +1086,11 @@ const uint16_t *get_filter_6tap_64p(struct fixed31_32 ratio)
 
 const uint16_t *get_filter_7tap_64p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_7tap_64p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_7tap_64p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_7tap_64p_150;
 	else
 		return filter_7tap_64p_183;
@@ -1098,11 +1098,11 @@ const uint16_t *get_filter_7tap_64p(struct fixed31_32 ratio)
 
 const uint16_t *get_filter_8tap_64p(struct fixed31_32 ratio)
 {
-	if (ratio.value < dal_fixed31_32_one.value)
+	if (ratio.value < dc_fixpt_one.value)
 		return filter_8tap_64p_upscale;
-	else if (ratio.value < dal_fixed31_32_from_fraction(4, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(4, 3).value)
 		return filter_8tap_64p_117;
-	else if (ratio.value < dal_fixed31_32_from_fraction(5, 3).value)
+	else if (ratio.value < dc_fixpt_from_fraction(5, 3).value)
 		return filter_8tap_64p_150;
 	else
 		return filter_8tap_64p_183;
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
index 162f6a6c4208..b139b4017820 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
@@ -26,27 +26,10 @@
 #include "dc_bios_types.h"
 #include "dce_stream_encoder.h"
 #include "reg_helper.h"
+#include "hw_shared.h"
+
 #define DC_LOGGER \
 		enc110->base.ctx->logger
-enum DP_PIXEL_ENCODING {
-DP_PIXEL_ENCODING_RGB444                 = 0x00000000,
-DP_PIXEL_ENCODING_YCBCR422               = 0x00000001,
-DP_PIXEL_ENCODING_YCBCR444               = 0x00000002,
-DP_PIXEL_ENCODING_RGB_WIDE_GAMUT         = 0x00000003,
-DP_PIXEL_ENCODING_Y_ONLY                 = 0x00000004,
-DP_PIXEL_ENCODING_YCBCR420               = 0x00000005,
-DP_PIXEL_ENCODING_RESERVED               = 0x00000006,
-};
-
-
-enum DP_COMPONENT_DEPTH {
-DP_COMPONENT_DEPTH_6BPC                  = 0x00000000,
-DP_COMPONENT_DEPTH_8BPC                  = 0x00000001,
-DP_COMPONENT_DEPTH_10BPC                 = 0x00000002,
-DP_COMPONENT_DEPTH_12BPC                 = 0x00000003,
-DP_COMPONENT_DEPTH_16BPC                 = 0x00000004,
-DP_COMPONENT_DEPTH_RESERVED              = 0x00000005,
-};
 
 
 #define REG(reg)\
@@ -80,7 +63,7 @@ enum {
 static void dce110_update_generic_info_packet(
 	struct dce110_stream_encoder *enc110,
 	uint32_t packet_index,
-	const struct encoder_info_packet *info_packet)
+	const struct dc_info_packet *info_packet)
 {
 	uint32_t regval;
 	/* TODOFPGA Figure out a proper number for max_retries polling for lock
@@ -89,7 +72,8 @@ static void dce110_update_generic_info_packet(
 	uint32_t max_retries = 50;
 
 	/*we need turn on clock before programming AFMT block*/
-	REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, 1);
+	if (REG(AFMT_CNTL))
+		REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, 1);
 
 	if (REG(AFMT_VBI_PACKET_CONTROL1)) {
 		if (packet_index >= 8)
@@ -151,7 +135,7 @@ static void dce110_update_generic_info_packet(
 			AFMT_GENERIC0_UPDATE, (packet_index == 0),
 			AFMT_GENERIC2_UPDATE, (packet_index == 2));
 	}
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	if (REG(AFMT_VBI_PACKET_CONTROL1)) {
 		switch (packet_index) {
 		case 0:
@@ -196,7 +180,7 @@ static void dce110_update_generic_info_packet(
 static void dce110_update_hdmi_info_packet(
 	struct dce110_stream_encoder *enc110,
 	uint32_t packet_index,
-	const struct encoder_info_packet *info_packet)
+	const struct dc_info_packet *info_packet)
 {
 	uint32_t cont, send, line;
 
@@ -245,7 +229,7 @@ static void dce110_update_hdmi_info_packet(
 				HDMI_GENERIC1_SEND, send,
 				HDMI_GENERIC1_LINE, line);
 		break;
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	case 4:
 		if (REG(HDMI_GENERIC_PACKET_CONTROL2))
 			REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL2,
@@ -290,7 +274,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute(
 	struct dc_crtc_timing *crtc_timing,
 	enum dc_color_space output_color_space)
 {
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	uint32_t h_active_start;
 	uint32_t v_active_start;
 	uint32_t misc0 = 0;
@@ -305,20 +289,15 @@ static void dce110_stream_encoder_dp_set_stream_attribute(
 
 	struct dce110_stream_encoder *enc110 = DCE110STRENC_FROM_STRENC(enc);
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
-	if (REG(DP_DB_CNTL))
-		REG_UPDATE(DP_DB_CNTL, DP_DB_DISABLE, 1);
-#endif
-
 	/* set pixel encoding */
 	switch (crtc_timing->pixel_encoding) {
 	case PIXEL_ENCODING_YCBCR422:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-				DP_PIXEL_ENCODING_YCBCR422);
+				DP_PIXEL_ENCODING_TYPE_YCBCR422);
 		break;
 	case PIXEL_ENCODING_YCBCR444:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-				DP_PIXEL_ENCODING_YCBCR444);
+				DP_PIXEL_ENCODING_TYPE_YCBCR444);
 
 		if (crtc_timing->flags.Y_ONLY)
 			if (crtc_timing->display_color_depth != COLOR_DEPTH_666)
@@ -326,7 +305,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute(
 				 * Color depth of Y-only could be
 				 * 8, 10, 12, 16 bits */
 				REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-						DP_PIXEL_ENCODING_Y_ONLY);
+						DP_PIXEL_ENCODING_TYPE_Y_ONLY);
 		/* Note: DP_MSA_MISC1 bit 7 is the indicator
 		 * of Y-only mode.
 		 * This bit is set in HW if register
@@ -334,22 +313,22 @@ static void dce110_stream_encoder_dp_set_stream_attribute(
 		break;
 	case PIXEL_ENCODING_YCBCR420:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-				DP_PIXEL_ENCODING_YCBCR420);
+				DP_PIXEL_ENCODING_TYPE_YCBCR420);
 		if (enc110->se_mask->DP_VID_M_DOUBLE_VALUE_EN)
 			REG_UPDATE(DP_VID_TIMING, DP_VID_M_DOUBLE_VALUE_EN, 1);
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 		if (enc110->se_mask->DP_VID_N_MUL)
 			REG_UPDATE(DP_VID_TIMING, DP_VID_N_MUL, 1);
 #endif
 		break;
 	default:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_PIXEL_ENCODING,
-				DP_PIXEL_ENCODING_RGB444);
+				DP_PIXEL_ENCODING_TYPE_RGB444);
 		break;
 	}
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	if (REG(DP_MSA_MISC))
 		misc1 = REG_READ(DP_MSA_MISC);
 #endif
@@ -363,27 +342,27 @@ static void dce110_stream_encoder_dp_set_stream_attribute(
 		break;
 	case COLOR_DEPTH_888:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
-				DP_COMPONENT_DEPTH_8BPC);
+				DP_COMPONENT_PIXEL_DEPTH_8BPC);
 		break;
 	case COLOR_DEPTH_101010:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
-				DP_COMPONENT_DEPTH_10BPC);
+				DP_COMPONENT_PIXEL_DEPTH_10BPC);
 
 		break;
 	case COLOR_DEPTH_121212:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
-				DP_COMPONENT_DEPTH_12BPC);
+				DP_COMPONENT_PIXEL_DEPTH_12BPC);
 		break;
 	default:
 		REG_UPDATE(DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH,
-				DP_COMPONENT_DEPTH_6BPC);
+				DP_COMPONENT_PIXEL_DEPTH_6BPC);
 		break;
 	}
 
 	/* set dynamic range and YCbCr range */
 
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	switch (crtc_timing->display_color_depth) {
 	case COLOR_DEPTH_666:
 		colorimetry_bpc = 0;
@@ -462,7 +441,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute(
 				DP_DYN_RANGE, dynamic_range_rgb,
 				DP_YCBCR_RANGE, dynamic_range_ycbcr);
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 		if (REG(DP_MSA_COLORIMETRY))
 			REG_SET(DP_MSA_COLORIMETRY, 0, DP_MSA_MISC0, misc0);
 
@@ -497,7 +476,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute(
 				crtc_timing->v_front_porch;
 
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 		/* start at begining of left border */
 		if (REG(DP_MSA_TIMING_PARAM2))
 			REG_SET_2(DP_MSA_TIMING_PARAM2, 0,
@@ -700,11 +679,11 @@ static void dce110_stream_encoder_set_mst_bandwidth(
 	struct fixed31_32 avg_time_slots_per_mtp)
 {
 	struct dce110_stream_encoder *enc110 = DCE110STRENC_FROM_STRENC(enc);
-	uint32_t x = dal_fixed31_32_floor(
+	uint32_t x = dc_fixpt_floor(
 		avg_time_slots_per_mtp);
-	uint32_t y = dal_fixed31_32_ceil(
-		dal_fixed31_32_shl(
-			dal_fixed31_32_sub_int(
+	uint32_t y = dc_fixpt_ceil(
+		dc_fixpt_shl(
+			dc_fixpt_sub_int(
 				avg_time_slots_per_mtp,
 				x),
 			26));
@@ -736,7 +715,8 @@ static void dce110_stream_encoder_update_hdmi_info_packets(
 			const uint32_t *content =
 				(const uint32_t *) &info_frame->avi.sb[0];
 			/*we need turn on clock before programming AFMT block*/
-			REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, 1);
+			if (REG(AFMT_CNTL))
+				REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, 1);
 
 			REG_WRITE(AFMT_AVI_INFO0, content[0]);
 
@@ -771,7 +751,7 @@ static void dce110_stream_encoder_update_hdmi_info_packets(
 		dce110_update_hdmi_info_packet(enc110, 3, &info_frame->hdrsmd);
 	}
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	if (enc110->se_mask->HDMI_DB_DISABLE) {
 		/* for bring up, disable dp double  TODO */
 		if (REG(HDMI_DB_CONTROL))
@@ -809,7 +789,7 @@ static void dce110_stream_encoder_stop_hdmi_info_packets(
 		HDMI_GENERIC1_LINE, 0,
 		HDMI_GENERIC1_SEND, 0);
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	/* stop generic packets 2 & 3 on HDMI */
 	if (REG(HDMI_GENERIC_PACKET_CONTROL2))
 		REG_SET_6(HDMI_GENERIC_PACKET_CONTROL2, 0,
@@ -836,7 +816,7 @@ static void dce110_stream_encoder_update_dp_info_packets(
 	const struct encoder_info_frame *info_frame)
 {
 	struct dce110_stream_encoder *enc110 = DCE110STRENC_FROM_STRENC(enc);
-	uint32_t value = REG_READ(DP_SEC_CNTL);
+	uint32_t value = 0;
 
 	if (info_frame->vsc.valid)
 		dce110_update_generic_info_packet(
@@ -870,6 +850,7 @@ static void dce110_stream_encoder_update_dp_info_packets(
 	* Therefore we need to enable master bit
 	* if at least on of the fields is not 0
 	*/
+	value = REG_READ(DP_SEC_CNTL);
 	if (value)
 		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
 }
@@ -879,7 +860,7 @@ static void dce110_stream_encoder_stop_dp_info_packets(
 {
 	/* stop generic packets on DP */
 	struct dce110_stream_encoder *enc110 = DCE110STRENC_FROM_STRENC(enc);
-	uint32_t value = REG_READ(DP_SEC_CNTL);
+	uint32_t value = 0;
 
 	if (enc110->se_mask->DP_SEC_AVI_ENABLE) {
 		REG_SET_7(DP_SEC_CNTL, 0,
@@ -892,25 +873,10 @@ static void dce110_stream_encoder_stop_dp_info_packets(
 			DP_SEC_STREAM_ENABLE, 0);
 	}
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
-	if (enc110->se_mask->DP_SEC_GSP7_ENABLE) {
-		REG_SET_10(DP_SEC_CNTL, 0,
-			DP_SEC_GSP0_ENABLE, 0,
-			DP_SEC_GSP1_ENABLE, 0,
-			DP_SEC_GSP2_ENABLE, 0,
-			DP_SEC_GSP3_ENABLE, 0,
-			DP_SEC_GSP4_ENABLE, 0,
-			DP_SEC_GSP5_ENABLE, 0,
-			DP_SEC_GSP6_ENABLE, 0,
-			DP_SEC_GSP7_ENABLE, 0,
-			DP_SEC_MPG_ENABLE, 0,
-			DP_SEC_STREAM_ENABLE, 0);
-	}
-#endif
 	/* this register shared with audio info frame.
 	 * therefore we need to keep master enabled
 	 * if at least one of the fields is not 0 */
-
+	value = REG_READ(DP_SEC_CNTL);
 	if (value)
 		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
 
@@ -1513,7 +1479,7 @@ static void dce110_se_disable_dp_audio(
 	struct stream_encoder *enc)
 {
 	struct dce110_stream_encoder *enc110 = DCE110STRENC_FROM_STRENC(enc);
-	uint32_t value = REG_READ(DP_SEC_CNTL);
+	uint32_t value = 0;
 
 	/* Disable Audio packets */
 	REG_UPDATE_5(DP_SEC_CNTL,
@@ -1525,6 +1491,7 @@ static void dce110_se_disable_dp_audio(
 
 	/* This register shared with encoder info frame. Therefore we need to
 	keep master enabled if at least on of the fields is not 0 */
+	value = REG_READ(DP_SEC_CNTL);
 	if (value != 0)
 		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c b/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c
index 832c5daada35..ab63d0d0304c 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_transform.c
@@ -41,7 +41,7 @@
 #define DC_LOGGER \
 	xfm_dce->base.ctx->logger
 
-#define IDENTITY_RATIO(ratio) (dal_fixed31_32_u2d19(ratio) == (1 << 19))
+#define IDENTITY_RATIO(ratio) (dc_fixpt_u2d19(ratio) == (1 << 19))
 #define GAMUT_MATRIX_SIZE 12
 #define SCL_PHASES 16
 
@@ -155,7 +155,7 @@ static void program_overscan(
 	int overscan_bottom = data->v_active
 			- data->recout.y - data->recout.height;
 
-	if (xfm_dce->base.ctx->dc->debug.surface_visual_confirm) {
+	if (xfm_dce->base.ctx->dc->debug.visual_confirm != VISUAL_CONFIRM_DISABLE) {
 		overscan_bottom += 2;
 		overscan_right += 2;
 	}
@@ -256,27 +256,27 @@ static void calculate_inits(
 	struct fixed31_32 v_init;
 
 	inits->h_int_scale_ratio =
-		dal_fixed31_32_u2d19(data->ratios.horz) << 5;
+		dc_fixpt_u2d19(data->ratios.horz) << 5;
 	inits->v_int_scale_ratio =
-		dal_fixed31_32_u2d19(data->ratios.vert) << 5;
+		dc_fixpt_u2d19(data->ratios.vert) << 5;
 
 	h_init =
-		dal_fixed31_32_div_int(
-			dal_fixed31_32_add(
+		dc_fixpt_div_int(
+			dc_fixpt_add(
 				data->ratios.horz,
-				dal_fixed31_32_from_int(data->taps.h_taps + 1)),
+				dc_fixpt_from_int(data->taps.h_taps + 1)),
 				2);
-	inits->h_init.integer = dal_fixed31_32_floor(h_init);
-	inits->h_init.fraction = dal_fixed31_32_u0d19(h_init) << 5;
+	inits->h_init.integer = dc_fixpt_floor(h_init);
+	inits->h_init.fraction = dc_fixpt_u0d19(h_init) << 5;
 
 	v_init =
-		dal_fixed31_32_div_int(
-			dal_fixed31_32_add(
+		dc_fixpt_div_int(
+			dc_fixpt_add(
 				data->ratios.vert,
-				dal_fixed31_32_from_int(data->taps.v_taps + 1)),
+				dc_fixpt_from_int(data->taps.v_taps + 1)),
 				2);
-	inits->v_init.integer = dal_fixed31_32_floor(v_init);
-	inits->v_init.fraction = dal_fixed31_32_u0d19(v_init) << 5;
+	inits->v_init.integer = dc_fixpt_floor(v_init);
+	inits->v_init.fraction = dc_fixpt_u0d19(v_init) << 5;
 }
 
 static void program_scl_ratios_inits(
diff --git a/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.c
index 41f83ecd7469..ec3221333011 100644
--- a/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce100/dce100_hw_sequencer.c
@@ -125,17 +125,54 @@ static void dce100_pplib_apply_display_requirements(
 	dc->prev_display_config = *pp_display_cfg;
 }
 
+/* unit: in_khz before mode set, get pixel clock from context. ASIC register
+ * may not be programmed yet
+ */
+static uint32_t get_max_pixel_clock_for_all_paths(
+	struct dc *dc,
+	struct dc_state *context)
+{
+	uint32_t max_pix_clk = 0;
+	int i;
+
+	for (i = 0; i < MAX_PIPES; i++) {
+		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
+
+		if (pipe_ctx->stream == NULL)
+			continue;
+
+		/* do not check under lay */
+		if (pipe_ctx->top_pipe)
+			continue;
+
+		if (pipe_ctx->stream_res.pix_clk_params.requested_pix_clk > max_pix_clk)
+			max_pix_clk =
+				pipe_ctx->stream_res.pix_clk_params.requested_pix_clk;
+	}
+
+	if (max_pix_clk == 0)
+		ASSERT(0);
+
+	return max_pix_clk;
+}
+
 void dce100_set_bandwidth(
 		struct dc *dc,
 		struct dc_state *context,
 		bool decrease_allowed)
 {
-	if (decrease_allowed || context->bw.dce.dispclk_khz > dc->current_state->bw.dce.dispclk_khz) {
-		dc->res_pool->display_clock->funcs->set_clock(
-				dc->res_pool->display_clock,
-				context->bw.dce.dispclk_khz * 115 / 100);
-		dc->current_state->bw.dce.dispclk_khz = context->bw.dce.dispclk_khz;
-	}
+	struct dc_clocks req_clks;
+
+	req_clks.dispclk_khz = context->bw.dce.dispclk_khz * 115 / 100;
+	req_clks.phyclk_khz = get_max_pixel_clock_for_all_paths(dc, context);
+
+	dce110_set_safe_displaymarks(&context->res_ctx, dc->res_pool);
+
+	dc->res_pool->dccg->funcs->update_clocks(
+			dc->res_pool->dccg,
+			&req_clks,
+			decrease_allowed);
+
 	dce100_pplib_apply_display_requirements(dc, context);
 }
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c b/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c
index 3092f76bdb75..8ed8eace42be 100644
--- a/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce100/dce100_resource.c
@@ -135,15 +135,15 @@ static const struct dce110_timing_generator_offsets dce100_tg_offsets[] = {
 	.reg_name = mm ## block ## id ## _ ## reg_name
 
 
-static const struct dce_disp_clk_registers disp_clk_regs = {
+static const struct dccg_registers disp_clk_regs = {
 		CLK_COMMON_REG_LIST_DCE_BASE()
 };
 
-static const struct dce_disp_clk_shift disp_clk_shift = {
+static const struct dccg_shift disp_clk_shift = {
 		CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(__SHIFT)
 };
 
-static const struct dce_disp_clk_mask disp_clk_mask = {
+static const struct dccg_mask disp_clk_mask = {
 		CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(_MASK)
 };
 
@@ -644,8 +644,8 @@ static void destruct(struct dce110_resource_pool *pool)
 			dce_aud_destroy(&pool->base.audios[i]);
 	}
 
-	if (pool->base.display_clock != NULL)
-		dce_disp_clk_destroy(&pool->base.display_clock);
+	if (pool->base.dccg != NULL)
+		dce_dccg_destroy(&pool->base.dccg);
 
 	if (pool->base.abm != NULL)
 				dce_abm_destroy(&pool->base.abm);
@@ -733,38 +733,6 @@ enum dc_status dce100_add_stream_to_ctx(
 	return result;
 }
 
-enum dc_status dce100_validate_guaranteed(
-		struct dc  *dc,
-		struct dc_stream_state *dc_stream,
-		struct dc_state *context)
-{
-	enum dc_status result = DC_ERROR_UNEXPECTED;
-
-	context->streams[0] = dc_stream;
-	dc_stream_retain(context->streams[0]);
-	context->stream_count++;
-
-	result = resource_map_pool_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = resource_map_clock_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = build_mapped_resource(dc, context, dc_stream);
-
-	if (result == DC_OK) {
-		validate_guaranteed_copy_streams(
-				context, dc->caps.max_streams);
-		result = resource_build_scaling_params_for_context(dc, context);
-	}
-
-	if (result == DC_OK)
-		if (!dce100_validate_bandwidth(dc, context))
-			result = DC_FAIL_BANDWIDTH_VALIDATE;
-
-	return result;
-}
-
 static void dce100_destroy_resource_pool(struct resource_pool **pool)
 {
 	struct dce110_resource_pool *dce110_pool = TO_DCE110_RES_POOL(*pool);
@@ -786,7 +754,6 @@ enum dc_status dce100_validate_plane(const struct dc_plane_state *plane_state, s
 static const struct resource_funcs dce100_res_pool_funcs = {
 	.destroy = dce100_destroy_resource_pool,
 	.link_enc_create = dce100_link_encoder_create,
-	.validate_guaranteed = dce100_validate_guaranteed,
 	.validate_bandwidth = dce100_validate_bandwidth,
 	.validate_plane = dce100_validate_plane,
 	.add_stream_to_ctx = dce100_add_stream_to_ctx,
@@ -850,11 +817,11 @@ static bool construct(
 		}
 	}
 
-	pool->base.display_clock = dce_disp_clk_create(ctx,
+	pool->base.dccg = dce_dccg_create(ctx,
 			&disp_clk_regs,
 			&disp_clk_shift,
 			&disp_clk_mask);
-	if (pool->base.display_clock == NULL) {
+	if (pool->base.dccg == NULL) {
 		dm_error("DC: failed to create display clock!\n");
 		BREAK_TO_DEBUGGER();
 		goto res_create_fail;
@@ -884,7 +851,7 @@ static bool construct(
 	 * max_clock_state
 	 */
 	if (dm_pp_get_static_clocks(ctx, &static_clk_info))
-		pool->base.display_clock->max_clks_state =
+		pool->base.dccg->max_clks_state =
 					static_clk_info.max_clocks_state;
 	{
 		struct irq_service_init_data init_data;
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c
index 9150d2694450..1f7f25013217 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c
@@ -121,10 +121,10 @@ static void reset_lb_on_vblank(struct dc_context *ctx)
 		frame_count = dm_read_reg(ctx, mmCRTC_STATUS_FRAME_COUNT);
 
 
-		for (retry = 100; retry > 0; retry--) {
+		for (retry = 10000; retry > 0; retry--) {
 			if (frame_count != dm_read_reg(ctx, mmCRTC_STATUS_FRAME_COUNT))
 				break;
-			msleep(1);
+			udelay(10);
 		}
 		if (!retry)
 			dm_error("Frame count did not increase for 100ms.\n");
@@ -143,22 +143,22 @@ static void wait_for_fbc_state_changed(
 	struct dce110_compressor *cp110,
 	bool enabled)
 {
-	uint8_t counter = 0;
+	uint32_t counter = 0;
 	uint32_t addr = mmFBC_STATUS;
 	uint32_t value;
 
-	while (counter < 10) {
+	while (counter < 1000) {
 		value = dm_read_reg(cp110->base.ctx, addr);
 		if (get_reg_field_value(
 			value,
 			FBC_STATUS,
 			FBC_ENABLE_STATUS) == enabled)
 			break;
-		msleep(10);
+		udelay(100);
 		counter++;
 	}
 
-	if (counter == 10) {
+	if (counter == 1000) {
 		DC_LOG_WARNING("%s: wait counter exceeded, changes to HW not applied",
 			__func__);
 	} else {
@@ -551,9 +551,7 @@ void dce110_compressor_construct(struct dce110_compressor *compressor,
 	compressor->base.lpt_channels_num = 0;
 	compressor->base.attached_inst = 0;
 	compressor->base.is_enabled = false;
-#if defined(CONFIG_DRM_AMD_DC_FBC)
 	compressor->base.funcs = &dce110_compressor_funcs;
 
-#endif
 }
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
index d0575999f172..33a14e163f88 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c
@@ -34,9 +34,7 @@
 #include "dce/dce_hwseq.h"
 #include "gpio_service_interface.h"
 
-#if defined(CONFIG_DRM_AMD_DC_FBC)
 #include "dce110_compressor.h"
-#endif
 
 #include "bios/bios_parser_helper.h"
 #include "timing_generator.h"
@@ -70,8 +68,9 @@
 
 #define CTX \
 	hws->ctx
-#define DC_LOGGER \
-	ctx->logger
+
+#define DC_LOGGER_INIT()
+
 #define REG(reg)\
 	hws->regs->reg
 
@@ -279,7 +278,9 @@ dce110_set_input_transfer_func(struct pipe_ctx *pipe_ctx,
 	build_prescale_params(&prescale_params, plane_state);
 	ipp->funcs->ipp_program_prescale(ipp, &prescale_params);
 
-	if (plane_state->gamma_correction && dce_use_lut(plane_state->format))
+	if (plane_state->gamma_correction &&
+			!plane_state->gamma_correction->is_identity &&
+			dce_use_lut(plane_state->format))
 		ipp->funcs->ipp_program_input_lut(ipp, plane_state->gamma_correction);
 
 	if (tf == NULL) {
@@ -506,19 +507,19 @@ dce110_translate_regamma_to_hw_format(const struct dc_transfer_func *output_tf,
 	rgb_resulted[hw_points - 1].green = output_tf->tf_pts.green[start_index];
 	rgb_resulted[hw_points - 1].blue = output_tf->tf_pts.blue[start_index];
 
-	arr_points[0].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2),
-					     dal_fixed31_32_from_int(region_start));
-	arr_points[1].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2),
-					     dal_fixed31_32_from_int(region_end));
+	arr_points[0].x = dc_fixpt_pow(dc_fixpt_from_int(2),
+					     dc_fixpt_from_int(region_start));
+	arr_points[1].x = dc_fixpt_pow(dc_fixpt_from_int(2),
+					     dc_fixpt_from_int(region_end));
 
 	y_r = rgb_resulted[0].red;
 	y_g = rgb_resulted[0].green;
 	y_b = rgb_resulted[0].blue;
 
-	y1_min = dal_fixed31_32_min(y_r, dal_fixed31_32_min(y_g, y_b));
+	y1_min = dc_fixpt_min(y_r, dc_fixpt_min(y_g, y_b));
 
 	arr_points[0].y = y1_min;
-	arr_points[0].slope = dal_fixed31_32_div(arr_points[0].y,
+	arr_points[0].slope = dc_fixpt_div(arr_points[0].y,
 						 arr_points[0].x);
 
 	y_r = rgb_resulted[hw_points - 1].red;
@@ -528,21 +529,21 @@ dce110_translate_regamma_to_hw_format(const struct dc_transfer_func *output_tf,
 	/* see comment above, m_arrPoints[1].y should be the Y value for the
 	 * region end (m_numOfHwPoints), not last HW point(m_numOfHwPoints - 1)
 	 */
-	y3_max = dal_fixed31_32_max(y_r, dal_fixed31_32_max(y_g, y_b));
+	y3_max = dc_fixpt_max(y_r, dc_fixpt_max(y_g, y_b));
 
 	arr_points[1].y = y3_max;
 
-	arr_points[1].slope = dal_fixed31_32_zero;
+	arr_points[1].slope = dc_fixpt_zero;
 
 	if (output_tf->tf == TRANSFER_FUNCTION_PQ) {
 		/* for PQ, we want to have a straight line from last HW X point,
 		 * and the slope to be such that we hit 1.0 at 10000 nits.
 		 */
-		const struct fixed31_32 end_value = dal_fixed31_32_from_int(125);
+		const struct fixed31_32 end_value = dc_fixpt_from_int(125);
 
-		arr_points[1].slope = dal_fixed31_32_div(
-				dal_fixed31_32_sub(dal_fixed31_32_one, arr_points[1].y),
-				dal_fixed31_32_sub(end_value, arr_points[1].x));
+		arr_points[1].slope = dc_fixpt_div(
+				dc_fixpt_sub(dc_fixpt_one, arr_points[1].y),
+				dc_fixpt_sub(end_value, arr_points[1].x));
 	}
 
 	regamma_params->hw_points_num = hw_points;
@@ -566,16 +567,16 @@ dce110_translate_regamma_to_hw_format(const struct dc_transfer_func *output_tf,
 	i = 1;
 
 	while (i != hw_points + 1) {
-		if (dal_fixed31_32_lt(rgb_plus_1->red, rgb->red))
+		if (dc_fixpt_lt(rgb_plus_1->red, rgb->red))
 			rgb_plus_1->red = rgb->red;
-		if (dal_fixed31_32_lt(rgb_plus_1->green, rgb->green))
+		if (dc_fixpt_lt(rgb_plus_1->green, rgb->green))
 			rgb_plus_1->green = rgb->green;
-		if (dal_fixed31_32_lt(rgb_plus_1->blue, rgb->blue))
+		if (dc_fixpt_lt(rgb_plus_1->blue, rgb->blue))
 			rgb_plus_1->blue = rgb->blue;
 
-		rgb->delta_red = dal_fixed31_32_sub(rgb_plus_1->red, rgb->red);
-		rgb->delta_green = dal_fixed31_32_sub(rgb_plus_1->green, rgb->green);
-		rgb->delta_blue = dal_fixed31_32_sub(rgb_plus_1->blue, rgb->blue);
+		rgb->delta_red = dc_fixpt_sub(rgb_plus_1->red, rgb->red);
+		rgb->delta_green = dc_fixpt_sub(rgb_plus_1->green, rgb->green);
+		rgb->delta_blue = dc_fixpt_sub(rgb_plus_1->blue, rgb->blue);
 
 		++rgb_plus_1;
 		++rgb;
@@ -664,16 +665,25 @@ static enum dc_status bios_parser_crtc_source_select(
 
 void dce110_update_info_frame(struct pipe_ctx *pipe_ctx)
 {
+	bool is_hdmi;
+	bool is_dp;
+
 	ASSERT(pipe_ctx->stream);
 
 	if (pipe_ctx->stream_res.stream_enc == NULL)
 		return;  /* this is not root pipe */
 
-	if (dc_is_hdmi_signal(pipe_ctx->stream->signal))
+	is_hdmi = dc_is_hdmi_signal(pipe_ctx->stream->signal);
+	is_dp = dc_is_dp_signal(pipe_ctx->stream->signal);
+
+	if (!is_hdmi && !is_dp)
+		return;
+
+	if (is_hdmi)
 		pipe_ctx->stream_res.stream_enc->funcs->update_hdmi_info_packets(
 			pipe_ctx->stream_res.stream_enc,
 			&pipe_ctx->stream_res.encoder_info_frame);
-	else if (dc_is_dp_signal(pipe_ctx->stream->signal))
+	else
 		pipe_ctx->stream_res.stream_enc->funcs->update_dp_info_packets(
 			pipe_ctx->stream_res.stream_enc,
 			&pipe_ctx->stream_res.encoder_info_frame);
@@ -851,6 +861,33 @@ void hwss_edp_power_control(
 
 	if (power_up != is_panel_powered_on(hwseq)) {
 		/* Send VBIOS command to prompt eDP panel power */
+		if (power_up) {
+			unsigned long long current_ts = dm_get_timestamp(ctx);
+			unsigned long long duration_in_ms =
+					div64_u64(dm_get_elapse_time_in_ns(
+							ctx,
+							current_ts,
+							link->link_trace.time_stamp.edp_poweroff), 1000000);
+			unsigned long long wait_time_ms = 0;
+
+			/* max 500ms from LCDVDD off to on */
+			unsigned long long edp_poweroff_time_ms = 500;
+
+			if (link->local_sink != NULL)
+				edp_poweroff_time_ms =
+						500 + link->local_sink->edid_caps.panel_patch.extra_t12_ms;
+			if (link->link_trace.time_stamp.edp_poweroff == 0)
+				wait_time_ms = edp_poweroff_time_ms;
+			else if (duration_in_ms < edp_poweroff_time_ms)
+				wait_time_ms = edp_poweroff_time_ms - duration_in_ms;
+
+			if (wait_time_ms) {
+				msleep(wait_time_ms);
+				dm_output_to_console("%s: wait %lld ms to power on eDP.\n",
+						__func__, wait_time_ms);
+			}
+
+		}
 
 		DC_LOG_HW_RESUME_S3(
 				"%s: Panel Power action: %s\n",
@@ -864,9 +901,14 @@ void hwss_edp_power_control(
 		cntl.coherent = false;
 		cntl.lanes_number = LANE_COUNT_FOUR;
 		cntl.hpd_sel = link->link_enc->hpd_source;
-
 		bp_result = link_transmitter_control(ctx->dc_bios, &cntl);
 
+		if (!power_up)
+			/*save driver power off time stamp*/
+			link->link_trace.time_stamp.edp_poweroff = dm_get_timestamp(ctx);
+		else
+			link->link_trace.time_stamp.edp_poweron = dm_get_timestamp(ctx);
+
 		if (bp_result != BP_RESULT_OK)
 			DC_LOG_ERROR(
 					"%s: Panel Power bp_result: %d\n",
@@ -942,19 +984,35 @@ void hwss_edp_backlight_control(
 		edp_receiver_ready_T9(link);
 }
 
-void dce110_disable_stream(struct pipe_ctx *pipe_ctx, int option)
+void dce110_enable_audio_stream(struct pipe_ctx *pipe_ctx)
 {
-	struct dc_stream_state *stream = pipe_ctx->stream;
-	struct dc_link *link = stream->sink->link;
-	struct dc *dc = pipe_ctx->stream->ctx->dc;
+	struct dc *core_dc = pipe_ctx->stream->ctx->dc;
+	/* notify audio driver for audio modes of monitor */
+	struct pp_smu_funcs_rv *pp_smu = core_dc->res_pool->pp_smu;
+	unsigned int i, num_audio = 1;
 
-	if (dc_is_hdmi_signal(pipe_ctx->stream->signal))
-		pipe_ctx->stream_res.stream_enc->funcs->stop_hdmi_info_packets(
-			pipe_ctx->stream_res.stream_enc);
+	if (pipe_ctx->stream_res.audio) {
+		for (i = 0; i < MAX_PIPES; i++) {
+			/*current_state not updated yet*/
+			if (core_dc->current_state->res_ctx.pipe_ctx[i].stream_res.audio != NULL)
+				num_audio++;
+		}
 
-	if (dc_is_dp_signal(pipe_ctx->stream->signal))
-		pipe_ctx->stream_res.stream_enc->funcs->stop_dp_info_packets(
-			pipe_ctx->stream_res.stream_enc);
+		pipe_ctx->stream_res.audio->funcs->az_enable(pipe_ctx->stream_res.audio);
+
+		if (num_audio == 1 && pp_smu != NULL && pp_smu->set_pme_wa_enable != NULL)
+			/*this is the first audio. apply the PME w/a in order to wake AZ from D3*/
+			pp_smu->set_pme_wa_enable(&pp_smu->pp_smu);
+		/* un-mute audio */
+		/* TODO: audio should be per stream rather than per link */
+		pipe_ctx->stream_res.stream_enc->funcs->audio_mute_control(
+			pipe_ctx->stream_res.stream_enc, false);
+	}
+}
+
+void dce110_disable_audio_stream(struct pipe_ctx *pipe_ctx, int option)
+{
+	struct dc *dc = pipe_ctx->stream->ctx->dc;
 
 	pipe_ctx->stream_res.stream_enc->funcs->audio_mute_control(
 			pipe_ctx->stream_res.stream_enc, true);
@@ -974,9 +1032,9 @@ void dce110_disable_stream(struct pipe_ctx *pipe_ctx, int option)
 		/*don't free audio if it is from retrain or internal disable stream*/
 		if (option == FREE_ACQUIRED_RESOURCE && dc->caps.dynamic_audio == true) {
 			/*we have to dynamic arbitrate the audio endpoints*/
-			pipe_ctx->stream_res.audio = NULL;
 			/*we free the resource, need reset is_audio_acquired*/
 			update_audio_usage(&dc->current_state->res_ctx, dc->res_pool, pipe_ctx->stream_res.audio, false);
+			pipe_ctx->stream_res.audio = NULL;
 		}
 
 		/* TODO: notify audio driver for if audio modes list changed
@@ -985,7 +1043,23 @@ void dce110_disable_stream(struct pipe_ctx *pipe_ctx, int option)
 		 * stream->stream_engine_id);
 		 */
 	}
+}
 
+void dce110_disable_stream(struct pipe_ctx *pipe_ctx, int option)
+{
+	struct dc_stream_state *stream = pipe_ctx->stream;
+	struct dc_link *link = stream->sink->link;
+	struct dc *dc = pipe_ctx->stream->ctx->dc;
+
+	if (dc_is_hdmi_signal(pipe_ctx->stream->signal))
+		pipe_ctx->stream_res.stream_enc->funcs->stop_hdmi_info_packets(
+			pipe_ctx->stream_res.stream_enc);
+
+	if (dc_is_dp_signal(pipe_ctx->stream->signal))
+		pipe_ctx->stream_res.stream_enc->funcs->stop_dp_info_packets(
+			pipe_ctx->stream_res.stream_enc);
+
+	dc->hwss.disable_audio_stream(pipe_ctx, option);
 
 	link->link_enc->funcs->connect_dig_be_to_fe(
 			link->link_enc,
@@ -1011,7 +1085,7 @@ void dce110_unblank_stream(struct pipe_ctx *pipe_ctx,
 
 	if (link->local_sink && link->local_sink->sink_signal == SIGNAL_TYPE_EDP) {
 		link->dc->hwss.edp_backlight_control(link, true);
-		stream->bl_pwm_level = 0;
+		stream->bl_pwm_level = EDP_BACKLIGHT_RAMP_DISABLE_LEVEL;
 	}
 }
 void dce110_blank_stream(struct pipe_ctx *pipe_ctx)
@@ -1176,13 +1250,13 @@ static void program_scaler(const struct dc *dc,
 {
 	struct tg_color color = {0};
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	/* TOFPGA */
 	if (pipe_ctx->plane_res.xfm->funcs->transform_set_pixel_storage_depth == NULL)
 		return;
 #endif
 
-	if (dc->debug.surface_visual_confirm)
+	if (dc->debug.visual_confirm == VISUAL_CONFIRM_SURFACE)
 		get_surface_visual_confirm_color(pipe_ctx, &color);
 	else
 		color_space_to_black_color(dc,
@@ -1203,7 +1277,7 @@ static void program_scaler(const struct dc *dc,
 		&pipe_ctx->plane_res.scl_data);
 }
 
-static enum dc_status dce110_prog_pixclk_crtc_otg(
+static enum dc_status dce110_enable_stream_timing(
 		struct pipe_ctx *pipe_ctx,
 		struct dc_state *context,
 		struct dc *dc)
@@ -1268,8 +1342,32 @@ static enum dc_status apply_single_controller_ctx_to_hw(
 	struct pipe_ctx *pipe_ctx_old = &dc->current_state->res_ctx.
 			pipe_ctx[pipe_ctx->pipe_idx];
 
+	if (pipe_ctx->stream_res.audio != NULL) {
+		struct audio_output audio_output;
+
+		build_audio_output(context, pipe_ctx, &audio_output);
+
+		if (dc_is_dp_signal(pipe_ctx->stream->signal))
+			pipe_ctx->stream_res.stream_enc->funcs->dp_audio_setup(
+					pipe_ctx->stream_res.stream_enc,
+					pipe_ctx->stream_res.audio->inst,
+					&pipe_ctx->stream->audio_info);
+		else
+			pipe_ctx->stream_res.stream_enc->funcs->hdmi_audio_setup(
+					pipe_ctx->stream_res.stream_enc,
+					pipe_ctx->stream_res.audio->inst,
+					&pipe_ctx->stream->audio_info,
+					&audio_output.crtc_info);
+
+		pipe_ctx->stream_res.audio->funcs->az_configure(
+				pipe_ctx->stream_res.audio,
+				pipe_ctx->stream->signal,
+				&audio_output.crtc_info,
+				&pipe_ctx->stream->audio_info);
+	}
+
 	/*  */
-	dc->hwss.prog_pixclk_crtc_otg(pipe_ctx, context, dc);
+	dc->hwss.enable_stream_timing(pipe_ctx, context, dc);
 
 	/* FPGA does not program backend */
 	if (IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) {
@@ -1382,7 +1480,7 @@ static void power_down_controllers(struct dc *dc)
 {
 	int i;
 
-	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+	for (i = 0; i < dc->res_pool->timing_generator_count; i++) {
 		dc->res_pool->timing_generators[i]->funcs->disable_crtc(
 				dc->res_pool->timing_generators[i]);
 	}
@@ -1411,10 +1509,8 @@ static void power_down_all_hw_blocks(struct dc *dc)
 
 	power_down_clock_sources(dc);
 
-#if defined(CONFIG_DRM_AMD_DC_FBC)
 	if (dc->fbc_compressor)
 		dc->fbc_compressor->funcs->disable_fbc(dc->fbc_compressor);
-#endif
 }
 
 static void disable_vga_and_power_gate_all_controllers(
@@ -1424,12 +1520,13 @@ static void disable_vga_and_power_gate_all_controllers(
 	struct timing_generator *tg;
 	struct dc_context *ctx = dc->ctx;
 
-	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+	for (i = 0; i < dc->res_pool->timing_generator_count; i++) {
 		tg = dc->res_pool->timing_generators[i];
 
 		if (tg->funcs->disable_vga)
 			tg->funcs->disable_vga(tg);
-
+	}
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
 		/* Enable CLOCK gating for each pipe BEFORE controller
 		 * powergating. */
 		enable_display_pipe_clock_gating(ctx,
@@ -1441,6 +1538,17 @@ static void disable_vga_and_power_gate_all_controllers(
 	}
 }
 
+static struct dc_link *get_link_for_edp(struct dc *dc)
+{
+	int i;
+
+	for (i = 0; i < dc->link_count; i++) {
+		if (dc->links[i]->connector_signal == SIGNAL_TYPE_EDP)
+			return dc->links[i];
+	}
+	return NULL;
+}
+
 static struct dc_link *get_link_for_edp_not_in_use(
 		struct dc *dc,
 		struct dc_state *context)
@@ -1475,20 +1583,21 @@ static struct dc_link *get_link_for_edp_not_in_use(
  */
 void dce110_enable_accelerated_mode(struct dc *dc, struct dc_state *context)
 {
-	struct dc_bios *dcb = dc->ctx->dc_bios;
-
-	/* vbios already light up eDP, so we can leverage vbios and skip eDP
-	 * programming
-	 */
-	bool can_eDP_fast_boot_optimize =
-			(dcb->funcs->get_vga_enabled_displays(dc->ctx->dc_bios) == ATOM_DISPLAY_LCD1_ACTIVE);
-
-	/* if OS doesn't light up eDP and eDP link is available, we want to disable */
 	struct dc_link *edp_link_to_turnoff = NULL;
+	struct dc_link *edp_link = get_link_for_edp(dc);
+	bool can_eDP_fast_boot_optimize = false;
+
+	if (edp_link) {
+		can_eDP_fast_boot_optimize =
+				edp_link->link_enc->funcs->is_dig_enabled(edp_link->link_enc);
+	}
 
 	if (can_eDP_fast_boot_optimize) {
 		edp_link_to_turnoff = get_link_for_edp_not_in_use(dc, context);
 
+		/* if OS doesn't light up eDP and eDP link is available, we want to disable
+		 * If resume from S4/S5, should optimization.
+		 */
 		if (!edp_link_to_turnoff)
 			dc->apply_edp_fast_boot_optimization = true;
 	}
@@ -1544,6 +1653,7 @@ static void dce110_set_displaymarks(
 			pipe_ctx->plane_res.mi,
 			context->bw.dce.nbp_state_change_wm_ns[num_pipes],
 			context->bw.dce.stutter_exit_wm_ns[num_pipes],
+			context->bw.dce.stutter_entry_wm_ns[num_pipes],
 			context->bw.dce.urgent_wm_ns[num_pipes],
 			total_dest_line_time_ns);
 		if (i == underlay_idx) {
@@ -1559,7 +1669,7 @@ static void dce110_set_displaymarks(
 	}
 }
 
-static void set_safe_displaymarks(
+void dce110_set_safe_displaymarks(
 		struct resource_context *res_ctx,
 		const struct resource_pool *pool)
 {
@@ -1569,6 +1679,7 @@ static void set_safe_displaymarks(
 		MAX_WATERMARK, MAX_WATERMARK, MAX_WATERMARK, MAX_WATERMARK };
 	struct dce_watermarks nbp_marks = {
 		SAFE_NBP_MARK, SAFE_NBP_MARK, SAFE_NBP_MARK, SAFE_NBP_MARK };
+	struct dce_watermarks min_marks = { 0, 0, 0, 0};
 
 	for (i = 0; i < MAX_PIPES; i++) {
 		if (res_ctx->pipe_ctx[i].stream == NULL || res_ctx->pipe_ctx[i].plane_res.mi == NULL)
@@ -1578,6 +1689,7 @@ static void set_safe_displaymarks(
 				res_ctx->pipe_ctx[i].plane_res.mi,
 				nbp_marks,
 				max_marks,
+				min_marks,
 				max_marks,
 				MAX_WATERMARK);
 
@@ -1641,9 +1753,7 @@ static void set_static_screen_control(struct pipe_ctx **pipe_ctx,
 	if (events->force_trigger)
 		value |= 0x1;
 
-#if defined(CONFIG_DRM_AMD_DC_FBC)
 	value |= 0x84;
-#endif
 
 	for (i = 0; i < num_pipes; i++)
 		pipe_ctx[i]->stream_res.tg->funcs->
@@ -1651,23 +1761,15 @@ static void set_static_screen_control(struct pipe_ctx **pipe_ctx,
 }
 
 /* unit: in_khz before mode set, get pixel clock from context. ASIC register
- * may not be programmed yet.
- * TODO: after mode set, pre_mode_set = false,
- * may read PLL register to get pixel clock
+ * may not be programmed yet
  */
 static uint32_t get_max_pixel_clock_for_all_paths(
 	struct dc *dc,
-	struct dc_state *context,
-	bool pre_mode_set)
+	struct dc_state *context)
 {
 	uint32_t max_pix_clk = 0;
 	int i;
 
-	if (!pre_mode_set) {
-		/* TODO: read ASIC register to get pixel clock */
-		ASSERT(0);
-	}
-
 	for (i = 0; i < MAX_PIPES; i++) {
 		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
 
@@ -1683,97 +1785,10 @@ static uint32_t get_max_pixel_clock_for_all_paths(
 				pipe_ctx->stream_res.pix_clk_params.requested_pix_clk;
 	}
 
-	if (max_pix_clk == 0)
-		ASSERT(0);
-
 	return max_pix_clk;
 }
 
 /*
- * Find clock state based on clock requested. if clock value is 0, simply
- * set clock state as requested without finding clock state by clock value
- */
-
-static void apply_min_clocks(
-	struct dc *dc,
-	struct dc_state *context,
-	enum dm_pp_clocks_state *clocks_state,
-	bool pre_mode_set)
-{
-	struct state_dependent_clocks req_clocks = {0};
-
-	if (!pre_mode_set) {
-		/* set clock_state without verification */
-		if (context->dis_clk->funcs->set_min_clocks_state) {
-			context->dis_clk->funcs->set_min_clocks_state(
-						context->dis_clk, *clocks_state);
-			return;
-		}
-
-		/* TODO: This is incorrect. Figure out how to fix. */
-		context->dis_clk->funcs->apply_clock_voltage_request(
-				context->dis_clk,
-				DM_PP_CLOCK_TYPE_DISPLAY_CLK,
-				context->dis_clk->cur_clocks_value.dispclk_in_khz,
-				pre_mode_set,
-				false);
-
-		context->dis_clk->funcs->apply_clock_voltage_request(
-				context->dis_clk,
-				DM_PP_CLOCK_TYPE_PIXELCLK,
-				context->dis_clk->cur_clocks_value.max_pixelclk_in_khz,
-				pre_mode_set,
-				false);
-
-		context->dis_clk->funcs->apply_clock_voltage_request(
-				context->dis_clk,
-				DM_PP_CLOCK_TYPE_DISPLAYPHYCLK,
-				context->dis_clk->cur_clocks_value.max_non_dp_phyclk_in_khz,
-				pre_mode_set,
-				false);
-		return;
-	}
-
-	/* get the required state based on state dependent clocks:
-	 * display clock and pixel clock
-	 */
-	req_clocks.display_clk_khz = context->bw.dce.dispclk_khz;
-
-	req_clocks.pixel_clk_khz = get_max_pixel_clock_for_all_paths(
-			dc, context, true);
-
-	if (context->dis_clk->funcs->get_required_clocks_state) {
-		*clocks_state = context->dis_clk->funcs->get_required_clocks_state(
-				context->dis_clk, &req_clocks);
-		context->dis_clk->funcs->set_min_clocks_state(
-			context->dis_clk, *clocks_state);
-	} else {
-		context->dis_clk->funcs->apply_clock_voltage_request(
-				context->dis_clk,
-				DM_PP_CLOCK_TYPE_DISPLAY_CLK,
-				req_clocks.display_clk_khz,
-				pre_mode_set,
-				false);
-
-		context->dis_clk->funcs->apply_clock_voltage_request(
-				context->dis_clk,
-				DM_PP_CLOCK_TYPE_PIXELCLK,
-				req_clocks.pixel_clk_khz,
-				pre_mode_set,
-				false);
-
-		context->dis_clk->funcs->apply_clock_voltage_request(
-				context->dis_clk,
-				DM_PP_CLOCK_TYPE_DISPLAYPHYCLK,
-				req_clocks.pixel_clk_khz,
-				pre_mode_set,
-				false);
-	}
-}
-
-#if defined(CONFIG_DRM_AMD_DC_FBC)
-
-/*
  *  Check if FBC can be enabled
  */
 static bool should_enable_fbc(struct dc *dc,
@@ -1803,6 +1818,9 @@ static bool should_enable_fbc(struct dc *dc,
 		}
 	}
 
+	/* Pipe context should be found */
+	ASSERT(pipe_ctx);
+
 	/* Only supports eDP */
 	if (pipe_ctx->stream->sink->link->connector_signal != SIGNAL_TYPE_EDP)
 		return false;
@@ -1848,7 +1866,6 @@ static void enable_fbc(struct dc *dc,
 		compr->funcs->enable_fbc(compr, &params);
 	}
 }
-#endif
 
 static void dce110_reset_hw_ctx_wrap(
 		struct dc *dc,
@@ -1901,97 +1918,12 @@ static void dce110_reset_hw_ctx_wrap(
 	}
 }
 
-
-enum dc_status dce110_apply_ctx_to_hw(
+static void dce110_setup_audio_dto(
 		struct dc *dc,
 		struct dc_state *context)
 {
-	struct dc_bios *dcb = dc->ctx->dc_bios;
-	enum dc_status status;
 	int i;
-	enum dm_pp_clocks_state clocks_state = DM_PP_CLOCKS_STATE_INVALID;
-
-	/* Reset old context */
-	/* look up the targets that have been removed since last commit */
-	dc->hwss.reset_hw_ctx_wrap(dc, context);
-
-	/* Skip applying if no targets */
-	if (context->stream_count <= 0)
-		return DC_OK;
-
-	/* Apply new context */
-	dcb->funcs->set_scratch_critical_state(dcb, true);
-
-	/* below is for real asic only */
-	for (i = 0; i < dc->res_pool->pipe_count; i++) {
-		struct pipe_ctx *pipe_ctx_old =
-					&dc->current_state->res_ctx.pipe_ctx[i];
-		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
-
-		if (pipe_ctx->stream == NULL || pipe_ctx->top_pipe)
-			continue;
-
-		if (pipe_ctx->stream == pipe_ctx_old->stream) {
-			if (pipe_ctx_old->clock_source != pipe_ctx->clock_source)
-				dce_crtc_switch_to_clk_src(dc->hwseq,
-						pipe_ctx->clock_source, i);
-			continue;
-		}
-
-		dc->hwss.enable_display_power_gating(
-				dc, i, dc->ctx->dc_bios,
-				PIPE_GATING_CONTROL_DISABLE);
-	}
-
-	set_safe_displaymarks(&context->res_ctx, dc->res_pool);
 
-#if defined(CONFIG_DRM_AMD_DC_FBC)
-	if (dc->fbc_compressor)
-		dc->fbc_compressor->funcs->disable_fbc(dc->fbc_compressor);
-#endif
-	/*TODO: when pplib works*/
-	apply_min_clocks(dc, context, &clocks_state, true);
-
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
-	if (dc->ctx->dce_version >= DCN_VERSION_1_0) {
-		if (context->bw.dcn.calc_clk.fclk_khz
-				> dc->current_state->bw.dcn.cur_clk.fclk_khz) {
-			struct dm_pp_clock_for_voltage_req clock;
-
-			clock.clk_type = DM_PP_CLOCK_TYPE_FCLK;
-			clock.clocks_in_khz = context->bw.dcn.calc_clk.fclk_khz;
-			dm_pp_apply_clock_for_voltage_request(dc->ctx, &clock);
-			dc->current_state->bw.dcn.cur_clk.fclk_khz = clock.clocks_in_khz;
-			context->bw.dcn.cur_clk.fclk_khz = clock.clocks_in_khz;
-		}
-		if (context->bw.dcn.calc_clk.dcfclk_khz
-				> dc->current_state->bw.dcn.cur_clk.dcfclk_khz) {
-			struct dm_pp_clock_for_voltage_req clock;
-
-			clock.clk_type = DM_PP_CLOCK_TYPE_DCFCLK;
-			clock.clocks_in_khz = context->bw.dcn.calc_clk.dcfclk_khz;
-			dm_pp_apply_clock_for_voltage_request(dc->ctx, &clock);
-			dc->current_state->bw.dcn.cur_clk.dcfclk_khz = clock.clocks_in_khz;
-			context->bw.dcn.cur_clk.dcfclk_khz = clock.clocks_in_khz;
-		}
-		if (context->bw.dcn.calc_clk.dispclk_khz
-				> dc->current_state->bw.dcn.cur_clk.dispclk_khz) {
-			dc->res_pool->display_clock->funcs->set_clock(
-					dc->res_pool->display_clock,
-					context->bw.dcn.calc_clk.dispclk_khz);
-			dc->current_state->bw.dcn.cur_clk.dispclk_khz =
-					context->bw.dcn.calc_clk.dispclk_khz;
-			context->bw.dcn.cur_clk.dispclk_khz =
-					context->bw.dcn.calc_clk.dispclk_khz;
-		}
-	} else
-#endif
-	if (context->bw.dce.dispclk_khz
-			> dc->current_state->bw.dce.dispclk_khz) {
-		dc->res_pool->display_clock->funcs->set_clock(
-				dc->res_pool->display_clock,
-				context->bw.dce.dispclk_khz * 115 / 100);
-	}
 	/* program audio wall clock. use HDMI as clock source if HDMI
 	 * audio active. Otherwise, use DP as clock source
 	 * first, loop to find any HDMI audio, if not, loop find DP audio
@@ -2065,6 +1997,52 @@ enum dc_status dce110_apply_ctx_to_hw(
 			}
 		}
 	}
+}
+
+enum dc_status dce110_apply_ctx_to_hw(
+		struct dc *dc,
+		struct dc_state *context)
+{
+	struct dc_bios *dcb = dc->ctx->dc_bios;
+	enum dc_status status;
+	int i;
+
+	/* Reset old context */
+	/* look up the targets that have been removed since last commit */
+	dc->hwss.reset_hw_ctx_wrap(dc, context);
+
+	/* Skip applying if no targets */
+	if (context->stream_count <= 0)
+		return DC_OK;
+
+	/* Apply new context */
+	dcb->funcs->set_scratch_critical_state(dcb, true);
+
+	/* below is for real asic only */
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx_old =
+					&dc->current_state->res_ctx.pipe_ctx[i];
+		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
+
+		if (pipe_ctx->stream == NULL || pipe_ctx->top_pipe)
+			continue;
+
+		if (pipe_ctx->stream == pipe_ctx_old->stream) {
+			if (pipe_ctx_old->clock_source != pipe_ctx->clock_source)
+				dce_crtc_switch_to_clk_src(dc->hwseq,
+						pipe_ctx->clock_source, i);
+			continue;
+		}
+
+		dc->hwss.enable_display_power_gating(
+				dc, i, dc->ctx->dc_bios,
+				PIPE_GATING_CONTROL_DISABLE);
+	}
+
+	if (dc->fbc_compressor)
+		dc->fbc_compressor->funcs->disable_fbc(dc->fbc_compressor);
+
+	dce110_setup_audio_dto(dc, context);
 
 	for (i = 0; i < dc->res_pool->pipe_count; i++) {
 		struct pipe_ctx *pipe_ctx_old =
@@ -2083,31 +2061,6 @@ enum dc_status dce110_apply_ctx_to_hw(
 		if (pipe_ctx->top_pipe)
 			continue;
 
-		if (context->res_ctx.pipe_ctx[i].stream_res.audio != NULL) {
-
-			struct audio_output audio_output;
-
-			build_audio_output(context, pipe_ctx, &audio_output);
-
-			if (dc_is_dp_signal(pipe_ctx->stream->signal))
-				pipe_ctx->stream_res.stream_enc->funcs->dp_audio_setup(
-						pipe_ctx->stream_res.stream_enc,
-						pipe_ctx->stream_res.audio->inst,
-						&pipe_ctx->stream->audio_info);
-			else
-				pipe_ctx->stream_res.stream_enc->funcs->hdmi_audio_setup(
-						pipe_ctx->stream_res.stream_enc,
-						pipe_ctx->stream_res.audio->inst,
-						&pipe_ctx->stream->audio_info,
-						&audio_output.crtc_info);
-
-			pipe_ctx->stream_res.audio->funcs->az_configure(
-					pipe_ctx->stream_res.audio,
-					pipe_ctx->stream->signal,
-					&audio_output.crtc_info,
-					&pipe_ctx->stream->audio_info);
-		}
-
 		status = apply_single_controller_ctx_to_hw(
 				pipe_ctx,
 				context,
@@ -2117,17 +2070,11 @@ enum dc_status dce110_apply_ctx_to_hw(
 			return status;
 	}
 
-	/* to save power */
-	apply_min_clocks(dc, context, &clocks_state, false);
-
 	dcb->funcs->set_scratch_critical_state(dcb, false);
 
-#if defined(CONFIG_DRM_AMD_DC_FBC)
 	if (dc->fbc_compressor)
 		enable_fbc(dc, context);
 
-#endif
-
 	return DC_OK;
 }
 
@@ -2221,74 +2168,6 @@ static void program_gamut_remap(struct pipe_ctx *pipe_ctx)
 
 	pipe_ctx->plane_res.xfm->funcs->transform_set_gamut_remap(pipe_ctx->plane_res.xfm, &adjust);
 }
-
-/**
- * TODO REMOVE, USE UPDATE INSTEAD
- */
-static void set_plane_config(
-	const struct dc *dc,
-	struct pipe_ctx *pipe_ctx,
-	struct resource_context *res_ctx)
-{
-	struct mem_input *mi = pipe_ctx->plane_res.mi;
-	struct dc_plane_state *plane_state = pipe_ctx->plane_state;
-	struct xfm_grph_csc_adjustment adjust;
-	struct out_csc_color_matrix tbl_entry;
-	unsigned int i;
-
-	memset(&adjust, 0, sizeof(adjust));
-	memset(&tbl_entry, 0, sizeof(tbl_entry));
-	adjust.gamut_adjust_type = GRAPHICS_GAMUT_ADJUST_TYPE_BYPASS;
-
-	dce_enable_fe_clock(dc->hwseq, mi->inst, true);
-
-	set_default_colors(pipe_ctx);
-	if (pipe_ctx->stream->csc_color_matrix.enable_adjustment == true) {
-		tbl_entry.color_space =
-			pipe_ctx->stream->output_color_space;
-
-		for (i = 0; i < 12; i++)
-			tbl_entry.regval[i] =
-			pipe_ctx->stream->csc_color_matrix.matrix[i];
-
-		pipe_ctx->plane_res.xfm->funcs->opp_set_csc_adjustment
-				(pipe_ctx->plane_res.xfm, &tbl_entry);
-	}
-
-	if (pipe_ctx->stream->gamut_remap_matrix.enable_remap == true) {
-		adjust.gamut_adjust_type = GRAPHICS_GAMUT_ADJUST_TYPE_SW;
-
-		for (i = 0; i < CSC_TEMPERATURE_MATRIX_SIZE; i++)
-			adjust.temperature_matrix[i] =
-				pipe_ctx->stream->gamut_remap_matrix.matrix[i];
-	}
-
-	pipe_ctx->plane_res.xfm->funcs->transform_set_gamut_remap(pipe_ctx->plane_res.xfm, &adjust);
-
-	pipe_ctx->plane_res.scl_data.lb_params.alpha_en = pipe_ctx->bottom_pipe != 0;
-	program_scaler(dc, pipe_ctx);
-
-	program_surface_visibility(dc, pipe_ctx);
-
-	mi->funcs->mem_input_program_surface_config(
-			mi,
-			plane_state->format,
-			&plane_state->tiling_info,
-			&plane_state->plane_size,
-			plane_state->rotation,
-			NULL,
-			false);
-	if (mi->funcs->set_blank)
-		mi->funcs->set_blank(mi, pipe_ctx->plane_state->visible);
-
-	if (dc->config.gpu_vm_support)
-		mi->funcs->mem_input_program_pte_vm(
-				pipe_ctx->plane_res.mi,
-				plane_state->format,
-				&plane_state->tiling_info,
-				plane_state->rotation);
-}
-
 static void update_plane_addr(const struct dc *dc,
 		struct pipe_ctx *pipe_ctx)
 {
@@ -2510,10 +2389,9 @@ static void init_hw(struct dc *dc)
 		abm->funcs->init_backlight(abm);
 		abm->funcs->abm_init(abm);
 	}
-#if defined(CONFIG_DRM_AMD_DC_FBC)
+
 	if (dc->fbc_compressor)
 		dc->fbc_compressor->funcs->power_up_fbc(dc->fbc_compressor);
-#endif
 
 }
 
@@ -2679,15 +2557,20 @@ static void dce110_set_bandwidth(
 		struct dc_state *context,
 		bool decrease_allowed)
 {
-	dce110_set_displaymarks(dc, context);
+	struct dc_clocks req_clks;
 
-	if (decrease_allowed || context->bw.dce.dispclk_khz > dc->current_state->bw.dce.dispclk_khz) {
-		dc->res_pool->display_clock->funcs->set_clock(
-				dc->res_pool->display_clock,
-				context->bw.dce.dispclk_khz * 115 / 100);
-		dc->current_state->bw.dce.dispclk_khz = context->bw.dce.dispclk_khz;
-	}
+	req_clks.dispclk_khz = context->bw.dce.dispclk_khz * 115 / 100;
+	req_clks.phyclk_khz = get_max_pixel_clock_for_all_paths(dc, context);
+
+	if (decrease_allowed)
+		dce110_set_displaymarks(dc, context);
+	else
+		dce110_set_safe_displaymarks(&context->res_ctx, dc->res_pool);
 
+	dc->res_pool->dccg->funcs->update_clocks(
+			dc->res_pool->dccg,
+			&req_clks,
+			decrease_allowed);
 	pplib_apply_display_requirements(dc, context);
 }
 
@@ -2699,8 +2582,9 @@ static void dce110_program_front_end_for_pipe(
 	struct dc_plane_state *plane_state = pipe_ctx->plane_state;
 	struct xfm_grph_csc_adjustment adjust;
 	struct out_csc_color_matrix tbl_entry;
+	unsigned int underlay_idx = dc->res_pool->underlay_pipe_index;
 	unsigned int i;
-	struct dc_context *ctx = dc->ctx;
+	DC_LOGGER_INIT();
 	memset(&tbl_entry, 0, sizeof(tbl_entry));
 
 	if (dc->current_state)
@@ -2739,14 +2623,14 @@ static void dce110_program_front_end_for_pipe(
 
 	program_scaler(dc, pipe_ctx);
 
-#if defined(CONFIG_DRM_AMD_DC_FBC)
-	if (dc->fbc_compressor && old_pipe->stream) {
+	/* fbc not applicable on Underlay pipe */
+	if (dc->fbc_compressor && old_pipe->stream &&
+	    pipe_ctx->pipe_idx != underlay_idx) {
 		if (plane_state->tiling_info.gfx8.array_mode == DC_ARRAY_LINEAR_GENERAL)
 			dc->fbc_compressor->funcs->disable_fbc(dc->fbc_compressor);
 		else
 			enable_fbc(dc, dc->current_state);
 	}
-#endif
 
 	mi->funcs->mem_input_program_surface_config(
 			mi,
@@ -2776,13 +2660,13 @@ static void dce110_program_front_end_for_pipe(
 		dc->hwss.set_output_transfer_func(pipe_ctx, pipe_ctx->stream);
 
 	DC_LOG_SURFACE(
-			"Pipe:%d 0x%x: addr hi:0x%x, "
+			"Pipe:%d %p: addr hi:0x%x, "
 			"addr low:0x%x, "
 			"src: %d, %d, %d,"
 			" %d; dst: %d, %d, %d, %d;"
 			"clip: %d, %d, %d, %d\n",
 			pipe_ctx->pipe_idx,
-			pipe_ctx->plane_state,
+			(void *) pipe_ctx->plane_state,
 			pipe_ctx->plane_state->address.grph.addr.high_part,
 			pipe_ctx->plane_state->address.grph.addr.low_part,
 			pipe_ctx->plane_state->src_rect.x,
@@ -2922,9 +2806,11 @@ void dce110_set_cursor_position(struct pipe_ctx *pipe_ctx)
 	struct dc_cursor_mi_param param = {
 		.pixel_clk_khz = pipe_ctx->stream->timing.pix_clk_khz,
 		.ref_clk_khz = pipe_ctx->stream->ctx->dc->res_pool->ref_clock_inKhz,
-		.viewport_x_start = pipe_ctx->plane_res.scl_data.viewport.x,
-		.viewport_width = pipe_ctx->plane_res.scl_data.viewport.width,
-		.h_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.horz
+		.viewport = pipe_ctx->plane_res.scl_data.viewport,
+		.h_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.horz,
+		.v_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.vert,
+		.rotation = pipe_ctx->plane_state->rotation,
+		.mirror = pipe_ctx->plane_state->horizontal_mirror
 	};
 
 	if (pipe_ctx->plane_state->address.type
@@ -2970,7 +2856,6 @@ static const struct hw_sequencer_funcs dce110_funcs = {
 	.init_hw = init_hw,
 	.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
 	.apply_ctx_for_surface = dce110_apply_ctx_for_surface,
-	.set_plane_config = set_plane_config,
 	.update_plane_addr = update_plane_addr,
 	.update_pending_status = dce110_update_pending_status,
 	.set_input_transfer_func = dce110_set_input_transfer_func,
@@ -2984,6 +2869,8 @@ static const struct hw_sequencer_funcs dce110_funcs = {
 	.disable_stream = dce110_disable_stream,
 	.unblank_stream = dce110_unblank_stream,
 	.blank_stream = dce110_blank_stream,
+	.enable_audio_stream = dce110_enable_audio_stream,
+	.disable_audio_stream = dce110_disable_audio_stream,
 	.enable_display_pipe_clock_gating = enable_display_pipe_clock_gating,
 	.enable_display_power_gating = dce110_enable_display_power_gating,
 	.disable_plane = dce110_power_down_fe,
@@ -2993,7 +2880,7 @@ static const struct hw_sequencer_funcs dce110_funcs = {
 	.get_position = get_position,
 	.set_static_screen_control = set_static_screen_control,
 	.reset_hw_ctx_wrap = dce110_reset_hw_ctx_wrap,
-	.prog_pixclk_crtc_otg = dce110_prog_pixclk_crtc_otg,
+	.enable_stream_timing = dce110_enable_stream_timing,
 	.setup_stereo = NULL,
 	.set_avmute = dce110_set_avmute,
 	.wait_for_mpcc_disconnect = dce110_wait_for_mpcc_disconnect,
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.h
index 5d7e9f516827..d6db3dbd9015 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.h
@@ -49,6 +49,10 @@ void dce110_unblank_stream(struct pipe_ctx *pipe_ctx,
 		struct dc_link_settings *link_settings);
 
 void dce110_blank_stream(struct pipe_ctx *pipe_ctx);
+
+void dce110_enable_audio_stream(struct pipe_ctx *pipe_ctx);
+void dce110_disable_audio_stream(struct pipe_ctx *pipe_ctx, int option);
+
 void dce110_update_info_frame(struct pipe_ctx *pipe_ctx);
 
 void dce110_set_avmute(struct pipe_ctx *pipe_ctx, bool enable);
@@ -56,6 +60,10 @@ void dce110_enable_accelerated_mode(struct dc *dc, struct dc_state *context);
 
 void dce110_power_down(struct dc *dc);
 
+void dce110_set_safe_displaymarks(
+		struct resource_context *res_ctx,
+		const struct resource_pool *pool);
+
 void dce110_fill_display_configs(
 	const struct dc_state *context,
 	struct dm_pp_display_configuration *pp_display_cfg);
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c
index 7bab8c6d2a73..9b9fc3d96c07 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c
@@ -923,6 +923,7 @@ void dce_mem_input_v_program_display_marks(
 	struct mem_input *mem_input,
 	struct dce_watermarks nbp,
 	struct dce_watermarks stutter,
+	struct dce_watermarks stutter_enter,
 	struct dce_watermarks urgent,
 	uint32_t total_dest_line_time_ns)
 {
@@ -1010,7 +1011,7 @@ void dce110_free_mem_input_v(
 {
 }
 
-static struct mem_input_funcs dce110_mem_input_v_funcs = {
+static const struct mem_input_funcs dce110_mem_input_v_funcs = {
 	.mem_input_program_display_marks =
 			dce_mem_input_v_program_display_marks,
 	.mem_input_program_chroma_display_marks =
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c
index b1f14be20fdf..1c902e49a712 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_resource.c
@@ -54,9 +54,8 @@
 
 #define DC_LOGGER \
 		dc->ctx->logger
-#if defined(CONFIG_DRM_AMD_DC_FBC)
+
 #include "dce110/dce110_compressor.h"
-#endif
 
 #include "reg_helper.h"
 
@@ -147,15 +146,15 @@ static const struct dce110_timing_generator_offsets dce110_tg_offsets[] = {
 #define SRI(reg_name, block, id)\
 	.reg_name = mm ## block ## id ## _ ## reg_name
 
-static const struct dce_disp_clk_registers disp_clk_regs = {
+static const struct dccg_registers disp_clk_regs = {
 		CLK_COMMON_REG_LIST_DCE_BASE()
 };
 
-static const struct dce_disp_clk_shift disp_clk_shift = {
+static const struct dccg_shift disp_clk_shift = {
 		CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(__SHIFT)
 };
 
-static const struct dce_disp_clk_mask disp_clk_mask = {
+static const struct dccg_mask disp_clk_mask = {
 		CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(_MASK)
 };
 
@@ -680,8 +679,8 @@ static void destruct(struct dce110_resource_pool *pool)
 	if (pool->base.dmcu != NULL)
 		dce_dmcu_destroy(&pool->base.dmcu);
 
-	if (pool->base.display_clock != NULL)
-		dce_disp_clk_destroy(&pool->base.display_clock);
+	if (pool->base.dccg != NULL)
+		dce_dccg_destroy(&pool->base.dccg);
 
 	if (pool->base.irqs != NULL) {
 		dal_irq_service_destroy(&pool->base.irqs);
@@ -795,43 +794,38 @@ static bool dce110_validate_bandwidth(
 
 	if (memcmp(&dc->current_state->bw.dce,
 			&context->bw.dce, sizeof(context->bw.dce))) {
-		struct log_entry log_entry;
-		dm_logger_open(
-			dc->ctx->logger,
-			&log_entry,
-			LOG_BANDWIDTH_CALCS);
-		dm_logger_append(&log_entry, "%s: finish,\n"
+
+		DC_LOG_BANDWIDTH_CALCS(
+			"%s: finish,\n"
+			"nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n"
+			"stutMark_b: %d stutMark_a: %d\n"
 			"nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n"
-			"stutMark_b: %d stutMark_a: %d\n",
+			"stutMark_b: %d stutMark_a: %d\n"
+			"nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n"
+			"stutMark_b: %d stutMark_a: %d stutter_mode_enable: %d\n"
+			"cstate: %d pstate: %d nbpstate: %d sync: %d dispclk: %d\n"
+			"sclk: %d sclk_sleep: %d yclk: %d blackout_recovery_time_us: %d\n"
+			,
 			__func__,
 			context->bw.dce.nbp_state_change_wm_ns[0].b_mark,
 			context->bw.dce.nbp_state_change_wm_ns[0].a_mark,
 			context->bw.dce.urgent_wm_ns[0].b_mark,
 			context->bw.dce.urgent_wm_ns[0].a_mark,
 			context->bw.dce.stutter_exit_wm_ns[0].b_mark,
-			context->bw.dce.stutter_exit_wm_ns[0].a_mark);
-		dm_logger_append(&log_entry,
-			"nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n"
-			"stutMark_b: %d stutMark_a: %d\n",
+			context->bw.dce.stutter_exit_wm_ns[0].a_mark,
 			context->bw.dce.nbp_state_change_wm_ns[1].b_mark,
 			context->bw.dce.nbp_state_change_wm_ns[1].a_mark,
 			context->bw.dce.urgent_wm_ns[1].b_mark,
 			context->bw.dce.urgent_wm_ns[1].a_mark,
 			context->bw.dce.stutter_exit_wm_ns[1].b_mark,
-			context->bw.dce.stutter_exit_wm_ns[1].a_mark);
-		dm_logger_append(&log_entry,
-			"nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n"
-			"stutMark_b: %d stutMark_a: %d stutter_mode_enable: %d\n",
+			context->bw.dce.stutter_exit_wm_ns[1].a_mark,
 			context->bw.dce.nbp_state_change_wm_ns[2].b_mark,
 			context->bw.dce.nbp_state_change_wm_ns[2].a_mark,
 			context->bw.dce.urgent_wm_ns[2].b_mark,
 			context->bw.dce.urgent_wm_ns[2].a_mark,
 			context->bw.dce.stutter_exit_wm_ns[2].b_mark,
 			context->bw.dce.stutter_exit_wm_ns[2].a_mark,
-			context->bw.dce.stutter_mode_enable);
-		dm_logger_append(&log_entry,
-			"cstate: %d pstate: %d nbpstate: %d sync: %d dispclk: %d\n"
-			"sclk: %d sclk_sleep: %d yclk: %d blackout_recovery_time_us: %d\n",
+			context->bw.dce.stutter_mode_enable,
 			context->bw.dce.cpuc_state_change_enable,
 			context->bw.dce.cpup_state_change_enable,
 			context->bw.dce.nbp_state_change_enable,
@@ -841,7 +835,6 @@ static bool dce110_validate_bandwidth(
 			context->bw.dce.sclk_deep_sleep_khz,
 			context->bw.dce.yclk_khz,
 			context->bw.dce.blackout_recovery_time_us);
-		dm_logger_close(&log_entry);
 	}
 	return result;
 }
@@ -930,38 +923,6 @@ static enum dc_status dce110_add_stream_to_ctx(
 	return result;
 }
 
-static enum dc_status dce110_validate_guaranteed(
-		struct dc *dc,
-		struct dc_stream_state *dc_stream,
-		struct dc_state *context)
-{
-	enum dc_status result = DC_ERROR_UNEXPECTED;
-
-	context->streams[0] = dc_stream;
-	dc_stream_retain(context->streams[0]);
-	context->stream_count++;
-
-	result = resource_map_pool_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = resource_map_clock_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = build_mapped_resource(dc, context, dc_stream);
-
-	if (result == DC_OK) {
-		validate_guaranteed_copy_streams(
-				context, dc->caps.max_streams);
-		result = resource_build_scaling_params_for_context(dc, context);
-	}
-
-	if (result == DC_OK)
-		if (!dce110_validate_bandwidth(dc, context))
-			result = DC_FAIL_BANDWIDTH_VALIDATE;
-
-	return result;
-}
-
 static struct pipe_ctx *dce110_acquire_underlay(
 		struct dc_state *context,
 		const struct resource_pool *pool,
@@ -1036,7 +997,6 @@ static void dce110_destroy_resource_pool(struct resource_pool **pool)
 static const struct resource_funcs dce110_res_pool_funcs = {
 	.destroy = dce110_destroy_resource_pool,
 	.link_enc_create = dce110_link_encoder_create,
-	.validate_guaranteed = dce110_validate_guaranteed,
 	.validate_bandwidth = dce110_validate_bandwidth,
 	.validate_plane = dce110_validate_plane,
 	.acquire_idle_pipe_for_layer = dce110_acquire_underlay,
@@ -1213,11 +1173,11 @@ static bool construct(
 		}
 	}
 
-	pool->base.display_clock = dce110_disp_clk_create(ctx,
+	pool->base.dccg = dce110_dccg_create(ctx,
 			&disp_clk_regs,
 			&disp_clk_shift,
 			&disp_clk_mask);
-	if (pool->base.display_clock == NULL) {
+	if (pool->base.dccg == NULL) {
 		dm_error("DC: failed to create display clock!\n");
 		BREAK_TO_DEBUGGER();
 		goto res_create_fail;
@@ -1247,7 +1207,7 @@ static bool construct(
 	 * max_clock_state
 	 */
 	if (dm_pp_get_static_clocks(ctx, &static_clk_info))
-		pool->base.display_clock->max_clks_state =
+		pool->base.dccg->max_clks_state =
 				static_clk_info.max_clocks_state;
 
 	{
@@ -1300,12 +1260,8 @@ static bool construct(
 		}
 	}
 
-#if defined(CONFIG_DRM_AMD_DC_FBC)
 	dc->fbc_compressor = dce110_compressor_create(ctx);
 
-
-
-#endif
 	if (!underlay_create(ctx, &pool->base))
 		goto res_create_fail;
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c
index be7153924a70..1b2fe0df347f 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c
@@ -431,14 +431,6 @@ void dce110_timing_generator_set_drr(
 			0,
 			CRTC_V_TOTAL_CONTROL,
 			CRTC_SET_V_TOTAL_MIN_MASK);
-		set_reg_field_value(v_total_min,
-				0,
-				CRTC_V_TOTAL_MIN,
-				CRTC_V_TOTAL_MIN);
-		set_reg_field_value(v_total_max,
-				0,
-				CRTC_V_TOTAL_MAX,
-				CRTC_V_TOTAL_MAX);
 		set_reg_field_value(v_total_cntl,
 				0,
 				CRTC_V_TOTAL_CONTROL,
@@ -447,6 +439,14 @@ void dce110_timing_generator_set_drr(
 				0,
 				CRTC_V_TOTAL_CONTROL,
 				CRTC_V_TOTAL_MAX_SEL);
+		set_reg_field_value(v_total_min,
+				0,
+				CRTC_V_TOTAL_MIN,
+				CRTC_V_TOTAL_MIN);
+		set_reg_field_value(v_total_max,
+				0,
+				CRTC_V_TOTAL_MAX,
+				CRTC_V_TOTAL_MAX);
 		set_reg_field_value(v_total_cntl,
 				0,
 				CRTC_V_TOTAL_CONTROL,
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator_v.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator_v.c
index 8ad04816e7d3..a3cef60380ed 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator_v.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator_v.c
@@ -648,12 +648,6 @@ static void dce110_timing_generator_v_disable_vga(
 	return;
 }
 
-static bool dce110_tg_v_is_blanked(struct timing_generator *tg)
-{
-	/* Signal comes from the primary pipe, underlay is never blanked. */
-	return false;
-}
-
 /** ********************************************************************************************
  *
  * DCE11 Timing Generator Constructor / Destructor
@@ -670,7 +664,6 @@ static const struct timing_generator_funcs dce110_tg_v_funcs = {
 		.set_early_control = dce110_timing_generator_v_set_early_control,
 		.wait_for_state = dce110_timing_generator_v_wait_for_state,
 		.set_blank = dce110_timing_generator_v_set_blank,
-		.is_blanked = dce110_tg_v_is_blanked,
 		.set_colors = dce110_timing_generator_v_set_colors,
 		.set_overscan_blank_color =
 				dce110_timing_generator_v_set_overscan_color_black,
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c
index 8ba3c12fc608..aa8d6b10d2c3 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_transform_v.c
@@ -235,7 +235,7 @@ static void program_overscan(
 	int overscan_right = data->h_active - data->recout.x - data->recout.width;
 	int overscan_bottom = data->v_active - data->recout.y - data->recout.height;
 
-	if (xfm_dce->base.ctx->dc->debug.surface_visual_confirm) {
+	if (xfm_dce->base.ctx->dc->debug.visual_confirm != VISUAL_CONFIRM_DISABLE) {
 		overscan_bottom += 2;
 		overscan_right += 2;
 	}
@@ -373,13 +373,13 @@ static void calculate_inits(
 	struct rect *chroma_viewport)
 {
 	inits->h_int_scale_ratio_luma =
-		dal_fixed31_32_u2d19(data->ratios.horz) << 5;
+		dc_fixpt_u2d19(data->ratios.horz) << 5;
 	inits->v_int_scale_ratio_luma =
-		dal_fixed31_32_u2d19(data->ratios.vert) << 5;
+		dc_fixpt_u2d19(data->ratios.vert) << 5;
 	inits->h_int_scale_ratio_chroma =
-		dal_fixed31_32_u2d19(data->ratios.horz_c) << 5;
+		dc_fixpt_u2d19(data->ratios.horz_c) << 5;
 	inits->v_int_scale_ratio_chroma =
-		dal_fixed31_32_u2d19(data->ratios.vert_c) << 5;
+		dc_fixpt_u2d19(data->ratios.vert_c) << 5;
 
 	inits->h_init_luma.integer = 1;
 	inits->v_init_luma.integer = 1;
diff --git a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c
index cd1e3f72c44e..30d5b32892d6 100644
--- a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.c
@@ -146,15 +146,15 @@ static const struct dce110_timing_generator_offsets dce112_tg_offsets[] = {
 	.reg_name = mm ## block ## id ## _ ## reg_name
 
 
-static const struct dce_disp_clk_registers disp_clk_regs = {
+static const struct dccg_registers disp_clk_regs = {
 		CLK_COMMON_REG_LIST_DCE_BASE()
 };
 
-static const struct dce_disp_clk_shift disp_clk_shift = {
+static const struct dccg_shift disp_clk_shift = {
 		CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(__SHIFT)
 };
 
-static const struct dce_disp_clk_mask disp_clk_mask = {
+static const struct dccg_mask disp_clk_mask = {
 		CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(_MASK)
 };
 
@@ -430,7 +430,7 @@ static struct stream_encoder *dce112_stream_encoder_create(
 
 	if (!enc110)
 		return NULL;
-	
+
 	dce110_stream_encoder_construct(enc110, ctx, ctx->dc_bios, eng_id,
 					&stream_enc_regs[eng_id],
 					&se_shift, &se_mask);
@@ -668,8 +668,8 @@ static void destruct(struct dce110_resource_pool *pool)
 	if (pool->base.dmcu != NULL)
 		dce_dmcu_destroy(&pool->base.dmcu);
 
-	if (pool->base.display_clock != NULL)
-		dce_disp_clk_destroy(&pool->base.display_clock);
+	if (pool->base.dccg != NULL)
+		dce_dccg_destroy(&pool->base.dccg);
 
 	if (pool->base.irqs != NULL) {
 		dal_irq_service_destroy(&pool->base.irqs);
@@ -744,43 +744,38 @@ bool dce112_validate_bandwidth(
 
 	if (memcmp(&dc->current_state->bw.dce,
 			&context->bw.dce, sizeof(context->bw.dce))) {
-		struct log_entry log_entry;
-		dm_logger_open(
-			dc->ctx->logger,
-			&log_entry,
-			LOG_BANDWIDTH_CALCS);
-		dm_logger_append(&log_entry, "%s: finish,\n"
+
+		DC_LOG_BANDWIDTH_CALCS(
+			"%s: finish,\n"
+			"nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n"
+			"stutMark_b: %d stutMark_a: %d\n"
+			"nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n"
+			"stutMark_b: %d stutMark_a: %d\n"
 			"nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n"
-			"stutMark_b: %d stutMark_a: %d\n",
+			"stutMark_b: %d stutMark_a: %d stutter_mode_enable: %d\n"
+			"cstate: %d pstate: %d nbpstate: %d sync: %d dispclk: %d\n"
+			"sclk: %d sclk_sleep: %d yclk: %d blackout_recovery_time_us: %d\n"
+			,
 			__func__,
 			context->bw.dce.nbp_state_change_wm_ns[0].b_mark,
 			context->bw.dce.nbp_state_change_wm_ns[0].a_mark,
 			context->bw.dce.urgent_wm_ns[0].b_mark,
 			context->bw.dce.urgent_wm_ns[0].a_mark,
 			context->bw.dce.stutter_exit_wm_ns[0].b_mark,
-			context->bw.dce.stutter_exit_wm_ns[0].a_mark);
-		dm_logger_append(&log_entry,
-			"nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n"
-			"stutMark_b: %d stutMark_a: %d\n",
+			context->bw.dce.stutter_exit_wm_ns[0].a_mark,
 			context->bw.dce.nbp_state_change_wm_ns[1].b_mark,
 			context->bw.dce.nbp_state_change_wm_ns[1].a_mark,
 			context->bw.dce.urgent_wm_ns[1].b_mark,
 			context->bw.dce.urgent_wm_ns[1].a_mark,
 			context->bw.dce.stutter_exit_wm_ns[1].b_mark,
-			context->bw.dce.stutter_exit_wm_ns[1].a_mark);
-		dm_logger_append(&log_entry,
-			"nbpMark_b: %d nbpMark_a: %d urgentMark_b: %d urgentMark_a: %d\n"
-			"stutMark_b: %d stutMark_a: %d stutter_mode_enable: %d\n",
+			context->bw.dce.stutter_exit_wm_ns[1].a_mark,
 			context->bw.dce.nbp_state_change_wm_ns[2].b_mark,
 			context->bw.dce.nbp_state_change_wm_ns[2].a_mark,
 			context->bw.dce.urgent_wm_ns[2].b_mark,
 			context->bw.dce.urgent_wm_ns[2].a_mark,
 			context->bw.dce.stutter_exit_wm_ns[2].b_mark,
 			context->bw.dce.stutter_exit_wm_ns[2].a_mark,
-			context->bw.dce.stutter_mode_enable);
-		dm_logger_append(&log_entry,
-			"cstate: %d pstate: %d nbpstate: %d sync: %d dispclk: %d\n"
-			"sclk: %d sclk_sleep: %d yclk: %d blackout_recovery_time_us: %d\n",
+			context->bw.dce.stutter_mode_enable,
 			context->bw.dce.cpuc_state_change_enable,
 			context->bw.dce.cpup_state_change_enable,
 			context->bw.dce.nbp_state_change_enable,
@@ -790,7 +785,6 @@ bool dce112_validate_bandwidth(
 			context->bw.dce.sclk_deep_sleep_khz,
 			context->bw.dce.yclk_khz,
 			context->bw.dce.blackout_recovery_time_us);
-		dm_logger_close(&log_entry);
 	}
 	return result;
 }
@@ -867,38 +861,6 @@ enum dc_status dce112_add_stream_to_ctx(
 	return result;
 }
 
-enum dc_status dce112_validate_guaranteed(
-		struct dc *dc,
-		struct dc_stream_state *stream,
-		struct dc_state *context)
-{
-	enum dc_status result = DC_ERROR_UNEXPECTED;
-
-	context->streams[0] = stream;
-	dc_stream_retain(context->streams[0]);
-	context->stream_count++;
-
-	result = resource_map_pool_resources(dc, context, stream);
-
-	if (result == DC_OK)
-		result = resource_map_phy_clock_resources(dc, context, stream);
-
-	if (result == DC_OK)
-		result = build_mapped_resource(dc, context, stream);
-
-	if (result == DC_OK) {
-		validate_guaranteed_copy_streams(
-				context, dc->caps.max_streams);
-		result = resource_build_scaling_params_for_context(dc, context);
-	}
-
-	if (result == DC_OK)
-		if (!dce112_validate_bandwidth(dc, context))
-			result = DC_FAIL_BANDWIDTH_VALIDATE;
-
-	return result;
-}
-
 enum dc_status dce112_validate_global(
 		struct dc *dc,
 		struct dc_state *context)
@@ -921,7 +883,6 @@ static void dce112_destroy_resource_pool(struct resource_pool **pool)
 static const struct resource_funcs dce112_res_pool_funcs = {
 	.destroy = dce112_destroy_resource_pool,
 	.link_enc_create = dce112_link_encoder_create,
-	.validate_guaranteed = dce112_validate_guaranteed,
 	.validate_bandwidth = dce112_validate_bandwidth,
 	.validate_plane = dce100_validate_plane,
 	.add_stream_to_ctx = dce112_add_stream_to_ctx,
@@ -1033,7 +994,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc)
 			eng_clks.data[0].clocks_in_khz;
 	clk_ranges.wm_clk_ranges[0].wm_max_eng_clk_in_khz =
 			eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz - 1;
-	clk_ranges.wm_clk_ranges[0].wm_min_memg_clk_in_khz =
+	clk_ranges.wm_clk_ranges[0].wm_min_mem_clk_in_khz =
 			mem_clks.data[0].clocks_in_khz;
 	clk_ranges.wm_clk_ranges[0].wm_max_mem_clk_in_khz =
 			mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz - 1;
@@ -1043,7 +1004,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc)
 			eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz;
 	/* 5 GHz instead of data[7].clockInKHz to cover Overdrive */
 	clk_ranges.wm_clk_ranges[1].wm_max_eng_clk_in_khz = 5000000;
-	clk_ranges.wm_clk_ranges[1].wm_min_memg_clk_in_khz =
+	clk_ranges.wm_clk_ranges[1].wm_min_mem_clk_in_khz =
 			mem_clks.data[0].clocks_in_khz;
 	clk_ranges.wm_clk_ranges[1].wm_max_mem_clk_in_khz =
 			mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz - 1;
@@ -1053,7 +1014,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc)
 			eng_clks.data[0].clocks_in_khz;
 	clk_ranges.wm_clk_ranges[2].wm_max_eng_clk_in_khz =
 			eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz - 1;
-	clk_ranges.wm_clk_ranges[2].wm_min_memg_clk_in_khz =
+	clk_ranges.wm_clk_ranges[2].wm_min_mem_clk_in_khz =
 			mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz;
 	/* 5 GHz instead of data[2].clockInKHz to cover Overdrive */
 	clk_ranges.wm_clk_ranges[2].wm_max_mem_clk_in_khz = 5000000;
@@ -1063,7 +1024,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc)
 			eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz;
 	/* 5 GHz instead of data[7].clockInKHz to cover Overdrive */
 	clk_ranges.wm_clk_ranges[3].wm_max_eng_clk_in_khz = 5000000;
-	clk_ranges.wm_clk_ranges[3].wm_min_memg_clk_in_khz =
+	clk_ranges.wm_clk_ranges[3].wm_min_mem_clk_in_khz =
 			mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz;
 	/* 5 GHz instead of data[2].clockInKHz to cover Overdrive */
 	clk_ranges.wm_clk_ranges[3].wm_max_mem_clk_in_khz = 5000000;
@@ -1157,11 +1118,11 @@ static bool construct(
 		}
 	}
 
-	pool->base.display_clock = dce112_disp_clk_create(ctx,
+	pool->base.dccg = dce112_dccg_create(ctx,
 			&disp_clk_regs,
 			&disp_clk_shift,
 			&disp_clk_mask);
-	if (pool->base.display_clock == NULL) {
+	if (pool->base.dccg == NULL) {
 		dm_error("DC: failed to create display clock!\n");
 		BREAK_TO_DEBUGGER();
 		goto res_create_fail;
@@ -1191,7 +1152,7 @@ static bool construct(
 	 * max_clock_state
 	 */
 	if (dm_pp_get_static_clocks(ctx, &static_clk_info))
-		pool->base.display_clock->max_clks_state =
+		pool->base.dccg->max_clks_state =
 				static_clk_info.max_clocks_state;
 
 	{
diff --git a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.h b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.h
index d5c19d34eb0a..95a403396219 100644
--- a/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/dce112/dce112_resource.h
@@ -42,11 +42,6 @@ enum dc_status dce112_validate_with_context(
 		struct dc_state *context,
 		struct dc_state *old_context);
 
-enum dc_status dce112_validate_guaranteed(
-		struct dc *dc,
-		struct dc_stream_state *dc_stream,
-		struct dc_state *context);
-
 bool dce112_validate_bandwidth(
 	struct dc *dc,
 	struct dc_state *context);
diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
index 4659a4bfabaa..8381f27a2361 100644
--- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_resource.c
@@ -494,8 +494,8 @@ static void destruct(struct dce110_resource_pool *pool)
 	if (pool->base.dmcu != NULL)
 		dce_dmcu_destroy(&pool->base.dmcu);
 
-	if (pool->base.display_clock != NULL)
-		dce_disp_clk_destroy(&pool->base.display_clock);
+	if (pool->base.dccg != NULL)
+		dce_dccg_destroy(&pool->base.dccg);
 }
 
 static void read_dce_straps(
@@ -652,7 +652,7 @@ static struct mem_input *dce120_mem_input_create(
 		return NULL;
 	}
 
-	dce112_mem_input_construct(dce_mi, ctx, inst, &mi_regs[inst], &mi_shifts, &mi_masks);
+	dce120_mem_input_construct(dce_mi, ctx, inst, &mi_regs[inst], &mi_shifts, &mi_masks);
 	return &dce_mi->base;
 }
 
@@ -684,7 +684,6 @@ static void dce120_destroy_resource_pool(struct resource_pool **pool)
 static const struct resource_funcs dce120_res_pool_funcs = {
 	.destroy = dce120_destroy_resource_pool,
 	.link_enc_create = dce120_link_encoder_create,
-	.validate_guaranteed = dce112_validate_guaranteed,
 	.validate_bandwidth = dce112_validate_bandwidth,
 	.validate_plane = dce100_validate_plane,
 	.add_stream_to_ctx = dce112_add_stream_to_ctx
@@ -776,7 +775,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc)
 			eng_clks.data[0].clocks_in_khz;
 	clk_ranges.wm_clk_ranges[0].wm_max_eng_clk_in_khz =
 			eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz - 1;
-	clk_ranges.wm_clk_ranges[0].wm_min_memg_clk_in_khz =
+	clk_ranges.wm_clk_ranges[0].wm_min_mem_clk_in_khz =
 			mem_clks.data[0].clocks_in_khz;
 	clk_ranges.wm_clk_ranges[0].wm_max_mem_clk_in_khz =
 			mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz - 1;
@@ -786,7 +785,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc)
 			eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz;
 	/* 5 GHz instead of data[7].clockInKHz to cover Overdrive */
 	clk_ranges.wm_clk_ranges[1].wm_max_eng_clk_in_khz = 5000000;
-	clk_ranges.wm_clk_ranges[1].wm_min_memg_clk_in_khz =
+	clk_ranges.wm_clk_ranges[1].wm_min_mem_clk_in_khz =
 			mem_clks.data[0].clocks_in_khz;
 	clk_ranges.wm_clk_ranges[1].wm_max_mem_clk_in_khz =
 			mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz - 1;
@@ -796,7 +795,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc)
 			eng_clks.data[0].clocks_in_khz;
 	clk_ranges.wm_clk_ranges[2].wm_max_eng_clk_in_khz =
 			eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz - 1;
-	clk_ranges.wm_clk_ranges[2].wm_min_memg_clk_in_khz =
+	clk_ranges.wm_clk_ranges[2].wm_min_mem_clk_in_khz =
 			mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz;
 	/* 5 GHz instead of data[2].clockInKHz to cover Overdrive */
 	clk_ranges.wm_clk_ranges[2].wm_max_mem_clk_in_khz = 5000000;
@@ -806,7 +805,7 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc)
 			eng_clks.data[eng_clks.num_levels*3/8].clocks_in_khz;
 	/* 5 GHz instead of data[7].clockInKHz to cover Overdrive */
 	clk_ranges.wm_clk_ranges[3].wm_max_eng_clk_in_khz = 5000000;
-	clk_ranges.wm_clk_ranges[3].wm_min_memg_clk_in_khz =
+	clk_ranges.wm_clk_ranges[3].wm_min_mem_clk_in_khz =
 			mem_clks.data[mem_clks.num_levels>>1].clocks_in_khz;
 	/* 5 GHz instead of data[2].clockInKHz to cover Overdrive */
 	clk_ranges.wm_clk_ranges[3].wm_max_mem_clk_in_khz = 5000000;
@@ -815,14 +814,25 @@ static void bw_calcs_data_update_from_pplib(struct dc *dc)
 	dm_pp_notify_wm_clock_changes(dc->ctx, &clk_ranges);
 }
 
+static uint32_t read_pipe_fuses(struct dc_context *ctx)
+{
+	uint32_t value = dm_read_reg_soc15(ctx, mmCC_DC_PIPE_DIS, 0);
+	/* VG20 support max 6 pipes */
+	value = value & 0x3f;
+	return value;
+}
+
 static bool construct(
 	uint8_t num_virtual_links,
 	struct dc *dc,
 	struct dce110_resource_pool *pool)
 {
 	unsigned int i;
+	int j;
 	struct dc_context *ctx = dc->ctx;
 	struct irq_service_init_data irq_init_data;
+	bool harvest_enabled = ASICREV_IS_VEGA20_P(ctx->asic_id.hw_internal_rev);
+	uint32_t pipe_fuses;
 
 	ctx->dc_bios->regs = &bios_regs;
 
@@ -884,11 +894,11 @@ static bool construct(
 		}
 	}
 
-	pool->base.display_clock = dce120_disp_clk_create(ctx);
-	if (pool->base.display_clock == NULL) {
+	pool->base.dccg = dce120_dccg_create(ctx);
+	if (pool->base.dccg == NULL) {
 		dm_error("DC: failed to create display clock!\n");
 		BREAK_TO_DEBUGGER();
-		goto disp_clk_create_fail;
+		goto dccg_create_fail;
 	}
 
 	pool->base.dmcu = dce_dmcu_create(ctx,
@@ -916,28 +926,41 @@ static bool construct(
 	if (!pool->base.irqs)
 		goto irqs_create_fail;
 
+	/* retrieve valid pipe fuses */
+	if (harvest_enabled)
+		pipe_fuses = read_pipe_fuses(ctx);
+
+	/* index to valid pipe resource */
+	j = 0;
 	for (i = 0; i < pool->base.pipe_count; i++) {
-		pool->base.timing_generators[i] =
+		if (harvest_enabled) {
+			if ((pipe_fuses & (1 << i)) != 0) {
+				dm_error("DC: skip invalid pipe %d!\n", i);
+				continue;
+			}
+		}
+
+		pool->base.timing_generators[j] =
 				dce120_timing_generator_create(
 					ctx,
 					i,
 					&dce120_tg_offsets[i]);
-		if (pool->base.timing_generators[i] == NULL) {
+		if (pool->base.timing_generators[j] == NULL) {
 			BREAK_TO_DEBUGGER();
 			dm_error("DC: failed to create tg!\n");
 			goto controller_create_fail;
 		}
 
-		pool->base.mis[i] = dce120_mem_input_create(ctx, i);
+		pool->base.mis[j] = dce120_mem_input_create(ctx, i);
 
-		if (pool->base.mis[i] == NULL) {
+		if (pool->base.mis[j] == NULL) {
 			BREAK_TO_DEBUGGER();
 			dm_error(
 				"DC: failed to create memory input!\n");
 			goto controller_create_fail;
 		}
 
-		pool->base.ipps[i] = dce120_ipp_create(ctx, i);
+		pool->base.ipps[j] = dce120_ipp_create(ctx, i);
 		if (pool->base.ipps[i] == NULL) {
 			BREAK_TO_DEBUGGER();
 			dm_error(
@@ -945,7 +968,7 @@ static bool construct(
 			goto controller_create_fail;
 		}
 
-		pool->base.transforms[i] = dce120_transform_create(ctx, i);
+		pool->base.transforms[j] = dce120_transform_create(ctx, i);
 		if (pool->base.transforms[i] == NULL) {
 			BREAK_TO_DEBUGGER();
 			dm_error(
@@ -953,16 +976,23 @@ static bool construct(
 			goto res_create_fail;
 		}
 
-		pool->base.opps[i] = dce120_opp_create(
+		pool->base.opps[j] = dce120_opp_create(
 			ctx,
 			i);
-		if (pool->base.opps[i] == NULL) {
+		if (pool->base.opps[j] == NULL) {
 			BREAK_TO_DEBUGGER();
 			dm_error(
 				"DC: failed to create output pixel processor!\n");
 		}
+
+		/* check next valid pipe */
+		j++;
 	}
 
+	/* valid pipe num */
+	pool->base.pipe_count = j;
+	pool->base.timing_generator_count = j;
+
 	if (!resource_construct(num_virtual_links, dc, &pool->base,
 			 &res_create_funcs))
 		goto res_create_fail;
@@ -981,7 +1011,7 @@ static bool construct(
 
 irqs_create_fail:
 controller_create_fail:
-disp_clk_create_fail:
+dccg_create_fail:
 clk_src_create_fail:
 res_create_fail:
 
diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c
index 7bee78172d85..2ea490f8482e 100644
--- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c
+++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c
@@ -570,12 +570,6 @@ void dce120_timing_generator_set_drr(
 				0x180);
 
 	} else {
-		CRTC_REG_UPDATE(
-				CRTC0_CRTC_V_TOTAL_MIN,
-				CRTC_V_TOTAL_MIN, 0);
-		CRTC_REG_UPDATE(
-				CRTC0_CRTC_V_TOTAL_MAX,
-				CRTC_V_TOTAL_MAX, 0);
 		CRTC_REG_SET_N(CRTC0_CRTC_V_TOTAL_CONTROL, 5,
 				FD(CRTC0_CRTC_V_TOTAL_CONTROL__CRTC_V_TOTAL_MIN_SEL), 0,
 				FD(CRTC0_CRTC_V_TOTAL_CONTROL__CRTC_V_TOTAL_MAX_SEL), 0,
@@ -583,6 +577,12 @@ void dce120_timing_generator_set_drr(
 				FD(CRTC0_CRTC_V_TOTAL_CONTROL__CRTC_FORCE_LOCK_TO_MASTER_VSYNC), 0,
 				FD(CRTC0_CRTC_V_TOTAL_CONTROL__CRTC_SET_V_TOTAL_MIN_MASK), 0);
 		CRTC_REG_UPDATE(
+				CRTC0_CRTC_V_TOTAL_MIN,
+				CRTC_V_TOTAL_MIN, 0);
+		CRTC_REG_UPDATE(
+				CRTC0_CRTC_V_TOTAL_MAX,
+				CRTC_V_TOTAL_MAX, 0);
+		CRTC_REG_UPDATE(
 				CRTC0_CRTC_STATIC_SCREEN_CONTROL,
 				CRTC_STATIC_SCREEN_EVENT_MASK,
 				0);
diff --git a/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c b/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c
index 5d854a37a978..2ac95ec2bf96 100644
--- a/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dce80/dce80_resource.c
@@ -153,15 +153,15 @@ static const struct dce110_timing_generator_offsets dce80_tg_offsets[] = {
 	.reg_name = mm ## block ## id ## _ ## reg_name
 
 
-static const struct dce_disp_clk_registers disp_clk_regs = {
+static const struct dccg_registers disp_clk_regs = {
 		CLK_COMMON_REG_LIST_DCE_BASE()
 };
 
-static const struct dce_disp_clk_shift disp_clk_shift = {
+static const struct dccg_shift disp_clk_shift = {
 		CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(__SHIFT)
 };
 
-static const struct dce_disp_clk_mask disp_clk_mask = {
+static const struct dccg_mask disp_clk_mask = {
 		CLK_COMMON_MASK_SH_LIST_DCE_COMMON_BASE(_MASK)
 };
 
@@ -683,31 +683,14 @@ static void destruct(struct dce110_resource_pool *pool)
 		}
 	}
 
-	if (pool->base.display_clock != NULL)
-		dce_disp_clk_destroy(&pool->base.display_clock);
+	if (pool->base.dccg != NULL)
+		dce_dccg_destroy(&pool->base.dccg);
 
 	if (pool->base.irqs != NULL) {
 		dal_irq_service_destroy(&pool->base.irqs);
 	}
 }
 
-static enum dc_status build_mapped_resource(
-		const struct dc *dc,
-		struct dc_state *context,
-		struct dc_stream_state *stream)
-{
-	struct pipe_ctx *pipe_ctx = resource_get_head_pipe_for_stream(&context->res_ctx, stream);
-
-	if (!pipe_ctx)
-		return DC_ERROR_UNEXPECTED;
-
-	dce110_resource_build_pipe_hw_param(pipe_ctx);
-
-	resource_build_info_frame(pipe_ctx);
-
-	return DC_OK;
-}
-
 bool dce80_validate_bandwidth(
 	struct dc *dc,
 	struct dc_state *context)
@@ -749,37 +732,6 @@ enum dc_status dce80_validate_global(
 	return DC_OK;
 }
 
-enum dc_status dce80_validate_guaranteed(
-		struct dc *dc,
-		struct dc_stream_state *dc_stream,
-		struct dc_state *context)
-{
-	enum dc_status result = DC_ERROR_UNEXPECTED;
-
-	context->streams[0] = dc_stream;
-	dc_stream_retain(context->streams[0]);
-	context->stream_count++;
-
-	result = resource_map_pool_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = resource_map_clock_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = build_mapped_resource(dc, context, dc_stream);
-
-	if (result == DC_OK) {
-		validate_guaranteed_copy_streams(
-				context, dc->caps.max_streams);
-		result = resource_build_scaling_params_for_context(dc, context);
-	}
-
-	if (result == DC_OK)
-		result = dce80_validate_bandwidth(dc, context);
-
-	return result;
-}
-
 static void dce80_destroy_resource_pool(struct resource_pool **pool)
 {
 	struct dce110_resource_pool *dce110_pool = TO_DCE110_RES_POOL(*pool);
@@ -792,7 +744,6 @@ static void dce80_destroy_resource_pool(struct resource_pool **pool)
 static const struct resource_funcs dce80_res_pool_funcs = {
 	.destroy = dce80_destroy_resource_pool,
 	.link_enc_create = dce80_link_encoder_create,
-	.validate_guaranteed = dce80_validate_guaranteed,
 	.validate_bandwidth = dce80_validate_bandwidth,
 	.validate_plane = dce100_validate_plane,
 	.add_stream_to_ctx = dce100_add_stream_to_ctx,
@@ -871,11 +822,11 @@ static bool dce80_construct(
 		}
 	}
 
-	pool->base.display_clock = dce_disp_clk_create(ctx,
+	pool->base.dccg = dce_dccg_create(ctx,
 			&disp_clk_regs,
 			&disp_clk_shift,
 			&disp_clk_mask);
-	if (pool->base.display_clock == NULL) {
+	if (pool->base.dccg == NULL) {
 		dm_error("DC: failed to create display clock!\n");
 		BREAK_TO_DEBUGGER();
 		goto res_create_fail;
@@ -901,7 +852,7 @@ static bool dce80_construct(
 		goto res_create_fail;
 	}
 	if (dm_pp_get_static_clocks(ctx, &static_clk_info))
-		pool->base.display_clock->max_clks_state =
+		pool->base.dccg->max_clks_state =
 					static_clk_info.max_clocks_state;
 
 	{
@@ -1055,11 +1006,11 @@ static bool dce81_construct(
 		}
 	}
 
-	pool->base.display_clock = dce_disp_clk_create(ctx,
+	pool->base.dccg = dce_dccg_create(ctx,
 			&disp_clk_regs,
 			&disp_clk_shift,
 			&disp_clk_mask);
-	if (pool->base.display_clock == NULL) {
+	if (pool->base.dccg == NULL) {
 		dm_error("DC: failed to create display clock!\n");
 		BREAK_TO_DEBUGGER();
 		goto res_create_fail;
@@ -1086,7 +1037,7 @@ static bool dce81_construct(
 	}
 
 	if (dm_pp_get_static_clocks(ctx, &static_clk_info))
-		pool->base.display_clock->max_clks_state =
+		pool->base.dccg->max_clks_state =
 					static_clk_info.max_clocks_state;
 
 	{
@@ -1236,11 +1187,11 @@ static bool dce83_construct(
 		}
 	}
 
-	pool->base.display_clock = dce_disp_clk_create(ctx,
+	pool->base.dccg = dce_dccg_create(ctx,
 			&disp_clk_regs,
 			&disp_clk_shift,
 			&disp_clk_mask);
-	if (pool->base.display_clock == NULL) {
+	if (pool->base.dccg == NULL) {
 		dm_error("DC: failed to create display clock!\n");
 		BREAK_TO_DEBUGGER();
 		goto res_create_fail;
@@ -1267,7 +1218,7 @@ static bool dce83_construct(
 	}
 
 	if (dm_pp_get_static_clocks(ctx, &static_clk_info))
-		pool->base.display_clock->max_clks_state =
+		pool->base.dccg->max_clks_state =
 					static_clk_info.max_clocks_state;
 
 	{
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/Makefile b/drivers/gpu/drm/amd/display/dc/dcn10/Makefile
index 5469bdfe19f3..84f52c63d95c 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/Makefile
@@ -26,7 +26,7 @@ DCN10 = dcn10_resource.o dcn10_ipp.o dcn10_hw_sequencer.o \
 		dcn10_dpp.o dcn10_opp.o dcn10_optc.o \
 		dcn10_hubp.o dcn10_mpc.o \
 		dcn10_dpp_dscl.o dcn10_dpp_cm.o dcn10_cm_common.o \
-		dcn10_hubbub.o
+		dcn10_hubbub.o dcn10_stream_encoder.o dcn10_link_encoder.o
 
 AMD_DAL_DCN10 = $(addprefix $(AMDDALPATH)/dc/dcn10/,$(DCN10))
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c
index 881a1bff94d2..5d95a997fd9f 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_cm_common.c
@@ -169,7 +169,7 @@ bool cm_helper_convert_to_custom_float(
 	}
 
 	if (fixpoint == true)
-		arr_points[1].custom_float_y = dal_fixed31_32_clamp_u0d14(arr_points[1].y);
+		arr_points[1].custom_float_y = dc_fixpt_clamp_u0d14(arr_points[1].y);
 	else if (!convert_to_custom_float_format(arr_points[1].y, &fmt,
 		&arr_points[1].custom_float_y)) {
 		BREAK_TO_DEBUGGER();
@@ -327,19 +327,19 @@ bool cm_helper_translate_curve_to_hw_format(
 	rgb_resulted[hw_points - 1].green = output_tf->tf_pts.green[start_index];
 	rgb_resulted[hw_points - 1].blue = output_tf->tf_pts.blue[start_index];
 
-	arr_points[0].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2),
-					     dal_fixed31_32_from_int(region_start));
-	arr_points[1].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2),
-					     dal_fixed31_32_from_int(region_end));
+	arr_points[0].x = dc_fixpt_pow(dc_fixpt_from_int(2),
+					     dc_fixpt_from_int(region_start));
+	arr_points[1].x = dc_fixpt_pow(dc_fixpt_from_int(2),
+					     dc_fixpt_from_int(region_end));
 
 	y_r = rgb_resulted[0].red;
 	y_g = rgb_resulted[0].green;
 	y_b = rgb_resulted[0].blue;
 
-	y1_min = dal_fixed31_32_min(y_r, dal_fixed31_32_min(y_g, y_b));
+	y1_min = dc_fixpt_min(y_r, dc_fixpt_min(y_g, y_b));
 
 	arr_points[0].y = y1_min;
-	arr_points[0].slope = dal_fixed31_32_div(arr_points[0].y, arr_points[0].x);
+	arr_points[0].slope = dc_fixpt_div(arr_points[0].y, arr_points[0].x);
 	y_r = rgb_resulted[hw_points - 1].red;
 	y_g = rgb_resulted[hw_points - 1].green;
 	y_b = rgb_resulted[hw_points - 1].blue;
@@ -347,35 +347,35 @@ bool cm_helper_translate_curve_to_hw_format(
 	/* see comment above, m_arrPoints[1].y should be the Y value for the
 	 * region end (m_numOfHwPoints), not last HW point(m_numOfHwPoints - 1)
 	 */
-	y3_max = dal_fixed31_32_max(y_r, dal_fixed31_32_max(y_g, y_b));
+	y3_max = dc_fixpt_max(y_r, dc_fixpt_max(y_g, y_b));
 
 	arr_points[1].y = y3_max;
 
-	arr_points[1].slope = dal_fixed31_32_zero;
+	arr_points[1].slope = dc_fixpt_zero;
 
 	if (output_tf->tf == TRANSFER_FUNCTION_PQ) {
 		/* for PQ, we want to have a straight line from last HW X point,
 		 * and the slope to be such that we hit 1.0 at 10000 nits.
 		 */
 		const struct fixed31_32 end_value =
-				dal_fixed31_32_from_int(125);
+				dc_fixpt_from_int(125);
 
-		arr_points[1].slope = dal_fixed31_32_div(
-			dal_fixed31_32_sub(dal_fixed31_32_one, arr_points[1].y),
-			dal_fixed31_32_sub(end_value, arr_points[1].x));
+		arr_points[1].slope = dc_fixpt_div(
+			dc_fixpt_sub(dc_fixpt_one, arr_points[1].y),
+			dc_fixpt_sub(end_value, arr_points[1].x));
 	}
 
 	lut_params->hw_points_num = hw_points;
 
-	i = 1;
-	for (k = 0; k < MAX_REGIONS_NUMBER && i < MAX_REGIONS_NUMBER; k++) {
+	k = 0;
+	for (i = 1; i < MAX_REGIONS_NUMBER; i++) {
 		if (seg_distr[k] != -1) {
 			lut_params->arr_curve_points[k].segments_num =
 					seg_distr[k];
 			lut_params->arr_curve_points[i].offset =
 					lut_params->arr_curve_points[k].offset + (1 << seg_distr[k]);
 		}
-		i++;
+		k++;
 	}
 
 	if (seg_distr[k] != -1)
@@ -386,24 +386,24 @@ bool cm_helper_translate_curve_to_hw_format(
 
 	i = 1;
 	while (i != hw_points + 1) {
-		if (dal_fixed31_32_lt(rgb_plus_1->red, rgb->red))
+		if (dc_fixpt_lt(rgb_plus_1->red, rgb->red))
 			rgb_plus_1->red = rgb->red;
-		if (dal_fixed31_32_lt(rgb_plus_1->green, rgb->green))
+		if (dc_fixpt_lt(rgb_plus_1->green, rgb->green))
 			rgb_plus_1->green = rgb->green;
-		if (dal_fixed31_32_lt(rgb_plus_1->blue, rgb->blue))
+		if (dc_fixpt_lt(rgb_plus_1->blue, rgb->blue))
 			rgb_plus_1->blue = rgb->blue;
 
-		rgb->delta_red   = dal_fixed31_32_sub(rgb_plus_1->red,   rgb->red);
-		rgb->delta_green = dal_fixed31_32_sub(rgb_plus_1->green, rgb->green);
-		rgb->delta_blue  = dal_fixed31_32_sub(rgb_plus_1->blue,  rgb->blue);
+		rgb->delta_red   = dc_fixpt_sub(rgb_plus_1->red,   rgb->red);
+		rgb->delta_green = dc_fixpt_sub(rgb_plus_1->green, rgb->green);
+		rgb->delta_blue  = dc_fixpt_sub(rgb_plus_1->blue,  rgb->blue);
 
 		if (fixpoint == true) {
-			rgb->delta_red_reg   = dal_fixed31_32_clamp_u0d10(rgb->delta_red);
-			rgb->delta_green_reg = dal_fixed31_32_clamp_u0d10(rgb->delta_green);
-			rgb->delta_blue_reg  = dal_fixed31_32_clamp_u0d10(rgb->delta_blue);
-			rgb->red_reg         = dal_fixed31_32_clamp_u0d14(rgb->red);
-			rgb->green_reg       = dal_fixed31_32_clamp_u0d14(rgb->green);
-			rgb->blue_reg        = dal_fixed31_32_clamp_u0d14(rgb->blue);
+			rgb->delta_red_reg   = dc_fixpt_clamp_u0d10(rgb->delta_red);
+			rgb->delta_green_reg = dc_fixpt_clamp_u0d10(rgb->delta_green);
+			rgb->delta_blue_reg  = dc_fixpt_clamp_u0d10(rgb->delta_blue);
+			rgb->red_reg         = dc_fixpt_clamp_u0d14(rgb->red);
+			rgb->green_reg       = dc_fixpt_clamp_u0d14(rgb->green);
+			rgb->blue_reg        = dc_fixpt_clamp_u0d14(rgb->blue);
 		}
 
 		++rgb_plus_1;
@@ -489,19 +489,19 @@ bool cm_helper_translate_curve_to_degamma_hw_format(
 	rgb_resulted[hw_points - 1].green = output_tf->tf_pts.green[start_index];
 	rgb_resulted[hw_points - 1].blue = output_tf->tf_pts.blue[start_index];
 
-	arr_points[0].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2),
-					     dal_fixed31_32_from_int(region_start));
-	arr_points[1].x = dal_fixed31_32_pow(dal_fixed31_32_from_int(2),
-					     dal_fixed31_32_from_int(region_end));
+	arr_points[0].x = dc_fixpt_pow(dc_fixpt_from_int(2),
+					     dc_fixpt_from_int(region_start));
+	arr_points[1].x = dc_fixpt_pow(dc_fixpt_from_int(2),
+					     dc_fixpt_from_int(region_end));
 
 	y_r = rgb_resulted[0].red;
 	y_g = rgb_resulted[0].green;
 	y_b = rgb_resulted[0].blue;
 
-	y1_min = dal_fixed31_32_min(y_r, dal_fixed31_32_min(y_g, y_b));
+	y1_min = dc_fixpt_min(y_r, dc_fixpt_min(y_g, y_b));
 
 	arr_points[0].y = y1_min;
-	arr_points[0].slope = dal_fixed31_32_div(arr_points[0].y, arr_points[0].x);
+	arr_points[0].slope = dc_fixpt_div(arr_points[0].y, arr_points[0].x);
 	y_r = rgb_resulted[hw_points - 1].red;
 	y_g = rgb_resulted[hw_points - 1].green;
 	y_b = rgb_resulted[hw_points - 1].blue;
@@ -509,35 +509,35 @@ bool cm_helper_translate_curve_to_degamma_hw_format(
 	/* see comment above, m_arrPoints[1].y should be the Y value for the
 	 * region end (m_numOfHwPoints), not last HW point(m_numOfHwPoints - 1)
 	 */
-	y3_max = dal_fixed31_32_max(y_r, dal_fixed31_32_max(y_g, y_b));
+	y3_max = dc_fixpt_max(y_r, dc_fixpt_max(y_g, y_b));
 
 	arr_points[1].y = y3_max;
 
-	arr_points[1].slope = dal_fixed31_32_zero;
+	arr_points[1].slope = dc_fixpt_zero;
 
 	if (output_tf->tf == TRANSFER_FUNCTION_PQ) {
 		/* for PQ, we want to have a straight line from last HW X point,
 		 * and the slope to be such that we hit 1.0 at 10000 nits.
 		 */
 		const struct fixed31_32 end_value =
-				dal_fixed31_32_from_int(125);
+				dc_fixpt_from_int(125);
 
-		arr_points[1].slope = dal_fixed31_32_div(
-			dal_fixed31_32_sub(dal_fixed31_32_one, arr_points[1].y),
-			dal_fixed31_32_sub(end_value, arr_points[1].x));
+		arr_points[1].slope = dc_fixpt_div(
+			dc_fixpt_sub(dc_fixpt_one, arr_points[1].y),
+			dc_fixpt_sub(end_value, arr_points[1].x));
 	}
 
 	lut_params->hw_points_num = hw_points;
 
-	i = 1;
-	for (k = 0; k < MAX_REGIONS_NUMBER && i < MAX_REGIONS_NUMBER; k++) {
+	k = 0;
+	for (i = 1; i < MAX_REGIONS_NUMBER; i++) {
 		if (seg_distr[k] != -1) {
 			lut_params->arr_curve_points[k].segments_num =
 					seg_distr[k];
 			lut_params->arr_curve_points[i].offset =
 					lut_params->arr_curve_points[k].offset + (1 << seg_distr[k]);
 		}
-		i++;
+		k++;
 	}
 
 	if (seg_distr[k] != -1)
@@ -548,16 +548,16 @@ bool cm_helper_translate_curve_to_degamma_hw_format(
 
 	i = 1;
 	while (i != hw_points + 1) {
-		if (dal_fixed31_32_lt(rgb_plus_1->red, rgb->red))
+		if (dc_fixpt_lt(rgb_plus_1->red, rgb->red))
 			rgb_plus_1->red = rgb->red;
-		if (dal_fixed31_32_lt(rgb_plus_1->green, rgb->green))
+		if (dc_fixpt_lt(rgb_plus_1->green, rgb->green))
 			rgb_plus_1->green = rgb->green;
-		if (dal_fixed31_32_lt(rgb_plus_1->blue, rgb->blue))
+		if (dc_fixpt_lt(rgb_plus_1->blue, rgb->blue))
 			rgb_plus_1->blue = rgb->blue;
 
-		rgb->delta_red   = dal_fixed31_32_sub(rgb_plus_1->red,   rgb->red);
-		rgb->delta_green = dal_fixed31_32_sub(rgb_plus_1->green, rgb->green);
-		rgb->delta_blue  = dal_fixed31_32_sub(rgb_plus_1->blue,  rgb->blue);
+		rgb->delta_red   = dc_fixpt_sub(rgb_plus_1->red,   rgb->red);
+		rgb->delta_green = dc_fixpt_sub(rgb_plus_1->green, rgb->green);
+		rgb->delta_blue  = dc_fixpt_sub(rgb_plus_1->blue,  rgb->blue);
 
 		++rgb_plus_1;
 		++rgb;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
index e305c28c98de..bf8b68f8db4f 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.c
@@ -98,6 +98,30 @@ enum gamut_remap_select {
 	GAMUT_REMAP_COMB_COEFF
 };
 
+void dpp_read_state(struct dpp *dpp_base,
+		struct dcn_dpp_state *s)
+{
+	struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base);
+
+	REG_GET(CM_IGAM_CONTROL,
+			CM_IGAM_LUT_MODE, &s->igam_lut_mode);
+	REG_GET(CM_IGAM_CONTROL,
+			CM_IGAM_INPUT_FORMAT, &s->igam_input_format);
+	REG_GET(CM_DGAM_CONTROL,
+			CM_DGAM_LUT_MODE, &s->dgam_lut_mode);
+	REG_GET(CM_RGAM_CONTROL,
+			CM_RGAM_LUT_MODE, &s->rgam_lut_mode);
+	REG_GET(CM_GAMUT_REMAP_CONTROL,
+			CM_GAMUT_REMAP_MODE, &s->gamut_remap_mode);
+
+	s->gamut_remap_c11_c12 = REG_READ(CM_GAMUT_REMAP_C11_C12);
+	s->gamut_remap_c13_c14 = REG_READ(CM_GAMUT_REMAP_C13_C14);
+	s->gamut_remap_c21_c22 = REG_READ(CM_GAMUT_REMAP_C21_C22);
+	s->gamut_remap_c23_c24 = REG_READ(CM_GAMUT_REMAP_C23_C24);
+	s->gamut_remap_c31_c32 = REG_READ(CM_GAMUT_REMAP_C31_C32);
+	s->gamut_remap_c33_c34 = REG_READ(CM_GAMUT_REMAP_C33_C34);
+}
+
 /* Program gamut remap in bypass mode */
 void dpp_set_gamut_remap_bypass(struct dcn10_dpp *dpp)
 {
@@ -106,10 +130,9 @@ void dpp_set_gamut_remap_bypass(struct dcn10_dpp *dpp)
 	/* Gamut remap in bypass */
 }
 
-#define IDENTITY_RATIO(ratio) (dal_fixed31_32_u2d19(ratio) == (1 << 19))
-
+#define IDENTITY_RATIO(ratio) (dc_fixpt_u2d19(ratio) == (1 << 19))
 
-bool dpp_get_optimal_number_of_taps(
+static bool dpp_get_optimal_number_of_taps(
 		struct dpp *dpp,
 		struct scaler_data *scl_data,
 		const struct scaling_taps *in_taps)
@@ -121,6 +144,18 @@ bool dpp_get_optimal_number_of_taps(
 	else
 		pixel_width = scl_data->viewport.width;
 
+	/* Some ASICs does not support  FP16 scaling, so we reject modes require this*/
+	if (scl_data->format == PIXEL_FORMAT_FP16 &&
+		dpp->caps->dscl_data_proc_format == DSCL_DATA_PRCESSING_FIXED_FORMAT &&
+		scl_data->ratios.horz.value != dc_fixpt_one.value &&
+		scl_data->ratios.vert.value != dc_fixpt_one.value)
+		return false;
+
+	if (scl_data->viewport.width > scl_data->h_active &&
+		dpp->ctx->dc->debug.max_downscale_src_width != 0 &&
+		scl_data->viewport.width > dpp->ctx->dc->debug.max_downscale_src_width)
+		return false;
+
 	/* TODO: add lb check */
 
 	/* No support for programming ratio of 4, drop to 3.99999.. */
@@ -257,7 +292,7 @@ void dpp1_cnv_setup (
 		struct dpp *dpp_base,
 		enum surface_pixel_format format,
 		enum expansion_mode mode,
-		struct csc_transform input_csc_color_matrix,
+		struct dc_csc_transform input_csc_color_matrix,
 		enum dc_color_space input_color_space)
 {
 	uint32_t pixel_format;
@@ -410,13 +445,13 @@ void dpp1_set_cursor_position(
 		uint32_t width)
 {
 	struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base);
-	int src_x_offset = pos->x - pos->x_hotspot - param->viewport_x_start;
+	int src_x_offset = pos->x - pos->x_hotspot - param->viewport.x;
 	uint32_t cur_en = pos->enable ? 1 : 0;
 
-	if (src_x_offset >= (int)param->viewport_width)
+	if (src_x_offset >= (int)param->viewport.width)
 		cur_en = 0;  /* not visible beyond right edge*/
 
-	if (src_x_offset + (int)width < 0)
+	if (src_x_offset + (int)width <= 0)
 		cur_en = 0;  /* not visible beyond left edge*/
 
 	REG_UPDATE(CURSOR0_CONTROL,
@@ -424,6 +459,18 @@ void dpp1_set_cursor_position(
 
 }
 
+void dpp1_cnv_set_optional_cursor_attributes(
+		struct dpp *dpp_base,
+		struct dpp_cursor_attributes *attr)
+{
+	struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base);
+
+	if (attr) {
+		REG_UPDATE(CURSOR0_FP_SCALE_BIAS,  CUR0_FP_BIAS,  attr->bias);
+		REG_UPDATE(CURSOR0_FP_SCALE_BIAS,  CUR0_FP_SCALE, attr->scale);
+	}
+}
+
 void dpp1_dppclk_control(
 		struct dpp *dpp_base,
 		bool dppclk_div,
@@ -443,6 +490,7 @@ void dpp1_dppclk_control(
 }
 
 static const struct dpp_funcs dcn10_dpp_funcs = {
+		.dpp_read_state = dpp_read_state,
 		.dpp_reset = dpp_reset,
 		.dpp_set_scaler = dpp1_dscl_set_scaler_manual_scale,
 		.dpp_get_optimal_number_of_taps = dpp_get_optimal_number_of_taps,
@@ -463,6 +511,7 @@ static const struct dpp_funcs dcn10_dpp_funcs = {
 		.dpp_full_bypass		= dpp1_full_bypass,
 		.set_cursor_attributes = dpp1_set_cursor_attributes,
 		.set_cursor_position = dpp1_set_cursor_position,
+		.set_optional_cursor_attributes = dpp1_cnv_set_optional_cursor_attributes,
 		.dpp_dppclk_control = dpp1_dppclk_control,
 		.dpp_set_hdr_multiplier = dpp1_set_hdr_multiplier,
 };
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
index 17b062a8f88a..e2889e61b18c 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp.h
@@ -44,6 +44,10 @@
 #define TF_REG_LIST_DCN(id) \
 	SRI(CM_GAMUT_REMAP_CONTROL, CM, id),\
 	SRI(CM_GAMUT_REMAP_C11_C12, CM, id),\
+	SRI(CM_GAMUT_REMAP_C13_C14, CM, id),\
+	SRI(CM_GAMUT_REMAP_C21_C22, CM, id),\
+	SRI(CM_GAMUT_REMAP_C23_C24, CM, id),\
+	SRI(CM_GAMUT_REMAP_C31_C32, CM, id),\
 	SRI(CM_GAMUT_REMAP_C33_C34, CM, id),\
 	SRI(DSCL_EXT_OVERSCAN_LEFT_RIGHT, DSCL, id), \
 	SRI(DSCL_EXT_OVERSCAN_TOP_BOTTOM, DSCL, id), \
@@ -108,11 +112,14 @@
 	SRI(CM_DGAM_LUT_DATA, CM, id), \
 	SRI(CM_CONTROL, CM, id), \
 	SRI(CM_DGAM_CONTROL, CM, id), \
+	SRI(CM_TEST_DEBUG_INDEX, CM, id), \
+	SRI(CM_TEST_DEBUG_DATA, CM, id), \
 	SRI(FORMAT_CONTROL, CNVC_CFG, id), \
 	SRI(CNVC_SURFACE_PIXEL_FORMAT, CNVC_CFG, id), \
 	SRI(CURSOR0_CONTROL, CNVC_CUR, id), \
 	SRI(CURSOR0_COLOR0, CNVC_CUR, id), \
 	SRI(CURSOR0_COLOR1, CNVC_CUR, id), \
+	SRI(CURSOR0_FP_SCALE_BIAS, CNVC_CUR, id), \
 	SRI(DPP_CONTROL, DPP_TOP, id), \
 	SRI(CM_HDR_MULT_COEF, CM, id)
 
@@ -175,6 +182,14 @@
 	TF_SF(CM0_CM_GAMUT_REMAP_CONTROL, CM_GAMUT_REMAP_MODE, mask_sh),\
 	TF_SF(CM0_CM_GAMUT_REMAP_C11_C12, CM_GAMUT_REMAP_C11, mask_sh),\
 	TF_SF(CM0_CM_GAMUT_REMAP_C11_C12, CM_GAMUT_REMAP_C12, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C13_C14, CM_GAMUT_REMAP_C13, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C13_C14, CM_GAMUT_REMAP_C14, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C21_C22, CM_GAMUT_REMAP_C21, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C21_C22, CM_GAMUT_REMAP_C22, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C23_C24, CM_GAMUT_REMAP_C23, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C23_C24, CM_GAMUT_REMAP_C24, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C31_C32, CM_GAMUT_REMAP_C31, mask_sh),\
+	TF_SF(CM0_CM_GAMUT_REMAP_C31_C32, CM_GAMUT_REMAP_C32, mask_sh),\
 	TF_SF(CM0_CM_GAMUT_REMAP_C33_C34, CM_GAMUT_REMAP_C33, mask_sh),\
 	TF_SF(CM0_CM_GAMUT_REMAP_C33_C34, CM_GAMUT_REMAP_C34, mask_sh),\
 	TF_SF(DSCL0_DSCL_EXT_OVERSCAN_LEFT_RIGHT, EXT_OVERSCAN_LEFT, mask_sh),\
@@ -300,6 +315,7 @@
 	TF_SF(CM0_CM_DGAM_LUT_INDEX, CM_DGAM_LUT_INDEX, mask_sh), \
 	TF_SF(CM0_CM_DGAM_LUT_DATA, CM_DGAM_LUT_DATA, mask_sh), \
 	TF_SF(CM0_CM_DGAM_CONTROL, CM_DGAM_LUT_MODE, mask_sh), \
+	TF_SF(CM0_CM_TEST_DEBUG_INDEX, CM_TEST_DEBUG_INDEX, mask_sh), \
 	TF_SF(CNVC_CFG0_FORMAT_CONTROL, CNVC_BYPASS, mask_sh), \
 	TF2_SF(CNVC_CFG0, FORMAT_CONTROL__ALPHA_EN, mask_sh), \
 	TF_SF(CNVC_CFG0_FORMAT_CONTROL, FORMAT_EXPANSION_MODE, mask_sh), \
@@ -309,6 +325,8 @@
 	TF_SF(CNVC_CUR0_CURSOR0_CONTROL, CUR0_ENABLE, mask_sh), \
 	TF_SF(CNVC_CUR0_CURSOR0_COLOR0, CUR0_COLOR0, mask_sh), \
 	TF_SF(CNVC_CUR0_CURSOR0_COLOR1, CUR0_COLOR1, mask_sh), \
+	TF_SF(CNVC_CUR0_CURSOR0_FP_SCALE_BIAS, CUR0_FP_BIAS, mask_sh), \
+	TF_SF(CNVC_CUR0_CURSOR0_FP_SCALE_BIAS, CUR0_FP_SCALE, mask_sh), \
 	TF_SF(DPP_TOP0_DPP_CONTROL, DPP_CLOCK_ENABLE, mask_sh), \
 	TF_SF(CM0_CM_HDR_MULT_COEF, CM_HDR_MULT_COEF, mask_sh)
 
@@ -417,6 +435,41 @@
 	TF_SF(CURSOR0_CURSOR_CONTROL, CURSOR_ENABLE, mask_sh), \
 	TF_SF(DPP_TOP0_DPP_CONTROL, DPPCLK_RATE_CONTROL, mask_sh)
 
+/*
+ *
+	DCN1 CM debug status register definition
+
+	register :ID9_CM_STATUS do
+	implement_ref :cm
+	map to:  :cmdebugind, at: j
+	width 32
+	disclosure   NEVER
+
+		field :ID9_VUPDATE_CFG, [0], R
+		field :ID9_IGAM_LUT_MODE, [2..1], R
+		field :ID9_BNS_BYPASS, [3], R
+		field :ID9_ICSC_MODE, [5..4], R
+		field :ID9_DGAM_LUT_MODE, [8..6], R
+		field :ID9_HDR_BYPASS, [9], R
+		field :ID9_GAMUT_REMAP_MODE, [11..10], R
+		field :ID9_RGAM_LUT_MODE, [14..12], R
+		#1 free bit
+		field :ID9_OCSC_MODE, [18..16], R
+		field :ID9_DENORM_MODE, [21..19], R
+		field :ID9_ROUND_TRUNC_MODE, [25..22], R
+		field :ID9_DITHER_EN, [26], R
+		field :ID9_DITHER_MODE, [28..27], R
+	end
+*/
+
+#define TF_DEBUG_REG_LIST_SH_DCN10 \
+	.CM_TEST_DEBUG_DATA_ID9_ICSC_MODE = 4, \
+	.CM_TEST_DEBUG_DATA_ID9_OCSC_MODE = 16
+
+#define TF_DEBUG_REG_LIST_MASK_DCN10 \
+	.CM_TEST_DEBUG_DATA_ID9_ICSC_MODE = 0x30, \
+	.CM_TEST_DEBUG_DATA_ID9_OCSC_MODE = 0x70000
+
 #define TF_REG_FIELD_LIST(type) \
 	type EXT_OVERSCAN_LEFT; \
 	type EXT_OVERSCAN_RIGHT; \
@@ -486,6 +539,14 @@
 	type CM_GAMUT_REMAP_MODE; \
 	type CM_GAMUT_REMAP_C11; \
 	type CM_GAMUT_REMAP_C12; \
+	type CM_GAMUT_REMAP_C13; \
+	type CM_GAMUT_REMAP_C14; \
+	type CM_GAMUT_REMAP_C21; \
+	type CM_GAMUT_REMAP_C22; \
+	type CM_GAMUT_REMAP_C23; \
+	type CM_GAMUT_REMAP_C24; \
+	type CM_GAMUT_REMAP_C31; \
+	type CM_GAMUT_REMAP_C32; \
 	type CM_GAMUT_REMAP_C33; \
 	type CM_GAMUT_REMAP_C34; \
 	type CM_COMA_C11; \
@@ -1010,12 +1071,17 @@
 	type CUR0_EXPANSION_MODE; \
 	type CUR0_ENABLE; \
 	type CM_BYPASS; \
+	type CM_TEST_DEBUG_INDEX; \
+	type CM_TEST_DEBUG_DATA_ID9_ICSC_MODE; \
+	type CM_TEST_DEBUG_DATA_ID9_OCSC_MODE;\
 	type FORMAT_CONTROL__ALPHA_EN; \
 	type CUR0_COLOR0; \
 	type CUR0_COLOR1; \
 	type DPPCLK_RATE_CONTROL; \
 	type DPP_CLOCK_ENABLE; \
-	type CM_HDR_MULT_COEF;
+	type CM_HDR_MULT_COEF; \
+	type CUR0_FP_BIAS; \
+	type CUR0_FP_SCALE;
 
 struct dcn_dpp_shift {
 	TF_REG_FIELD_LIST(uint8_t)
@@ -1054,6 +1120,10 @@ struct dcn_dpp_mask {
 	uint32_t RECOUT_SIZE; \
 	uint32_t CM_GAMUT_REMAP_CONTROL; \
 	uint32_t CM_GAMUT_REMAP_C11_C12; \
+	uint32_t CM_GAMUT_REMAP_C13_C14; \
+	uint32_t CM_GAMUT_REMAP_C21_C22; \
+	uint32_t CM_GAMUT_REMAP_C23_C24; \
+	uint32_t CM_GAMUT_REMAP_C31_C32; \
 	uint32_t CM_GAMUT_REMAP_C33_C34; \
 	uint32_t CM_COMA_C11_C12; \
 	uint32_t CM_COMA_C33_C34; \
@@ -1255,6 +1325,8 @@ struct dcn_dpp_mask {
 	uint32_t CM_IGAM_LUT_RW_CONTROL; \
 	uint32_t CM_IGAM_LUT_RW_INDEX; \
 	uint32_t CM_IGAM_LUT_SEQ_COLOR; \
+	uint32_t CM_TEST_DEBUG_INDEX; \
+	uint32_t CM_TEST_DEBUG_DATA; \
 	uint32_t FORMAT_CONTROL; \
 	uint32_t CNVC_SURFACE_PIXEL_FORMAT; \
 	uint32_t CURSOR_CONTROL; \
@@ -1262,7 +1334,8 @@ struct dcn_dpp_mask {
 	uint32_t CURSOR0_COLOR0; \
 	uint32_t CURSOR0_COLOR1; \
 	uint32_t DPP_CONTROL; \
-	uint32_t CM_HDR_MULT_COEF;
+	uint32_t CM_HDR_MULT_COEF; \
+	uint32_t CURSOR0_FP_SCALE_BIAS;
 
 struct dcn_dpp_registers {
 	DPP_COMMON_REG_VARIABLE_LIST
@@ -1289,8 +1362,8 @@ struct dcn10_dpp {
 
 enum dcn10_input_csc_select {
 	INPUT_CSC_SELECT_BYPASS = 0,
-	INPUT_CSC_SELECT_ICSC,
-	INPUT_CSC_SELECT_COMA
+	INPUT_CSC_SELECT_ICSC = 1,
+	INPUT_CSC_SELECT_COMA = 2
 };
 
 void dpp1_set_cursor_attributes(
@@ -1303,6 +1376,10 @@ void dpp1_set_cursor_position(
 		const struct dc_cursor_mi_param *param,
 		uint32_t width);
 
+void dpp1_cnv_set_optional_cursor_attributes(
+			struct dpp *dpp_base,
+			struct dpp_cursor_attributes *attr);
+
 bool dpp1_dscl_is_lb_conf_valid(
 		int ceil_vratio,
 		int num_partitions,
@@ -1357,12 +1434,11 @@ void dpp1_set_degamma(
 		enum ipp_degamma_mode mode);
 
 void dpp1_set_degamma_pwl(struct dpp *dpp_base,
-								 const struct pwl_params *params);
+		const struct pwl_params *params);
+
 
-bool dpp_get_optimal_number_of_taps(
-		struct dpp *dpp,
-		struct scaler_data *scl_data,
-		const struct scaling_taps *in_taps);
+void dpp_read_state(struct dpp *dpp_base,
+		struct dcn_dpp_state *s);
 
 void dpp_reset(struct dpp *dpp_base);
 
@@ -1408,7 +1484,7 @@ void dpp1_cnv_setup (
 		struct dpp *dpp_base,
 		enum surface_pixel_format format,
 		enum expansion_mode mode,
-		struct csc_transform input_csc_color_matrix,
+		struct dc_csc_transform input_csc_color_matrix,
 		enum dc_color_space input_color_space);
 
 void dpp1_full_bypass(struct dpp *dpp_base);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c
index fb32975e4b67..116977eb24e2 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_cm.c
@@ -216,41 +216,55 @@ static void dpp1_cm_program_color_matrix(
 		struct dcn10_dpp *dpp,
 		const uint16_t *regval)
 {
-	uint32_t mode;
+	uint32_t ocsc_mode;
+	uint32_t cur_mode;
 	struct color_matrices_reg gam_regs;
 
-	REG_GET(CM_OCSC_CONTROL, CM_OCSC_MODE, &mode);
-
 	if (regval == NULL) {
 		BREAK_TO_DEBUGGER();
 		return;
 	}
-	mode = 4;
+
+	/* determine which CSC matrix (ocsc or comb) we are using
+	 * currently.  select the alternate set to double buffer
+	 * the CSC update so CSC is updated on frame boundary
+	 */
+	REG_SET(CM_TEST_DEBUG_INDEX, 0,
+			CM_TEST_DEBUG_INDEX, 9);
+
+	REG_GET(CM_TEST_DEBUG_DATA,
+			CM_TEST_DEBUG_DATA_ID9_OCSC_MODE, &cur_mode);
+
+	if (cur_mode != 4)
+		ocsc_mode = 4;
+	else
+		ocsc_mode = 5;
+
+
 	gam_regs.shifts.csc_c11 = dpp->tf_shift->CM_OCSC_C11;
 	gam_regs.masks.csc_c11  = dpp->tf_mask->CM_OCSC_C11;
 	gam_regs.shifts.csc_c12 = dpp->tf_shift->CM_OCSC_C12;
 	gam_regs.masks.csc_c12 = dpp->tf_mask->CM_OCSC_C12;
 
-	if (mode == 4) {
+	if (ocsc_mode == 4) {
 
 		gam_regs.csc_c11_c12 = REG(CM_OCSC_C11_C12);
 		gam_regs.csc_c33_c34 = REG(CM_OCSC_C33_C34);
 
-		cm_helper_program_color_matrices(
-				dpp->base.ctx,
-				regval,
-				&gam_regs);
-
 	} else {
 
 		gam_regs.csc_c11_c12 = REG(CM_COMB_C11_C12);
 		gam_regs.csc_c33_c34 = REG(CM_COMB_C33_C34);
 
-		cm_helper_program_color_matrices(
-				dpp->base.ctx,
-				regval,
-				&gam_regs);
 	}
+
+	cm_helper_program_color_matrices(
+			dpp->base.ctx,
+			regval,
+			&gam_regs);
+
+	REG_SET(CM_OCSC_CONTROL, 0, CM_OCSC_MODE, ocsc_mode);
+
 }
 
 void dpp1_cm_set_output_csc_default(
@@ -260,15 +274,14 @@ void dpp1_cm_set_output_csc_default(
 	struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base);
 	const uint16_t *regval = NULL;
 	int arr_size;
-	uint32_t ocsc_mode = 4;
 
 	regval = find_color_matrix(colorspace, &arr_size);
 	if (regval == NULL) {
 		BREAK_TO_DEBUGGER();
 		return;
 	}
+
 	dpp1_cm_program_color_matrix(dpp, regval);
-	REG_SET(CM_OCSC_CONTROL, 0, CM_OCSC_MODE, ocsc_mode);
 }
 
 static void dpp1_cm_get_reg_field(
@@ -329,9 +342,8 @@ void dpp1_cm_set_output_csc_adjustment(
 		const uint16_t *regval)
 {
 	struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base);
-	uint32_t ocsc_mode = 4;
+
 	dpp1_cm_program_color_matrix(dpp, regval);
-	REG_SET(CM_OCSC_CONTROL, 0, CM_OCSC_MODE, ocsc_mode);
 }
 
 void dpp1_cm_power_on_regamma_lut(struct dpp *dpp_base,
@@ -437,17 +449,18 @@ void dpp1_cm_program_regamma_lutb_settings(
 void dpp1_program_input_csc(
 		struct dpp *dpp_base,
 		enum dc_color_space color_space,
-		enum dcn10_input_csc_select select,
+		enum dcn10_input_csc_select input_select,
 		const struct out_csc_color_matrix *tbl_entry)
 {
 	struct dcn10_dpp *dpp = TO_DCN10_DPP(dpp_base);
 	int i;
 	int arr_size = sizeof(dcn10_input_csc_matrix)/sizeof(struct dcn10_input_csc_matrix);
 	const uint16_t *regval = NULL;
-	uint32_t selection = 1;
+	uint32_t cur_select = 0;
+	enum dcn10_input_csc_select select;
 	struct color_matrices_reg gam_regs;
 
-	if (select == INPUT_CSC_SELECT_BYPASS) {
+	if (input_select == INPUT_CSC_SELECT_BYPASS) {
 		REG_SET(CM_ICSC_CONTROL, 0, CM_ICSC_MODE, 0);
 		return;
 	}
@@ -467,36 +480,45 @@ void dpp1_program_input_csc(
 		regval = tbl_entry->regval;
 	}
 
-	if (select == INPUT_CSC_SELECT_COMA)
-		selection = 2;
-	REG_SET(CM_ICSC_CONTROL, 0,
-			CM_ICSC_MODE, selection);
+	/* determine which CSC matrix (icsc or coma) we are using
+	 * currently.  select the alternate set to double buffer
+	 * the CSC update so CSC is updated on frame boundary
+	 */
+	REG_SET(CM_TEST_DEBUG_INDEX, 0,
+			CM_TEST_DEBUG_INDEX, 9);
+
+	REG_GET(CM_TEST_DEBUG_DATA,
+			CM_TEST_DEBUG_DATA_ID9_ICSC_MODE, &cur_select);
+
+	if (cur_select != INPUT_CSC_SELECT_ICSC)
+		select = INPUT_CSC_SELECT_ICSC;
+	else
+		select = INPUT_CSC_SELECT_COMA;
 
 	gam_regs.shifts.csc_c11 = dpp->tf_shift->CM_ICSC_C11;
 	gam_regs.masks.csc_c11  = dpp->tf_mask->CM_ICSC_C11;
 	gam_regs.shifts.csc_c12 = dpp->tf_shift->CM_ICSC_C12;
 	gam_regs.masks.csc_c12 = dpp->tf_mask->CM_ICSC_C12;
 
-
 	if (select == INPUT_CSC_SELECT_ICSC) {
 
 		gam_regs.csc_c11_c12 = REG(CM_ICSC_C11_C12);
 		gam_regs.csc_c33_c34 = REG(CM_ICSC_C33_C34);
 
-		cm_helper_program_color_matrices(
-				dpp->base.ctx,
-				regval,
-				&gam_regs);
 	} else {
 
 		gam_regs.csc_c11_c12 = REG(CM_COMA_C11_C12);
 		gam_regs.csc_c33_c34 = REG(CM_COMA_C33_C34);
 
-		cm_helper_program_color_matrices(
-				dpp->base.ctx,
-				regval,
-				&gam_regs);
 	}
+
+	cm_helper_program_color_matrices(
+			dpp->base.ctx,
+			regval,
+			&gam_regs);
+
+	REG_SET(CM_ICSC_CONTROL, 0,
+				CM_ICSC_MODE, select);
 }
 
 //keep here for now, decide multi dce support later
@@ -789,13 +811,13 @@ void dpp1_program_input_lut(
 	REG_UPDATE(CM_IGAM_LUT_RW_INDEX, CM_IGAM_LUT_RW_INDEX, 0);
 	for (i = 0; i < gamma->num_entries; i++) {
 		REG_SET(CM_IGAM_LUT_SEQ_COLOR, 0, CM_IGAM_LUT_SEQ_COLOR,
-				dal_fixed31_32_round(
+				dc_fixpt_round(
 					gamma->entries.red[i]));
 		REG_SET(CM_IGAM_LUT_SEQ_COLOR, 0, CM_IGAM_LUT_SEQ_COLOR,
-				dal_fixed31_32_round(
+				dc_fixpt_round(
 					gamma->entries.green[i]));
 		REG_SET(CM_IGAM_LUT_SEQ_COLOR, 0, CM_IGAM_LUT_SEQ_COLOR,
-				dal_fixed31_32_round(
+				dc_fixpt_round(
 					gamma->entries.blue[i]));
 	}
 	// Power off LUT memory
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c
index 3eb824debf43..4a863a5dab41 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_dpp_dscl.c
@@ -169,7 +169,7 @@ static enum dscl_mode_sel dpp1_dscl_get_dscl_mode(
 		const struct scaler_data *data,
 		bool dbg_always_scale)
 {
-	const long long one = dal_fixed31_32_one.value;
+	const long long one = dc_fixpt_one.value;
 
 	if (dpp_base->caps->dscl_data_proc_format == DSCL_DATA_PRCESSING_FIXED_FORMAT) {
 		/* DSCL is processing data in fixed format */
@@ -464,8 +464,8 @@ static enum lb_memory_config dpp1_dscl_find_lb_memory_config(struct dcn10_dpp *d
 	int num_part_y, num_part_c;
 	int vtaps = scl_data->taps.v_taps;
 	int vtaps_c = scl_data->taps.v_taps_c;
-	int ceil_vratio = dal_fixed31_32_ceil(scl_data->ratios.vert);
-	int ceil_vratio_c = dal_fixed31_32_ceil(scl_data->ratios.vert_c);
+	int ceil_vratio = dc_fixpt_ceil(scl_data->ratios.vert);
+	int ceil_vratio_c = dc_fixpt_ceil(scl_data->ratios.vert_c);
 	enum lb_memory_config mem_cfg = LB_MEMORY_CONFIG_0;
 
 	if (dpp->base.ctx->dc->debug.use_max_lb)
@@ -565,52 +565,52 @@ static void dpp1_dscl_set_manual_ratio_init(
 	uint32_t init_int = 0;
 
 	REG_SET(SCL_HORZ_FILTER_SCALE_RATIO, 0,
-			SCL_H_SCALE_RATIO, dal_fixed31_32_u2d19(data->ratios.horz) << 5);
+			SCL_H_SCALE_RATIO, dc_fixpt_u3d19(data->ratios.horz) << 5);
 
 	REG_SET(SCL_VERT_FILTER_SCALE_RATIO, 0,
-			SCL_V_SCALE_RATIO, dal_fixed31_32_u2d19(data->ratios.vert) << 5);
+			SCL_V_SCALE_RATIO, dc_fixpt_u3d19(data->ratios.vert) << 5);
 
 	REG_SET(SCL_HORZ_FILTER_SCALE_RATIO_C, 0,
-			SCL_H_SCALE_RATIO_C, dal_fixed31_32_u2d19(data->ratios.horz_c) << 5);
+			SCL_H_SCALE_RATIO_C, dc_fixpt_u3d19(data->ratios.horz_c) << 5);
 
 	REG_SET(SCL_VERT_FILTER_SCALE_RATIO_C, 0,
-			SCL_V_SCALE_RATIO_C, dal_fixed31_32_u2d19(data->ratios.vert_c) << 5);
+			SCL_V_SCALE_RATIO_C, dc_fixpt_u3d19(data->ratios.vert_c) << 5);
 
 	/*
 	 * 0.24 format for fraction, first five bits zeroed
 	 */
-	init_frac = dal_fixed31_32_u0d19(data->inits.h) << 5;
-	init_int = dal_fixed31_32_floor(data->inits.h);
+	init_frac = dc_fixpt_u0d19(data->inits.h) << 5;
+	init_int = dc_fixpt_floor(data->inits.h);
 	REG_SET_2(SCL_HORZ_FILTER_INIT, 0,
 		SCL_H_INIT_FRAC, init_frac,
 		SCL_H_INIT_INT, init_int);
 
-	init_frac = dal_fixed31_32_u0d19(data->inits.h_c) << 5;
-	init_int = dal_fixed31_32_floor(data->inits.h_c);
+	init_frac = dc_fixpt_u0d19(data->inits.h_c) << 5;
+	init_int = dc_fixpt_floor(data->inits.h_c);
 	REG_SET_2(SCL_HORZ_FILTER_INIT_C, 0,
 		SCL_H_INIT_FRAC_C, init_frac,
 		SCL_H_INIT_INT_C, init_int);
 
-	init_frac = dal_fixed31_32_u0d19(data->inits.v) << 5;
-	init_int = dal_fixed31_32_floor(data->inits.v);
+	init_frac = dc_fixpt_u0d19(data->inits.v) << 5;
+	init_int = dc_fixpt_floor(data->inits.v);
 	REG_SET_2(SCL_VERT_FILTER_INIT, 0,
 		SCL_V_INIT_FRAC, init_frac,
 		SCL_V_INIT_INT, init_int);
 
-	init_frac = dal_fixed31_32_u0d19(data->inits.v_bot) << 5;
-	init_int = dal_fixed31_32_floor(data->inits.v_bot);
+	init_frac = dc_fixpt_u0d19(data->inits.v_bot) << 5;
+	init_int = dc_fixpt_floor(data->inits.v_bot);
 	REG_SET_2(SCL_VERT_FILTER_INIT_BOT, 0,
 		SCL_V_INIT_FRAC_BOT, init_frac,
 		SCL_V_INIT_INT_BOT, init_int);
 
-	init_frac = dal_fixed31_32_u0d19(data->inits.v_c) << 5;
-	init_int = dal_fixed31_32_floor(data->inits.v_c);
+	init_frac = dc_fixpt_u0d19(data->inits.v_c) << 5;
+	init_int = dc_fixpt_floor(data->inits.v_c);
 	REG_SET_2(SCL_VERT_FILTER_INIT_C, 0,
 		SCL_V_INIT_FRAC_C, init_frac,
 		SCL_V_INIT_INT_C, init_int);
 
-	init_frac = dal_fixed31_32_u0d19(data->inits.v_c_bot) << 5;
-	init_int = dal_fixed31_32_floor(data->inits.v_c_bot);
+	init_frac = dc_fixpt_u0d19(data->inits.v_c_bot) << 5;
+	init_int = dc_fixpt_floor(data->inits.v_c_bot);
 	REG_SET_2(SCL_VERT_FILTER_INIT_BOT_C, 0,
 		SCL_V_INIT_FRAC_BOT_C, init_frac,
 		SCL_V_INIT_INT_BOT_C, init_int);
@@ -621,6 +621,10 @@ static void dpp1_dscl_set_manual_ratio_init(
 static void dpp1_dscl_set_recout(
 			struct dcn10_dpp *dpp, const struct rect *recout)
 {
+	int visual_confirm_on = 0;
+	if (dpp->base.ctx->dc->debug.visual_confirm != VISUAL_CONFIRM_DISABLE)
+		visual_confirm_on = 1;
+
 	REG_SET_2(RECOUT_START, 0,
 		/* First pixel of RECOUT */
 			 RECOUT_START_X, recout->x,
@@ -632,8 +636,7 @@ static void dpp1_dscl_set_recout(
 			 RECOUT_WIDTH, recout->width,
 		/* Number of RECOUT vertical lines */
 			 RECOUT_HEIGHT, recout->height
-			 - dpp->base.ctx->dc->debug.surface_visual_confirm * 4 *
-			 (dpp->base.inst + 1));
+			 - visual_confirm_on * 4 * (dpp->base.inst + 1));
 }
 
 /* Main function to program scaler and line buffer in manual scaling mode */
@@ -655,6 +658,12 @@ void dpp1_dscl_set_scaler_manual_scale(
 
 	dpp->scl_data = *scl_data;
 
+	/* Autocal off */
+	REG_SET_3(DSCL_AUTOCAL, 0,
+		AUTOCAL_MODE, AUTOCAL_MODE_OFF,
+		AUTOCAL_NUM_PIPE, 0,
+		AUTOCAL_PIPE_ID, 0);
+
 	/* Recout */
 	dpp1_dscl_set_recout(dpp, &scl_data->recout);
 
@@ -678,12 +687,6 @@ void dpp1_dscl_set_scaler_manual_scale(
 	if (dscl_mode == DSCL_MODE_SCALING_444_BYPASS)
 		return;
 
-	/* Autocal off */
-	REG_SET_3(DSCL_AUTOCAL, 0,
-		AUTOCAL_MODE, AUTOCAL_MODE_OFF,
-		AUTOCAL_NUM_PIPE, 0,
-		AUTOCAL_PIPE_ID, 0);
-
 	/* Black offsets */
 	if (ycbcr)
 		REG_SET_2(SCL_BLACK_OFFSET, 0,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c
index 738f67ffd1b4..1ea91e153d3a 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.c
@@ -190,10 +190,17 @@ static uint32_t convert_and_clamp(
 }
 
 
+void hubbub1_wm_change_req_wa(struct hubbub *hubbub)
+{
+	REG_UPDATE_SEQ(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL,
+			DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, 0, 1);
+}
+
 void hubbub1_program_watermarks(
 		struct hubbub *hubbub,
 		struct dcn_watermark_set *watermarks,
-		unsigned int refclk_mhz)
+		unsigned int refclk_mhz,
+		bool safe_to_lower)
 {
 	uint32_t force_en = hubbub->ctx->dc->debug.disable_stutter ? 1 : 0;
 	/*
@@ -202,191 +209,259 @@ void hubbub1_program_watermarks(
 	 */
 	uint32_t prog_wm_value;
 
-	REG_UPDATE(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL,
-			DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, 0);
 
 	/* Repeat for water mark set A, B, C and D. */
 	/* clock state A */
-	prog_wm_value = convert_and_clamp(watermarks->a.urgent_ns,
-			refclk_mhz, 0x1fffff);
-	REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A, prog_wm_value);
-
-	DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_A calculated =%d\n"
-		"HW register value = 0x%x\n",
-		watermarks->a.urgent_ns, prog_wm_value);
+	if (safe_to_lower || watermarks->a.urgent_ns > hubbub->watermarks.a.urgent_ns) {
+		hubbub->watermarks.a.urgent_ns = watermarks->a.urgent_ns;
+		prog_wm_value = convert_and_clamp(watermarks->a.urgent_ns,
+				refclk_mhz, 0x1fffff);
+		REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A, prog_wm_value);
 
-	prog_wm_value = convert_and_clamp(watermarks->a.pte_meta_urgent_ns,
-			refclk_mhz, 0x1fffff);
-	REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_A, prog_wm_value);
-	DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_A calculated =%d\n"
-		"HW register value = 0x%x\n",
-		watermarks->a.pte_meta_urgent_ns, prog_wm_value);
+		DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_A calculated =%d\n"
+			"HW register value = 0x%x\n",
+			watermarks->a.urgent_ns, prog_wm_value);
+	}
 
-	if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A)) {
-		prog_wm_value = convert_and_clamp(
-				watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns,
+	if (safe_to_lower || watermarks->a.pte_meta_urgent_ns > hubbub->watermarks.a.pte_meta_urgent_ns) {
+		hubbub->watermarks.a.pte_meta_urgent_ns = watermarks->a.pte_meta_urgent_ns;
+		prog_wm_value = convert_and_clamp(watermarks->a.pte_meta_urgent_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A, prog_wm_value);
-		DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_A calculated =%d\n"
+		REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_A, prog_wm_value);
+		DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_A calculated =%d\n"
 			"HW register value = 0x%x\n",
-			watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
+			watermarks->a.pte_meta_urgent_ns, prog_wm_value);
+	}
 
+	if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A)) {
+		if (safe_to_lower || watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns
+				> hubbub->watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns) {
+			hubbub->watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns =
+					watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns;
+			prog_wm_value = convert_and_clamp(
+					watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns,
+					refclk_mhz, 0x1fffff);
+			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A, prog_wm_value);
+			DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_A calculated =%d\n"
+				"HW register value = 0x%x\n",
+				watermarks->a.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
+		}
 
+		if (safe_to_lower || watermarks->a.cstate_pstate.cstate_exit_ns
+				> hubbub->watermarks.a.cstate_pstate.cstate_exit_ns) {
+			hubbub->watermarks.a.cstate_pstate.cstate_exit_ns =
+					watermarks->a.cstate_pstate.cstate_exit_ns;
+			prog_wm_value = convert_and_clamp(
+					watermarks->a.cstate_pstate.cstate_exit_ns,
+					refclk_mhz, 0x1fffff);
+			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_A, prog_wm_value);
+			DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_A calculated =%d\n"
+				"HW register value = 0x%x\n",
+				watermarks->a.cstate_pstate.cstate_exit_ns, prog_wm_value);
+		}
+	}
+
+	if (safe_to_lower || watermarks->a.cstate_pstate.pstate_change_ns
+			> hubbub->watermarks.a.cstate_pstate.pstate_change_ns) {
+		hubbub->watermarks.a.cstate_pstate.pstate_change_ns =
+				watermarks->a.cstate_pstate.pstate_change_ns;
 		prog_wm_value = convert_and_clamp(
-				watermarks->a.cstate_pstate.cstate_exit_ns,
+				watermarks->a.cstate_pstate.pstate_change_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_A, prog_wm_value);
-		DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_A calculated =%d\n"
-			"HW register value = 0x%x\n",
-			watermarks->a.cstate_pstate.cstate_exit_ns, prog_wm_value);
+		REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A, prog_wm_value);
+		DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_A calculated =%d\n"
+			"HW register value = 0x%x\n\n",
+			watermarks->a.cstate_pstate.pstate_change_ns, prog_wm_value);
 	}
 
-	prog_wm_value = convert_and_clamp(
-			watermarks->a.cstate_pstate.pstate_change_ns,
-			refclk_mhz, 0x1fffff);
-	REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A, prog_wm_value);
-	DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_A calculated =%d\n"
-		"HW register value = 0x%x\n\n",
-		watermarks->a.cstate_pstate.pstate_change_ns, prog_wm_value);
-
-
 	/* clock state B */
-	prog_wm_value = convert_and_clamp(
-			watermarks->b.urgent_ns, refclk_mhz, 0x1fffff);
-	REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_B, prog_wm_value);
-	DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_B calculated =%d\n"
-		"HW register value = 0x%x\n",
-		watermarks->b.urgent_ns, prog_wm_value);
-
-
-	prog_wm_value = convert_and_clamp(
-			watermarks->b.pte_meta_urgent_ns,
-			refclk_mhz, 0x1fffff);
-	REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_B, prog_wm_value);
-	DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_B calculated =%d\n"
-		"HW register value = 0x%x\n",
-		watermarks->b.pte_meta_urgent_ns, prog_wm_value);
+	if (safe_to_lower || watermarks->b.urgent_ns > hubbub->watermarks.b.urgent_ns) {
+		hubbub->watermarks.b.urgent_ns = watermarks->b.urgent_ns;
+		prog_wm_value = convert_and_clamp(watermarks->b.urgent_ns,
+				refclk_mhz, 0x1fffff);
+		REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_B, prog_wm_value);
 
+		DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_B calculated =%d\n"
+			"HW register value = 0x%x\n",
+			watermarks->b.urgent_ns, prog_wm_value);
+	}
 
-	if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B)) {
-		prog_wm_value = convert_and_clamp(
-				watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns,
+	if (safe_to_lower || watermarks->b.pte_meta_urgent_ns > hubbub->watermarks.b.pte_meta_urgent_ns) {
+		hubbub->watermarks.b.pte_meta_urgent_ns = watermarks->b.pte_meta_urgent_ns;
+		prog_wm_value = convert_and_clamp(watermarks->b.pte_meta_urgent_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B, prog_wm_value);
-		DC_LOG_BANDWIDTH_CALCS("SR_ENTER_WATERMARK_B calculated =%d\n"
+		REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_B, prog_wm_value);
+		DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_B calculated =%d\n"
 			"HW register value = 0x%x\n",
-			watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
+			watermarks->b.pte_meta_urgent_ns, prog_wm_value);
+	}
+
+	if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B)) {
+		if (safe_to_lower || watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns
+				> hubbub->watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns) {
+			hubbub->watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns =
+					watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns;
+			prog_wm_value = convert_and_clamp(
+					watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns,
+					refclk_mhz, 0x1fffff);
+			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_B, prog_wm_value);
+			DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_B calculated =%d\n"
+				"HW register value = 0x%x\n",
+				watermarks->b.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
+		}
 
+		if (safe_to_lower || watermarks->b.cstate_pstate.cstate_exit_ns
+				> hubbub->watermarks.b.cstate_pstate.cstate_exit_ns) {
+			hubbub->watermarks.b.cstate_pstate.cstate_exit_ns =
+					watermarks->b.cstate_pstate.cstate_exit_ns;
+			prog_wm_value = convert_and_clamp(
+					watermarks->b.cstate_pstate.cstate_exit_ns,
+					refclk_mhz, 0x1fffff);
+			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_B, prog_wm_value);
+			DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_B calculated =%d\n"
+				"HW register value = 0x%x\n",
+				watermarks->b.cstate_pstate.cstate_exit_ns, prog_wm_value);
+		}
+	}
 
+	if (safe_to_lower || watermarks->b.cstate_pstate.pstate_change_ns
+			> hubbub->watermarks.b.cstate_pstate.pstate_change_ns) {
+		hubbub->watermarks.b.cstate_pstate.pstate_change_ns =
+				watermarks->b.cstate_pstate.pstate_change_ns;
 		prog_wm_value = convert_and_clamp(
-				watermarks->b.cstate_pstate.cstate_exit_ns,
+				watermarks->b.cstate_pstate.pstate_change_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_B, prog_wm_value);
-		DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_B calculated =%d\n"
-			"HW register value = 0x%x\n",
-			watermarks->b.cstate_pstate.cstate_exit_ns, prog_wm_value);
+		REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B, prog_wm_value);
+		DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_B calculated =%d\n"
+			"HW register value = 0x%x\n\n",
+			watermarks->b.cstate_pstate.pstate_change_ns, prog_wm_value);
 	}
 
-	prog_wm_value = convert_and_clamp(
-			watermarks->b.cstate_pstate.pstate_change_ns,
-			refclk_mhz, 0x1fffff);
-	REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B, prog_wm_value);
-	DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_B calculated =%d\n\n"
-		"HW register value = 0x%x\n",
-		watermarks->b.cstate_pstate.pstate_change_ns, prog_wm_value);
-
 	/* clock state C */
-	prog_wm_value = convert_and_clamp(
-			watermarks->c.urgent_ns, refclk_mhz, 0x1fffff);
-	REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_C, prog_wm_value);
-	DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_C calculated =%d\n"
-		"HW register value = 0x%x\n",
-		watermarks->c.urgent_ns, prog_wm_value);
-
-
-	prog_wm_value = convert_and_clamp(
-			watermarks->c.pte_meta_urgent_ns,
-			refclk_mhz, 0x1fffff);
-	REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_C, prog_wm_value);
-	DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_C calculated =%d\n"
-		"HW register value = 0x%x\n",
-		watermarks->c.pte_meta_urgent_ns, prog_wm_value);
+	if (safe_to_lower || watermarks->c.urgent_ns > hubbub->watermarks.c.urgent_ns) {
+		hubbub->watermarks.c.urgent_ns = watermarks->c.urgent_ns;
+		prog_wm_value = convert_and_clamp(watermarks->c.urgent_ns,
+				refclk_mhz, 0x1fffff);
+		REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_C, prog_wm_value);
 
+		DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_C calculated =%d\n"
+			"HW register value = 0x%x\n",
+			watermarks->c.urgent_ns, prog_wm_value);
+	}
 
-	if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C)) {
-		prog_wm_value = convert_and_clamp(
-				watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns,
+	if (safe_to_lower || watermarks->c.pte_meta_urgent_ns > hubbub->watermarks.c.pte_meta_urgent_ns) {
+		hubbub->watermarks.c.pte_meta_urgent_ns = watermarks->c.pte_meta_urgent_ns;
+		prog_wm_value = convert_and_clamp(watermarks->c.pte_meta_urgent_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C, prog_wm_value);
-		DC_LOG_BANDWIDTH_CALCS("SR_ENTER_WATERMARK_C calculated =%d\n"
+		REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_C, prog_wm_value);
+		DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_C calculated =%d\n"
 			"HW register value = 0x%x\n",
-			watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
+			watermarks->c.pte_meta_urgent_ns, prog_wm_value);
+	}
 
+	if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C)) {
+		if (safe_to_lower || watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns
+				> hubbub->watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns) {
+			hubbub->watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns =
+					watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns;
+			prog_wm_value = convert_and_clamp(
+					watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns,
+					refclk_mhz, 0x1fffff);
+			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_C, prog_wm_value);
+			DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_C calculated =%d\n"
+				"HW register value = 0x%x\n",
+				watermarks->c.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
+		}
 
+		if (safe_to_lower || watermarks->c.cstate_pstate.cstate_exit_ns
+				> hubbub->watermarks.c.cstate_pstate.cstate_exit_ns) {
+			hubbub->watermarks.c.cstate_pstate.cstate_exit_ns =
+					watermarks->c.cstate_pstate.cstate_exit_ns;
+			prog_wm_value = convert_and_clamp(
+					watermarks->c.cstate_pstate.cstate_exit_ns,
+					refclk_mhz, 0x1fffff);
+			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_C, prog_wm_value);
+			DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_C calculated =%d\n"
+				"HW register value = 0x%x\n",
+				watermarks->c.cstate_pstate.cstate_exit_ns, prog_wm_value);
+		}
+	}
+
+	if (safe_to_lower || watermarks->c.cstate_pstate.pstate_change_ns
+			> hubbub->watermarks.c.cstate_pstate.pstate_change_ns) {
+		hubbub->watermarks.c.cstate_pstate.pstate_change_ns =
+				watermarks->c.cstate_pstate.pstate_change_ns;
 		prog_wm_value = convert_and_clamp(
-				watermarks->c.cstate_pstate.cstate_exit_ns,
+				watermarks->c.cstate_pstate.pstate_change_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_C, prog_wm_value);
-		DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_C calculated =%d\n"
-			"HW register value = 0x%x\n",
-			watermarks->c.cstate_pstate.cstate_exit_ns, prog_wm_value);
+		REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C, prog_wm_value);
+		DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_C calculated =%d\n"
+			"HW register value = 0x%x\n\n",
+			watermarks->c.cstate_pstate.pstate_change_ns, prog_wm_value);
 	}
 
-	prog_wm_value = convert_and_clamp(
-			watermarks->c.cstate_pstate.pstate_change_ns,
-			refclk_mhz, 0x1fffff);
-	REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C, prog_wm_value);
-	DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_C calculated =%d\n\n"
-		"HW register value = 0x%x\n",
-		watermarks->c.cstate_pstate.pstate_change_ns, prog_wm_value);
-
 	/* clock state D */
-	prog_wm_value = convert_and_clamp(
-			watermarks->d.urgent_ns, refclk_mhz, 0x1fffff);
-	REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_D, prog_wm_value);
-	DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_D calculated =%d\n"
-		"HW register value = 0x%x\n",
-		watermarks->d.urgent_ns, prog_wm_value);
-
-	prog_wm_value = convert_and_clamp(
-			watermarks->d.pte_meta_urgent_ns,
-			refclk_mhz, 0x1fffff);
-	REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_D, prog_wm_value);
-	DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_D calculated =%d\n"
-		"HW register value = 0x%x\n",
-		watermarks->d.pte_meta_urgent_ns, prog_wm_value);
-
-
-	if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D)) {
-		prog_wm_value = convert_and_clamp(
-				watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns,
+	if (safe_to_lower || watermarks->d.urgent_ns > hubbub->watermarks.d.urgent_ns) {
+		hubbub->watermarks.d.urgent_ns = watermarks->d.urgent_ns;
+		prog_wm_value = convert_and_clamp(watermarks->d.urgent_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D, prog_wm_value);
-		DC_LOG_BANDWIDTH_CALCS("SR_ENTER_WATERMARK_D calculated =%d\n"
-			"HW register value = 0x%x\n",
-			watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
+		REG_WRITE(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_D, prog_wm_value);
 
+		DC_LOG_BANDWIDTH_CALCS("URGENCY_WATERMARK_D calculated =%d\n"
+			"HW register value = 0x%x\n",
+			watermarks->d.urgent_ns, prog_wm_value);
+	}
 
-		prog_wm_value = convert_and_clamp(
-				watermarks->d.cstate_pstate.cstate_exit_ns,
+	if (safe_to_lower || watermarks->d.pte_meta_urgent_ns > hubbub->watermarks.d.pte_meta_urgent_ns) {
+		hubbub->watermarks.d.pte_meta_urgent_ns = watermarks->d.pte_meta_urgent_ns;
+		prog_wm_value = convert_and_clamp(watermarks->d.pte_meta_urgent_ns,
 				refclk_mhz, 0x1fffff);
-		REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_D, prog_wm_value);
-		DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_D calculated =%d\n"
+		REG_WRITE(DCHUBBUB_ARB_PTE_META_URGENCY_WATERMARK_D, prog_wm_value);
+		DC_LOG_BANDWIDTH_CALCS("PTE_META_URGENCY_WATERMARK_D calculated =%d\n"
 			"HW register value = 0x%x\n",
-			watermarks->d.cstate_pstate.cstate_exit_ns, prog_wm_value);
+			watermarks->d.pte_meta_urgent_ns, prog_wm_value);
 	}
 
+	if (REG(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D)) {
+		if (safe_to_lower || watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns
+				> hubbub->watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns) {
+			hubbub->watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns =
+					watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns;
+			prog_wm_value = convert_and_clamp(
+					watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns,
+					refclk_mhz, 0x1fffff);
+			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_D, prog_wm_value);
+			DC_LOG_BANDWIDTH_CALCS("SR_ENTER_EXIT_WATERMARK_D calculated =%d\n"
+				"HW register value = 0x%x\n",
+				watermarks->d.cstate_pstate.cstate_enter_plus_exit_ns, prog_wm_value);
+		}
 
-	prog_wm_value = convert_and_clamp(
-			watermarks->d.cstate_pstate.pstate_change_ns,
-			refclk_mhz, 0x1fffff);
-	REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, prog_wm_value);
-	DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_D calculated =%d\n"
-		"HW register value = 0x%x\n\n",
-		watermarks->d.cstate_pstate.pstate_change_ns, prog_wm_value);
+		if (safe_to_lower || watermarks->d.cstate_pstate.cstate_exit_ns
+				> hubbub->watermarks.d.cstate_pstate.cstate_exit_ns) {
+			hubbub->watermarks.d.cstate_pstate.cstate_exit_ns =
+					watermarks->d.cstate_pstate.cstate_exit_ns;
+			prog_wm_value = convert_and_clamp(
+					watermarks->d.cstate_pstate.cstate_exit_ns,
+					refclk_mhz, 0x1fffff);
+			REG_WRITE(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_D, prog_wm_value);
+			DC_LOG_BANDWIDTH_CALCS("SR_EXIT_WATERMARK_D calculated =%d\n"
+				"HW register value = 0x%x\n",
+				watermarks->d.cstate_pstate.cstate_exit_ns, prog_wm_value);
+		}
+	}
 
-	REG_UPDATE(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL,
-			DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, 1);
+	if (safe_to_lower || watermarks->d.cstate_pstate.pstate_change_ns
+			> hubbub->watermarks.d.cstate_pstate.pstate_change_ns) {
+		hubbub->watermarks.d.cstate_pstate.pstate_change_ns =
+				watermarks->d.cstate_pstate.pstate_change_ns;
+		prog_wm_value = convert_and_clamp(
+				watermarks->d.cstate_pstate.pstate_change_ns,
+				refclk_mhz, 0x1fffff);
+		REG_WRITE(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, prog_wm_value);
+		DC_LOG_BANDWIDTH_CALCS("DRAM_CLK_CHANGE_WATERMARK_D calculated =%d\n"
+			"HW register value = 0x%x\n\n",
+			watermarks->d.cstate_pstate.pstate_change_ns, prog_wm_value);
+	}
 
 	REG_UPDATE(DCHUBBUB_ARB_SAT_LEVEL,
 			DCHUBBUB_ARB_SAT_LEVEL, 60 * refclk_mhz);
@@ -408,6 +483,11 @@ void hubbub1_update_dchub(
 	struct hubbub *hubbub,
 	struct dchub_init_data *dh_data)
 {
+	if (REG(DCHUBBUB_SDPIF_FB_TOP) == 0) {
+		ASSERT(false);
+		/*should not come here*/
+		return;
+	}
 	/* TODO: port code from dal2 */
 	switch (dh_data->fb_mode) {
 	case FRAME_BUFFER_MODE_ZFB_ONLY:
@@ -476,8 +556,235 @@ void hubbub1_toggle_watermark_change_req(struct hubbub *hubbub)
 			DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, watermark_change_req);
 }
 
+void hubbub1_soft_reset(struct hubbub *hubbub, bool reset)
+{
+	uint32_t reset_en = reset ? 1 : 0;
+
+	REG_UPDATE(DCHUBBUB_SOFT_RESET,
+			DCHUBBUB_GLOBAL_SOFT_RESET, reset_en);
+}
+
+static bool hubbub1_dcc_support_swizzle(
+		enum swizzle_mode_values swizzle,
+		unsigned int bytes_per_element,
+		enum segment_order *segment_order_horz,
+		enum segment_order *segment_order_vert)
+{
+	bool standard_swizzle = false;
+	bool display_swizzle = false;
+
+	switch (swizzle) {
+	case DC_SW_4KB_S:
+	case DC_SW_64KB_S:
+	case DC_SW_VAR_S:
+	case DC_SW_4KB_S_X:
+	case DC_SW_64KB_S_X:
+	case DC_SW_VAR_S_X:
+		standard_swizzle = true;
+		break;
+	case DC_SW_4KB_D:
+	case DC_SW_64KB_D:
+	case DC_SW_VAR_D:
+	case DC_SW_4KB_D_X:
+	case DC_SW_64KB_D_X:
+	case DC_SW_VAR_D_X:
+		display_swizzle = true;
+		break;
+	default:
+		break;
+	}
+
+	if (bytes_per_element == 1 && standard_swizzle) {
+		*segment_order_horz = segment_order__contiguous;
+		*segment_order_vert = segment_order__na;
+		return true;
+	}
+	if (bytes_per_element == 2 && standard_swizzle) {
+		*segment_order_horz = segment_order__non_contiguous;
+		*segment_order_vert = segment_order__contiguous;
+		return true;
+	}
+	if (bytes_per_element == 4 && standard_swizzle) {
+		*segment_order_horz = segment_order__non_contiguous;
+		*segment_order_vert = segment_order__contiguous;
+		return true;
+	}
+	if (bytes_per_element == 8 && standard_swizzle) {
+		*segment_order_horz = segment_order__na;
+		*segment_order_vert = segment_order__contiguous;
+		return true;
+	}
+	if (bytes_per_element == 8 && display_swizzle) {
+		*segment_order_horz = segment_order__contiguous;
+		*segment_order_vert = segment_order__non_contiguous;
+		return true;
+	}
+
+	return false;
+}
+
+static bool hubbub1_dcc_support_pixel_format(
+		enum surface_pixel_format format,
+		unsigned int *bytes_per_element)
+{
+	/* DML: get_bytes_per_element */
+	switch (format) {
+	case SURFACE_PIXEL_FORMAT_GRPH_ARGB1555:
+	case SURFACE_PIXEL_FORMAT_GRPH_RGB565:
+		*bytes_per_element = 2;
+		return true;
+	case SURFACE_PIXEL_FORMAT_GRPH_ARGB8888:
+	case SURFACE_PIXEL_FORMAT_GRPH_ABGR8888:
+	case SURFACE_PIXEL_FORMAT_GRPH_ARGB2101010:
+	case SURFACE_PIXEL_FORMAT_GRPH_ABGR2101010:
+		*bytes_per_element = 4;
+		return true;
+	case SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616:
+	case SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616F:
+	case SURFACE_PIXEL_FORMAT_GRPH_ABGR16161616F:
+		*bytes_per_element = 8;
+		return true;
+	default:
+		return false;
+	}
+}
+
+static void hubbub1_get_blk256_size(unsigned int *blk256_width, unsigned int *blk256_height,
+		unsigned int bytes_per_element)
+{
+	/* copied from DML.  might want to refactor DML to leverage from DML */
+	/* DML : get_blk256_size */
+	if (bytes_per_element == 1) {
+		*blk256_width = 16;
+		*blk256_height = 16;
+	} else if (bytes_per_element == 2) {
+		*blk256_width = 16;
+		*blk256_height = 8;
+	} else if (bytes_per_element == 4) {
+		*blk256_width = 8;
+		*blk256_height = 8;
+	} else if (bytes_per_element == 8) {
+		*blk256_width = 8;
+		*blk256_height = 4;
+	}
+}
+
+static void hubbub1_det_request_size(
+		unsigned int height,
+		unsigned int width,
+		unsigned int bpe,
+		bool *req128_horz_wc,
+		bool *req128_vert_wc)
+{
+	unsigned int detile_buf_size = 164 * 1024;  /* 164KB for DCN1.0 */
+
+	unsigned int blk256_height = 0;
+	unsigned int blk256_width = 0;
+	unsigned int swath_bytes_horz_wc, swath_bytes_vert_wc;
+
+	hubbub1_get_blk256_size(&blk256_width, &blk256_height, bpe);
+
+	swath_bytes_horz_wc = height * blk256_height * bpe;
+	swath_bytes_vert_wc = width * blk256_width * bpe;
+
+	*req128_horz_wc = (2 * swath_bytes_horz_wc <= detile_buf_size) ?
+			false : /* full 256B request */
+			true; /* half 128b request */
+
+	*req128_vert_wc = (2 * swath_bytes_vert_wc <= detile_buf_size) ?
+			false : /* full 256B request */
+			true; /* half 128b request */
+}
+
+static bool hubbub1_get_dcc_compression_cap(struct hubbub *hubbub,
+		const struct dc_dcc_surface_param *input,
+		struct dc_surface_dcc_cap *output)
+{
+	struct dc *dc = hubbub->ctx->dc;
+	/* implement section 1.6.2.1 of DCN1_Programming_Guide.docx */
+	enum dcc_control dcc_control;
+	unsigned int bpe;
+	enum segment_order segment_order_horz, segment_order_vert;
+	bool req128_horz_wc, req128_vert_wc;
+
+	memset(output, 0, sizeof(*output));
+
+	if (dc->debug.disable_dcc == DCC_DISABLE)
+		return false;
+
+	if (!hubbub->funcs->dcc_support_pixel_format(input->format, &bpe))
+		return false;
+
+	if (!hubbub->funcs->dcc_support_swizzle(input->swizzle_mode, bpe,
+			&segment_order_horz, &segment_order_vert))
+		return false;
+
+	hubbub1_det_request_size(input->surface_size.height,  input->surface_size.width,
+			bpe, &req128_horz_wc, &req128_vert_wc);
+
+	if (!req128_horz_wc && !req128_vert_wc) {
+		dcc_control = dcc_control__256_256_xxx;
+	} else if (input->scan == SCAN_DIRECTION_HORIZONTAL) {
+		if (!req128_horz_wc)
+			dcc_control = dcc_control__256_256_xxx;
+		else if (segment_order_horz == segment_order__contiguous)
+			dcc_control = dcc_control__128_128_xxx;
+		else
+			dcc_control = dcc_control__256_64_64;
+	} else if (input->scan == SCAN_DIRECTION_VERTICAL) {
+		if (!req128_vert_wc)
+			dcc_control = dcc_control__256_256_xxx;
+		else if (segment_order_vert == segment_order__contiguous)
+			dcc_control = dcc_control__128_128_xxx;
+		else
+			dcc_control = dcc_control__256_64_64;
+	} else {
+		if ((req128_horz_wc &&
+			segment_order_horz == segment_order__non_contiguous) ||
+			(req128_vert_wc &&
+			segment_order_vert == segment_order__non_contiguous))
+			/* access_dir not known, must use most constraining */
+			dcc_control = dcc_control__256_64_64;
+		else
+			/* reg128 is true for either horz and vert
+			 * but segment_order is contiguous
+			 */
+			dcc_control = dcc_control__128_128_xxx;
+	}
+
+	if (dc->debug.disable_dcc == DCC_HALF_REQ_DISALBE &&
+		dcc_control != dcc_control__256_256_xxx)
+		return false;
+
+	switch (dcc_control) {
+	case dcc_control__256_256_xxx:
+		output->grph.rgb.max_uncompressed_blk_size = 256;
+		output->grph.rgb.max_compressed_blk_size = 256;
+		output->grph.rgb.independent_64b_blks = false;
+		break;
+	case dcc_control__128_128_xxx:
+		output->grph.rgb.max_uncompressed_blk_size = 128;
+		output->grph.rgb.max_compressed_blk_size = 128;
+		output->grph.rgb.independent_64b_blks = false;
+		break;
+	case dcc_control__256_64_64:
+		output->grph.rgb.max_uncompressed_blk_size = 256;
+		output->grph.rgb.max_compressed_blk_size = 64;
+		output->grph.rgb.independent_64b_blks = true;
+		break;
+	}
+
+	output->capable = true;
+	output->const_color_support = false;
+
+	return true;
+}
+
 static const struct hubbub_funcs hubbub1_funcs = {
-	.update_dchub = hubbub1_update_dchub
+	.update_dchub = hubbub1_update_dchub,
+	.dcc_support_swizzle = hubbub1_dcc_support_swizzle,
+	.dcc_support_pixel_format = hubbub1_dcc_support_pixel_format,
+	.get_dcc_compression_cap = hubbub1_get_dcc_compression_cap,
 };
 
 void hubbub1_construct(struct hubbub *hubbub,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h
index a16e908821a0..d6e596eef4c5 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h
@@ -27,6 +27,7 @@
 #define __DC_HUBBUB_DCN10_H__
 
 #include "core_types.h"
+#include "dchubbub.h"
 
 #define HUBHUB_REG_LIST_DCN()\
 	SR(DCHUBBUB_ARB_DATA_URGENCY_WATERMARK_A),\
@@ -47,7 +48,8 @@
 	SR(DCHUBBUB_ARB_DF_REQ_OUTSTAND),\
 	SR(DCHUBBUB_GLOBAL_TIMER_CNTL), \
 	SR(DCHUBBUB_TEST_DEBUG_INDEX), \
-	SR(DCHUBBUB_TEST_DEBUG_DATA)
+	SR(DCHUBBUB_TEST_DEBUG_DATA),\
+	SR(DCHUBBUB_SOFT_RESET)
 
 #define HUBBUB_SR_WATERMARK_REG_LIST()\
 	SR(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A),\
@@ -104,6 +106,7 @@ struct dcn_hubbub_registers {
 	uint32_t DCHUBBUB_SDPIF_AGP_BOT;
 	uint32_t DCHUBBUB_SDPIF_AGP_TOP;
 	uint32_t DCHUBBUB_CRC_CTRL;
+	uint32_t DCHUBBUB_SOFT_RESET;
 };
 
 /* set field name */
@@ -113,6 +116,7 @@ struct dcn_hubbub_registers {
 
 #define HUBBUB_MASK_SH_LIST_DCN(mask_sh)\
 		HUBBUB_SF(DCHUBBUB_GLOBAL_TIMER_CNTL, DCHUBBUB_GLOBAL_TIMER_ENABLE, mask_sh), \
+		HUBBUB_SF(DCHUBBUB_SOFT_RESET, DCHUBBUB_GLOBAL_SOFT_RESET, mask_sh), \
 		HUBBUB_SF(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL, DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, mask_sh), \
 		HUBBUB_SF(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL, DCHUBBUB_ARB_WATERMARK_CHANGE_DONE_INTERRUPT_DISABLE, mask_sh), \
 		HUBBUB_SF(DCHUBBUB_ARB_DRAM_STATE_CNTL, DCHUBBUB_ARB_ALLOW_SELF_REFRESH_FORCE_VALUE, mask_sh), \
@@ -142,6 +146,7 @@ struct dcn_hubbub_registers {
 		type DCHUBBUB_ARB_SAT_LEVEL;\
 		type DCHUBBUB_ARB_MIN_REQ_OUTSTAND;\
 		type DCHUBBUB_GLOBAL_TIMER_REFDIV;\
+		type DCHUBBUB_GLOBAL_SOFT_RESET; \
 		type SDPIF_FB_TOP;\
 		type SDPIF_FB_BASE;\
 		type SDPIF_FB_OFFSET;\
@@ -173,12 +178,6 @@ struct dcn_hubbub_wm {
 	struct dcn_hubbub_wm_set sets[4];
 };
 
-struct hubbub_funcs {
-	void (*update_dchub)(
-			struct hubbub *hubbub,
-			struct dchub_init_data *dh_data);
-};
-
 struct hubbub {
 	const struct hubbub_funcs *funcs;
 	struct dc_context *ctx;
@@ -186,6 +185,7 @@ struct hubbub {
 	const struct dcn_hubbub_shift *shifts;
 	const struct dcn_hubbub_mask *masks;
 	unsigned int debug_test_index_pstate;
+	struct dcn_watermark_set watermarks;
 };
 
 void hubbub1_update_dchub(
@@ -195,10 +195,13 @@ void hubbub1_update_dchub(
 bool hubbub1_verify_allow_pstate_change_high(
 	struct hubbub *hubbub);
 
+void hubbub1_wm_change_req_wa(struct hubbub *hubbub);
+
 void hubbub1_program_watermarks(
 		struct hubbub *hubbub,
 		struct dcn_watermark_set *watermarks,
-		unsigned int refclk_mhz);
+		unsigned int refclk_mhz,
+		bool safe_to_lower);
 
 void hubbub1_toggle_watermark_change_req(
 		struct hubbub *hubbub);
@@ -206,6 +209,7 @@ void hubbub1_toggle_watermark_change_req(
 void hubbub1_wm_read_state(struct hubbub *hubbub,
 		struct dcn_hubbub_wm *wm);
 
+void hubbub1_soft_reset(struct hubbub *hubbub, bool reset);
 void hubbub1_construct(struct hubbub *hubbub,
 	struct dc_context *ctx,
 	const struct dcn_hubbub_registers *hubbub_regs,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
index 39b72f696ae9..332354ca6529 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c
@@ -78,6 +78,27 @@ static void hubp1_disconnect(struct hubp *hubp)
 			CURSOR_ENABLE, 0);
 }
 
+static void hubp1_disable_control(struct hubp *hubp, bool disable_hubp)
+{
+	struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
+	uint32_t disable = disable_hubp ? 1 : 0;
+
+	REG_UPDATE(DCHUBP_CNTL,
+			HUBP_DISABLE, disable);
+}
+
+static unsigned int hubp1_get_underflow_status(struct hubp *hubp)
+{
+	uint32_t hubp_underflow = 0;
+	struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
+
+	REG_GET(DCHUBP_CNTL,
+		HUBP_UNDERFLOW_STATUS,
+		&hubp_underflow);
+
+	return hubp_underflow;
+}
+
 static void hubp1_set_hubp_blank_en(struct hubp *hubp, bool blank)
 {
 	struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
@@ -131,21 +152,22 @@ void hubp1_program_tiling(
 			PIPE_ALIGNED, info->gfx9.pipe_aligned);
 }
 
-void hubp1_program_size_and_rotation(
+void hubp1_program_size(
 	struct hubp *hubp,
-	enum dc_rotation_angle rotation,
 	enum surface_pixel_format format,
 	const union plane_size *plane_size,
-	struct dc_plane_dcc_param *dcc,
-	bool horizontal_mirror)
+	struct dc_plane_dcc_param *dcc)
 {
 	struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
-	uint32_t pitch, meta_pitch, pitch_c, meta_pitch_c, mirror;
+	uint32_t pitch, meta_pitch, pitch_c, meta_pitch_c;
 
 	/* Program data and meta surface pitch (calculation from addrlib)
 	 * 444 or 420 luma
 	 */
-	if (format >= SURFACE_PIXEL_FORMAT_VIDEO_BEGIN) {
+	if (format >= SURFACE_PIXEL_FORMAT_VIDEO_BEGIN && format < SURFACE_PIXEL_FORMAT_SUBSAMPLE_END) {
+		ASSERT(plane_size->video.chroma_pitch != 0);
+		/* Chroma pitch zero can cause system hang! */
+
 		pitch = plane_size->video.luma_pitch - 1;
 		meta_pitch = dcc->video.meta_pitch_l - 1;
 		pitch_c = plane_size->video.chroma_pitch - 1;
@@ -168,13 +190,22 @@ void hubp1_program_size_and_rotation(
 	if (format >= SURFACE_PIXEL_FORMAT_VIDEO_BEGIN)
 		REG_UPDATE_2(DCSURF_SURFACE_PITCH_C,
 			PITCH_C, pitch_c, META_PITCH_C, meta_pitch_c);
+}
+
+void hubp1_program_rotation(
+	struct hubp *hubp,
+	enum dc_rotation_angle rotation,
+	bool horizontal_mirror)
+{
+	struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
+	uint32_t mirror;
+
 
 	if (horizontal_mirror)
 		mirror = 1;
 	else
 		mirror = 0;
 
-
 	/* Program rotation angle and horz mirror - no mirror */
 	if (rotation == ROTATION_ANGLE_0)
 		REG_UPDATE_2(DCSURF_SURFACE_CONFIG,
@@ -372,11 +403,15 @@ bool hubp1_program_surface_flip_and_addr(
 		if (address->grph_stereo.right_addr.quad_part == 0)
 			break;
 
-		REG_UPDATE_4(DCSURF_SURFACE_CONTROL,
+		REG_UPDATE_8(DCSURF_SURFACE_CONTROL,
 				PRIMARY_SURFACE_TMZ, address->tmz_surface,
 				PRIMARY_SURFACE_TMZ_C, address->tmz_surface,
 				PRIMARY_META_SURFACE_TMZ, address->tmz_surface,
-				PRIMARY_META_SURFACE_TMZ_C, address->tmz_surface);
+				PRIMARY_META_SURFACE_TMZ_C, address->tmz_surface,
+				SECONDARY_SURFACE_TMZ, address->tmz_surface,
+				SECONDARY_SURFACE_TMZ_C, address->tmz_surface,
+				SECONDARY_META_SURFACE_TMZ, address->tmz_surface,
+				SECONDARY_META_SURFACE_TMZ_C, address->tmz_surface);
 
 		if (address->grph_stereo.right_meta_addr.quad_part != 0) {
 
@@ -422,9 +457,6 @@ bool hubp1_program_surface_flip_and_addr(
 
 	hubp->request_address = *address;
 
-	if (flip_immediate)
-		hubp->current_address = *address;
-
 	return true;
 }
 
@@ -435,9 +467,11 @@ void hubp1_dcc_control(struct hubp *hubp, bool enable,
 	uint32_t dcc_ind_64b_blk = independent_64b_blks ? 1 : 0;
 	struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
 
-	REG_UPDATE_2(DCSURF_SURFACE_CONTROL,
+	REG_UPDATE_4(DCSURF_SURFACE_CONTROL,
 			PRIMARY_SURFACE_DCC_EN, dcc_en,
-			PRIMARY_SURFACE_DCC_IND_64B_BLK, dcc_ind_64b_blk);
+			PRIMARY_SURFACE_DCC_IND_64B_BLK, dcc_ind_64b_blk,
+			SECONDARY_SURFACE_DCC_EN, dcc_en,
+			SECONDARY_SURFACE_DCC_IND_64B_BLK, dcc_ind_64b_blk);
 }
 
 void hubp1_program_surface_config(
@@ -451,8 +485,8 @@ void hubp1_program_surface_config(
 {
 	hubp1_dcc_control(hubp, dcc->enable, dcc->grph.independent_64b_blks);
 	hubp1_program_tiling(hubp, tiling_info, format);
-	hubp1_program_size_and_rotation(
-			hubp, rotation, format, plane_size, dcc, horizontal_mirror);
+	hubp1_program_size(hubp, format, plane_size, dcc);
+	hubp1_program_rotation(hubp, rotation, horizontal_mirror);
 	hubp1_program_pixel_format(hubp, format);
 }
 
@@ -535,11 +569,13 @@ void hubp1_program_deadline(
 	REG_SET(VBLANK_PARAMETERS_3, 0,
 		REFCYC_PER_META_CHUNK_VBLANK_L, dlg_attr->refcyc_per_meta_chunk_vblank_l);
 
-	REG_SET(NOM_PARAMETERS_0, 0,
-		DST_Y_PER_PTE_ROW_NOM_L, dlg_attr->dst_y_per_pte_row_nom_l);
+	if (REG(NOM_PARAMETERS_0))
+		REG_SET(NOM_PARAMETERS_0, 0,
+			DST_Y_PER_PTE_ROW_NOM_L, dlg_attr->dst_y_per_pte_row_nom_l);
 
-	REG_SET(NOM_PARAMETERS_1, 0,
-		REFCYC_PER_PTE_GROUP_NOM_L, dlg_attr->refcyc_per_pte_group_nom_l);
+	if (REG(NOM_PARAMETERS_1))
+		REG_SET(NOM_PARAMETERS_1, 0,
+			REFCYC_PER_PTE_GROUP_NOM_L, dlg_attr->refcyc_per_pte_group_nom_l);
 
 	REG_SET(NOM_PARAMETERS_4, 0,
 		DST_Y_PER_META_ROW_NOM_L, dlg_attr->dst_y_per_meta_row_nom_l);
@@ -568,11 +604,13 @@ void hubp1_program_deadline(
 	REG_SET(VBLANK_PARAMETERS_4, 0,
 		REFCYC_PER_META_CHUNK_VBLANK_C, dlg_attr->refcyc_per_meta_chunk_vblank_c);
 
-	REG_SET(NOM_PARAMETERS_2, 0,
-		DST_Y_PER_PTE_ROW_NOM_C, dlg_attr->dst_y_per_pte_row_nom_c);
+	if (REG(NOM_PARAMETERS_2))
+		REG_SET(NOM_PARAMETERS_2, 0,
+			DST_Y_PER_PTE_ROW_NOM_C, dlg_attr->dst_y_per_pte_row_nom_c);
 
-	REG_SET(NOM_PARAMETERS_3, 0,
-		REFCYC_PER_PTE_GROUP_NOM_C, dlg_attr->refcyc_per_pte_group_nom_c);
+	if (REG(NOM_PARAMETERS_3))
+		REG_SET(NOM_PARAMETERS_3, 0,
+			REFCYC_PER_PTE_GROUP_NOM_C, dlg_attr->refcyc_per_pte_group_nom_c);
 
 	REG_SET(NOM_PARAMETERS_6, 0,
 		DST_Y_PER_META_ROW_NOM_C, dlg_attr->dst_y_per_meta_row_nom_c);
@@ -609,6 +647,13 @@ void hubp1_program_deadline(
 	REG_SET(DCN_SURF1_TTU_CNTL1, 0,
 		REFCYC_PER_REQ_DELIVERY_PRE,
 		ttu_attr->refcyc_per_req_delivery_pre_c);
+
+	REG_SET_3(DCN_CUR0_TTU_CNTL0, 0,
+		REFCYC_PER_REQ_DELIVERY, ttu_attr->refcyc_per_req_delivery_cur0,
+		QoS_LEVEL_FIXED, ttu_attr->qos_level_fixed_cur0,
+		QoS_RAMP_DISABLE, ttu_attr->qos_ramp_disable_cur0);
+	REG_SET(DCN_CUR0_TTU_CNTL1, 0,
+		REFCYC_PER_REQ_DELIVERY_PRE, ttu_attr->refcyc_per_req_delivery_pre_cur0);
 }
 
 static void hubp1_setup(
@@ -647,7 +692,6 @@ bool hubp1_is_flip_pending(struct hubp *hubp)
 	if (earliest_inuse_address.grph.addr.quad_part != hubp->request_address.grph.addr.quad_part)
 		return true;
 
-	hubp->current_address = hubp->request_address;
 	return false;
 }
 
@@ -752,9 +796,159 @@ void min_set_viewport(
 		  PRI_VIEWPORT_Y_START_C, viewport_c->y);
 }
 
-void hubp1_read_state(struct dcn10_hubp *hubp1,
-		struct dcn_hubp_state *s)
+void hubp1_read_state(struct hubp *hubp)
 {
+	struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
+	struct dcn_hubp_state *s = &hubp1->state;
+	struct _vcs_dpi_display_dlg_regs_st *dlg_attr = &s->dlg_attr;
+	struct _vcs_dpi_display_ttu_regs_st *ttu_attr = &s->ttu_attr;
+	struct _vcs_dpi_display_rq_regs_st *rq_regs = &s->rq_regs;
+
+	/* Requester */
+	REG_GET(HUBPRET_CONTROL,
+			DET_BUF_PLANE1_BASE_ADDRESS, &rq_regs->plane1_base_address);
+	REG_GET_4(DCN_EXPANSION_MODE,
+			DRQ_EXPANSION_MODE, &rq_regs->drq_expansion_mode,
+			PRQ_EXPANSION_MODE, &rq_regs->prq_expansion_mode,
+			MRQ_EXPANSION_MODE, &rq_regs->mrq_expansion_mode,
+			CRQ_EXPANSION_MODE, &rq_regs->crq_expansion_mode);
+	REG_GET_8(DCHUBP_REQ_SIZE_CONFIG,
+		CHUNK_SIZE, &rq_regs->rq_regs_l.chunk_size,
+		MIN_CHUNK_SIZE, &rq_regs->rq_regs_l.min_chunk_size,
+		META_CHUNK_SIZE, &rq_regs->rq_regs_l.meta_chunk_size,
+		MIN_META_CHUNK_SIZE, &rq_regs->rq_regs_l.min_meta_chunk_size,
+		DPTE_GROUP_SIZE, &rq_regs->rq_regs_l.dpte_group_size,
+		MPTE_GROUP_SIZE, &rq_regs->rq_regs_l.mpte_group_size,
+		SWATH_HEIGHT, &rq_regs->rq_regs_l.swath_height,
+		PTE_ROW_HEIGHT_LINEAR, &rq_regs->rq_regs_l.pte_row_height_linear);
+	REG_GET_8(DCHUBP_REQ_SIZE_CONFIG_C,
+		CHUNK_SIZE_C, &rq_regs->rq_regs_c.chunk_size,
+		MIN_CHUNK_SIZE_C, &rq_regs->rq_regs_c.min_chunk_size,
+		META_CHUNK_SIZE_C, &rq_regs->rq_regs_c.meta_chunk_size,
+		MIN_META_CHUNK_SIZE_C, &rq_regs->rq_regs_c.min_meta_chunk_size,
+		DPTE_GROUP_SIZE_C, &rq_regs->rq_regs_c.dpte_group_size,
+		MPTE_GROUP_SIZE_C, &rq_regs->rq_regs_c.mpte_group_size,
+		SWATH_HEIGHT_C, &rq_regs->rq_regs_c.swath_height,
+		PTE_ROW_HEIGHT_LINEAR_C, &rq_regs->rq_regs_c.pte_row_height_linear);
+
+	/* DLG - Per hubp */
+	REG_GET_2(BLANK_OFFSET_0,
+		REFCYC_H_BLANK_END, &dlg_attr->refcyc_h_blank_end,
+		DLG_V_BLANK_END, &dlg_attr->dlg_vblank_end);
+
+	REG_GET(BLANK_OFFSET_1,
+		MIN_DST_Y_NEXT_START, &dlg_attr->min_dst_y_next_start);
+
+	REG_GET(DST_DIMENSIONS,
+		REFCYC_PER_HTOTAL, &dlg_attr->refcyc_per_htotal);
+
+	REG_GET_2(DST_AFTER_SCALER,
+		REFCYC_X_AFTER_SCALER, &dlg_attr->refcyc_x_after_scaler,
+		DST_Y_AFTER_SCALER, &dlg_attr->dst_y_after_scaler);
+
+	if (REG(PREFETCH_SETTINS))
+		REG_GET_2(PREFETCH_SETTINS,
+			DST_Y_PREFETCH, &dlg_attr->dst_y_prefetch,
+			VRATIO_PREFETCH, &dlg_attr->vratio_prefetch);
+	else
+		REG_GET_2(PREFETCH_SETTINGS,
+			DST_Y_PREFETCH, &dlg_attr->dst_y_prefetch,
+			VRATIO_PREFETCH, &dlg_attr->vratio_prefetch);
+
+	REG_GET_2(VBLANK_PARAMETERS_0,
+		DST_Y_PER_VM_VBLANK, &dlg_attr->dst_y_per_vm_vblank,
+		DST_Y_PER_ROW_VBLANK, &dlg_attr->dst_y_per_row_vblank);
+
+	REG_GET(REF_FREQ_TO_PIX_FREQ,
+		REF_FREQ_TO_PIX_FREQ, &dlg_attr->ref_freq_to_pix_freq);
+
+	/* DLG - Per luma/chroma */
+	REG_GET(VBLANK_PARAMETERS_1,
+		REFCYC_PER_PTE_GROUP_VBLANK_L, &dlg_attr->refcyc_per_pte_group_vblank_l);
+
+	REG_GET(VBLANK_PARAMETERS_3,
+		REFCYC_PER_META_CHUNK_VBLANK_L, &dlg_attr->refcyc_per_meta_chunk_vblank_l);
+
+	if (REG(NOM_PARAMETERS_0))
+		REG_GET(NOM_PARAMETERS_0,
+			DST_Y_PER_PTE_ROW_NOM_L, &dlg_attr->dst_y_per_pte_row_nom_l);
+
+	if (REG(NOM_PARAMETERS_1))
+		REG_GET(NOM_PARAMETERS_1,
+			REFCYC_PER_PTE_GROUP_NOM_L, &dlg_attr->refcyc_per_pte_group_nom_l);
+
+	REG_GET(NOM_PARAMETERS_4,
+		DST_Y_PER_META_ROW_NOM_L, &dlg_attr->dst_y_per_meta_row_nom_l);
+
+	REG_GET(NOM_PARAMETERS_5,
+		REFCYC_PER_META_CHUNK_NOM_L, &dlg_attr->refcyc_per_meta_chunk_nom_l);
+
+	REG_GET_2(PER_LINE_DELIVERY_PRE,
+		REFCYC_PER_LINE_DELIVERY_PRE_L, &dlg_attr->refcyc_per_line_delivery_pre_l,
+		REFCYC_PER_LINE_DELIVERY_PRE_C, &dlg_attr->refcyc_per_line_delivery_pre_c);
+
+	REG_GET_2(PER_LINE_DELIVERY,
+		REFCYC_PER_LINE_DELIVERY_L, &dlg_attr->refcyc_per_line_delivery_l,
+		REFCYC_PER_LINE_DELIVERY_C, &dlg_attr->refcyc_per_line_delivery_c);
+
+	if (REG(PREFETCH_SETTINS_C))
+		REG_GET(PREFETCH_SETTINS_C,
+			VRATIO_PREFETCH_C, &dlg_attr->vratio_prefetch_c);
+	else
+		REG_GET(PREFETCH_SETTINGS_C,
+			VRATIO_PREFETCH_C, &dlg_attr->vratio_prefetch_c);
+
+	REG_GET(VBLANK_PARAMETERS_2,
+		REFCYC_PER_PTE_GROUP_VBLANK_C, &dlg_attr->refcyc_per_pte_group_vblank_c);
+
+	REG_GET(VBLANK_PARAMETERS_4,
+		REFCYC_PER_META_CHUNK_VBLANK_C, &dlg_attr->refcyc_per_meta_chunk_vblank_c);
+
+	if (REG(NOM_PARAMETERS_2))
+		REG_GET(NOM_PARAMETERS_2,
+			DST_Y_PER_PTE_ROW_NOM_C, &dlg_attr->dst_y_per_pte_row_nom_c);
+
+	if (REG(NOM_PARAMETERS_3))
+		REG_GET(NOM_PARAMETERS_3,
+			REFCYC_PER_PTE_GROUP_NOM_C, &dlg_attr->refcyc_per_pte_group_nom_c);
+
+	REG_GET(NOM_PARAMETERS_6,
+		DST_Y_PER_META_ROW_NOM_C, &dlg_attr->dst_y_per_meta_row_nom_c);
+
+	REG_GET(NOM_PARAMETERS_7,
+		REFCYC_PER_META_CHUNK_NOM_C, &dlg_attr->refcyc_per_meta_chunk_nom_c);
+
+	/* TTU - per hubp */
+	REG_GET_2(DCN_TTU_QOS_WM,
+		QoS_LEVEL_LOW_WM, &ttu_attr->qos_level_low_wm,
+		QoS_LEVEL_HIGH_WM, &ttu_attr->qos_level_high_wm);
+
+	REG_GET_2(DCN_GLOBAL_TTU_CNTL,
+		MIN_TTU_VBLANK, &ttu_attr->min_ttu_vblank,
+		QoS_LEVEL_FLIP, &ttu_attr->qos_level_flip);
+
+	/* TTU - per luma/chroma */
+	/* Assumed surf0 is luma and 1 is chroma */
+
+	REG_GET_3(DCN_SURF0_TTU_CNTL0,
+		REFCYC_PER_REQ_DELIVERY, &ttu_attr->refcyc_per_req_delivery_l,
+		QoS_LEVEL_FIXED, &ttu_attr->qos_level_fixed_l,
+		QoS_RAMP_DISABLE, &ttu_attr->qos_ramp_disable_l);
+
+	REG_GET(DCN_SURF0_TTU_CNTL1,
+		REFCYC_PER_REQ_DELIVERY_PRE,
+		&ttu_attr->refcyc_per_req_delivery_pre_l);
+
+	REG_GET_3(DCN_SURF1_TTU_CNTL0,
+		REFCYC_PER_REQ_DELIVERY, &ttu_attr->refcyc_per_req_delivery_c,
+		QoS_LEVEL_FIXED, &ttu_attr->qos_level_fixed_c,
+		QoS_RAMP_DISABLE, &ttu_attr->qos_ramp_disable_c);
+
+	REG_GET(DCN_SURF1_TTU_CNTL1,
+		REFCYC_PER_REQ_DELIVERY_PRE,
+		&ttu_attr->refcyc_per_req_delivery_pre_c);
+
+	/* Rest of hubp */
 	REG_GET(DCSURF_SURFACE_CONFIG,
 			SURFACE_PIXEL_FORMAT, &s->pixel_format);
 
@@ -870,9 +1064,11 @@ void hubp1_cursor_set_position(
 		const struct dc_cursor_mi_param *param)
 {
 	struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
-	int src_x_offset = pos->x - pos->x_hotspot - param->viewport_x_start;
+	int src_x_offset = pos->x - pos->x_hotspot - param->viewport.x;
+	int x_hotspot = pos->x_hotspot;
+	int y_hotspot = pos->y_hotspot;
+	uint32_t dst_x_offset;
 	uint32_t cur_en = pos->enable ? 1 : 0;
-	uint32_t dst_x_offset = (src_x_offset >= 0) ? src_x_offset : 0;
 
 	/*
 	 * Guard aganst cursor_set_position() from being called with invalid
@@ -884,20 +1080,32 @@ void hubp1_cursor_set_position(
 	if (hubp->curs_attr.address.quad_part == 0)
 		return;
 
+	if (param->rotation == ROTATION_ANGLE_90 || param->rotation == ROTATION_ANGLE_270) {
+		src_x_offset = pos->y - pos->y_hotspot - param->viewport.x;
+		y_hotspot = pos->x_hotspot;
+		x_hotspot = pos->y_hotspot;
+	}
+
+	if (param->mirror) {
+		x_hotspot = param->viewport.width - x_hotspot;
+		src_x_offset = param->viewport.x + param->viewport.width - src_x_offset;
+	}
+
+	dst_x_offset = (src_x_offset >= 0) ? src_x_offset : 0;
 	dst_x_offset *= param->ref_clk_khz;
 	dst_x_offset /= param->pixel_clk_khz;
 
 	ASSERT(param->h_scale_ratio.value);
 
 	if (param->h_scale_ratio.value)
-		dst_x_offset = dal_fixed31_32_floor(dal_fixed31_32_div(
-				dal_fixed31_32_from_int(dst_x_offset),
+		dst_x_offset = dc_fixpt_floor(dc_fixpt_div(
+				dc_fixpt_from_int(dst_x_offset),
 				param->h_scale_ratio));
 
-	if (src_x_offset >= (int)param->viewport_width)
+	if (src_x_offset >= (int)param->viewport.width)
 		cur_en = 0;  /* not visible beyond right edge*/
 
-	if (src_x_offset + (int)hubp->curs_attr.width < 0)
+	if (src_x_offset + (int)hubp->curs_attr.width <= 0)
 		cur_en = 0;  /* not visible beyond left edge*/
 
 	if (cur_en && REG_READ(CURSOR_SURFACE_ADDRESS) == 0)
@@ -911,8 +1119,8 @@ void hubp1_cursor_set_position(
 			CURSOR_Y_POSITION, pos->y);
 
 	REG_SET_2(CURSOR_HOT_SPOT, 0,
-			CURSOR_HOT_SPOT_X, pos->x_hotspot,
-			CURSOR_HOT_SPOT_Y, pos->y_hotspot);
+			CURSOR_HOT_SPOT_X, x_hotspot,
+			CURSOR_HOT_SPOT_Y, y_hotspot);
 
 	REG_SET(CURSOR_DST_OFFSET, 0,
 			CURSOR_DST_X_OFFSET, dst_x_offset);
@@ -934,7 +1142,7 @@ void hubp1_vtg_sel(struct hubp *hubp, uint32_t otg_inst)
 	REG_UPDATE(DCHUBP_CNTL, HUBP_VTG_SEL, otg_inst);
 }
 
-static struct hubp_funcs dcn10_hubp_funcs = {
+static const struct hubp_funcs dcn10_hubp_funcs = {
 	.hubp_program_surface_flip_and_addr =
 			hubp1_program_surface_flip_and_addr,
 	.hubp_program_surface_config =
@@ -952,6 +1160,10 @@ static struct hubp_funcs dcn10_hubp_funcs = {
 	.hubp_disconnect = hubp1_disconnect,
 	.hubp_clk_cntl = hubp1_clk_cntl,
 	.hubp_vtg_sel = hubp1_vtg_sel,
+	.hubp_read_state = hubp1_read_state,
+	.hubp_disable_control =  hubp1_disable_control,
+	.hubp_get_underflow_status = hubp1_get_underflow_status,
+
 };
 
 /*****************************************/
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
index 4a3703e12ea1..f689feace82d 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h
@@ -30,6 +30,7 @@
 #define TO_DCN10_HUBP(hubp)\
 	container_of(hubp, struct dcn10_hubp, base)
 
+/* Register address initialization macro for all ASICs (including those with reduced functionality) */
 #define HUBP_REG_LIST_DCN(id)\
 	SRI(DCHUBP_CNTL, HUBP, id),\
 	SRI(HUBPREQ_DEBUG_DB, HUBP, id),\
@@ -78,16 +79,12 @@
 	SRI(REF_FREQ_TO_PIX_FREQ, HUBPREQ, id),\
 	SRI(VBLANK_PARAMETERS_1, HUBPREQ, id),\
 	SRI(VBLANK_PARAMETERS_3, HUBPREQ, id),\
-	SRI(NOM_PARAMETERS_0, HUBPREQ, id),\
-	SRI(NOM_PARAMETERS_1, HUBPREQ, id),\
 	SRI(NOM_PARAMETERS_4, HUBPREQ, id),\
 	SRI(NOM_PARAMETERS_5, HUBPREQ, id),\
 	SRI(PER_LINE_DELIVERY_PRE, HUBPREQ, id),\
 	SRI(PER_LINE_DELIVERY, HUBPREQ, id),\
 	SRI(VBLANK_PARAMETERS_2, HUBPREQ, id),\
 	SRI(VBLANK_PARAMETERS_4, HUBPREQ, id),\
-	SRI(NOM_PARAMETERS_2, HUBPREQ, id),\
-	SRI(NOM_PARAMETERS_3, HUBPREQ, id),\
 	SRI(NOM_PARAMETERS_6, HUBPREQ, id),\
 	SRI(NOM_PARAMETERS_7, HUBPREQ, id),\
 	SRI(DCN_TTU_QOS_WM, HUBPREQ, id),\
@@ -96,11 +93,21 @@
 	SRI(DCN_SURF0_TTU_CNTL1, HUBPREQ, id),\
 	SRI(DCN_SURF1_TTU_CNTL0, HUBPREQ, id),\
 	SRI(DCN_SURF1_TTU_CNTL1, HUBPREQ, id),\
-	SRI(DCN_VM_MX_L1_TLB_CNTL, HUBPREQ, id),\
+	SRI(DCN_CUR0_TTU_CNTL0, HUBPREQ, id),\
+	SRI(DCN_CUR0_TTU_CNTL1, HUBPREQ, id),\
 	SRI(HUBP_CLK_CNTL, HUBP, id)
 
+/* Register address initialization macro for ASICs with VM */
+#define HUBP_REG_LIST_DCN_VM(id)\
+	SRI(NOM_PARAMETERS_0, HUBPREQ, id),\
+	SRI(NOM_PARAMETERS_1, HUBPREQ, id),\
+	SRI(NOM_PARAMETERS_2, HUBPREQ, id),\
+	SRI(NOM_PARAMETERS_3, HUBPREQ, id),\
+	SRI(DCN_VM_MX_L1_TLB_CNTL, HUBPREQ, id)
+
 #define HUBP_REG_LIST_DCN10(id)\
 	HUBP_REG_LIST_DCN(id),\
+	HUBP_REG_LIST_DCN_VM(id),\
 	SRI(PREFETCH_SETTINS, HUBPREQ, id),\
 	SRI(PREFETCH_SETTINS_C, HUBPREQ, id),\
 	SRI(DCN_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_MSB, HUBPREQ, id),\
@@ -198,6 +205,8 @@
 	uint32_t DCN_SURF0_TTU_CNTL1; \
 	uint32_t DCN_SURF1_TTU_CNTL0; \
 	uint32_t DCN_SURF1_TTU_CNTL1; \
+	uint32_t DCN_CUR0_TTU_CNTL0; \
+	uint32_t DCN_CUR0_TTU_CNTL1; \
 	uint32_t DCN_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_MSB; \
 	uint32_t DCN_VM_CONTEXT0_PAGE_TABLE_BASE_ADDR_LSB; \
 	uint32_t DCN_VM_CONTEXT0_PAGE_TABLE_START_ADDR_MSB; \
@@ -237,12 +246,14 @@
 #define HUBP_SF(reg_name, field_name, post_fix)\
 	.field_name = reg_name ## __ ## field_name ## post_fix
 
+/* Mask/shift struct generation macro for all ASICs (including those with reduced functionality) */
 #define HUBP_MASK_SH_LIST_DCN(mask_sh)\
 	HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_BLANK_EN, mask_sh),\
 	HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_TTU_DISABLE, mask_sh),\
 	HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_UNDERFLOW_STATUS, mask_sh),\
 	HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_NO_OUTSTANDING_REQ, mask_sh),\
 	HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_VTG_SEL, mask_sh),\
+	HUBP_SF(HUBP0_DCHUBP_CNTL, HUBP_DISABLE, mask_sh),\
 	HUBP_SF(HUBP0_DCSURF_ADDR_CONFIG, NUM_PIPES, mask_sh),\
 	HUBP_SF(HUBP0_DCSURF_ADDR_CONFIG, NUM_BANKS, mask_sh),\
 	HUBP_SF(HUBP0_DCSURF_ADDR_CONFIG, PIPE_INTERLEAVE, mask_sh),\
@@ -257,8 +268,6 @@
 	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_PITCH, META_PITCH, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_PITCH_C, PITCH_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_PITCH_C, META_PITCH_C, mask_sh),\
-	HUBP_SF(HUBP0_DCSURF_SURFACE_CONFIG, ROTATION_ANGLE, mask_sh),\
-	HUBP_SF(HUBP0_DCSURF_SURFACE_CONFIG, H_MIRROR_EN, mask_sh),\
 	HUBP_SF(HUBP0_DCSURF_SURFACE_CONFIG, SURFACE_PIXEL_FORMAT, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_FLIP_CONTROL, SURFACE_FLIP_TYPE, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_FLIP_CONTROL, SURFACE_FLIP_PENDING, mask_sh),\
@@ -301,6 +310,12 @@
 	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, PRIMARY_META_SURFACE_TMZ_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, PRIMARY_SURFACE_DCC_EN, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, PRIMARY_SURFACE_DCC_IND_64B_BLK, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, SECONDARY_SURFACE_TMZ, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, SECONDARY_SURFACE_TMZ_C, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, SECONDARY_META_SURFACE_TMZ, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, SECONDARY_META_SURFACE_TMZ_C, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, SECONDARY_SURFACE_DCC_EN, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCSURF_SURFACE_CONTROL, SECONDARY_SURFACE_DCC_IND_64B_BLK, mask_sh),\
 	HUBP_SF(HUBPRET0_HUBPRET_CONTROL, DET_BUF_PLANE1_BASE_ADDRESS, mask_sh),\
 	HUBP_SF(HUBPRET0_HUBPRET_CONTROL, CROSSBAR_SRC_CB_B, mask_sh),\
 	HUBP_SF(HUBPRET0_HUBPRET_CONTROL, CROSSBAR_SRC_CR_R, mask_sh),\
@@ -335,8 +350,6 @@
 	HUBP_SF(HUBPREQ0_REF_FREQ_TO_PIX_FREQ, REF_FREQ_TO_PIX_FREQ, mask_sh),\
 	HUBP_SF(HUBPREQ0_VBLANK_PARAMETERS_1, REFCYC_PER_PTE_GROUP_VBLANK_L, mask_sh),\
 	HUBP_SF(HUBPREQ0_VBLANK_PARAMETERS_3, REFCYC_PER_META_CHUNK_VBLANK_L, mask_sh),\
-	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_0, DST_Y_PER_PTE_ROW_NOM_L, mask_sh),\
-	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_1, REFCYC_PER_PTE_GROUP_NOM_L, mask_sh),\
 	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_4, DST_Y_PER_META_ROW_NOM_L, mask_sh),\
 	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_5, REFCYC_PER_META_CHUNK_NOM_L, mask_sh),\
 	HUBP_SF(HUBPREQ0_PER_LINE_DELIVERY_PRE, REFCYC_PER_LINE_DELIVERY_PRE_L, mask_sh),\
@@ -345,8 +358,6 @@
 	HUBP_SF(HUBPREQ0_PER_LINE_DELIVERY, REFCYC_PER_LINE_DELIVERY_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_VBLANK_PARAMETERS_2, REFCYC_PER_PTE_GROUP_VBLANK_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_VBLANK_PARAMETERS_4, REFCYC_PER_META_CHUNK_VBLANK_C, mask_sh),\
-	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_2, DST_Y_PER_PTE_ROW_NOM_C, mask_sh),\
-	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_3, REFCYC_PER_PTE_GROUP_NOM_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_6, DST_Y_PER_META_ROW_NOM_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_7, REFCYC_PER_META_CHUNK_NOM_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCN_TTU_QOS_WM, QoS_LEVEL_LOW_WM, mask_sh),\
@@ -357,12 +368,26 @@
 	HUBP_SF(HUBPREQ0_DCN_SURF0_TTU_CNTL0, QoS_LEVEL_FIXED, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCN_SURF0_TTU_CNTL0, QoS_RAMP_DISABLE, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCN_SURF0_TTU_CNTL1, REFCYC_PER_REQ_DELIVERY_PRE, mask_sh),\
+	HUBP_SF(HUBP0_HUBP_CLK_CNTL, HUBP_CLOCK_ENABLE, mask_sh)
+
+/* Mask/shift struct generation macro for ASICs with VM */
+#define HUBP_MASK_SH_LIST_DCN_VM(mask_sh)\
+	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_0, DST_Y_PER_PTE_ROW_NOM_L, mask_sh),\
+	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_1, REFCYC_PER_PTE_GROUP_NOM_L, mask_sh),\
+	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_2, DST_Y_PER_PTE_ROW_NOM_C, mask_sh),\
+	HUBP_SF(HUBPREQ0_NOM_PARAMETERS_3, REFCYC_PER_PTE_GROUP_NOM_C, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCN_VM_MX_L1_TLB_CNTL, ENABLE_L1_TLB, mask_sh),\
 	HUBP_SF(HUBPREQ0_DCN_VM_MX_L1_TLB_CNTL, SYSTEM_ACCESS_MODE, mask_sh),\
-	HUBP_SF(HUBP0_HUBP_CLK_CNTL, HUBP_CLOCK_ENABLE, mask_sh)
+	HUBP_SF(HUBPREQ0_DCN_CUR0_TTU_CNTL0, REFCYC_PER_REQ_DELIVERY, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCN_CUR0_TTU_CNTL0, QoS_LEVEL_FIXED, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCN_CUR0_TTU_CNTL0, QoS_RAMP_DISABLE, mask_sh),\
+	HUBP_SF(HUBPREQ0_DCN_CUR0_TTU_CNTL1, REFCYC_PER_REQ_DELIVERY_PRE, mask_sh)
 
 #define HUBP_MASK_SH_LIST_DCN10(mask_sh)\
 	HUBP_MASK_SH_LIST_DCN(mask_sh),\
+	HUBP_MASK_SH_LIST_DCN_VM(mask_sh),\
+	HUBP_SF(HUBP0_DCSURF_SURFACE_CONFIG, ROTATION_ANGLE, mask_sh),\
+	HUBP_SF(HUBP0_DCSURF_SURFACE_CONFIG, H_MIRROR_EN, mask_sh),\
 	HUBP_SF(HUBPREQ0_PREFETCH_SETTINS, DST_Y_PREFETCH, mask_sh),\
 	HUBP_SF(HUBPREQ0_PREFETCH_SETTINS, VRATIO_PREFETCH, mask_sh),\
 	HUBP_SF(HUBPREQ0_PREFETCH_SETTINS_C, VRATIO_PREFETCH_C, mask_sh),\
@@ -403,6 +428,7 @@
 
 #define DCN_HUBP_REG_FIELD_LIST(type) \
 	type HUBP_BLANK_EN;\
+	type HUBP_DISABLE;\
 	type HUBP_TTU_DISABLE;\
 	type HUBP_NO_OUTSTANDING_REQ;\
 	type HUBP_VTG_SEL;\
@@ -469,6 +495,8 @@
 	type SECONDARY_META_SURFACE_TMZ_C;\
 	type PRIMARY_SURFACE_DCC_EN;\
 	type PRIMARY_SURFACE_DCC_IND_64B_BLK;\
+	type SECONDARY_SURFACE_DCC_EN;\
+	type SECONDARY_SURFACE_DCC_IND_64B_BLK;\
 	type DET_BUF_PLANE1_BASE_ADDRESS;\
 	type CROSSBAR_SRC_CB_B;\
 	type CROSSBAR_SRC_CR_R;\
@@ -601,8 +629,29 @@ struct dcn_mi_mask {
 	DCN_HUBP_REG_FIELD_LIST(uint32_t);
 };
 
+struct dcn_hubp_state {
+	struct _vcs_dpi_display_dlg_regs_st dlg_attr;
+	struct _vcs_dpi_display_ttu_regs_st ttu_attr;
+	struct _vcs_dpi_display_rq_regs_st rq_regs;
+	uint32_t pixel_format;
+	uint32_t inuse_addr_hi;
+	uint32_t viewport_width;
+	uint32_t viewport_height;
+	uint32_t rotation_angle;
+	uint32_t h_mirror_en;
+	uint32_t sw_mode;
+	uint32_t dcc_en;
+	uint32_t blank_en;
+	uint32_t underflow_status;
+	uint32_t ttu_disable;
+	uint32_t min_ttu_vblank;
+	uint32_t qos_level_low_wm;
+	uint32_t qos_level_high_wm;
+};
+
 struct dcn10_hubp {
 	struct hubp base;
+	struct dcn_hubp_state state;
 	const struct dcn_mi_registers *hubp_regs;
 	const struct dcn_mi_shift *hubp_shift;
 	const struct dcn_mi_mask *hubp_mask;
@@ -630,12 +679,15 @@ void hubp1_program_pixel_format(
 	struct hubp *hubp,
 	enum surface_pixel_format format);
 
-void hubp1_program_size_and_rotation(
+void hubp1_program_size(
 	struct hubp *hubp,
-	enum dc_rotation_angle rotation,
 	enum surface_pixel_format format,
 	const union plane_size *plane_size,
-	struct dc_plane_dcc_param *dcc,
+	struct dc_plane_dcc_param *dcc);
+
+void hubp1_program_rotation(
+	struct hubp *hubp,
+	enum dc_rotation_angle rotation,
 	bool horizontal_mirror);
 
 void hubp1_program_tiling(
@@ -680,26 +732,9 @@ void dcn10_hubp_construct(
 	const struct dcn_mi_shift *hubp_shift,
 	const struct dcn_mi_mask *hubp_mask);
 
-
-struct dcn_hubp_state {
-	uint32_t pixel_format;
-	uint32_t inuse_addr_hi;
-	uint32_t viewport_width;
-	uint32_t viewport_height;
-	uint32_t rotation_angle;
-	uint32_t h_mirror_en;
-	uint32_t sw_mode;
-	uint32_t dcc_en;
-	uint32_t blank_en;
-	uint32_t underflow_status;
-	uint32_t ttu_disable;
-	uint32_t min_ttu_vblank;
-	uint32_t qos_level_low_wm;
-	uint32_t qos_level_high_wm;
-};
-void hubp1_read_state(struct dcn10_hubp *hubp1,
-		struct dcn_hubp_state *s);
+void hubp1_read_state(struct hubp *hubp);
 
 enum cursor_pitch hubp1_get_cursor_pitch(unsigned int pitch);
 
+
 #endif
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 8b0f6b8a5627..c87f6e603055 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -45,8 +45,8 @@
 #include "dcn10_hubbub.h"
 #include "dcn10_cm_common.h"
 
-#define DC_LOGGER \
-	ctx->logger
+#define DC_LOGGER_INIT(logger)
+
 #define CTX \
 	hws->ctx
 #define REG(reg)\
@@ -56,16 +56,17 @@
 #define FN(reg_name, field_name) \
 	hws->shifts->field_name, hws->masks->field_name
 
+/*print is 17 wide, first two characters are spaces*/
 #define DTN_INFO_MICRO_SEC(ref_cycle) \
 	print_microsec(dc_ctx, ref_cycle)
 
 void print_microsec(struct dc_context *dc_ctx, uint32_t ref_cycle)
 {
-	static const uint32_t ref_clk_mhz = 48;
-	static const unsigned int frac = 10;
+	const uint32_t ref_clk_mhz = dc_ctx->dc->res_pool->ref_clock_inKhz / 1000;
+	static const unsigned int frac = 1000;
 	uint32_t us_x10 = (ref_cycle * frac) / ref_clk_mhz;
 
-	DTN_INFO("%d.%d \t ",
+	DTN_INFO("  %11d.%03d",
 			us_x10 / frac,
 			us_x10 % frac);
 }
@@ -92,14 +93,14 @@ void dcn10_log_hubbub_state(struct dc *dc)
 
 	hubbub1_wm_read_state(dc->res_pool->hubbub, &wm);
 
-	DTN_INFO("HUBBUB WM: \t data_urgent \t pte_meta_urgent \t "
-			"sr_enter \t sr_exit \t dram_clk_change \n");
+	DTN_INFO("HUBBUB WM:      data_urgent  pte_meta_urgent"
+			"         sr_enter          sr_exit  dram_clk_change\n");
 
 	for (i = 0; i < 4; i++) {
 		struct dcn_hubbub_wm_set *s;
 
 		s = &wm.sets[i];
-		DTN_INFO("WM_Set[%d]:\t ", s->wm_set);
+		DTN_INFO("WM_Set[%d]:", s->wm_set);
 		DTN_INFO_MICRO_SEC(s->data_urgent);
 		DTN_INFO_MICRO_SEC(s->pte_meta_urgent);
 		DTN_INFO_MICRO_SEC(s->sr_enter);
@@ -111,6 +112,121 @@ void dcn10_log_hubbub_state(struct dc *dc)
 	DTN_INFO("\n");
 }
 
+static void dcn10_log_hubp_states(struct dc *dc)
+{
+	struct dc_context *dc_ctx = dc->ctx;
+	struct resource_pool *pool = dc->res_pool;
+	int i;
+
+	DTN_INFO("HUBP:  format  addr_hi  width  height"
+			"  rot  mir  sw_mode  dcc_en  blank_en  ttu_dis  underflow"
+			"   min_ttu_vblank       qos_low_wm      qos_high_wm\n");
+	for (i = 0; i < pool->pipe_count; i++) {
+		struct hubp *hubp = pool->hubps[i];
+		struct dcn_hubp_state *s = &(TO_DCN10_HUBP(hubp)->state);
+
+		hubp->funcs->hubp_read_state(hubp);
+
+		if (!s->blank_en) {
+			DTN_INFO("[%2d]:  %5xh  %6xh  %5d  %6d  %2xh  %2xh  %6xh"
+					"  %6d  %8d  %7d  %8xh",
+					hubp->inst,
+					s->pixel_format,
+					s->inuse_addr_hi,
+					s->viewport_width,
+					s->viewport_height,
+					s->rotation_angle,
+					s->h_mirror_en,
+					s->sw_mode,
+					s->dcc_en,
+					s->blank_en,
+					s->ttu_disable,
+					s->underflow_status);
+			DTN_INFO_MICRO_SEC(s->min_ttu_vblank);
+			DTN_INFO_MICRO_SEC(s->qos_level_low_wm);
+			DTN_INFO_MICRO_SEC(s->qos_level_high_wm);
+			DTN_INFO("\n");
+		}
+	}
+
+	DTN_INFO("\n=========RQ========\n");
+	DTN_INFO("HUBP:  drq_exp_m  prq_exp_m  mrq_exp_m  crq_exp_m  plane1_ba  L:chunk_s  min_chu_s  meta_ch_s"
+		"  min_m_c_s  dpte_gr_s  mpte_gr_s  swath_hei  pte_row_h  C:chunk_s  min_chu_s  meta_ch_s"
+		"  min_m_c_s  dpte_gr_s  mpte_gr_s  swath_hei  pte_row_h\n");
+	for (i = 0; i < pool->pipe_count; i++) {
+		struct dcn_hubp_state *s = &(TO_DCN10_HUBP(pool->hubps[i])->state);
+		struct _vcs_dpi_display_rq_regs_st *rq_regs = &s->rq_regs;
+
+		if (!s->blank_en)
+			DTN_INFO("[%2d]:  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh\n",
+				pool->hubps[i]->inst, rq_regs->drq_expansion_mode, rq_regs->prq_expansion_mode, rq_regs->mrq_expansion_mode,
+				rq_regs->crq_expansion_mode, rq_regs->plane1_base_address, rq_regs->rq_regs_l.chunk_size,
+				rq_regs->rq_regs_l.min_chunk_size, rq_regs->rq_regs_l.meta_chunk_size,
+				rq_regs->rq_regs_l.min_meta_chunk_size, rq_regs->rq_regs_l.dpte_group_size,
+				rq_regs->rq_regs_l.mpte_group_size, rq_regs->rq_regs_l.swath_height,
+				rq_regs->rq_regs_l.pte_row_height_linear, rq_regs->rq_regs_c.chunk_size, rq_regs->rq_regs_c.min_chunk_size,
+				rq_regs->rq_regs_c.meta_chunk_size, rq_regs->rq_regs_c.min_meta_chunk_size,
+				rq_regs->rq_regs_c.dpte_group_size, rq_regs->rq_regs_c.mpte_group_size,
+				rq_regs->rq_regs_c.swath_height, rq_regs->rq_regs_c.pte_row_height_linear);
+	}
+
+	DTN_INFO("========DLG========\n");
+	DTN_INFO("HUBP:  rc_hbe     dlg_vbe    min_d_y_n  rc_per_ht  rc_x_a_s "
+			"  dst_y_a_s  dst_y_pf   dst_y_vvb  dst_y_rvb  dst_y_vfl  dst_y_rfl  rf_pix_fq"
+			"  vratio_pf  vrat_pf_c  rc_pg_vbl  rc_pg_vbc  rc_mc_vbl  rc_mc_vbc  rc_pg_fll"
+			"  rc_pg_flc  rc_mc_fll  rc_mc_flc  pr_nom_l   pr_nom_c   rc_pg_nl   rc_pg_nc "
+			"  mr_nom_l   mr_nom_c   rc_mc_nl   rc_mc_nc   rc_ld_pl   rc_ld_pc   rc_ld_l  "
+			"  rc_ld_c    cha_cur0   ofst_cur1  cha_cur1   vr_af_vc0  ddrq_limt  x_rt_dlay"
+			"  x_rp_dlay  x_rr_sfl\n");
+	for (i = 0; i < pool->pipe_count; i++) {
+		struct dcn_hubp_state *s = &(TO_DCN10_HUBP(pool->hubps[i])->state);
+		struct _vcs_dpi_display_dlg_regs_st *dlg_regs = &s->dlg_attr;
+
+		if (!s->blank_en)
+			DTN_INFO("[%2d]:  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh"
+				"%  8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh"
+				"  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh\n",
+				pool->hubps[i]->inst, dlg_regs->refcyc_h_blank_end, dlg_regs->dlg_vblank_end, dlg_regs->min_dst_y_next_start,
+				dlg_regs->refcyc_per_htotal, dlg_regs->refcyc_x_after_scaler, dlg_regs->dst_y_after_scaler,
+				dlg_regs->dst_y_prefetch, dlg_regs->dst_y_per_vm_vblank, dlg_regs->dst_y_per_row_vblank,
+				dlg_regs->dst_y_per_vm_flip, dlg_regs->dst_y_per_row_flip, dlg_regs->ref_freq_to_pix_freq,
+				dlg_regs->vratio_prefetch, dlg_regs->vratio_prefetch_c, dlg_regs->refcyc_per_pte_group_vblank_l,
+				dlg_regs->refcyc_per_pte_group_vblank_c, dlg_regs->refcyc_per_meta_chunk_vblank_l,
+				dlg_regs->refcyc_per_meta_chunk_vblank_c, dlg_regs->refcyc_per_pte_group_flip_l,
+				dlg_regs->refcyc_per_pte_group_flip_c, dlg_regs->refcyc_per_meta_chunk_flip_l,
+				dlg_regs->refcyc_per_meta_chunk_flip_c, dlg_regs->dst_y_per_pte_row_nom_l,
+				dlg_regs->dst_y_per_pte_row_nom_c, dlg_regs->refcyc_per_pte_group_nom_l,
+				dlg_regs->refcyc_per_pte_group_nom_c, dlg_regs->dst_y_per_meta_row_nom_l,
+				dlg_regs->dst_y_per_meta_row_nom_c, dlg_regs->refcyc_per_meta_chunk_nom_l,
+				dlg_regs->refcyc_per_meta_chunk_nom_c, dlg_regs->refcyc_per_line_delivery_pre_l,
+				dlg_regs->refcyc_per_line_delivery_pre_c, dlg_regs->refcyc_per_line_delivery_l,
+				dlg_regs->refcyc_per_line_delivery_c, dlg_regs->chunk_hdl_adjust_cur0, dlg_regs->dst_y_offset_cur1,
+				dlg_regs->chunk_hdl_adjust_cur1, dlg_regs->vready_after_vcount0, dlg_regs->dst_y_delta_drq_limit,
+				dlg_regs->xfc_reg_transfer_delay, dlg_regs->xfc_reg_precharge_delay,
+				dlg_regs->xfc_reg_remote_surface_flip_latency);
+	}
+
+	DTN_INFO("========TTU========\n");
+	DTN_INFO("HUBP:  qos_ll_wm  qos_lh_wm  mn_ttu_vb  qos_l_flp  rc_rd_p_l  rc_rd_l    rc_rd_p_c"
+			"  rc_rd_c    rc_rd_c0   rc_rd_pc0  rc_rd_c1   rc_rd_pc1  qos_lf_l   qos_rds_l"
+			"  qos_lf_c   qos_rds_c  qos_lf_c0  qos_rds_c0 qos_lf_c1  qos_rds_c1\n");
+	for (i = 0; i < pool->pipe_count; i++) {
+		struct dcn_hubp_state *s = &(TO_DCN10_HUBP(pool->hubps[i])->state);
+		struct _vcs_dpi_display_ttu_regs_st *ttu_regs = &s->ttu_attr;
+
+		if (!s->blank_en)
+			DTN_INFO("[%2d]:  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh  %8xh\n",
+				pool->hubps[i]->inst, ttu_regs->qos_level_low_wm, ttu_regs->qos_level_high_wm, ttu_regs->min_ttu_vblank,
+				ttu_regs->qos_level_flip, ttu_regs->refcyc_per_req_delivery_pre_l, ttu_regs->refcyc_per_req_delivery_l,
+				ttu_regs->refcyc_per_req_delivery_pre_c, ttu_regs->refcyc_per_req_delivery_c, ttu_regs->refcyc_per_req_delivery_cur0,
+				ttu_regs->refcyc_per_req_delivery_pre_cur0, ttu_regs->refcyc_per_req_delivery_cur1,
+				ttu_regs->refcyc_per_req_delivery_pre_cur1, ttu_regs->qos_level_fixed_l, ttu_regs->qos_ramp_disable_l,
+				ttu_regs->qos_level_fixed_c, ttu_regs->qos_ramp_disable_c, ttu_regs->qos_level_fixed_cur0,
+				ttu_regs->qos_ramp_disable_cur0, ttu_regs->qos_level_fixed_cur1, ttu_regs->qos_ramp_disable_cur1);
+	}
+	DTN_INFO("\n");
+}
+
 void dcn10_log_hw_state(struct dc *dc)
 {
 	struct dc_context *dc_ctx = dc->ctx;
@@ -121,41 +237,64 @@ void dcn10_log_hw_state(struct dc *dc)
 
 	dcn10_log_hubbub_state(dc);
 
-	DTN_INFO("HUBP:\t format \t addr_hi \t width \t height \t "
-			"rotation \t mirror \t  sw_mode \t "
-			"dcc_en \t blank_en \t ttu_dis \t underflow \t "
-			"min_ttu_vblank \t qos_low_wm \t qos_high_wm \n");
+	dcn10_log_hubp_states(dc);
 
+	DTN_INFO("DPP:    IGAM format  IGAM mode    DGAM mode    RGAM mode"
+			"  GAMUT mode  C11 C12   C13 C14   C21 C22   C23 C24   "
+			"C31 C32   C33 C34\n");
 	for (i = 0; i < pool->pipe_count; i++) {
-		struct hubp *hubp = pool->hubps[i];
-		struct dcn_hubp_state s;
-
-		hubp1_read_state(TO_DCN10_HUBP(hubp), &s);
-
-		DTN_INFO("[%d]:\t %xh \t %xh \t %d \t %d \t "
-				"%xh \t %xh \t %xh \t "
-				"%d \t %d \t %d \t %xh \t",
-				hubp->inst,
-				s.pixel_format,
-				s.inuse_addr_hi,
-				s.viewport_width,
-				s.viewport_height,
-				s.rotation_angle,
-				s.h_mirror_en,
-				s.sw_mode,
-				s.dcc_en,
-				s.blank_en,
-				s.ttu_disable,
-				s.underflow_status);
-		DTN_INFO_MICRO_SEC(s.min_ttu_vblank);
-		DTN_INFO_MICRO_SEC(s.qos_level_low_wm);
-		DTN_INFO_MICRO_SEC(s.qos_level_high_wm);
+		struct dpp *dpp = pool->dpps[i];
+		struct dcn_dpp_state s;
+
+		dpp->funcs->dpp_read_state(dpp, &s);
+
+		DTN_INFO("[%2d]:  %11xh  %-11s  %-11s  %-11s"
+				"%8x    %08xh %08xh %08xh %08xh %08xh %08xh",
+				dpp->inst,
+				s.igam_input_format,
+				(s.igam_lut_mode == 0) ? "BypassFixed" :
+					((s.igam_lut_mode == 1) ? "BypassFloat" :
+					((s.igam_lut_mode == 2) ? "RAM" :
+					((s.igam_lut_mode == 3) ? "RAM" :
+								 "Unknown"))),
+				(s.dgam_lut_mode == 0) ? "Bypass" :
+					((s.dgam_lut_mode == 1) ? "sRGB" :
+					((s.dgam_lut_mode == 2) ? "Ycc" :
+					((s.dgam_lut_mode == 3) ? "RAM" :
+					((s.dgam_lut_mode == 4) ? "RAM" :
+								 "Unknown")))),
+				(s.rgam_lut_mode == 0) ? "Bypass" :
+					((s.rgam_lut_mode == 1) ? "sRGB" :
+					((s.rgam_lut_mode == 2) ? "Ycc" :
+					((s.rgam_lut_mode == 3) ? "RAM" :
+					((s.rgam_lut_mode == 4) ? "RAM" :
+								 "Unknown")))),
+				s.gamut_remap_mode,
+				s.gamut_remap_c11_c12,
+				s.gamut_remap_c13_c14,
+				s.gamut_remap_c21_c22,
+				s.gamut_remap_c23_c24,
+				s.gamut_remap_c31_c32,
+				s.gamut_remap_c33_c34);
 		DTN_INFO("\n");
 	}
 	DTN_INFO("\n");
 
-	DTN_INFO("OTG:\t v_bs \t v_be \t v_ss \t v_se \t vpol \t vmax \t vmin \t "
-			"h_bs \t h_be \t h_ss \t h_se \t hpol \t htot \t vtot \t underflow\n");
+	DTN_INFO("MPCC:  OPP  DPP  MPCCBOT  MODE  ALPHA_MODE  PREMULT  OVERLAP_ONLY  IDLE\n");
+	for (i = 0; i < pool->pipe_count; i++) {
+		struct mpcc_state s = {0};
+
+		pool->mpc->funcs->read_mpcc_state(pool->mpc, i, &s);
+		if (s.opp_id != 0xf)
+			DTN_INFO("[%2d]:  %2xh  %2xh  %6xh  %4d  %10d  %7d  %12d  %4d\n",
+				i, s.opp_id, s.dpp_id, s.bot_mpcc_id,
+				s.mode, s.alpha_mode, s.pre_multiplied_alpha, s.overlap_only,
+				s.idle);
+	}
+	DTN_INFO("\n");
+
+	DTN_INFO("OTG:  v_bs  v_be  v_ss  v_se  vpol  vmax  vmin  vmax_sel  vmin_sel"
+			"  h_bs  h_be  h_ss  h_se  hpol  htot  vtot  underflow\n");
 
 	for (i = 0; i < pool->timing_generator_count; i++) {
 		struct timing_generator *tg = pool->timing_generators[i];
@@ -167,9 +306,8 @@ void dcn10_log_hw_state(struct dc *dc)
 		if ((s.otg_enabled & 1) == 0)
 			continue;
 
-		DTN_INFO("[%d]:\t %d \t %d \t %d \t %d \t "
-				"%d \t %d \t %d \t %d \t %d \t %d \t "
-				"%d \t %d \t %d \t %d \t %d \t ",
+		DTN_INFO("[%d]: %5d %5d %5d %5d %5d %5d %5d %9d %9d %5d %5d %5d"
+				" %5d %5d %5d %5d  %9d\n",
 				tg->inst,
 				s.v_blank_start,
 				s.v_blank_end,
@@ -178,6 +316,8 @@ void dcn10_log_hw_state(struct dc *dc)
 				s.v_sync_a_pol,
 				s.v_total_max,
 				s.v_total_min,
+				s.v_total_max_sel,
+				s.v_total_min_sel,
 				s.h_blank_start,
 				s.h_blank_end,
 				s.h_sync_a_start,
@@ -186,10 +326,25 @@ void dcn10_log_hw_state(struct dc *dc)
 				s.h_total,
 				s.v_total,
 				s.underflow_occurred_status);
-		DTN_INFO("\n");
+
+		// Clear underflow for debug purposes
+		// We want to keep underflow sticky bit on for the longevity tests outside of test environment.
+		// This function is called only from Windows or Diags test environment, hence it's safe to clear
+		// it from here without affecting the original intent.
+		tg->funcs->clear_optc_underflow(tg);
 	}
 	DTN_INFO("\n");
 
+	DTN_INFO("\nCALCULATED Clocks: dcfclk_khz:%d  dcfclk_deep_sleep_khz:%d  dispclk_khz:%d\n"
+		"dppclk_khz:%d  max_supported_dppclk_khz:%d  fclk_khz:%d  socclk_khz:%d\n\n",
+			dc->current_state->bw.dcn.clk.dcfclk_khz,
+			dc->current_state->bw.dcn.clk.dcfclk_deep_sleep_khz,
+			dc->current_state->bw.dcn.clk.dispclk_khz,
+			dc->current_state->bw.dcn.clk.dppclk_khz,
+			dc->current_state->bw.dcn.clk.max_supported_dppclk_khz,
+			dc->current_state->bw.dcn.clk.fclk_khz,
+			dc->current_state->bw.dcn.clk.socclk_khz);
+
 	log_mpc_crc(dc);
 
 	DTN_INFO_END();
@@ -260,6 +415,8 @@ static void dpp_pg_control(
 
 	if (hws->ctx->dc->debug.disable_dpp_power_gate)
 		return;
+	if (REG(DOMAIN1_PG_CONFIG) == 0)
+		return;
 
 	switch (dpp_inst) {
 	case 0: /* DPP0 */
@@ -310,6 +467,8 @@ static void hubp_pg_control(
 
 	if (hws->ctx->dc->debug.disable_hubp_power_gate)
 		return;
+	if (REG(DOMAIN0_PG_CONFIG) == 0)
+		return;
 
 	switch (hubp_inst) {
 	case 0: /* DCHUBP0 */
@@ -354,7 +513,7 @@ static void power_on_plane(
 	struct dce_hwseq *hws,
 	int plane_id)
 {
-	struct dc_context *ctx = hws->ctx;
+	DC_LOGGER_INIT(hws->ctx->logger);
 	if (REG(DC_IP_REQUEST_CNTL)) {
 		REG_SET(DC_IP_REQUEST_CNTL, 0,
 				IP_REQUEST_EN, 1);
@@ -461,7 +620,7 @@ static void false_optc_underflow_wa(
 		tg->funcs->clear_optc_underflow(tg);
 }
 
-static enum dc_status dcn10_prog_pixclk_crtc_otg(
+static enum dc_status dcn10_enable_stream_timing(
 		struct pipe_ctx *pipe_ctx,
 		struct dc_state *context,
 		struct dc *dc)
@@ -553,7 +712,7 @@ static void reset_back_end_for_pipe(
 		struct dc_state *context)
 {
 	int i;
-	struct dc_context *ctx = dc->ctx;
+	DC_LOGGER_INIT(dc->ctx->logger);
 	if (pipe_ctx->stream_res.stream_enc == NULL) {
 		pipe_ctx->stream = NULL;
 		return;
@@ -564,19 +723,7 @@ static void reset_back_end_for_pipe(
 		if (!pipe_ctx->stream->dpms_off)
 			core_link_disable_stream(pipe_ctx, FREE_ACQUIRED_RESOURCE);
 		else if (pipe_ctx->stream_res.audio) {
-			/*
-			 * if stream is already disabled outside of commit streams path,
-			 * audio disable was skipped. Need to do it here
-			 */
-			pipe_ctx->stream_res.audio->funcs->az_disable(pipe_ctx->stream_res.audio);
-
-			if (dc->caps.dynamic_audio == true) {
-				/*we have to dynamic arbitrate the audio endpoints*/
-				pipe_ctx->stream_res.audio = NULL;
-				/*we free the resource, need reset is_audio_acquired*/
-				update_audio_usage(&dc->current_state->res_ctx, dc->res_pool, pipe_ctx->stream_res.audio, false);
-			}
-
+			dc->hwss.disable_audio_stream(pipe_ctx, FREE_ACQUIRED_RESOURCE);
 		}
 
 	}
@@ -603,7 +750,91 @@ static void reset_back_end_for_pipe(
 					pipe_ctx->pipe_idx, pipe_ctx->stream_res.tg->inst);
 }
 
-static void dcn10_verify_allow_pstate_change_high(struct dc *dc)
+static bool dcn10_hw_wa_force_recovery(struct dc *dc)
+{
+	struct hubp *hubp ;
+	unsigned int i;
+	bool need_recover = true;
+
+	if (!dc->debug.recovery_enabled)
+		return false;
+
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx =
+			&dc->current_state->res_ctx.pipe_ctx[i];
+		if (pipe_ctx != NULL) {
+			hubp = pipe_ctx->plane_res.hubp;
+			if (hubp != NULL) {
+				if (hubp->funcs->hubp_get_underflow_status(hubp) != 0) {
+					/* one pipe underflow, we will reset all the pipes*/
+					need_recover = true;
+				}
+			}
+		}
+	}
+	if (!need_recover)
+		return false;
+	/*
+	DCHUBP_CNTL:HUBP_BLANK_EN=1
+	DCHUBBUB_SOFT_RESET:DCHUBBUB_GLOBAL_SOFT_RESET=1
+	DCHUBP_CNTL:HUBP_DISABLE=1
+	DCHUBP_CNTL:HUBP_DISABLE=0
+	DCHUBBUB_SOFT_RESET:DCHUBBUB_GLOBAL_SOFT_RESET=0
+	DCSURF_PRIMARY_SURFACE_ADDRESS
+	DCHUBP_CNTL:HUBP_BLANK_EN=0
+	*/
+
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx =
+			&dc->current_state->res_ctx.pipe_ctx[i];
+		if (pipe_ctx != NULL) {
+			hubp = pipe_ctx->plane_res.hubp;
+			/*DCHUBP_CNTL:HUBP_BLANK_EN=1*/
+			if (hubp != NULL)
+				hubp->funcs->set_hubp_blank_en(hubp, true);
+		}
+	}
+	/*DCHUBBUB_SOFT_RESET:DCHUBBUB_GLOBAL_SOFT_RESET=1*/
+	hubbub1_soft_reset(dc->res_pool->hubbub, true);
+
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx =
+			&dc->current_state->res_ctx.pipe_ctx[i];
+		if (pipe_ctx != NULL) {
+			hubp = pipe_ctx->plane_res.hubp;
+			/*DCHUBP_CNTL:HUBP_DISABLE=1*/
+			if (hubp != NULL)
+				hubp->funcs->hubp_disable_control(hubp, true);
+		}
+	}
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx =
+			&dc->current_state->res_ctx.pipe_ctx[i];
+		if (pipe_ctx != NULL) {
+			hubp = pipe_ctx->plane_res.hubp;
+			/*DCHUBP_CNTL:HUBP_DISABLE=0*/
+			if (hubp != NULL)
+				hubp->funcs->hubp_disable_control(hubp, true);
+		}
+	}
+	/*DCHUBBUB_SOFT_RESET:DCHUBBUB_GLOBAL_SOFT_RESET=0*/
+	hubbub1_soft_reset(dc->res_pool->hubbub, false);
+	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+		struct pipe_ctx *pipe_ctx =
+			&dc->current_state->res_ctx.pipe_ctx[i];
+		if (pipe_ctx != NULL) {
+			hubp = pipe_ctx->plane_res.hubp;
+			/*DCHUBP_CNTL:HUBP_BLANK_EN=0*/
+			if (hubp != NULL)
+				hubp->funcs->set_hubp_blank_en(hubp, true);
+		}
+	}
+	return true;
+
+}
+
+
+void dcn10_verify_allow_pstate_change_high(struct dc *dc)
 {
 	static bool should_log_hw_state; /* prevent hw state log by default */
 
@@ -611,13 +842,17 @@ static void dcn10_verify_allow_pstate_change_high(struct dc *dc)
 		if (should_log_hw_state) {
 			dcn10_log_hw_state(dc);
 		}
-
 		BREAK_TO_DEBUGGER();
+		if (dcn10_hw_wa_force_recovery(dc)) {
+		/*check again*/
+			if (!hubbub1_verify_allow_pstate_change_high(dc->res_pool->hubbub))
+				BREAK_TO_DEBUGGER();
+		}
 	}
 }
 
 /* trigger HW to start disconnect plane from stream on the next vsync */
-static void plane_atomic_disconnect(struct dc *dc, struct pipe_ctx *pipe_ctx)
+void hwss1_plane_atomic_disconnect(struct dc *dc, struct pipe_ctx *pipe_ctx)
 {
 	struct hubp *hubp = pipe_ctx->plane_res.hubp;
 	int dpp_id = pipe_ctx->plane_res.dpp->inst;
@@ -634,7 +869,8 @@ static void plane_atomic_disconnect(struct dc *dc, struct pipe_ctx *pipe_ctx)
 		return;
 
 	mpc->funcs->remove_mpcc(mpc, mpc_tree_params, mpcc_to_remove);
-	opp->mpcc_disconnect_pending[pipe_ctx->plane_res.mpcc_inst] = true;
+	if (opp != NULL)
+		opp->mpcc_disconnect_pending[pipe_ctx->plane_res.mpcc_inst] = true;
 
 	dc->optimized_required = true;
 
@@ -649,7 +885,7 @@ static void plane_atomic_power_down(struct dc *dc, struct pipe_ctx *pipe_ctx)
 {
 	struct dce_hwseq *hws = dc->hwseq;
 	struct dpp *dpp = pipe_ctx->plane_res.dpp;
-	struct dc_context *ctx = dc->ctx;
+	DC_LOGGER_INIT(dc->ctx->logger);
 
 	if (REG(DC_IP_REQUEST_CNTL)) {
 		REG_SET(DC_IP_REQUEST_CNTL, 0,
@@ -699,7 +935,7 @@ static void plane_atomic_disable(struct dc *dc, struct pipe_ctx *pipe_ctx)
 
 static void dcn10_disable_plane(struct dc *dc, struct pipe_ctx *pipe_ctx)
 {
-	struct dc_context *ctx = dc->ctx;
+	DC_LOGGER_INIT(dc->ctx->logger);
 
 	if (!pipe_ctx->plane_res.hubp || pipe_ctx->plane_res.hubp->power_gated)
 		return;
@@ -779,7 +1015,7 @@ static void dcn10_init_hw(struct dc *dc)
 	/* Reset all MPCC muxes */
 	dc->res_pool->mpc->funcs->mpc_init(dc->res_pool->mpc);
 
-	for (i = 0; i < dc->res_pool->pipe_count; i++) {
+	for (i = 0; i < dc->res_pool->timing_generator_count; i++) {
 		struct timing_generator *tg = dc->res_pool->timing_generators[i];
 		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
 		struct hubp *hubp = dc->res_pool->hubps[i];
@@ -800,7 +1036,7 @@ static void dcn10_init_hw(struct dc *dc)
 		dc->res_pool->opps[i]->mpcc_disconnect_pending[pipe_ctx->plane_res.mpcc_inst] = true;
 		pipe_ctx->stream_res.opp = dc->res_pool->opps[i];
 
-		plane_atomic_disconnect(dc, pipe_ctx);
+		hwss1_plane_atomic_disconnect(dc, pipe_ctx);
 	}
 
 	for (i = 0; i < dc->res_pool->pipe_count; i++) {
@@ -921,12 +1157,19 @@ static void dcn10_update_plane_addr(const struct dc *dc, struct pipe_ctx *pipe_c
 
 	if (plane_state == NULL)
 		return;
+
 	addr_patched = patch_address_for_sbs_tb_stereo(pipe_ctx, &addr);
+
 	pipe_ctx->plane_res.hubp->funcs->hubp_program_surface_flip_and_addr(
 			pipe_ctx->plane_res.hubp,
 			&plane_state->address,
 			plane_state->flip_immediate);
+
 	plane_state->status.requested_address = plane_state->address;
+
+	if (plane_state->flip_immediate)
+		plane_state->status.current_address = plane_state->address;
+
 	if (addr_patched)
 		pipe_ctx->plane_state->address.grph_stereo.left_addr = addr;
 }
@@ -945,9 +1188,8 @@ static bool dcn10_set_input_transfer_func(struct pipe_ctx *pipe_ctx,
 		tf = plane_state->in_transfer_func;
 
 	if (plane_state->gamma_correction &&
-		plane_state->gamma_correction->is_identity)
-		dpp_base->funcs->dpp_set_degamma(dpp_base, IPP_DEGAMMA_MODE_BYPASS);
-	else if (plane_state->gamma_correction && dce_use_lut(plane_state->format))
+		!plane_state->gamma_correction->is_identity
+			&& dce_use_lut(plane_state->format))
 		dpp_base->funcs->dpp_program_input_lut(dpp_base, plane_state->gamma_correction);
 
 	if (tf == NULL)
@@ -1113,10 +1355,11 @@ static void dcn10_enable_per_frame_crtc_position_reset(
 
 	DC_SYNC_INFO("Setting up\n");
 	for (i = 0; i < group_size; i++)
-		grouped_pipes[i]->stream_res.tg->funcs->enable_crtc_reset(
-				grouped_pipes[i]->stream_res.tg,
-				grouped_pipes[i]->stream->triggered_crtc_reset.event_source->status.primary_otg_inst,
-				&grouped_pipes[i]->stream->triggered_crtc_reset);
+		if (grouped_pipes[i]->stream_res.tg->funcs->enable_crtc_reset)
+			grouped_pipes[i]->stream_res.tg->funcs->enable_crtc_reset(
+					grouped_pipes[i]->stream_res.tg,
+					grouped_pipes[i]->stream->triggered_crtc_reset.event_source->status.primary_otg_inst,
+					&grouped_pipes[i]->stream->triggered_crtc_reset);
 
 	DC_SYNC_INFO("Waiting for trigger\n");
 
@@ -1433,7 +1676,7 @@ static void program_csc_matrix(struct pipe_ctx *pipe_ctx,
 	}
 }
 
-static void program_output_csc(struct dc *dc,
+static void dcn10_program_output_csc(struct dc *dc,
 		struct pipe_ctx *pipe_ctx,
 		enum dc_color_space colorspace,
 		uint16_t *matrix,
@@ -1532,6 +1775,43 @@ static void dcn10_get_surface_visual_confirm_color(
 	}
 }
 
+static void dcn10_get_hdr_visual_confirm_color(
+		struct pipe_ctx *pipe_ctx,
+		struct tg_color *color)
+{
+	uint32_t color_value = MAX_TG_COLOR_VALUE;
+
+	// Determine the overscan color based on the top-most (desktop) plane's context
+	struct pipe_ctx *top_pipe_ctx  = pipe_ctx;
+
+	while (top_pipe_ctx->top_pipe != NULL)
+		top_pipe_ctx = top_pipe_ctx->top_pipe;
+
+	switch (top_pipe_ctx->plane_res.scl_data.format) {
+	case PIXEL_FORMAT_ARGB2101010:
+		if (top_pipe_ctx->stream->out_transfer_func->tf == TRANSFER_FUNCTION_UNITY) {
+			/* HDR10, ARGB2101010 - set boarder color to red */
+			color->color_r_cr = color_value;
+		}
+		break;
+	case PIXEL_FORMAT_FP16:
+		if (top_pipe_ctx->stream->out_transfer_func->tf == TRANSFER_FUNCTION_PQ) {
+			/* HDR10, FP16 - set boarder color to blue */
+			color->color_b_cb = color_value;
+		} else if (top_pipe_ctx->stream->out_transfer_func->tf == TRANSFER_FUNCTION_GAMMA22) {
+			/* FreeSync 2 HDR - set boarder color to green */
+			color->color_g_y = color_value;
+		}
+		break;
+	default:
+		/* SDR - set boarder color to Gray */
+		color->color_r_cr = color_value/2;
+		color->color_b_cb = color_value/2;
+		color->color_g_y = color_value/2;
+		break;
+	}
+}
+
 static uint16_t fixed_point_to_int_frac(
 	struct fixed31_32 arg,
 	uint8_t integer_bits,
@@ -1542,22 +1822,22 @@ static uint16_t fixed_point_to_int_frac(
 
 	uint16_t result;
 
-	uint16_t d = (uint16_t)dal_fixed31_32_floor(
-		dal_fixed31_32_abs(
+	uint16_t d = (uint16_t)dc_fixpt_floor(
+		dc_fixpt_abs(
 			arg));
 
 	if (d <= (uint16_t)(1 << integer_bits) - (1 / (uint16_t)divisor))
-		numerator = (uint16_t)dal_fixed31_32_floor(
-			dal_fixed31_32_mul_int(
+		numerator = (uint16_t)dc_fixpt_floor(
+			dc_fixpt_mul_int(
 				arg,
 				divisor));
 	else {
-		numerator = dal_fixed31_32_floor(
-			dal_fixed31_32_sub(
-				dal_fixed31_32_from_int(
+		numerator = dc_fixpt_floor(
+			dc_fixpt_sub(
+				dc_fixpt_from_int(
 					1LL << integer_bits),
-				dal_fixed31_32_recip(
-					dal_fixed31_32_from_int(
+				dc_fixpt_recip(
+					dc_fixpt_from_int(
 						divisor))));
 	}
 
@@ -1567,8 +1847,8 @@ static uint16_t fixed_point_to_int_frac(
 		result = (uint16_t)(
 		(1 << (integer_bits + fractional_bits + 1)) + numerator);
 
-	if ((result != 0) && dal_fixed31_32_lt(
-		arg, dal_fixed31_32_zero))
+	if ((result != 0) && dc_fixpt_lt(
+		arg, dc_fixpt_zero))
 		result |= 1 << (integer_bits + fractional_bits);
 
 	return result;
@@ -1582,8 +1862,8 @@ void build_prescale_params(struct  dc_bias_and_scale *bias_and_scale,
 			&& plane_state->input_csc_color_matrix.enable_adjustment
 			&& plane_state->coeff_reduction_factor.value != 0) {
 		bias_and_scale->scale_blue = fixed_point_to_int_frac(
-			dal_fixed31_32_mul(plane_state->coeff_reduction_factor,
-					dal_fixed31_32_from_fraction(256, 255)),
+			dc_fixpt_mul(plane_state->coeff_reduction_factor,
+					dc_fixpt_from_fraction(256, 255)),
 				2,
 				13);
 		bias_and_scale->scale_red = bias_and_scale->scale_blue;
@@ -1612,26 +1892,31 @@ static void update_dpp(struct dpp *dpp, struct dc_plane_state *plane_state)
 		dpp->funcs->dpp_program_bias_and_scale(dpp, &bns_params);
 }
 
-
-static void update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx)
+static void dcn10_update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx)
 {
 	struct hubp *hubp = pipe_ctx->plane_res.hubp;
-	struct mpcc_blnd_cfg blnd_cfg;
+	struct mpcc_blnd_cfg blnd_cfg = {0};
 	bool per_pixel_alpha = pipe_ctx->plane_state->per_pixel_alpha && pipe_ctx->bottom_pipe;
 	int mpcc_id;
 	struct mpcc *new_mpcc;
 	struct mpc *mpc = dc->res_pool->mpc;
 	struct mpc_tree *mpc_tree_params = &(pipe_ctx->stream_res.opp->mpc_tree_params);
 
+
+
 	/* TODO: proper fix once fpga works */
 
-	if (dc->debug.surface_visual_confirm)
+	if (dc->debug.visual_confirm == VISUAL_CONFIRM_HDR) {
+		dcn10_get_hdr_visual_confirm_color(
+				pipe_ctx, &blnd_cfg.black_color);
+	} else if (dc->debug.visual_confirm == VISUAL_CONFIRM_SURFACE) {
 		dcn10_get_surface_visual_confirm_color(
 				pipe_ctx, &blnd_cfg.black_color);
-	else
+	} else {
 		color_space_to_black_color(
-			dc, pipe_ctx->stream->output_color_space,
-			&blnd_cfg.black_color);
+				dc, pipe_ctx->stream->output_color_space,
+				&blnd_cfg.black_color);
+	}
 
 	if (per_pixel_alpha)
 		blnd_cfg.alpha_mode = MPCC_ALPHA_BLEND_MODE_PER_PIXEL_ALPHA;
@@ -1649,6 +1934,7 @@ static void update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx)
 			pipe_ctx->stream->output_color_space)
 					&& per_pixel_alpha;
 
+
 	/*
 	 * TODO: remove hack
 	 * Note: currently there is a bug in init_hw such that
@@ -1659,6 +1945,12 @@ static void update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx)
 	 */
 	mpcc_id = hubp->inst;
 
+	/* If there is no full update, don't need to touch MPC tree*/
+	if (!pipe_ctx->plane_state->update_flags.bits.full_update) {
+		mpc->funcs->update_blending(mpc, &blnd_cfg, mpcc_id);
+		return;
+	}
+
 	/* check if this MPCC is already being used */
 	new_mpcc = mpc->funcs->get_mpcc_for_dpp(mpc_tree_params, mpcc_id);
 	/* remove MPCC if being used */
@@ -1713,18 +2005,17 @@ static void update_dchubp_dpp(
 	 * divided by 2
 	 */
 	if (plane_state->update_flags.bits.full_update) {
-		bool should_divided_by_2 = context->bw.dcn.calc_clk.dppclk_khz <=
-				context->bw.dcn.cur_clk.dispclk_khz / 2;
+		bool should_divided_by_2 = context->bw.dcn.clk.dppclk_khz <=
+				dc->res_pool->dccg->clks.dispclk_khz / 2;
 
 		dpp->funcs->dpp_dppclk_control(
 				dpp,
 				should_divided_by_2,
 				true);
 
-		dc->current_state->bw.dcn.cur_clk.dppclk_khz =
-				should_divided_by_2 ?
-				context->bw.dcn.cur_clk.dispclk_khz / 2 :
-				context->bw.dcn.cur_clk.dispclk_khz;
+		dc->res_pool->dccg->clks.dppclk_khz = should_divided_by_2 ?
+						dc->res_pool->dccg->clks.dispclk_khz / 2 :
+							dc->res_pool->dccg->clks.dispclk_khz;
 	}
 
 	/* TODO: Need input parameter to tell current DCHUB pipe tie to which OTG
@@ -1750,7 +2041,7 @@ static void update_dchubp_dpp(
 
 	if (plane_state->update_flags.bits.full_update ||
 		plane_state->update_flags.bits.per_pixel_alpha_change)
-		update_mpcc(dc, pipe_ctx);
+		dc->hwss.update_mpcc(dc, pipe_ctx);
 
 	if (plane_state->update_flags.bits.full_update ||
 		plane_state->update_flags.bits.per_pixel_alpha_change ||
@@ -1777,7 +2068,7 @@ static void update_dchubp_dpp(
 		/*gamut remap*/
 		program_gamut_remap(pipe_ctx);
 
-		program_output_csc(dc,
+		dc->hwss.program_output_csc(dc,
 				pipe_ctx,
 				pipe_ctx->stream->output_color_space,
 				pipe_ctx->stream->csc_color_matrix.matrix,
@@ -1810,40 +2101,41 @@ static void update_dchubp_dpp(
 		hubp->funcs->set_blank(hubp, false);
 }
 
-static void dcn10_otg_blank(
+static void dcn10_blank_pixel_data(
 		struct dc *dc,
-		struct stream_resource stream_res,
-		struct dc_stream_state *stream,
+		struct pipe_ctx *pipe_ctx,
 		bool blank)
 {
 	enum dc_color_space color_space;
 	struct tg_color black_color = {0};
+	struct stream_resource *stream_res = &pipe_ctx->stream_res;
+	struct dc_stream_state *stream = pipe_ctx->stream;
 
 	/* program otg blank color */
 	color_space = stream->output_color_space;
 	color_space_to_black_color(dc, color_space, &black_color);
 
-	if (stream_res.tg->funcs->set_blank_color)
-		stream_res.tg->funcs->set_blank_color(
-				stream_res.tg,
+	if (stream_res->tg->funcs->set_blank_color)
+		stream_res->tg->funcs->set_blank_color(
+				stream_res->tg,
 				&black_color);
 
 	if (!blank) {
-		if (stream_res.tg->funcs->set_blank)
-			stream_res.tg->funcs->set_blank(stream_res.tg, blank);
-		if (stream_res.abm)
-			stream_res.abm->funcs->set_abm_level(stream_res.abm, stream->abm_level);
+		if (stream_res->tg->funcs->set_blank)
+			stream_res->tg->funcs->set_blank(stream_res->tg, blank);
+		if (stream_res->abm)
+			stream_res->abm->funcs->set_abm_level(stream_res->abm, stream->abm_level);
 	} else if (blank) {
-		if (stream_res.abm)
-			stream_res.abm->funcs->set_abm_immediate_disable(stream_res.abm);
-		if (stream_res.tg->funcs->set_blank)
-			stream_res.tg->funcs->set_blank(stream_res.tg, blank);
+		if (stream_res->abm)
+			stream_res->abm->funcs->set_abm_immediate_disable(stream_res->abm);
+		if (stream_res->tg->funcs->set_blank)
+			stream_res->tg->funcs->set_blank(stream_res->tg, blank);
 	}
 }
 
 static void set_hdr_multiplier(struct pipe_ctx *pipe_ctx)
 {
-	struct fixed31_32 multiplier = dal_fixed31_32_from_fraction(
+	struct fixed31_32 multiplier = dc_fixpt_from_fraction(
 			pipe_ctx->plane_state->sdr_white_level, 80);
 	uint32_t hw_mult = 0x1f000; // 1.0 default multiplier
 	struct custom_float_format fmt;
@@ -1859,6 +2151,33 @@ static void set_hdr_multiplier(struct pipe_ctx *pipe_ctx)
 			pipe_ctx->plane_res.dpp, hw_mult);
 }
 
+void dcn10_program_pipe(
+		struct dc *dc,
+		struct pipe_ctx *pipe_ctx,
+		struct dc_state *context)
+{
+	if (pipe_ctx->plane_state->update_flags.bits.full_update)
+		dcn10_enable_plane(dc, pipe_ctx, context);
+
+	update_dchubp_dpp(dc, pipe_ctx, context);
+
+	set_hdr_multiplier(pipe_ctx);
+
+	if (pipe_ctx->plane_state->update_flags.bits.full_update ||
+			pipe_ctx->plane_state->update_flags.bits.in_transfer_func_change ||
+			pipe_ctx->plane_state->update_flags.bits.gamma_change)
+		dc->hwss.set_input_transfer_func(pipe_ctx, pipe_ctx->plane_state);
+
+	/* dcn10_translate_regamma_to_hw_format takes 750us to finish
+	 * only do gamma programming for full update.
+	 * TODO: This can be further optimized/cleaned up
+	 * Always call this for now since it does memcmp inside before
+	 * doing heavy calculation and programming
+	 */
+	if (pipe_ctx->plane_state->update_flags.bits.full_update)
+		dc->hwss.set_output_transfer_func(pipe_ctx, pipe_ctx->stream);
+}
+
 static void program_all_pipe_in_tree(
 		struct dc *dc,
 		struct pipe_ctx *pipe_ctx,
@@ -1876,31 +2195,12 @@ static void program_all_pipe_in_tree(
 		pipe_ctx->stream_res.tg->funcs->program_global_sync(
 				pipe_ctx->stream_res.tg);
 
-		dcn10_otg_blank(dc, pipe_ctx->stream_res,
-				pipe_ctx->stream, blank);
+		dc->hwss.blank_pixel_data(dc, pipe_ctx, blank);
+
 	}
 
 	if (pipe_ctx->plane_state != NULL) {
-		if (pipe_ctx->plane_state->update_flags.bits.full_update)
-			dcn10_enable_plane(dc, pipe_ctx, context);
-
-		update_dchubp_dpp(dc, pipe_ctx, context);
-
-		set_hdr_multiplier(pipe_ctx);
-
-		if (pipe_ctx->plane_state->update_flags.bits.full_update ||
-				pipe_ctx->plane_state->update_flags.bits.in_transfer_func_change ||
-				pipe_ctx->plane_state->update_flags.bits.gamma_change)
-			dc->hwss.set_input_transfer_func(pipe_ctx, pipe_ctx->plane_state);
-
-		/* dcn10_translate_regamma_to_hw_format takes 750us to finish
-		 * only do gamma programming for full update.
-		 * TODO: This can be further optimized/cleaned up
-		 * Always call this for now since it does memcmp inside before
-		 * doing heavy calculation and programming
-		 */
-		if (pipe_ctx->plane_state->update_flags.bits.full_update)
-			dc->hwss.set_output_transfer_func(pipe_ctx, pipe_ctx->stream);
+		dcn10_program_pipe(dc, pipe_ctx, context);
 	}
 
 	if (pipe_ctx->bottom_pipe != NULL && pipe_ctx->bottom_pipe != pipe_ctx) {
@@ -1914,12 +2214,12 @@ static void dcn10_pplib_apply_display_requirements(
 {
 	struct dm_pp_display_configuration *pp_display_cfg = &context->pp_display_cfg;
 
-	pp_display_cfg->min_engine_clock_khz = context->bw.dcn.cur_clk.dcfclk_khz;
-	pp_display_cfg->min_memory_clock_khz = context->bw.dcn.cur_clk.fclk_khz;
-	pp_display_cfg->min_engine_clock_deep_sleep_khz = context->bw.dcn.cur_clk.dcfclk_deep_sleep_khz;
-	pp_display_cfg->min_dcfc_deep_sleep_clock_khz = context->bw.dcn.cur_clk.dcfclk_deep_sleep_khz;
-	pp_display_cfg->min_dcfclock_khz = context->bw.dcn.cur_clk.dcfclk_khz;
-	pp_display_cfg->disp_clk_khz = context->bw.dcn.cur_clk.dispclk_khz;
+	pp_display_cfg->min_engine_clock_khz = dc->res_pool->dccg->clks.dcfclk_khz;
+	pp_display_cfg->min_memory_clock_khz = dc->res_pool->dccg->clks.fclk_khz;
+	pp_display_cfg->min_engine_clock_deep_sleep_khz = dc->res_pool->dccg->clks.dcfclk_deep_sleep_khz;
+	pp_display_cfg->min_dcfc_deep_sleep_clock_khz = dc->res_pool->dccg->clks.dcfclk_deep_sleep_khz;
+	pp_display_cfg->min_dcfclock_khz = dc->res_pool->dccg->clks.dcfclk_khz;
+	pp_display_cfg->disp_clk_khz = dc->res_pool->dccg->clks.dispclk_khz;
 	dce110_fill_display_configs(context, pp_display_cfg);
 
 	if (memcmp(&dc->prev_display_config, pp_display_cfg, sizeof(
@@ -1981,11 +2281,9 @@ static void dcn10_apply_ctx_for_surface(
 	int i;
 	struct timing_generator *tg;
 	bool removed_pipe[4] = { false };
-	unsigned int ref_clk_mhz = dc->res_pool->ref_clock_inKhz/1000;
-	bool program_water_mark = false;
-	struct dc_context *ctx = dc->ctx;
 	struct pipe_ctx *top_pipe_to_program =
 			find_top_pipe_for_stream(dc, context, stream);
+	DC_LOGGER_INIT(dc->ctx->logger);
 
 	if (!top_pipe_to_program)
 		return;
@@ -1996,7 +2294,7 @@ static void dcn10_apply_ctx_for_surface(
 
 	if (num_planes == 0) {
 		/* OTG blank before remove all front end */
-		dcn10_otg_blank(dc, top_pipe_to_program->stream_res, top_pipe_to_program->stream, true);
+		dc->hwss.blank_pixel_data(dc, top_pipe_to_program, true);
 	}
 
 	/* Disconnect unused mpcc */
@@ -2027,11 +2325,10 @@ static void dcn10_apply_ctx_for_surface(
 			old_pipe_ctx->plane_state &&
 			old_pipe_ctx->stream_res.tg == tg) {
 
-			plane_atomic_disconnect(dc, old_pipe_ctx);
+			dc->hwss.plane_atomic_disconnect(dc, old_pipe_ctx);
 			removed_pipe[i] = true;
 
-			DC_LOG_DC(
-					"Reset mpcc for pipe %d\n",
+			DC_LOG_DC("Reset mpcc for pipe %d\n",
 					old_pipe_ctx->pipe_idx);
 		}
 	}
@@ -2044,248 +2341,41 @@ static void dcn10_apply_ctx_for_surface(
 	if (num_planes == 0)
 		false_optc_underflow_wa(dc, stream, tg);
 
-	for (i = 0; i < dc->res_pool->pipe_count; i++) {
-		struct pipe_ctx *old_pipe_ctx =
-				&dc->current_state->res_ctx.pipe_ctx[i];
-		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
-
-		if (pipe_ctx->stream == stream &&
-				pipe_ctx->plane_state &&
-			pipe_ctx->plane_state->update_flags.bits.full_update)
-			program_water_mark = true;
-
+	for (i = 0; i < dc->res_pool->pipe_count; i++)
 		if (removed_pipe[i])
-			dcn10_disable_plane(dc, old_pipe_ctx);
-	}
-
-	if (program_water_mark) {
-		if (dc->debug.sanity_checks) {
-			/* pstate stuck check after watermark update */
-			dcn10_verify_allow_pstate_change_high(dc);
-		}
-
-		/* watermark is for all pipes */
-		hubbub1_program_watermarks(dc->res_pool->hubbub,
-				&context->bw.dcn.watermarks, ref_clk_mhz);
-
-		if (dc->debug.sanity_checks) {
-			/* pstate stuck check after watermark update */
-			dcn10_verify_allow_pstate_change_high(dc);
-		}
-	}
-/*	DC_LOG_BANDWIDTH_CALCS(dc->ctx->logger,
-			"\n============== Watermark parameters ==============\n"
-			"a.urgent_ns: %d \n"
-			"a.cstate_enter_plus_exit: %d \n"
-			"a.cstate_exit: %d \n"
-			"a.pstate_change: %d \n"
-			"a.pte_meta_urgent: %d \n"
-			"b.urgent_ns: %d \n"
-			"b.cstate_enter_plus_exit: %d \n"
-			"b.cstate_exit: %d \n"
-			"b.pstate_change: %d \n"
-			"b.pte_meta_urgent: %d \n",
-			context->bw.dcn.watermarks.a.urgent_ns,
-			context->bw.dcn.watermarks.a.cstate_pstate.cstate_enter_plus_exit_ns,
-			context->bw.dcn.watermarks.a.cstate_pstate.cstate_exit_ns,
-			context->bw.dcn.watermarks.a.cstate_pstate.pstate_change_ns,
-			context->bw.dcn.watermarks.a.pte_meta_urgent_ns,
-			context->bw.dcn.watermarks.b.urgent_ns,
-			context->bw.dcn.watermarks.b.cstate_pstate.cstate_enter_plus_exit_ns,
-			context->bw.dcn.watermarks.b.cstate_pstate.cstate_exit_ns,
-			context->bw.dcn.watermarks.b.cstate_pstate.pstate_change_ns,
-			context->bw.dcn.watermarks.b.pte_meta_urgent_ns
-			);
-	DC_LOG_BANDWIDTH_CALCS(dc->ctx->logger,
-			"\nc.urgent_ns: %d \n"
-			"c.cstate_enter_plus_exit: %d \n"
-			"c.cstate_exit: %d \n"
-			"c.pstate_change: %d \n"
-			"c.pte_meta_urgent: %d \n"
-			"d.urgent_ns: %d \n"
-			"d.cstate_enter_plus_exit: %d \n"
-			"d.cstate_exit: %d \n"
-			"d.pstate_change: %d \n"
-			"d.pte_meta_urgent: %d \n"
-			"========================================================\n",
-			context->bw.dcn.watermarks.c.urgent_ns,
-			context->bw.dcn.watermarks.c.cstate_pstate.cstate_enter_plus_exit_ns,
-			context->bw.dcn.watermarks.c.cstate_pstate.cstate_exit_ns,
-			context->bw.dcn.watermarks.c.cstate_pstate.pstate_change_ns,
-			context->bw.dcn.watermarks.c.pte_meta_urgent_ns,
-			context->bw.dcn.watermarks.d.urgent_ns,
-			context->bw.dcn.watermarks.d.cstate_pstate.cstate_enter_plus_exit_ns,
-			context->bw.dcn.watermarks.d.cstate_pstate.cstate_exit_ns,
-			context->bw.dcn.watermarks.d.cstate_pstate.pstate_change_ns,
-			context->bw.dcn.watermarks.d.pte_meta_urgent_ns
-			);
-*/
-}
-
-static inline bool should_set_clock(bool decrease_allowed, int calc_clk, int cur_clk)
-{
-	return ((decrease_allowed && calc_clk < cur_clk) || calc_clk > cur_clk);
-}
-
-static int determine_dppclk_threshold(struct dc *dc, struct dc_state *context)
-{
-	bool request_dpp_div = context->bw.dcn.calc_clk.dispclk_khz >
-			context->bw.dcn.calc_clk.dppclk_khz;
-	bool dispclk_increase = context->bw.dcn.calc_clk.dispclk_khz >
-			context->bw.dcn.cur_clk.dispclk_khz;
-	int disp_clk_threshold = context->bw.dcn.calc_clk.max_supported_dppclk_khz;
-	bool cur_dpp_div = context->bw.dcn.cur_clk.dispclk_khz >
-			context->bw.dcn.cur_clk.dppclk_khz;
-
-	/* increase clock, looking for div is 0 for current, request div is 1*/
-	if (dispclk_increase) {
-		/* already divided by 2, no need to reach target clk with 2 steps*/
-		if (cur_dpp_div)
-			return context->bw.dcn.calc_clk.dispclk_khz;
-
-		/* request disp clk is lower than maximum supported dpp clk,
-		 * no need to reach target clk with two steps.
-		 */
-		if (context->bw.dcn.calc_clk.dispclk_khz <= disp_clk_threshold)
-			return context->bw.dcn.calc_clk.dispclk_khz;
-
-		/* target dpp clk not request divided by 2, still within threshold */
-		if (!request_dpp_div)
-			return context->bw.dcn.calc_clk.dispclk_khz;
-
-	} else {
-		/* decrease clock, looking for current dppclk divided by 2,
-		 * request dppclk not divided by 2.
-		 */
-
-		/* current dpp clk not divided by 2, no need to ramp*/
-		if (!cur_dpp_div)
-			return context->bw.dcn.calc_clk.dispclk_khz;
-
-		/* current disp clk is lower than current maximum dpp clk,
-		 * no need to ramp
-		 */
-		if (context->bw.dcn.cur_clk.dispclk_khz <= disp_clk_threshold)
-			return context->bw.dcn.calc_clk.dispclk_khz;
-
-		/* request dpp clk need to be divided by 2 */
-		if (request_dpp_div)
-			return context->bw.dcn.calc_clk.dispclk_khz;
-	}
+			dcn10_disable_plane(dc, &dc->current_state->res_ctx.pipe_ctx[i]);
 
-	return disp_clk_threshold;
-}
-
-static void ramp_up_dispclk_with_dpp(struct dc *dc, struct dc_state *context)
-{
-	int i;
-	bool request_dpp_div = context->bw.dcn.calc_clk.dispclk_khz >
-				context->bw.dcn.calc_clk.dppclk_khz;
-
-	int dispclk_to_dpp_threshold = determine_dppclk_threshold(dc, context);
-
-	/* set disp clk to dpp clk threshold */
-	dc->res_pool->display_clock->funcs->set_clock(
-			dc->res_pool->display_clock,
-			dispclk_to_dpp_threshold);
-
-	/* update request dpp clk division option */
-	for (i = 0; i < dc->res_pool->pipe_count; i++) {
-		struct pipe_ctx *pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i];
-
-		if (!pipe_ctx->plane_state)
-			continue;
-
-		pipe_ctx->plane_res.dpp->funcs->dpp_dppclk_control(
-				pipe_ctx->plane_res.dpp,
-				request_dpp_div,
-				true);
-	}
-
-	/* If target clk not same as dppclk threshold, set to target clock */
-	if (dispclk_to_dpp_threshold != context->bw.dcn.calc_clk.dispclk_khz) {
-		dc->res_pool->display_clock->funcs->set_clock(
-				dc->res_pool->display_clock,
-				context->bw.dcn.calc_clk.dispclk_khz);
-	}
-
-	context->bw.dcn.cur_clk.dispclk_khz =
-			context->bw.dcn.calc_clk.dispclk_khz;
-	context->bw.dcn.cur_clk.dppclk_khz =
-			context->bw.dcn.calc_clk.dppclk_khz;
-	context->bw.dcn.cur_clk.max_supported_dppclk_khz =
-			context->bw.dcn.calc_clk.max_supported_dppclk_khz;
+	if (dc->hwseq->wa.DEGVIDCN10_254)
+		hubbub1_wm_change_req_wa(dc->res_pool->hubbub);
 }
 
 static void dcn10_set_bandwidth(
 		struct dc *dc,
 		struct dc_state *context,
-		bool decrease_allowed)
+		bool safe_to_lower)
 {
-	struct pp_smu_display_requirement_rv *smu_req_cur =
-			&dc->res_pool->pp_smu_req;
-	struct pp_smu_display_requirement_rv smu_req = *smu_req_cur;
-	struct pp_smu_funcs_rv *pp_smu = dc->res_pool->pp_smu;
-
-	if (dc->debug.sanity_checks) {
+	if (dc->debug.sanity_checks)
 		dcn10_verify_allow_pstate_change_high(dc);
-	}
-
-	if (IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment))
-		return;
-
-	if (should_set_clock(
-			decrease_allowed,
-			context->bw.dcn.calc_clk.dcfclk_khz,
-			dc->current_state->bw.dcn.cur_clk.dcfclk_khz)) {
-		context->bw.dcn.cur_clk.dcfclk_khz =
-				context->bw.dcn.calc_clk.dcfclk_khz;
-		smu_req.hard_min_dcefclk_khz =
-				context->bw.dcn.calc_clk.dcfclk_khz;
-	}
 
-	if (should_set_clock(
-			decrease_allowed,
-			context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz,
-			dc->current_state->bw.dcn.cur_clk.dcfclk_deep_sleep_khz)) {
-		context->bw.dcn.cur_clk.dcfclk_deep_sleep_khz =
-				context->bw.dcn.calc_clk.dcfclk_deep_sleep_khz;
-	}
-
-	if (should_set_clock(
-			decrease_allowed,
-			context->bw.dcn.calc_clk.fclk_khz,
-			dc->current_state->bw.dcn.cur_clk.fclk_khz)) {
-		context->bw.dcn.cur_clk.fclk_khz =
-				context->bw.dcn.calc_clk.fclk_khz;
-		smu_req.hard_min_fclk_khz = context->bw.dcn.calc_clk.fclk_khz;
-	}
-
-	smu_req.display_count = context->stream_count;
-
-	if (pp_smu->set_display_requirement)
-		pp_smu->set_display_requirement(&pp_smu->pp_smu, &smu_req);
+	if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) {
+		if (context->stream_count == 0)
+			context->bw.dcn.clk.phyclk_khz = 0;
 
-	*smu_req_cur = smu_req;
+		dc->res_pool->dccg->funcs->update_clocks(
+				dc->res_pool->dccg,
+				&context->bw.dcn.clk,
+				safe_to_lower);
 
-	/* make sure dcf clk is before dpp clk to
-	 * make sure we have enough voltage to run dpp clk
-	 */
-	if (should_set_clock(
-			decrease_allowed,
-			context->bw.dcn.calc_clk.dispclk_khz,
-			dc->current_state->bw.dcn.cur_clk.dispclk_khz)) {
-
-		ramp_up_dispclk_with_dpp(dc, context);
+		dcn10_pplib_apply_display_requirements(dc, context);
 	}
 
-	dcn10_pplib_apply_display_requirements(dc, context);
+	hubbub1_program_watermarks(dc->res_pool->hubbub,
+			&context->bw.dcn.watermarks,
+			dc->res_pool->ref_clock_inKhz / 1000,
+			true);
 
-	if (dc->debug.sanity_checks) {
+	if (dc->debug.sanity_checks)
 		dcn10_verify_allow_pstate_change_high(dc);
-	}
-
-	/* need to fix this function.  not doing the right thing here */
 }
 
 static void set_drr(struct pipe_ctx **pipe_ctx,
@@ -2335,15 +2425,6 @@ static void set_static_screen_control(struct pipe_ctx **pipe_ctx,
 			set_static_screen_control(pipe_ctx[i]->stream_res.tg, value);
 }
 
-static void set_plane_config(
-	const struct dc *dc,
-	struct pipe_ctx *pipe_ctx,
-	struct resource_context *res_ctx)
-{
-	/* TODO */
-	program_gamut_remap(pipe_ctx);
-}
-
 static void dcn10_config_stereo_parameters(
 		struct dc_stream_state *stream, struct crtc_stereo_flags *flags)
 {
@@ -2459,16 +2540,20 @@ static void dcn10_update_pending_status(struct pipe_ctx *pipe_ctx)
 {
 	struct dc_plane_state *plane_state = pipe_ctx->plane_state;
 	struct timing_generator *tg = pipe_ctx->stream_res.tg;
+	bool flip_pending;
 
 	if (plane_state == NULL)
 		return;
 
-	plane_state->status.is_flip_pending =
-			pipe_ctx->plane_res.hubp->funcs->hubp_is_flip_pending(
+	flip_pending = pipe_ctx->plane_res.hubp->funcs->hubp_is_flip_pending(
 					pipe_ctx->plane_res.hubp);
 
-	plane_state->status.current_address = pipe_ctx->plane_res.hubp->current_address;
-	if (pipe_ctx->plane_res.hubp->current_address.type == PLN_ADDR_TYPE_GRPH_STEREO &&
+	plane_state->status.is_flip_pending = flip_pending;
+
+	if (!flip_pending)
+		plane_state->status.current_address = plane_state->status.requested_address;
+
+	if (plane_state->status.current_address.type == PLN_ADDR_TYPE_GRPH_STEREO &&
 			tg->funcs->is_stereo_left_eye) {
 		plane_state->status.is_right_eye =
 				!tg->funcs->is_stereo_left_eye(pipe_ctx->stream_res.tg);
@@ -2477,8 +2562,14 @@ static void dcn10_update_pending_status(struct pipe_ctx *pipe_ctx)
 
 static void dcn10_update_dchub(struct dce_hwseq *hws, struct dchub_init_data *dh_data)
 {
-	if (hws->ctx->dc->res_pool->hubbub != NULL)
-		hubbub1_update_dchub(hws->ctx->dc->res_pool->hubbub, dh_data);
+	if (hws->ctx->dc->res_pool->hubbub != NULL) {
+		struct hubp *hubp = hws->ctx->dc->res_pool->hubps[0];
+
+		if (hubp->funcs->hubp_update_dchub)
+			hubp->funcs->hubp_update_dchub(hubp, dh_data);
+		else
+			hubbub1_update_dchub(hws->ctx->dc->res_pool->hubbub, dh_data);
+	}
 }
 
 static void dcn10_set_cursor_position(struct pipe_ctx *pipe_ctx)
@@ -2489,9 +2580,11 @@ static void dcn10_set_cursor_position(struct pipe_ctx *pipe_ctx)
 	struct dc_cursor_mi_param param = {
 		.pixel_clk_khz = pipe_ctx->stream->timing.pix_clk_khz,
 		.ref_clk_khz = pipe_ctx->stream->ctx->dc->res_pool->ref_clock_inKhz,
-		.viewport_x_start = pipe_ctx->plane_res.scl_data.viewport.x,
-		.viewport_width = pipe_ctx->plane_res.scl_data.viewport.width,
-		.h_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.horz
+		.viewport = pipe_ctx->plane_res.scl_data.viewport,
+		.h_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.horz,
+		.v_scale_ratio = pipe_ctx->plane_res.scl_data.ratios.vert,
+		.rotation = pipe_ctx->plane_state->rotation,
+		.mirror = pipe_ctx->plane_state->horizontal_mirror
 	};
 
 	if (pipe_ctx->plane_state->address.type
@@ -2515,18 +2608,47 @@ static void dcn10_set_cursor_attribute(struct pipe_ctx *pipe_ctx)
 		pipe_ctx->plane_res.dpp, attributes->color_format);
 }
 
+static void dcn10_set_cursor_sdr_white_level(struct pipe_ctx *pipe_ctx)
+{
+	uint32_t sdr_white_level = pipe_ctx->stream->cursor_attributes.sdr_white_level;
+	struct fixed31_32 multiplier;
+	struct dpp_cursor_attributes opt_attr = { 0 };
+	uint32_t hw_scale = 0x3c00; // 1.0 default multiplier
+	struct custom_float_format fmt;
+
+	if (!pipe_ctx->plane_res.dpp->funcs->set_optional_cursor_attributes)
+		return;
+
+	fmt.exponenta_bits = 5;
+	fmt.mantissa_bits = 10;
+	fmt.sign = true;
+
+	if (sdr_white_level > 80) {
+		multiplier = dc_fixpt_from_fraction(sdr_white_level, 80);
+		convert_to_custom_float_format(multiplier, &fmt, &hw_scale);
+	}
+
+	opt_attr.scale = hw_scale;
+	opt_attr.bias = 0;
+
+	pipe_ctx->plane_res.dpp->funcs->set_optional_cursor_attributes(
+			pipe_ctx->plane_res.dpp, &opt_attr);
+}
+
 static const struct hw_sequencer_funcs dcn10_funcs = {
 	.program_gamut_remap = program_gamut_remap,
 	.program_csc_matrix = program_csc_matrix,
 	.init_hw = dcn10_init_hw,
 	.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
 	.apply_ctx_for_surface = dcn10_apply_ctx_for_surface,
-	.set_plane_config = set_plane_config,
 	.update_plane_addr = dcn10_update_plane_addr,
+	.plane_atomic_disconnect = hwss1_plane_atomic_disconnect,
 	.update_dchub = dcn10_update_dchub,
+	.update_mpcc = dcn10_update_mpcc,
 	.update_pending_status = dcn10_update_pending_status,
 	.set_input_transfer_func = dcn10_set_input_transfer_func,
 	.set_output_transfer_func = dcn10_set_output_transfer_func,
+	.program_output_csc = dcn10_program_output_csc,
 	.power_down = dce110_power_down,
 	.enable_accelerated_mode = dce110_enable_accelerated_mode,
 	.enable_timing_synchronization = dcn10_enable_timing_synchronization,
@@ -2536,12 +2658,15 @@ static const struct hw_sequencer_funcs dcn10_funcs = {
 	.disable_stream = dce110_disable_stream,
 	.unblank_stream = dce110_unblank_stream,
 	.blank_stream = dce110_blank_stream,
+	.enable_audio_stream = dce110_enable_audio_stream,
+	.disable_audio_stream = dce110_disable_audio_stream,
 	.enable_display_power_gating = dcn10_dummy_display_power_gating,
 	.disable_plane = dcn10_disable_plane,
+	.blank_pixel_data = dcn10_blank_pixel_data,
 	.pipe_control_lock = dcn10_pipe_control_lock,
 	.set_bandwidth = dcn10_set_bandwidth,
 	.reset_hw_ctx_wrap = reset_hw_ctx_wrap,
-	.prog_pixclk_crtc_otg = dcn10_prog_pixclk_crtc_otg,
+	.enable_stream_timing = dcn10_enable_stream_timing,
 	.set_drr = set_drr,
 	.get_position = get_position,
 	.set_static_screen_control = set_static_screen_control,
@@ -2557,7 +2682,8 @@ static const struct hw_sequencer_funcs dcn10_funcs = {
 	.edp_power_control = hwss_edp_power_control,
 	.edp_wait_for_hpd_ready = hwss_edp_wait_for_hpd_ready,
 	.set_cursor_position = dcn10_set_cursor_position,
-	.set_cursor_attribute = dcn10_set_cursor_attribute
+	.set_cursor_attribute = dcn10_set_cursor_attribute,
+	.set_cursor_sdr_white_level = dcn10_set_cursor_sdr_white_level
 };
 
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h
index 6c526b5095d9..7139fb73e966 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h
@@ -37,4 +37,13 @@ extern void fill_display_configs(
 
 bool is_rgb_cspace(enum dc_color_space output_color_space);
 
+void hwss1_plane_atomic_disconnect(struct dc *dc, struct pipe_ctx *pipe_ctx);
+
+void dcn10_verify_allow_pstate_change_high(struct dc *dc);
+
+void dcn10_program_pipe(
+		struct dc *dc,
+		struct pipe_ctx *pipe_ctx,
+		struct dc_state *context);
+
 #endif /* __DC_HWSS_DCN10_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c
new file mode 100644
index 000000000000..6f675206a136
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.c
@@ -0,0 +1,1363 @@
+/*
+ * Copyright 2012-15 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include "reg_helper.h"
+
+#include "core_types.h"
+#include "link_encoder.h"
+#include "dcn10_link_encoder.h"
+#include "stream_encoder.h"
+#include "i2caux_interface.h"
+#include "dc_bios_types.h"
+
+#include "gpio_service_interface.h"
+
+#define CTX \
+	enc10->base.ctx
+#define DC_LOGGER \
+	enc10->base.ctx->logger
+
+#define REG(reg)\
+	(enc10->link_regs->reg)
+
+#undef FN
+#define FN(reg_name, field_name) \
+	enc10->link_shift->field_name, enc10->link_mask->field_name
+
+
+/*
+ * @brief
+ * Trigger Source Select
+ * ASIC-dependent, actual values for register programming
+ */
+#define DCN10_DIG_FE_SOURCE_SELECT_INVALID 0x0
+#define DCN10_DIG_FE_SOURCE_SELECT_DIGA 0x1
+#define DCN10_DIG_FE_SOURCE_SELECT_DIGB 0x2
+#define DCN10_DIG_FE_SOURCE_SELECT_DIGC 0x4
+#define DCN10_DIG_FE_SOURCE_SELECT_DIGD 0x08
+#define DCN10_DIG_FE_SOURCE_SELECT_DIGE 0x10
+#define DCN10_DIG_FE_SOURCE_SELECT_DIGF 0x20
+#define DCN10_DIG_FE_SOURCE_SELECT_DIGG 0x40
+
+enum {
+	DP_MST_UPDATE_MAX_RETRY = 50
+};
+
+static const struct link_encoder_funcs dcn10_lnk_enc_funcs = {
+	.validate_output_with_stream =
+		dcn10_link_encoder_validate_output_with_stream,
+	.hw_init = dcn10_link_encoder_hw_init,
+	.setup = dcn10_link_encoder_setup,
+	.enable_tmds_output = dcn10_link_encoder_enable_tmds_output,
+	.enable_dp_output = dcn10_link_encoder_enable_dp_output,
+	.enable_dp_mst_output = dcn10_link_encoder_enable_dp_mst_output,
+	.disable_output = dcn10_link_encoder_disable_output,
+	.dp_set_lane_settings = dcn10_link_encoder_dp_set_lane_settings,
+	.dp_set_phy_pattern = dcn10_link_encoder_dp_set_phy_pattern,
+	.update_mst_stream_allocation_table =
+		dcn10_link_encoder_update_mst_stream_allocation_table,
+	.psr_program_dp_dphy_fast_training =
+			dcn10_psr_program_dp_dphy_fast_training,
+	.psr_program_secondary_packet = dcn10_psr_program_secondary_packet,
+	.connect_dig_be_to_fe = dcn10_link_encoder_connect_dig_be_to_fe,
+	.enable_hpd = dcn10_link_encoder_enable_hpd,
+	.disable_hpd = dcn10_link_encoder_disable_hpd,
+	.is_dig_enabled = dcn10_is_dig_enabled,
+	.destroy = dcn10_link_encoder_destroy
+};
+
+static enum bp_result link_transmitter_control(
+	struct dcn10_link_encoder *enc10,
+	struct bp_transmitter_control *cntl)
+{
+	enum bp_result result;
+	struct dc_bios *bp = enc10->base.ctx->dc_bios;
+
+	result = bp->funcs->transmitter_control(bp, cntl);
+
+	return result;
+}
+
+static void enable_phy_bypass_mode(
+	struct dcn10_link_encoder *enc10,
+	bool enable)
+{
+	/* This register resides in DP back end block;
+	 * transmitter is used for the offset
+	 */
+	REG_UPDATE(DP_DPHY_CNTL, DPHY_BYPASS, enable);
+
+}
+
+static void disable_prbs_symbols(
+	struct dcn10_link_encoder *enc10,
+	bool disable)
+{
+	/* This register resides in DP back end block;
+	 * transmitter is used for the offset
+	 */
+	REG_UPDATE_4(DP_DPHY_CNTL,
+			DPHY_ATEST_SEL_LANE0, disable,
+			DPHY_ATEST_SEL_LANE1, disable,
+			DPHY_ATEST_SEL_LANE2, disable,
+			DPHY_ATEST_SEL_LANE3, disable);
+}
+
+static void disable_prbs_mode(
+	struct dcn10_link_encoder *enc10)
+{
+	REG_UPDATE(DP_DPHY_PRBS_CNTL, DPHY_PRBS_EN, 0);
+}
+
+static void program_pattern_symbols(
+	struct dcn10_link_encoder *enc10,
+	uint16_t pattern_symbols[8])
+{
+	/* This register resides in DP back end block;
+	 * transmitter is used for the offset
+	 */
+	REG_SET_3(DP_DPHY_SYM0, 0,
+			DPHY_SYM1, pattern_symbols[0],
+			DPHY_SYM2, pattern_symbols[1],
+			DPHY_SYM3, pattern_symbols[2]);
+
+	/* This register resides in DP back end block;
+	 * transmitter is used for the offset
+	 */
+	REG_SET_3(DP_DPHY_SYM1, 0,
+			DPHY_SYM4, pattern_symbols[3],
+			DPHY_SYM5, pattern_symbols[4],
+			DPHY_SYM6, pattern_symbols[5]);
+
+	/* This register resides in DP back end block;
+	 * transmitter is used for the offset
+	 */
+	REG_SET_2(DP_DPHY_SYM2, 0,
+			DPHY_SYM7, pattern_symbols[6],
+			DPHY_SYM8, pattern_symbols[7]);
+}
+
+static void set_dp_phy_pattern_d102(
+	struct dcn10_link_encoder *enc10)
+{
+	/* Disable PHY Bypass mode to setup the test pattern */
+	enable_phy_bypass_mode(enc10, false);
+
+	/* For 10-bit PRBS or debug symbols
+	 * please use the following sequence:
+	 *
+	 * Enable debug symbols on the lanes
+	 */
+	disable_prbs_symbols(enc10, true);
+
+	/* Disable PRBS mode */
+	disable_prbs_mode(enc10);
+
+	/* Program debug symbols to be output */
+	{
+		uint16_t pattern_symbols[8] = {
+			0x2AA, 0x2AA, 0x2AA, 0x2AA,
+			0x2AA, 0x2AA, 0x2AA, 0x2AA
+		};
+
+		program_pattern_symbols(enc10, pattern_symbols);
+	}
+
+	/* Enable phy bypass mode to enable the test pattern */
+
+	enable_phy_bypass_mode(enc10, true);
+}
+
+static void set_link_training_complete(
+	struct dcn10_link_encoder *enc10,
+	bool complete)
+{
+	/* This register resides in DP back end block;
+	 * transmitter is used for the offset
+	 */
+	REG_UPDATE(DP_LINK_CNTL, DP_LINK_TRAINING_COMPLETE, complete);
+
+}
+
+void dcn10_link_encoder_set_dp_phy_pattern_training_pattern(
+	struct link_encoder *enc,
+	uint32_t index)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	/* Write Training Pattern */
+
+	REG_WRITE(DP_DPHY_TRAINING_PATTERN_SEL, index);
+
+	/* Set HW Register Training Complete to false */
+
+	set_link_training_complete(enc10, false);
+
+	/* Disable PHY Bypass mode to output Training Pattern */
+
+	enable_phy_bypass_mode(enc10, false);
+
+	/* Disable PRBS mode */
+	disable_prbs_mode(enc10);
+}
+
+static void setup_panel_mode(
+	struct dcn10_link_encoder *enc10,
+	enum dp_panel_mode panel_mode)
+{
+	uint32_t value;
+
+	ASSERT(REG(DP_DPHY_INTERNAL_CTRL));
+	value = REG_READ(DP_DPHY_INTERNAL_CTRL);
+
+	switch (panel_mode) {
+	case DP_PANEL_MODE_EDP:
+		value = 0x1;
+		break;
+	case DP_PANEL_MODE_SPECIAL:
+		value = 0x11;
+		break;
+	default:
+		value = 0x0;
+		break;
+	}
+
+	REG_WRITE(DP_DPHY_INTERNAL_CTRL, value);
+}
+
+static void set_dp_phy_pattern_symbol_error(
+	struct dcn10_link_encoder *enc10)
+{
+	/* Disable PHY Bypass mode to setup the test pattern */
+	enable_phy_bypass_mode(enc10, false);
+
+	/* program correct panel mode*/
+	setup_panel_mode(enc10, DP_PANEL_MODE_DEFAULT);
+
+	/* A PRBS23 pattern is used for most DP electrical measurements. */
+
+	/* Enable PRBS symbols on the lanes */
+	disable_prbs_symbols(enc10, false);
+
+	/* For PRBS23 Set bit DPHY_PRBS_SEL=1 and Set bit DPHY_PRBS_EN=1 */
+	REG_UPDATE_2(DP_DPHY_PRBS_CNTL,
+			DPHY_PRBS_SEL, 1,
+			DPHY_PRBS_EN, 1);
+
+	/* Enable phy bypass mode to enable the test pattern */
+	enable_phy_bypass_mode(enc10, true);
+}
+
+static void set_dp_phy_pattern_prbs7(
+	struct dcn10_link_encoder *enc10)
+{
+	/* Disable PHY Bypass mode to setup the test pattern */
+	enable_phy_bypass_mode(enc10, false);
+
+	/* A PRBS7 pattern is used for most DP electrical measurements. */
+
+	/* Enable PRBS symbols on the lanes */
+	disable_prbs_symbols(enc10, false);
+
+	/* For PRBS7 Set bit DPHY_PRBS_SEL=0 and Set bit DPHY_PRBS_EN=1 */
+	REG_UPDATE_2(DP_DPHY_PRBS_CNTL,
+			DPHY_PRBS_SEL, 0,
+			DPHY_PRBS_EN, 1);
+
+	/* Enable phy bypass mode to enable the test pattern */
+	enable_phy_bypass_mode(enc10, true);
+}
+
+static void set_dp_phy_pattern_80bit_custom(
+	struct dcn10_link_encoder *enc10,
+	const uint8_t *pattern)
+{
+	/* Disable PHY Bypass mode to setup the test pattern */
+	enable_phy_bypass_mode(enc10, false);
+
+	/* Enable debug symbols on the lanes */
+
+	disable_prbs_symbols(enc10, true);
+
+	/* Enable PHY bypass mode to enable the test pattern */
+	/* TODO is it really needed ? */
+
+	enable_phy_bypass_mode(enc10, true);
+
+	/* Program 80 bit custom pattern */
+	{
+		uint16_t pattern_symbols[8];
+
+		pattern_symbols[0] =
+			((pattern[1] & 0x03) << 8) | pattern[0];
+		pattern_symbols[1] =
+			((pattern[2] & 0x0f) << 6) | ((pattern[1] >> 2) & 0x3f);
+		pattern_symbols[2] =
+			((pattern[3] & 0x3f) << 4) | ((pattern[2] >> 4) & 0x0f);
+		pattern_symbols[3] =
+			(pattern[4] << 2) | ((pattern[3] >> 6) & 0x03);
+		pattern_symbols[4] =
+			((pattern[6] & 0x03) << 8) | pattern[5];
+		pattern_symbols[5] =
+			((pattern[7] & 0x0f) << 6) | ((pattern[6] >> 2) & 0x3f);
+		pattern_symbols[6] =
+			((pattern[8] & 0x3f) << 4) | ((pattern[7] >> 4) & 0x0f);
+		pattern_symbols[7] =
+			(pattern[9] << 2) | ((pattern[8] >> 6) & 0x03);
+
+		program_pattern_symbols(enc10, pattern_symbols);
+	}
+
+	/* Enable phy bypass mode to enable the test pattern */
+
+	enable_phy_bypass_mode(enc10, true);
+}
+
+static void set_dp_phy_pattern_hbr2_compliance_cp2520_2(
+	struct dcn10_link_encoder *enc10,
+	unsigned int cp2520_pattern)
+{
+
+	/* previously there is a register DP_HBR2_EYE_PATTERN
+	 * that is enabled to get the pattern.
+	 * But it does not work with the latest spec change,
+	 * so we are programming the following registers manually.
+	 *
+	 * The following settings have been confirmed
+	 * by Nick Chorney and Sandra Liu
+	 */
+
+	/* Disable PHY Bypass mode to setup the test pattern */
+
+	enable_phy_bypass_mode(enc10, false);
+
+	/* Setup DIG encoder in DP SST mode */
+	enc10->base.funcs->setup(&enc10->base, SIGNAL_TYPE_DISPLAY_PORT);
+
+	/* ensure normal panel mode. */
+	setup_panel_mode(enc10, DP_PANEL_MODE_DEFAULT);
+
+	/* no vbid after BS (SR)
+	 * DP_LINK_FRAMING_CNTL changed history Sandra Liu
+	 * 11000260 / 11000104 / 110000FC
+	 */
+	REG_UPDATE_3(DP_LINK_FRAMING_CNTL,
+			DP_IDLE_BS_INTERVAL, 0xFC,
+			DP_VBID_DISABLE, 1,
+			DP_VID_ENHANCED_FRAME_MODE, 1);
+
+	/* swap every BS with SR */
+	REG_UPDATE(DP_DPHY_SCRAM_CNTL, DPHY_SCRAMBLER_BS_COUNT, 0);
+
+	/* select cp2520 patterns */
+	if (REG(DP_DPHY_HBR2_PATTERN_CONTROL))
+		REG_UPDATE(DP_DPHY_HBR2_PATTERN_CONTROL,
+				DP_DPHY_HBR2_PATTERN_CONTROL, cp2520_pattern);
+	else
+		/* pre-DCE11 can only generate CP2520 pattern 2 */
+		ASSERT(cp2520_pattern == 2);
+
+	/* set link training complete */
+	set_link_training_complete(enc10, true);
+
+	/* disable video stream */
+	REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, 0);
+
+	/* Disable PHY Bypass mode to setup the test pattern */
+	enable_phy_bypass_mode(enc10, false);
+}
+
+static void set_dp_phy_pattern_passthrough_mode(
+	struct dcn10_link_encoder *enc10,
+	enum dp_panel_mode panel_mode)
+{
+	/* program correct panel mode */
+	setup_panel_mode(enc10, panel_mode);
+
+	/* restore LINK_FRAMING_CNTL and DPHY_SCRAMBLER_BS_COUNT
+	 * in case we were doing HBR2 compliance pattern before
+	 */
+	REG_UPDATE_3(DP_LINK_FRAMING_CNTL,
+			DP_IDLE_BS_INTERVAL, 0x2000,
+			DP_VBID_DISABLE, 0,
+			DP_VID_ENHANCED_FRAME_MODE, 1);
+
+	REG_UPDATE(DP_DPHY_SCRAM_CNTL, DPHY_SCRAMBLER_BS_COUNT, 0x1FF);
+
+	/* set link training complete */
+	set_link_training_complete(enc10, true);
+
+	/* Disable PHY Bypass mode to setup the test pattern */
+	enable_phy_bypass_mode(enc10, false);
+
+	/* Disable PRBS mode */
+	disable_prbs_mode(enc10);
+}
+
+/* return value is bit-vector */
+static uint8_t get_frontend_source(
+	enum engine_id engine)
+{
+	switch (engine) {
+	case ENGINE_ID_DIGA:
+		return DCN10_DIG_FE_SOURCE_SELECT_DIGA;
+	case ENGINE_ID_DIGB:
+		return DCN10_DIG_FE_SOURCE_SELECT_DIGB;
+	case ENGINE_ID_DIGC:
+		return DCN10_DIG_FE_SOURCE_SELECT_DIGC;
+	case ENGINE_ID_DIGD:
+		return DCN10_DIG_FE_SOURCE_SELECT_DIGD;
+	case ENGINE_ID_DIGE:
+		return DCN10_DIG_FE_SOURCE_SELECT_DIGE;
+	case ENGINE_ID_DIGF:
+		return DCN10_DIG_FE_SOURCE_SELECT_DIGF;
+	case ENGINE_ID_DIGG:
+		return DCN10_DIG_FE_SOURCE_SELECT_DIGG;
+	default:
+		ASSERT_CRITICAL(false);
+		return DCN10_DIG_FE_SOURCE_SELECT_INVALID;
+	}
+}
+
+void configure_encoder(
+	struct dcn10_link_encoder *enc10,
+	const struct dc_link_settings *link_settings)
+{
+	/* set number of lanes */
+	REG_SET(DP_CONFIG, 0,
+			DP_UDI_LANES, link_settings->lane_count - LANE_COUNT_ONE);
+
+	/* setup scrambler */
+	REG_UPDATE(DP_DPHY_SCRAM_CNTL, DPHY_SCRAMBLER_ADVANCE, 1);
+}
+
+void dcn10_psr_program_dp_dphy_fast_training(struct link_encoder *enc,
+			bool exit_link_training_required)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+	if (exit_link_training_required)
+		REG_UPDATE(DP_DPHY_FAST_TRAINING,
+				DPHY_RX_FAST_TRAINING_CAPABLE, 1);
+	else {
+		REG_UPDATE(DP_DPHY_FAST_TRAINING,
+				DPHY_RX_FAST_TRAINING_CAPABLE, 0);
+		/*In DCE 11, we are able to pre-program a Force SR register
+		 * to be able to trigger SR symbol after 5 idle patterns
+		 * transmitted. Upon PSR Exit, DMCU can trigger
+		 * DPHY_LOAD_BS_COUNT_START = 1. Upon writing 1 to
+		 * DPHY_LOAD_BS_COUNT_START and the internal counter
+		 * reaches DPHY_LOAD_BS_COUNT, the next BS symbol will be
+		 * replaced by SR symbol once.
+		 */
+
+		REG_UPDATE(DP_DPHY_BS_SR_SWAP_CNTL, DPHY_LOAD_BS_COUNT, 0x5);
+	}
+}
+
+void dcn10_psr_program_secondary_packet(struct link_encoder *enc,
+			unsigned int sdp_transmit_line_num_deadline)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+	REG_UPDATE_2(DP_SEC_CNTL1,
+		DP_SEC_GSP0_LINE_NUM, sdp_transmit_line_num_deadline,
+		DP_SEC_GSP0_PRIORITY, 1);
+}
+
+bool dcn10_is_dig_enabled(struct link_encoder *enc)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	uint32_t value;
+
+	REG_GET(DIG_BE_EN_CNTL, DIG_ENABLE, &value);
+	return value;
+}
+
+static void link_encoder_disable(struct dcn10_link_encoder *enc10)
+{
+	/* reset training pattern */
+	REG_SET(DP_DPHY_TRAINING_PATTERN_SEL, 0,
+			DPHY_TRAINING_PATTERN_SEL, 0);
+
+	/* reset training complete */
+	REG_UPDATE(DP_LINK_CNTL, DP_LINK_TRAINING_COMPLETE, 0);
+
+	/* reset panel mode */
+	setup_panel_mode(enc10, DP_PANEL_MODE_DEFAULT);
+}
+
+static void hpd_initialize(
+	struct dcn10_link_encoder *enc10)
+{
+	/* Associate HPD with DIG_BE */
+	enum hpd_source_id hpd_source = enc10->base.hpd_source;
+
+	REG_UPDATE(DIG_BE_CNTL, DIG_HPD_SELECT, hpd_source);
+}
+
+bool dcn10_link_encoder_validate_dvi_output(
+	const struct dcn10_link_encoder *enc10,
+	enum signal_type connector_signal,
+	enum signal_type signal,
+	const struct dc_crtc_timing *crtc_timing)
+{
+	uint32_t max_pixel_clock = TMDS_MAX_PIXEL_CLOCK;
+
+	if (signal == SIGNAL_TYPE_DVI_DUAL_LINK)
+		max_pixel_clock *= 2;
+
+	/* This handles the case of HDMI downgrade to DVI we don't want to
+	 * we don't want to cap the pixel clock if the DDI is not DVI.
+	 */
+	if (connector_signal != SIGNAL_TYPE_DVI_DUAL_LINK &&
+			connector_signal != SIGNAL_TYPE_DVI_SINGLE_LINK)
+		max_pixel_clock = enc10->base.features.max_hdmi_pixel_clock;
+
+	/* DVI only support RGB pixel encoding */
+	if (crtc_timing->pixel_encoding != PIXEL_ENCODING_RGB)
+		return false;
+
+	/*connect DVI via adpater's HDMI connector*/
+	if ((connector_signal == SIGNAL_TYPE_DVI_SINGLE_LINK ||
+		connector_signal == SIGNAL_TYPE_HDMI_TYPE_A) &&
+		signal != SIGNAL_TYPE_HDMI_TYPE_A &&
+		crtc_timing->pix_clk_khz > TMDS_MAX_PIXEL_CLOCK)
+		return false;
+	if (crtc_timing->pix_clk_khz < TMDS_MIN_PIXEL_CLOCK)
+		return false;
+
+	if (crtc_timing->pix_clk_khz > max_pixel_clock)
+		return false;
+
+	/* DVI supports 6/8bpp single-link and 10/16bpp dual-link */
+	switch (crtc_timing->display_color_depth) {
+	case COLOR_DEPTH_666:
+	case COLOR_DEPTH_888:
+	break;
+	case COLOR_DEPTH_101010:
+	case COLOR_DEPTH_161616:
+		if (signal != SIGNAL_TYPE_DVI_DUAL_LINK)
+			return false;
+	break;
+	default:
+		return false;
+	}
+
+	return true;
+}
+
+static bool dcn10_link_encoder_validate_hdmi_output(
+	const struct dcn10_link_encoder *enc10,
+	const struct dc_crtc_timing *crtc_timing,
+	int adjusted_pix_clk_khz)
+{
+	enum dc_color_depth max_deep_color =
+			enc10->base.features.max_hdmi_deep_color;
+
+	if (max_deep_color < crtc_timing->display_color_depth)
+		return false;
+
+	if (crtc_timing->display_color_depth < COLOR_DEPTH_888)
+		return false;
+	if (adjusted_pix_clk_khz < TMDS_MIN_PIXEL_CLOCK)
+		return false;
+
+	if ((adjusted_pix_clk_khz == 0) ||
+		(adjusted_pix_clk_khz > enc10->base.features.max_hdmi_pixel_clock))
+		return false;
+
+	/* DCE11 HW does not support 420 */
+	if (!enc10->base.features.ycbcr420_supported &&
+			crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR420)
+		return false;
+
+	if (!enc10->base.features.flags.bits.HDMI_6GB_EN &&
+		adjusted_pix_clk_khz >= 300000)
+		return false;
+	if (enc10->base.ctx->dc->debug.hdmi20_disable &&
+		crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR420)
+		return false;
+	return true;
+}
+
+bool dcn10_link_encoder_validate_dp_output(
+	const struct dcn10_link_encoder *enc10,
+	const struct dc_crtc_timing *crtc_timing)
+{
+	/* default RGB only */
+	if (crtc_timing->pixel_encoding == PIXEL_ENCODING_RGB)
+		return true;
+
+	if (enc10->base.features.flags.bits.IS_YCBCR_CAPABLE)
+		return true;
+
+	/* for DCE 8.x or later DP Y-only feature,
+	 * we need ASIC cap + FeatureSupportDPYonly, not support 666
+	 */
+	if (crtc_timing->flags.Y_ONLY &&
+		enc10->base.features.flags.bits.IS_YCBCR_CAPABLE &&
+		crtc_timing->display_color_depth != COLOR_DEPTH_666)
+		return true;
+
+	return false;
+}
+
+void dcn10_link_encoder_construct(
+	struct dcn10_link_encoder *enc10,
+	const struct encoder_init_data *init_data,
+	const struct encoder_feature_support *enc_features,
+	const struct dcn10_link_enc_registers *link_regs,
+	const struct dcn10_link_enc_aux_registers *aux_regs,
+	const struct dcn10_link_enc_hpd_registers *hpd_regs,
+	const struct dcn10_link_enc_shift *link_shift,
+	const struct dcn10_link_enc_mask *link_mask)
+{
+	struct bp_encoder_cap_info bp_cap_info = {0};
+	const struct dc_vbios_funcs *bp_funcs = init_data->ctx->dc_bios->funcs;
+	enum bp_result result = BP_RESULT_OK;
+
+	enc10->base.funcs = &dcn10_lnk_enc_funcs;
+	enc10->base.ctx = init_data->ctx;
+	enc10->base.id = init_data->encoder;
+
+	enc10->base.hpd_source = init_data->hpd_source;
+	enc10->base.connector = init_data->connector;
+
+	enc10->base.preferred_engine = ENGINE_ID_UNKNOWN;
+
+	enc10->base.features = *enc_features;
+
+	enc10->base.transmitter = init_data->transmitter;
+
+	/* set the flag to indicate whether driver poll the I2C data pin
+	 * while doing the DP sink detect
+	 */
+
+/*	if (dal_adapter_service_is_feature_supported(as,
+		FEATURE_DP_SINK_DETECT_POLL_DATA_PIN))
+		enc10->base.features.flags.bits.
+			DP_SINK_DETECT_POLL_DATA_PIN = true;*/
+
+	enc10->base.output_signals =
+		SIGNAL_TYPE_DVI_SINGLE_LINK |
+		SIGNAL_TYPE_DVI_DUAL_LINK |
+		SIGNAL_TYPE_LVDS |
+		SIGNAL_TYPE_DISPLAY_PORT |
+		SIGNAL_TYPE_DISPLAY_PORT_MST |
+		SIGNAL_TYPE_EDP |
+		SIGNAL_TYPE_HDMI_TYPE_A;
+
+	/* For DCE 8.0 and 8.1, by design, UNIPHY is hardwired to DIG_BE.
+	 * SW always assign DIG_FE 1:1 mapped to DIG_FE for non-MST UNIPHY.
+	 * SW assign DIG_FE to non-MST UNIPHY first and MST last. So prefer
+	 * DIG is per UNIPHY and used by SST DP, eDP, HDMI, DVI and LVDS.
+	 * Prefer DIG assignment is decided by board design.
+	 * For DCE 8.0, there are only max 6 UNIPHYs, we assume board design
+	 * and VBIOS will filter out 7 UNIPHY for DCE 8.0.
+	 * By this, adding DIGG should not hurt DCE 8.0.
+	 * This will let DCE 8.1 share DCE 8.0 as much as possible
+	 */
+
+	enc10->link_regs = link_regs;
+	enc10->aux_regs = aux_regs;
+	enc10->hpd_regs = hpd_regs;
+	enc10->link_shift = link_shift;
+	enc10->link_mask = link_mask;
+
+	switch (enc10->base.transmitter) {
+	case TRANSMITTER_UNIPHY_A:
+		enc10->base.preferred_engine = ENGINE_ID_DIGA;
+	break;
+	case TRANSMITTER_UNIPHY_B:
+		enc10->base.preferred_engine = ENGINE_ID_DIGB;
+	break;
+	case TRANSMITTER_UNIPHY_C:
+		enc10->base.preferred_engine = ENGINE_ID_DIGC;
+	break;
+	case TRANSMITTER_UNIPHY_D:
+		enc10->base.preferred_engine = ENGINE_ID_DIGD;
+	break;
+	case TRANSMITTER_UNIPHY_E:
+		enc10->base.preferred_engine = ENGINE_ID_DIGE;
+	break;
+	case TRANSMITTER_UNIPHY_F:
+		enc10->base.preferred_engine = ENGINE_ID_DIGF;
+	break;
+	case TRANSMITTER_UNIPHY_G:
+		enc10->base.preferred_engine = ENGINE_ID_DIGG;
+	break;
+	default:
+		ASSERT_CRITICAL(false);
+		enc10->base.preferred_engine = ENGINE_ID_UNKNOWN;
+	}
+
+	/* default to one to mirror Windows behavior */
+	enc10->base.features.flags.bits.HDMI_6GB_EN = 1;
+
+	result = bp_funcs->get_encoder_cap_info(enc10->base.ctx->dc_bios,
+						enc10->base.id, &bp_cap_info);
+
+	/* Override features with DCE-specific values */
+	if (result == BP_RESULT_OK) {
+		enc10->base.features.flags.bits.IS_HBR2_CAPABLE =
+				bp_cap_info.DP_HBR2_EN;
+		enc10->base.features.flags.bits.IS_HBR3_CAPABLE =
+				bp_cap_info.DP_HBR3_EN;
+		enc10->base.features.flags.bits.HDMI_6GB_EN = bp_cap_info.HDMI_6GB_EN;
+	} else {
+		DC_LOG_WARNING("%s: Failed to get encoder_cap_info from VBIOS with error code %d!\n",
+				__func__,
+				result);
+	}
+	if (enc10->base.ctx->dc->debug.hdmi20_disable) {
+		enc10->base.features.flags.bits.HDMI_6GB_EN = 0;
+	}
+}
+
+bool dcn10_link_encoder_validate_output_with_stream(
+	struct link_encoder *enc,
+	const struct dc_stream_state *stream)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	bool is_valid;
+
+	switch (stream->signal) {
+	case SIGNAL_TYPE_DVI_SINGLE_LINK:
+	case SIGNAL_TYPE_DVI_DUAL_LINK:
+		is_valid = dcn10_link_encoder_validate_dvi_output(
+			enc10,
+			stream->sink->link->connector_signal,
+			stream->signal,
+			&stream->timing);
+	break;
+	case SIGNAL_TYPE_HDMI_TYPE_A:
+		is_valid = dcn10_link_encoder_validate_hdmi_output(
+				enc10,
+				&stream->timing,
+				stream->phy_pix_clk);
+	break;
+	case SIGNAL_TYPE_DISPLAY_PORT:
+	case SIGNAL_TYPE_DISPLAY_PORT_MST:
+		is_valid = dcn10_link_encoder_validate_dp_output(
+					enc10, &stream->timing);
+	break;
+	case SIGNAL_TYPE_EDP:
+		is_valid = (stream->timing.pixel_encoding == PIXEL_ENCODING_RGB) ? true : false;
+	break;
+	case SIGNAL_TYPE_VIRTUAL:
+		is_valid = true;
+		break;
+	default:
+		is_valid = false;
+	break;
+	}
+
+	return is_valid;
+}
+
+void dcn10_link_encoder_hw_init(
+	struct link_encoder *enc)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	struct bp_transmitter_control cntl = { 0 };
+	enum bp_result result;
+
+	cntl.action = TRANSMITTER_CONTROL_INIT;
+	cntl.engine_id = ENGINE_ID_UNKNOWN;
+	cntl.transmitter = enc10->base.transmitter;
+	cntl.connector_obj_id = enc10->base.connector;
+	cntl.lanes_number = LANE_COUNT_FOUR;
+	cntl.coherent = false;
+	cntl.hpd_sel = enc10->base.hpd_source;
+
+	if (enc10->base.connector.id == CONNECTOR_ID_EDP)
+		cntl.signal = SIGNAL_TYPE_EDP;
+
+	result = link_transmitter_control(enc10, &cntl);
+
+	if (result != BP_RESULT_OK) {
+		DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n",
+			__func__);
+		BREAK_TO_DEBUGGER();
+		return;
+	}
+
+	if (enc10->base.connector.id == CONNECTOR_ID_LVDS) {
+		cntl.action = TRANSMITTER_CONTROL_BACKLIGHT_BRIGHTNESS;
+
+		result = link_transmitter_control(enc10, &cntl);
+
+		ASSERT(result == BP_RESULT_OK);
+
+	}
+	dcn10_aux_initialize(enc10);
+
+	/* reinitialize HPD.
+	 * hpd_initialize() will pass DIG_FE id to HW context.
+	 * All other routine within HW context will use fe_engine_offset
+	 * as DIG_FE id even caller pass DIG_FE id.
+	 * So this routine must be called first.
+	 */
+	hpd_initialize(enc10);
+}
+
+void dcn10_link_encoder_destroy(struct link_encoder **enc)
+{
+	kfree(TO_DCN10_LINK_ENC(*enc));
+	*enc = NULL;
+}
+
+void dcn10_link_encoder_setup(
+	struct link_encoder *enc,
+	enum signal_type signal)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+	switch (signal) {
+	case SIGNAL_TYPE_EDP:
+	case SIGNAL_TYPE_DISPLAY_PORT:
+		/* DP SST */
+		REG_UPDATE(DIG_BE_CNTL, DIG_MODE, 0);
+		break;
+	case SIGNAL_TYPE_LVDS:
+		/* LVDS */
+		REG_UPDATE(DIG_BE_CNTL, DIG_MODE, 1);
+		break;
+	case SIGNAL_TYPE_DVI_SINGLE_LINK:
+	case SIGNAL_TYPE_DVI_DUAL_LINK:
+		/* TMDS-DVI */
+		REG_UPDATE(DIG_BE_CNTL, DIG_MODE, 2);
+		break;
+	case SIGNAL_TYPE_HDMI_TYPE_A:
+		/* TMDS-HDMI */
+		REG_UPDATE(DIG_BE_CNTL, DIG_MODE, 3);
+		break;
+	case SIGNAL_TYPE_DISPLAY_PORT_MST:
+		/* DP MST */
+		REG_UPDATE(DIG_BE_CNTL, DIG_MODE, 5);
+		break;
+	default:
+		ASSERT_CRITICAL(false);
+		/* invalid mode ! */
+		break;
+	}
+
+}
+
+/* TODO: still need depth or just pass in adjusted pixel clock? */
+void dcn10_link_encoder_enable_tmds_output(
+	struct link_encoder *enc,
+	enum clock_source_id clock_source,
+	enum dc_color_depth color_depth,
+	enum signal_type signal,
+	uint32_t pixel_clock)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	struct bp_transmitter_control cntl = { 0 };
+	enum bp_result result;
+
+	/* Enable the PHY */
+
+	cntl.action = TRANSMITTER_CONTROL_ENABLE;
+	cntl.engine_id = enc->preferred_engine;
+	cntl.transmitter = enc10->base.transmitter;
+	cntl.pll_id = clock_source;
+	cntl.signal = signal;
+	if (cntl.signal == SIGNAL_TYPE_DVI_DUAL_LINK)
+		cntl.lanes_number = 8;
+	else
+		cntl.lanes_number = 4;
+
+	cntl.hpd_sel = enc10->base.hpd_source;
+
+	cntl.pixel_clock = pixel_clock;
+	cntl.color_depth = color_depth;
+
+	result = link_transmitter_control(enc10, &cntl);
+
+	if (result != BP_RESULT_OK) {
+		DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n",
+			__func__);
+		BREAK_TO_DEBUGGER();
+	}
+}
+
+/* enables DP PHY output */
+void dcn10_link_encoder_enable_dp_output(
+	struct link_encoder *enc,
+	const struct dc_link_settings *link_settings,
+	enum clock_source_id clock_source)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	struct bp_transmitter_control cntl = { 0 };
+	enum bp_result result;
+
+	/* Enable the PHY */
+
+	/* number_of_lanes is used for pixel clock adjust,
+	 * but it's not passed to asic_control.
+	 * We need to set number of lanes manually.
+	 */
+	configure_encoder(enc10, link_settings);
+
+	cntl.action = TRANSMITTER_CONTROL_ENABLE;
+	cntl.engine_id = enc->preferred_engine;
+	cntl.transmitter = enc10->base.transmitter;
+	cntl.pll_id = clock_source;
+	cntl.signal = SIGNAL_TYPE_DISPLAY_PORT;
+	cntl.lanes_number = link_settings->lane_count;
+	cntl.hpd_sel = enc10->base.hpd_source;
+	cntl.pixel_clock = link_settings->link_rate
+						* LINK_RATE_REF_FREQ_IN_KHZ;
+	/* TODO: check if undefined works */
+	cntl.color_depth = COLOR_DEPTH_UNDEFINED;
+
+	result = link_transmitter_control(enc10, &cntl);
+
+	if (result != BP_RESULT_OK) {
+		DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n",
+			__func__);
+		BREAK_TO_DEBUGGER();
+	}
+}
+
+/* enables DP PHY output in MST mode */
+void dcn10_link_encoder_enable_dp_mst_output(
+	struct link_encoder *enc,
+	const struct dc_link_settings *link_settings,
+	enum clock_source_id clock_source)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	struct bp_transmitter_control cntl = { 0 };
+	enum bp_result result;
+
+	/* Enable the PHY */
+
+	/* number_of_lanes is used for pixel clock adjust,
+	 * but it's not passed to asic_control.
+	 * We need to set number of lanes manually.
+	 */
+	configure_encoder(enc10, link_settings);
+
+	cntl.action = TRANSMITTER_CONTROL_ENABLE;
+	cntl.engine_id = ENGINE_ID_UNKNOWN;
+	cntl.transmitter = enc10->base.transmitter;
+	cntl.pll_id = clock_source;
+	cntl.signal = SIGNAL_TYPE_DISPLAY_PORT_MST;
+	cntl.lanes_number = link_settings->lane_count;
+	cntl.hpd_sel = enc10->base.hpd_source;
+	cntl.pixel_clock = link_settings->link_rate
+						* LINK_RATE_REF_FREQ_IN_KHZ;
+	/* TODO: check if undefined works */
+	cntl.color_depth = COLOR_DEPTH_UNDEFINED;
+
+	result = link_transmitter_control(enc10, &cntl);
+
+	if (result != BP_RESULT_OK) {
+		DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n",
+			__func__);
+		BREAK_TO_DEBUGGER();
+	}
+}
+/*
+ * @brief
+ * Disable transmitter and its encoder
+ */
+void dcn10_link_encoder_disable_output(
+	struct link_encoder *enc,
+	enum signal_type signal)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	struct bp_transmitter_control cntl = { 0 };
+	enum bp_result result;
+
+	if (!dcn10_is_dig_enabled(enc)) {
+		/* OF_SKIP_POWER_DOWN_INACTIVE_ENCODER */
+	/*in DP_Alt_No_Connect case, we turn off the dig already,
+	after excuation the PHY w/a sequence, not allow touch PHY any more*/
+		return;
+	}
+	/* Power-down RX and disable GPU PHY should be paired.
+	 * Disabling PHY without powering down RX may cause
+	 * symbol lock loss, on which we will get DP Sink interrupt.
+	 */
+
+	/* There is a case for the DP active dongles
+	 * where we want to disable the PHY but keep RX powered,
+	 * for those we need to ignore DP Sink interrupt
+	 * by checking lane count that has been set
+	 * on the last do_enable_output().
+	 */
+
+	/* disable transmitter */
+	cntl.action = TRANSMITTER_CONTROL_DISABLE;
+	cntl.transmitter = enc10->base.transmitter;
+	cntl.hpd_sel = enc10->base.hpd_source;
+	cntl.signal = signal;
+	cntl.connector_obj_id = enc10->base.connector;
+
+	result = link_transmitter_control(enc10, &cntl);
+
+	if (result != BP_RESULT_OK) {
+		DC_LOG_ERROR("%s: Failed to execute VBIOS command table!\n",
+			__func__);
+		BREAK_TO_DEBUGGER();
+		return;
+	}
+
+	/* disable encoder */
+	if (dc_is_dp_signal(signal))
+		link_encoder_disable(enc10);
+}
+
+void dcn10_link_encoder_dp_set_lane_settings(
+	struct link_encoder *enc,
+	const struct link_training_settings *link_settings)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	union dpcd_training_lane_set training_lane_set = { { 0 } };
+	int32_t lane = 0;
+	struct bp_transmitter_control cntl = { 0 };
+
+	if (!link_settings) {
+		BREAK_TO_DEBUGGER();
+		return;
+	}
+
+	cntl.action = TRANSMITTER_CONTROL_SET_VOLTAGE_AND_PREEMPASIS;
+	cntl.transmitter = enc10->base.transmitter;
+	cntl.connector_obj_id = enc10->base.connector;
+	cntl.lanes_number = link_settings->link_settings.lane_count;
+	cntl.hpd_sel = enc10->base.hpd_source;
+	cntl.pixel_clock = link_settings->link_settings.link_rate *
+						LINK_RATE_REF_FREQ_IN_KHZ;
+
+	for (lane = 0; lane < link_settings->link_settings.lane_count; lane++) {
+		/* translate lane settings */
+
+		training_lane_set.bits.VOLTAGE_SWING_SET =
+			link_settings->lane_settings[lane].VOLTAGE_SWING;
+		training_lane_set.bits.PRE_EMPHASIS_SET =
+			link_settings->lane_settings[lane].PRE_EMPHASIS;
+
+		/* post cursor 2 setting only applies to HBR2 link rate */
+		if (link_settings->link_settings.link_rate == LINK_RATE_HIGH2) {
+			/* this is passed to VBIOS
+			 * to program post cursor 2 level
+			 */
+			training_lane_set.bits.POST_CURSOR2_SET =
+				link_settings->lane_settings[lane].POST_CURSOR2;
+		}
+
+		cntl.lane_select = lane;
+		cntl.lane_settings = training_lane_set.raw;
+
+		/* call VBIOS table to set voltage swing and pre-emphasis */
+		link_transmitter_control(enc10, &cntl);
+	}
+}
+
+/* set DP PHY test and training patterns */
+void dcn10_link_encoder_dp_set_phy_pattern(
+	struct link_encoder *enc,
+	const struct encoder_set_dp_phy_pattern_param *param)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+	switch (param->dp_phy_pattern) {
+	case DP_TEST_PATTERN_TRAINING_PATTERN1:
+		dcn10_link_encoder_set_dp_phy_pattern_training_pattern(enc, 0);
+		break;
+	case DP_TEST_PATTERN_TRAINING_PATTERN2:
+		dcn10_link_encoder_set_dp_phy_pattern_training_pattern(enc, 1);
+		break;
+	case DP_TEST_PATTERN_TRAINING_PATTERN3:
+		dcn10_link_encoder_set_dp_phy_pattern_training_pattern(enc, 2);
+		break;
+	case DP_TEST_PATTERN_TRAINING_PATTERN4:
+		dcn10_link_encoder_set_dp_phy_pattern_training_pattern(enc, 3);
+		break;
+	case DP_TEST_PATTERN_D102:
+		set_dp_phy_pattern_d102(enc10);
+		break;
+	case DP_TEST_PATTERN_SYMBOL_ERROR:
+		set_dp_phy_pattern_symbol_error(enc10);
+		break;
+	case DP_TEST_PATTERN_PRBS7:
+		set_dp_phy_pattern_prbs7(enc10);
+		break;
+	case DP_TEST_PATTERN_80BIT_CUSTOM:
+		set_dp_phy_pattern_80bit_custom(
+			enc10, param->custom_pattern);
+		break;
+	case DP_TEST_PATTERN_CP2520_1:
+		set_dp_phy_pattern_hbr2_compliance_cp2520_2(enc10, 1);
+		break;
+	case DP_TEST_PATTERN_CP2520_2:
+		set_dp_phy_pattern_hbr2_compliance_cp2520_2(enc10, 2);
+		break;
+	case DP_TEST_PATTERN_CP2520_3:
+		set_dp_phy_pattern_hbr2_compliance_cp2520_2(enc10, 3);
+		break;
+	case DP_TEST_PATTERN_VIDEO_MODE: {
+		set_dp_phy_pattern_passthrough_mode(
+			enc10, param->dp_panel_mode);
+		break;
+	}
+
+	default:
+		/* invalid phy pattern */
+		ASSERT_CRITICAL(false);
+		break;
+	}
+}
+
+static void fill_stream_allocation_row_info(
+	const struct link_mst_stream_allocation *stream_allocation,
+	uint32_t *src,
+	uint32_t *slots)
+{
+	const struct stream_encoder *stream_enc = stream_allocation->stream_enc;
+
+	if (stream_enc) {
+		*src = stream_enc->id;
+		*slots = stream_allocation->slot_count;
+	} else {
+		*src = 0;
+		*slots = 0;
+	}
+}
+
+/* programs DP MST VC payload allocation */
+void dcn10_link_encoder_update_mst_stream_allocation_table(
+	struct link_encoder *enc,
+	const struct link_mst_stream_allocation_table *table)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	uint32_t value0 = 0;
+	uint32_t value1 = 0;
+	uint32_t value2 = 0;
+	uint32_t slots = 0;
+	uint32_t src = 0;
+	uint32_t retries = 0;
+
+	/* For CZ, there are only 3 pipes. So Virtual channel is up 3.*/
+
+	/* --- Set MSE Stream Attribute -
+	 * Setup VC Payload Table on Tx Side,
+	 * Issue allocation change trigger
+	 * to commit payload on both tx and rx side
+	 */
+
+	/* we should clean-up table each time */
+
+	if (table->stream_count >= 1) {
+		fill_stream_allocation_row_info(
+			&table->stream_allocations[0],
+			&src,
+			&slots);
+	} else {
+		src = 0;
+		slots = 0;
+	}
+
+	REG_UPDATE_2(DP_MSE_SAT0,
+			DP_MSE_SAT_SRC0, src,
+			DP_MSE_SAT_SLOT_COUNT0, slots);
+
+	if (table->stream_count >= 2) {
+		fill_stream_allocation_row_info(
+			&table->stream_allocations[1],
+			&src,
+			&slots);
+	} else {
+		src = 0;
+		slots = 0;
+	}
+
+	REG_UPDATE_2(DP_MSE_SAT0,
+			DP_MSE_SAT_SRC1, src,
+			DP_MSE_SAT_SLOT_COUNT1, slots);
+
+	if (table->stream_count >= 3) {
+		fill_stream_allocation_row_info(
+			&table->stream_allocations[2],
+			&src,
+			&slots);
+	} else {
+		src = 0;
+		slots = 0;
+	}
+
+	REG_UPDATE_2(DP_MSE_SAT1,
+			DP_MSE_SAT_SRC2, src,
+			DP_MSE_SAT_SLOT_COUNT2, slots);
+
+	if (table->stream_count >= 4) {
+		fill_stream_allocation_row_info(
+			&table->stream_allocations[3],
+			&src,
+			&slots);
+	} else {
+		src = 0;
+		slots = 0;
+	}
+
+	REG_UPDATE_2(DP_MSE_SAT1,
+			DP_MSE_SAT_SRC3, src,
+			DP_MSE_SAT_SLOT_COUNT3, slots);
+
+	/* --- wait for transaction finish */
+
+	/* send allocation change trigger (ACT) ?
+	 * this step first sends the ACT,
+	 * then double buffers the SAT into the hardware
+	 * making the new allocation active on the DP MST mode link
+	 */
+
+	/* DP_MSE_SAT_UPDATE:
+	 * 0 - No Action
+	 * 1 - Update SAT with trigger
+	 * 2 - Update SAT without trigger
+	 */
+	REG_UPDATE(DP_MSE_SAT_UPDATE,
+			DP_MSE_SAT_UPDATE, 1);
+
+	/* wait for update to complete
+	 * (i.e. DP_MSE_SAT_UPDATE field is reset to 0)
+	 * then wait for the transmission
+	 * of at least 16 MTP headers on immediate local link.
+	 * i.e. DP_MSE_16_MTP_KEEPOUT field (read only) is reset to 0
+	 * a value of 1 indicates that DP MST mode
+	 * is in the 16 MTP keepout region after a VC has been added.
+	 * MST stream bandwidth (VC rate) can be configured
+	 * after this bit is cleared
+	 */
+	do {
+		udelay(10);
+
+		value0 = REG_READ(DP_MSE_SAT_UPDATE);
+
+		REG_GET(DP_MSE_SAT_UPDATE,
+				DP_MSE_SAT_UPDATE, &value1);
+
+		REG_GET(DP_MSE_SAT_UPDATE,
+				DP_MSE_16_MTP_KEEPOUT, &value2);
+
+		/* bit field DP_MSE_SAT_UPDATE is set to 1 already */
+		if (!value1 && !value2)
+			break;
+		++retries;
+	} while (retries < DP_MST_UPDATE_MAX_RETRY);
+}
+
+void dcn10_link_encoder_connect_dig_be_to_fe(
+	struct link_encoder *enc,
+	enum engine_id engine,
+	bool connect)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+	uint32_t field;
+
+	if (engine != ENGINE_ID_UNKNOWN) {
+
+		REG_GET(DIG_BE_CNTL, DIG_FE_SOURCE_SELECT, &field);
+
+		if (connect)
+			field |= get_frontend_source(engine);
+		else
+			field &= ~get_frontend_source(engine);
+
+		REG_UPDATE(DIG_BE_CNTL, DIG_FE_SOURCE_SELECT, field);
+	}
+}
+
+
+#define HPD_REG(reg)\
+	(enc10->hpd_regs->reg)
+
+#define HPD_REG_READ(reg_name) \
+		dm_read_reg(CTX, HPD_REG(reg_name))
+
+#define HPD_REG_UPDATE_N(reg_name, n, ...)	\
+		generic_reg_update_ex(CTX, \
+				HPD_REG(reg_name), \
+				HPD_REG_READ(reg_name), \
+				n, __VA_ARGS__)
+
+#define HPD_REG_UPDATE(reg_name, field, val)	\
+		HPD_REG_UPDATE_N(reg_name, 1, \
+				FN(reg_name, field), val)
+
+void dcn10_link_encoder_enable_hpd(struct link_encoder *enc)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+	HPD_REG_UPDATE(DC_HPD_CONTROL,
+			DC_HPD_EN, 1);
+}
+
+void dcn10_link_encoder_disable_hpd(struct link_encoder *enc)
+{
+	struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+	HPD_REG_UPDATE(DC_HPD_CONTROL,
+			DC_HPD_EN, 0);
+}
+
+
+#define AUX_REG(reg)\
+	(enc10->aux_regs->reg)
+
+#define AUX_REG_READ(reg_name) \
+		dm_read_reg(CTX, AUX_REG(reg_name))
+
+#define AUX_REG_UPDATE_N(reg_name, n, ...)	\
+		generic_reg_update_ex(CTX, \
+				AUX_REG(reg_name), \
+				AUX_REG_READ(reg_name), \
+				n, __VA_ARGS__)
+
+#define AUX_REG_UPDATE(reg_name, field, val)	\
+		AUX_REG_UPDATE_N(reg_name, 1, \
+				FN(reg_name, field), val)
+
+#define AUX_REG_UPDATE_2(reg, f1, v1, f2, v2)	\
+		AUX_REG_UPDATE_N(reg, 2,\
+				FN(reg, f1), v1,\
+				FN(reg, f2), v2)
+
+void dcn10_aux_initialize(struct dcn10_link_encoder *enc10)
+{
+	enum hpd_source_id hpd_source = enc10->base.hpd_source;
+
+	AUX_REG_UPDATE_2(AUX_CONTROL,
+			AUX_HPD_SEL, hpd_source,
+			AUX_LS_READ_EN, 0);
+
+	/* 1/4 window (the maximum allowed) */
+	AUX_REG_UPDATE(AUX_DPHY_RX_CONTROL0,
+			AUX_RX_RECEIVE_WINDOW, 1);
+}
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h
new file mode 100644
index 000000000000..49ead12b2532
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_link_encoder.h
@@ -0,0 +1,341 @@
+/*
+ * Copyright 2012-15 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ *  and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __DC_LINK_ENCODER__DCN10_H__
+#define __DC_LINK_ENCODER__DCN10_H__
+
+#include "link_encoder.h"
+
+#define TO_DCN10_LINK_ENC(link_encoder)\
+	container_of(link_encoder, struct dcn10_link_encoder, base)
+
+
+#define AUX_REG_LIST(id)\
+	SRI(AUX_CONTROL, DP_AUX, id), \
+	SRI(AUX_DPHY_RX_CONTROL0, DP_AUX, id)
+
+#define HPD_REG_LIST(id)\
+	SRI(DC_HPD_CONTROL, HPD, id)
+
+#define LE_DCN_COMMON_REG_LIST(id) \
+	SRI(DIG_BE_CNTL, DIG, id), \
+	SRI(DIG_BE_EN_CNTL, DIG, id), \
+	SRI(TMDS_CTL_BITS, DIG, id), \
+	SRI(DP_CONFIG, DP, id), \
+	SRI(DP_DPHY_CNTL, DP, id), \
+	SRI(DP_DPHY_PRBS_CNTL, DP, id), \
+	SRI(DP_DPHY_SCRAM_CNTL, DP, id),\
+	SRI(DP_DPHY_SYM0, DP, id), \
+	SRI(DP_DPHY_SYM1, DP, id), \
+	SRI(DP_DPHY_SYM2, DP, id), \
+	SRI(DP_DPHY_TRAINING_PATTERN_SEL, DP, id), \
+	SRI(DP_LINK_CNTL, DP, id), \
+	SRI(DP_LINK_FRAMING_CNTL, DP, id), \
+	SRI(DP_MSE_SAT0, DP, id), \
+	SRI(DP_MSE_SAT1, DP, id), \
+	SRI(DP_MSE_SAT2, DP, id), \
+	SRI(DP_MSE_SAT_UPDATE, DP, id), \
+	SRI(DP_SEC_CNTL, DP, id), \
+	SRI(DP_VID_STREAM_CNTL, DP, id), \
+	SRI(DP_DPHY_FAST_TRAINING, DP, id), \
+	SRI(DP_SEC_CNTL1, DP, id), \
+	SRI(DP_DPHY_BS_SR_SWAP_CNTL, DP, id), \
+	SRI(DP_DPHY_INTERNAL_CTRL, DP, id), \
+	SRI(DP_DPHY_HBR2_PATTERN_CONTROL, DP, id)
+
+
+#define LE_DCN10_REG_LIST(id)\
+	LE_DCN_COMMON_REG_LIST(id)
+
+struct dcn10_link_enc_aux_registers {
+	uint32_t AUX_CONTROL;
+	uint32_t AUX_DPHY_RX_CONTROL0;
+};
+
+struct dcn10_link_enc_hpd_registers {
+	uint32_t DC_HPD_CONTROL;
+};
+
+struct dcn10_link_enc_registers {
+	uint32_t DIG_BE_CNTL;
+	uint32_t DIG_BE_EN_CNTL;
+	uint32_t DP_CONFIG;
+	uint32_t DP_DPHY_CNTL;
+	uint32_t DP_DPHY_INTERNAL_CTRL;
+	uint32_t DP_DPHY_PRBS_CNTL;
+	uint32_t DP_DPHY_SCRAM_CNTL;
+	uint32_t DP_DPHY_SYM0;
+	uint32_t DP_DPHY_SYM1;
+	uint32_t DP_DPHY_SYM2;
+	uint32_t DP_DPHY_TRAINING_PATTERN_SEL;
+	uint32_t DP_LINK_CNTL;
+	uint32_t DP_LINK_FRAMING_CNTL;
+	uint32_t DP_MSE_SAT0;
+	uint32_t DP_MSE_SAT1;
+	uint32_t DP_MSE_SAT2;
+	uint32_t DP_MSE_SAT_UPDATE;
+	uint32_t DP_SEC_CNTL;
+	uint32_t DP_VID_STREAM_CNTL;
+	uint32_t DP_DPHY_FAST_TRAINING;
+	uint32_t DP_DPHY_BS_SR_SWAP_CNTL;
+	uint32_t DP_DPHY_HBR2_PATTERN_CONTROL;
+	uint32_t DP_SEC_CNTL1;
+	uint32_t TMDS_CTL_BITS;
+};
+
+#define LE_SF(reg_name, field_name, post_fix)\
+	.field_name = reg_name ## __ ## field_name ## post_fix
+
+#define LINK_ENCODER_MASK_SH_LIST_DCN10(mask_sh)\
+	LE_SF(DIG0_DIG_BE_EN_CNTL, DIG_ENABLE, mask_sh),\
+	LE_SF(DIG0_DIG_BE_CNTL, DIG_HPD_SELECT, mask_sh),\
+	LE_SF(DIG0_DIG_BE_CNTL, DIG_MODE, mask_sh),\
+	LE_SF(DIG0_DIG_BE_CNTL, DIG_FE_SOURCE_SELECT, mask_sh),\
+	LE_SF(DIG0_TMDS_CTL_BITS, TMDS_CTL0, mask_sh), \
+	LE_SF(DP0_DP_DPHY_CNTL, DPHY_BYPASS, mask_sh),\
+	LE_SF(DP0_DP_DPHY_CNTL, DPHY_ATEST_SEL_LANE0, mask_sh),\
+	LE_SF(DP0_DP_DPHY_CNTL, DPHY_ATEST_SEL_LANE1, mask_sh),\
+	LE_SF(DP0_DP_DPHY_CNTL, DPHY_ATEST_SEL_LANE2, mask_sh),\
+	LE_SF(DP0_DP_DPHY_CNTL, DPHY_ATEST_SEL_LANE3, mask_sh),\
+	LE_SF(DP0_DP_DPHY_PRBS_CNTL, DPHY_PRBS_EN, mask_sh),\
+	LE_SF(DP0_DP_DPHY_PRBS_CNTL, DPHY_PRBS_SEL, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM0, DPHY_SYM1, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM0, DPHY_SYM2, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM0, DPHY_SYM3, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM1, DPHY_SYM4, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM1, DPHY_SYM5, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM1, DPHY_SYM6, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM2, DPHY_SYM7, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SYM2, DPHY_SYM8, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SCRAM_CNTL, DPHY_SCRAMBLER_BS_COUNT, mask_sh),\
+	LE_SF(DP0_DP_DPHY_SCRAM_CNTL, DPHY_SCRAMBLER_ADVANCE, mask_sh),\
+	LE_SF(DP0_DP_DPHY_FAST_TRAINING, DPHY_RX_FAST_TRAINING_CAPABLE, mask_sh),\
+	LE_SF(DP0_DP_DPHY_BS_SR_SWAP_CNTL, DPHY_LOAD_BS_COUNT, mask_sh),\
+	LE_SF(DP0_DP_DPHY_TRAINING_PATTERN_SEL, DPHY_TRAINING_PATTERN_SEL, mask_sh),\
+	LE_SF(DP0_DP_DPHY_HBR2_PATTERN_CONTROL, DP_DPHY_HBR2_PATTERN_CONTROL, mask_sh),\
+	LE_SF(DP0_DP_LINK_CNTL, DP_LINK_TRAINING_COMPLETE, mask_sh),\
+	LE_SF(DP0_DP_LINK_FRAMING_CNTL, DP_IDLE_BS_INTERVAL, mask_sh),\
+	LE_SF(DP0_DP_LINK_FRAMING_CNTL, DP_VBID_DISABLE, mask_sh),\
+	LE_SF(DP0_DP_LINK_FRAMING_CNTL, DP_VID_ENHANCED_FRAME_MODE, mask_sh),\
+	LE_SF(DP0_DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, mask_sh),\
+	LE_SF(DP0_DP_CONFIG, DP_UDI_LANES, mask_sh),\
+	LE_SF(DP0_DP_SEC_CNTL1, DP_SEC_GSP0_LINE_NUM, mask_sh),\
+	LE_SF(DP0_DP_SEC_CNTL1, DP_SEC_GSP0_PRIORITY, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT0, DP_MSE_SAT_SRC0, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT0, DP_MSE_SAT_SRC1, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT0, DP_MSE_SAT_SLOT_COUNT0, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT0, DP_MSE_SAT_SLOT_COUNT1, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT1, DP_MSE_SAT_SRC2, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT1, DP_MSE_SAT_SRC3, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT1, DP_MSE_SAT_SLOT_COUNT2, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT1, DP_MSE_SAT_SLOT_COUNT3, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT_UPDATE, DP_MSE_SAT_UPDATE, mask_sh),\
+	LE_SF(DP0_DP_MSE_SAT_UPDATE, DP_MSE_16_MTP_KEEPOUT, mask_sh),\
+	LE_SF(DP_AUX0_AUX_CONTROL, AUX_HPD_SEL, mask_sh),\
+	LE_SF(DP_AUX0_AUX_CONTROL, AUX_LS_READ_EN, mask_sh),\
+	LE_SF(DP_AUX0_AUX_DPHY_RX_CONTROL0, AUX_RX_RECEIVE_WINDOW, mask_sh),\
+	LE_SF(HPD0_DC_HPD_CONTROL, DC_HPD_EN, mask_sh)
+
+#define DCN_LINK_ENCODER_REG_FIELD_LIST(type) \
+	type DIG_ENABLE;\
+	type DIG_HPD_SELECT;\
+	type DIG_MODE;\
+	type DIG_FE_SOURCE_SELECT;\
+	type DPHY_BYPASS;\
+	type DPHY_ATEST_SEL_LANE0;\
+	type DPHY_ATEST_SEL_LANE1;\
+	type DPHY_ATEST_SEL_LANE2;\
+	type DPHY_ATEST_SEL_LANE3;\
+	type DPHY_PRBS_EN;\
+	type DPHY_PRBS_SEL;\
+	type DPHY_SYM1;\
+	type DPHY_SYM2;\
+	type DPHY_SYM3;\
+	type DPHY_SYM4;\
+	type DPHY_SYM5;\
+	type DPHY_SYM6;\
+	type DPHY_SYM7;\
+	type DPHY_SYM8;\
+	type DPHY_SCRAMBLER_BS_COUNT;\
+	type DPHY_SCRAMBLER_ADVANCE;\
+	type DPHY_RX_FAST_TRAINING_CAPABLE;\
+	type DPHY_LOAD_BS_COUNT;\
+	type DPHY_TRAINING_PATTERN_SEL;\
+	type DP_DPHY_HBR2_PATTERN_CONTROL;\
+	type DP_LINK_TRAINING_COMPLETE;\
+	type DP_IDLE_BS_INTERVAL;\
+	type DP_VBID_DISABLE;\
+	type DP_VID_ENHANCED_FRAME_MODE;\
+	type DP_VID_STREAM_ENABLE;\
+	type DP_UDI_LANES;\
+	type DP_SEC_GSP0_LINE_NUM;\
+	type DP_SEC_GSP0_PRIORITY;\
+	type DP_MSE_SAT_SRC0;\
+	type DP_MSE_SAT_SRC1;\
+	type DP_MSE_SAT_SRC2;\
+	type DP_MSE_SAT_SRC3;\
+	type DP_MSE_SAT_SLOT_COUNT0;\
+	type DP_MSE_SAT_SLOT_COUNT1;\
+	type DP_MSE_SAT_SLOT_COUNT2;\
+	type DP_MSE_SAT_SLOT_COUNT3;\
+	type DP_MSE_SAT_UPDATE;\
+	type DP_MSE_16_MTP_KEEPOUT;\
+	type DC_HPD_EN;\
+	type TMDS_CTL0;\
+	type AUX_HPD_SEL;\
+	type AUX_LS_READ_EN;\
+	type AUX_RX_RECEIVE_WINDOW
+
+struct dcn10_link_enc_shift {
+	DCN_LINK_ENCODER_REG_FIELD_LIST(uint8_t);
+};
+
+struct dcn10_link_enc_mask {
+	DCN_LINK_ENCODER_REG_FIELD_LIST(uint32_t);
+};
+
+struct dcn10_link_encoder {
+	struct link_encoder base;
+	const struct dcn10_link_enc_registers *link_regs;
+	const struct dcn10_link_enc_aux_registers *aux_regs;
+	const struct dcn10_link_enc_hpd_registers *hpd_regs;
+	const struct dcn10_link_enc_shift *link_shift;
+	const struct dcn10_link_enc_mask *link_mask;
+};
+
+
+void dcn10_link_encoder_construct(
+	struct dcn10_link_encoder *enc10,
+	const struct encoder_init_data *init_data,
+	const struct encoder_feature_support *enc_features,
+	const struct dcn10_link_enc_registers *link_regs,
+	const struct dcn10_link_enc_aux_registers *aux_regs,
+	const struct dcn10_link_enc_hpd_registers *hpd_regs,
+	const struct dcn10_link_enc_shift *link_shift,
+	const struct dcn10_link_enc_mask *link_mask);
+
+bool dcn10_link_encoder_validate_dvi_output(
+	const struct dcn10_link_encoder *enc10,
+	enum signal_type connector_signal,
+	enum signal_type signal,
+	const struct dc_crtc_timing *crtc_timing);
+
+bool dcn10_link_encoder_validate_rgb_output(
+	const struct dcn10_link_encoder *enc10,
+	const struct dc_crtc_timing *crtc_timing);
+
+bool dcn10_link_encoder_validate_dp_output(
+	const struct dcn10_link_encoder *enc10,
+	const struct dc_crtc_timing *crtc_timing);
+
+bool dcn10_link_encoder_validate_wireless_output(
+	const struct dcn10_link_encoder *enc10,
+	const struct dc_crtc_timing *crtc_timing);
+
+bool dcn10_link_encoder_validate_output_with_stream(
+	struct link_encoder *enc,
+	const struct dc_stream_state *stream);
+
+/****************** HW programming ************************/
+
+/* initialize HW */  /* why do we initialze aux in here? */
+void dcn10_link_encoder_hw_init(struct link_encoder *enc);
+
+void dcn10_link_encoder_destroy(struct link_encoder **enc);
+
+/* program DIG_MODE in DIG_BE */
+/* TODO can this be combined with enable_output? */
+void dcn10_link_encoder_setup(
+	struct link_encoder *enc,
+	enum signal_type signal);
+
+void configure_encoder(
+	struct dcn10_link_encoder *enc10,
+	const struct dc_link_settings *link_settings);
+
+/* enables TMDS PHY output */
+/* TODO: still need depth or just pass in adjusted pixel clock? */
+void dcn10_link_encoder_enable_tmds_output(
+	struct link_encoder *enc,
+	enum clock_source_id clock_source,
+	enum dc_color_depth color_depth,
+	enum signal_type signal,
+	uint32_t pixel_clock);
+
+/* enables DP PHY output */
+void dcn10_link_encoder_enable_dp_output(
+	struct link_encoder *enc,
+	const struct dc_link_settings *link_settings,
+	enum clock_source_id clock_source);
+
+/* enables DP PHY output in MST mode */
+void dcn10_link_encoder_enable_dp_mst_output(
+	struct link_encoder *enc,
+	const struct dc_link_settings *link_settings,
+	enum clock_source_id clock_source);
+
+/* disable PHY output */
+void dcn10_link_encoder_disable_output(
+	struct link_encoder *enc,
+	enum signal_type signal);
+
+/* set DP lane settings */
+void dcn10_link_encoder_dp_set_lane_settings(
+	struct link_encoder *enc,
+	const struct link_training_settings *link_settings);
+
+void dcn10_link_encoder_dp_set_phy_pattern(
+	struct link_encoder *enc,
+	const struct encoder_set_dp_phy_pattern_param *param);
+
+/* programs DP MST VC payload allocation */
+void dcn10_link_encoder_update_mst_stream_allocation_table(
+	struct link_encoder *enc,
+	const struct link_mst_stream_allocation_table *table);
+
+void dcn10_link_encoder_connect_dig_be_to_fe(
+	struct link_encoder *enc,
+	enum engine_id engine,
+	bool connect);
+
+void dcn10_link_encoder_set_dp_phy_pattern_training_pattern(
+	struct link_encoder *enc,
+	uint32_t index);
+
+void dcn10_link_encoder_enable_hpd(struct link_encoder *enc);
+
+void dcn10_link_encoder_disable_hpd(struct link_encoder *enc);
+
+void dcn10_psr_program_dp_dphy_fast_training(struct link_encoder *enc,
+			bool exit_link_training_required);
+
+void dcn10_psr_program_secondary_packet(struct link_encoder *enc,
+			unsigned int sdp_transmit_line_num_deadline);
+
+bool dcn10_is_dig_enabled(struct link_encoder *enc);
+
+void dcn10_aux_initialize(struct dcn10_link_encoder *enc10);
+
+#endif /* __DC_LINK_ENCODER__DCN10_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c
index 179890b1a8c4..958994edf2c4 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c
@@ -65,6 +65,7 @@ static void mpc1_update_blending(
 	int mpcc_id)
 {
 	struct dcn10_mpc *mpc10 = TO_DCN10_MPC(mpc);
+	struct mpcc *mpcc = mpc1_get_mpcc(mpc, mpcc_id);
 
 	REG_UPDATE_5(MPCC_CONTROL[mpcc_id],
 			MPCC_ALPHA_BLND_MODE,		blnd_cfg->alpha_mode,
@@ -74,6 +75,7 @@ static void mpc1_update_blending(
 			MPCC_GLOBAL_GAIN,		blnd_cfg->global_gain);
 
 	mpc1_set_bg_color(mpc, &blnd_cfg->black_color, mpcc_id);
+	mpcc->blnd_cfg = *blnd_cfg;
 }
 
 void mpc1_update_stereo_mix(
@@ -235,8 +237,7 @@ struct mpcc *mpc1_insert_plane(
 	}
 
 	/* update the blending configuration */
-	new_mpcc->blnd_cfg = *blnd_cfg;
-	mpc->funcs->update_blending(mpc, &new_mpcc->blnd_cfg, mpcc_id);
+	mpc->funcs->update_blending(mpc, blnd_cfg, mpcc_id);
 
 	/* update the stereo mix settings, if provided */
 	if (sm_cfg != NULL) {
@@ -409,7 +410,26 @@ void mpc1_init_mpcc_list_from_hw(
 	}
 }
 
-const struct mpc_funcs dcn10_mpc_funcs = {
+void mpc1_read_mpcc_state(
+		struct mpc *mpc,
+		int mpcc_inst,
+		struct mpcc_state *s)
+{
+	struct dcn10_mpc *mpc10 = TO_DCN10_MPC(mpc);
+
+	REG_GET(MPCC_OPP_ID[mpcc_inst], MPCC_OPP_ID, &s->opp_id);
+	REG_GET(MPCC_TOP_SEL[mpcc_inst], MPCC_TOP_SEL, &s->dpp_id);
+	REG_GET(MPCC_BOT_SEL[mpcc_inst], MPCC_BOT_SEL, &s->bot_mpcc_id);
+	REG_GET_4(MPCC_CONTROL[mpcc_inst], MPCC_MODE, &s->mode,
+			MPCC_ALPHA_BLND_MODE, &s->alpha_mode,
+			MPCC_ALPHA_MULTIPLIED_MODE, &s->pre_multiplied_alpha,
+			MPCC_BLND_ACTIVE_OVERLAP_ONLY, &s->overlap_only);
+	REG_GET_2(MPCC_STATUS[mpcc_inst], MPCC_IDLE, &s->idle,
+			MPCC_BUSY, &s->busy);
+}
+
+static const struct mpc_funcs dcn10_mpc_funcs = {
+	.read_mpcc_state = mpc1_read_mpcc_state,
 	.insert_plane = mpc1_insert_plane,
 	.remove_mpcc = mpc1_remove_mpcc,
 	.mpc_init = mpc1_mpc_init,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.h
index 267a2995ef6e..d3d16c4cbea3 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.h
@@ -183,4 +183,9 @@ struct mpcc *mpc1_get_mpcc_for_dpp(
 	struct mpc_tree *tree,
 	int dpp_id);
 
+void mpc1_read_mpcc_state(
+		struct mpc *mpc,
+		int mpcc_inst,
+		struct mpcc_state *s);
+
 #endif
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c
index 77a1a9d541a4..ab958cff3b76 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_opp.c
@@ -385,7 +385,7 @@ void opp1_destroy(struct output_pixel_processor **opp)
 	*opp = NULL;
 }
 
-static struct opp_funcs dcn10_opp_funcs = {
+static const struct opp_funcs dcn10_opp_funcs = {
 		.opp_set_dyn_expansion = opp1_set_dyn_expansion,
 		.opp_program_fmt = opp1_program_fmt,
 		.opp_program_bit_depth_reduction = opp1_program_bit_depth_reduction,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
index 4bf64d1b2c60..411f89218e01 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
@@ -93,6 +93,81 @@ static void optc1_disable_stereo(struct timing_generator *optc)
 		OTG_3D_STRUCTURE_STEREO_SEL_OVR, 0);
 }
 
+static uint32_t get_start_vline(struct timing_generator *optc, const struct dc_crtc_timing *dc_crtc_timing)
+{
+	struct dc_crtc_timing patched_crtc_timing;
+	int vesa_sync_start;
+	int asic_blank_end;
+	int interlace_factor;
+	int vertical_line_start;
+
+	patched_crtc_timing = *dc_crtc_timing;
+	optc1_apply_front_porch_workaround(optc, &patched_crtc_timing);
+
+	vesa_sync_start = patched_crtc_timing.h_addressable +
+			patched_crtc_timing.h_border_right +
+			patched_crtc_timing.h_front_porch;
+
+	asic_blank_end = patched_crtc_timing.h_total -
+			vesa_sync_start -
+			patched_crtc_timing.h_border_left;
+
+	interlace_factor = patched_crtc_timing.flags.INTERLACE ? 2 : 1;
+
+	vesa_sync_start = patched_crtc_timing.v_addressable +
+			patched_crtc_timing.v_border_bottom +
+			patched_crtc_timing.v_front_porch;
+
+	asic_blank_end = (patched_crtc_timing.v_total -
+			vesa_sync_start -
+			patched_crtc_timing.v_border_top)
+			* interlace_factor;
+
+	vertical_line_start = asic_blank_end - optc->dlg_otg_param.vstartup_start + 1;
+	if (vertical_line_start < 0) {
+		ASSERT(0);
+		vertical_line_start = 0;
+	}
+
+	return vertical_line_start;
+}
+
+void optc1_program_vline_interrupt(
+		struct timing_generator *optc,
+		const struct dc_crtc_timing *dc_crtc_timing,
+		unsigned long long vsync_delta)
+{
+
+	struct optc *optc1 = DCN10TG_FROM_TG(optc);
+
+	unsigned long long req_delta_tens_of_usec = div64_u64((vsync_delta + 9999), 10000);
+	unsigned long long pix_clk_hundreds_khz = div64_u64((dc_crtc_timing->pix_clk_khz + 99), 100);
+	uint32_t req_delta_lines = (uint32_t) div64_u64(
+			(req_delta_tens_of_usec * pix_clk_hundreds_khz + dc_crtc_timing->h_total - 1),
+								dc_crtc_timing->h_total);
+
+	uint32_t vsync_line = get_start_vline(optc, dc_crtc_timing);
+	uint32_t start_line = 0;
+	uint32_t endLine = 0;
+
+	if (req_delta_lines != 0)
+		req_delta_lines--;
+
+	if (req_delta_lines > vsync_line)
+		start_line = dc_crtc_timing->v_total - (req_delta_lines - vsync_line) - 1;
+	else
+		start_line = vsync_line - req_delta_lines;
+
+	endLine = start_line + 2;
+
+	if (endLine >= dc_crtc_timing->v_total)
+		endLine = 2;
+
+	REG_SET_2(OTG_VERTICAL_INTERRUPT0_POSITION, 0,
+			OTG_VERTICAL_INTERRUPT0_LINE_START, start_line,
+			OTG_VERTICAL_INTERRUPT0_LINE_END, endLine);
+}
+
 /**
  * program_timing_generator   used by mode timing set
  * Program CRTC Timing Registers - OTG_H_*, OTG_V_*, Pixel repetition.
@@ -285,7 +360,7 @@ void optc1_program_timing(
 
 }
 
-static void optc1_set_blank_data_double_buffer(struct timing_generator *optc, bool enable)
+void optc1_set_blank_data_double_buffer(struct timing_generator *optc, bool enable)
 {
 	struct optc *optc1 = DCN10TG_FROM_TG(optc);
 
@@ -780,17 +855,17 @@ void optc1_set_drr(
 				OTG_SET_V_TOTAL_MIN_MASK_EN, 0,
 				OTG_SET_V_TOTAL_MIN_MASK, 0);
 	} else {
-		REG_SET(OTG_V_TOTAL_MIN, 0,
-			OTG_V_TOTAL_MIN, 0);
-
-		REG_SET(OTG_V_TOTAL_MAX, 0,
-			OTG_V_TOTAL_MAX, 0);
-
 		REG_UPDATE_4(OTG_V_TOTAL_CONTROL,
 				OTG_SET_V_TOTAL_MIN_MASK, 0,
 				OTG_V_TOTAL_MIN_SEL, 0,
 				OTG_V_TOTAL_MAX_SEL, 0,
 				OTG_FORCE_LOCK_ON_EVENT, 0);
+
+		REG_SET(OTG_V_TOTAL_MIN, 0,
+			OTG_V_TOTAL_MIN, 0);
+
+		REG_SET(OTG_V_TOTAL_MAX, 0,
+			OTG_V_TOTAL_MAX, 0);
 	}
 }
 
@@ -1154,6 +1229,12 @@ void optc1_read_otg_state(struct optc *optc1,
 	REG_GET(OTG_V_TOTAL_MIN,
 			OTG_V_TOTAL_MIN, &s->v_total_min);
 
+	REG_GET(OTG_V_TOTAL_CONTROL,
+			OTG_V_TOTAL_MAX_SEL, &s->v_total_max_sel);
+
+	REG_GET(OTG_V_TOTAL_CONTROL,
+			OTG_V_TOTAL_MIN_SEL, &s->v_total_min_sel);
+
 	REG_GET_2(OTG_V_SYNC_A,
 			OTG_V_SYNC_A_START, &s->v_sync_a_start,
 			OTG_V_SYNC_A_END, &s->v_sync_a_end);
@@ -1176,20 +1257,51 @@ void optc1_read_otg_state(struct optc *optc1,
 			OPTC_UNDERFLOW_OCCURRED_STATUS, &s->underflow_occurred_status);
 }
 
-static void optc1_clear_optc_underflow(struct timing_generator *optc)
+bool optc1_get_otg_active_size(struct timing_generator *optc,
+		uint32_t *otg_active_width,
+		uint32_t *otg_active_height)
+{
+	uint32_t otg_enabled;
+	uint32_t v_blank_start;
+	uint32_t v_blank_end;
+	uint32_t h_blank_start;
+	uint32_t h_blank_end;
+	struct optc *optc1 = DCN10TG_FROM_TG(optc);
+
+
+	REG_GET(OTG_CONTROL,
+			OTG_MASTER_EN, &otg_enabled);
+
+	if (otg_enabled == 0)
+		return false;
+
+	REG_GET_2(OTG_V_BLANK_START_END,
+			OTG_V_BLANK_START, &v_blank_start,
+			OTG_V_BLANK_END, &v_blank_end);
+
+	REG_GET_2(OTG_H_BLANK_START_END,
+			OTG_H_BLANK_START, &h_blank_start,
+			OTG_H_BLANK_END, &h_blank_end);
+
+	*otg_active_width = v_blank_start - v_blank_end;
+	*otg_active_height = h_blank_start - h_blank_end;
+	return true;
+}
+
+void optc1_clear_optc_underflow(struct timing_generator *optc)
 {
 	struct optc *optc1 = DCN10TG_FROM_TG(optc);
 
 	REG_UPDATE(OPTC_INPUT_GLOBAL_CONTROL, OPTC_UNDERFLOW_CLEAR, 1);
 }
 
-static void optc1_tg_init(struct timing_generator *optc)
+void optc1_tg_init(struct timing_generator *optc)
 {
 	optc1_set_blank_data_double_buffer(optc, true);
 	optc1_clear_optc_underflow(optc);
 }
 
-static bool optc1_is_tg_enabled(struct timing_generator *optc)
+bool optc1_is_tg_enabled(struct timing_generator *optc)
 {
 	struct optc *optc1 = DCN10TG_FROM_TG(optc);
 	uint32_t otg_enabled = 0;
@@ -1200,7 +1312,7 @@ static bool optc1_is_tg_enabled(struct timing_generator *optc)
 
 }
 
-static bool optc1_is_optc_underflow_occurred(struct timing_generator *optc)
+bool optc1_is_optc_underflow_occurred(struct timing_generator *optc)
 {
 	struct optc *optc1 = DCN10TG_FROM_TG(optc);
 	uint32_t underflow_occurred = 0;
@@ -1212,9 +1324,76 @@ static bool optc1_is_optc_underflow_occurred(struct timing_generator *optc)
 	return (underflow_occurred == 1);
 }
 
+bool optc1_configure_crc(struct timing_generator *optc,
+			  const struct crc_params *params)
+{
+	struct optc *optc1 = DCN10TG_FROM_TG(optc);
+
+	/* Cannot configure crc on a CRTC that is disabled */
+	if (!optc1_is_tg_enabled(optc))
+		return false;
+
+	REG_WRITE(OTG_CRC_CNTL, 0);
+
+	if (!params->enable)
+		return true;
+
+	/* Program frame boundaries */
+	/* Window A x axis start and end. */
+	REG_UPDATE_2(OTG_CRC0_WINDOWA_X_CONTROL,
+			OTG_CRC0_WINDOWA_X_START, params->windowa_x_start,
+			OTG_CRC0_WINDOWA_X_END, params->windowa_x_end);
+
+	/* Window A y axis start and end. */
+	REG_UPDATE_2(OTG_CRC0_WINDOWA_Y_CONTROL,
+			OTG_CRC0_WINDOWA_Y_START, params->windowa_y_start,
+			OTG_CRC0_WINDOWA_Y_END, params->windowa_y_end);
+
+	/* Window B x axis start and end. */
+	REG_UPDATE_2(OTG_CRC0_WINDOWB_X_CONTROL,
+			OTG_CRC0_WINDOWB_X_START, params->windowb_x_start,
+			OTG_CRC0_WINDOWB_X_END, params->windowb_x_end);
+
+	/* Window B y axis start and end. */
+	REG_UPDATE_2(OTG_CRC0_WINDOWB_Y_CONTROL,
+			OTG_CRC0_WINDOWB_Y_START, params->windowb_y_start,
+			OTG_CRC0_WINDOWB_Y_END, params->windowb_y_end);
+
+	/* Set crc mode and selection, and enable. Only using CRC0*/
+	REG_UPDATE_3(OTG_CRC_CNTL,
+			OTG_CRC_CONT_EN, params->continuous_mode ? 1 : 0,
+			OTG_CRC0_SELECT, params->selection,
+			OTG_CRC_EN, 1);
+
+	return true;
+}
+
+bool optc1_get_crc(struct timing_generator *optc,
+		    uint32_t *r_cr, uint32_t *g_y, uint32_t *b_cb)
+{
+	uint32_t field = 0;
+	struct optc *optc1 = DCN10TG_FROM_TG(optc);
+
+	REG_GET(OTG_CRC_CNTL, OTG_CRC_EN, &field);
+
+	/* Early return if CRC is not enabled for this CRTC */
+	if (!field)
+		return false;
+
+	REG_GET_2(OTG_CRC0_DATA_RG,
+			CRC0_R_CR, r_cr,
+			CRC0_G_Y, g_y);
+
+	REG_GET(OTG_CRC0_DATA_B,
+			CRC0_B_CB, b_cb);
+
+	return true;
+}
+
 static const struct timing_generator_funcs dcn10_tg_funcs = {
 		.validate_timing = optc1_validate_timing,
 		.program_timing = optc1_program_timing,
+		.program_vline_interrupt = optc1_program_vline_interrupt,
 		.program_global_sync = optc1_program_global_sync,
 		.enable_crtc = optc1_enable_crtc,
 		.disable_crtc = optc1_disable_crtc,
@@ -1223,6 +1402,7 @@ static const struct timing_generator_funcs dcn10_tg_funcs = {
 		.get_position = optc1_get_position,
 		.get_frame_count = optc1_get_vblank_counter,
 		.get_scanoutpos = optc1_get_crtc_scanoutpos,
+		.get_otg_active_size = optc1_get_otg_active_size,
 		.set_early_control = optc1_set_early_control,
 		/* used by enable_timing_synchronization. Not need for FPGA */
 		.wait_for_state = optc1_wait_for_state,
@@ -1246,6 +1426,8 @@ static const struct timing_generator_funcs dcn10_tg_funcs = {
 		.is_tg_enabled = optc1_is_tg_enabled,
 		.is_optc_underflow_occurred = optc1_is_optc_underflow_occurred,
 		.clear_optc_underflow = optc1_clear_optc_underflow,
+		.get_crc = optc1_get_crc,
+		.configure_crc = optc1_configure_crc,
 };
 
 void dcn10_timing_generator_init(struct optc *optc1)
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
index d25e7bf0d0d7..c1b114209fe8 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
@@ -65,6 +65,8 @@
 	SRI(OTG_NOM_VERT_POSITION, OTG, inst),\
 	SRI(OTG_BLACK_COLOR, OTG, inst),\
 	SRI(OTG_CLOCK_CONTROL, OTG, inst),\
+	SRI(OTG_VERTICAL_INTERRUPT0_CONTROL, OTG, inst),\
+	SRI(OTG_VERTICAL_INTERRUPT0_POSITION, OTG, inst),\
 	SRI(OTG_VERTICAL_INTERRUPT2_CONTROL, OTG, inst),\
 	SRI(OTG_VERTICAL_INTERRUPT2_POSITION, OTG, inst),\
 	SRI(OPTC_INPUT_CLOCK_CONTROL, ODM, inst),\
@@ -73,7 +75,14 @@
 	SRI(CONTROL, VTG, inst),\
 	SRI(OTG_VERT_SYNC_CONTROL, OTG, inst),\
 	SRI(OTG_MASTER_UPDATE_MODE, OTG, inst),\
-	SRI(OTG_GSL_CONTROL, OTG, inst)
+	SRI(OTG_GSL_CONTROL, OTG, inst),\
+	SRI(OTG_CRC_CNTL, OTG, inst),\
+	SRI(OTG_CRC0_DATA_RG, OTG, inst),\
+	SRI(OTG_CRC0_DATA_B, OTG, inst),\
+	SRI(OTG_CRC0_WINDOWA_X_CONTROL, OTG, inst),\
+	SRI(OTG_CRC0_WINDOWA_Y_CONTROL, OTG, inst),\
+	SRI(OTG_CRC0_WINDOWB_X_CONTROL, OTG, inst),\
+	SRI(OTG_CRC0_WINDOWB_Y_CONTROL, OTG, inst)
 
 #define TG_COMMON_REG_LIST_DCN1_0(inst) \
 	TG_COMMON_REG_LIST_DCN(inst),\
@@ -124,6 +133,8 @@ struct dcn_optc_registers {
 	uint32_t OTG_TEST_PATTERN_CONTROL;
 	uint32_t OTG_TEST_PATTERN_COLOR;
 	uint32_t OTG_CLOCK_CONTROL;
+	uint32_t OTG_VERTICAL_INTERRUPT0_CONTROL;
+	uint32_t OTG_VERTICAL_INTERRUPT0_POSITION;
 	uint32_t OTG_VERTICAL_INTERRUPT2_CONTROL;
 	uint32_t OTG_VERTICAL_INTERRUPT2_POSITION;
 	uint32_t OPTC_INPUT_CLOCK_CONTROL;
@@ -134,6 +145,13 @@ struct dcn_optc_registers {
 	uint32_t OTG_GSL_WINDOW_X;
 	uint32_t OTG_GSL_WINDOW_Y;
 	uint32_t OTG_VUPDATE_KEEPOUT;
+	uint32_t OTG_CRC_CNTL;
+	uint32_t OTG_CRC0_DATA_RG;
+	uint32_t OTG_CRC0_DATA_B;
+	uint32_t OTG_CRC0_WINDOWA_X_CONTROL;
+	uint32_t OTG_CRC0_WINDOWA_Y_CONTROL;
+	uint32_t OTG_CRC0_WINDOWB_X_CONTROL;
+	uint32_t OTG_CRC0_WINDOWB_Y_CONTROL;
 };
 
 #define TG_COMMON_MASK_SH_LIST_DCN(mask_sh)\
@@ -206,6 +224,9 @@ struct dcn_optc_registers {
 	SF(OTG0_OTG_CLOCK_CONTROL, OTG_CLOCK_EN, mask_sh),\
 	SF(OTG0_OTG_CLOCK_CONTROL, OTG_CLOCK_ON, mask_sh),\
 	SF(OTG0_OTG_CLOCK_CONTROL, OTG_CLOCK_GATE_DIS, mask_sh),\
+	SF(OTG0_OTG_VERTICAL_INTERRUPT0_CONTROL, OTG_VERTICAL_INTERRUPT0_INT_ENABLE, mask_sh),\
+	SF(OTG0_OTG_VERTICAL_INTERRUPT0_POSITION, OTG_VERTICAL_INTERRUPT0_LINE_START, mask_sh),\
+	SF(OTG0_OTG_VERTICAL_INTERRUPT0_POSITION, OTG_VERTICAL_INTERRUPT0_LINE_END, mask_sh),\
 	SF(OTG0_OTG_VERTICAL_INTERRUPT2_CONTROL, OTG_VERTICAL_INTERRUPT2_INT_ENABLE, mask_sh),\
 	SF(OTG0_OTG_VERTICAL_INTERRUPT2_POSITION, OTG_VERTICAL_INTERRUPT2_LINE_START, mask_sh),\
 	SF(ODM0_OPTC_INPUT_CLOCK_CONTROL, OPTC_INPUT_CLK_EN, mask_sh),\
@@ -225,7 +246,21 @@ struct dcn_optc_registers {
 	SF(OTG0_OTG_GSL_CONTROL, OTG_GSL2_EN, mask_sh),\
 	SF(OTG0_OTG_GSL_CONTROL, OTG_GSL_MASTER_EN, mask_sh),\
 	SF(OTG0_OTG_GSL_CONTROL, OTG_GSL_FORCE_DELAY, mask_sh),\
-	SF(OTG0_OTG_GSL_CONTROL, OTG_GSL_CHECK_ALL_FIELDS, mask_sh)
+	SF(OTG0_OTG_GSL_CONTROL, OTG_GSL_CHECK_ALL_FIELDS, mask_sh),\
+	SF(OTG0_OTG_CRC_CNTL, OTG_CRC_CONT_EN, mask_sh),\
+	SF(OTG0_OTG_CRC_CNTL, OTG_CRC0_SELECT, mask_sh),\
+	SF(OTG0_OTG_CRC_CNTL, OTG_CRC_EN, mask_sh),\
+	SF(OTG0_OTG_CRC0_DATA_RG, CRC0_R_CR, mask_sh),\
+	SF(OTG0_OTG_CRC0_DATA_RG, CRC0_G_Y, mask_sh),\
+	SF(OTG0_OTG_CRC0_DATA_B, CRC0_B_CB, mask_sh),\
+	SF(OTG0_OTG_CRC0_WINDOWA_X_CONTROL, OTG_CRC0_WINDOWA_X_START, mask_sh),\
+	SF(OTG0_OTG_CRC0_WINDOWA_X_CONTROL, OTG_CRC0_WINDOWA_X_END, mask_sh),\
+	SF(OTG0_OTG_CRC0_WINDOWA_Y_CONTROL, OTG_CRC0_WINDOWA_Y_START, mask_sh),\
+	SF(OTG0_OTG_CRC0_WINDOWA_Y_CONTROL, OTG_CRC0_WINDOWA_Y_END, mask_sh),\
+	SF(OTG0_OTG_CRC0_WINDOWB_X_CONTROL, OTG_CRC0_WINDOWB_X_START, mask_sh),\
+	SF(OTG0_OTG_CRC0_WINDOWB_X_CONTROL, OTG_CRC0_WINDOWB_X_END, mask_sh),\
+	SF(OTG0_OTG_CRC0_WINDOWB_Y_CONTROL, OTG_CRC0_WINDOWB_Y_START, mask_sh),\
+	SF(OTG0_OTG_CRC0_WINDOWB_Y_CONTROL, OTG_CRC0_WINDOWB_Y_END, mask_sh)
 
 
 #define TG_COMMON_MASK_SH_LIST_DCN1_0(mask_sh)\
@@ -323,6 +358,9 @@ struct dcn_optc_registers {
 	type OTG_CLOCK_EN;\
 	type OTG_CLOCK_ON;\
 	type OTG_CLOCK_GATE_DIS;\
+	type OTG_VERTICAL_INTERRUPT0_INT_ENABLE;\
+	type OTG_VERTICAL_INTERRUPT0_LINE_START;\
+	type OTG_VERTICAL_INTERRUPT0_LINE_END;\
 	type OTG_VERTICAL_INTERRUPT2_INT_ENABLE;\
 	type OTG_VERTICAL_INTERRUPT2_LINE_START;\
 	type OPTC_INPUT_CLK_EN;\
@@ -353,7 +391,22 @@ struct dcn_optc_registers {
 	type OTG_MASTER_UPDATE_LOCK_GSL_EN;\
 	type MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_START_OFFSET;\
 	type MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_END_OFFSET;\
-	type OTG_MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_EN;
+	type OTG_MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_EN;\
+	type OTG_CRC_CONT_EN;\
+	type OTG_CRC0_SELECT;\
+	type OTG_CRC_EN;\
+	type CRC0_R_CR;\
+	type CRC0_G_Y;\
+	type CRC0_B_CB;\
+	type OTG_CRC0_WINDOWA_X_START;\
+	type OTG_CRC0_WINDOWA_X_END;\
+	type OTG_CRC0_WINDOWA_Y_START;\
+	type OTG_CRC0_WINDOWA_Y_END;\
+	type OTG_CRC0_WINDOWB_X_START;\
+	type OTG_CRC0_WINDOWB_X_END;\
+	type OTG_CRC0_WINDOWB_Y_START;\
+	type OTG_CRC0_WINDOWB_Y_END;
+
 
 #define TG_REG_FIELD_LIST(type) \
 	TG_REG_FIELD_LIST_DCN1_0(type)
@@ -396,6 +449,8 @@ struct dcn_otg_state {
 	uint32_t v_total;
 	uint32_t v_total_max;
 	uint32_t v_total_min;
+	uint32_t v_total_min_sel;
+	uint32_t v_total_max_sel;
 	uint32_t v_sync_a_start;
 	uint32_t v_sync_a_end;
 	uint32_t h_blank_start;
@@ -420,6 +475,10 @@ void optc1_program_timing(
 	const struct dc_crtc_timing *dc_crtc_timing,
 	bool use_vbios);
 
+void optc1_program_vline_interrupt(struct timing_generator *optc,
+		const struct dc_crtc_timing *dc_crtc_timing,
+		unsigned long long vsync_delta);
+
 void optc1_program_global_sync(
 		struct timing_generator *optc);
 
@@ -481,4 +540,29 @@ void optc1_program_stereo(struct timing_generator *optc,
 
 bool optc1_is_stereo_left_eye(struct timing_generator *optc);
 
+void optc1_clear_optc_underflow(struct timing_generator *optc);
+
+void optc1_tg_init(struct timing_generator *optc);
+
+bool optc1_is_tg_enabled(struct timing_generator *optc);
+
+bool optc1_is_optc_underflow_occurred(struct timing_generator *optc);
+
+void optc1_set_blank_data_double_buffer(struct timing_generator *optc, bool enable);
+
+bool optc1_get_otg_active_size(struct timing_generator *optc,
+		uint32_t *otg_active_width,
+		uint32_t *otg_active_height);
+
+void optc1_enable_crtc_reset(
+		struct timing_generator *optc,
+		int source_tg_inst,
+		struct crtc_trigger_info *crtc_tp);
+
+bool optc1_configure_crc(struct timing_generator *optc,
+			  const struct crc_params *params);
+
+bool optc1_get_crc(struct timing_generator *optc,
+		    uint32_t *r_cr, uint32_t *g_y, uint32_t *b_cb);
+
 #endif /* __DC_TIMING_GENERATOR_DCN10_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
index 02bd664aed3e..84581b3c392b 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c
@@ -38,8 +38,8 @@
 #include "dcn10/dcn10_hw_sequencer.h"
 #include "dce110/dce110_hw_sequencer.h"
 #include "dcn10/dcn10_opp.h"
-#include "dce/dce_link_encoder.h"
-#include "dce/dce_stream_encoder.h"
+#include "dcn10/dcn10_link_encoder.h"
+#include "dcn10/dcn10_stream_encoder.h"
 #include "dce/dce_clocks.h"
 #include "dce/dce_clock_source.h"
 #include "dce/dce_audio.h"
@@ -65,6 +65,68 @@
 #include "dce/dce_abm.h"
 #include "dce/dce_dmcu.h"
 
+const struct _vcs_dpi_ip_params_st dcn1_0_ip = {
+	.rob_buffer_size_kbytes = 64,
+	.det_buffer_size_kbytes = 164,
+	.dpte_buffer_size_in_pte_reqs = 42,
+	.dpp_output_buffer_pixels = 2560,
+	.opp_output_buffer_lines = 1,
+	.pixel_chunk_size_kbytes = 8,
+	.pte_enable = 1,
+	.pte_chunk_size_kbytes = 2,
+	.meta_chunk_size_kbytes = 2,
+	.writeback_chunk_size_kbytes = 2,
+	.line_buffer_size_bits = 589824,
+	.max_line_buffer_lines = 12,
+	.IsLineBufferBppFixed = 0,
+	.LineBufferFixedBpp = -1,
+	.writeback_luma_buffer_size_kbytes = 12,
+	.writeback_chroma_buffer_size_kbytes = 8,
+	.max_num_dpp = 4,
+	.max_num_wb = 2,
+	.max_dchub_pscl_bw_pix_per_clk = 4,
+	.max_pscl_lb_bw_pix_per_clk = 2,
+	.max_lb_vscl_bw_pix_per_clk = 4,
+	.max_vscl_hscl_bw_pix_per_clk = 4,
+	.max_hscl_ratio = 4,
+	.max_vscl_ratio = 4,
+	.hscl_mults = 4,
+	.vscl_mults = 4,
+	.max_hscl_taps = 8,
+	.max_vscl_taps = 8,
+	.dispclk_ramp_margin_percent = 1,
+	.underscan_factor = 1.10,
+	.min_vblank_lines = 14,
+	.dppclk_delay_subtotal = 90,
+	.dispclk_delay_subtotal = 42,
+	.dcfclk_cstate_latency = 10,
+	.max_inter_dcn_tile_repeaters = 8,
+	.can_vstartup_lines_exceed_vsync_plus_back_porch_lines_minus_one = 0,
+	.bug_forcing_LC_req_same_size_fixed = 0,
+};
+
+const struct _vcs_dpi_soc_bounding_box_st dcn1_0_soc = {
+	.sr_exit_time_us = 9.0,
+	.sr_enter_plus_exit_time_us = 11.0,
+	.urgent_latency_us = 4.0,
+	.writeback_latency_us = 12.0,
+	.ideal_dram_bw_after_urgent_percent = 80.0,
+	.max_request_size_bytes = 256,
+	.downspread_percent = 0.5,
+	.dram_page_open_time_ns = 50.0,
+	.dram_rw_turnaround_time_ns = 17.5,
+	.dram_return_buffer_per_channel_bytes = 8192,
+	.round_trip_ping_latency_dcfclk_cycles = 128,
+	.urgent_out_of_order_return_per_channel_bytes = 256,
+	.channel_interleave_bytes = 256,
+	.num_banks = 8,
+	.num_chans = 2,
+	.vmm_page_size_bytes = 4096,
+	.dram_clock_change_latency_us = 17.0,
+	.writeback_dram_clock_change_latency_us = 23.0,
+	.return_bus_width_bytes = 64,
+};
+
 #ifndef mmDP0_DP_DPHY_INTERNAL_CTRL
 	#define mmDP0_DP_DPHY_INTERNAL_CTRL		0x210f
 	#define mmDP0_DP_DPHY_INTERNAL_CTRL_BASE_IDX	2
@@ -166,36 +228,22 @@ static const struct dce_abm_mask abm_mask = {
 
 #define stream_enc_regs(id)\
 [id] = {\
-	SE_DCN_REG_LIST(id),\
-	.TMDS_CNTL = 0,\
-	.AFMT_AVI_INFO0 = 0,\
-	.AFMT_AVI_INFO1 = 0,\
-	.AFMT_AVI_INFO2 = 0,\
-	.AFMT_AVI_INFO3 = 0,\
+	SE_DCN_REG_LIST(id)\
 }
 
-static const struct dce110_stream_enc_registers stream_enc_regs[] = {
+static const struct dcn10_stream_enc_registers stream_enc_regs[] = {
 	stream_enc_regs(0),
 	stream_enc_regs(1),
 	stream_enc_regs(2),
 	stream_enc_regs(3),
 };
 
-static const struct dce_stream_encoder_shift se_shift = {
+static const struct dcn10_stream_encoder_shift se_shift = {
 		SE_COMMON_MASK_SH_LIST_DCN10(__SHIFT)
 };
 
-static const struct dce_stream_encoder_mask se_mask = {
-		SE_COMMON_MASK_SH_LIST_DCN10(_MASK),
-		.AFMT_GENERIC0_UPDATE = 0,
-		.AFMT_GENERIC2_UPDATE = 0,
-		.DP_DYN_RANGE = 0,
-		.DP_YCBCR_RANGE = 0,
-		.HDMI_AVI_INFO_SEND = 0,
-		.HDMI_AVI_INFO_CONT = 0,
-		.HDMI_AVI_INFO_LINE = 0,
-		.DP_SEC_AVI_ENABLE = 0,
-		.AFMT_AVI_INFO_VERSION = 0
+static const struct dcn10_stream_encoder_mask se_mask = {
+		SE_COMMON_MASK_SH_LIST_DCN10(_MASK)
 };
 
 #define audio_regs(id)\
@@ -228,13 +276,11 @@ static const struct dce_aduio_mask audio_mask = {
 	AUX_REG_LIST(id)\
 }
 
-static const struct dce110_link_enc_aux_registers link_enc_aux_regs[] = {
+static const struct dcn10_link_enc_aux_registers link_enc_aux_regs[] = {
 		aux_regs(0),
 		aux_regs(1),
 		aux_regs(2),
-		aux_regs(3),
-		aux_regs(4),
-		aux_regs(5)
+		aux_regs(3)
 };
 
 #define hpd_regs(id)\
@@ -242,13 +288,11 @@ static const struct dce110_link_enc_aux_registers link_enc_aux_regs[] = {
 	HPD_REG_LIST(id)\
 }
 
-static const struct dce110_link_enc_hpd_registers link_enc_hpd_regs[] = {
+static const struct dcn10_link_enc_hpd_registers link_enc_hpd_regs[] = {
 		hpd_regs(0),
 		hpd_regs(1),
 		hpd_regs(2),
-		hpd_regs(3),
-		hpd_regs(4),
-		hpd_regs(5)
+		hpd_regs(3)
 };
 
 #define link_regs(id)\
@@ -257,14 +301,19 @@ static const struct dce110_link_enc_hpd_registers link_enc_hpd_regs[] = {
 	SRI(DP_DPHY_INTERNAL_CTRL, DP, id) \
 }
 
-static const struct dce110_link_enc_registers link_enc_regs[] = {
+static const struct dcn10_link_enc_registers link_enc_regs[] = {
 	link_regs(0),
 	link_regs(1),
 	link_regs(2),
-	link_regs(3),
-	link_regs(4),
-	link_regs(5),
-	link_regs(6),
+	link_regs(3)
+};
+
+static const struct dcn10_link_enc_shift le_shift = {
+		LINK_ENCODER_MASK_SH_LIST_DCN10(__SHIFT)
+};
+
+static const struct dcn10_link_enc_mask le_mask = {
+		LINK_ENCODER_MASK_SH_LIST_DCN10(_MASK)
 };
 
 #define ipp_regs(id)\
@@ -320,11 +369,14 @@ static const struct dcn_dpp_registers tf_regs[] = {
 };
 
 static const struct dcn_dpp_shift tf_shift = {
-	TF_REG_LIST_SH_MASK_DCN10(__SHIFT)
+	TF_REG_LIST_SH_MASK_DCN10(__SHIFT),
+	TF_DEBUG_REG_LIST_SH_DCN10
+
 };
 
 static const struct dcn_dpp_mask tf_mask = {
 	TF_REG_LIST_SH_MASK_DCN10(_MASK),
+	TF_DEBUG_REG_LIST_MASK_DCN10
 };
 
 static const struct dcn_mpc_registers mpc_regs = {
@@ -427,6 +479,7 @@ static const struct dce110_clk_src_mask cs_mask = {
 
 static const struct resource_caps res_cap = {
 		.num_timing_generator = 4,
+		.num_opp = 4,
 		.num_video_plane = 4,
 		.num_audio = 4,
 		.num_stream_encoder = 4,
@@ -446,7 +499,7 @@ static const struct dc_debug debug_defaults_drv = {
 		 */
 		.min_disp_clk_khz = 100000,
 
-		.disable_pplib_clock_request = true,
+		.disable_pplib_clock_request = false,
 		.disable_pplib_wm_range = false,
 		.pplib_wm_report_mode = WM_REPORT_DEFAULT,
 		.pipe_split_policy = MPC_SPLIT_AVOID_MULT_DISP,
@@ -457,6 +510,8 @@ static const struct dc_debug debug_defaults_drv = {
 		.vsr_support = true,
 		.performance_trace = false,
 		.az_endpoint_mute_only = true,
+		.recovery_enabled = false, /*enable this by default after testing.*/
+		.max_downscale_src_width = 3840,
 };
 
 static const struct dc_debug debug_defaults_diags = {
@@ -592,20 +647,22 @@ static const struct encoder_feature_support link_enc_feature = {
 struct link_encoder *dcn10_link_encoder_create(
 	const struct encoder_init_data *enc_init_data)
 {
-	struct dce110_link_encoder *enc110 =
-		kzalloc(sizeof(struct dce110_link_encoder), GFP_KERNEL);
+	struct dcn10_link_encoder *enc10 =
+		kzalloc(sizeof(struct dcn10_link_encoder), GFP_KERNEL);
 
-	if (!enc110)
+	if (!enc10)
 		return NULL;
 
-	dce110_link_encoder_construct(enc110,
+	dcn10_link_encoder_construct(enc10,
 				      enc_init_data,
 				      &link_enc_feature,
 				      &link_enc_regs[enc_init_data->transmitter],
 				      &link_enc_aux_regs[enc_init_data->channel - 1],
-				      &link_enc_hpd_regs[enc_init_data->hpd_source]);
+				      &link_enc_hpd_regs[enc_init_data->hpd_source],
+				      &le_shift,
+				      &le_mask);
 
-	return &enc110->base;
+	return &enc10->base;
 }
 
 struct clock_source *dcn10_clock_source_create(
@@ -650,16 +707,16 @@ static struct stream_encoder *dcn10_stream_encoder_create(
 	enum engine_id eng_id,
 	struct dc_context *ctx)
 {
-	struct dce110_stream_encoder *enc110 =
-		kzalloc(sizeof(struct dce110_stream_encoder), GFP_KERNEL);
+	struct dcn10_stream_encoder *enc1 =
+		kzalloc(sizeof(struct dcn10_stream_encoder), GFP_KERNEL);
 
-	if (!enc110)
+	if (!enc1)
 		return NULL;
 
-	dce110_stream_encoder_construct(enc110, ctx, ctx->dc_bios, eng_id,
+	dcn10_stream_encoder_construct(enc1, ctx, ctx->dc_bios, eng_id,
 					&stream_enc_regs[eng_id],
 					&se_shift, &se_mask);
-	return &enc110->base;
+	return &enc1->base;
 }
 
 static const struct dce_hwseq_registers hwseq_reg = {
@@ -686,6 +743,7 @@ static struct dce_hwseq *dcn10_hwseq_create(
 		hws->masks = &hwseq_mask;
 		hws->wa.DEGVIDCN10_253 = true;
 		hws->wa.false_optc_underflow = true;
+		hws->wa.DEGVIDCN10_254 = true;
 	}
 	return hws;
 }
@@ -796,8 +854,8 @@ static void destruct(struct dcn10_resource_pool *pool)
 	if (pool->base.dmcu != NULL)
 		dce_dmcu_destroy(&pool->base.dmcu);
 
-	if (pool->base.display_clock != NULL)
-		dce_disp_clk_destroy(&pool->base.display_clock);
+	if (pool->base.dccg != NULL)
+		dce_dccg_destroy(&pool->base.dccg);
 
 	kfree(pool->base.pp_smu);
 }
@@ -918,36 +976,6 @@ enum dc_status dcn10_add_stream_to_ctx(
 	return result;
 }
 
-enum dc_status dcn10_validate_guaranteed(
-		struct dc *dc,
-		struct dc_stream_state *dc_stream,
-		struct dc_state *context)
-{
-	enum dc_status result = DC_ERROR_UNEXPECTED;
-
-	context->streams[0] = dc_stream;
-	dc_stream_retain(context->streams[0]);
-	context->stream_count++;
-
-	result = resource_map_pool_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = resource_map_phy_clock_resources(dc, context, dc_stream);
-
-	if (result == DC_OK)
-		result = build_mapped_resource(dc, context, dc_stream);
-
-	if (result == DC_OK) {
-		validate_guaranteed_copy_streams(
-				context, dc->caps.max_streams);
-		result = resource_build_scaling_params_for_context(dc, context);
-	}
-	if (result == DC_OK && !dcn_validate_bandwidth(dc, context))
-		return DC_FAIL_BANDWIDTH_VALIDATE;
-
-	return result;
-}
-
 static struct pipe_ctx *dcn10_acquire_idle_pipe_for_layer(
 		struct dc_state *context,
 		const struct resource_pool *pool,
@@ -978,235 +1006,16 @@ static struct pipe_ctx *dcn10_acquire_idle_pipe_for_layer(
 	return idle_pipe;
 }
 
-enum dcc_control {
-	dcc_control__256_256_xxx,
-	dcc_control__128_128_xxx,
-	dcc_control__256_64_64,
-};
-
-enum segment_order {
-	segment_order__na,
-	segment_order__contiguous,
-	segment_order__non_contiguous,
-};
-
-static bool dcc_support_pixel_format(
-		enum surface_pixel_format format,
-		unsigned int *bytes_per_element)
-{
-	/* DML: get_bytes_per_element */
-	switch (format) {
-	case SURFACE_PIXEL_FORMAT_GRPH_ARGB1555:
-	case SURFACE_PIXEL_FORMAT_GRPH_RGB565:
-		*bytes_per_element = 2;
-		return true;
-	case SURFACE_PIXEL_FORMAT_GRPH_ARGB8888:
-	case SURFACE_PIXEL_FORMAT_GRPH_ABGR8888:
-	case SURFACE_PIXEL_FORMAT_GRPH_ARGB2101010:
-	case SURFACE_PIXEL_FORMAT_GRPH_ABGR2101010:
-		*bytes_per_element = 4;
-		return true;
-	case SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616:
-	case SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616F:
-	case SURFACE_PIXEL_FORMAT_GRPH_ABGR16161616F:
-		*bytes_per_element = 8;
-		return true;
-	default:
-		return false;
-	}
-}
-
-static bool dcc_support_swizzle(
-		enum swizzle_mode_values swizzle,
-		unsigned int bytes_per_element,
-		enum segment_order *segment_order_horz,
-		enum segment_order *segment_order_vert)
-{
-	bool standard_swizzle = false;
-	bool display_swizzle = false;
-
-	switch (swizzle) {
-	case DC_SW_4KB_S:
-	case DC_SW_64KB_S:
-	case DC_SW_VAR_S:
-	case DC_SW_4KB_S_X:
-	case DC_SW_64KB_S_X:
-	case DC_SW_VAR_S_X:
-		standard_swizzle = true;
-		break;
-	case DC_SW_4KB_D:
-	case DC_SW_64KB_D:
-	case DC_SW_VAR_D:
-	case DC_SW_4KB_D_X:
-	case DC_SW_64KB_D_X:
-	case DC_SW_VAR_D_X:
-		display_swizzle = true;
-		break;
-	default:
-		break;
-	}
-
-	if (bytes_per_element == 1 && standard_swizzle) {
-		*segment_order_horz = segment_order__contiguous;
-		*segment_order_vert = segment_order__na;
-		return true;
-	}
-	if (bytes_per_element == 2 && standard_swizzle) {
-		*segment_order_horz = segment_order__non_contiguous;
-		*segment_order_vert = segment_order__contiguous;
-		return true;
-	}
-	if (bytes_per_element == 4 && standard_swizzle) {
-		*segment_order_horz = segment_order__non_contiguous;
-		*segment_order_vert = segment_order__contiguous;
-		return true;
-	}
-	if (bytes_per_element == 8 && standard_swizzle) {
-		*segment_order_horz = segment_order__na;
-		*segment_order_vert = segment_order__contiguous;
-		return true;
-	}
-	if (bytes_per_element == 8 && display_swizzle) {
-		*segment_order_horz = segment_order__contiguous;
-		*segment_order_vert = segment_order__non_contiguous;
-		return true;
-	}
-
-	return false;
-}
-
-static void get_blk256_size(unsigned int *blk256_width, unsigned int *blk256_height,
-		unsigned int bytes_per_element)
-{
-	/* copied from DML.  might want to refactor DML to leverage from DML */
-	/* DML : get_blk256_size */
-	if (bytes_per_element == 1) {
-		*blk256_width = 16;
-		*blk256_height = 16;
-	} else if (bytes_per_element == 2) {
-		*blk256_width = 16;
-		*blk256_height = 8;
-	} else if (bytes_per_element == 4) {
-		*blk256_width = 8;
-		*blk256_height = 8;
-	} else if (bytes_per_element == 8) {
-		*blk256_width = 8;
-		*blk256_height = 4;
-	}
-}
-
-static void det_request_size(
-		unsigned int height,
-		unsigned int width,
-		unsigned int bpe,
-		bool *req128_horz_wc,
-		bool *req128_vert_wc)
-{
-	unsigned int detile_buf_size = 164 * 1024;  /* 164KB for DCN1.0 */
-
-	unsigned int blk256_height = 0;
-	unsigned int blk256_width = 0;
-	unsigned int swath_bytes_horz_wc, swath_bytes_vert_wc;
-
-	get_blk256_size(&blk256_width, &blk256_height, bpe);
-
-	swath_bytes_horz_wc = height * blk256_height * bpe;
-	swath_bytes_vert_wc = width * blk256_width * bpe;
-
-	*req128_horz_wc = (2 * swath_bytes_horz_wc <= detile_buf_size) ?
-			false : /* full 256B request */
-			true; /* half 128b request */
-
-	*req128_vert_wc = (2 * swath_bytes_vert_wc <= detile_buf_size) ?
-			false : /* full 256B request */
-			true; /* half 128b request */
-}
-
-static bool get_dcc_compression_cap(const struct dc *dc,
+static bool dcn10_get_dcc_compression_cap(const struct dc *dc,
 		const struct dc_dcc_surface_param *input,
 		struct dc_surface_dcc_cap *output)
 {
-	/* implement section 1.6.2.1 of DCN1_Programming_Guide.docx */
-	enum dcc_control dcc_control;
-	unsigned int bpe;
-	enum segment_order segment_order_horz, segment_order_vert;
-	bool req128_horz_wc, req128_vert_wc;
-
-	memset(output, 0, sizeof(*output));
-
-	if (dc->debug.disable_dcc == DCC_DISABLE)
-		return false;
-
-	if (!dcc_support_pixel_format(input->format,
-			&bpe))
-		return false;
-
-	if (!dcc_support_swizzle(input->swizzle_mode, bpe,
-			&segment_order_horz, &segment_order_vert))
-		return false;
-
-	det_request_size(input->surface_size.height,  input->surface_size.width,
-			bpe, &req128_horz_wc, &req128_vert_wc);
-
-	if (!req128_horz_wc && !req128_vert_wc) {
-		dcc_control = dcc_control__256_256_xxx;
-	} else if (input->scan == SCAN_DIRECTION_HORIZONTAL) {
-		if (!req128_horz_wc)
-			dcc_control = dcc_control__256_256_xxx;
-		else if (segment_order_horz == segment_order__contiguous)
-			dcc_control = dcc_control__128_128_xxx;
-		else
-			dcc_control = dcc_control__256_64_64;
-	} else if (input->scan == SCAN_DIRECTION_VERTICAL) {
-		if (!req128_vert_wc)
-			dcc_control = dcc_control__256_256_xxx;
-		else if (segment_order_vert == segment_order__contiguous)
-			dcc_control = dcc_control__128_128_xxx;
-		else
-			dcc_control = dcc_control__256_64_64;
-	} else {
-		if ((req128_horz_wc &&
-			segment_order_horz == segment_order__non_contiguous) ||
-			(req128_vert_wc &&
-			segment_order_vert == segment_order__non_contiguous))
-			/* access_dir not known, must use most constraining */
-			dcc_control = dcc_control__256_64_64;
-		else
-			/* reg128 is true for either horz and vert
-			 * but segment_order is contiguous
-			 */
-			dcc_control = dcc_control__128_128_xxx;
-	}
-
-	if (dc->debug.disable_dcc == DCC_HALF_REQ_DISALBE &&
-		dcc_control != dcc_control__256_256_xxx)
-		return false;
-
-	switch (dcc_control) {
-	case dcc_control__256_256_xxx:
-		output->grph.rgb.max_uncompressed_blk_size = 256;
-		output->grph.rgb.max_compressed_blk_size = 256;
-		output->grph.rgb.independent_64b_blks = false;
-		break;
-	case dcc_control__128_128_xxx:
-		output->grph.rgb.max_uncompressed_blk_size = 128;
-		output->grph.rgb.max_compressed_blk_size = 128;
-		output->grph.rgb.independent_64b_blks = false;
-		break;
-	case dcc_control__256_64_64:
-		output->grph.rgb.max_uncompressed_blk_size = 256;
-		output->grph.rgb.max_compressed_blk_size = 64;
-		output->grph.rgb.independent_64b_blks = true;
-		break;
-	}
-
-	output->capable = true;
-	output->const_color_support = false;
-
-	return true;
+	return dc->res_pool->hubbub->funcs->get_dcc_compression_cap(
+			dc->res_pool->hubbub,
+			input,
+			output);
 }
 
-
 static void dcn10_destroy_resource_pool(struct resource_pool **pool)
 {
 	struct dcn10_resource_pool *dcn10_pool = TO_DCN10_RES_POOL(*pool);
@@ -1226,14 +1035,13 @@ static enum dc_status dcn10_validate_plane(const struct dc_plane_state *plane_st
 	return DC_OK;
 }
 
-static struct dc_cap_funcs cap_funcs = {
-	.get_dcc_compression_cap = get_dcc_compression_cap
+static const struct dc_cap_funcs cap_funcs = {
+	.get_dcc_compression_cap = dcn10_get_dcc_compression_cap
 };
 
-static struct resource_funcs dcn10_res_pool_funcs = {
+static const struct resource_funcs dcn10_res_pool_funcs = {
 	.destroy = dcn10_destroy_resource_pool,
 	.link_enc_create = dcn10_link_encoder_create,
-	.validate_guaranteed = dcn10_validate_guaranteed,
 	.validate_bandwidth = dcn_validate_bandwidth,
 	.acquire_idle_pipe_for_layer = dcn10_acquire_idle_pipe_for_layer,
 	.validate_plane = dcn10_validate_plane,
@@ -1282,6 +1090,7 @@ static bool construct(
 	dc->caps.max_cursor_size = 256;
 	dc->caps.max_slave_planes = 1;
 	dc->caps.is_apu = true;
+	dc->caps.post_blend_color_processing = false;
 
 	if (dc->ctx->dce_environment == DCE_ENV_PRODUCTION_DRV)
 		dc->debug = debug_defaults_drv;
@@ -1325,8 +1134,8 @@ static bool construct(
 		}
 	}
 
-	pool->base.display_clock = dce120_disp_clk_create(ctx);
-	if (pool->base.display_clock == NULL) {
+	pool->base.dccg = dcn1_dccg_create(ctx);
+	if (pool->base.dccg == NULL) {
 		dm_error("DC: failed to create display clock!\n");
 		BREAK_TO_DEBUGGER();
 		goto fail;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
new file mode 100644
index 000000000000..6f9078f3c4d3
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.c
@@ -0,0 +1,1470 @@
+/*
+ * Copyright 2012-15 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ *  and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+
+#include "dc_bios_types.h"
+#include "dcn10_stream_encoder.h"
+#include "reg_helper.h"
+#include "hw_shared.h"
+
+#define DC_LOGGER \
+		enc1->base.ctx->logger
+
+
+#define REG(reg)\
+	(enc1->regs->reg)
+
+#undef FN
+#define FN(reg_name, field_name) \
+	enc1->se_shift->field_name, enc1->se_mask->field_name
+
+#define VBI_LINE_0 0
+#define DP_BLANK_MAX_RETRY 20
+#define HDMI_CLOCK_CHANNEL_RATE_MORE_340M 340000
+
+
+enum {
+	DP_MST_UPDATE_MAX_RETRY = 50
+};
+
+#define CTX \
+	enc1->base.ctx
+
+void enc1_update_generic_info_packet(
+	struct dcn10_stream_encoder *enc1,
+	uint32_t packet_index,
+	const struct dc_info_packet *info_packet)
+{
+	uint32_t regval;
+	/* TODOFPGA Figure out a proper number for max_retries polling for lock
+	 * use 50 for now.
+	 */
+	uint32_t max_retries = 50;
+
+	/*we need turn on clock before programming AFMT block*/
+	REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, 1);
+
+	if (packet_index >= 8)
+		ASSERT(0);
+
+	/* poll dig_update_lock is not locked -> asic internal signal
+	 * assume otg master lock will unlock it
+	 */
+/*		REG_WAIT(AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_LOCK_STATUS,
+			0, 10, max_retries);*/
+
+	/* check if HW reading GSP memory */
+	REG_WAIT(AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_CONFLICT,
+			0, 10, max_retries);
+
+	/* HW does is not reading GSP memory not reading too long ->
+	 * something wrong. clear GPS memory access and notify?
+	 * hw SW is writing to GSP memory
+	 */
+	REG_UPDATE(AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_CONFLICT_CLR, 1);
+
+	/* choose which generic packet to use */
+	regval = REG_READ(AFMT_VBI_PACKET_CONTROL);
+	REG_UPDATE(AFMT_VBI_PACKET_CONTROL,
+			AFMT_GENERIC_INDEX, packet_index);
+
+	/* write generic packet header
+	 * (4th byte is for GENERIC0 only)
+	 */
+	REG_SET_4(AFMT_GENERIC_HDR, 0,
+			AFMT_GENERIC_HB0, info_packet->hb0,
+			AFMT_GENERIC_HB1, info_packet->hb1,
+			AFMT_GENERIC_HB2, info_packet->hb2,
+			AFMT_GENERIC_HB3, info_packet->hb3);
+
+	/* write generic packet contents
+	 * (we never use last 4 bytes)
+	 * there are 8 (0-7) mmDIG0_AFMT_GENERIC0_x registers
+	 */
+	{
+		const uint32_t *content =
+			(const uint32_t *) &info_packet->sb[0];
+
+		REG_WRITE(AFMT_GENERIC_0, *content++);
+		REG_WRITE(AFMT_GENERIC_1, *content++);
+		REG_WRITE(AFMT_GENERIC_2, *content++);
+		REG_WRITE(AFMT_GENERIC_3, *content++);
+		REG_WRITE(AFMT_GENERIC_4, *content++);
+		REG_WRITE(AFMT_GENERIC_5, *content++);
+		REG_WRITE(AFMT_GENERIC_6, *content++);
+		REG_WRITE(AFMT_GENERIC_7, *content);
+	}
+
+	switch (packet_index) {
+	case 0:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC0_FRAME_UPDATE, 1);
+		break;
+	case 1:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC1_FRAME_UPDATE, 1);
+		break;
+	case 2:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC2_FRAME_UPDATE, 1);
+		break;
+	case 3:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC3_FRAME_UPDATE, 1);
+		break;
+	case 4:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC4_FRAME_UPDATE, 1);
+		break;
+	case 5:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC5_FRAME_UPDATE, 1);
+		break;
+	case 6:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC6_FRAME_UPDATE, 1);
+		break;
+	case 7:
+		REG_UPDATE(AFMT_VBI_PACKET_CONTROL1,
+				AFMT_GENERIC7_FRAME_UPDATE, 1);
+		break;
+	default:
+		break;
+	}
+}
+
+static void enc1_update_hdmi_info_packet(
+	struct dcn10_stream_encoder *enc1,
+	uint32_t packet_index,
+	const struct dc_info_packet *info_packet)
+{
+	uint32_t cont, send, line;
+
+	if (info_packet->valid) {
+		enc1_update_generic_info_packet(
+			enc1,
+			packet_index,
+			info_packet);
+
+		/* enable transmission of packet(s) -
+		 * packet transmission begins on the next frame
+		 */
+		cont = 1;
+		/* send packet(s) every frame */
+		send = 1;
+		/* select line number to send packets on */
+		line = 2;
+	} else {
+		cont = 0;
+		send = 0;
+		line = 0;
+	}
+
+	/* choose which generic packet control to use */
+	switch (packet_index) {
+	case 0:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL0,
+				HDMI_GENERIC0_CONT, cont,
+				HDMI_GENERIC0_SEND, send,
+				HDMI_GENERIC0_LINE, line);
+		break;
+	case 1:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL0,
+				HDMI_GENERIC1_CONT, cont,
+				HDMI_GENERIC1_SEND, send,
+				HDMI_GENERIC1_LINE, line);
+		break;
+	case 2:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL1,
+				HDMI_GENERIC0_CONT, cont,
+				HDMI_GENERIC0_SEND, send,
+				HDMI_GENERIC0_LINE, line);
+		break;
+	case 3:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL1,
+				HDMI_GENERIC1_CONT, cont,
+				HDMI_GENERIC1_SEND, send,
+				HDMI_GENERIC1_LINE, line);
+		break;
+	case 4:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL2,
+				HDMI_GENERIC0_CONT, cont,
+				HDMI_GENERIC0_SEND, send,
+				HDMI_GENERIC0_LINE, line);
+		break;
+	case 5:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL2,
+				HDMI_GENERIC1_CONT, cont,
+				HDMI_GENERIC1_SEND, send,
+				HDMI_GENERIC1_LINE, line);
+		break;
+	case 6:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL3,
+				HDMI_GENERIC0_CONT, cont,
+				HDMI_GENERIC0_SEND, send,
+				HDMI_GENERIC0_LINE, line);
+		break;
+	case 7:
+		REG_UPDATE_3(HDMI_GENERIC_PACKET_CONTROL3,
+				HDMI_GENERIC1_CONT, cont,
+				HDMI_GENERIC1_SEND, send,
+				HDMI_GENERIC1_LINE, line);
+		break;
+	default:
+		/* invalid HW packet index */
+		DC_LOG_WARNING(
+			"Invalid HW packet index: %s()\n",
+			__func__);
+		return;
+	}
+}
+
+/* setup stream encoder in dp mode */
+void enc1_stream_encoder_dp_set_stream_attribute(
+	struct stream_encoder *enc,
+	struct dc_crtc_timing *crtc_timing,
+	enum dc_color_space output_color_space)
+{
+	uint32_t h_active_start;
+	uint32_t v_active_start;
+	uint32_t misc0 = 0;
+	uint32_t misc1 = 0;
+	uint32_t h_blank;
+	uint32_t h_back_porch;
+	uint8_t synchronous_clock = 0; /* asynchronous mode */
+	uint8_t colorimetry_bpc;
+	uint8_t dynamic_range_rgb = 0; /*full range*/
+	uint8_t dynamic_range_ycbcr = 1; /*bt709*/
+	uint8_t dp_pixel_encoding = 0;
+	uint8_t dp_component_depth = 0;
+
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	/* set pixel encoding */
+	switch (crtc_timing->pixel_encoding) {
+	case PIXEL_ENCODING_YCBCR422:
+		dp_pixel_encoding = DP_PIXEL_ENCODING_TYPE_YCBCR422;
+		break;
+	case PIXEL_ENCODING_YCBCR444:
+		dp_pixel_encoding = DP_PIXEL_ENCODING_TYPE_YCBCR444;
+
+		if (crtc_timing->flags.Y_ONLY)
+			if (crtc_timing->display_color_depth != COLOR_DEPTH_666)
+				/* HW testing only, no use case yet.
+				 * Color depth of Y-only could be
+				 * 8, 10, 12, 16 bits
+				 */
+				dp_pixel_encoding = DP_PIXEL_ENCODING_TYPE_Y_ONLY;
+
+		/* Note: DP_MSA_MISC1 bit 7 is the indicator
+		 * of Y-only mode.
+		 * This bit is set in HW if register
+		 * DP_PIXEL_ENCODING is programmed to 0x4
+		 */
+		break;
+	case PIXEL_ENCODING_YCBCR420:
+		dp_pixel_encoding = DP_PIXEL_ENCODING_TYPE_YCBCR420;
+		REG_UPDATE(DP_VID_TIMING, DP_VID_N_MUL, 1);
+		break;
+	default:
+		dp_pixel_encoding = DP_PIXEL_ENCODING_TYPE_RGB444;
+		break;
+	}
+
+	misc1 = REG_READ(DP_MSA_MISC);
+	/* For YCbCr420 and BT2020 Colorimetry Formats, VSC SDP shall be used.
+	 * When MISC1, bit 6, is Set to 1, a Source device uses a VSC SDP to indicate the
+	 * Pixel Encoding/Colorimetry Format and that a Sink device shall ignore MISC1, bit 7,
+	 * and MISC0, bits 7:1 (MISC1, bit 7, and MISC0, bits 7:1, become "don't care").
+	 */
+	if ((crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR420) ||
+			(output_color_space == COLOR_SPACE_2020_YCBCR) ||
+			(output_color_space == COLOR_SPACE_2020_RGB_FULLRANGE) ||
+			(output_color_space == COLOR_SPACE_2020_RGB_LIMITEDRANGE))
+		misc1 = misc1 | 0x40;
+	else
+		misc1 = misc1 & ~0x40;
+
+	/* set color depth */
+	switch (crtc_timing->display_color_depth) {
+	case COLOR_DEPTH_666:
+		dp_component_depth = DP_COMPONENT_PIXEL_DEPTH_6BPC;
+		break;
+	case COLOR_DEPTH_888:
+		dp_component_depth = DP_COMPONENT_PIXEL_DEPTH_8BPC;
+		break;
+	case COLOR_DEPTH_101010:
+		dp_component_depth = DP_COMPONENT_PIXEL_DEPTH_10BPC;
+		break;
+	case COLOR_DEPTH_121212:
+		dp_component_depth = DP_COMPONENT_PIXEL_DEPTH_12BPC;
+		break;
+	case COLOR_DEPTH_161616:
+		dp_component_depth = DP_COMPONENT_PIXEL_DEPTH_16BPC;
+		break;
+	default:
+		dp_component_depth = DP_COMPONENT_PIXEL_DEPTH_6BPC;
+		break;
+	}
+
+	/* Set DP pixel encoding and component depth */
+	REG_UPDATE_2(DP_PIXEL_FORMAT,
+			DP_PIXEL_ENCODING, dp_pixel_encoding,
+			DP_COMPONENT_DEPTH, dp_component_depth);
+
+	/* set dynamic range and YCbCr range */
+
+	switch (crtc_timing->display_color_depth) {
+	case COLOR_DEPTH_666:
+		colorimetry_bpc = 0;
+		break;
+	case COLOR_DEPTH_888:
+		colorimetry_bpc = 1;
+		break;
+	case COLOR_DEPTH_101010:
+		colorimetry_bpc = 2;
+		break;
+	case COLOR_DEPTH_121212:
+		colorimetry_bpc = 3;
+		break;
+	default:
+		colorimetry_bpc = 0;
+		break;
+	}
+
+	misc0 = misc0 | synchronous_clock;
+	misc0 = colorimetry_bpc << 5;
+
+	switch (output_color_space) {
+	case COLOR_SPACE_SRGB:
+		misc1 = misc1 & ~0x80; /* bit7 = 0*/
+		dynamic_range_rgb = 0; /*full range*/
+		break;
+	case COLOR_SPACE_SRGB_LIMITED:
+		misc0 = misc0 | 0x8; /* bit3=1 */
+		misc1 = misc1 & ~0x80; /* bit7 = 0*/
+		dynamic_range_rgb = 1; /*limited range*/
+		break;
+	case COLOR_SPACE_YCBCR601:
+	case COLOR_SPACE_YCBCR601_LIMITED:
+		misc0 = misc0 | 0x8; /* bit3=1, bit4=0 */
+		misc1 = misc1 & ~0x80; /* bit7 = 0*/
+		dynamic_range_ycbcr = 0; /*bt601*/
+		if (crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR422)
+			misc0 = misc0 | 0x2; /* bit2=0, bit1=1 */
+		else if (crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR444)
+			misc0 = misc0 | 0x4; /* bit2=1, bit1=0 */
+		break;
+	case COLOR_SPACE_YCBCR709:
+	case COLOR_SPACE_YCBCR709_LIMITED:
+		misc0 = misc0 | 0x18; /* bit3=1, bit4=1 */
+		misc1 = misc1 & ~0x80; /* bit7 = 0*/
+		dynamic_range_ycbcr = 1; /*bt709*/
+		if (crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR422)
+			misc0 = misc0 | 0x2; /* bit2=0, bit1=1 */
+		else if (crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR444)
+			misc0 = misc0 | 0x4; /* bit2=1, bit1=0 */
+		break;
+	case COLOR_SPACE_2020_RGB_LIMITEDRANGE:
+		dynamic_range_rgb = 1; /*limited range*/
+		break;
+	case COLOR_SPACE_2020_RGB_FULLRANGE:
+	case COLOR_SPACE_2020_YCBCR:
+	case COLOR_SPACE_XR_RGB:
+	case COLOR_SPACE_MSREF_SCRGB:
+	case COLOR_SPACE_ADOBERGB:
+	case COLOR_SPACE_DCIP3:
+	case COLOR_SPACE_XV_YCC_709:
+	case COLOR_SPACE_XV_YCC_601:
+	case COLOR_SPACE_DISPLAYNATIVE:
+	case COLOR_SPACE_DOLBYVISION:
+	case COLOR_SPACE_APPCTRL:
+	case COLOR_SPACE_CUSTOMPOINTS:
+	case COLOR_SPACE_UNKNOWN:
+		/* do nothing */
+		break;
+	}
+
+	REG_SET(DP_MSA_COLORIMETRY, 0, DP_MSA_MISC0, misc0);
+	REG_WRITE(DP_MSA_MISC, misc1);   /* MSA_MISC1 */
+
+	/* dcn new register
+	 * dc_crtc_timing is vesa dmt struct. data from edid
+	 */
+	REG_SET_2(DP_MSA_TIMING_PARAM1, 0,
+			DP_MSA_HTOTAL, crtc_timing->h_total,
+			DP_MSA_VTOTAL, crtc_timing->v_total);
+
+	/* calculate from vesa timing parameters
+	 * h_active_start related to leading edge of sync
+	 */
+
+	h_blank = crtc_timing->h_total - crtc_timing->h_border_left -
+			crtc_timing->h_addressable - crtc_timing->h_border_right;
+
+	h_back_porch = h_blank - crtc_timing->h_front_porch -
+			crtc_timing->h_sync_width;
+
+	/* start at beginning of left border */
+	h_active_start = crtc_timing->h_sync_width + h_back_porch;
+
+
+	v_active_start = crtc_timing->v_total - crtc_timing->v_border_top -
+			crtc_timing->v_addressable - crtc_timing->v_border_bottom -
+			crtc_timing->v_front_porch;
+
+
+	/* start at beginning of left border */
+	REG_SET_2(DP_MSA_TIMING_PARAM2, 0,
+		DP_MSA_HSTART, h_active_start,
+		DP_MSA_VSTART, v_active_start);
+
+	REG_SET_4(DP_MSA_TIMING_PARAM3, 0,
+			DP_MSA_HSYNCWIDTH,
+			crtc_timing->h_sync_width,
+			DP_MSA_HSYNCPOLARITY,
+			!crtc_timing->flags.HSYNC_POSITIVE_POLARITY,
+			DP_MSA_VSYNCWIDTH,
+			crtc_timing->v_sync_width,
+			DP_MSA_VSYNCPOLARITY,
+			!crtc_timing->flags.VSYNC_POSITIVE_POLARITY);
+
+	/* HWDITH include border or overscan */
+	REG_SET_2(DP_MSA_TIMING_PARAM4, 0,
+		DP_MSA_HWIDTH, crtc_timing->h_border_left +
+		crtc_timing->h_addressable + crtc_timing->h_border_right,
+		DP_MSA_VHEIGHT, crtc_timing->v_border_top +
+		crtc_timing->v_addressable + crtc_timing->v_border_bottom);
+}
+
+static void enc1_stream_encoder_set_stream_attribute_helper(
+		struct dcn10_stream_encoder *enc1,
+		struct dc_crtc_timing *crtc_timing)
+{
+	switch (crtc_timing->pixel_encoding) {
+	case PIXEL_ENCODING_YCBCR422:
+		REG_UPDATE(DIG_FE_CNTL, TMDS_PIXEL_ENCODING, 1);
+		break;
+	default:
+		REG_UPDATE(DIG_FE_CNTL, TMDS_PIXEL_ENCODING, 0);
+		break;
+	}
+	REG_UPDATE(DIG_FE_CNTL, TMDS_COLOR_FORMAT, 0);
+}
+
+/* setup stream encoder in hdmi mode */
+void enc1_stream_encoder_hdmi_set_stream_attribute(
+	struct stream_encoder *enc,
+	struct dc_crtc_timing *crtc_timing,
+	int actual_pix_clk_khz,
+	bool enable_audio)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	struct bp_encoder_control cntl = {0};
+
+	cntl.action = ENCODER_CONTROL_SETUP;
+	cntl.engine_id = enc1->base.id;
+	cntl.signal = SIGNAL_TYPE_HDMI_TYPE_A;
+	cntl.enable_dp_audio = enable_audio;
+	cntl.pixel_clock = actual_pix_clk_khz;
+	cntl.lanes_number = LANE_COUNT_FOUR;
+
+	if (enc1->base.bp->funcs->encoder_control(
+			enc1->base.bp, &cntl) != BP_RESULT_OK)
+		return;
+
+	enc1_stream_encoder_set_stream_attribute_helper(enc1, crtc_timing);
+
+	/* setup HDMI engine */
+	REG_UPDATE_5(HDMI_CONTROL,
+		HDMI_PACKET_GEN_VERSION, 1,
+		HDMI_KEEPOUT_MODE, 1,
+		HDMI_DEEP_COLOR_ENABLE, 0,
+		HDMI_DATA_SCRAMBLE_EN, 0,
+		HDMI_CLOCK_CHANNEL_RATE, 0);
+
+
+	switch (crtc_timing->display_color_depth) {
+	case COLOR_DEPTH_888:
+		REG_UPDATE(HDMI_CONTROL, HDMI_DEEP_COLOR_DEPTH, 0);
+		break;
+	case COLOR_DEPTH_101010:
+		if (crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR422) {
+			REG_UPDATE_2(HDMI_CONTROL,
+					HDMI_DEEP_COLOR_DEPTH, 1,
+					HDMI_DEEP_COLOR_ENABLE, 0);
+		} else {
+			REG_UPDATE_2(HDMI_CONTROL,
+					HDMI_DEEP_COLOR_DEPTH, 1,
+					HDMI_DEEP_COLOR_ENABLE, 1);
+			}
+		break;
+	case COLOR_DEPTH_121212:
+		if (crtc_timing->pixel_encoding == PIXEL_ENCODING_YCBCR422) {
+			REG_UPDATE_2(HDMI_CONTROL,
+					HDMI_DEEP_COLOR_DEPTH, 2,
+					HDMI_DEEP_COLOR_ENABLE, 0);
+		} else {
+			REG_UPDATE_2(HDMI_CONTROL,
+					HDMI_DEEP_COLOR_DEPTH, 2,
+					HDMI_DEEP_COLOR_ENABLE, 1);
+			}
+		break;
+	case COLOR_DEPTH_161616:
+		REG_UPDATE_2(HDMI_CONTROL,
+				HDMI_DEEP_COLOR_DEPTH, 3,
+				HDMI_DEEP_COLOR_ENABLE, 1);
+		break;
+	default:
+		break;
+	}
+
+	if (actual_pix_clk_khz >= HDMI_CLOCK_CHANNEL_RATE_MORE_340M) {
+		/* enable HDMI data scrambler
+		 * HDMI_CLOCK_CHANNEL_RATE_MORE_340M
+		 * Clock channel frequency is 1/4 of character rate.
+		 */
+		REG_UPDATE_2(HDMI_CONTROL,
+			HDMI_DATA_SCRAMBLE_EN, 1,
+			HDMI_CLOCK_CHANNEL_RATE, 1);
+	} else if (crtc_timing->flags.LTE_340MCSC_SCRAMBLE) {
+
+		/* TODO: New feature for DCE11, still need to implement */
+
+		/* enable HDMI data scrambler
+		 * HDMI_CLOCK_CHANNEL_FREQ_EQUAL_TO_CHAR_RATE
+		 * Clock channel frequency is the same
+		 * as character rate
+		 */
+		REG_UPDATE_2(HDMI_CONTROL,
+			HDMI_DATA_SCRAMBLE_EN, 1,
+			HDMI_CLOCK_CHANNEL_RATE, 0);
+	}
+
+
+	REG_UPDATE_3(HDMI_VBI_PACKET_CONTROL,
+		HDMI_GC_CONT, 1,
+		HDMI_GC_SEND, 1,
+		HDMI_NULL_SEND, 1);
+
+	/* following belongs to audio */
+	REG_UPDATE(HDMI_INFOFRAME_CONTROL0, HDMI_AUDIO_INFO_SEND, 1);
+
+	REG_UPDATE(AFMT_INFOFRAME_CONTROL0, AFMT_AUDIO_INFO_UPDATE, 1);
+
+	REG_UPDATE(HDMI_INFOFRAME_CONTROL1, HDMI_AUDIO_INFO_LINE,
+				VBI_LINE_0 + 2);
+
+	REG_UPDATE(HDMI_GC, HDMI_GC_AVMUTE, 0);
+}
+
+/* setup stream encoder in dvi mode */
+void enc1_stream_encoder_dvi_set_stream_attribute(
+	struct stream_encoder *enc,
+	struct dc_crtc_timing *crtc_timing,
+	bool is_dual_link)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	struct bp_encoder_control cntl = {0};
+
+	cntl.action = ENCODER_CONTROL_SETUP;
+	cntl.engine_id = enc1->base.id;
+	cntl.signal = is_dual_link ?
+			SIGNAL_TYPE_DVI_DUAL_LINK : SIGNAL_TYPE_DVI_SINGLE_LINK;
+	cntl.enable_dp_audio = false;
+	cntl.pixel_clock = crtc_timing->pix_clk_khz;
+	cntl.lanes_number = (is_dual_link) ? LANE_COUNT_EIGHT : LANE_COUNT_FOUR;
+
+	if (enc1->base.bp->funcs->encoder_control(
+			enc1->base.bp, &cntl) != BP_RESULT_OK)
+		return;
+
+	ASSERT(crtc_timing->pixel_encoding == PIXEL_ENCODING_RGB);
+	ASSERT(crtc_timing->display_color_depth == COLOR_DEPTH_888);
+	enc1_stream_encoder_set_stream_attribute_helper(enc1, crtc_timing);
+}
+
+void enc1_stream_encoder_set_mst_bandwidth(
+	struct stream_encoder *enc,
+	struct fixed31_32 avg_time_slots_per_mtp)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	uint32_t x = dc_fixpt_floor(
+		avg_time_slots_per_mtp);
+	uint32_t y = dc_fixpt_ceil(
+		dc_fixpt_shl(
+			dc_fixpt_sub_int(
+				avg_time_slots_per_mtp,
+				x),
+			26));
+
+	REG_SET_2(DP_MSE_RATE_CNTL, 0,
+		DP_MSE_RATE_X, x,
+		DP_MSE_RATE_Y, y);
+
+	/* wait for update to be completed on the link */
+	/* i.e. DP_MSE_RATE_UPDATE_PENDING field (read only) */
+	/* is reset to 0 (not pending) */
+	REG_WAIT(DP_MSE_RATE_UPDATE, DP_MSE_RATE_UPDATE_PENDING,
+			0,
+			10, DP_MST_UPDATE_MAX_RETRY);
+}
+
+static void enc1_stream_encoder_update_hdmi_info_packets(
+	struct stream_encoder *enc,
+	const struct encoder_info_frame *info_frame)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	/* for bring up, disable dp double  TODO */
+	REG_UPDATE(HDMI_DB_CONTROL, HDMI_DB_DISABLE, 1);
+
+	enc1_update_hdmi_info_packet(enc1, 0, &info_frame->avi);
+	enc1_update_hdmi_info_packet(enc1, 1, &info_frame->vendor);
+	enc1_update_hdmi_info_packet(enc1, 2, &info_frame->gamut);
+	enc1_update_hdmi_info_packet(enc1, 3, &info_frame->spd);
+	enc1_update_hdmi_info_packet(enc1, 4, &info_frame->hdrsmd);
+}
+
+static void enc1_stream_encoder_stop_hdmi_info_packets(
+	struct stream_encoder *enc)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	/* stop generic packets 0 & 1 on HDMI */
+	REG_SET_6(HDMI_GENERIC_PACKET_CONTROL0, 0,
+		HDMI_GENERIC1_CONT, 0,
+		HDMI_GENERIC1_LINE, 0,
+		HDMI_GENERIC1_SEND, 0,
+		HDMI_GENERIC0_CONT, 0,
+		HDMI_GENERIC0_LINE, 0,
+		HDMI_GENERIC0_SEND, 0);
+
+	/* stop generic packets 2 & 3 on HDMI */
+	REG_SET_6(HDMI_GENERIC_PACKET_CONTROL1, 0,
+		HDMI_GENERIC0_CONT, 0,
+		HDMI_GENERIC0_LINE, 0,
+		HDMI_GENERIC0_SEND, 0,
+		HDMI_GENERIC1_CONT, 0,
+		HDMI_GENERIC1_LINE, 0,
+		HDMI_GENERIC1_SEND, 0);
+
+	/* stop generic packets 2 & 3 on HDMI */
+	REG_SET_6(HDMI_GENERIC_PACKET_CONTROL2, 0,
+		HDMI_GENERIC0_CONT, 0,
+		HDMI_GENERIC0_LINE, 0,
+		HDMI_GENERIC0_SEND, 0,
+		HDMI_GENERIC1_CONT, 0,
+		HDMI_GENERIC1_LINE, 0,
+		HDMI_GENERIC1_SEND, 0);
+
+	REG_SET_6(HDMI_GENERIC_PACKET_CONTROL3, 0,
+		HDMI_GENERIC0_CONT, 0,
+		HDMI_GENERIC0_LINE, 0,
+		HDMI_GENERIC0_SEND, 0,
+		HDMI_GENERIC1_CONT, 0,
+		HDMI_GENERIC1_LINE, 0,
+		HDMI_GENERIC1_SEND, 0);
+}
+
+void enc1_stream_encoder_update_dp_info_packets(
+	struct stream_encoder *enc,
+	const struct encoder_info_frame *info_frame)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	uint32_t value = 0;
+
+	if (info_frame->vsc.valid)
+		enc1_update_generic_info_packet(
+					enc1,
+					0,  /* packetIndex */
+					&info_frame->vsc);
+
+	if (info_frame->spd.valid)
+		enc1_update_generic_info_packet(
+				enc1,
+				2,  /* packetIndex */
+				&info_frame->spd);
+
+	if (info_frame->hdrsmd.valid)
+		enc1_update_generic_info_packet(
+				enc1,
+				3,  /* packetIndex */
+				&info_frame->hdrsmd);
+
+	/* enable/disable transmission of packet(s).
+	 * If enabled, packet transmission begins on the next frame
+	 */
+	REG_UPDATE(DP_SEC_CNTL, DP_SEC_GSP0_ENABLE, info_frame->vsc.valid);
+	REG_UPDATE(DP_SEC_CNTL, DP_SEC_GSP2_ENABLE, info_frame->spd.valid);
+	REG_UPDATE(DP_SEC_CNTL, DP_SEC_GSP3_ENABLE, info_frame->hdrsmd.valid);
+
+
+	/* This bit is the master enable bit.
+	 * When enabling secondary stream engine,
+	 * this master bit must also be set.
+	 * This register shared with audio info frame.
+	 * Therefore we need to enable master bit
+	 * if at least on of the fields is not 0
+	 */
+	value = REG_READ(DP_SEC_CNTL);
+	if (value)
+		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
+}
+
+void enc1_stream_encoder_stop_dp_info_packets(
+	struct stream_encoder *enc)
+{
+	/* stop generic packets on DP */
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	uint32_t value = 0;
+
+	REG_SET_10(DP_SEC_CNTL, 0,
+		DP_SEC_GSP0_ENABLE, 0,
+		DP_SEC_GSP1_ENABLE, 0,
+		DP_SEC_GSP2_ENABLE, 0,
+		DP_SEC_GSP3_ENABLE, 0,
+		DP_SEC_GSP4_ENABLE, 0,
+		DP_SEC_GSP5_ENABLE, 0,
+		DP_SEC_GSP6_ENABLE, 0,
+		DP_SEC_GSP7_ENABLE, 0,
+		DP_SEC_MPG_ENABLE, 0,
+		DP_SEC_STREAM_ENABLE, 0);
+
+	/* this register shared with audio info frame.
+	 * therefore we need to keep master enabled
+	 * if at least one of the fields is not 0 */
+	value = REG_READ(DP_SEC_CNTL);
+	if (value)
+		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
+
+}
+
+void enc1_stream_encoder_dp_blank(
+	struct stream_encoder *enc)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	uint32_t retries = 0;
+	uint32_t  reg1 = 0;
+	uint32_t max_retries = DP_BLANK_MAX_RETRY * 10;
+
+	/* Note: For CZ, we are changing driver default to disable
+	 * stream deferred to next VBLANK. If results are positive, we
+	 * will make the same change to all DCE versions. There are a
+	 * handful of panels that cannot handle disable stream at
+	 * HBLANK and will result in a white line flash across the
+	 * screen on stream disable.
+	 */
+	REG_GET(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, &reg1);
+	if ((reg1 & 0x1) == 0)
+		/*stream not enabled*/
+		return;
+	/* Specify the video stream disable point
+	 * (2 = start of the next vertical blank)
+	 */
+	REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_DIS_DEFER, 2);
+	/* Larger delay to wait until VBLANK - use max retry of
+	 * 10us*3000=30ms. This covers 16.6ms of typical 60 Hz mode +
+	 * a little more because we may not trust delay accuracy.
+	 */
+	max_retries = DP_BLANK_MAX_RETRY * 150;
+
+	/* disable DP stream */
+	REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, 0);
+
+	/* the encoder stops sending the video stream
+	 * at the start of the vertical blanking.
+	 * Poll for DP_VID_STREAM_STATUS == 0
+	 */
+
+	REG_WAIT(DP_VID_STREAM_CNTL, DP_VID_STREAM_STATUS,
+			0,
+			10, max_retries);
+
+	ASSERT(retries <= max_retries);
+
+	/* Tell the DP encoder to ignore timing from CRTC, must be done after
+	 * the polling. If we set DP_STEER_FIFO_RESET before DP stream blank is
+	 * complete, stream status will be stuck in video stream enabled state,
+	 * i.e. DP_VID_STREAM_STATUS stuck at 1.
+	 */
+
+	REG_UPDATE(DP_STEER_FIFO, DP_STEER_FIFO_RESET, true);
+}
+
+/* output video stream to link encoder */
+void enc1_stream_encoder_dp_unblank(
+	struct stream_encoder *enc,
+	const struct encoder_unblank_param *param)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	if (param->link_settings.link_rate != LINK_RATE_UNKNOWN) {
+		uint32_t n_vid = 0x8000;
+		uint32_t m_vid;
+
+		/* M / N = Fstream / Flink
+		 * m_vid / n_vid = pixel rate / link rate
+		 */
+
+		uint64_t m_vid_l = n_vid;
+
+		m_vid_l *= param->pixel_clk_khz;
+		m_vid_l = div_u64(m_vid_l,
+			param->link_settings.link_rate
+				* LINK_RATE_REF_FREQ_IN_KHZ);
+
+		m_vid = (uint32_t) m_vid_l;
+
+		/* enable auto measurement */
+
+		REG_UPDATE(DP_VID_TIMING, DP_VID_M_N_GEN_EN, 0);
+
+		/* auto measurement need 1 full 0x8000 symbol cycle to kick in,
+		 * therefore program initial value for Mvid and Nvid
+		 */
+
+		REG_UPDATE(DP_VID_N, DP_VID_N, n_vid);
+
+		REG_UPDATE(DP_VID_M, DP_VID_M, m_vid);
+
+		REG_UPDATE(DP_VID_TIMING, DP_VID_M_N_GEN_EN, 1);
+	}
+
+	/* set DIG_START to 0x1 to resync FIFO */
+
+	REG_UPDATE(DIG_FE_CNTL, DIG_START, 1);
+
+	/* switch DP encoder to CRTC data */
+
+	REG_UPDATE(DP_STEER_FIFO, DP_STEER_FIFO_RESET, 0);
+
+	/* wait 100us for DIG/DP logic to prime
+	 * (i.e. a few video lines)
+	 */
+	udelay(100);
+
+	/* the hardware would start sending video at the start of the next DP
+	 * frame (i.e. rising edge of the vblank).
+	 * NOTE: We used to program DP_VID_STREAM_DIS_DEFER = 2 here, but this
+	 * register has no effect on enable transition! HW always guarantees
+	 * VID_STREAM enable at start of next frame, and this is not
+	 * programmable
+	 */
+
+	REG_UPDATE(DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, true);
+}
+
+void enc1_stream_encoder_set_avmute(
+	struct stream_encoder *enc,
+	bool enable)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	unsigned int value = enable ? 1 : 0;
+
+	REG_UPDATE(HDMI_GC, HDMI_GC_AVMUTE, value);
+}
+
+
+#define DP_SEC_AUD_N__DP_SEC_AUD_N__DEFAULT 0x8000
+#define DP_SEC_TIMESTAMP__DP_SEC_TIMESTAMP_MODE__AUTO_CALC 1
+
+#include "include/audio_types.h"
+
+/**
+* speakersToChannels
+*
+* @brief
+*  translate speakers to channels
+*
+*  FL  - Front Left
+*  FR  - Front Right
+*  RL  - Rear Left
+*  RR  - Rear Right
+*  RC  - Rear Center
+*  FC  - Front Center
+*  FLC - Front Left Center
+*  FRC - Front Right Center
+*  RLC - Rear Left Center
+*  RRC - Rear Right Center
+*  LFE - Low Freq Effect
+*
+*               FC
+*          FLC      FRC
+*    FL                    FR
+*
+*                    LFE
+*              ()
+*
+*
+*    RL                    RR
+*          RLC      RRC
+*               RC
+*
+*             ch  8   7   6   5   4   3   2   1
+* 0b00000011      -   -   -   -   -   -   FR  FL
+* 0b00000111      -   -   -   -   -   LFE FR  FL
+* 0b00001011      -   -   -   -   FC  -   FR  FL
+* 0b00001111      -   -   -   -   FC  LFE FR  FL
+* 0b00010011      -   -   -   RC  -   -   FR  FL
+* 0b00010111      -   -   -   RC  -   LFE FR  FL
+* 0b00011011      -   -   -   RC  FC  -   FR  FL
+* 0b00011111      -   -   -   RC  FC  LFE FR  FL
+* 0b00110011      -   -   RR  RL  -   -   FR  FL
+* 0b00110111      -   -   RR  RL  -   LFE FR  FL
+* 0b00111011      -   -   RR  RL  FC  -   FR  FL
+* 0b00111111      -   -   RR  RL  FC  LFE FR  FL
+* 0b01110011      -   RC  RR  RL  -   -   FR  FL
+* 0b01110111      -   RC  RR  RL  -   LFE FR  FL
+* 0b01111011      -   RC  RR  RL  FC  -   FR  FL
+* 0b01111111      -   RC  RR  RL  FC  LFE FR  FL
+* 0b11110011      RRC RLC RR  RL  -   -   FR  FL
+* 0b11110111      RRC RLC RR  RL  -   LFE FR  FL
+* 0b11111011      RRC RLC RR  RL  FC  -   FR  FL
+* 0b11111111      RRC RLC RR  RL  FC  LFE FR  FL
+* 0b11000011      FRC FLC -   -   -   -   FR  FL
+* 0b11000111      FRC FLC -   -   -   LFE FR  FL
+* 0b11001011      FRC FLC -   -   FC  -   FR  FL
+* 0b11001111      FRC FLC -   -   FC  LFE FR  FL
+* 0b11010011      FRC FLC -   RC  -   -   FR  FL
+* 0b11010111      FRC FLC -   RC  -   LFE FR  FL
+* 0b11011011      FRC FLC -   RC  FC  -   FR  FL
+* 0b11011111      FRC FLC -   RC  FC  LFE FR  FL
+* 0b11110011      FRC FLC RR  RL  -   -   FR  FL
+* 0b11110111      FRC FLC RR  RL  -   LFE FR  FL
+* 0b11111011      FRC FLC RR  RL  FC  -   FR  FL
+* 0b11111111      FRC FLC RR  RL  FC  LFE FR  FL
+*
+* @param
+*  speakers - speaker information as it comes from CEA audio block
+*/
+/* translate speakers to channels */
+
+union audio_cea_channels {
+	uint8_t all;
+	struct audio_cea_channels_bits {
+		uint32_t FL:1;
+		uint32_t FR:1;
+		uint32_t LFE:1;
+		uint32_t FC:1;
+		uint32_t RL_RC:1;
+		uint32_t RR:1;
+		uint32_t RC_RLC_FLC:1;
+		uint32_t RRC_FRC:1;
+	} channels;
+};
+
+struct audio_clock_info {
+	/* pixel clock frequency*/
+	uint32_t pixel_clock_in_10khz;
+	/* N - 32KHz audio */
+	uint32_t n_32khz;
+	/* CTS - 32KHz audio*/
+	uint32_t cts_32khz;
+	uint32_t n_44khz;
+	uint32_t cts_44khz;
+	uint32_t n_48khz;
+	uint32_t cts_48khz;
+};
+
+/* 25.2MHz/1.001*/
+/* 25.2MHz/1.001*/
+/* 25.2MHz*/
+/* 27MHz */
+/* 27MHz*1.001*/
+/* 27MHz*1.001*/
+/* 54MHz*/
+/* 54MHz*1.001*/
+/* 74.25MHz/1.001*/
+/* 74.25MHz*/
+/* 148.5MHz/1.001*/
+/* 148.5MHz*/
+
+static const struct audio_clock_info audio_clock_info_table[16] = {
+	{2517, 4576, 28125, 7007, 31250, 6864, 28125},
+	{2518, 4576, 28125, 7007, 31250, 6864, 28125},
+	{2520, 4096, 25200, 6272, 28000, 6144, 25200},
+	{2700, 4096, 27000, 6272, 30000, 6144, 27000},
+	{2702, 4096, 27027, 6272, 30030, 6144, 27027},
+	{2703, 4096, 27027, 6272, 30030, 6144, 27027},
+	{5400, 4096, 54000, 6272, 60000, 6144, 54000},
+	{5405, 4096, 54054, 6272, 60060, 6144, 54054},
+	{7417, 11648, 210937, 17836, 234375, 11648, 140625},
+	{7425, 4096, 74250, 6272, 82500, 6144, 74250},
+	{14835, 11648, 421875, 8918, 234375, 5824, 140625},
+	{14850, 4096, 148500, 6272, 165000, 6144, 148500},
+	{29670, 5824, 421875, 4459, 234375, 5824, 281250},
+	{29700, 3072, 222750, 4704, 247500, 5120, 247500},
+	{59340, 5824, 843750, 8918, 937500, 5824, 562500},
+	{59400, 3072, 445500, 9408, 990000, 6144, 594000}
+};
+
+static const struct audio_clock_info audio_clock_info_table_36bpc[14] = {
+	{2517,  9152,  84375,  7007,  48875,  9152,  56250},
+	{2518,  9152,  84375,  7007,  48875,  9152,  56250},
+	{2520,  4096,  37800,  6272,  42000,  6144,  37800},
+	{2700,  4096,  40500,  6272,  45000,  6144,  40500},
+	{2702,  8192,  81081,  6272,  45045,  8192,  54054},
+	{2703,  8192,  81081,  6272,  45045,  8192,  54054},
+	{5400,  4096,  81000,  6272,  90000,  6144,  81000},
+	{5405,  4096,  81081,  6272,  90090,  6144,  81081},
+	{7417, 11648, 316406, 17836, 351562, 11648, 210937},
+	{7425, 4096, 111375,  6272, 123750,  6144, 111375},
+	{14835, 11648, 632812, 17836, 703125, 11648, 421875},
+	{14850, 4096, 222750,  6272, 247500,  6144, 222750},
+	{29670, 5824, 632812,  8918, 703125,  5824, 421875},
+	{29700, 4096, 445500,  4704, 371250,  5120, 371250}
+};
+
+static const struct audio_clock_info audio_clock_info_table_48bpc[14] = {
+	{2517,  4576,  56250,  7007,  62500,  6864,  56250},
+	{2518,  4576,  56250,  7007,  62500,  6864,  56250},
+	{2520,  4096,  50400,  6272,  56000,  6144,  50400},
+	{2700,  4096,  54000,  6272,  60000,  6144,  54000},
+	{2702,  4096,  54054,  6267,  60060,  8192,  54054},
+	{2703,  4096,  54054,  6272,  60060,  8192,  54054},
+	{5400,  4096, 108000,  6272, 120000,  6144, 108000},
+	{5405,  4096, 108108,  6272, 120120,  6144, 108108},
+	{7417, 11648, 421875, 17836, 468750, 11648, 281250},
+	{7425,  4096, 148500,  6272, 165000,  6144, 148500},
+	{14835, 11648, 843750,  8918, 468750, 11648, 281250},
+	{14850, 4096, 297000,  6272, 330000,  6144, 297000},
+	{29670, 5824, 843750,  4459, 468750,  5824, 562500},
+	{29700, 3072, 445500,  4704, 495000,  5120, 495000}
+
+
+};
+
+static union audio_cea_channels speakers_to_channels(
+	struct audio_speaker_flags speaker_flags)
+{
+	union audio_cea_channels cea_channels = {0};
+
+	/* these are one to one */
+	cea_channels.channels.FL = speaker_flags.FL_FR;
+	cea_channels.channels.FR = speaker_flags.FL_FR;
+	cea_channels.channels.LFE = speaker_flags.LFE;
+	cea_channels.channels.FC = speaker_flags.FC;
+
+	/* if Rear Left and Right exist move RC speaker to channel 7
+	 * otherwise to channel 5
+	 */
+	if (speaker_flags.RL_RR) {
+		cea_channels.channels.RL_RC = speaker_flags.RL_RR;
+		cea_channels.channels.RR = speaker_flags.RL_RR;
+		cea_channels.channels.RC_RLC_FLC = speaker_flags.RC;
+	} else {
+		cea_channels.channels.RL_RC = speaker_flags.RC;
+	}
+
+	/* FRONT Left Right Center and REAR Left Right Center are exclusive */
+	if (speaker_flags.FLC_FRC) {
+		cea_channels.channels.RC_RLC_FLC = speaker_flags.FLC_FRC;
+		cea_channels.channels.RRC_FRC = speaker_flags.FLC_FRC;
+	} else {
+		cea_channels.channels.RC_RLC_FLC = speaker_flags.RLC_RRC;
+		cea_channels.channels.RRC_FRC = speaker_flags.RLC_RRC;
+	}
+
+	return cea_channels;
+}
+
+static void get_audio_clock_info(
+	enum dc_color_depth color_depth,
+	uint32_t crtc_pixel_clock_in_khz,
+	uint32_t actual_pixel_clock_in_khz,
+	struct audio_clock_info *audio_clock_info)
+{
+	const struct audio_clock_info *clock_info;
+	uint32_t index;
+	uint32_t crtc_pixel_clock_in_10khz = crtc_pixel_clock_in_khz / 10;
+	uint32_t audio_array_size;
+
+	switch (color_depth) {
+	case COLOR_DEPTH_161616:
+		clock_info = audio_clock_info_table_48bpc;
+		audio_array_size = ARRAY_SIZE(
+				audio_clock_info_table_48bpc);
+		break;
+	case COLOR_DEPTH_121212:
+		clock_info = audio_clock_info_table_36bpc;
+		audio_array_size = ARRAY_SIZE(
+				audio_clock_info_table_36bpc);
+		break;
+	default:
+		clock_info = audio_clock_info_table;
+		audio_array_size = ARRAY_SIZE(
+				audio_clock_info_table);
+		break;
+	}
+
+	if (clock_info != NULL) {
+		/* search for exact pixel clock in table */
+		for (index = 0; index < audio_array_size; index++) {
+			if (clock_info[index].pixel_clock_in_10khz >
+				crtc_pixel_clock_in_10khz)
+				break;  /* not match */
+			else if (clock_info[index].pixel_clock_in_10khz ==
+					crtc_pixel_clock_in_10khz) {
+				/* match found */
+				*audio_clock_info = clock_info[index];
+				return;
+			}
+		}
+	}
+
+	/* not found */
+	if (actual_pixel_clock_in_khz == 0)
+		actual_pixel_clock_in_khz = crtc_pixel_clock_in_khz;
+
+	/* See HDMI spec  the table entry under
+	 *  pixel clock of "Other". */
+	audio_clock_info->pixel_clock_in_10khz =
+			actual_pixel_clock_in_khz / 10;
+	audio_clock_info->cts_32khz = actual_pixel_clock_in_khz;
+	audio_clock_info->cts_44khz = actual_pixel_clock_in_khz;
+	audio_clock_info->cts_48khz = actual_pixel_clock_in_khz;
+
+	audio_clock_info->n_32khz = 4096;
+	audio_clock_info->n_44khz = 6272;
+	audio_clock_info->n_48khz = 6144;
+}
+
+static void enc1_se_audio_setup(
+	struct stream_encoder *enc,
+	unsigned int az_inst,
+	struct audio_info *audio_info)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	uint32_t speakers = 0;
+	uint32_t channels = 0;
+
+	ASSERT(audio_info);
+	if (audio_info == NULL)
+		/* This should not happen.it does so we don't get BSOD*/
+		return;
+
+	speakers = audio_info->flags.info.ALLSPEAKERS;
+	channels = speakers_to_channels(audio_info->flags.speaker_flags).all;
+
+	/* setup the audio stream source select (audio -> dig mapping) */
+	REG_SET(AFMT_AUDIO_SRC_CONTROL, 0, AFMT_AUDIO_SRC_SELECT, az_inst);
+
+	/* Channel allocation */
+	REG_UPDATE(AFMT_AUDIO_PACKET_CONTROL2, AFMT_AUDIO_CHANNEL_ENABLE, channels);
+}
+
+static void enc1_se_setup_hdmi_audio(
+	struct stream_encoder *enc,
+	const struct audio_crtc_info *crtc_info)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	struct audio_clock_info audio_clock_info = {0};
+
+	/* HDMI_AUDIO_PACKET_CONTROL */
+	REG_UPDATE(HDMI_AUDIO_PACKET_CONTROL,
+			HDMI_AUDIO_DELAY_EN, 1);
+
+	/* AFMT_AUDIO_PACKET_CONTROL */
+	REG_UPDATE(AFMT_AUDIO_PACKET_CONTROL, AFMT_60958_CS_UPDATE, 1);
+
+	/* AFMT_AUDIO_PACKET_CONTROL2 */
+	REG_UPDATE_2(AFMT_AUDIO_PACKET_CONTROL2,
+			AFMT_AUDIO_LAYOUT_OVRD, 0,
+			AFMT_60958_OSF_OVRD, 0);
+
+	/* HDMI_ACR_PACKET_CONTROL */
+	REG_UPDATE_3(HDMI_ACR_PACKET_CONTROL,
+			HDMI_ACR_AUTO_SEND, 1,
+			HDMI_ACR_SOURCE, 0,
+			HDMI_ACR_AUDIO_PRIORITY, 0);
+
+	/* Program audio clock sample/regeneration parameters */
+	get_audio_clock_info(crtc_info->color_depth,
+			     crtc_info->requested_pixel_clock,
+			     crtc_info->calculated_pixel_clock,
+			     &audio_clock_info);
+	DC_LOG_HW_AUDIO(
+			"\n%s:Input::requested_pixel_clock = %d"	\
+			"calculated_pixel_clock = %d \n", __func__,	\
+			crtc_info->requested_pixel_clock,		\
+			crtc_info->calculated_pixel_clock);
+
+	/* HDMI_ACR_32_0__HDMI_ACR_CTS_32_MASK */
+	REG_UPDATE(HDMI_ACR_32_0, HDMI_ACR_CTS_32, audio_clock_info.cts_32khz);
+
+	/* HDMI_ACR_32_1__HDMI_ACR_N_32_MASK */
+	REG_UPDATE(HDMI_ACR_32_1, HDMI_ACR_N_32, audio_clock_info.n_32khz);
+
+	/* HDMI_ACR_44_0__HDMI_ACR_CTS_44_MASK */
+	REG_UPDATE(HDMI_ACR_44_0, HDMI_ACR_CTS_44, audio_clock_info.cts_44khz);
+
+	/* HDMI_ACR_44_1__HDMI_ACR_N_44_MASK */
+	REG_UPDATE(HDMI_ACR_44_1, HDMI_ACR_N_44, audio_clock_info.n_44khz);
+
+	/* HDMI_ACR_48_0__HDMI_ACR_CTS_48_MASK */
+	REG_UPDATE(HDMI_ACR_48_0, HDMI_ACR_CTS_48, audio_clock_info.cts_48khz);
+
+	/* HDMI_ACR_48_1__HDMI_ACR_N_48_MASK */
+	REG_UPDATE(HDMI_ACR_48_1, HDMI_ACR_N_48, audio_clock_info.n_48khz);
+
+	/* Video driver cannot know in advance which sample rate will
+	 * be used by HD Audio driver
+	 * HDMI_ACR_PACKET_CONTROL__HDMI_ACR_N_MULTIPLE field is
+	 * programmed below in interruppt callback
+	 */
+
+	/* AFMT_60958_0__AFMT_60958_CS_CHANNEL_NUMBER_L_MASK &
+	 * AFMT_60958_0__AFMT_60958_CS_CLOCK_ACCURACY_MASK
+	 */
+	REG_UPDATE_2(AFMT_60958_0,
+			AFMT_60958_CS_CHANNEL_NUMBER_L, 1,
+			AFMT_60958_CS_CLOCK_ACCURACY, 0);
+
+	/* AFMT_60958_1 AFMT_60958_CS_CHALNNEL_NUMBER_R */
+	REG_UPDATE(AFMT_60958_1, AFMT_60958_CS_CHANNEL_NUMBER_R, 2);
+
+	/* AFMT_60958_2 now keep this settings until
+	 * Programming guide comes out
+	 */
+	REG_UPDATE_6(AFMT_60958_2,
+			AFMT_60958_CS_CHANNEL_NUMBER_2, 3,
+			AFMT_60958_CS_CHANNEL_NUMBER_3, 4,
+			AFMT_60958_CS_CHANNEL_NUMBER_4, 5,
+			AFMT_60958_CS_CHANNEL_NUMBER_5, 6,
+			AFMT_60958_CS_CHANNEL_NUMBER_6, 7,
+			AFMT_60958_CS_CHANNEL_NUMBER_7, 8);
+}
+
+static void enc1_se_setup_dp_audio(
+	struct stream_encoder *enc)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	/* --- DP Audio packet configurations --- */
+
+	/* ATP Configuration */
+	REG_SET(DP_SEC_AUD_N, 0,
+			DP_SEC_AUD_N, DP_SEC_AUD_N__DP_SEC_AUD_N__DEFAULT);
+
+	/* Async/auto-calc timestamp mode */
+	REG_SET(DP_SEC_TIMESTAMP, 0, DP_SEC_TIMESTAMP_MODE,
+			DP_SEC_TIMESTAMP__DP_SEC_TIMESTAMP_MODE__AUTO_CALC);
+
+	/* --- The following are the registers
+	 *  copied from the SetupHDMI ---
+	 */
+
+	/* AFMT_AUDIO_PACKET_CONTROL */
+	REG_UPDATE(AFMT_AUDIO_PACKET_CONTROL, AFMT_60958_CS_UPDATE, 1);
+
+	/* AFMT_AUDIO_PACKET_CONTROL2 */
+	/* Program the ATP and AIP next */
+	REG_UPDATE_2(AFMT_AUDIO_PACKET_CONTROL2,
+			AFMT_AUDIO_LAYOUT_OVRD, 0,
+			AFMT_60958_OSF_OVRD, 0);
+
+	/* AFMT_INFOFRAME_CONTROL0 */
+	REG_UPDATE(AFMT_INFOFRAME_CONTROL0, AFMT_AUDIO_INFO_UPDATE, 1);
+
+	/* AFMT_60958_0__AFMT_60958_CS_CLOCK_ACCURACY_MASK */
+	REG_UPDATE(AFMT_60958_0, AFMT_60958_CS_CLOCK_ACCURACY, 0);
+}
+
+static void enc1_se_enable_audio_clock(
+	struct stream_encoder *enc,
+	bool enable)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	if (REG(AFMT_CNTL) == 0)
+		return;   /* DCE8/10 does not have this register */
+
+	REG_UPDATE(AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, !!enable);
+
+	/* wait for AFMT clock to turn on,
+	 * expectation: this should complete in 1-2 reads
+	 *
+	 * REG_WAIT(AFMT_CNTL, AFMT_AUDIO_CLOCK_ON, !!enable, 1, 10);
+	 *
+	 * TODO: wait for clock_on does not work well. May need HW
+	 * program sequence. But audio seems work normally even without wait
+	 * for clock_on status change
+	 */
+}
+
+static void enc1_se_enable_dp_audio(
+	struct stream_encoder *enc)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	/* Enable Audio packets */
+	REG_UPDATE(DP_SEC_CNTL, DP_SEC_ASP_ENABLE, 1);
+
+	/* Program the ATP and AIP next */
+	REG_UPDATE_2(DP_SEC_CNTL,
+			DP_SEC_ATP_ENABLE, 1,
+			DP_SEC_AIP_ENABLE, 1);
+
+	/* Program STREAM_ENABLE after all the other enables. */
+	REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
+}
+
+static void enc1_se_disable_dp_audio(
+	struct stream_encoder *enc)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	uint32_t value = 0;
+
+	/* Disable Audio packets */
+	REG_UPDATE_5(DP_SEC_CNTL,
+			DP_SEC_ASP_ENABLE, 0,
+			DP_SEC_ATP_ENABLE, 0,
+			DP_SEC_AIP_ENABLE, 0,
+			DP_SEC_ACM_ENABLE, 0,
+			DP_SEC_STREAM_ENABLE, 0);
+
+	/* This register shared with encoder info frame. Therefore we need to
+	 * keep master enabled if at least on of the fields is not 0
+	 */
+	value = REG_READ(DP_SEC_CNTL);
+	if (value != 0)
+		REG_UPDATE(DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, 1);
+
+}
+
+void enc1_se_audio_mute_control(
+	struct stream_encoder *enc,
+	bool mute)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+
+	REG_UPDATE(AFMT_AUDIO_PACKET_CONTROL, AFMT_AUDIO_SAMPLE_SEND, !mute);
+}
+
+void enc1_se_dp_audio_setup(
+	struct stream_encoder *enc,
+	unsigned int az_inst,
+	struct audio_info *info)
+{
+	enc1_se_audio_setup(enc, az_inst, info);
+}
+
+void enc1_se_dp_audio_enable(
+	struct stream_encoder *enc)
+{
+	enc1_se_enable_audio_clock(enc, true);
+	enc1_se_setup_dp_audio(enc);
+	enc1_se_enable_dp_audio(enc);
+}
+
+void enc1_se_dp_audio_disable(
+	struct stream_encoder *enc)
+{
+	enc1_se_disable_dp_audio(enc);
+	enc1_se_enable_audio_clock(enc, false);
+}
+
+void enc1_se_hdmi_audio_setup(
+	struct stream_encoder *enc,
+	unsigned int az_inst,
+	struct audio_info *info,
+	struct audio_crtc_info *audio_crtc_info)
+{
+	enc1_se_enable_audio_clock(enc, true);
+	enc1_se_setup_hdmi_audio(enc, audio_crtc_info);
+	enc1_se_audio_setup(enc, az_inst, info);
+}
+
+void enc1_se_hdmi_audio_disable(
+	struct stream_encoder *enc)
+{
+	enc1_se_enable_audio_clock(enc, false);
+}
+
+
+void enc1_setup_stereo_sync(
+	struct stream_encoder *enc,
+	int tg_inst, bool enable)
+{
+	struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
+	REG_UPDATE(DIG_FE_CNTL, DIG_STEREOSYNC_SELECT, tg_inst);
+	REG_UPDATE(DIG_FE_CNTL, DIG_STEREOSYNC_GATE_EN, !enable);
+}
+
+
+static const struct stream_encoder_funcs dcn10_str_enc_funcs = {
+	.dp_set_stream_attribute =
+		enc1_stream_encoder_dp_set_stream_attribute,
+	.hdmi_set_stream_attribute =
+		enc1_stream_encoder_hdmi_set_stream_attribute,
+	.dvi_set_stream_attribute =
+		enc1_stream_encoder_dvi_set_stream_attribute,
+	.set_mst_bandwidth =
+		enc1_stream_encoder_set_mst_bandwidth,
+	.update_hdmi_info_packets =
+		enc1_stream_encoder_update_hdmi_info_packets,
+	.stop_hdmi_info_packets =
+		enc1_stream_encoder_stop_hdmi_info_packets,
+	.update_dp_info_packets =
+		enc1_stream_encoder_update_dp_info_packets,
+	.stop_dp_info_packets =
+		enc1_stream_encoder_stop_dp_info_packets,
+	.dp_blank =
+		enc1_stream_encoder_dp_blank,
+	.dp_unblank =
+		enc1_stream_encoder_dp_unblank,
+	.audio_mute_control = enc1_se_audio_mute_control,
+
+	.dp_audio_setup = enc1_se_dp_audio_setup,
+	.dp_audio_enable = enc1_se_dp_audio_enable,
+	.dp_audio_disable = enc1_se_dp_audio_disable,
+
+	.hdmi_audio_setup = enc1_se_hdmi_audio_setup,
+	.hdmi_audio_disable = enc1_se_hdmi_audio_disable,
+	.setup_stereo_sync  = enc1_setup_stereo_sync,
+	.set_avmute = enc1_stream_encoder_set_avmute,
+};
+
+void dcn10_stream_encoder_construct(
+	struct dcn10_stream_encoder *enc1,
+	struct dc_context *ctx,
+	struct dc_bios *bp,
+	enum engine_id eng_id,
+	const struct dcn10_stream_enc_registers *regs,
+	const struct dcn10_stream_encoder_shift *se_shift,
+	const struct dcn10_stream_encoder_mask *se_mask)
+{
+	enc1->base.funcs = &dcn10_str_enc_funcs;
+	enc1->base.ctx = ctx;
+	enc1->base.id = eng_id;
+	enc1->base.bp = bp;
+	enc1->regs = regs;
+	enc1->se_shift = se_shift;
+	enc1->se_mask = se_mask;
+}
+
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h
new file mode 100644
index 000000000000..6b3e4ded155b
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_stream_encoder.h
@@ -0,0 +1,524 @@
+/*
+ * Copyright 2012-15 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ *  and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __DC_STREAM_ENCODER_DCN10_H__
+#define __DC_STREAM_ENCODER_DCN10_H__
+
+#include "stream_encoder.h"
+
+#define DCN10STRENC_FROM_STRENC(stream_encoder)\
+	container_of(stream_encoder, struct dcn10_stream_encoder, base)
+
+#define SE_COMMON_DCN_REG_LIST(id) \
+	SRI(AFMT_CNTL, DIG, id), \
+	SRI(AFMT_GENERIC_0, DIG, id), \
+	SRI(AFMT_GENERIC_1, DIG, id), \
+	SRI(AFMT_GENERIC_2, DIG, id), \
+	SRI(AFMT_GENERIC_3, DIG, id), \
+	SRI(AFMT_GENERIC_4, DIG, id), \
+	SRI(AFMT_GENERIC_5, DIG, id), \
+	SRI(AFMT_GENERIC_6, DIG, id), \
+	SRI(AFMT_GENERIC_7, DIG, id), \
+	SRI(AFMT_GENERIC_HDR, DIG, id), \
+	SRI(AFMT_INFOFRAME_CONTROL0, DIG, id), \
+	SRI(AFMT_VBI_PACKET_CONTROL, DIG, id), \
+	SRI(AFMT_VBI_PACKET_CONTROL1, DIG, id), \
+	SRI(AFMT_AUDIO_PACKET_CONTROL, DIG, id), \
+	SRI(AFMT_AUDIO_PACKET_CONTROL2, DIG, id), \
+	SRI(AFMT_AUDIO_SRC_CONTROL, DIG, id), \
+	SRI(AFMT_60958_0, DIG, id), \
+	SRI(AFMT_60958_1, DIG, id), \
+	SRI(AFMT_60958_2, DIG, id), \
+	SRI(DIG_FE_CNTL, DIG, id), \
+	SRI(HDMI_CONTROL, DIG, id), \
+	SRI(HDMI_DB_CONTROL, DIG, id), \
+	SRI(HDMI_GC, DIG, id), \
+	SRI(HDMI_GENERIC_PACKET_CONTROL0, DIG, id), \
+	SRI(HDMI_GENERIC_PACKET_CONTROL1, DIG, id), \
+	SRI(HDMI_GENERIC_PACKET_CONTROL2, DIG, id), \
+	SRI(HDMI_GENERIC_PACKET_CONTROL3, DIG, id), \
+	SRI(HDMI_INFOFRAME_CONTROL0, DIG, id), \
+	SRI(HDMI_INFOFRAME_CONTROL1, DIG, id), \
+	SRI(HDMI_VBI_PACKET_CONTROL, DIG, id), \
+	SRI(HDMI_AUDIO_PACKET_CONTROL, DIG, id),\
+	SRI(HDMI_ACR_PACKET_CONTROL, DIG, id),\
+	SRI(HDMI_ACR_32_0, DIG, id),\
+	SRI(HDMI_ACR_32_1, DIG, id),\
+	SRI(HDMI_ACR_44_0, DIG, id),\
+	SRI(HDMI_ACR_44_1, DIG, id),\
+	SRI(HDMI_ACR_48_0, DIG, id),\
+	SRI(HDMI_ACR_48_1, DIG, id),\
+	SRI(DP_DB_CNTL, DP, id), \
+	SRI(DP_MSA_MISC, DP, id), \
+	SRI(DP_MSA_COLORIMETRY, DP, id), \
+	SRI(DP_MSA_TIMING_PARAM1, DP, id), \
+	SRI(DP_MSA_TIMING_PARAM2, DP, id), \
+	SRI(DP_MSA_TIMING_PARAM3, DP, id), \
+	SRI(DP_MSA_TIMING_PARAM4, DP, id), \
+	SRI(DP_MSE_RATE_CNTL, DP, id), \
+	SRI(DP_MSE_RATE_UPDATE, DP, id), \
+	SRI(DP_PIXEL_FORMAT, DP, id), \
+	SRI(DP_SEC_CNTL, DP, id), \
+	SRI(DP_STEER_FIFO, DP, id), \
+	SRI(DP_VID_M, DP, id), \
+	SRI(DP_VID_N, DP, id), \
+	SRI(DP_VID_STREAM_CNTL, DP, id), \
+	SRI(DP_VID_TIMING, DP, id), \
+	SRI(DP_SEC_AUD_N, DP, id), \
+	SRI(DP_SEC_TIMESTAMP, DP, id)
+
+#define SE_DCN_REG_LIST(id)\
+	SE_COMMON_DCN_REG_LIST(id)
+
+
+struct dcn10_stream_enc_registers {
+	uint32_t AFMT_CNTL;
+	uint32_t AFMT_AVI_INFO0;
+	uint32_t AFMT_AVI_INFO1;
+	uint32_t AFMT_AVI_INFO2;
+	uint32_t AFMT_AVI_INFO3;
+	uint32_t AFMT_GENERIC_0;
+	uint32_t AFMT_GENERIC_1;
+	uint32_t AFMT_GENERIC_2;
+	uint32_t AFMT_GENERIC_3;
+	uint32_t AFMT_GENERIC_4;
+	uint32_t AFMT_GENERIC_5;
+	uint32_t AFMT_GENERIC_6;
+	uint32_t AFMT_GENERIC_7;
+	uint32_t AFMT_GENERIC_HDR;
+	uint32_t AFMT_INFOFRAME_CONTROL0;
+	uint32_t AFMT_VBI_PACKET_CONTROL;
+	uint32_t AFMT_VBI_PACKET_CONTROL1;
+	uint32_t AFMT_AUDIO_PACKET_CONTROL;
+	uint32_t AFMT_AUDIO_PACKET_CONTROL2;
+	uint32_t AFMT_AUDIO_SRC_CONTROL;
+	uint32_t AFMT_60958_0;
+	uint32_t AFMT_60958_1;
+	uint32_t AFMT_60958_2;
+	uint32_t DIG_FE_CNTL;
+	uint32_t DP_MSE_RATE_CNTL;
+	uint32_t DP_MSE_RATE_UPDATE;
+	uint32_t DP_PIXEL_FORMAT;
+	uint32_t DP_SEC_CNTL;
+	uint32_t DP_STEER_FIFO;
+	uint32_t DP_VID_M;
+	uint32_t DP_VID_N;
+	uint32_t DP_VID_STREAM_CNTL;
+	uint32_t DP_VID_TIMING;
+	uint32_t DP_SEC_AUD_N;
+	uint32_t DP_SEC_TIMESTAMP;
+	uint32_t HDMI_CONTROL;
+	uint32_t HDMI_GC;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL0;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL1;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL2;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL3;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL4;
+	uint32_t HDMI_GENERIC_PACKET_CONTROL5;
+	uint32_t HDMI_INFOFRAME_CONTROL0;
+	uint32_t HDMI_INFOFRAME_CONTROL1;
+	uint32_t HDMI_VBI_PACKET_CONTROL;
+	uint32_t HDMI_AUDIO_PACKET_CONTROL;
+	uint32_t HDMI_ACR_PACKET_CONTROL;
+	uint32_t HDMI_ACR_32_0;
+	uint32_t HDMI_ACR_32_1;
+	uint32_t HDMI_ACR_44_0;
+	uint32_t HDMI_ACR_44_1;
+	uint32_t HDMI_ACR_48_0;
+	uint32_t HDMI_ACR_48_1;
+	uint32_t DP_DB_CNTL;
+	uint32_t DP_MSA_MISC;
+	uint32_t DP_MSA_COLORIMETRY;
+	uint32_t DP_MSA_TIMING_PARAM1;
+	uint32_t DP_MSA_TIMING_PARAM2;
+	uint32_t DP_MSA_TIMING_PARAM3;
+	uint32_t DP_MSA_TIMING_PARAM4;
+	uint32_t HDMI_DB_CONTROL;
+};
+
+
+#define SE_SF(reg_name, field_name, post_fix)\
+	.field_name = reg_name ## __ ## field_name ## post_fix
+
+#define SE_COMMON_MASK_SH_LIST_SOC_BASE(mask_sh)\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_INDEX, mask_sh),\
+	SE_SF(DIG0_AFMT_GENERIC_HDR, AFMT_GENERIC_HB0, mask_sh),\
+	SE_SF(DIG0_AFMT_GENERIC_HDR, AFMT_GENERIC_HB1, mask_sh),\
+	SE_SF(DIG0_AFMT_GENERIC_HDR, AFMT_GENERIC_HB2, mask_sh),\
+	SE_SF(DIG0_AFMT_GENERIC_HDR, AFMT_GENERIC_HB3, mask_sh),\
+	SE_SF(DP0_DP_PIXEL_FORMAT, DP_PIXEL_ENCODING, mask_sh),\
+	SE_SF(DP0_DP_PIXEL_FORMAT, DP_COMPONENT_DEPTH, mask_sh),\
+	SE_SF(DIG0_HDMI_CONTROL, HDMI_PACKET_GEN_VERSION, mask_sh),\
+	SE_SF(DIG0_HDMI_CONTROL, HDMI_KEEPOUT_MODE, mask_sh),\
+	SE_SF(DIG0_HDMI_CONTROL, HDMI_DEEP_COLOR_ENABLE, mask_sh),\
+	SE_SF(DIG0_HDMI_CONTROL, HDMI_DEEP_COLOR_DEPTH, mask_sh),\
+	SE_SF(DIG0_HDMI_CONTROL, HDMI_DATA_SCRAMBLE_EN, mask_sh),\
+	SE_SF(DIG0_HDMI_VBI_PACKET_CONTROL, HDMI_GC_CONT, mask_sh),\
+	SE_SF(DIG0_HDMI_VBI_PACKET_CONTROL, HDMI_GC_SEND, mask_sh),\
+	SE_SF(DIG0_HDMI_VBI_PACKET_CONTROL, HDMI_NULL_SEND, mask_sh),\
+	SE_SF(DIG0_HDMI_INFOFRAME_CONTROL0, HDMI_AUDIO_INFO_SEND, mask_sh),\
+	SE_SF(DIG0_AFMT_INFOFRAME_CONTROL0, AFMT_AUDIO_INFO_UPDATE, mask_sh),\
+	SE_SF(DIG0_HDMI_INFOFRAME_CONTROL1, HDMI_AUDIO_INFO_LINE, mask_sh),\
+	SE_SF(DIG0_HDMI_GC, HDMI_GC_AVMUTE, mask_sh),\
+	SE_SF(DP0_DP_MSE_RATE_CNTL, DP_MSE_RATE_X, mask_sh),\
+	SE_SF(DP0_DP_MSE_RATE_CNTL, DP_MSE_RATE_Y, mask_sh),\
+	SE_SF(DP0_DP_MSE_RATE_UPDATE, DP_MSE_RATE_UPDATE_PENDING, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP0_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_STREAM_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP1_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP2_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP3_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_MPG_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_VID_STREAM_CNTL, DP_VID_STREAM_DIS_DEFER, mask_sh),\
+	SE_SF(DP0_DP_VID_STREAM_CNTL, DP_VID_STREAM_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_VID_STREAM_CNTL, DP_VID_STREAM_STATUS, mask_sh),\
+	SE_SF(DP0_DP_STEER_FIFO, DP_STEER_FIFO_RESET, mask_sh),\
+	SE_SF(DP0_DP_VID_TIMING, DP_VID_M_N_GEN_EN, mask_sh),\
+	SE_SF(DP0_DP_VID_N, DP_VID_N, mask_sh),\
+	SE_SF(DP0_DP_VID_M, DP_VID_M, mask_sh),\
+	SE_SF(DIG0_DIG_FE_CNTL, DIG_START, mask_sh),\
+	SE_SF(DIG0_AFMT_AUDIO_SRC_CONTROL, AFMT_AUDIO_SRC_SELECT, mask_sh),\
+	SE_SF(DIG0_AFMT_AUDIO_PACKET_CONTROL2, AFMT_AUDIO_CHANNEL_ENABLE, mask_sh),\
+	SE_SF(DIG0_HDMI_AUDIO_PACKET_CONTROL, HDMI_AUDIO_PACKETS_PER_LINE, mask_sh),\
+	SE_SF(DIG0_HDMI_AUDIO_PACKET_CONTROL, HDMI_AUDIO_DELAY_EN, mask_sh),\
+	SE_SF(DIG0_AFMT_AUDIO_PACKET_CONTROL, AFMT_60958_CS_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_AUDIO_PACKET_CONTROL2, AFMT_AUDIO_LAYOUT_OVRD, mask_sh),\
+	SE_SF(DIG0_AFMT_AUDIO_PACKET_CONTROL2, AFMT_60958_OSF_OVRD, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_PACKET_CONTROL, HDMI_ACR_AUTO_SEND, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_PACKET_CONTROL, HDMI_ACR_SOURCE, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_PACKET_CONTROL, HDMI_ACR_AUDIO_PRIORITY, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_32_0, HDMI_ACR_CTS_32, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_32_1, HDMI_ACR_N_32, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_44_0, HDMI_ACR_CTS_44, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_44_1, HDMI_ACR_N_44, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_48_0, HDMI_ACR_CTS_48, mask_sh),\
+	SE_SF(DIG0_HDMI_ACR_48_1, HDMI_ACR_N_48, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_0, AFMT_60958_CS_CHANNEL_NUMBER_L, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_0, AFMT_60958_CS_CLOCK_ACCURACY, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_1, AFMT_60958_CS_CHANNEL_NUMBER_R, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_2, AFMT_60958_CS_CHANNEL_NUMBER_2, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_2, AFMT_60958_CS_CHANNEL_NUMBER_3, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_2, AFMT_60958_CS_CHANNEL_NUMBER_4, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_2, AFMT_60958_CS_CHANNEL_NUMBER_5, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_2, AFMT_60958_CS_CHANNEL_NUMBER_6, mask_sh),\
+	SE_SF(DIG0_AFMT_60958_2, AFMT_60958_CS_CHANNEL_NUMBER_7, mask_sh),\
+	SE_SF(DP0_DP_SEC_AUD_N, DP_SEC_AUD_N, mask_sh),\
+	SE_SF(DP0_DP_SEC_TIMESTAMP, DP_SEC_TIMESTAMP_MODE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_ASP_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_ATP_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_AIP_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_ACM_ENABLE, mask_sh),\
+	SE_SF(DIG0_AFMT_AUDIO_PACKET_CONTROL, AFMT_AUDIO_SAMPLE_SEND, mask_sh),\
+	SE_SF(DIG0_AFMT_CNTL, AFMT_AUDIO_CLOCK_EN, mask_sh),\
+	SE_SF(DIG0_HDMI_CONTROL, HDMI_CLOCK_CHANNEL_RATE, mask_sh),\
+	SE_SF(DIG0_DIG_FE_CNTL, TMDS_PIXEL_ENCODING, mask_sh),\
+	SE_SF(DIG0_DIG_FE_CNTL, TMDS_COLOR_FORMAT, mask_sh),\
+	SE_SF(DIG0_DIG_FE_CNTL, DIG_STEREOSYNC_SELECT, mask_sh),\
+	SE_SF(DIG0_DIG_FE_CNTL, DIG_STEREOSYNC_GATE_EN, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_LOCK_STATUS, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_CONFLICT, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL, AFMT_GENERIC_CONFLICT_CLR, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC0_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC1_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC2_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC3_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC4_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC5_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC6_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC7_FRAME_UPDATE_PENDING, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC0_FRAME_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC1_FRAME_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC2_FRAME_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC3_FRAME_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC4_FRAME_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC5_FRAME_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC6_FRAME_UPDATE, mask_sh),\
+	SE_SF(DIG0_AFMT_VBI_PACKET_CONTROL1, AFMT_GENERIC7_FRAME_UPDATE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP4_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP5_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP6_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_SEC_CNTL, DP_SEC_GSP7_ENABLE, mask_sh),\
+	SE_SF(DP0_DP_DB_CNTL, DP_DB_DISABLE, mask_sh),\
+	SE_SF(DP0_DP_MSA_COLORIMETRY, DP_MSA_MISC0, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM1, DP_MSA_HTOTAL, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM1, DP_MSA_VTOTAL, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM2, DP_MSA_HSTART, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM2, DP_MSA_VSTART, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM3, DP_MSA_HSYNCWIDTH, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM3, DP_MSA_HSYNCPOLARITY, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM3, DP_MSA_VSYNCWIDTH, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM3, DP_MSA_VSYNCPOLARITY, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM4, DP_MSA_HWIDTH, mask_sh),\
+	SE_SF(DP0_DP_MSA_TIMING_PARAM4, DP_MSA_VHEIGHT, mask_sh),\
+	SE_SF(DIG0_HDMI_DB_CONTROL, HDMI_DB_DISABLE, mask_sh),\
+	SE_SF(DP0_DP_VID_TIMING, DP_VID_N_MUL, mask_sh)
+
+#define SE_COMMON_MASK_SH_LIST_SOC(mask_sh)\
+	SE_COMMON_MASK_SH_LIST_SOC_BASE(mask_sh)
+
+#define SE_COMMON_MASK_SH_LIST_DCN10(mask_sh)\
+	SE_COMMON_MASK_SH_LIST_SOC(mask_sh),\
+	SE_SF(DIG0_HDMI_GENERIC_PACKET_CONTROL0, HDMI_GENERIC0_CONT, mask_sh),\
+	SE_SF(DIG0_HDMI_GENERIC_PACKET_CONTROL0, HDMI_GENERIC0_SEND, mask_sh),\
+	SE_SF(DIG0_HDMI_GENERIC_PACKET_CONTROL0, HDMI_GENERIC0_LINE, mask_sh),\
+	SE_SF(DIG0_HDMI_GENERIC_PACKET_CONTROL0, HDMI_GENERIC1_CONT, mask_sh),\
+	SE_SF(DIG0_HDMI_GENERIC_PACKET_CONTROL0, HDMI_GENERIC1_SEND, mask_sh),\
+	SE_SF(DIG0_HDMI_GENERIC_PACKET_CONTROL0, HDMI_GENERIC1_LINE, mask_sh)
+
+
+#define SE_REG_FIELD_LIST_DCN1_0(type) \
+	type AFMT_GENERIC_INDEX;\
+	type AFMT_GENERIC_HB0;\
+	type AFMT_GENERIC_HB1;\
+	type AFMT_GENERIC_HB2;\
+	type AFMT_GENERIC_HB3;\
+	type AFMT_GENERIC_LOCK_STATUS;\
+	type AFMT_GENERIC_CONFLICT;\
+	type AFMT_GENERIC_CONFLICT_CLR;\
+	type AFMT_GENERIC0_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC1_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC2_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC3_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC4_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC5_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC6_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC7_FRAME_UPDATE_PENDING;\
+	type AFMT_GENERIC0_FRAME_UPDATE;\
+	type AFMT_GENERIC1_FRAME_UPDATE;\
+	type AFMT_GENERIC2_FRAME_UPDATE;\
+	type AFMT_GENERIC3_FRAME_UPDATE;\
+	type AFMT_GENERIC4_FRAME_UPDATE;\
+	type AFMT_GENERIC5_FRAME_UPDATE;\
+	type AFMT_GENERIC6_FRAME_UPDATE;\
+	type AFMT_GENERIC7_FRAME_UPDATE;\
+	type HDMI_GENERIC0_CONT;\
+	type HDMI_GENERIC0_SEND;\
+	type HDMI_GENERIC0_LINE;\
+	type HDMI_GENERIC1_CONT;\
+	type HDMI_GENERIC1_SEND;\
+	type HDMI_GENERIC1_LINE;\
+	type HDMI_GENERIC2_CONT;\
+	type HDMI_GENERIC2_SEND;\
+	type HDMI_GENERIC2_LINE;\
+	type HDMI_GENERIC3_CONT;\
+	type HDMI_GENERIC3_SEND;\
+	type HDMI_GENERIC3_LINE;\
+	type HDMI_GENERIC4_CONT;\
+	type HDMI_GENERIC4_SEND;\
+	type HDMI_GENERIC4_LINE;\
+	type HDMI_GENERIC5_CONT;\
+	type HDMI_GENERIC5_SEND;\
+	type HDMI_GENERIC5_LINE;\
+	type HDMI_GENERIC6_CONT;\
+	type HDMI_GENERIC6_SEND;\
+	type HDMI_GENERIC6_LINE;\
+	type HDMI_GENERIC7_CONT;\
+	type HDMI_GENERIC7_SEND;\
+	type HDMI_GENERIC7_LINE;\
+	type DP_PIXEL_ENCODING;\
+	type DP_COMPONENT_DEPTH;\
+	type HDMI_PACKET_GEN_VERSION;\
+	type HDMI_KEEPOUT_MODE;\
+	type HDMI_DEEP_COLOR_ENABLE;\
+	type HDMI_CLOCK_CHANNEL_RATE;\
+	type HDMI_DEEP_COLOR_DEPTH;\
+	type HDMI_GC_CONT;\
+	type HDMI_GC_SEND;\
+	type HDMI_NULL_SEND;\
+	type HDMI_DATA_SCRAMBLE_EN;\
+	type HDMI_AUDIO_INFO_SEND;\
+	type AFMT_AUDIO_INFO_UPDATE;\
+	type HDMI_AUDIO_INFO_LINE;\
+	type HDMI_GC_AVMUTE;\
+	type DP_MSE_RATE_X;\
+	type DP_MSE_RATE_Y;\
+	type DP_MSE_RATE_UPDATE_PENDING;\
+	type DP_SEC_GSP0_ENABLE;\
+	type DP_SEC_STREAM_ENABLE;\
+	type DP_SEC_GSP1_ENABLE;\
+	type DP_SEC_GSP2_ENABLE;\
+	type DP_SEC_GSP3_ENABLE;\
+	type DP_SEC_GSP4_ENABLE;\
+	type DP_SEC_GSP5_ENABLE;\
+	type DP_SEC_GSP6_ENABLE;\
+	type DP_SEC_GSP7_ENABLE;\
+	type DP_SEC_MPG_ENABLE;\
+	type DP_VID_STREAM_DIS_DEFER;\
+	type DP_VID_STREAM_ENABLE;\
+	type DP_VID_STREAM_STATUS;\
+	type DP_STEER_FIFO_RESET;\
+	type DP_VID_M_N_GEN_EN;\
+	type DP_VID_N;\
+	type DP_VID_M;\
+	type DIG_START;\
+	type AFMT_AUDIO_SRC_SELECT;\
+	type AFMT_AUDIO_CHANNEL_ENABLE;\
+	type HDMI_AUDIO_PACKETS_PER_LINE;\
+	type HDMI_AUDIO_DELAY_EN;\
+	type AFMT_60958_CS_UPDATE;\
+	type AFMT_AUDIO_LAYOUT_OVRD;\
+	type AFMT_60958_OSF_OVRD;\
+	type HDMI_ACR_AUTO_SEND;\
+	type HDMI_ACR_SOURCE;\
+	type HDMI_ACR_AUDIO_PRIORITY;\
+	type HDMI_ACR_CTS_32;\
+	type HDMI_ACR_N_32;\
+	type HDMI_ACR_CTS_44;\
+	type HDMI_ACR_N_44;\
+	type HDMI_ACR_CTS_48;\
+	type HDMI_ACR_N_48;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_L;\
+	type AFMT_60958_CS_CLOCK_ACCURACY;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_R;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_2;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_3;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_4;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_5;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_6;\
+	type AFMT_60958_CS_CHANNEL_NUMBER_7;\
+	type DP_SEC_AUD_N;\
+	type DP_SEC_TIMESTAMP_MODE;\
+	type DP_SEC_ASP_ENABLE;\
+	type DP_SEC_ATP_ENABLE;\
+	type DP_SEC_AIP_ENABLE;\
+	type DP_SEC_ACM_ENABLE;\
+	type AFMT_AUDIO_SAMPLE_SEND;\
+	type AFMT_AUDIO_CLOCK_EN;\
+	type TMDS_PIXEL_ENCODING;\
+	type TMDS_COLOR_FORMAT;\
+	type DIG_STEREOSYNC_SELECT;\
+	type DIG_STEREOSYNC_GATE_EN;\
+	type DP_DB_DISABLE;\
+	type DP_MSA_MISC0;\
+	type DP_MSA_HTOTAL;\
+	type DP_MSA_VTOTAL;\
+	type DP_MSA_HSTART;\
+	type DP_MSA_VSTART;\
+	type DP_MSA_HSYNCWIDTH;\
+	type DP_MSA_HSYNCPOLARITY;\
+	type DP_MSA_VSYNCWIDTH;\
+	type DP_MSA_VSYNCPOLARITY;\
+	type DP_MSA_HWIDTH;\
+	type DP_MSA_VHEIGHT;\
+	type HDMI_DB_DISABLE;\
+	type DP_VID_N_MUL;\
+	type DP_VID_M_DOUBLE_VALUE_EN
+
+struct dcn10_stream_encoder_shift {
+	SE_REG_FIELD_LIST_DCN1_0(uint8_t);
+};
+
+struct dcn10_stream_encoder_mask {
+	SE_REG_FIELD_LIST_DCN1_0(uint32_t);
+};
+
+struct dcn10_stream_encoder {
+	struct stream_encoder base;
+	const struct dcn10_stream_enc_registers *regs;
+	const struct dcn10_stream_encoder_shift *se_shift;
+	const struct dcn10_stream_encoder_mask *se_mask;
+};
+
+void dcn10_stream_encoder_construct(
+	struct dcn10_stream_encoder *enc1,
+	struct dc_context *ctx,
+	struct dc_bios *bp,
+	enum engine_id eng_id,
+	const struct dcn10_stream_enc_registers *regs,
+	const struct dcn10_stream_encoder_shift *se_shift,
+	const struct dcn10_stream_encoder_mask *se_mask);
+
+void enc1_update_generic_info_packet(
+	struct dcn10_stream_encoder *enc1,
+	uint32_t packet_index,
+	const struct dc_info_packet *info_packet);
+
+void enc1_stream_encoder_dp_set_stream_attribute(
+	struct stream_encoder *enc,
+	struct dc_crtc_timing *crtc_timing,
+	enum dc_color_space output_color_space);
+
+void enc1_stream_encoder_hdmi_set_stream_attribute(
+	struct stream_encoder *enc,
+	struct dc_crtc_timing *crtc_timing,
+	int actual_pix_clk_khz,
+	bool enable_audio);
+
+void enc1_stream_encoder_dvi_set_stream_attribute(
+	struct stream_encoder *enc,
+	struct dc_crtc_timing *crtc_timing,
+	bool is_dual_link);
+
+void enc1_stream_encoder_set_mst_bandwidth(
+	struct stream_encoder *enc,
+	struct fixed31_32 avg_time_slots_per_mtp);
+
+void enc1_stream_encoder_update_dp_info_packets(
+	struct stream_encoder *enc,
+	const struct encoder_info_frame *info_frame);
+
+void enc1_stream_encoder_stop_dp_info_packets(
+	struct stream_encoder *enc);
+
+void enc1_stream_encoder_dp_blank(
+	struct stream_encoder *enc);
+
+void enc1_stream_encoder_dp_unblank(
+	struct stream_encoder *enc,
+	const struct encoder_unblank_param *param);
+
+void enc1_setup_stereo_sync(
+	struct stream_encoder *enc,
+	int tg_inst, bool enable);
+
+void enc1_stream_encoder_set_avmute(
+	struct stream_encoder *enc,
+	bool enable);
+
+void enc1_se_audio_mute_control(
+	struct stream_encoder *enc,
+	bool mute);
+
+void enc1_se_dp_audio_setup(
+	struct stream_encoder *enc,
+	unsigned int az_inst,
+	struct audio_info *info);
+
+void enc1_se_dp_audio_enable(
+	struct stream_encoder *enc);
+
+void enc1_se_dp_audio_disable(
+	struct stream_encoder *enc);
+
+void enc1_se_hdmi_audio_setup(
+	struct stream_encoder *enc,
+	unsigned int az_inst,
+	struct audio_info *info,
+	struct audio_crtc_info *audio_crtc_info);
+
+void enc1_se_hdmi_audio_disable(
+	struct stream_encoder *enc);
+
+#endif /* __DC_STREAM_ENCODER_DCN10_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dm_helpers.h b/drivers/gpu/drm/amd/display/dc/dm_helpers.h
index 034369fbb9e2..5d4527d03045 100644
--- a/drivers/gpu/drm/amd/display/dc/dm_helpers.h
+++ b/drivers/gpu/drm/amd/display/dc/dm_helpers.h
@@ -40,6 +40,14 @@ enum dc_edid_status dm_helpers_parse_edid_caps(
 	const struct dc_edid *edid,
 	struct dc_edid_caps *edid_caps);
 
+
+/*
+ * Update DP branch info
+ */
+void dm_helpers_dp_update_branch_info(
+		struct dc_context *ctx,
+		const struct dc_link *link);
+
 /*
  * Writes payload allocation table in immediate downstream device.
  */
@@ -103,6 +111,9 @@ bool dm_helpers_submit_i2c(
 		const struct dc_link *link,
 		struct i2c_command *cmd);
 
+bool dm_helpers_is_dp_sink_present(
+		struct dc_link *link);
+
 enum dc_edid_status dm_helpers_read_local_edid(
 		struct dc_context *ctx,
 		struct dc_link *link,
diff --git a/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h b/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h
index eac4bfe12257..58ed2055ef9f 100644
--- a/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h
+++ b/drivers/gpu/drm/amd/display/dc/dm_pp_smu.h
@@ -40,7 +40,7 @@ enum wm_set_id {
 	WM_B,
 	WM_C,
 	WM_D,
-	WM_COUNT,
+	WM_SET_COUNT,
 };
 
 struct pp_smu_wm_set_range {
@@ -53,10 +53,10 @@ struct pp_smu_wm_set_range {
 
 struct pp_smu_wm_range_sets {
 	uint32_t num_reader_wm_sets;
-	struct pp_smu_wm_set_range reader_wm_sets[WM_COUNT];
+	struct pp_smu_wm_set_range reader_wm_sets[WM_SET_COUNT];
 
 	uint32_t num_writer_wm_sets;
-	struct pp_smu_wm_set_range writer_wm_sets[WM_COUNT];
+	struct pp_smu_wm_set_range writer_wm_sets[WM_SET_COUNT];
 };
 
 struct pp_smu_display_requirement_rv {
diff --git a/drivers/gpu/drm/amd/display/dc/dm_services.h b/drivers/gpu/drm/amd/display/dc/dm_services.h
index 22e7ee7dcd26..eb5ab3978e84 100644
--- a/drivers/gpu/drm/amd/display/dc/dm_services.h
+++ b/drivers/gpu/drm/amd/display/dc/dm_services.h
@@ -339,7 +339,14 @@ bool dm_dmcu_set_pipe(struct dc_context *ctx, unsigned int controller_id);
 #define dm_log_to_buffer(buffer, size, fmt, args)\
 	vsnprintf(buffer, size, fmt, args)
 
-unsigned long long dm_get_timestamp(struct dc_context *ctx);
+static inline unsigned long long dm_get_timestamp(struct dc_context *ctx)
+{
+	return ktime_get_raw_ns();
+}
+
+unsigned long long dm_get_elapse_time_in_ns(struct dc_context *ctx,
+		unsigned long long current_time_stamp,
+		unsigned long long last_time_stamp);
 
 /*
  * performance tracing
@@ -351,10 +358,6 @@ void dm_perf_trace_timestamp(const char *func_name, unsigned int line);
 /*
  * Debug and verification hooks
  */
-bool dm_helpers_dc_conn_log(
-		struct dc_context *ctx,
-		struct log_entry *entry,
-		enum dc_log_type event);
 
 void dm_dtn_log_begin(struct dc_context *ctx);
 void dm_dtn_log_append_v(struct dc_context *ctx, const char *msg, ...);
diff --git a/drivers/gpu/drm/amd/display/dc/dm_services_types.h b/drivers/gpu/drm/amd/display/dc/dm_services_types.h
index ab8c77d4e6df..2b83f922ac02 100644
--- a/drivers/gpu/drm/amd/display/dc/dm_services_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dm_services_types.h
@@ -137,7 +137,7 @@ struct dm_pp_clock_range_for_wm_set {
 	enum dm_pp_wm_set_id wm_set_id;
 	uint32_t wm_min_eng_clk_in_khz;
 	uint32_t wm_max_eng_clk_in_khz;
-	uint32_t wm_min_memg_clk_in_khz;
+	uint32_t wm_min_mem_clk_in_khz;
 	uint32_t wm_max_mem_clk_in_khz;
 };
 
@@ -150,7 +150,7 @@ struct dm_pp_clock_range_for_dmif_wm_set_soc15 {
 	enum dm_pp_wm_set_id wm_set_id;
 	uint32_t wm_min_dcfclk_clk_in_khz;
 	uint32_t wm_max_dcfclk_clk_in_khz;
-	uint32_t wm_min_memg_clk_in_khz;
+	uint32_t wm_min_mem_clk_in_khz;
 	uint32_t wm_max_mem_clk_in_khz;
 };
 
@@ -158,7 +158,7 @@ struct dm_pp_clock_range_for_mcif_wm_set_soc15 {
 	enum dm_pp_wm_set_id wm_set_id;
 	uint32_t wm_min_socclk_clk_in_khz;
 	uint32_t wm_max_socclk_clk_in_khz;
-	uint32_t wm_min_memg_clk_in_khz;
+	uint32_t wm_min_mem_clk_in_khz;
 	uint32_t wm_max_mem_clk_in_khz;
 };
 
diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile
index f83a608f93e9..d97ca6528f9d 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile
@@ -36,11 +36,10 @@ CFLAGS_display_mode_lib.o := $(dml_ccflags)
 CFLAGS_display_pipe_clocks.o := $(dml_ccflags)
 CFLAGS_dml1_display_rq_dlg_calc.o := $(dml_ccflags)
 CFLAGS_display_rq_dlg_helpers.o := $(dml_ccflags)
-CFLAGS_soc_bounding_box.o := $(dml_ccflags)
 CFLAGS_dml_common_defs.o := $(dml_ccflags)
 
 DML = display_mode_lib.o display_rq_dlg_helpers.o dml1_display_rq_dlg_calc.o \
-	  soc_bounding_box.o dml_common_defs.o
+	dml_common_defs.o
 
 AMD_DAL_DML = $(addprefix $(AMDDALPATH)/dc/dml/,$(DML))
 
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_enums.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_enums.h
index b1ad3553f900..47c19f8fe7d1 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_enums.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_enums.h
@@ -108,4 +108,17 @@ enum output_standard {
 	dm_std_uninitialized = 0, dm_std_cvtr2, dm_std_cvt
 };
 
+enum mpc_combine_affinity {
+	dm_mpc_always_when_possible,
+	dm_mpc_reduce_voltage,
+	dm_mpc_reduce_voltage_and_clocks
+};
+
+enum self_refresh_affinity {
+	dm_try_to_allow_self_refresh_and_mclk_switch,
+	dm_allow_self_refresh_and_mclk_switch,
+	dm_allow_self_refresh,
+	dm_neither_self_refresh_nor_mclk_switch
+};
+
 #endif
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c
index c109b2c34c8f..dddeb0d4db8f 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.c
@@ -26,75 +26,30 @@
 #include "display_mode_lib.h"
 #include "dc_features.h"
 
+extern const struct _vcs_dpi_ip_params_st dcn1_0_ip;
+extern const struct _vcs_dpi_soc_bounding_box_st dcn1_0_soc;
+
 static void set_soc_bounding_box(struct _vcs_dpi_soc_bounding_box_st *soc, enum dml_project project)
 {
-	if (project == DML_PROJECT_RAVEN1) {
-		soc->sr_exit_time_us = 9.0;
-		soc->sr_enter_plus_exit_time_us = 11.0;
-		soc->urgent_latency_us = 4.0;
-		soc->writeback_latency_us = 12.0;
-		soc->ideal_dram_bw_after_urgent_percent = 80.0;
-		soc->max_request_size_bytes = 256;
-		soc->downspread_percent = 0.5;
-		soc->dram_page_open_time_ns = 50.0;
-		soc->dram_rw_turnaround_time_ns = 17.5;
-		soc->dram_return_buffer_per_channel_bytes = 8192;
-		soc->round_trip_ping_latency_dcfclk_cycles = 128;
-		soc->urgent_out_of_order_return_per_channel_bytes = 256;
-		soc->channel_interleave_bytes = 256;
-		soc->num_banks = 8;
-		soc->num_chans = 2;
-		soc->vmm_page_size_bytes = 4096;
-		soc->dram_clock_change_latency_us = 17.0;
-		soc->writeback_dram_clock_change_latency_us = 23.0;
-		soc->return_bus_width_bytes = 64;
-	} else {
-		BREAK_TO_DEBUGGER(); /* Invalid Project Specified */
+	switch (project) {
+	case DML_PROJECT_RAVEN1:
+		*soc = dcn1_0_soc;
+		break;
+	default:
+		ASSERT(0);
+		break;
 	}
 }
 
 static void set_ip_params(struct _vcs_dpi_ip_params_st *ip, enum dml_project project)
 {
-	if (project == DML_PROJECT_RAVEN1) {
-		ip->rob_buffer_size_kbytes = 64;
-		ip->det_buffer_size_kbytes = 164;
-		ip->dpte_buffer_size_in_pte_reqs = 42;
-		ip->dpp_output_buffer_pixels = 2560;
-		ip->opp_output_buffer_lines = 1;
-		ip->pixel_chunk_size_kbytes = 8;
-		ip->pte_enable = 1;
-		ip->pte_chunk_size_kbytes = 2;
-		ip->meta_chunk_size_kbytes = 2;
-		ip->writeback_chunk_size_kbytes = 2;
-		ip->line_buffer_size_bits = 589824;
-		ip->max_line_buffer_lines = 12;
-		ip->IsLineBufferBppFixed = 0;
-		ip->LineBufferFixedBpp = -1;
-		ip->writeback_luma_buffer_size_kbytes = 12;
-		ip->writeback_chroma_buffer_size_kbytes = 8;
-		ip->max_num_dpp = 4;
-		ip->max_num_wb = 2;
-		ip->max_dchub_pscl_bw_pix_per_clk = 4;
-		ip->max_pscl_lb_bw_pix_per_clk = 2;
-		ip->max_lb_vscl_bw_pix_per_clk = 4;
-		ip->max_vscl_hscl_bw_pix_per_clk = 4;
-		ip->max_hscl_ratio = 4;
-		ip->max_vscl_ratio = 4;
-		ip->hscl_mults = 4;
-		ip->vscl_mults = 4;
-		ip->max_hscl_taps = 8;
-		ip->max_vscl_taps = 8;
-		ip->dispclk_ramp_margin_percent = 1;
-		ip->underscan_factor = 1.10;
-		ip->min_vblank_lines = 14;
-		ip->dppclk_delay_subtotal = 90;
-		ip->dispclk_delay_subtotal = 42;
-		ip->dcfclk_cstate_latency = 10;
-		ip->max_inter_dcn_tile_repeaters = 8;
-		ip->can_vstartup_lines_exceed_vsync_plus_back_porch_lines_minus_one = 0;
-		ip->bug_forcing_LC_req_same_size_fixed = 0;
-	} else {
-		BREAK_TO_DEBUGGER(); /* Invalid Project Specified */
+	switch (project) {
+	case DML_PROJECT_RAVEN1:
+		*ip = dcn1_0_ip;
+		break;
+	default:
+		ASSERT(0);
+		break;
 	}
 }
 
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h
index 3c2abcb8a1b0..635206248889 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_lib.h
@@ -27,7 +27,6 @@
 
 
 #include "dml_common_defs.h"
-#include "soc_bounding_box.h"
 #include "dml1_display_rq_dlg_calc.h"
 
 enum dml_project {
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
index 09affa16cc43..cbafce649e33 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h
@@ -25,136 +25,155 @@
 #ifndef __DISPLAY_MODE_STRUCTS_H__
 #define __DISPLAY_MODE_STRUCTS_H__
 
-typedef struct _vcs_dpi_voltage_scaling_st	voltage_scaling_st;
-typedef struct _vcs_dpi_soc_bounding_box_st	soc_bounding_box_st;
-typedef struct _vcs_dpi_ip_params_st	ip_params_st;
-typedef struct _vcs_dpi_display_pipe_source_params_st	display_pipe_source_params_st;
-typedef struct _vcs_dpi_display_output_params_st	display_output_params_st;
-typedef struct _vcs_dpi_display_bandwidth_st	display_bandwidth_st;
-typedef struct _vcs_dpi_scaler_ratio_depth_st	scaler_ratio_depth_st;
-typedef struct _vcs_dpi_scaler_taps_st	scaler_taps_st;
-typedef struct _vcs_dpi_display_pipe_dest_params_st	display_pipe_dest_params_st;
-typedef struct _vcs_dpi_display_pipe_params_st	display_pipe_params_st;
-typedef struct _vcs_dpi_display_clocks_and_cfg_st	display_clocks_and_cfg_st;
-typedef struct _vcs_dpi_display_e2e_pipe_params_st	display_e2e_pipe_params_st;
-typedef struct _vcs_dpi_dchub_buffer_sizing_st	dchub_buffer_sizing_st;
-typedef struct _vcs_dpi_watermarks_perf_st	watermarks_perf_st;
-typedef struct _vcs_dpi_cstate_pstate_watermarks_st	cstate_pstate_watermarks_st;
-typedef struct _vcs_dpi_wm_calc_pipe_params_st	wm_calc_pipe_params_st;
-typedef struct _vcs_dpi_vratio_pre_st	vratio_pre_st;
-typedef struct _vcs_dpi_display_data_rq_misc_params_st	display_data_rq_misc_params_st;
-typedef struct _vcs_dpi_display_data_rq_sizing_params_st	display_data_rq_sizing_params_st;
-typedef struct _vcs_dpi_display_data_rq_dlg_params_st	display_data_rq_dlg_params_st;
-typedef struct _vcs_dpi_display_cur_rq_dlg_params_st	display_cur_rq_dlg_params_st;
-typedef struct _vcs_dpi_display_rq_dlg_params_st	display_rq_dlg_params_st;
-typedef struct _vcs_dpi_display_rq_sizing_params_st	display_rq_sizing_params_st;
-typedef struct _vcs_dpi_display_rq_misc_params_st	display_rq_misc_params_st;
-typedef struct _vcs_dpi_display_rq_params_st	display_rq_params_st;
-typedef struct _vcs_dpi_display_dlg_regs_st	display_dlg_regs_st;
-typedef struct _vcs_dpi_display_ttu_regs_st	display_ttu_regs_st;
-typedef struct _vcs_dpi_display_data_rq_regs_st	display_data_rq_regs_st;
-typedef struct _vcs_dpi_display_rq_regs_st	display_rq_regs_st;
-typedef struct _vcs_dpi_display_dlg_sys_params_st	display_dlg_sys_params_st;
-typedef struct _vcs_dpi_display_dlg_prefetch_param_st	display_dlg_prefetch_param_st;
-typedef struct _vcs_dpi_display_pipe_clock_st	display_pipe_clock_st;
-typedef struct _vcs_dpi_display_arb_params_st	display_arb_params_st;
+typedef struct _vcs_dpi_voltage_scaling_st voltage_scaling_st;
+typedef struct _vcs_dpi_soc_bounding_box_st soc_bounding_box_st;
+typedef struct _vcs_dpi_ip_params_st ip_params_st;
+typedef struct _vcs_dpi_display_pipe_source_params_st display_pipe_source_params_st;
+typedef struct _vcs_dpi_display_output_params_st display_output_params_st;
+typedef struct _vcs_dpi_display_bandwidth_st display_bandwidth_st;
+typedef struct _vcs_dpi_scaler_ratio_depth_st scaler_ratio_depth_st;
+typedef struct _vcs_dpi_scaler_taps_st scaler_taps_st;
+typedef struct _vcs_dpi_display_pipe_dest_params_st display_pipe_dest_params_st;
+typedef struct _vcs_dpi_display_pipe_params_st display_pipe_params_st;
+typedef struct _vcs_dpi_display_clocks_and_cfg_st display_clocks_and_cfg_st;
+typedef struct _vcs_dpi_display_e2e_pipe_params_st display_e2e_pipe_params_st;
+typedef struct _vcs_dpi_dchub_buffer_sizing_st dchub_buffer_sizing_st;
+typedef struct _vcs_dpi_watermarks_perf_st watermarks_perf_st;
+typedef struct _vcs_dpi_cstate_pstate_watermarks_st cstate_pstate_watermarks_st;
+typedef struct _vcs_dpi_wm_calc_pipe_params_st wm_calc_pipe_params_st;
+typedef struct _vcs_dpi_vratio_pre_st vratio_pre_st;
+typedef struct _vcs_dpi_display_data_rq_misc_params_st display_data_rq_misc_params_st;
+typedef struct _vcs_dpi_display_data_rq_sizing_params_st display_data_rq_sizing_params_st;
+typedef struct _vcs_dpi_display_data_rq_dlg_params_st display_data_rq_dlg_params_st;
+typedef struct _vcs_dpi_display_cur_rq_dlg_params_st display_cur_rq_dlg_params_st;
+typedef struct _vcs_dpi_display_rq_dlg_params_st display_rq_dlg_params_st;
+typedef struct _vcs_dpi_display_rq_sizing_params_st display_rq_sizing_params_st;
+typedef struct _vcs_dpi_display_rq_misc_params_st display_rq_misc_params_st;
+typedef struct _vcs_dpi_display_rq_params_st display_rq_params_st;
+typedef struct _vcs_dpi_display_dlg_regs_st display_dlg_regs_st;
+typedef struct _vcs_dpi_display_ttu_regs_st display_ttu_regs_st;
+typedef struct _vcs_dpi_display_data_rq_regs_st display_data_rq_regs_st;
+typedef struct _vcs_dpi_display_rq_regs_st display_rq_regs_st;
+typedef struct _vcs_dpi_display_dlg_sys_params_st display_dlg_sys_params_st;
+typedef struct _vcs_dpi_display_dlg_prefetch_param_st display_dlg_prefetch_param_st;
+typedef struct _vcs_dpi_display_pipe_clock_st display_pipe_clock_st;
+typedef struct _vcs_dpi_display_arb_params_st display_arb_params_st;
 
 struct _vcs_dpi_voltage_scaling_st {
 	int state;
 	double dscclk_mhz;
 	double dcfclk_mhz;
 	double socclk_mhz;
-	double dram_speed_mhz;
+	double dram_speed_mts;
 	double fabricclk_mhz;
 	double dispclk_mhz;
-	double dram_bw_per_chan_gbps;
 	double phyclk_mhz;
 	double dppclk_mhz;
 };
 
-struct	_vcs_dpi_soc_bounding_box_st	{
-	double	sr_exit_time_us;
-	double	sr_enter_plus_exit_time_us;
-	double	urgent_latency_us;
-	double	writeback_latency_us;
-	double	ideal_dram_bw_after_urgent_percent;
-	unsigned int	max_request_size_bytes;
-	double	downspread_percent;
-	double	dram_page_open_time_ns;
-	double	dram_rw_turnaround_time_ns;
-	double	dram_return_buffer_per_channel_bytes;
-	double	dram_channel_width_bytes;
+struct _vcs_dpi_soc_bounding_box_st {
+	double sr_exit_time_us;
+	double sr_enter_plus_exit_time_us;
+	double urgent_latency_us;
+	double urgent_latency_pixel_data_only_us;
+	double urgent_latency_pixel_mixed_with_vm_data_us;
+	double urgent_latency_vm_data_only_us;
+	double writeback_latency_us;
+	double ideal_dram_bw_after_urgent_percent;
+	double pct_ideal_dram_sdp_bw_after_urgent_pixel_only; // PercentOfIdealDRAMFabricAndSDPPortBWReceivedAfterUrgLatencyPixelDataOnly
+	double pct_ideal_dram_sdp_bw_after_urgent_pixel_and_vm;
+	double pct_ideal_dram_sdp_bw_after_urgent_vm_only;
+	double max_avg_sdp_bw_use_normal_percent;
+	double max_avg_dram_bw_use_normal_percent;
+	unsigned int max_request_size_bytes;
+	double downspread_percent;
+	double dram_page_open_time_ns;
+	double dram_rw_turnaround_time_ns;
+	double dram_return_buffer_per_channel_bytes;
+	double dram_channel_width_bytes;
 	double fabric_datapath_to_dcn_data_return_bytes;
 	double dcn_downspread_percent;
 	double dispclk_dppclk_vco_speed_mhz;
 	double dfs_vco_period_ps;
-	unsigned int	round_trip_ping_latency_dcfclk_cycles;
-	unsigned int	urgent_out_of_order_return_per_channel_bytes;
-	unsigned int	channel_interleave_bytes;
-	unsigned int	num_banks;
-	unsigned int	num_chans;
-	unsigned int	vmm_page_size_bytes;
-	double	dram_clock_change_latency_us;
-	double	writeback_dram_clock_change_latency_us;
-	unsigned int	return_bus_width_bytes;
-	unsigned int	voltage_override;
-	double	xfc_bus_transport_time_us;
-	double	xfc_xbuf_latency_tolerance_us;
+	unsigned int urgent_out_of_order_return_per_channel_pixel_only_bytes;
+	unsigned int urgent_out_of_order_return_per_channel_pixel_and_vm_bytes;
+	unsigned int urgent_out_of_order_return_per_channel_vm_only_bytes;
+	unsigned int round_trip_ping_latency_dcfclk_cycles;
+	unsigned int urgent_out_of_order_return_per_channel_bytes;
+	unsigned int channel_interleave_bytes;
+	unsigned int num_banks;
+	unsigned int num_chans;
+	unsigned int vmm_page_size_bytes;
+	double dram_clock_change_latency_us;
+	double writeback_dram_clock_change_latency_us;
+	unsigned int return_bus_width_bytes;
+	unsigned int voltage_override;
+	double xfc_bus_transport_time_us;
+	double xfc_xbuf_latency_tolerance_us;
+	int use_urgent_burst_bw;
+	double max_hscl_ratio;
+	double max_vscl_ratio;
 	struct _vcs_dpi_voltage_scaling_st clock_limits[7];
 };
 
-struct	_vcs_dpi_ip_params_st	{
-	unsigned int	max_inter_dcn_tile_repeaters;
-	unsigned int	num_dsc;
-	unsigned int	odm_capable;
-	unsigned int	rob_buffer_size_kbytes;
-	unsigned int	det_buffer_size_kbytes;
-	unsigned int	dpte_buffer_size_in_pte_reqs;
-	unsigned int	pde_proc_buffer_size_64k_reqs;
-	unsigned int	dpp_output_buffer_pixels;
-	unsigned int	opp_output_buffer_lines;
-	unsigned int	pixel_chunk_size_kbytes;
-	unsigned char	pte_enable;
-	unsigned int	pte_chunk_size_kbytes;
-	unsigned int	meta_chunk_size_kbytes;
-	unsigned int	writeback_chunk_size_kbytes;
-	unsigned int	line_buffer_size_bits;
-	unsigned int	max_line_buffer_lines;
-	unsigned int	writeback_luma_buffer_size_kbytes;
-	unsigned int	writeback_chroma_buffer_size_kbytes;
-	unsigned int	writeback_chroma_line_buffer_width_pixels;
-	unsigned int	max_page_table_levels;
-	unsigned int	max_num_dpp;
-	unsigned int	max_num_otg;
-	unsigned int	cursor_chunk_size;
-	unsigned int	cursor_buffer_size;
-	unsigned int	max_num_wb;
-	unsigned int	max_dchub_pscl_bw_pix_per_clk;
-	unsigned int	max_pscl_lb_bw_pix_per_clk;
-	unsigned int	max_lb_vscl_bw_pix_per_clk;
-	unsigned int	max_vscl_hscl_bw_pix_per_clk;
-	double	max_hscl_ratio;
-	double	max_vscl_ratio;
-	unsigned int	hscl_mults;
-	unsigned int	vscl_mults;
-	unsigned int	max_hscl_taps;
-	unsigned int	max_vscl_taps;
-	unsigned int	xfc_supported;
-	unsigned int	xfc_fill_constant_bytes;
-	double	dispclk_ramp_margin_percent;
-	double	xfc_fill_bw_overhead_percent;
-	double	underscan_factor;
-	unsigned int	min_vblank_lines;
-	unsigned int	dppclk_delay_subtotal;
-	unsigned int	dispclk_delay_subtotal;
-	unsigned int	dcfclk_cstate_latency;
-	unsigned int	dppclk_delay_scl;
-	unsigned int	dppclk_delay_scl_lb_only;
-	unsigned int	dppclk_delay_cnvc_formatter;
-	unsigned int	dppclk_delay_cnvc_cursor;
-	unsigned int	is_line_buffer_bpp_fixed;
-	unsigned int	line_buffer_fixed_bpp;
-	unsigned int	dcc_supported;
+struct _vcs_dpi_ip_params_st {
+	bool gpuvm_enable;
+	bool hostvm_enable;
+	unsigned int gpuvm_max_page_table_levels;
+	unsigned int hostvm_max_page_table_levels;
+	unsigned int hostvm_cached_page_table_levels;
+	unsigned int pte_group_size_bytes;
+	unsigned int max_inter_dcn_tile_repeaters;
+	unsigned int num_dsc;
+	unsigned int odm_capable;
+	unsigned int rob_buffer_size_kbytes;
+	unsigned int det_buffer_size_kbytes;
+	unsigned int dpte_buffer_size_in_pte_reqs;
+	unsigned int pde_proc_buffer_size_64k_reqs;
+	unsigned int dpp_output_buffer_pixels;
+	unsigned int opp_output_buffer_lines;
+	unsigned int pixel_chunk_size_kbytes;
+	unsigned char pte_enable;
+	unsigned int pte_chunk_size_kbytes;
+	unsigned int meta_chunk_size_kbytes;
+	unsigned int writeback_chunk_size_kbytes;
+	unsigned int line_buffer_size_bits;
+	unsigned int max_line_buffer_lines;
+	unsigned int writeback_luma_buffer_size_kbytes;
+	unsigned int writeback_chroma_buffer_size_kbytes;
+	unsigned int writeback_chroma_line_buffer_width_pixels;
+	unsigned int max_page_table_levels;
+	unsigned int max_num_dpp;
+	unsigned int max_num_otg;
+	unsigned int cursor_chunk_size;
+	unsigned int cursor_buffer_size;
+	unsigned int max_num_wb;
+	unsigned int max_dchub_pscl_bw_pix_per_clk;
+	unsigned int max_pscl_lb_bw_pix_per_clk;
+	unsigned int max_lb_vscl_bw_pix_per_clk;
+	unsigned int max_vscl_hscl_bw_pix_per_clk;
+	double max_hscl_ratio;
+	double max_vscl_ratio;
+	unsigned int hscl_mults;
+	unsigned int vscl_mults;
+	unsigned int max_hscl_taps;
+	unsigned int max_vscl_taps;
+	unsigned int xfc_supported;
+	unsigned int xfc_fill_constant_bytes;
+	double dispclk_ramp_margin_percent;
+	double xfc_fill_bw_overhead_percent;
+	double underscan_factor;
+	unsigned int min_vblank_lines;
+	unsigned int dppclk_delay_subtotal;
+	unsigned int dispclk_delay_subtotal;
+	unsigned int dcfclk_cstate_latency;
+	unsigned int dppclk_delay_scl;
+	unsigned int dppclk_delay_scl_lb_only;
+	unsigned int dppclk_delay_cnvc_formatter;
+	unsigned int dppclk_delay_cnvc_cursor;
+	unsigned int is_line_buffer_bpp_fixed;
+	unsigned int line_buffer_fixed_bpp;
+	unsigned int dcc_supported;
 
 	unsigned int IsLineBufferBppFixed;
 	unsigned int LineBufferFixedBpp;
@@ -169,41 +188,45 @@ struct _vcs_dpi_display_xfc_params_st {
 	int xfc_slv_chunk_size_bytes;
 };
 
-struct	_vcs_dpi_display_pipe_source_params_st	{
-	int	source_format;
-	unsigned char	dcc;
-	unsigned int	dcc_override;
-	unsigned int	dcc_rate;
-	unsigned char	dcc_use_global;
-	unsigned char	vm;
-	unsigned char	vm_levels_force_en;
-	unsigned int	vm_levels_force;
-	int	source_scan;
-	int	sw_mode;
-	int	macro_tile_size;
-	unsigned char	is_display_sw;
-	unsigned int	viewport_width;
-	unsigned int	viewport_height;
-	unsigned int	viewport_y_y;
-	unsigned int	viewport_y_c;
-	unsigned int	viewport_width_c;
-	unsigned int	viewport_height_c;
-	unsigned int	data_pitch;
-	unsigned int	data_pitch_c;
-	unsigned int	meta_pitch;
-	unsigned int	meta_pitch_c;
-	unsigned int	cur0_src_width;
-	int	cur0_bpp;
-	unsigned int	cur1_src_width;
-	int	cur1_bpp;
-	int	num_cursors;
-	unsigned char	is_hsplit;
-	unsigned char	dynamic_metadata_enable;
-	unsigned int	dynamic_metadata_lines_before_active;
-	unsigned int	dynamic_metadata_xmit_bytes;
-	unsigned int	hsplit_grp;
-	unsigned char	xfc_enable;
-	unsigned char	xfc_slave;
+struct _vcs_dpi_display_pipe_source_params_st {
+	int source_format;
+	unsigned char dcc;
+	unsigned int dcc_override;
+	unsigned int dcc_rate;
+	unsigned char dcc_use_global;
+	unsigned char vm;
+	bool gpuvm;    // gpuvm enabled
+	bool hostvm;    // hostvm enabled
+	bool gpuvm_levels_force_en;
+	unsigned int gpuvm_levels_force;
+	bool hostvm_levels_force_en;
+	unsigned int hostvm_levels_force;
+	int source_scan;
+	int sw_mode;
+	int macro_tile_size;
+	unsigned char is_display_sw;
+	unsigned int viewport_width;
+	unsigned int viewport_height;
+	unsigned int viewport_y_y;
+	unsigned int viewport_y_c;
+	unsigned int viewport_width_c;
+	unsigned int viewport_height_c;
+	unsigned int data_pitch;
+	unsigned int data_pitch_c;
+	unsigned int meta_pitch;
+	unsigned int meta_pitch_c;
+	unsigned int cur0_src_width;
+	int cur0_bpp;
+	unsigned int cur1_src_width;
+	int cur1_bpp;
+	int num_cursors;
+	unsigned char is_hsplit;
+	unsigned char dynamic_metadata_enable;
+	unsigned int dynamic_metadata_lines_before_active;
+	unsigned int dynamic_metadata_xmit_bytes;
+	unsigned int hsplit_grp;
+	unsigned char xfc_enable;
+	unsigned char xfc_slave;
 	struct _vcs_dpi_display_xfc_params_st xfc_params;
 };
 struct writeback_st {
@@ -215,338 +238,340 @@ struct writeback_st {
 	int wb_vtaps_luma;
 	int wb_htaps_chroma;
 	int wb_vtaps_chroma;
-	int wb_hratio;
-	int wb_vratio;
-};
-
-struct	_vcs_dpi_display_output_params_st	{
-	int	dp_lanes;
-	int	output_bpp;
-	int	dsc_enable;
-	int	wb_enable;
-	int	opp_input_bpc;
-	int	output_type;
-	int	output_format;
-	int	output_standard;
-	int	dsc_slices;
+	double wb_hratio;
+	double wb_vratio;
+};
+
+struct _vcs_dpi_display_output_params_st {
+	int dp_lanes;
+	int output_bpp;
+	int dsc_enable;
+	int wb_enable;
+	int num_active_wb;
+	int output_bpc;
+	int output_type;
+	int output_format;
+	int output_standard;
+	int dsc_slices;
 	struct writeback_st wb;
 };
 
-struct	_vcs_dpi_display_bandwidth_st	{
-	double	total_bw_consumed_gbps;
-	double	guaranteed_urgent_return_bw_gbps;
-};
-
-struct	_vcs_dpi_scaler_ratio_depth_st	{
-	double	hscl_ratio;
-	double	vscl_ratio;
-	double	hscl_ratio_c;
-	double	vscl_ratio_c;
-	double	vinit;
-	double	vinit_c;
-	double	vinit_bot;
-	double	vinit_bot_c;
-	int	lb_depth;
-	int	scl_enable;
-};
-
-struct	_vcs_dpi_scaler_taps_st	{
-	unsigned int	htaps;
-	unsigned int	vtaps;
-	unsigned int	htaps_c;
-	unsigned int	vtaps_c;
-};
-
-struct	_vcs_dpi_display_pipe_dest_params_st	{
-	unsigned int	recout_width;
-	unsigned int	recout_height;
-	unsigned int	full_recout_width;
-	unsigned int	full_recout_height;
-	unsigned int	hblank_start;
-	unsigned int	hblank_end;
-	unsigned int	vblank_start;
-	unsigned int	vblank_end;
-	unsigned int	htotal;
-	unsigned int	vtotal;
-	unsigned int	vactive;
-	unsigned int	hactive;
-	unsigned int	vstartup_start;
-	unsigned int	vupdate_offset;
-	unsigned int	vupdate_width;
-	unsigned int	vready_offset;
-	unsigned char	interlaced;
-	unsigned char	underscan;
-	double	pixel_rate_mhz;
-	unsigned char	synchronized_vblank_all_planes;
-	unsigned char	otg_inst;
-	unsigned char	odm_split_cnt;
-	unsigned char	odm_combine;
-};
-
-struct	_vcs_dpi_display_pipe_params_st	{
-	display_pipe_source_params_st	src;
-	display_pipe_dest_params_st	dest;
-	scaler_ratio_depth_st	scale_ratio_depth;
-	scaler_taps_st	scale_taps;
-};
-
-struct	_vcs_dpi_display_clocks_and_cfg_st	{
-	int	voltage;
-	double	dppclk_mhz;
-	double	refclk_mhz;
-	double	dispclk_mhz;
-	double	dcfclk_mhz;
-	double	socclk_mhz;
-};
-
-struct	_vcs_dpi_display_e2e_pipe_params_st	{
-	display_pipe_params_st	pipe;
-	display_output_params_st	dout;
-	display_clocks_and_cfg_st	clks_cfg;
-};
-
-struct	_vcs_dpi_dchub_buffer_sizing_st	{
-	unsigned int	swath_width_y;
-	unsigned int	swath_height_y;
-	unsigned int	swath_height_c;
-	unsigned int	detail_buffer_size_y;
-};
-
-struct	_vcs_dpi_watermarks_perf_st	{
-	double	stutter_eff_in_active_region_percent;
-	double	urgent_latency_supported_us;
-	double	non_urgent_latency_supported_us;
-	double	dram_clock_change_margin_us;
-	double	dram_access_eff_percent;
-};
-
-struct	_vcs_dpi_cstate_pstate_watermarks_st	{
-	double	cstate_exit_us;
-	double	cstate_enter_plus_exit_us;
-	double	pstate_change_us;
-};
-
-struct	_vcs_dpi_wm_calc_pipe_params_st	{
-	unsigned int	num_dpp;
-	int	voltage;
-	int	output_type;
-	double	dcfclk_mhz;
-	double	socclk_mhz;
-	double	dppclk_mhz;
-	double	pixclk_mhz;
-	unsigned char	interlace_en;
-	unsigned char	pte_enable;
-	unsigned char	dcc_enable;
-	double	dcc_rate;
-	double	bytes_per_pixel_c;
-	double	bytes_per_pixel_y;
-	unsigned int	swath_width_y;
-	unsigned int	swath_height_y;
-	unsigned int	swath_height_c;
-	unsigned int	det_buffer_size_y;
-	double	h_ratio;
-	double	v_ratio;
-	unsigned int	h_taps;
-	unsigned int	h_total;
-	unsigned int	v_total;
-	unsigned int	v_active;
-	unsigned int	e2e_index;
-	double	display_pipe_line_delivery_time;
-	double	read_bw;
-	unsigned int	lines_in_det_y;
-	unsigned int	lines_in_det_y_rounded_down_to_swath;
-	double	full_det_buffering_time;
-	double	dcfclk_deepsleep_mhz_per_plane;
-};
-
-struct	_vcs_dpi_vratio_pre_st	{
-	double	vratio_pre_l;
-	double	vratio_pre_c;
-};
-
-struct	_vcs_dpi_display_data_rq_misc_params_st	{
-	unsigned int	full_swath_bytes;
-	unsigned int	stored_swath_bytes;
-	unsigned int	blk256_height;
-	unsigned int	blk256_width;
-	unsigned int	req_height;
-	unsigned int	req_width;
-};
-
-struct	_vcs_dpi_display_data_rq_sizing_params_st	{
-	unsigned int	chunk_bytes;
-	unsigned int	min_chunk_bytes;
-	unsigned int	meta_chunk_bytes;
-	unsigned int	min_meta_chunk_bytes;
-	unsigned int	mpte_group_bytes;
-	unsigned int	dpte_group_bytes;
-};
-
-struct	_vcs_dpi_display_data_rq_dlg_params_st	{
-	unsigned int	swath_width_ub;
-	unsigned int	swath_height;
-	unsigned int	req_per_swath_ub;
-	unsigned int	meta_pte_bytes_per_frame_ub;
-	unsigned int	dpte_req_per_row_ub;
-	unsigned int	dpte_groups_per_row_ub;
-	unsigned int	dpte_row_height;
-	unsigned int	dpte_bytes_per_row_ub;
-	unsigned int	meta_chunks_per_row_ub;
-	unsigned int	meta_req_per_row_ub;
-	unsigned int	meta_row_height;
-	unsigned int	meta_bytes_per_row_ub;
-};
-
-struct	_vcs_dpi_display_cur_rq_dlg_params_st	{
-	unsigned char	enable;
-	unsigned int	swath_height;
-	unsigned int	req_per_line;
-};
-
-struct	_vcs_dpi_display_rq_dlg_params_st	{
-	display_data_rq_dlg_params_st	rq_l;
-	display_data_rq_dlg_params_st	rq_c;
-	display_cur_rq_dlg_params_st	rq_cur0;
-};
-
-struct	_vcs_dpi_display_rq_sizing_params_st	{
-	display_data_rq_sizing_params_st	rq_l;
-	display_data_rq_sizing_params_st	rq_c;
-};
-
-struct	_vcs_dpi_display_rq_misc_params_st	{
-	display_data_rq_misc_params_st	rq_l;
-	display_data_rq_misc_params_st	rq_c;
-};
-
-struct	_vcs_dpi_display_rq_params_st	{
-	unsigned char	yuv420;
-	unsigned char	yuv420_10bpc;
-	display_rq_misc_params_st	misc;
-	display_rq_sizing_params_st	sizing;
-	display_rq_dlg_params_st	dlg;
-};
-
-struct	_vcs_dpi_display_dlg_regs_st	{
-	unsigned int	refcyc_h_blank_end;
-	unsigned int	dlg_vblank_end;
-	unsigned int	min_dst_y_next_start;
-	unsigned int	refcyc_per_htotal;
-	unsigned int	refcyc_x_after_scaler;
-	unsigned int	dst_y_after_scaler;
-	unsigned int	dst_y_prefetch;
-	unsigned int	dst_y_per_vm_vblank;
-	unsigned int	dst_y_per_row_vblank;
-	unsigned int	dst_y_per_vm_flip;
-	unsigned int	dst_y_per_row_flip;
-	unsigned int	ref_freq_to_pix_freq;
-	unsigned int	vratio_prefetch;
-	unsigned int	vratio_prefetch_c;
-	unsigned int	refcyc_per_pte_group_vblank_l;
-	unsigned int	refcyc_per_pte_group_vblank_c;
-	unsigned int	refcyc_per_meta_chunk_vblank_l;
-	unsigned int	refcyc_per_meta_chunk_vblank_c;
-	unsigned int	refcyc_per_pte_group_flip_l;
-	unsigned int	refcyc_per_pte_group_flip_c;
-	unsigned int	refcyc_per_meta_chunk_flip_l;
-	unsigned int	refcyc_per_meta_chunk_flip_c;
-	unsigned int	dst_y_per_pte_row_nom_l;
-	unsigned int	dst_y_per_pte_row_nom_c;
-	unsigned int	refcyc_per_pte_group_nom_l;
-	unsigned int	refcyc_per_pte_group_nom_c;
-	unsigned int	dst_y_per_meta_row_nom_l;
-	unsigned int	dst_y_per_meta_row_nom_c;
-	unsigned int	refcyc_per_meta_chunk_nom_l;
-	unsigned int	refcyc_per_meta_chunk_nom_c;
-	unsigned int	refcyc_per_line_delivery_pre_l;
-	unsigned int	refcyc_per_line_delivery_pre_c;
-	unsigned int	refcyc_per_line_delivery_l;
-	unsigned int	refcyc_per_line_delivery_c;
-	unsigned int	chunk_hdl_adjust_cur0;
-	unsigned int	chunk_hdl_adjust_cur1;
-	unsigned int	vready_after_vcount0;
-	unsigned int	dst_y_offset_cur0;
-	unsigned int	dst_y_offset_cur1;
-	unsigned int	xfc_reg_transfer_delay;
-	unsigned int	xfc_reg_precharge_delay;
-	unsigned int	xfc_reg_remote_surface_flip_latency;
-	unsigned int	xfc_reg_prefetch_margin;
-	unsigned int	dst_y_delta_drq_limit;
-};
-
-struct	_vcs_dpi_display_ttu_regs_st	{
-	unsigned int	qos_level_low_wm;
-	unsigned int	qos_level_high_wm;
-	unsigned int	min_ttu_vblank;
-	unsigned int	qos_level_flip;
-	unsigned int	refcyc_per_req_delivery_l;
-	unsigned int	refcyc_per_req_delivery_c;
-	unsigned int	refcyc_per_req_delivery_cur0;
-	unsigned int	refcyc_per_req_delivery_cur1;
-	unsigned int	refcyc_per_req_delivery_pre_l;
-	unsigned int	refcyc_per_req_delivery_pre_c;
-	unsigned int	refcyc_per_req_delivery_pre_cur0;
-	unsigned int	refcyc_per_req_delivery_pre_cur1;
-	unsigned int	qos_level_fixed_l;
-	unsigned int	qos_level_fixed_c;
-	unsigned int	qos_level_fixed_cur0;
-	unsigned int	qos_level_fixed_cur1;
-	unsigned int	qos_ramp_disable_l;
-	unsigned int	qos_ramp_disable_c;
-	unsigned int	qos_ramp_disable_cur0;
-	unsigned int	qos_ramp_disable_cur1;
-};
-
-struct	_vcs_dpi_display_data_rq_regs_st	{
-	unsigned int	chunk_size;
-	unsigned int	min_chunk_size;
-	unsigned int	meta_chunk_size;
-	unsigned int	min_meta_chunk_size;
-	unsigned int	dpte_group_size;
-	unsigned int	mpte_group_size;
-	unsigned int	swath_height;
-	unsigned int	pte_row_height_linear;
-};
-
-struct	_vcs_dpi_display_rq_regs_st	{
-	display_data_rq_regs_st	rq_regs_l;
-	display_data_rq_regs_st	rq_regs_c;
-	unsigned int	drq_expansion_mode;
-	unsigned int	prq_expansion_mode;
-	unsigned int	mrq_expansion_mode;
-	unsigned int	crq_expansion_mode;
-	unsigned int	plane1_base_address;
-};
-
-struct	_vcs_dpi_display_dlg_sys_params_st	{
-	double	t_mclk_wm_us;
-	double	t_urg_wm_us;
-	double	t_sr_wm_us;
-	double	t_extra_us;
-	double	mem_trip_us;
-	double	t_srx_delay_us;
-	double	deepsleep_dcfclk_mhz;
-	double	total_flip_bw;
-	unsigned int	total_flip_bytes;
-};
-
-struct	_vcs_dpi_display_dlg_prefetch_param_st	{
-	double	prefetch_bw;
-	unsigned int	flip_bytes;
-};
-
-struct	_vcs_dpi_display_pipe_clock_st	{
-	double	dcfclk_mhz;
-	double	dispclk_mhz;
-	double	socclk_mhz;
-	double	dscclk_mhz[6];
-	double	dppclk_mhz[6];
-};
-
-struct	_vcs_dpi_display_arb_params_st	{
-	int	max_req_outstanding;
-	int	min_req_outstanding;
-	int	sat_level_us;
+struct _vcs_dpi_display_bandwidth_st {
+	double total_bw_consumed_gbps;
+	double guaranteed_urgent_return_bw_gbps;
+};
+
+struct _vcs_dpi_scaler_ratio_depth_st {
+	double hscl_ratio;
+	double vscl_ratio;
+	double hscl_ratio_c;
+	double vscl_ratio_c;
+	double vinit;
+	double vinit_c;
+	double vinit_bot;
+	double vinit_bot_c;
+	int lb_depth;
+	int scl_enable;
+};
+
+struct _vcs_dpi_scaler_taps_st {
+	unsigned int htaps;
+	unsigned int vtaps;
+	unsigned int htaps_c;
+	unsigned int vtaps_c;
+};
+
+struct _vcs_dpi_display_pipe_dest_params_st {
+	unsigned int recout_width;
+	unsigned int recout_height;
+	unsigned int full_recout_width;
+	unsigned int full_recout_height;
+	unsigned int hblank_start;
+	unsigned int hblank_end;
+	unsigned int vblank_start;
+	unsigned int vblank_end;
+	unsigned int htotal;
+	unsigned int vtotal;
+	unsigned int vactive;
+	unsigned int hactive;
+	unsigned int vstartup_start;
+	unsigned int vupdate_offset;
+	unsigned int vupdate_width;
+	unsigned int vready_offset;
+	unsigned char interlaced;
+	unsigned char underscan;
+	double pixel_rate_mhz;
+	unsigned char synchronized_vblank_all_planes;
+	unsigned char otg_inst;
+	unsigned char odm_split_cnt;
+	unsigned char odm_combine;
+	unsigned char use_maximum_vstartup;
+};
+
+struct _vcs_dpi_display_pipe_params_st {
+	display_pipe_source_params_st src;
+	display_pipe_dest_params_st dest;
+	scaler_ratio_depth_st scale_ratio_depth;
+	scaler_taps_st scale_taps;
+};
+
+struct _vcs_dpi_display_clocks_and_cfg_st {
+	int voltage;
+	double dppclk_mhz;
+	double refclk_mhz;
+	double dispclk_mhz;
+	double dcfclk_mhz;
+	double socclk_mhz;
+};
+
+struct _vcs_dpi_display_e2e_pipe_params_st {
+	display_pipe_params_st pipe;
+	display_output_params_st dout;
+	display_clocks_and_cfg_st clks_cfg;
+};
+
+struct _vcs_dpi_dchub_buffer_sizing_st {
+	unsigned int swath_width_y;
+	unsigned int swath_height_y;
+	unsigned int swath_height_c;
+	unsigned int detail_buffer_size_y;
+};
+
+struct _vcs_dpi_watermarks_perf_st {
+	double stutter_eff_in_active_region_percent;
+	double urgent_latency_supported_us;
+	double non_urgent_latency_supported_us;
+	double dram_clock_change_margin_us;
+	double dram_access_eff_percent;
+};
+
+struct _vcs_dpi_cstate_pstate_watermarks_st {
+	double cstate_exit_us;
+	double cstate_enter_plus_exit_us;
+	double pstate_change_us;
+};
+
+struct _vcs_dpi_wm_calc_pipe_params_st {
+	unsigned int num_dpp;
+	int voltage;
+	int output_type;
+	double dcfclk_mhz;
+	double socclk_mhz;
+	double dppclk_mhz;
+	double pixclk_mhz;
+	unsigned char interlace_en;
+	unsigned char pte_enable;
+	unsigned char dcc_enable;
+	double dcc_rate;
+	double bytes_per_pixel_c;
+	double bytes_per_pixel_y;
+	unsigned int swath_width_y;
+	unsigned int swath_height_y;
+	unsigned int swath_height_c;
+	unsigned int det_buffer_size_y;
+	double h_ratio;
+	double v_ratio;
+	unsigned int h_taps;
+	unsigned int h_total;
+	unsigned int v_total;
+	unsigned int v_active;
+	unsigned int e2e_index;
+	double display_pipe_line_delivery_time;
+	double read_bw;
+	unsigned int lines_in_det_y;
+	unsigned int lines_in_det_y_rounded_down_to_swath;
+	double full_det_buffering_time;
+	double dcfclk_deepsleep_mhz_per_plane;
+};
+
+struct _vcs_dpi_vratio_pre_st {
+	double vratio_pre_l;
+	double vratio_pre_c;
+};
+
+struct _vcs_dpi_display_data_rq_misc_params_st {
+	unsigned int full_swath_bytes;
+	unsigned int stored_swath_bytes;
+	unsigned int blk256_height;
+	unsigned int blk256_width;
+	unsigned int req_height;
+	unsigned int req_width;
+};
+
+struct _vcs_dpi_display_data_rq_sizing_params_st {
+	unsigned int chunk_bytes;
+	unsigned int min_chunk_bytes;
+	unsigned int meta_chunk_bytes;
+	unsigned int min_meta_chunk_bytes;
+	unsigned int mpte_group_bytes;
+	unsigned int dpte_group_bytes;
+};
+
+struct _vcs_dpi_display_data_rq_dlg_params_st {
+	unsigned int swath_width_ub;
+	unsigned int swath_height;
+	unsigned int req_per_swath_ub;
+	unsigned int meta_pte_bytes_per_frame_ub;
+	unsigned int dpte_req_per_row_ub;
+	unsigned int dpte_groups_per_row_ub;
+	unsigned int dpte_row_height;
+	unsigned int dpte_bytes_per_row_ub;
+	unsigned int meta_chunks_per_row_ub;
+	unsigned int meta_req_per_row_ub;
+	unsigned int meta_row_height;
+	unsigned int meta_bytes_per_row_ub;
+};
+
+struct _vcs_dpi_display_cur_rq_dlg_params_st {
+	unsigned char enable;
+	unsigned int swath_height;
+	unsigned int req_per_line;
+};
+
+struct _vcs_dpi_display_rq_dlg_params_st {
+	display_data_rq_dlg_params_st rq_l;
+	display_data_rq_dlg_params_st rq_c;
+	display_cur_rq_dlg_params_st rq_cur0;
+};
+
+struct _vcs_dpi_display_rq_sizing_params_st {
+	display_data_rq_sizing_params_st rq_l;
+	display_data_rq_sizing_params_st rq_c;
+};
+
+struct _vcs_dpi_display_rq_misc_params_st {
+	display_data_rq_misc_params_st rq_l;
+	display_data_rq_misc_params_st rq_c;
+};
+
+struct _vcs_dpi_display_rq_params_st {
+	unsigned char yuv420;
+	unsigned char yuv420_10bpc;
+	display_rq_misc_params_st misc;
+	display_rq_sizing_params_st sizing;
+	display_rq_dlg_params_st dlg;
+};
+
+struct _vcs_dpi_display_dlg_regs_st {
+	unsigned int refcyc_h_blank_end;
+	unsigned int dlg_vblank_end;
+	unsigned int min_dst_y_next_start;
+	unsigned int refcyc_per_htotal;
+	unsigned int refcyc_x_after_scaler;
+	unsigned int dst_y_after_scaler;
+	unsigned int dst_y_prefetch;
+	unsigned int dst_y_per_vm_vblank;
+	unsigned int dst_y_per_row_vblank;
+	unsigned int dst_y_per_vm_flip;
+	unsigned int dst_y_per_row_flip;
+	unsigned int ref_freq_to_pix_freq;
+	unsigned int vratio_prefetch;
+	unsigned int vratio_prefetch_c;
+	unsigned int refcyc_per_pte_group_vblank_l;
+	unsigned int refcyc_per_pte_group_vblank_c;
+	unsigned int refcyc_per_meta_chunk_vblank_l;
+	unsigned int refcyc_per_meta_chunk_vblank_c;
+	unsigned int refcyc_per_pte_group_flip_l;
+	unsigned int refcyc_per_pte_group_flip_c;
+	unsigned int refcyc_per_meta_chunk_flip_l;
+	unsigned int refcyc_per_meta_chunk_flip_c;
+	unsigned int dst_y_per_pte_row_nom_l;
+	unsigned int dst_y_per_pte_row_nom_c;
+	unsigned int refcyc_per_pte_group_nom_l;
+	unsigned int refcyc_per_pte_group_nom_c;
+	unsigned int dst_y_per_meta_row_nom_l;
+	unsigned int dst_y_per_meta_row_nom_c;
+	unsigned int refcyc_per_meta_chunk_nom_l;
+	unsigned int refcyc_per_meta_chunk_nom_c;
+	unsigned int refcyc_per_line_delivery_pre_l;
+	unsigned int refcyc_per_line_delivery_pre_c;
+	unsigned int refcyc_per_line_delivery_l;
+	unsigned int refcyc_per_line_delivery_c;
+	unsigned int chunk_hdl_adjust_cur0;
+	unsigned int chunk_hdl_adjust_cur1;
+	unsigned int vready_after_vcount0;
+	unsigned int dst_y_offset_cur0;
+	unsigned int dst_y_offset_cur1;
+	unsigned int xfc_reg_transfer_delay;
+	unsigned int xfc_reg_precharge_delay;
+	unsigned int xfc_reg_remote_surface_flip_latency;
+	unsigned int xfc_reg_prefetch_margin;
+	unsigned int dst_y_delta_drq_limit;
+};
+
+struct _vcs_dpi_display_ttu_regs_st {
+	unsigned int qos_level_low_wm;
+	unsigned int qos_level_high_wm;
+	unsigned int min_ttu_vblank;
+	unsigned int qos_level_flip;
+	unsigned int refcyc_per_req_delivery_l;
+	unsigned int refcyc_per_req_delivery_c;
+	unsigned int refcyc_per_req_delivery_cur0;
+	unsigned int refcyc_per_req_delivery_cur1;
+	unsigned int refcyc_per_req_delivery_pre_l;
+	unsigned int refcyc_per_req_delivery_pre_c;
+	unsigned int refcyc_per_req_delivery_pre_cur0;
+	unsigned int refcyc_per_req_delivery_pre_cur1;
+	unsigned int qos_level_fixed_l;
+	unsigned int qos_level_fixed_c;
+	unsigned int qos_level_fixed_cur0;
+	unsigned int qos_level_fixed_cur1;
+	unsigned int qos_ramp_disable_l;
+	unsigned int qos_ramp_disable_c;
+	unsigned int qos_ramp_disable_cur0;
+	unsigned int qos_ramp_disable_cur1;
+};
+
+struct _vcs_dpi_display_data_rq_regs_st {
+	unsigned int chunk_size;
+	unsigned int min_chunk_size;
+	unsigned int meta_chunk_size;
+	unsigned int min_meta_chunk_size;
+	unsigned int dpte_group_size;
+	unsigned int mpte_group_size;
+	unsigned int swath_height;
+	unsigned int pte_row_height_linear;
+};
+
+struct _vcs_dpi_display_rq_regs_st {
+	display_data_rq_regs_st rq_regs_l;
+	display_data_rq_regs_st rq_regs_c;
+	unsigned int drq_expansion_mode;
+	unsigned int prq_expansion_mode;
+	unsigned int mrq_expansion_mode;
+	unsigned int crq_expansion_mode;
+	unsigned int plane1_base_address;
+};
+
+struct _vcs_dpi_display_dlg_sys_params_st {
+	double t_mclk_wm_us;
+	double t_urg_wm_us;
+	double t_sr_wm_us;
+	double t_extra_us;
+	double mem_trip_us;
+	double t_srx_delay_us;
+	double deepsleep_dcfclk_mhz;
+	double total_flip_bw;
+	unsigned int total_flip_bytes;
+};
+
+struct _vcs_dpi_display_dlg_prefetch_param_st {
+	double prefetch_bw;
+	unsigned int flip_bytes;
+};
+
+struct _vcs_dpi_display_pipe_clock_st {
+	double dcfclk_mhz;
+	double dispclk_mhz;
+	double socclk_mhz;
+	double dscclk_mhz[6];
+	double dppclk_mhz[6];
+};
+
+struct _vcs_dpi_display_arb_params_st {
+	int max_req_outstanding;
+	int min_req_outstanding;
+	int sat_level_us;
 };
 
 #endif /*__DISPLAY_MODE_STRUCTS_H__*/
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h b/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h
index f9cf08357989..e8ce08567cd8 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h
@@ -35,6 +35,16 @@ static inline double dml_min(double a, double b)
 	return (double) dcn_bw_min2(a, b);
 }
 
+static inline double dml_min3(double a, double b, double c)
+{
+	return dml_min(dml_min(a, b), c);
+}
+
+static inline double dml_min4(double a, double b, double c, double d)
+{
+	return dml_min(dml_min(a, b), dml_min(c, d));
+}
+
 static inline double dml_max(double a, double b)
 {
 	return (double) dcn_bw_max2(a, b);
diff --git a/drivers/gpu/drm/amd/display/dc/dml/soc_bounding_box.c b/drivers/gpu/drm/amd/display/dc/dml/soc_bounding_box.c
deleted file mode 100644
index 324239c77958..000000000000
--- a/drivers/gpu/drm/amd/display/dc/dml/soc_bounding_box.c
+++ /dev/null
@@ -1,79 +0,0 @@
-/*
- * Copyright 2017 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: AMD
- *
- */
-#include "soc_bounding_box.h"
-#include "display_mode_lib.h"
-#include "dc_features.h"
-
-#include "dml_inline_defs.h"
-
-/*
- * NOTE:
- *   This file is gcc-parseable HW gospel, coming straight from HW engineers.
- *
- * It doesn't adhere to Linux kernel style and sometimes will do things in odd
- * ways. Unless there is something clearly wrong with it the code should
- * remain as-is as it provides us with a guarantee from HW that it is correct.
- */
-
-void dml_socbb_set_latencies(soc_bounding_box_st *to_box, soc_bounding_box_st *from_box)
-{
-	to_box->dram_clock_change_latency_us = from_box->dram_clock_change_latency_us;
-	to_box->sr_exit_time_us = from_box->sr_exit_time_us;
-	to_box->sr_enter_plus_exit_time_us = from_box->sr_enter_plus_exit_time_us;
-	to_box->urgent_latency_us = from_box->urgent_latency_us;
-	to_box->writeback_latency_us = from_box->writeback_latency_us;
-}
-
-voltage_scaling_st dml_socbb_voltage_scaling(
-		const soc_bounding_box_st *soc,
-		enum voltage_state voltage)
-{
-	const voltage_scaling_st *voltage_state;
-	const voltage_scaling_st * const voltage_end = soc->clock_limits + DC__VOLTAGE_STATES;
-
-	for (voltage_state = soc->clock_limits;
-			voltage_state < voltage_end && voltage_state->state != voltage;
-			voltage_state++) {
-	}
-
-	if (voltage_state < voltage_end)
-		return *voltage_state;
-	return soc->clock_limits[DC__VOLTAGE_STATES - 1];
-}
-
-double dml_socbb_return_bw_mhz(soc_bounding_box_st *box, enum voltage_state voltage)
-{
-	double return_bw;
-
-	voltage_scaling_st state = dml_socbb_voltage_scaling(box, voltage);
-
-	return_bw = dml_min((double) box->return_bus_width_bytes * state.dcfclk_mhz,
-			state.dram_bw_per_chan_gbps * 1000.0 * (double) box->num_chans
-					* box->ideal_dram_bw_after_urgent_percent / 100.0);
-
-	return_bw = dml_min((double) box->return_bus_width_bytes * state.fabricclk_mhz, return_bw);
-
-	return return_bw;
-}
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/Makefile b/drivers/gpu/drm/amd/display/dc/gpio/Makefile
index 562ee189d780..b9d9930a4974 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/gpio/Makefile
@@ -61,7 +61,7 @@ AMD_DISPLAY_FILES += $(AMD_DAL_GPIO_DCE120)
 ###############################################################################
 # DCN 1x
 ###############################################################################
-ifdef CONFIG_DRM_AMD_DC_DCN1_0
+ifdef CONFIG_X86
 GPIO_DCN10 = hw_translate_dcn10.o hw_factory_dcn10.o
 
 AMD_DAL_GPIO_DCN10 = $(addprefix $(AMDDALPATH)/dc/gpio/dcn10/,$(GPIO_DCN10))
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/ddc_regs.h b/drivers/gpu/drm/amd/display/dc/gpio/ddc_regs.h
index 9c4a56c738c0..bf40725f982f 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/ddc_regs.h
+++ b/drivers/gpu/drm/amd/display/dc/gpio/ddc_regs.h
@@ -82,13 +82,16 @@
 	DDC_GPIO_I2C_REG_LIST(cd),\
 	.ddc_setup = 0
 
-#define DDC_MASK_SH_LIST(mask_sh) \
+#define DDC_MASK_SH_LIST_COMMON(mask_sh) \
 		SF_DDC(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_ENABLE, mask_sh),\
 		SF_DDC(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_EDID_DETECT_ENABLE, mask_sh),\
 		SF_DDC(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_EDID_DETECT_MODE, mask_sh),\
 		SF_DDC(DC_GPIO_DDC1_MASK, DC_GPIO_DDC1DATA_PD_EN, mask_sh),\
 		SF_DDC(DC_GPIO_DDC1_MASK, DC_GPIO_DDC1CLK_PD_EN, mask_sh),\
-		SF_DDC(DC_GPIO_DDC1_MASK, AUX_PAD1_MODE, mask_sh),\
+		SF_DDC(DC_GPIO_DDC1_MASK, AUX_PAD1_MODE, mask_sh)
+
+#define DDC_MASK_SH_LIST(mask_sh) \
+		DDC_MASK_SH_LIST_COMMON(mask_sh),\
 		SF_DDC(DC_GPIO_I2CPAD_MASK, DC_GPIO_SDA_PD_DIS, mask_sh),\
 		SF_DDC(DC_GPIO_I2CPAD_MASK, DC_GPIO_SCL_PD_DIS, mask_sh)
 
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c b/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c
index 80038e0e610f..f20161c5706d 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c
+++ b/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c
@@ -98,7 +98,8 @@ struct gpio_service *dal_gpio_service_create(
 			if (number_of_bits) {
 				uint32_t index_of_uint = 0;
 
-				slot = kzalloc(number_of_uints * sizeof(uint32_t),
+				slot = kcalloc(number_of_uints,
+					       sizeof(uint32_t),
 					       GFP_KERNEL);
 
 				if (!slot) {
@@ -374,6 +375,7 @@ struct gpio *dal_gpio_create_irq(
 	case GPIO_ID_GPIO_PAD:
 	break;
 	default:
+		id = GPIO_ID_HPD;
 		ASSERT_CRITICAL(false);
 		return NULL;
 	}
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c b/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c
index 87b580fa4bc9..83df779984e5 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c
+++ b/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c
@@ -43,7 +43,7 @@
 #include "dce80/hw_factory_dce80.h"
 #include "dce110/hw_factory_dce110.h"
 #include "dce120/hw_factory_dce120.h"
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 #include "dcn10/hw_factory_dcn10.h"
 #endif
 
@@ -75,12 +75,13 @@ bool dal_hw_factory_init(
 		return true;
 	case DCE_VERSION_11_0:
 	case DCE_VERSION_11_2:
+	case DCE_VERSION_11_22:
 		dal_hw_factory_dce110_init(factory);
 		return true;
 	case DCE_VERSION_12_0:
 		dal_hw_factory_dce120_init(factory);
 		return true;
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	case DCN_VERSION_1_0:
 		dal_hw_factory_dcn10_init(factory);
 		return true;
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c b/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c
index 0ae8ace25739..e7541310480b 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c
+++ b/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c
@@ -43,7 +43,7 @@
 #include "dce80/hw_translate_dce80.h"
 #include "dce110/hw_translate_dce110.h"
 #include "dce120/hw_translate_dce120.h"
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 #include "dcn10/hw_translate_dcn10.h"
 #endif
 
@@ -72,12 +72,13 @@ bool dal_hw_translate_init(
 	case DCE_VERSION_10_0:
 	case DCE_VERSION_11_0:
 	case DCE_VERSION_11_2:
+	case DCE_VERSION_11_22:
 		dal_hw_translate_dce110_init(translate);
 		return true;
 	case DCE_VERSION_12_0:
 		dal_hw_translate_dce120_init(translate);
 		return true;
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	case DCN_VERSION_1_0:
 		dal_hw_translate_dcn10_init(translate);
 		return true;
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/Makefile b/drivers/gpu/drm/amd/display/dc/i2caux/Makefile
index 352885cb4d07..a851d07f0190 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/Makefile
@@ -71,7 +71,7 @@ AMD_DISPLAY_FILES += $(AMD_DAL_I2CAUX_DCE112)
 ###############################################################################
 # DCN 1.0 family
 ###############################################################################
-ifdef CONFIG_DRM_AMD_DC_DCN1_0
+ifdef CONFIG_X86
 I2CAUX_DCN1 = i2caux_dcn10.o
 
 AMD_DAL_I2CAUX_DCN1 = $(addprefix $(AMDDALPATH)/dc/i2caux/dcn10/,$(I2CAUX_DCN1))
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.c b/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.c
index bb526ad326e5..0afd2fa57bbe 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.c
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.c
@@ -128,8 +128,20 @@ static void process_read_reply(
 			ctx->status =
 				I2CAUX_TRANSACTION_STATUS_FAILED_PROTOCOL_ERROR;
 			ctx->operation_succeeded = false;
+		} else if (ctx->returned_byte < ctx->current_read_length) {
+			ctx->current_read_length -= ctx->returned_byte;
+
+			ctx->offset += ctx->returned_byte;
+
+			++ctx->invalid_reply_retry_aux_on_ack;
+
+			if (ctx->invalid_reply_retry_aux_on_ack >
+				AUX_INVALID_REPLY_RETRY_COUNTER) {
+				ctx->status =
+				I2CAUX_TRANSACTION_STATUS_FAILED_PROTOCOL_ERROR;
+				ctx->operation_succeeded = false;
+			}
 		} else {
-			ctx->current_read_length = ctx->returned_byte;
 			ctx->status = I2CAUX_TRANSACTION_STATUS_SUCCEEDED;
 			ctx->transaction_complete = true;
 			ctx->operation_succeeded = true;
@@ -157,6 +169,10 @@ static void process_read_reply(
 			ctx->operation_succeeded = false;
 		}
 	break;
+	case AUX_TRANSACTION_REPLY_HPD_DISCON:
+		ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON;
+		ctx->operation_succeeded = false;
+	break;
 	default:
 		ctx->status = I2CAUX_TRANSACTION_STATUS_UNKNOWN;
 		ctx->operation_succeeded = false;
@@ -215,6 +231,10 @@ static void process_read_request(
 			 * so we should not wait here */
 		}
 	break;
+	case AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON:
+		ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON;
+		ctx->operation_succeeded = false;
+	break;
 	default:
 		ctx->status = I2CAUX_TRANSACTION_STATUS_UNKNOWN;
 		ctx->operation_succeeded = false;
@@ -282,7 +302,6 @@ static bool read_command(
 				ctx.operation_succeeded);
 	}
 
-	request->payload.length = ctx.reply.length;
 	return ctx.operation_succeeded;
 }
 
@@ -370,6 +389,10 @@ static void process_write_reply(
 			ctx->operation_succeeded = false;
 		}
 	break;
+	case AUX_TRANSACTION_REPLY_HPD_DISCON:
+		ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON;
+		ctx->operation_succeeded = false;
+	break;
 	default:
 		ctx->status = I2CAUX_TRANSACTION_STATUS_UNKNOWN;
 		ctx->operation_succeeded = false;
@@ -422,6 +445,10 @@ static void process_write_request(
 			 * so we should not wait here */
 		}
 	break;
+	case AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON:
+		ctx->status = I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON;
+		ctx->operation_succeeded = false;
+	break;
 	default:
 		ctx->status = I2CAUX_TRANSACTION_STATUS_UNKNOWN;
 		ctx->operation_succeeded = false;
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.h b/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.h
index 8e71324ccb10..c33a2898d967 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.h
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/aux_engine.h
@@ -26,46 +26,7 @@
 #ifndef __DAL_AUX_ENGINE_H__
 #define __DAL_AUX_ENGINE_H__
 
-enum aux_transaction_type {
-	AUX_TRANSACTION_TYPE_DP,
-	AUX_TRANSACTION_TYPE_I2C
-};
-
-struct aux_request_transaction_data {
-	enum aux_transaction_type type;
-	enum i2caux_transaction_action action;
-	/* 20-bit AUX channel transaction address */
-	uint32_t address;
-	/* delay, in 100-microsecond units */
-	uint8_t delay;
-	uint32_t length;
-	uint8_t *data;
-};
-
-enum aux_transaction_reply {
-	AUX_TRANSACTION_REPLY_AUX_ACK = 0x00,
-	AUX_TRANSACTION_REPLY_AUX_NACK = 0x01,
-	AUX_TRANSACTION_REPLY_AUX_DEFER = 0x02,
-
-	AUX_TRANSACTION_REPLY_I2C_ACK = 0x00,
-	AUX_TRANSACTION_REPLY_I2C_NACK = 0x10,
-	AUX_TRANSACTION_REPLY_I2C_DEFER = 0x20,
-
-	AUX_TRANSACTION_REPLY_INVALID = 0xFF
-};
-
-struct aux_reply_transaction_data {
-	enum aux_transaction_reply status;
-	uint32_t length;
-	uint8_t *data;
-};
-
-enum aux_channel_operation_result {
-	AUX_CHANNEL_OPERATION_SUCCEEDED,
-	AUX_CHANNEL_OPERATION_FAILED_REASON_UNKNOWN,
-	AUX_CHANNEL_OPERATION_FAILED_INVALID_REPLY,
-	AUX_CHANNEL_OPERATION_FAILED_TIMEOUT
-};
+#include "dc_ddc_types.h"
 
 struct aux_engine;
 
@@ -83,6 +44,12 @@ struct aux_engine_funcs {
 	void (*process_channel_reply)(
 		struct aux_engine *engine,
 		struct aux_reply_transaction_data *reply);
+	int (*read_channel_reply)(
+		struct aux_engine *engine,
+		uint32_t size,
+		uint8_t *buffer,
+		uint8_t *reply_result,
+		uint32_t *sw_status);
 	enum aux_channel_operation_result (*get_channel_status)(
 		struct aux_engine *engine,
 		uint8_t *returned_bytes);
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce100/i2caux_dce100.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce100/i2caux_dce100.c
index e8d3781deaed..8b704ab0471c 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/dce100/i2caux_dce100.c
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce100/i2caux_dce100.c
@@ -97,6 +97,7 @@ struct i2caux *dal_i2caux_dce100_create(
 
 	dal_i2caux_dce110_construct(i2caux_dce110,
 				    ctx,
+				    ARRAY_SIZE(dce100_aux_regs),
 				    dce100_aux_regs,
 				    dce100_hw_engine_regs,
 				    &i2c_shift,
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/aux_engine_dce110.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/aux_engine_dce110.c
index 5f47f6c007ac..ae5caa97caca 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/aux_engine_dce110.c
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/aux_engine_dce110.c
@@ -198,27 +198,27 @@ static void submit_channel_request(
 		((request->type == AUX_TRANSACTION_TYPE_I2C) &&
 		((request->action == I2CAUX_TRANSACTION_ACTION_I2C_WRITE) ||
 		 (request->action == I2CAUX_TRANSACTION_ACTION_I2C_WRITE_MOT)));
+	if (REG(AUXN_IMPCAL)) {
+		/* clear_aux_error */
+		REG_UPDATE_SEQ(AUXN_IMPCAL, AUXN_CALOUT_ERROR_AK,
+				1,
+				0);
 
-	/* clear_aux_error */
-	REG_UPDATE_SEQ(AUXN_IMPCAL, AUXN_CALOUT_ERROR_AK,
-			1,
-			0);
-
-	REG_UPDATE_SEQ(AUXP_IMPCAL, AUXP_CALOUT_ERROR_AK,
-			1,
-			0);
-
-	/* force_default_calibrate */
-	REG_UPDATE_1BY1_2(AUXN_IMPCAL,
-			AUXN_IMPCAL_ENABLE, 1,
-			AUXN_IMPCAL_OVERRIDE_ENABLE, 0);
+		REG_UPDATE_SEQ(AUXP_IMPCAL, AUXP_CALOUT_ERROR_AK,
+				1,
+				0);
 
-	/* bug? why AUXN update EN and OVERRIDE_EN 1 by 1 while AUX P toggles OVERRIDE? */
+		/* force_default_calibrate */
+		REG_UPDATE_1BY1_2(AUXN_IMPCAL,
+				AUXN_IMPCAL_ENABLE, 1,
+				AUXN_IMPCAL_OVERRIDE_ENABLE, 0);
 
-	REG_UPDATE_SEQ(AUXP_IMPCAL, AUXP_IMPCAL_OVERRIDE_ENABLE,
-			1,
-			0);
+		/* bug? why AUXN update EN and OVERRIDE_EN 1 by 1 while AUX P toggles OVERRIDE? */
 
+		REG_UPDATE_SEQ(AUXP_IMPCAL, AUXP_IMPCAL_OVERRIDE_ENABLE,
+				1,
+				0);
+	}
 	/* set the delay and the number of bytes to write */
 
 	/* The length include
@@ -275,55 +275,92 @@ static void submit_channel_request(
 	REG_UPDATE(AUX_SW_CONTROL, AUX_SW_GO, 1);
 }
 
-static void process_channel_reply(
-	struct aux_engine *engine,
-	struct aux_reply_transaction_data *reply)
+static int read_channel_reply(struct aux_engine *engine, uint32_t size,
+			      uint8_t *buffer, uint8_t *reply_result,
+			      uint32_t *sw_status)
 {
 	struct aux_engine_dce110 *aux110 = FROM_AUX_ENGINE(engine);
+	uint32_t bytes_replied;
+	uint32_t reply_result_32;
 
-	/* Need to do a read to get the number of bytes to process
-	 * Alternatively, this information can be passed -
-	 * but that causes coupling which isn't good either. */
+	*sw_status = REG_GET(AUX_SW_STATUS, AUX_SW_REPLY_BYTE_COUNT,
+			     &bytes_replied);
 
-	uint32_t bytes_replied;
-	uint32_t value;
+	/* In case HPD is LOW, exit AUX transaction */
+	if ((*sw_status & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK))
+		return -1;
 
-	value = REG_GET(AUX_SW_STATUS,
-			AUX_SW_REPLY_BYTE_COUNT, &bytes_replied);
+	/* Need at least the status byte */
+	if (!bytes_replied)
+		return -1;
 
-	if (bytes_replied) {
-		uint32_t reply_result;
+	REG_UPDATE_1BY1_3(AUX_SW_DATA,
+			  AUX_SW_INDEX, 0,
+			  AUX_SW_AUTOINCREMENT_DISABLE, 1,
+			  AUX_SW_DATA_RW, 1);
 
-		REG_UPDATE_1BY1_3(AUX_SW_DATA,
-				AUX_SW_INDEX, 0,
-				AUX_SW_AUTOINCREMENT_DISABLE, 1,
-				AUX_SW_DATA_RW, 1);
+	REG_GET(AUX_SW_DATA, AUX_SW_DATA, &reply_result_32);
+	reply_result_32 = reply_result_32 >> 4;
+	*reply_result = (uint8_t)reply_result_32;
 
-		REG_GET(AUX_SW_DATA,
-				AUX_SW_DATA, &reply_result);
+	if (reply_result_32 == 0) { /* ACK */
+		uint32_t i = 0;
 
-		reply_result = reply_result >> 4;
+		/* First byte was already used to get the command status */
+		--bytes_replied;
 
-		switch (reply_result) {
-		case 0: /* ACK */ {
-			uint32_t i = 0;
+		/* Do not overflow buffer */
+		if (bytes_replied > size)
+			return -1;
 
-			/* first byte was already used
-			 * to get the command status */
-			--bytes_replied;
+		while (i < bytes_replied) {
+			uint32_t aux_sw_data_val;
 
-			while (i < bytes_replied) {
-				uint32_t aux_sw_data_val;
+			REG_GET(AUX_SW_DATA, AUX_SW_DATA, &aux_sw_data_val);
+			buffer[i] = aux_sw_data_val;
+			++i;
+		}
 
-				REG_GET(AUX_SW_DATA,
-						AUX_SW_DATA, &aux_sw_data_val);
+		return i;
+	}
 
-				reply->data[i] = aux_sw_data_val;
-				++i;
-			}
+	return 0;
+}
 
-			reply->status = AUX_TRANSACTION_REPLY_AUX_ACK;
+static void process_channel_reply(
+	struct aux_engine *engine,
+	struct aux_reply_transaction_data *reply)
+{
+	int bytes_replied;
+	uint8_t reply_result;
+	uint32_t sw_status;
+
+	bytes_replied = read_channel_reply(engine, reply->length, reply->data,
+					   &reply_result, &sw_status);
+
+	/* in case HPD is LOW, exit AUX transaction */
+	if ((sw_status & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK)) {
+		reply->status = AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON;
+		return;
+	}
+
+	if (bytes_replied < 0) {
+		/* Need to handle an error case...
+		 * Hopefully, upper layer function won't call this function if
+		 * the number of bytes in the reply was 0, because there was
+		 * surely an error that was asserted that should have been
+		 * handled for hot plug case, this could happens
+		 */
+		if (!(sw_status & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK)) {
+			reply->status = AUX_TRANSACTION_REPLY_INVALID;
+			ASSERT_CRITICAL(false);
+			return;
 		}
+	} else {
+
+		switch (reply_result) {
+		case 0: /* ACK */
+			reply->status = AUX_TRANSACTION_REPLY_AUX_ACK;
 		break;
 		case 1: /* NACK */
 			reply->status = AUX_TRANSACTION_REPLY_AUX_NACK;
@@ -340,15 +377,6 @@ static void process_channel_reply(
 		default:
 			reply->status = AUX_TRANSACTION_REPLY_INVALID;
 		}
-	} else {
-		/* Need to handle an error case...
-		 * hopefully, upper layer function won't call this function
-		 * if the number of bytes in the reply was 0
-		 * because there was surely an error that was asserted
-		 * that should have been handled
-		 * for hot plug case, this could happens*/
-		if (!(value & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK))
-			ASSERT_CRITICAL(false);
 	}
 }
 
@@ -371,6 +399,10 @@ static enum aux_channel_operation_result get_channel_status(
 	value = REG_WAIT(AUX_SW_STATUS, AUX_SW_DONE, 1,
 				10, aux110->timeout_period/10);
 
+	/* in case HPD is LOW, exit AUX transaction */
+	if ((value & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK))
+		return AUX_CHANNEL_OPERATION_FAILED_HPD_DISCON;
+
 	/* Note that the following bits are set in 'status.bits'
 	 * during CTS 4.2.1.2 (FW 3.3.1):
 	 * AUX_SW_RX_MIN_COUNT_VIOL, AUX_SW_RX_INVALID_STOP,
@@ -402,10 +434,10 @@ static enum aux_channel_operation_result get_channel_status(
 			return AUX_CHANNEL_OPERATION_SUCCEEDED;
 		}
 	} else {
-		/*time_elapsed >= aux_engine->timeout_period */
-		if (!(value & AUX_SW_STATUS__AUX_SW_HPD_DISCON_MASK))
-			ASSERT_CRITICAL(false);
-
+		/*time_elapsed >= aux_engine->timeout_period
+		 *  AUX_SW_STATUS__AUX_SW_HPD_DISCON = at this point
+		 */
+		ASSERT_CRITICAL(false);
 		return AUX_CHANNEL_OPERATION_FAILED_TIMEOUT;
 	}
 }
@@ -415,6 +447,7 @@ static const struct aux_engine_funcs aux_engine_funcs = {
 	.acquire_engine = acquire_engine,
 	.submit_channel_request = submit_channel_request,
 	.process_channel_reply = process_channel_reply,
+	.read_channel_reply = read_channel_reply,
 	.get_channel_status = get_channel_status,
 	.is_engine_available = is_engine_available,
 };
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c
index abd0095ced30..9cbe1a7a6bcb 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.c
@@ -62,12 +62,7 @@ enum dc_i2c_arbitration {
 	DC_I2C_ARBITRATION__DC_I2C_SW_PRIORITY_HIGH
 };
 
-enum {
-	/* No timeout in HW
-	 * (timeout implemented in SW by querying status) */
-	I2C_SETUP_TIME_LIMIT = 255,
-	I2C_HW_BUFFER_SIZE = 538
-};
+
 
 /*
  * @brief
@@ -152,6 +147,11 @@ static bool setup_engine(
 	struct i2c_engine *i2c_engine)
 {
 	struct i2c_hw_engine_dce110 *hw_engine = FROM_I2C_ENGINE(i2c_engine);
+	uint32_t i2c_setup_limit = I2C_SETUP_TIME_LIMIT_DCE;
+	uint32_t  reset_length = 0;
+
+	if (hw_engine->base.base.setup_limit != 0)
+		i2c_setup_limit = hw_engine->base.base.setup_limit;
 
 	/* Program pin select */
 	REG_UPDATE_6(
@@ -164,11 +164,15 @@ static bool setup_engine(
 			DC_I2C_DDC_SELECT, hw_engine->engine_id);
 
 	/* Program time limit */
-	REG_UPDATE_N(
-			SETUP, 2,
-			FN(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_TIME_LIMIT), I2C_SETUP_TIME_LIMIT,
-			FN(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_ENABLE), 1);
-
+	if (hw_engine->base.base.send_reset_length == 0) {
+		/*pre-dcn*/
+		REG_UPDATE_N(
+				SETUP, 2,
+				FN(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_TIME_LIMIT), i2c_setup_limit,
+				FN(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_ENABLE), 1);
+	} else {
+		reset_length = hw_engine->base.base.send_reset_length;
+	}
 	/* Program HW priority
 	 * set to High - interrupt software I2C at any time
 	 * Enable restart of SW I2C that was interrupted by HW
@@ -527,7 +531,7 @@ static void construct(
 	REG_GET(MICROSECOND_TIME_BASE_DIV, XTAL_REF_DIV, &xtal_ref_div);
 
 	if (xtal_ref_div == 0) {
-		DC_LOG_WARNING("Invalid base timer divider\n",
+		DC_LOG_WARNING("Invalid base timer divider [%s]\n",
 				__func__);
 		xtal_ref_div = 2;
 	}
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.h b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.h
index 5bb04085f670..fea2946906ed 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.h
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2c_hw_engine_dce110.h
@@ -192,6 +192,7 @@ struct i2c_hw_engine_dce110 {
 	/* number of pending transactions (before GO) */
 	uint32_t transaction_count;
 	uint32_t engine_keep_power_up_count;
+	uint32_t i2_setup_time_limit;
 };
 
 struct i2c_hw_engine_dce110_create_arg {
@@ -207,4 +208,11 @@ struct i2c_hw_engine_dce110_create_arg {
 struct i2c_engine *dal_i2c_hw_engine_dce110_create(
 	const struct i2c_hw_engine_dce110_create_arg *arg);
 
+enum {
+	I2C_SETUP_TIME_LIMIT_DCE = 255,
+	I2C_SETUP_TIME_LIMIT_DCN = 3,
+	I2C_HW_BUFFER_SIZE = 538,
+	I2C_SEND_RESET_LENGTH_9 = 9,
+	I2C_SEND_RESET_LENGTH_10 = 10,
+};
 #endif
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.c
index 2a047f8ca0e9..1d748ac1d6d6 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.c
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.c
@@ -43,6 +43,9 @@
 #include "i2c_sw_engine_dce110.h"
 #include "i2c_hw_engine_dce110.h"
 #include "aux_engine_dce110.h"
+#include "../../dc.h"
+#include "dc_types.h"
+
 
 /*
  * Post-requisites: headers required by this unit
@@ -199,6 +202,7 @@ static const struct dce110_i2c_hw_engine_mask i2c_mask = {
 void dal_i2caux_dce110_construct(
 	struct i2caux_dce110 *i2caux_dce110,
 	struct dc_context *ctx,
+	unsigned int num_i2caux_inst,
 	const struct dce110_aux_registers aux_regs[],
 	const struct dce110_i2c_hw_engine_registers i2c_hw_engine_regs[],
 	const struct dce110_i2c_hw_engine_shift *i2c_shift,
@@ -249,9 +253,22 @@ void dal_i2caux_dce110_construct(
 
 		base->i2c_hw_engines[line_id] =
 			dal_i2c_hw_engine_dce110_create(&hw_arg_dce110);
-
+		if (base->i2c_hw_engines[line_id] != NULL) {
+			switch (ctx->dce_version) {
+			case DCN_VERSION_1_0:
+				base->i2c_hw_engines[line_id]->setup_limit =
+					I2C_SETUP_TIME_LIMIT_DCN;
+				base->i2c_hw_engines[line_id]->send_reset_length  = 0;
+			break;
+			default:
+				base->i2c_hw_engines[line_id]->setup_limit =
+					I2C_SETUP_TIME_LIMIT_DCE;
+				base->i2c_hw_engines[line_id]->send_reset_length  = 0;
+				break;
+			}
+		}
 		++i;
-	} while (i < ARRAY_SIZE(hw_ddc_lines));
+	} while (i < num_i2caux_inst);
 
 	/* Create AUX engines for all lines which has assisted HW AUX
 	 * 'i' (loop counter) used as DDC/AUX engine_id */
@@ -272,7 +289,7 @@ void dal_i2caux_dce110_construct(
 			dal_aux_engine_dce110_create(&aux_init_data);
 
 		++i;
-	} while (i < ARRAY_SIZE(hw_aux_lines));
+	} while (i < num_i2caux_inst);
 
 	/*TODO Generic I2C SW and HW*/
 }
@@ -303,6 +320,7 @@ struct i2caux *dal_i2caux_dce110_create(
 
 	dal_i2caux_dce110_construct(i2caux_dce110,
 				    ctx,
+				    ARRAY_SIZE(dce110_aux_regs),
 				    dce110_aux_regs,
 				    i2c_hw_engine_regs,
 				    &i2c_shift,
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.h b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.h
index 1b1f71c60ac9..d3d8cc58666a 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.h
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce110/i2caux_dce110.h
@@ -45,6 +45,7 @@ struct i2caux *dal_i2caux_dce110_create(
 void dal_i2caux_dce110_construct(
 	struct i2caux_dce110 *i2caux_dce110,
 	struct dc_context *ctx,
+	unsigned int num_i2caux_inst,
 	const struct dce110_aux_registers *aux_regs,
 	const struct dce110_i2c_hw_engine_registers *i2c_hw_engine_regs,
 	const struct dce110_i2c_hw_engine_shift *i2c_shift,
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce112/i2caux_dce112.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce112/i2caux_dce112.c
index dafc1a727f7f..a9db04738724 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/dce112/i2caux_dce112.c
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce112/i2caux_dce112.c
@@ -93,6 +93,7 @@ static void construct(
 {
 	dal_i2caux_dce110_construct(i2caux_dce110,
 				    ctx,
+				    ARRAY_SIZE(dce112_aux_regs),
 				    dce112_aux_regs,
 				    dce112_hw_engine_regs,
 				    &i2c_shift,
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dce120/i2caux_dce120.c b/drivers/gpu/drm/amd/display/dc/i2caux/dce120/i2caux_dce120.c
index 0e7b18260027..6a4f344c1db4 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/dce120/i2caux_dce120.c
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/dce120/i2caux_dce120.c
@@ -111,6 +111,7 @@ struct i2caux *dal_i2caux_dce120_create(
 
 	dal_i2caux_dce110_construct(i2caux_dce110,
 				    ctx,
+				    ARRAY_SIZE(dce120_aux_regs),
 				    dce120_aux_regs,
 				    dce120_hw_engine_regs,
 				    &i2c_shift,
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/dcn10/i2caux_dcn10.c b/drivers/gpu/drm/amd/display/dc/i2caux/dcn10/i2caux_dcn10.c
index e44a8901f38b..a59c1f50c1e8 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/dcn10/i2caux_dcn10.c
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/dcn10/i2caux_dcn10.c
@@ -111,6 +111,7 @@ struct i2caux *dal_i2caux_dcn10_create(
 
 	dal_i2caux_dce110_construct(i2caux_dce110,
 				    ctx,
+				    ARRAY_SIZE(dcn10_aux_regs),
 				    dcn10_aux_regs,
 				    dcn10_hw_engine_regs,
 				    &i2c_shift,
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/engine.h b/drivers/gpu/drm/amd/display/dc/i2caux/engine.h
index 33de8a8834dc..1e8a1585e401 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/engine.h
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/engine.h
@@ -26,6 +26,8 @@
 #ifndef __DAL_ENGINE_H__
 #define __DAL_ENGINE_H__
 
+#include "dc_ddc_types.h"
+
 enum i2caux_transaction_operation {
 	I2CAUX_TRANSACTION_READ,
 	I2CAUX_TRANSACTION_WRITE
@@ -53,7 +55,8 @@ enum i2caux_transaction_status {
 	I2CAUX_TRANSACTION_STATUS_FAILED_INCOMPLETE,
 	I2CAUX_TRANSACTION_STATUS_FAILED_OPERATION,
 	I2CAUX_TRANSACTION_STATUS_FAILED_INVALID_OPERATION,
-	I2CAUX_TRANSACTION_STATUS_FAILED_BUFFER_OVERFLOW
+	I2CAUX_TRANSACTION_STATUS_FAILED_BUFFER_OVERFLOW,
+	I2CAUX_TRANSACTION_STATUS_FAILED_HPD_DISCON
 };
 
 struct i2caux_transaction_request {
@@ -75,19 +78,6 @@ enum i2c_default_speed {
 	I2CAUX_DEFAULT_I2C_SW_SPEED = 50
 };
 
-enum i2caux_transaction_action {
-	I2CAUX_TRANSACTION_ACTION_I2C_WRITE = 0x00,
-	I2CAUX_TRANSACTION_ACTION_I2C_READ = 0x10,
-	I2CAUX_TRANSACTION_ACTION_I2C_STATUS_REQUEST = 0x20,
-
-	I2CAUX_TRANSACTION_ACTION_I2C_WRITE_MOT = 0x40,
-	I2CAUX_TRANSACTION_ACTION_I2C_READ_MOT = 0x50,
-	I2CAUX_TRANSACTION_ACTION_I2C_STATUS_REQUEST_MOT = 0x60,
-
-	I2CAUX_TRANSACTION_ACTION_DP_WRITE = 0x80,
-	I2CAUX_TRANSACTION_ACTION_DP_READ = 0x90
-};
-
 struct engine;
 
 struct engine_funcs {
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/i2c_engine.h b/drivers/gpu/drm/amd/display/dc/i2caux/i2c_engine.h
index 58fc0f25eceb..ded6ea34b714 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/i2c_engine.h
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/i2c_engine.h
@@ -86,6 +86,8 @@ struct i2c_engine {
 	struct engine base;
 	const struct i2c_engine_funcs *funcs;
 	uint32_t timeout_delay;
+	uint32_t setup_limit;
+	uint32_t send_reset_length;
 };
 
 void dal_i2c_engine_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c b/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c
index 5cbf6626b8d4..f7ed355fc84f 100644
--- a/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c
+++ b/drivers/gpu/drm/amd/display/dc/i2caux/i2caux.c
@@ -59,7 +59,7 @@
 
 #include "dce120/i2caux_dce120.h"
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 #include "dcn10/i2caux_dcn10.h"
 #endif
 
@@ -83,6 +83,7 @@ struct i2caux *dal_i2caux_create(
 	case DCE_VERSION_8_3:
 		return dal_i2caux_dce80_create(ctx);
 	case DCE_VERSION_11_2:
+	case DCE_VERSION_11_22:
 		return dal_i2caux_dce112_create(ctx);
 	case DCE_VERSION_11_0:
 		return dal_i2caux_dce110_create(ctx);
@@ -90,7 +91,7 @@ struct i2caux *dal_i2caux_create(
 		return dal_i2caux_dce100_create(ctx);
 	case DCE_VERSION_12_0:
 		return dal_i2caux_dce120_create(ctx);
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 	case DCN_VERSION_1_0:
 		return dal_i2caux_dcn10_create(ctx);
 #endif
@@ -253,7 +254,6 @@ bool dal_i2caux_submit_aux_command(
 			break;
 		}
 
-		cmd->payloads->length = request.payload.length;
 		++index_of_payload;
 	}
 
diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_types.h b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
index 8c51ad70cace..4446652a9a9e 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/core_types.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
@@ -33,7 +33,7 @@
 #include "dc_bios_types.h"
 #include "mem_input.h"
 #include "hubp.h"
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 #include "mpc.h"
 #endif
 
@@ -95,11 +95,6 @@ struct resource_funcs {
 	struct link_encoder *(*link_enc_create)(
 			const struct encoder_init_data *init);
 
-	enum dc_status (*validate_guaranteed)(
-					struct dc *dc,
-					struct dc_stream_state *stream,
-					struct dc_state *context);
-
 	bool (*validate_bandwidth)(
 					struct dc *dc,
 					struct dc_state *context);
@@ -167,7 +162,7 @@ struct resource_pool {
 	unsigned int audio_count;
 	struct audio_support audio_support;
 
-	struct display_clock *display_clock;
+	struct dccg *dccg;
 	struct irq_service *irqs;
 
 	struct abm *abm;
@@ -226,7 +221,7 @@ struct pipe_ctx {
 	struct pipe_ctx *top_pipe;
 	struct pipe_ctx *bottom_pipe;
 
-#ifdef CONFIG_DRM_AMD_DC_DCN1_0
+#ifdef CONFIG_X86
 	struct _vcs_dpi_display_dlg_regs_st dlg_regs;
 	struct _vcs_dpi_display_ttu_regs_st ttu_regs;
 	struct _vcs_dpi_display_rq_regs_st rq_regs;
@@ -250,6 +245,7 @@ struct dce_bw_output {
 	bool all_displays_in_sync;
 	struct dce_watermarks urgent_wm_ns[MAX_PIPES];
 	struct dce_watermarks stutter_exit_wm_ns[MAX_PIPES];
+	struct dce_watermarks stutter_entry_wm_ns[MAX_PIPES];
 	struct dce_watermarks nbp_state_change_wm_ns[MAX_PIPES];
 	int sclk_khz;
 	int sclk_deep_sleep_khz;
@@ -259,8 +255,7 @@ struct dce_bw_output {
 };
 
 struct dcn_bw_output {
-	struct dc_clocks cur_clk;
-	struct dc_clocks calc_clk;
+	struct dc_clocks clk;
 	struct dcn_watermark_set watermarks;
 };
 
@@ -281,11 +276,11 @@ struct dc_state {
 
 	/* Note: these are big structures, do *not* put on stack! */
 	struct dm_pp_display_configuration pp_display_cfg;
-#ifdef CONFIG_DRM_AMD_DC_DCN1_0
+#ifdef CONFIG_X86
 	struct dcn_bw_internal_vars dcn_bw_vars;
 #endif
 
-	struct display_clock *dis_clk;
+	struct dccg *dis_clk;
 
 	struct kref refcount;
 };
diff --git a/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h b/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h
index 090b7a8dd67b..538b83303b86 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/dc_link_ddc.h
@@ -102,21 +102,13 @@ bool dal_ddc_service_query_ddc_data(
 		uint8_t *read_buf,
 		uint32_t read_size);
 
-ssize_t dal_ddc_service_read_dpcd_data(
-		struct ddc_service *ddc,
-		bool i2c,
-		enum i2c_mot_mode mot,
-		uint32_t address,
-		uint8_t *data,
-		uint32_t len);
-
-enum ddc_result dal_ddc_service_write_dpcd_data(
-		struct ddc_service *ddc,
-		bool i2c,
-		enum i2c_mot_mode mot,
-		uint32_t address,
-		const uint8_t *data,
-		uint32_t len);
+int dc_link_aux_transfer(struct ddc_service *ddc,
+			     unsigned int address,
+			     uint8_t *reply,
+			     void *buffer,
+			     unsigned int size,
+			     enum aux_transaction_type type,
+			     enum i2caux_transaction_action action);
 
 void dal_ddc_service_write_scdc_data(
 		struct ddc_service *ddc_service,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h b/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h
index a9bfe9ff8ce6..eece165206f9 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/dce_calcs.h
@@ -42,6 +42,8 @@ enum bw_calcs_version {
 	BW_CALCS_VERSION_CARRIZO,
 	BW_CALCS_VERSION_POLARIS10,
 	BW_CALCS_VERSION_POLARIS11,
+	BW_CALCS_VERSION_POLARIS12,
+	BW_CALCS_VERSION_VEGAM,
 	BW_CALCS_VERSION_STONEY,
 	BW_CALCS_VERSION_VEGA10
 };
diff --git a/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h b/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h
index 132d18d4b293..ddbb673caa08 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h
@@ -625,7 +625,7 @@ bool dcn_validate_bandwidth(
 
 unsigned int dcn_find_dcfclk_suits_all(
 	const struct dc *dc,
-	struct clocks_value *clocks);
+	struct dc_clocks *clocks);
 
 void dcn_bw_update_from_pplib(struct dc *dc);
 void dcn_bw_notify_pplib_of_wm_ranges(struct dc *dc);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h
new file mode 100644
index 000000000000..02f757dd70d4
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h
@@ -0,0 +1,64 @@
+/*
+ * Copyright 2012-15 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __DAL_DCHUBBUB_H__
+#define __DAL_DCHUBBUB_H__
+
+
+enum dcc_control {
+	dcc_control__256_256_xxx,
+	dcc_control__128_128_xxx,
+	dcc_control__256_64_64,
+};
+
+enum segment_order {
+	segment_order__na,
+	segment_order__contiguous,
+	segment_order__non_contiguous,
+};
+
+
+struct hubbub_funcs {
+	void (*update_dchub)(
+			struct hubbub *hubbub,
+			struct dchub_init_data *dh_data);
+
+	bool (*get_dcc_compression_cap)(struct hubbub *hubbub,
+			const struct dc_dcc_surface_param *input,
+			struct dc_surface_dcc_cap *output);
+
+	bool (*dcc_support_swizzle)(
+			enum swizzle_mode_values swizzle,
+			unsigned int bytes_per_element,
+			enum segment_order *segment_order_horz,
+			enum segment_order *segment_order_vert);
+
+	bool (*dcc_support_pixel_format)(
+			enum surface_pixel_format format,
+			unsigned int *bytes_per_element);
+};
+
+
+#endif
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/display_clock.h b/drivers/gpu/drm/amd/display/dc/inc/hw/display_clock.h
index f5f69cd81f6f..3c7ccb68ecdb 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/display_clock.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/display_clock.h
@@ -27,23 +27,7 @@
 #define __DISPLAY_CLOCK_H__
 
 #include "dm_services_types.h"
-
-
-struct clocks_value {
-	int dispclk_in_khz;
-	int max_pixelclk_in_khz;
-	int max_non_dp_phyclk_in_khz;
-	int max_dp_phyclk_in_khz;
-	bool dispclk_notify_pplib_done;
-	bool pixelclk_notify_pplib_done;
-	bool phyclk_notigy_pplib_done;
-	int dcfclock_in_khz;
-	int dppclk_in_khz;
-	int mclk_in_khz;
-	int phyclk_in_khz;
-	int common_vdd_level;
-};
-
+#include "dc.h"
 
 /* Structure containing all state-dependent clocks
  * (dependent on "enum clocks_state") */
@@ -52,34 +36,23 @@ struct state_dependent_clocks {
 	int pixel_clk_khz;
 };
 
-struct display_clock {
+struct dccg {
 	struct dc_context *ctx;
 	const struct display_clock_funcs *funcs;
 
 	enum dm_pp_clocks_state max_clks_state;
 	enum dm_pp_clocks_state cur_min_clks_state;
-	struct clocks_value cur_clocks_value;
+	struct dc_clocks clks;
 };
 
 struct display_clock_funcs {
-	int (*set_clock)(struct display_clock *disp_clk,
+	void (*update_clocks)(struct dccg *dccg,
+			struct dc_clocks *new_clocks,
+			bool safe_to_lower);
+	int (*set_dispclk)(struct dccg *dccg,
 		int requested_clock_khz);
 
-	enum dm_pp_clocks_state (*get_required_clocks_state)(
-		struct display_clock *disp_clk,
-		struct state_dependent_clocks *req_clocks);
-
-	bool (*set_min_clocks_state)(struct display_clock *disp_clk,
-		enum dm_pp_clocks_state dm_pp_clocks_state);
-
-	int (*get_dp_ref_clk_frequency)(struct display_clock *disp_clk);
-
-	bool (*apply_clock_voltage_request)(
-		struct display_clock *disp_clk,
-		enum dm_pp_clock_type clocks_type,
-		int clocks_in_khz,
-		bool pre_mode_set,
-		bool update_dp_phyclk);
+	int (*get_dp_ref_clk_frequency)(struct dccg *dccg);
 };
 
 #endif /* __DISPLAY_CLOCK_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
index 99995608b620..74ad94b0e4f0 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
@@ -44,7 +44,23 @@ struct dpp_grph_csc_adjustment {
 	enum graphics_gamut_adjust_type gamut_adjust_type;
 };
 
+struct dcn_dpp_state {
+	uint32_t igam_lut_mode;
+	uint32_t igam_input_format;
+	uint32_t dgam_lut_mode;
+	uint32_t rgam_lut_mode;
+	uint32_t gamut_remap_mode;
+	uint32_t gamut_remap_c11_c12;
+	uint32_t gamut_remap_c13_c14;
+	uint32_t gamut_remap_c21_c22;
+	uint32_t gamut_remap_c23_c24;
+	uint32_t gamut_remap_c31_c32;
+	uint32_t gamut_remap_c33_c34;
+};
+
 struct dpp_funcs {
+	void (*dpp_read_state)(struct dpp *dpp, struct dcn_dpp_state *s);
+
 	void (*dpp_reset)(struct dpp *dpp);
 
 	void (*dpp_set_scaler)(struct dpp *dpp,
@@ -117,7 +133,7 @@ struct dpp_funcs {
 			struct dpp *dpp_base,
 			enum surface_pixel_format format,
 			enum expansion_mode mode,
-			struct csc_transform input_csc_color_matrix,
+			struct dc_csc_transform input_csc_color_matrix,
 			enum dc_color_space input_color_space);
 
 	void (*dpp_full_bypass)(struct dpp *dpp_base);
@@ -135,6 +151,9 @@ struct dpp_funcs {
 	void (*dpp_set_hdr_multiplier)(
 			struct dpp *dpp_base,
 			uint32_t multiplier);
+	void (*set_optional_cursor_attributes)(
+			struct dpp *dpp_base,
+			struct dpp_cursor_attributes *attr);
 
 	void (*dpp_dppclk_control)(
 			struct dpp *dpp_base,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
index 9ced254e652c..4f3f9e68ccfa 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
@@ -43,10 +43,9 @@ enum cursor_lines_per_chunk {
 };
 
 struct hubp {
-	struct hubp_funcs *funcs;
+	const struct hubp_funcs *funcs;
 	struct dc_context *ctx;
 	struct dc_plane_address request_address;
-	struct dc_plane_address current_address;
 	int inst;
 
 	/* run time states */
@@ -56,7 +55,6 @@ struct hubp {
 	bool power_gated;
 };
 
-
 struct hubp_funcs {
 	void (*hubp_setup)(
 			struct hubp *hubp,
@@ -121,6 +119,9 @@ struct hubp_funcs {
 
 	void (*hubp_clk_cntl)(struct hubp *hubp, bool enable);
 	void (*hubp_vtg_sel)(struct hubp *hubp, uint32_t otg_inst);
+	void (*hubp_read_state)(struct hubp *hubp);
+	void (*hubp_disable_control)(struct hubp *hubp, bool disable_hubp);
+	unsigned int (*hubp_get_underflow_status)(struct hubp *hubp);
 
 };
 
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
index b22158190262..cf7433ebf91a 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hw_shared.h
@@ -140,11 +140,6 @@ enum opp_regamma {
 	OPP_REGAMMA_USER
 };
 
-struct csc_transform {
-	uint16_t matrix[12];
-	bool enable_adjustment;
-};
-
 struct dc_bias_and_scale {
 	uint16_t scale_red;
 	uint16_t bias_red;
@@ -191,4 +186,9 @@ enum controller_dp_test_pattern {
 	CONTROLLER_DP_TEST_PATTERN_COLORSQUARES_CEA
 };
 
+enum dc_lut_mode {
+	LUT_BYPASS,
+	LUT_RAM_A,
+	LUT_RAM_B
+};
 #endif /* __DAL_HW_SHARED_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/ipp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/ipp.h
index 2109eac20a3d..b2fa4c4cd920 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/ipp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/ipp.h
@@ -87,7 +87,7 @@ struct ipp_funcs {
 		struct input_pixel_processor *ipp,
 		enum surface_pixel_format format,
 		enum expansion_mode mode,
-		struct csc_transform input_csc_color_matrix,
+		struct dc_csc_transform input_csc_color_matrix,
 		enum dc_color_space input_color_space);
 
 	/* DCE function to setup IPP.  TODO: see if we can consolidate to setup */
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h b/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h
index 54d8a1386142..cf6df2e7beb2 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h
@@ -149,6 +149,7 @@ struct link_encoder_funcs {
 		bool connect);
 	void (*enable_hpd)(struct link_encoder *enc);
 	void (*disable_hpd)(struct link_encoder *enc);
+	bool (*is_dig_enabled)(struct link_encoder *enc);
 	void (*destroy)(struct link_encoder **enc);
 };
 
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h b/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h
index 3e1e7e6a8792..da89c2edb07c 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h
@@ -64,7 +64,7 @@ struct stutter_modes {
 };
 
 struct mem_input {
-	struct mem_input_funcs *funcs;
+	const struct mem_input_funcs *funcs;
 	struct dc_context *ctx;
 	struct dc_plane_address request_address;
 	struct dc_plane_address current_address;
@@ -104,6 +104,7 @@ struct mem_input_funcs {
 		struct mem_input *mem_input,
 		struct dce_watermarks nbp,
 		struct dce_watermarks stutter,
+		struct dce_watermarks stutter_enter,
 		struct dce_watermarks urgent,
 		uint32_t total_dest_line_time_ns);
 
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h b/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h
index 23a8d5e53a89..caf74e3c836f 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h
@@ -105,7 +105,24 @@ struct mpc {
 	struct mpcc mpcc_array[MAX_MPCC];
 };
 
+struct mpcc_state {
+	uint32_t opp_id;
+	uint32_t dpp_id;
+	uint32_t bot_mpcc_id;
+	uint32_t mode;
+	uint32_t alpha_mode;
+	uint32_t pre_multiplied_alpha;
+	uint32_t overlap_only;
+	uint32_t idle;
+	uint32_t busy;
+};
+
 struct mpc_funcs {
+	void (*read_mpcc_state)(
+			struct mpc *mpc,
+			int mpcc_inst,
+			struct mpcc_state *s);
+
 	/*
 	 * Insert DPP into MPC tree based on specified blending position.
 	 * Only used for planes that are part of blending chain for OPP output
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h b/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
index b5db1692393c..cfa7ec9517ae 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
@@ -29,31 +29,40 @@
 #define STREAM_ENCODER_H_
 
 #include "audio_types.h"
+#include "hw_shared.h"
 
 struct dc_bios;
 struct dc_context;
 struct dc_crtc_timing;
 
-struct encoder_info_packet {
-	bool valid;
-	uint8_t hb0;
-	uint8_t hb1;
-	uint8_t hb2;
-	uint8_t hb3;
-	uint8_t sb[32];
+enum dp_pixel_encoding_type {
+	DP_PIXEL_ENCODING_TYPE_RGB444		= 0x00000000,
+	DP_PIXEL_ENCODING_TYPE_YCBCR422		= 0x00000001,
+	DP_PIXEL_ENCODING_TYPE_YCBCR444		= 0x00000002,
+	DP_PIXEL_ENCODING_TYPE_RGB_WIDE_GAMUT	= 0x00000003,
+	DP_PIXEL_ENCODING_TYPE_Y_ONLY		= 0x00000004,
+	DP_PIXEL_ENCODING_TYPE_YCBCR420		= 0x00000005
+};
+
+enum dp_component_depth {
+	DP_COMPONENT_PIXEL_DEPTH_6BPC		= 0x00000000,
+	DP_COMPONENT_PIXEL_DEPTH_8BPC		= 0x00000001,
+	DP_COMPONENT_PIXEL_DEPTH_10BPC		= 0x00000002,
+	DP_COMPONENT_PIXEL_DEPTH_12BPC		= 0x00000003,
+	DP_COMPONENT_PIXEL_DEPTH_16BPC		= 0x00000004
 };
 
 struct encoder_info_frame {
 	/* auxiliary video information */
-	struct encoder_info_packet avi;
-	struct encoder_info_packet gamut;
-	struct encoder_info_packet vendor;
+	struct dc_info_packet avi;
+	struct dc_info_packet gamut;
+	struct dc_info_packet vendor;
 	/* source product description */
-	struct encoder_info_packet spd;
+	struct dc_info_packet spd;
 	/* video stream configuration */
-	struct encoder_info_packet vsc;
+	struct dc_info_packet vsc;
 	/* HDR Static MetaData */
-	struct encoder_info_packet hdrsmd;
+	struct dc_info_packet hdrsmd;
 };
 
 struct encoder_unblank_param {
@@ -147,6 +156,7 @@ struct stream_encoder_funcs {
 
 	void (*set_avmute)(
 		struct stream_encoder *enc, bool enable);
+
 };
 
 #endif /* STREAM_ENCODER_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
index 3217b5bf6c7a..af700c7dac50 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
@@ -140,6 +140,9 @@ struct timing_generator_funcs {
 	void (*program_timing)(struct timing_generator *tg,
 							const struct dc_crtc_timing *timing,
 							bool use_vbios);
+	void (*program_vline_interrupt)(struct timing_generator *optc,
+			const struct dc_crtc_timing *dc_crtc_timing,
+			unsigned long long vsync_delta);
 	bool (*enable_crtc)(struct timing_generator *tg);
 	bool (*disable_crtc)(struct timing_generator *tg);
 	bool (*is_counter_moving)(struct timing_generator *tg);
@@ -153,6 +156,9 @@ struct timing_generator_funcs {
 		uint32_t *v_blank_end,
 		uint32_t *h_position,
 		uint32_t *v_position);
+	bool (*get_otg_active_size)(struct timing_generator *optc,
+			uint32_t *otg_active_width,
+			uint32_t *otg_active_height);
 	void (*set_early_control)(struct timing_generator *tg,
 							   uint32_t early_cntl);
 	void (*wait_for_state)(struct timing_generator *tg,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h b/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h
index c5b3623bcbd9..fecc80c47c26 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h
@@ -252,7 +252,7 @@ struct transform_funcs {
 			struct transform *xfm_base,
 			enum surface_pixel_format format,
 			enum expansion_mode mode,
-			struct csc_transform input_csc_color_matrix,
+			struct dc_csc_transform input_csc_color_matrix,
 			enum dc_color_space input_color_space);
 
 	void (*ipp_full_bypass)(struct transform *xfm_base);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
index e764cbad881b..a14ce4de80b2 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
@@ -32,6 +32,8 @@
 #include "inc/hw/link_encoder.h"
 #include "core_status.h"
 
+#define EDP_BACKLIGHT_RAMP_DISABLE_LEVEL 0xFFFFFFFF
+
 enum pipe_gating_control {
 	PIPE_GATING_CONTROL_DISABLE = 0,
 	PIPE_GATING_CONTROL_ENABLE,
@@ -42,6 +44,7 @@ struct dce_hwseq_wa {
 	bool blnd_crtc_trigger;
 	bool DEGVIDCN10_253;
 	bool false_optc_underflow;
+	bool DEGVIDCN10_254;
 };
 
 struct hwseq_wa_state {
@@ -63,6 +66,7 @@ struct dchub_init_data;
 struct dc_static_screen_events;
 struct resource_pool;
 struct resource_context;
+struct stream_resource;
 
 struct hw_sequencer_funcs {
 
@@ -80,11 +84,6 @@ struct hw_sequencer_funcs {
 			int num_planes,
 			struct dc_state *context);
 
-	void (*set_plane_config)(
-			const struct dc *dc,
-			struct pipe_ctx *pipe_ctx,
-			struct resource_context *res_ctx);
-
 	void (*program_gamut_remap)(
 			struct pipe_ctx *pipe_ctx);
 
@@ -93,14 +92,28 @@ struct hw_sequencer_funcs {
 			enum dc_color_space colorspace,
 			uint16_t *matrix);
 
+	void (*program_output_csc)(struct dc *dc,
+			struct pipe_ctx *pipe_ctx,
+			enum dc_color_space colorspace,
+			uint16_t *matrix,
+			int opp_id);
+
 	void (*update_plane_addr)(
 		const struct dc *dc,
 		struct pipe_ctx *pipe_ctx);
 
+	void (*plane_atomic_disconnect)(
+		struct dc *dc,
+		struct pipe_ctx *pipe_ctx);
+
 	void (*update_dchub)(
 		struct dce_hwseq *hws,
 		struct dchub_init_data *dh_data);
 
+	void (*update_mpcc)(
+		struct dc *dc,
+		struct pipe_ctx *pipe_ctx);
+
 	void (*update_pending_status)(
 			struct pipe_ctx *pipe_ctx);
 
@@ -150,15 +163,24 @@ struct hw_sequencer_funcs {
 			struct dc_link_settings *link_settings);
 
 	void (*blank_stream)(struct pipe_ctx *pipe_ctx);
+
+	void (*enable_audio_stream)(struct pipe_ctx *pipe_ctx);
+
+	void (*disable_audio_stream)(struct pipe_ctx *pipe_ctx, int option);
+
 	void (*pipe_control_lock)(
 				struct dc *dc,
 				struct pipe_ctx *pipe,
 				bool lock);
+	void (*blank_pixel_data)(
+			struct dc *dc,
+			struct pipe_ctx *pipe_ctx,
+			bool blank);
 
 	void (*set_bandwidth)(
 			struct dc *dc,
 			struct dc_state *context,
-			bool decrease_allowed);
+			bool safe_to_lower);
 
 	void (*set_drr)(struct pipe_ctx **pipe_ctx, int num_pipes,
 			int vmin, int vmax);
@@ -169,7 +191,7 @@ struct hw_sequencer_funcs {
 	void (*set_static_screen_control)(struct pipe_ctx **pipe_ctx,
 			int num_pipes, const struct dc_static_screen_events *events);
 
-	enum dc_status (*prog_pixclk_crtc_otg)(
+	enum dc_status (*enable_stream_timing)(
 			struct pipe_ctx *pipe_ctx,
 			struct dc_state *context,
 			struct dc *dc);
@@ -201,6 +223,8 @@ struct hw_sequencer_funcs {
 
 	void (*set_cursor_position)(struct pipe_ctx *pipe);
 	void (*set_cursor_attribute)(struct pipe_ctx *pipe);
+	void (*set_cursor_sdr_white_level)(struct pipe_ctx *pipe);
+
 };
 
 void color_space_to_black_color(
diff --git a/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h b/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h
index 77eb72874e90..cf5a84b9e27c 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h
@@ -183,6 +183,36 @@
 				FN(reg_name, f4), v4, \
 				FN(reg_name, f5), v5)
 
+#define REG_GET_6(reg_name, f1, v1, f2, v2, f3, v3, f4, v4, f5, v5, f6, v6)	\
+		generic_reg_get6(CTX, REG(reg_name), \
+				FN(reg_name, f1), v1, \
+				FN(reg_name, f2), v2, \
+				FN(reg_name, f3), v3, \
+				FN(reg_name, f4), v4, \
+				FN(reg_name, f5), v5, \
+				FN(reg_name, f6), v6)
+
+#define REG_GET_7(reg_name, f1, v1, f2, v2, f3, v3, f4, v4, f5, v5, f6, v6, f7, v7)	\
+		generic_reg_get7(CTX, REG(reg_name), \
+				FN(reg_name, f1), v1, \
+				FN(reg_name, f2), v2, \
+				FN(reg_name, f3), v3, \
+				FN(reg_name, f4), v4, \
+				FN(reg_name, f5), v5, \
+				FN(reg_name, f6), v6, \
+				FN(reg_name, f7), v7)
+
+#define REG_GET_8(reg_name, f1, v1, f2, v2, f3, v3, f4, v4, f5, v5, f6, v6, f7, v7, f8, v8)	\
+		generic_reg_get8(CTX, REG(reg_name), \
+				FN(reg_name, f1), v1, \
+				FN(reg_name, f2), v2, \
+				FN(reg_name, f3), v3, \
+				FN(reg_name, f4), v4, \
+				FN(reg_name, f5), v5, \
+				FN(reg_name, f6), v6, \
+				FN(reg_name, f7), v7, \
+				FN(reg_name, f8), v8)
+
 /* macro to poll and wait for a register field to read back given value */
 
 #define REG_WAIT(reg_name, field, val, delay_between_poll_us, max_try)	\
@@ -389,4 +419,76 @@ uint32_t generic_reg_get5(const struct dc_context *ctx, uint32_t addr,
 		uint8_t shift4, uint32_t mask4, uint32_t *field_value4,
 		uint8_t shift5, uint32_t mask5, uint32_t *field_value5);
 
+uint32_t generic_reg_get6(const struct dc_context *ctx, uint32_t addr,
+		uint8_t shift1, uint32_t mask1, uint32_t *field_value1,
+		uint8_t shift2, uint32_t mask2, uint32_t *field_value2,
+		uint8_t shift3, uint32_t mask3, uint32_t *field_value3,
+		uint8_t shift4, uint32_t mask4, uint32_t *field_value4,
+		uint8_t shift5, uint32_t mask5, uint32_t *field_value5,
+		uint8_t shift6, uint32_t mask6, uint32_t *field_value6);
+
+uint32_t generic_reg_get7(const struct dc_context *ctx, uint32_t addr,
+		uint8_t shift1, uint32_t mask1, uint32_t *field_value1,
+		uint8_t shift2, uint32_t mask2, uint32_t *field_value2,
+		uint8_t shift3, uint32_t mask3, uint32_t *field_value3,
+		uint8_t shift4, uint32_t mask4, uint32_t *field_value4,
+		uint8_t shift5, uint32_t mask5, uint32_t *field_value5,
+		uint8_t shift6, uint32_t mask6, uint32_t *field_value6,
+		uint8_t shift7, uint32_t mask7, uint32_t *field_value7);
+
+uint32_t generic_reg_get8(const struct dc_context *ctx, uint32_t addr,
+		uint8_t shift1, uint32_t mask1, uint32_t *field_value1,
+		uint8_t shift2, uint32_t mask2, uint32_t *field_value2,
+		uint8_t shift3, uint32_t mask3, uint32_t *field_value3,
+		uint8_t shift4, uint32_t mask4, uint32_t *field_value4,
+		uint8_t shift5, uint32_t mask5, uint32_t *field_value5,
+		uint8_t shift6, uint32_t mask6, uint32_t *field_value6,
+		uint8_t shift7, uint32_t mask7, uint32_t *field_value7,
+		uint8_t shift8, uint32_t mask8, uint32_t *field_value8);
+
+
+/* indirect register access */
+
+#define IX_REG_SET_N(index_reg_name, data_reg_name, index, n, initial_val, ...)	\
+		generic_indirect_reg_update_ex(CTX, \
+				REG(index_reg_name), REG(data_reg_name), IND_REG(index), \
+				initial_val, \
+				n, __VA_ARGS__)
+
+#define IX_REG_SET_2(index_reg_name, data_reg_name, index, init_value, f1, v1, f2, v2)	\
+		IX_REG_SET_N(index_reg_name, data_reg_name, index, 2, init_value, \
+				FN(reg, f1), v1,\
+				FN(reg, f2), v2)
+
+
+#define IX_REG_READ(index_reg_name, data_reg_name, index) \
+		generic_read_indirect_reg(CTX, REG(index_reg_name), REG(data_reg_name), IND_REG(index))
+
+
+
+#define IX_REG_UPDATE_N(index_reg_name, data_reg_name, index, n, ...)	\
+		generic_indirect_reg_update_ex(CTX, \
+				REG(index_reg_name), REG(data_reg_name), IND_REG(index), \
+				IX_REG_READ(index_reg_name, data_reg_name, index), \
+				n, __VA_ARGS__)
+
+#define IX_REG_UPDATE_2(index_reg_name, data_reg_name, index, f1, v1, f2, v2)	\
+		IX_REG_UPDATE_N(index_reg_name, data_reg_name, index, 2,\
+				FN(reg, f1), v1,\
+				FN(reg, f2), v2)
+
+void generic_write_indirect_reg(const struct dc_context *ctx,
+		uint32_t addr_index, uint32_t addr_data,
+		uint32_t index, uint32_t data);
+
+uint32_t generic_read_indirect_reg(const struct dc_context *ctx,
+		uint32_t addr_index, uint32_t addr_data,
+		uint32_t index);
+
+uint32_t generic_indirect_reg_update_ex(const struct dc_context *ctx,
+		uint32_t addr_index, uint32_t addr_data,
+		uint32_t index, uint32_t reg_val, int n,
+		uint8_t shift1, uint32_t mask1, uint32_t field_value1,
+		...);
+
 #endif /* DRIVERS_GPU_DRM_AMD_DC_DEV_DC_INC_REG_HELPER_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/inc/resource.h b/drivers/gpu/drm/amd/display/dc/inc/resource.h
index 5467332faf7b..e92facbd038f 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/resource.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/resource.h
@@ -38,6 +38,7 @@ enum dce_version resource_parse_asic_id(
 
 struct resource_caps {
 	int num_timing_generator;
+	int num_opp;
 	int num_video_plane;
 	int num_audio;
 	int num_stream_encoder;
@@ -139,10 +140,6 @@ bool resource_validate_attach_surfaces(
 		struct dc_state *context,
 		const struct resource_pool *pool);
 
-void validate_guaranteed_copy_streams(
-		struct dc_state *context,
-		int max_streams);
-
 void resource_validate_ctx_update_pointer_after_copy(
 		const struct dc_state *src_ctx,
 		struct dc_state *dst_ctx);
diff --git a/drivers/gpu/drm/amd/display/dc/irq/Makefile b/drivers/gpu/drm/amd/display/dc/irq/Makefile
index 498515aad4a5..a76ee600ecee 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/irq/Makefile
@@ -60,7 +60,7 @@ AMD_DISPLAY_FILES += $(AMD_DAL_IRQ_DCE12)
 ###############################################################################
 # DCN 1x
 ###############################################################################
-ifdef CONFIG_DRM_AMD_DC_DCN1_0
+ifdef CONFIG_X86
 IRQ_DCN1 = irq_service_dcn10.o
 
 AMD_DAL_IRQ_DCN1 = $(addprefix $(AMDDALPATH)/dc/irq/dcn10/,$(IRQ_DCN1))
diff --git a/drivers/gpu/drm/amd/display/dc/irq/irq_service.c b/drivers/gpu/drm/amd/display/dc/irq/irq_service.c
index dcdfa0f01551..ae3fd0a235ba 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/irq_service.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/irq_service.c
@@ -36,7 +36,7 @@
 #include "dce120/irq_service_dce120.h"
 
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 #include "dcn10/irq_service_dcn10.h"
 #endif
 
@@ -78,7 +78,7 @@ const struct irq_source_info *find_irq_source_info(
 	struct irq_service *irq_service,
 	enum dc_irq_source source)
 {
-	if (source > DAL_IRQ_SOURCES_NUMBER || source < DC_IRQ_SOURCE_INVALID)
+	if (source >= DAL_IRQ_SOURCES_NUMBER || source < DC_IRQ_SOURCE_INVALID)
 		return NULL;
 
 	return &irq_service->info[source];
diff --git a/drivers/gpu/drm/amd/display/dc/irq_types.h b/drivers/gpu/drm/amd/display/dc/irq_types.h
index a506c2e939f5..0b5f3a278c22 100644
--- a/drivers/gpu/drm/amd/display/dc/irq_types.h
+++ b/drivers/gpu/drm/amd/display/dc/irq_types.h
@@ -26,6 +26,8 @@
 #ifndef __DAL_IRQ_TYPES_H__
 #define __DAL_IRQ_TYPES_H__
 
+#include "os_types.h"
+
 struct dc_context;
 
 typedef void (*interrupt_handler)(void *);
@@ -135,6 +137,13 @@ enum dc_irq_source {
 	DC_IRQ_SOURCE_VBLANK5,
 	DC_IRQ_SOURCE_VBLANK6,
 
+	DC_IRQ_SOURCE_DC1_VLINE0,
+	DC_IRQ_SOURCE_DC2_VLINE0,
+	DC_IRQ_SOURCE_DC3_VLINE0,
+	DC_IRQ_SOURCE_DC4_VLINE0,
+	DC_IRQ_SOURCE_DC5_VLINE0,
+	DC_IRQ_SOURCE_DC6_VLINE0,
+
 	DAL_IRQ_SOURCES_NUMBER
 };
 
diff --git a/drivers/gpu/drm/amd/display/dc/os_types.h b/drivers/gpu/drm/amd/display/dc/os_types.h
index a407892905af..c9fce9066ad8 100644
--- a/drivers/gpu/drm/amd/display/dc/os_types.h
+++ b/drivers/gpu/drm/amd/display/dc/os_types.h
@@ -48,7 +48,7 @@
 
 #define dm_error(fmt, ...) DRM_ERROR(fmt, ##__VA_ARGS__)
 
-#if defined(CONFIG_DRM_AMD_DC_DCN1_0)
+#ifdef CONFIG_X86
 #include <asm/fpu/api.h>
 #endif
 
diff --git a/drivers/gpu/drm/amd/display/include/dal_asic_id.h b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
index 9831cb5eaa7c..25029ed42d89 100644
--- a/drivers/gpu/drm/amd/display/include/dal_asic_id.h
+++ b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
@@ -86,6 +86,7 @@
 #define VI_POLARIS10_P_A0 80
 #define VI_POLARIS11_M_A0 90
 #define VI_POLARIS12_V_A0 100
+#define VI_VEGAM_A0 110
 
 #define VI_UNKNOWN 0xFF
 
@@ -98,7 +99,9 @@
 		(eChipRev < VI_POLARIS11_M_A0))
 #define ASIC_REV_IS_POLARIS11_M(eChipRev) ((eChipRev >= VI_POLARIS11_M_A0) &&  \
 		(eChipRev < VI_POLARIS12_V_A0))
-#define ASIC_REV_IS_POLARIS12_V(eChipRev) (eChipRev >= VI_POLARIS12_V_A0)
+#define ASIC_REV_IS_POLARIS12_V(eChipRev) ((eChipRev >= VI_POLARIS12_V_A0) && \
+		(eChipRev < VI_VEGAM_A0))
+#define ASIC_REV_IS_VEGAM(eChipRev) (eChipRev >= VI_VEGAM_A0)
 
 /* DCE11 */
 #define CZ_CARRIZO_A0 0x01
@@ -110,12 +113,19 @@
 	((rev >= STONEY_A0) && (rev < CZ_UNKNOWN))
 
 /* DCE12 */
+#define AI_UNKNOWN 0xFF
 
 #define AI_GREENLAND_P_A0 1
 #define AI_GREENLAND_P_A1 2
+#define AI_UNKNOWN 0xFF
 
-#define ASICREV_IS_GREENLAND_M(eChipRev)  (eChipRev < AI_UNKNOWN)
-#define ASICREV_IS_GREENLAND_P(eChipRev)  (eChipRev < AI_UNKNOWN)
+#define AI_VEGA12_P_A0 20
+#define AI_VEGA20_P_A0 40
+#define ASICREV_IS_GREENLAND_M(eChipRev)  (eChipRev < AI_VEGA12_P_A0)
+#define ASICREV_IS_GREENLAND_P(eChipRev)  (eChipRev < AI_VEGA12_P_A0)
+
+#define ASICREV_IS_VEGA12_P(eChipRev) ((eChipRev >= AI_VEGA12_P_A0) && (eChipRev < AI_VEGA20_P_A0))
+#define ASICREV_IS_VEGA20_P(eChipRev) ((eChipRev >= AI_VEGA20_P_A0) && (eChipRev < AI_UNKNOWN))
 
 /* DCN1_0 */
 #define INTERNAL_REV_RAVEN_A0             0x00    /* First spin of Raven */
diff --git a/drivers/gpu/drm/amd/display/include/dal_types.h b/drivers/gpu/drm/amd/display/include/dal_types.h
index fa543965feb5..840142b65f8b 100644
--- a/drivers/gpu/drm/amd/display/include/dal_types.h
+++ b/drivers/gpu/drm/amd/display/include/dal_types.h
@@ -40,6 +40,7 @@ enum dce_version {
 	DCE_VERSION_10_0,
 	DCE_VERSION_11_0,
 	DCE_VERSION_11_2,
+	DCE_VERSION_11_22,
 	DCE_VERSION_12_0,
 	DCE_VERSION_MAX,
 	DCN_VERSION_1_0,
diff --git a/drivers/gpu/drm/amd/display/include/ddc_service_types.h b/drivers/gpu/drm/amd/display/include/ddc_service_types.h
index 019e7a095ea1..d968956a10cd 100644
--- a/drivers/gpu/drm/amd/display/include/ddc_service_types.h
+++ b/drivers/gpu/drm/amd/display/include/ddc_service_types.h
@@ -40,7 +40,8 @@ enum ddc_result {
 	DDC_RESULT_FAILED_INCOMPLETE,
 	DDC_RESULT_FAILED_OPERATION,
 	DDC_RESULT_FAILED_INVALID_OPERATION,
-	DDC_RESULT_FAILED_BUFFER_OVERFLOW
+	DDC_RESULT_FAILED_BUFFER_OVERFLOW,
+	DDC_RESULT_FAILED_HPD_DISCON
 };
 
 enum ddc_service_type {
diff --git a/drivers/gpu/drm/amd/display/include/dpcd_defs.h b/drivers/gpu/drm/amd/display/include/dpcd_defs.h
index d8e52e3b8e3c..1c66166d0a94 100644
--- a/drivers/gpu/drm/amd/display/include/dpcd_defs.h
+++ b/drivers/gpu/drm/amd/display/include/dpcd_defs.h
@@ -27,6 +27,9 @@
 #define __DAL_DPCD_DEFS_H__
 
 #include <drm/drm_dp_helper.h>
+#ifndef DP_SINK_HW_REVISION_START // can remove this once the define gets into linux drm_dp_helper.h
+#define DP_SINK_HW_REVISION_START 0x409
+#endif
 
 enum dpcd_revision {
 	DPCD_REV_10 = 0x10,
diff --git a/drivers/gpu/drm/amd/display/include/fixed31_32.h b/drivers/gpu/drm/amd/display/include/fixed31_32.h
index 0de258622c12..52a73332befb 100644
--- a/drivers/gpu/drm/amd/display/include/fixed31_32.h
+++ b/drivers/gpu/drm/amd/display/include/fixed31_32.h
@@ -26,9 +26,20 @@
 #ifndef __DAL_FIXED31_32_H__
 #define __DAL_FIXED31_32_H__
 
-#include "os_types.h"
+#ifndef LLONG_MAX
+#define LLONG_MAX 9223372036854775807ll
+#endif
+#ifndef LLONG_MIN
+#define LLONG_MIN (-LLONG_MAX - 1ll)
+#endif
 
 #define FIXED31_32_BITS_PER_FRACTIONAL_PART 32
+#ifndef LLONG_MIN
+#define LLONG_MIN (1LL<<63)
+#endif
+#ifndef LLONG_MAX
+#define LLONG_MAX (-1LL>>1)
+#endif
 
 /*
  * @brief
@@ -44,24 +55,25 @@
  */
 
 struct fixed31_32 {
-	int64_t value;
+	long long value;
 };
 
+
 /*
  * @brief
  * Useful constants
  */
 
-static const struct fixed31_32 dal_fixed31_32_zero = { 0 };
-static const struct fixed31_32 dal_fixed31_32_epsilon = { 1LL };
-static const struct fixed31_32 dal_fixed31_32_half = { 0x80000000LL };
-static const struct fixed31_32 dal_fixed31_32_one = { 0x100000000LL };
+static const struct fixed31_32 dc_fixpt_zero = { 0 };
+static const struct fixed31_32 dc_fixpt_epsilon = { 1LL };
+static const struct fixed31_32 dc_fixpt_half = { 0x80000000LL };
+static const struct fixed31_32 dc_fixpt_one = { 0x100000000LL };
 
-static const struct fixed31_32 dal_fixed31_32_pi = { 13493037705LL };
-static const struct fixed31_32 dal_fixed31_32_two_pi = { 26986075409LL };
-static const struct fixed31_32 dal_fixed31_32_e = { 11674931555LL };
-static const struct fixed31_32 dal_fixed31_32_ln2 = { 2977044471LL };
-static const struct fixed31_32 dal_fixed31_32_ln2_div_2 = { 1488522236LL };
+static const struct fixed31_32 dc_fixpt_pi = { 13493037705LL };
+static const struct fixed31_32 dc_fixpt_two_pi = { 26986075409LL };
+static const struct fixed31_32 dc_fixpt_e = { 11674931555LL };
+static const struct fixed31_32 dc_fixpt_ln2 = { 2977044471LL };
+static const struct fixed31_32 dc_fixpt_ln2_div_2 = { 1488522236LL };
 
 /*
  * @brief
@@ -72,24 +84,19 @@ static const struct fixed31_32 dal_fixed31_32_ln2_div_2 = { 1488522236LL };
  * @brief
  * result = numerator / denominator
  */
-struct fixed31_32 dal_fixed31_32_from_fraction(
-	int64_t numerator,
-	int64_t denominator);
+struct fixed31_32 dc_fixpt_from_fraction(long long numerator, long long denominator);
 
 /*
  * @brief
  * result = arg
  */
-struct fixed31_32 dal_fixed31_32_from_int_nonconst(int64_t arg);
-static inline struct fixed31_32 dal_fixed31_32_from_int(int64_t arg)
+static inline struct fixed31_32 dc_fixpt_from_int(int arg)
 {
-	if (__builtin_constant_p(arg)) {
-		struct fixed31_32 res;
-		BUILD_BUG_ON((LONG_MIN > arg) || (arg > LONG_MAX));
-		res.value = arg << FIXED31_32_BITS_PER_FRACTIONAL_PART;
-		return res;
-	} else
-		return dal_fixed31_32_from_int_nonconst(arg);
+	struct fixed31_32 res;
+
+	res.value = (long long) arg << FIXED31_32_BITS_PER_FRACTIONAL_PART;
+
+	return res;
 }
 
 /*
@@ -101,7 +108,7 @@ static inline struct fixed31_32 dal_fixed31_32_from_int(int64_t arg)
  * @brief
  * result = -arg
  */
-static inline struct fixed31_32 dal_fixed31_32_neg(struct fixed31_32 arg)
+static inline struct fixed31_32 dc_fixpt_neg(struct fixed31_32 arg)
 {
 	struct fixed31_32 res;
 
@@ -114,10 +121,10 @@ static inline struct fixed31_32 dal_fixed31_32_neg(struct fixed31_32 arg)
  * @brief
  * result = abs(arg) := (arg >= 0) ? arg : -arg
  */
-static inline struct fixed31_32 dal_fixed31_32_abs(struct fixed31_32 arg)
+static inline struct fixed31_32 dc_fixpt_abs(struct fixed31_32 arg)
 {
 	if (arg.value < 0)
-		return dal_fixed31_32_neg(arg);
+		return dc_fixpt_neg(arg);
 	else
 		return arg;
 }
@@ -131,8 +138,7 @@ static inline struct fixed31_32 dal_fixed31_32_abs(struct fixed31_32 arg)
  * @brief
  * result = arg1 < arg2
  */
-static inline bool dal_fixed31_32_lt(struct fixed31_32 arg1,
-				     struct fixed31_32 arg2)
+static inline bool dc_fixpt_lt(struct fixed31_32 arg1, struct fixed31_32 arg2)
 {
 	return arg1.value < arg2.value;
 }
@@ -141,8 +147,7 @@ static inline bool dal_fixed31_32_lt(struct fixed31_32 arg1,
  * @brief
  * result = arg1 <= arg2
  */
-static inline bool dal_fixed31_32_le(struct fixed31_32 arg1,
-				     struct fixed31_32 arg2)
+static inline bool dc_fixpt_le(struct fixed31_32 arg1, struct fixed31_32 arg2)
 {
 	return arg1.value <= arg2.value;
 }
@@ -151,8 +156,7 @@ static inline bool dal_fixed31_32_le(struct fixed31_32 arg1,
  * @brief
  * result = arg1 == arg2
  */
-static inline bool dal_fixed31_32_eq(struct fixed31_32 arg1,
-				     struct fixed31_32 arg2)
+static inline bool dc_fixpt_eq(struct fixed31_32 arg1, struct fixed31_32 arg2)
 {
 	return arg1.value == arg2.value;
 }
@@ -161,8 +165,7 @@ static inline bool dal_fixed31_32_eq(struct fixed31_32 arg1,
  * @brief
  * result = min(arg1, arg2) := (arg1 <= arg2) ? arg1 : arg2
  */
-static inline struct fixed31_32 dal_fixed31_32_min(struct fixed31_32 arg1,
-						   struct fixed31_32 arg2)
+static inline struct fixed31_32 dc_fixpt_min(struct fixed31_32 arg1, struct fixed31_32 arg2)
 {
 	if (arg1.value <= arg2.value)
 		return arg1;
@@ -174,8 +177,7 @@ static inline struct fixed31_32 dal_fixed31_32_min(struct fixed31_32 arg1,
  * @brief
  * result = max(arg1, arg2) := (arg1 <= arg2) ? arg2 : arg1
  */
-static inline struct fixed31_32 dal_fixed31_32_max(struct fixed31_32 arg1,
-						   struct fixed31_32 arg2)
+static inline struct fixed31_32 dc_fixpt_max(struct fixed31_32 arg1, struct fixed31_32 arg2)
 {
 	if (arg1.value <= arg2.value)
 		return arg2;
@@ -189,14 +191,14 @@ static inline struct fixed31_32 dal_fixed31_32_max(struct fixed31_32 arg1,
  * result = | arg, when min_value < arg < max_value
  *          | max_value, when arg >= max_value
  */
-static inline struct fixed31_32 dal_fixed31_32_clamp(
+static inline struct fixed31_32 dc_fixpt_clamp(
 	struct fixed31_32 arg,
 	struct fixed31_32 min_value,
 	struct fixed31_32 max_value)
 {
-	if (dal_fixed31_32_le(arg, min_value))
+	if (dc_fixpt_le(arg, min_value))
 		return min_value;
-	else if (dal_fixed31_32_le(max_value, arg))
+	else if (dc_fixpt_le(max_value, arg))
 		return max_value;
 	else
 		return arg;
@@ -211,21 +213,30 @@ static inline struct fixed31_32 dal_fixed31_32_clamp(
  * @brief
  * result = arg << shift
  */
-struct fixed31_32 dal_fixed31_32_shl(
-	struct fixed31_32 arg,
-	uint8_t shift);
+static inline struct fixed31_32 dc_fixpt_shl(struct fixed31_32 arg, unsigned char shift)
+{
+	ASSERT(((arg.value >= 0) && (arg.value <= LLONG_MAX >> shift)) ||
+		((arg.value < 0) && (arg.value >= ~(LLONG_MAX >> shift))));
+
+	arg.value = arg.value << shift;
+
+	return arg;
+}
 
 /*
  * @brief
  * result = arg >> shift
  */
-static inline struct fixed31_32 dal_fixed31_32_shr(
-	struct fixed31_32 arg,
-	uint8_t shift)
+static inline struct fixed31_32 dc_fixpt_shr(struct fixed31_32 arg, unsigned char shift)
 {
-	struct fixed31_32 res;
-	res.value = arg.value >> shift;
-	return res;
+	bool negative = arg.value < 0;
+
+	if (negative)
+		arg.value = -arg.value;
+	arg.value = arg.value >> shift;
+	if (negative)
+		arg.value = -arg.value;
+	return arg;
 }
 
 /*
@@ -237,38 +248,50 @@ static inline struct fixed31_32 dal_fixed31_32_shr(
  * @brief
  * result = arg1 + arg2
  */
-struct fixed31_32 dal_fixed31_32_add(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2);
+static inline struct fixed31_32 dc_fixpt_add(struct fixed31_32 arg1, struct fixed31_32 arg2)
+{
+	struct fixed31_32 res;
+
+	ASSERT(((arg1.value >= 0) && (LLONG_MAX - arg1.value >= arg2.value)) ||
+		((arg1.value < 0) && (LLONG_MIN - arg1.value <= arg2.value)));
+
+	res.value = arg1.value + arg2.value;
+
+	return res;
+}
 
 /*
  * @brief
  * result = arg1 + arg2
  */
-static inline struct fixed31_32 dal_fixed31_32_add_int(struct fixed31_32 arg1,
-						       int32_t arg2)
+static inline struct fixed31_32 dc_fixpt_add_int(struct fixed31_32 arg1, int arg2)
 {
-	return dal_fixed31_32_add(arg1,
-				  dal_fixed31_32_from_int(arg2));
+	return dc_fixpt_add(arg1, dc_fixpt_from_int(arg2));
 }
 
 /*
  * @brief
  * result = arg1 - arg2
  */
-struct fixed31_32 dal_fixed31_32_sub(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2);
+static inline struct fixed31_32 dc_fixpt_sub(struct fixed31_32 arg1, struct fixed31_32 arg2)
+{
+	struct fixed31_32 res;
+
+	ASSERT(((arg2.value >= 0) && (LLONG_MIN + arg2.value <= arg1.value)) ||
+		((arg2.value < 0) && (LLONG_MAX + arg2.value >= arg1.value)));
+
+	res.value = arg1.value - arg2.value;
+
+	return res;
+}
 
 /*
  * @brief
  * result = arg1 - arg2
  */
-static inline struct fixed31_32 dal_fixed31_32_sub_int(struct fixed31_32 arg1,
-						       int32_t arg2)
+static inline struct fixed31_32 dc_fixpt_sub_int(struct fixed31_32 arg1, int arg2)
 {
-	return dal_fixed31_32_sub(arg1,
-				  dal_fixed31_32_from_int(arg2));
+	return dc_fixpt_sub(arg1, dc_fixpt_from_int(arg2));
 }
 
 
@@ -281,49 +304,40 @@ static inline struct fixed31_32 dal_fixed31_32_sub_int(struct fixed31_32 arg1,
  * @brief
  * result = arg1 * arg2
  */
-struct fixed31_32 dal_fixed31_32_mul(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2);
+struct fixed31_32 dc_fixpt_mul(struct fixed31_32 arg1, struct fixed31_32 arg2);
 
 
 /*
  * @brief
  * result = arg1 * arg2
  */
-static inline struct fixed31_32 dal_fixed31_32_mul_int(struct fixed31_32 arg1,
-						       int32_t arg2)
+static inline struct fixed31_32 dc_fixpt_mul_int(struct fixed31_32 arg1, int arg2)
 {
-	return dal_fixed31_32_mul(arg1,
-				  dal_fixed31_32_from_int(arg2));
+	return dc_fixpt_mul(arg1, dc_fixpt_from_int(arg2));
 }
 
 /*
  * @brief
  * result = square(arg) := arg * arg
  */
-struct fixed31_32 dal_fixed31_32_sqr(
-	struct fixed31_32 arg);
+struct fixed31_32 dc_fixpt_sqr(struct fixed31_32 arg);
 
 /*
  * @brief
  * result = arg1 / arg2
  */
-static inline struct fixed31_32 dal_fixed31_32_div_int(struct fixed31_32 arg1,
-						       int64_t arg2)
+static inline struct fixed31_32 dc_fixpt_div_int(struct fixed31_32 arg1, long long arg2)
 {
-	return dal_fixed31_32_from_fraction(arg1.value,
-					    dal_fixed31_32_from_int(arg2).value);
+	return dc_fixpt_from_fraction(arg1.value, dc_fixpt_from_int(arg2).value);
 }
 
 /*
  * @brief
  * result = arg1 / arg2
  */
-static inline struct fixed31_32 dal_fixed31_32_div(struct fixed31_32 arg1,
-						   struct fixed31_32 arg2)
+static inline struct fixed31_32 dc_fixpt_div(struct fixed31_32 arg1, struct fixed31_32 arg2)
 {
-	return dal_fixed31_32_from_fraction(arg1.value,
-					    arg2.value);
+	return dc_fixpt_from_fraction(arg1.value, arg2.value);
 }
 
 /*
@@ -338,8 +352,7 @@ static inline struct fixed31_32 dal_fixed31_32_div(struct fixed31_32 arg1,
  * @note
  * No special actions taken in case argument is zero.
  */
-struct fixed31_32 dal_fixed31_32_recip(
-	struct fixed31_32 arg);
+struct fixed31_32 dc_fixpt_recip(struct fixed31_32 arg);
 
 /*
  * @brief
@@ -354,8 +367,7 @@ struct fixed31_32 dal_fixed31_32_recip(
  * Argument specified in radians,
  * internally it's normalized to [-2pi...2pi] range.
  */
-struct fixed31_32 dal_fixed31_32_sinc(
-	struct fixed31_32 arg);
+struct fixed31_32 dc_fixpt_sinc(struct fixed31_32 arg);
 
 /*
  * @brief
@@ -365,8 +377,7 @@ struct fixed31_32 dal_fixed31_32_sinc(
  * Argument specified in radians,
  * internally it's normalized to [-2pi...2pi] range.
  */
-struct fixed31_32 dal_fixed31_32_sin(
-	struct fixed31_32 arg);
+struct fixed31_32 dc_fixpt_sin(struct fixed31_32 arg);
 
 /*
  * @brief
@@ -378,8 +389,7 @@ struct fixed31_32 dal_fixed31_32_sin(
  * passing arguments outside that range
  * will cause incorrect result!
  */
-struct fixed31_32 dal_fixed31_32_cos(
-	struct fixed31_32 arg);
+struct fixed31_32 dc_fixpt_cos(struct fixed31_32 arg);
 
 /*
  * @brief
@@ -393,8 +403,7 @@ struct fixed31_32 dal_fixed31_32_cos(
  * @note
  * Currently, function is verified for abs(arg) <= 1.
  */
-struct fixed31_32 dal_fixed31_32_exp(
-	struct fixed31_32 arg);
+struct fixed31_32 dc_fixpt_exp(struct fixed31_32 arg);
 
 /*
  * @brief
@@ -406,8 +415,7 @@ struct fixed31_32 dal_fixed31_32_exp(
  * Currently, no special actions taken
  * in case of invalid argument(s). Take care!
  */
-struct fixed31_32 dal_fixed31_32_log(
-	struct fixed31_32 arg);
+struct fixed31_32 dc_fixpt_log(struct fixed31_32 arg);
 
 /*
  * @brief
@@ -421,9 +429,13 @@ struct fixed31_32 dal_fixed31_32_log(
  * @note
  * Currently, abs(arg1) should be less than 1. Take care!
  */
-struct fixed31_32 dal_fixed31_32_pow(
-	struct fixed31_32 arg1,
-	struct fixed31_32 arg2);
+static inline struct fixed31_32 dc_fixpt_pow(struct fixed31_32 arg1, struct fixed31_32 arg2)
+{
+	return dc_fixpt_exp(
+		dc_fixpt_mul(
+			dc_fixpt_log(arg1),
+			arg2));
+}
 
 /*
  * @brief
@@ -434,22 +446,56 @@ struct fixed31_32 dal_fixed31_32_pow(
  * @brief
  * result = floor(arg) := greatest integer lower than or equal to arg
  */
-int32_t dal_fixed31_32_floor(
-	struct fixed31_32 arg);
+static inline int dc_fixpt_floor(struct fixed31_32 arg)
+{
+	unsigned long long arg_value = arg.value > 0 ? arg.value : -arg.value;
+
+	if (arg.value >= 0)
+		return (int)(arg_value >> FIXED31_32_BITS_PER_FRACTIONAL_PART);
+	else
+		return -(int)(arg_value >> FIXED31_32_BITS_PER_FRACTIONAL_PART);
+}
 
 /*
  * @brief
  * result = round(arg) := integer nearest to arg
  */
-int32_t dal_fixed31_32_round(
-	struct fixed31_32 arg);
+static inline int dc_fixpt_round(struct fixed31_32 arg)
+{
+	unsigned long long arg_value = arg.value > 0 ? arg.value : -arg.value;
+
+	const long long summand = dc_fixpt_half.value;
+
+	ASSERT(LLONG_MAX - (long long)arg_value >= summand);
+
+	arg_value += summand;
+
+	if (arg.value >= 0)
+		return (int)(arg_value >> FIXED31_32_BITS_PER_FRACTIONAL_PART);
+	else
+		return -(int)(arg_value >> FIXED31_32_BITS_PER_FRACTIONAL_PART);
+}
 
 /*
  * @brief
  * result = ceil(arg) := lowest integer greater than or equal to arg
  */
-int32_t dal_fixed31_32_ceil(
-	struct fixed31_32 arg);
+static inline int dc_fixpt_ceil(struct fixed31_32 arg)
+{
+	unsigned long long arg_value = arg.value > 0 ? arg.value : -arg.value;
+
+	const long long summand = dc_fixpt_one.value -
+		dc_fixpt_epsilon.value;
+
+	ASSERT(LLONG_MAX - (long long)arg_value >= summand);
+
+	arg_value += summand;
+
+	if (arg.value >= 0)
+		return (int)(arg_value >> FIXED31_32_BITS_PER_FRACTIONAL_PART);
+	else
+		return -(int)(arg_value >> FIXED31_32_BITS_PER_FRACTIONAL_PART);
+}
 
 /* the following two function are used in scaler hw programming to convert fixed
  * point value to format 2 bits from integer part and 19 bits from fractional
@@ -457,20 +503,33 @@ int32_t dal_fixed31_32_ceil(
  * fractional
  */
 
-uint32_t dal_fixed31_32_u2d19(
-	struct fixed31_32 arg);
+unsigned int dc_fixpt_u3d19(struct fixed31_32 arg);
 
-uint32_t dal_fixed31_32_u0d19(
-	struct fixed31_32 arg);
+unsigned int dc_fixpt_u2d19(struct fixed31_32 arg);
 
+unsigned int dc_fixpt_u0d19(struct fixed31_32 arg);
 
-uint32_t dal_fixed31_32_clamp_u0d14(
-	struct fixed31_32 arg);
+unsigned int dc_fixpt_clamp_u0d14(struct fixed31_32 arg);
 
-uint32_t dal_fixed31_32_clamp_u0d10(
-	struct fixed31_32 arg);
+unsigned int dc_fixpt_clamp_u0d10(struct fixed31_32 arg);
 
-int32_t dal_fixed31_32_s4d19(
-	struct fixed31_32 arg);
+int dc_fixpt_s4d19(struct fixed31_32 arg);
+
+static inline struct fixed31_32 dc_fixpt_truncate(struct fixed31_32 arg, unsigned int frac_bits)
+{
+	bool negative = arg.value < 0;
+
+	if (frac_bits >= FIXED31_32_BITS_PER_FRACTIONAL_PART) {
+		ASSERT(frac_bits == FIXED31_32_BITS_PER_FRACTIONAL_PART);
+		return arg;
+	}
+
+	if (negative)
+		arg.value = -arg.value;
+	arg.value &= (~0LL) << (FIXED31_32_BITS_PER_FRACTIONAL_PART - frac_bits);
+	if (negative)
+		arg.value = -arg.value;
+	return arg;
+}
 
 #endif
diff --git a/drivers/gpu/drm/amd/display/include/fixed32_32.h b/drivers/gpu/drm/amd/display/include/fixed32_32.h
deleted file mode 100644
index 9c70341fe026..000000000000
--- a/drivers/gpu/drm/amd/display/include/fixed32_32.h
+++ /dev/null
@@ -1,129 +0,0 @@
-/*
- * Copyright 2012-15 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: AMD
- *
- */
-
-
-#ifndef __DAL_FIXED32_32_H__
-#define __DAL_FIXED32_32_H__
-
-#include "os_types.h"
-
-struct fixed32_32 {
-	uint64_t value;
-};
-
-static const struct fixed32_32 dal_fixed32_32_zero = { 0 };
-static const struct fixed32_32 dal_fixed32_32_one = { 0x100000000LL };
-static const struct fixed32_32 dal_fixed32_32_half = { 0x80000000LL };
-
-struct fixed32_32 dal_fixed32_32_from_fraction(uint32_t n, uint32_t d);
-static inline struct fixed32_32 dal_fixed32_32_from_int(uint32_t value)
-{
-	struct fixed32_32 fx;
-
-	fx.value = (uint64_t)value<<32;
-	return fx;
-}
-
-struct fixed32_32 dal_fixed32_32_add(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs);
-struct fixed32_32 dal_fixed32_32_add_int(
-	struct fixed32_32 lhs,
-	uint32_t rhs);
-struct fixed32_32 dal_fixed32_32_sub(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs);
-struct fixed32_32 dal_fixed32_32_sub_int(
-	struct fixed32_32 lhs,
-	uint32_t rhs);
-struct fixed32_32 dal_fixed32_32_mul(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs);
-struct fixed32_32 dal_fixed32_32_mul_int(
-	struct fixed32_32 lhs,
-	uint32_t rhs);
-struct fixed32_32 dal_fixed32_32_div(
-	struct fixed32_32 lhs,
-	struct fixed32_32 rhs);
-struct fixed32_32 dal_fixed32_32_div_int(
-	struct fixed32_32 lhs,
-	uint32_t rhs);
-
-static inline struct fixed32_32 dal_fixed32_32_min(struct fixed32_32 lhs,
-						   struct fixed32_32 rhs)
-{
-	return (lhs.value < rhs.value) ? lhs : rhs;
-}
-
-static inline struct fixed32_32 dal_fixed32_32_max(struct fixed32_32 lhs,
-						   struct fixed32_32 rhs)
-{
-	return (lhs.value > rhs.value) ? lhs : rhs;
-}
-
-static inline bool dal_fixed32_32_gt(struct fixed32_32 lhs, struct fixed32_32 rhs)
-{
-	return lhs.value > rhs.value;
-}
-
-static inline bool dal_fixed32_32_gt_int(struct fixed32_32 lhs, uint32_t rhs)
-{
-	return lhs.value > ((uint64_t)rhs<<32);
-}
-
-static inline bool dal_fixed32_32_lt(struct fixed32_32 lhs, struct fixed32_32 rhs)
-{
-	return lhs.value < rhs.value;
-}
-
-static inline bool dal_fixed32_32_lt_int(struct fixed32_32 lhs, uint32_t rhs)
-{
-	return lhs.value < ((uint64_t)rhs<<32);
-}
-
-static inline bool dal_fixed32_32_le(struct fixed32_32 lhs, struct fixed32_32 rhs)
-{
-	return lhs.value <= rhs.value;
-}
-
-static inline bool dal_fixed32_32_le_int(struct fixed32_32 lhs, uint32_t rhs)
-{
-	return lhs.value <= ((uint64_t)rhs<<32);
-}
-
-static inline bool dal_fixed32_32_eq(struct fixed32_32 lhs, struct fixed32_32 rhs)
-{
-	return lhs.value == rhs.value;
-}
-
-uint32_t dal_fixed32_32_ceil(struct fixed32_32 value);
-static inline uint32_t dal_fixed32_32_floor(struct fixed32_32 value)
-{
-	return value.value>>32;
-}
-
-uint32_t dal_fixed32_32_round(struct fixed32_32 value);
-
-#endif
diff --git a/drivers/gpu/drm/amd/display/include/grph_object_defs.h b/drivers/gpu/drm/amd/display/include/grph_object_defs.h
index 2941b882b0b6..58bb42ed85ca 100644
--- a/drivers/gpu/drm/amd/display/include/grph_object_defs.h
+++ b/drivers/gpu/drm/amd/display/include/grph_object_defs.h
@@ -37,6 +37,10 @@
  * ********************************************************************
  */
 
+#define MAX_CONNECTOR_NUMBER_PER_SLOT	(16)
+#define MAX_BOARD_SLOTS					(4)
+#define INVALID_CONNECTOR_INDEX			((unsigned int)(-1))
+
 /* HPD unit id - HW direct translation */
 enum hpd_source_id {
 	HPD_SOURCEID1 = 0,
@@ -136,5 +140,47 @@ enum sync_source {
 	SYNC_SOURCE_DUAL_GPU_PIN
 };
 
+/* connector sizes in millimeters - from BiosParserTypes.hpp */
+#define CONNECTOR_SIZE_DVI			40
+#define CONNECTOR_SIZE_VGA			32
+#define CONNECTOR_SIZE_HDMI			16
+#define CONNECTOR_SIZE_DP			16
+#define CONNECTOR_SIZE_MINI_DP			9
+#define CONNECTOR_SIZE_UNKNOWN			30
+
+enum connector_layout_type {
+	CONNECTOR_LAYOUT_TYPE_UNKNOWN,
+	CONNECTOR_LAYOUT_TYPE_DVI_D,
+	CONNECTOR_LAYOUT_TYPE_DVI_I,
+	CONNECTOR_LAYOUT_TYPE_VGA,
+	CONNECTOR_LAYOUT_TYPE_HDMI,
+	CONNECTOR_LAYOUT_TYPE_DP,
+	CONNECTOR_LAYOUT_TYPE_MINI_DP,
+};
+struct connector_layout_info {
+	struct graphics_object_id connector_id;
+	enum connector_layout_type connector_type;
+	unsigned int length;
+	unsigned int position;  /* offset in mm from right side of the board */
+};
+
+/* length and width in mm */
+struct slot_layout_info {
+	unsigned int length;
+	unsigned int width;
+	unsigned int num_of_connectors;
+	struct connector_layout_info connectors[MAX_CONNECTOR_NUMBER_PER_SLOT];
+};
+
+struct board_layout_info {
+	unsigned int num_of_slots;
 
+	/* indicates valid information in bracket layout structure. */
+	unsigned int is_number_of_slots_valid : 1;
+	unsigned int is_slots_size_valid : 1;
+	unsigned int is_connector_offsets_valid : 1;
+	unsigned int is_connector_lengths_valid : 1;
+
+	struct slot_layout_info slots[MAX_BOARD_SLOTS];
+};
 #endif
diff --git a/drivers/gpu/drm/amd/display/include/grph_object_id.h b/drivers/gpu/drm/amd/display/include/grph_object_id.h
index c4197432eb7c..33b3d755fe65 100644
--- a/drivers/gpu/drm/amd/display/include/grph_object_id.h
+++ b/drivers/gpu/drm/amd/display/include/grph_object_id.h
@@ -197,6 +197,11 @@ enum transmitter_color_depth {
 	TRANSMITTER_COLOR_DEPTH_48       /* 16 bits */
 };
 
+enum dp_alt_mode {
+	DP_Alt_mode__Unknown = 0,
+	DP_Alt_mode__Connect,
+	DP_Alt_mode__NoConnect,
+};
 /*
  *****************************************************************************
  * graphics_object_id struct
@@ -287,4 +292,15 @@ static inline enum engine_id dal_graphics_object_id_get_engine_id(
 		return (enum engine_id) id.id;
 	return ENGINE_ID_UNKNOWN;
 }
+
+static inline bool dal_graphics_object_id_equal(
+	struct graphics_object_id id_1,
+	struct graphics_object_id id_2)
+{
+	if ((id_1.id == id_2.id) && (id_1.enum_id == id_2.enum_id) &&
+		(id_1.type == id_2.type)) {
+		return true;
+	}
+	return false;
+}
 #endif
diff --git a/drivers/gpu/drm/amd/display/include/logger_interface.h b/drivers/gpu/drm/amd/display/include/logger_interface.h
index 28dee960d509..e3c79616682d 100644
--- a/drivers/gpu/drm/amd/display/include/logger_interface.h
+++ b/drivers/gpu/drm/amd/display/include/logger_interface.h
@@ -40,47 +40,7 @@ struct dc_state;
  *
  */
 
-struct dal_logger *dal_logger_create(struct dc_context *ctx, uint32_t log_mask);
-
-uint32_t dal_logger_destroy(struct dal_logger **logger);
-
-void dm_logger_flush_buffer(struct dal_logger *logger, bool should_warn);
-
-void dm_logger_write(
-		struct dal_logger *logger,
-		enum dc_log_type log_type,
-		const char *msg,
-		...);
-
-void dm_logger_append(
-		struct log_entry *entry,
-		const char *msg,
-		...);
-
-void dm_logger_append_va(
-		struct log_entry *entry,
-		const char *msg,
-		va_list args);
-
-void dm_logger_open(
-		struct dal_logger *logger,
-		struct log_entry *entry,
-		enum dc_log_type log_type);
-
-void dm_logger_close(struct log_entry *entry);
-
-void dc_conn_log(struct dc_context *ctx,
-		const struct dc_link *link,
-		uint8_t *hex_data,
-		int hex_data_count,
-		enum dc_log_type event,
-		const char *msg,
-		...);
-
-void logger_write(struct dal_logger *logger,
-		enum dc_log_type log_type,
-		const char *msg,
-		void *paralist);
+void dc_conn_log_hex_linux(const uint8_t *hex_data, int hex_data_count);
 
 void pre_surface_trace(
 		struct dc *dc,
@@ -106,28 +66,31 @@ void context_clock_trace(
  * marked by this macro.
  * Note that the message will be printed exactly once for every function
  * it is used in order to avoid repeating of the same message. */
+
 #define DAL_LOGGER_NOT_IMPL(fmt, ...) \
-{ \
-	static bool print_not_impl = true; \
-\
-	if (print_not_impl == true) { \
-		print_not_impl = false; \
-		dm_logger_write(ctx->logger, LOG_WARNING, \
-		"DAL_NOT_IMPL: " fmt, ##__VA_ARGS__); \
-	} \
-}
+	do { \
+		static bool print_not_impl = true; \
+		if (print_not_impl == true) { \
+			print_not_impl = false; \
+			DRM_WARN("DAL_NOT_IMPL: " fmt, ##__VA_ARGS__); \
+		} \
+	} while (0)
 
 /******************************************************************************
  * Convenience macros to save on typing.
  *****************************************************************************/
 
 #define DC_ERROR(...) \
-	dm_logger_write(dc_ctx->logger, LOG_ERROR, \
-		__VA_ARGS__)
+		do { \
+			(void)(dc_ctx); \
+			DC_LOG_ERROR(__VA_ARGS__); \
+		} while (0)
 
 #define DC_SYNC_INFO(...) \
-	dm_logger_write(dc_ctx->logger, LOG_SYNC, \
-		__VA_ARGS__)
+		do { \
+			(void)(dc_ctx); \
+			DC_LOG_SYNC(__VA_ARGS__); \
+		} while (0)
 
 /* Connectivity log format:
  * [time stamp]   [drm] [Major_minor] [connector name] message.....
@@ -137,20 +100,30 @@ void context_clock_trace(
  */
 
 #define CONN_DATA_DETECT(link, hex_data, hex_len, ...) \
-		dc_conn_log(link->ctx, link, hex_data, hex_len, \
-				LOG_EVENT_DETECTION, ##__VA_ARGS__)
+		do { \
+			(void)(link); \
+			dc_conn_log_hex_linux(hex_data, hex_len); \
+			DC_LOG_EVENT_DETECTION(__VA_ARGS__); \
+		} while (0)
 
 #define CONN_DATA_LINK_LOSS(link, hex_data, hex_len, ...) \
-		dc_conn_log(link->ctx, link, hex_data, hex_len, \
-				LOG_EVENT_LINK_LOSS, ##__VA_ARGS__)
+		do { \
+			(void)(link); \
+			dc_conn_log_hex_linux(hex_data, hex_len); \
+			DC_LOG_EVENT_LINK_LOSS(__VA_ARGS__); \
+		} while (0)
 
 #define CONN_MSG_LT(link, ...) \
-		dc_conn_log(link->ctx, link, NULL, 0, \
-				LOG_EVENT_LINK_TRAINING, ##__VA_ARGS__)
+		do { \
+			(void)(link); \
+			DC_LOG_EVENT_LINK_TRAINING(__VA_ARGS__); \
+		} while (0)
 
 #define CONN_MSG_MODE(link, ...) \
-		dc_conn_log(link->ctx, link, NULL, 0, \
-				LOG_EVENT_MODE_SET, ##__VA_ARGS__)
+		do { \
+			(void)(link); \
+			DC_LOG_EVENT_MODE_SET(__VA_ARGS__); \
+		} while (0)
 
 /*
  * Display Test Next logging
@@ -165,29 +138,21 @@ void context_clock_trace(
 	dm_dtn_log_end(dc_ctx)
 
 #define PERFORMANCE_TRACE_START() \
-	unsigned long long perf_trc_start_stmp = dm_get_timestamp(dc->ctx); \
-	unsigned long long perf_trc_start_log_msk = dc->ctx->logger->mask; \
-	unsigned int perf_trc_start_log_flags = dc->ctx->logger->flags.value; \
-	if (dc->debug.performance_trace) {\
-		dm_logger_flush_buffer(dc->ctx->logger, false);\
-		dc->ctx->logger->mask = 1<<LOG_PERF_TRACE;\
-		dc->ctx->logger->flags.bits.ENABLE_CONSOLE = 0;\
-		dc->ctx->logger->flags.bits.ENABLE_BUFFER = 1;\
-	}
-
-#define PERFORMANCE_TRACE_END() do {\
-	unsigned long long perf_trc_end_stmp = dm_get_timestamp(dc->ctx);\
-	if (dc->debug.performance_trace) {\
-		dm_logger_write(dc->ctx->logger, \
-				LOG_PERF_TRACE, \
-				"%s duration: %d ticks\n", __func__,\
+	unsigned long long perf_trc_start_stmp = dm_get_timestamp(dc->ctx)
+
+#define PERFORMANCE_TRACE_END() \
+	do { \
+		unsigned long long perf_trc_end_stmp = dm_get_timestamp(dc->ctx); \
+		if (dc->debug.performance_trace) { \
+			DC_LOG_PERF_TRACE("%s duration: %lld ticks\n", __func__, \
 				perf_trc_end_stmp - perf_trc_start_stmp); \
-		if (perf_trc_start_log_msk != 1<<LOG_PERF_TRACE) {\
-			dc->ctx->logger->mask = perf_trc_start_log_msk;\
-			dc->ctx->logger->flags.value = perf_trc_start_log_flags;\
-			dm_logger_flush_buffer(dc->ctx->logger, false);\
 		} \
-	} \
-} while (0)
+	} while (0)
+
+#define DISPLAY_STATS_BEGIN(entry) (void)(entry)
+
+#define DISPLAY_STATS(msg, ...) DC_LOG_PERF_TRACE(msg, __VA_ARGS__)
+
+#define DISPLAY_STATS_END(entry) (void)(entry)
 
 #endif /* __DAL_LOGGER_INTERFACE_H__ */
diff --git a/drivers/gpu/drm/amd/display/include/logger_types.h b/drivers/gpu/drm/amd/display/include/logger_types.h
index 427796bdc14a..ad3695e67b76 100644
--- a/drivers/gpu/drm/amd/display/include/logger_types.h
+++ b/drivers/gpu/drm/amd/display/include/logger_types.h
@@ -29,39 +29,39 @@
 #include "os_types.h"
 
 #define MAX_NAME_LEN 32
-#define DC_LOG_ERROR(a, ...) dm_logger_write(DC_LOGGER, LOG_ERROR, a, ## __VA_ARGS__)
-#define DC_LOG_WARNING(a, ...) dm_logger_write(DC_LOGGER, LOG_WARNING, a, ## __VA_ARGS__)
-#define DC_LOG_DEBUG(a, ...) dm_logger_write(DC_LOGGER, LOG_DEBUG, a, ## __VA_ARGS__)
-#define DC_LOG_DC(a, ...) dm_logger_write(DC_LOGGER, LOG_DC, a, ## __VA_ARGS__)
-#define DC_LOG_DTN(a, ...) dm_logger_write(DC_LOGGER, LOG_DTN, a, ## __VA_ARGS__)
-#define DC_LOG_SURFACE(a, ...) dm_logger_write(DC_LOGGER, LOG_SURFACE, a, ## __VA_ARGS__)
-#define DC_LOG_HW_HOTPLUG(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_HOTPLUG, a, ## __VA_ARGS__)
-#define DC_LOG_HW_LINK_TRAINING(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_LINK_TRAINING, a, ## __VA_ARGS__)
-#define DC_LOG_HW_SET_MODE(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_SET_MODE, a, ## __VA_ARGS__)
-#define DC_LOG_HW_RESUME_S3(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_RESUME_S3, a, ## __VA_ARGS__)
-#define DC_LOG_HW_AUDIO(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_AUDIO, a, ## __VA_ARGS__)
-#define DC_LOG_HW_HPD_IRQ(a, ...) dm_logger_write(DC_LOGGER, LOG_HW_HPD_IRQ, a, ## __VA_ARGS__)
-#define DC_LOG_MST(a, ...) dm_logger_write(DC_LOGGER, LOG_MST, a, ## __VA_ARGS__)
-#define DC_LOG_SCALER(a, ...) dm_logger_write(DC_LOGGER, LOG_SCALER, a, ## __VA_ARGS__)
-#define DC_LOG_BIOS(a, ...) dm_logger_write(DC_LOGGER, LOG_BIOS, a, ## __VA_ARGS__)
-#define DC_LOG_BANDWIDTH_CALCS(a, ...) dm_logger_write(DC_LOGGER, LOG_BANDWIDTH_CALCS, a, ## __VA_ARGS__)
-#define DC_LOG_BANDWIDTH_VALIDATION(a, ...) dm_logger_write(DC_LOGGER, LOG_BANDWIDTH_VALIDATION, a, ## __VA_ARGS__)
-#define DC_LOG_I2C_AUX(a, ...) dm_logger_write(DC_LOGGER, LOG_I2C_AUX, a, ## __VA_ARGS__)
-#define DC_LOG_SYNC(a, ...) dm_logger_write(DC_LOGGER, LOG_SYNC, a, ## __VA_ARGS__)
-#define DC_LOG_BACKLIGHT(a, ...) dm_logger_write(DC_LOGGER, LOG_BACKLIGHT, a, ## __VA_ARGS__)
-#define DC_LOG_FEATURE_OVERRIDE(a, ...) dm_logger_write(DC_LOGGER, LOG_FEATURE_OVERRIDE, a, ## __VA_ARGS__)
-#define DC_LOG_DETECTION_EDID_PARSER(a, ...) dm_logger_write(DC_LOGGER, LOG_DETECTION_EDID_PARSER, a, ## __VA_ARGS__)
-#define DC_LOG_DETECTION_DP_CAPS(a, ...) dm_logger_write(DC_LOGGER, LOG_DETECTION_DP_CAPS, a, ## __VA_ARGS__)
-#define DC_LOG_RESOURCE(a, ...) dm_logger_write(DC_LOGGER, LOG_RESOURCE, a, ## __VA_ARGS__)
-#define DC_LOG_DML(a, ...) dm_logger_write(DC_LOGGER, LOG_DML, a, ## __VA_ARGS__)
-#define DC_LOG_EVENT_MODE_SET(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_MODE_SET, a, ## __VA_ARGS__)
-#define DC_LOG_EVENT_DETECTION(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_DETECTION, a, ## __VA_ARGS__)
-#define DC_LOG_EVENT_LINK_TRAINING(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_LINK_TRAINING, a, ## __VA_ARGS__)
-#define DC_LOG_EVENT_LINK_LOSS(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_LINK_LOSS, a, ## __VA_ARGS__)
-#define DC_LOG_EVENT_UNDERFLOW(a, ...) dm_logger_write(DC_LOGGER, LOG_EVENT_UNDERFLOW, a, ## __VA_ARGS__)
-#define DC_LOG_IF_TRACE(a, ...) dm_logger_write(DC_LOGGER, LOG_IF_TRACE, a, ## __VA_ARGS__)
-#define DC_LOG_PERF_TRACE(a, ...) dm_logger_write(DC_LOGGER, LOG_PERF_TRACE, a, ## __VA_ARGS__)
 
+#define DC_LOG_ERROR(...) DRM_ERROR(__VA_ARGS__)
+#define DC_LOG_WARNING(...) DRM_WARN(__VA_ARGS__)
+#define DC_LOG_DEBUG(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_DC(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_DTN(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_SURFACE(...) pr_debug("[SURFACE]:"__VA_ARGS__)
+#define DC_LOG_HW_HOTPLUG(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_HW_LINK_TRAINING(...) pr_debug("[HW_LINK_TRAINING]:"__VA_ARGS__)
+#define DC_LOG_HW_SET_MODE(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_HW_RESUME_S3(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_HW_AUDIO(...) pr_debug("[HW_AUDIO]:"__VA_ARGS__)
+#define DC_LOG_HW_HPD_IRQ(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_MST(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_SCALER(...) pr_debug("[SCALER]:"__VA_ARGS__)
+#define DC_LOG_BIOS(...) pr_debug("[BIOS]:"__VA_ARGS__)
+#define DC_LOG_BANDWIDTH_CALCS(...) pr_debug("[BANDWIDTH_CALCS]:"__VA_ARGS__)
+#define DC_LOG_BANDWIDTH_VALIDATION(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_I2C_AUX(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_SYNC(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_BACKLIGHT(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_FEATURE_OVERRIDE(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_DETECTION_EDID_PARSER(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_DETECTION_DP_CAPS(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_RESOURCE(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_DML(...) pr_debug("[DML]:"__VA_ARGS__)
+#define DC_LOG_EVENT_MODE_SET(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_EVENT_DETECTION(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_EVENT_LINK_TRAINING(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_EVENT_LINK_LOSS(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_EVENT_UNDERFLOW(...) DRM_DEBUG_KMS(__VA_ARGS__)
+#define DC_LOG_IF_TRACE(...) pr_debug("[IF_TRACE]:"__VA_ARGS__)
+#define DC_LOG_PERF_TRACE(...) DRM_DEBUG_KMS(__VA_ARGS__)
 
 struct dal_logger;
 
@@ -98,7 +98,7 @@ enum dc_log_type {
 	LOG_EVENT_UNDERFLOW,
 	LOG_IF_TRACE,
 	LOG_PERF_TRACE,
-	LOG_PROFILING,
+	LOG_DISPLAYSTATS,
 
 	LOG_SECTION_TOTAL_COUNT
 };
@@ -138,63 +138,4 @@ enum dc_log_type {
 		(1 << LOG_HW_AUDIO)| \
 		(1 << LOG_BANDWIDTH_CALCS)*/
 
-union logger_flags {
-	struct {
-		uint32_t ENABLE_CONSOLE:1; /* Print to console */
-		uint32_t ENABLE_BUFFER:1; /* Print to buffer */
-		uint32_t RESERVED:30;
-	} bits;
-	uint32_t value;
-};
-
-struct log_entry {
-	struct dal_logger *logger;
-	enum dc_log_type type;
-
-	char *buf;
-	uint32_t buf_offset;
-	uint32_t max_buf_bytes;
-};
-
-/**
-* Structure for enumerating log types
-*/
-struct dc_log_type_info {
-	enum dc_log_type type;
-	char name[MAX_NAME_LEN];
-};
-
-/* Structure for keeping track of offsets, buffer, etc */
-
-#define DAL_LOGGER_BUFFER_MAX_SIZE 2048
-
-/*Connectivity log needs to output EDID, which needs at lease 256x3 bytes,
- * change log line size to 896 to meet the request.
- */
-#define LOG_MAX_LINE_SIZE 896
-
-struct dal_logger {
-
-	/* How far into the circular buffer has been read by dsat
-	 * Read offset should never cross write offset. Write \0's to
-	 * read data just to be sure?
-	 */
-	uint32_t buffer_read_offset;
-
-	/* How far into the circular buffer we have written
-	 * Write offset should never cross read offset
-	 */
-	uint32_t buffer_write_offset;
-
-	uint32_t open_count;
-
-	char *log_buffer;	/* Pointer to malloc'ed buffer */
-	uint32_t log_buffer_size; /* Size of circular buffer */
-
-	uint32_t mask; /*array of masks for major elements*/
-
-	union logger_flags flags;
-	struct dc_context *ctx;
-};
-
 #endif /* __DAL_LOGGER_TYPES_H__ */
diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
index e7e374f56864..ee69c949bfbf 100644
--- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
+++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.c
@@ -43,7 +43,7 @@ static bool de_pq_initialized; /* = false; */
 /* one-time setup of X points */
 void setup_x_points_distribution(void)
 {
-	struct fixed31_32 region_size = dal_fixed31_32_from_int(128);
+	struct fixed31_32 region_size = dc_fixpt_from_int(128);
 	int32_t segment;
 	uint32_t seg_offset;
 	uint32_t index;
@@ -53,8 +53,8 @@ void setup_x_points_distribution(void)
 	coordinates_x[MAX_HW_POINTS + 1].x = region_size;
 
 	for (segment = 6; segment > (6 - NUM_REGIONS); segment--) {
-		region_size = dal_fixed31_32_div_int(region_size, 2);
-		increment = dal_fixed31_32_div_int(region_size,
+		region_size = dc_fixpt_div_int(region_size, 2);
+		increment = dc_fixpt_div_int(region_size,
 						NUM_PTS_IN_REGION);
 		seg_offset = (segment + (NUM_REGIONS - 7)) * NUM_PTS_IN_REGION;
 		coordinates_x[seg_offset].x = region_size;
@@ -62,7 +62,7 @@ void setup_x_points_distribution(void)
 		for (index = seg_offset + 1;
 				index < seg_offset + NUM_PTS_IN_REGION;
 				index++) {
-			coordinates_x[index].x = dal_fixed31_32_add
+			coordinates_x[index].x = dc_fixpt_add
 					(coordinates_x[index-1].x, increment);
 		}
 	}
@@ -72,65 +72,122 @@ static void compute_pq(struct fixed31_32 in_x, struct fixed31_32 *out_y)
 {
 	/* consts for PQ gamma formula. */
 	const struct fixed31_32 m1 =
-		dal_fixed31_32_from_fraction(159301758, 1000000000);
+		dc_fixpt_from_fraction(159301758, 1000000000);
 	const struct fixed31_32 m2 =
-		dal_fixed31_32_from_fraction(7884375, 100000);
+		dc_fixpt_from_fraction(7884375, 100000);
 	const struct fixed31_32 c1 =
-		dal_fixed31_32_from_fraction(8359375, 10000000);
+		dc_fixpt_from_fraction(8359375, 10000000);
 	const struct fixed31_32 c2 =
-		dal_fixed31_32_from_fraction(188515625, 10000000);
+		dc_fixpt_from_fraction(188515625, 10000000);
 	const struct fixed31_32 c3 =
-		dal_fixed31_32_from_fraction(186875, 10000);
+		dc_fixpt_from_fraction(186875, 10000);
 
 	struct fixed31_32 l_pow_m1;
 	struct fixed31_32 base;
 
-	if (dal_fixed31_32_lt(in_x, dal_fixed31_32_zero))
-		in_x = dal_fixed31_32_zero;
+	if (dc_fixpt_lt(in_x, dc_fixpt_zero))
+		in_x = dc_fixpt_zero;
 
-	l_pow_m1 = dal_fixed31_32_pow(in_x, m1);
-	base = dal_fixed31_32_div(
-			dal_fixed31_32_add(c1,
-					(dal_fixed31_32_mul(c2, l_pow_m1))),
-			dal_fixed31_32_add(dal_fixed31_32_one,
-					(dal_fixed31_32_mul(c3, l_pow_m1))));
-	*out_y = dal_fixed31_32_pow(base, m2);
+	l_pow_m1 = dc_fixpt_pow(in_x, m1);
+	base = dc_fixpt_div(
+			dc_fixpt_add(c1,
+					(dc_fixpt_mul(c2, l_pow_m1))),
+			dc_fixpt_add(dc_fixpt_one,
+					(dc_fixpt_mul(c3, l_pow_m1))));
+	*out_y = dc_fixpt_pow(base, m2);
 }
 
 static void compute_de_pq(struct fixed31_32 in_x, struct fixed31_32 *out_y)
 {
 	/* consts for dePQ gamma formula. */
 	const struct fixed31_32 m1 =
-		dal_fixed31_32_from_fraction(159301758, 1000000000);
+		dc_fixpt_from_fraction(159301758, 1000000000);
 	const struct fixed31_32 m2 =
-		dal_fixed31_32_from_fraction(7884375, 100000);
+		dc_fixpt_from_fraction(7884375, 100000);
 	const struct fixed31_32 c1 =
-		dal_fixed31_32_from_fraction(8359375, 10000000);
+		dc_fixpt_from_fraction(8359375, 10000000);
 	const struct fixed31_32 c2 =
-		dal_fixed31_32_from_fraction(188515625, 10000000);
+		dc_fixpt_from_fraction(188515625, 10000000);
 	const struct fixed31_32 c3 =
-		dal_fixed31_32_from_fraction(186875, 10000);
+		dc_fixpt_from_fraction(186875, 10000);
 
 	struct fixed31_32 l_pow_m1;
 	struct fixed31_32 base, div;
 
 
-	if (dal_fixed31_32_lt(in_x, dal_fixed31_32_zero))
-		in_x = dal_fixed31_32_zero;
+	if (dc_fixpt_lt(in_x, dc_fixpt_zero))
+		in_x = dc_fixpt_zero;
 
-	l_pow_m1 = dal_fixed31_32_pow(in_x,
-			dal_fixed31_32_div(dal_fixed31_32_one, m2));
-	base = dal_fixed31_32_sub(l_pow_m1, c1);
+	l_pow_m1 = dc_fixpt_pow(in_x,
+			dc_fixpt_div(dc_fixpt_one, m2));
+	base = dc_fixpt_sub(l_pow_m1, c1);
 
-	if (dal_fixed31_32_lt(base, dal_fixed31_32_zero))
-		base = dal_fixed31_32_zero;
+	if (dc_fixpt_lt(base, dc_fixpt_zero))
+		base = dc_fixpt_zero;
 
-	div = dal_fixed31_32_sub(c2, dal_fixed31_32_mul(c3, l_pow_m1));
+	div = dc_fixpt_sub(c2, dc_fixpt_mul(c3, l_pow_m1));
 
-	*out_y = dal_fixed31_32_pow(dal_fixed31_32_div(base, div),
-			dal_fixed31_32_div(dal_fixed31_32_one, m1));
+	*out_y = dc_fixpt_pow(dc_fixpt_div(base, div),
+			dc_fixpt_div(dc_fixpt_one, m1));
 
 }
+
+/*de gamma, none linear to linear*/
+static void compute_hlg_oetf(struct fixed31_32 in_x, bool is_light0_12, struct fixed31_32 *out_y)
+{
+	struct fixed31_32 a;
+	struct fixed31_32 b;
+	struct fixed31_32 c;
+	struct fixed31_32 threshold;
+	struct fixed31_32 reference_white_level;
+
+	a = dc_fixpt_from_fraction(17883277, 100000000);
+	if (is_light0_12) {
+		/*light 0-12*/
+		b = dc_fixpt_from_fraction(28466892, 100000000);
+		c = dc_fixpt_from_fraction(55991073, 100000000);
+		threshold = dc_fixpt_one;
+		reference_white_level = dc_fixpt_half;
+	} else {
+		/*light 0-1*/
+		b = dc_fixpt_from_fraction(2372241, 100000000);
+		c = dc_fixpt_add(dc_fixpt_one, dc_fixpt_from_fraction(429347, 100000000));
+		threshold = dc_fixpt_from_fraction(1, 12);
+		reference_white_level = dc_fixpt_pow(dc_fixpt_from_fraction(3, 1), dc_fixpt_half);
+	}
+	if (dc_fixpt_lt(threshold, in_x))
+		*out_y = dc_fixpt_add(c, dc_fixpt_mul(a, dc_fixpt_log(dc_fixpt_sub(in_x, b))));
+	else
+		*out_y = dc_fixpt_mul(dc_fixpt_pow(in_x, dc_fixpt_half), reference_white_level);
+}
+
+/*re gamma, linear to none linear*/
+static void compute_hlg_eotf(struct fixed31_32 in_x, bool is_light0_12, struct fixed31_32 *out_y)
+{
+	struct fixed31_32 a;
+	struct fixed31_32 b;
+	struct fixed31_32 c;
+	struct fixed31_32 reference_white_level;
+
+	a = dc_fixpt_from_fraction(17883277, 100000000);
+	if (is_light0_12) {
+		/*light 0-12*/
+		b = dc_fixpt_from_fraction(28466892, 100000000);
+		c = dc_fixpt_from_fraction(55991073, 100000000);
+		reference_white_level = dc_fixpt_from_fraction(4, 1);
+	} else {
+		/*light 0-1*/
+		b = dc_fixpt_from_fraction(2372241, 100000000);
+		c = dc_fixpt_add(dc_fixpt_one, dc_fixpt_from_fraction(429347, 100000000));
+		reference_white_level = dc_fixpt_from_fraction(1, 3);
+	}
+	if (dc_fixpt_lt(dc_fixpt_half, in_x))
+		*out_y = dc_fixpt_add(dc_fixpt_exp(dc_fixpt_div(dc_fixpt_sub(in_x, c), a)), b);
+	else
+		*out_y = dc_fixpt_mul(dc_fixpt_pow(in_x, dc_fixpt_from_fraction(2, 1)), reference_white_level);
+}
+
+
 /* one-time pre-compute PQ values - only for sdr_white_level 80 */
 void precompute_pq(void)
 {
@@ -138,14 +195,14 @@ void precompute_pq(void)
 	struct fixed31_32 x;
 	const struct hw_x_point *coord_x = coordinates_x + 32;
 	struct fixed31_32 scaling_factor =
-			dal_fixed31_32_from_fraction(80, 10000);
+			dc_fixpt_from_fraction(80, 10000);
 
 	/* pow function has problems with arguments too small */
 	for (i = 0; i < 32; i++)
-		pq_table[i] = dal_fixed31_32_zero;
+		pq_table[i] = dc_fixpt_zero;
 
 	for (i = 32; i <= MAX_HW_POINTS; i++) {
-		x = dal_fixed31_32_mul(coord_x->x, scaling_factor);
+		x = dc_fixpt_mul(coord_x->x, scaling_factor);
 		compute_pq(x, &pq_table[i]);
 		++coord_x;
 	}
@@ -158,7 +215,7 @@ void precompute_de_pq(void)
 	struct fixed31_32  y;
 	uint32_t begin_index, end_index;
 
-	struct fixed31_32 scaling_factor = dal_fixed31_32_from_int(125);
+	struct fixed31_32 scaling_factor = dc_fixpt_from_int(125);
 
 	/* X points is 2^-25 to 2^7
 	 * De-gamma X is 2^-12 to 2^0 – we are skipping first -12-(-25) = 13 regions
@@ -167,11 +224,11 @@ void precompute_de_pq(void)
 	end_index = begin_index + 12 * NUM_PTS_IN_REGION;
 
 	for (i = 0; i <= begin_index; i++)
-		de_pq_table[i] = dal_fixed31_32_zero;
+		de_pq_table[i] = dc_fixpt_zero;
 
 	for (; i <= end_index; i++) {
 		compute_de_pq(coordinates_x[i].x, &y);
-		de_pq_table[i] = dal_fixed31_32_mul(y, scaling_factor);
+		de_pq_table[i] = dc_fixpt_mul(y, scaling_factor);
 	}
 
 	for (; i <= MAX_HW_POINTS; i++)
@@ -185,25 +242,25 @@ struct dividers {
 
 static void build_coefficients(struct gamma_coefficients *coefficients, bool is_2_4)
 {
-		static const int32_t numerator01[] = { 31308, 180000};
-		static const int32_t numerator02[] = { 12920, 4500};
-		static const int32_t numerator03[] = { 55, 99};
-		static const int32_t numerator04[] = { 55, 99};
-		static const int32_t numerator05[] = { 2400, 2200};
+	static const int32_t numerator01[] = { 31308, 180000};
+	static const int32_t numerator02[] = { 12920, 4500};
+	static const int32_t numerator03[] = { 55, 99};
+	static const int32_t numerator04[] = { 55, 99};
+	static const int32_t numerator05[] = { 2400, 2200};
 
-		uint32_t i = 0;
-		uint32_t index = is_2_4 == true ? 0:1;
+	uint32_t i = 0;
+	uint32_t index = is_2_4 == true ? 0:1;
 
 	do {
-		coefficients->a0[i] = dal_fixed31_32_from_fraction(
+		coefficients->a0[i] = dc_fixpt_from_fraction(
 			numerator01[index], 10000000);
-		coefficients->a1[i] = dal_fixed31_32_from_fraction(
+		coefficients->a1[i] = dc_fixpt_from_fraction(
 			numerator02[index], 1000);
-		coefficients->a2[i] = dal_fixed31_32_from_fraction(
+		coefficients->a2[i] = dc_fixpt_from_fraction(
 			numerator03[index], 1000);
-		coefficients->a3[i] = dal_fixed31_32_from_fraction(
+		coefficients->a3[i] = dc_fixpt_from_fraction(
 			numerator04[index], 1000);
-		coefficients->user_gamma[i] = dal_fixed31_32_from_fraction(
+		coefficients->user_gamma[i] = dc_fixpt_from_fraction(
 			numerator05[index], 1000);
 
 		++i;
@@ -218,33 +275,33 @@ static struct fixed31_32 translate_from_linear_space(
 	struct fixed31_32 a3,
 	struct fixed31_32 gamma)
 {
-	const struct fixed31_32 one = dal_fixed31_32_from_int(1);
+	const struct fixed31_32 one = dc_fixpt_from_int(1);
 
-	if (dal_fixed31_32_lt(one, arg))
+	if (dc_fixpt_lt(one, arg))
 		return one;
 
-	if (dal_fixed31_32_le(arg, dal_fixed31_32_neg(a0)))
-		return dal_fixed31_32_sub(
+	if (dc_fixpt_le(arg, dc_fixpt_neg(a0)))
+		return dc_fixpt_sub(
 			a2,
-			dal_fixed31_32_mul(
-				dal_fixed31_32_add(
+			dc_fixpt_mul(
+				dc_fixpt_add(
 					one,
 					a3),
-				dal_fixed31_32_pow(
-					dal_fixed31_32_neg(arg),
-					dal_fixed31_32_recip(gamma))));
-	else if (dal_fixed31_32_le(a0, arg))
-		return dal_fixed31_32_sub(
-			dal_fixed31_32_mul(
-				dal_fixed31_32_add(
+				dc_fixpt_pow(
+					dc_fixpt_neg(arg),
+					dc_fixpt_recip(gamma))));
+	else if (dc_fixpt_le(a0, arg))
+		return dc_fixpt_sub(
+			dc_fixpt_mul(
+				dc_fixpt_add(
 					one,
 					a3),
-				dal_fixed31_32_pow(
+				dc_fixpt_pow(
 					arg,
-					dal_fixed31_32_recip(gamma))),
+					dc_fixpt_recip(gamma))),
 			a2);
 	else
-		return dal_fixed31_32_mul(
+		return dc_fixpt_mul(
 			arg,
 			a1);
 }
@@ -259,25 +316,25 @@ static struct fixed31_32 translate_to_linear_space(
 {
 	struct fixed31_32 linear;
 
-	a0 = dal_fixed31_32_mul(a0, a1);
-	if (dal_fixed31_32_le(arg, dal_fixed31_32_neg(a0)))
+	a0 = dc_fixpt_mul(a0, a1);
+	if (dc_fixpt_le(arg, dc_fixpt_neg(a0)))
 
-		linear = dal_fixed31_32_neg(
-				 dal_fixed31_32_pow(
-				 dal_fixed31_32_div(
-				 dal_fixed31_32_sub(a2, arg),
-				 dal_fixed31_32_add(
-				 dal_fixed31_32_one, a3)), gamma));
+		linear = dc_fixpt_neg(
+				 dc_fixpt_pow(
+				 dc_fixpt_div(
+				 dc_fixpt_sub(a2, arg),
+				 dc_fixpt_add(
+				 dc_fixpt_one, a3)), gamma));
 
-	else if (dal_fixed31_32_le(dal_fixed31_32_neg(a0), arg) &&
-			 dal_fixed31_32_le(arg, a0))
-		linear = dal_fixed31_32_div(arg, a1);
+	else if (dc_fixpt_le(dc_fixpt_neg(a0), arg) &&
+			 dc_fixpt_le(arg, a0))
+		linear = dc_fixpt_div(arg, a1);
 	else
-		linear =  dal_fixed31_32_pow(
-					dal_fixed31_32_div(
-					dal_fixed31_32_add(a2, arg),
-					dal_fixed31_32_add(
-					dal_fixed31_32_one, a3)), gamma);
+		linear =  dc_fixpt_pow(
+					dc_fixpt_div(
+					dc_fixpt_add(a2, arg),
+					dc_fixpt_add(
+					dc_fixpt_one, a3)), gamma);
 
 	return linear;
 }
@@ -352,8 +409,8 @@ static bool find_software_points(
 				right = axis_x[max_number - 1].b;
 		}
 
-		if (dal_fixed31_32_le(left, hw_point) &&
-			dal_fixed31_32_le(hw_point, right)) {
+		if (dc_fixpt_le(left, hw_point) &&
+			dc_fixpt_le(hw_point, right)) {
 			*index_to_start = i;
 			*index_left = i;
 
@@ -366,7 +423,7 @@ static bool find_software_points(
 
 			return true;
 		} else if ((i == *index_to_start) &&
-			dal_fixed31_32_le(hw_point, left)) {
+			dc_fixpt_le(hw_point, left)) {
 			*index_to_start = i;
 			*index_left = i;
 			*index_right = i;
@@ -375,7 +432,7 @@ static bool find_software_points(
 
 			return true;
 		} else if ((i == max_number - 1) &&
-			dal_fixed31_32_le(right, hw_point)) {
+			dc_fixpt_le(right, hw_point)) {
 			*index_to_start = i;
 			*index_left = i;
 			*index_right = i;
@@ -457,17 +514,17 @@ static bool build_custom_gamma_mapping_coefficients_worker(
 		}
 
 		if (hw_pos == HW_POINT_POSITION_MIDDLE)
-			point->coeff = dal_fixed31_32_div(
-				dal_fixed31_32_sub(
+			point->coeff = dc_fixpt_div(
+				dc_fixpt_sub(
 					coord_x,
 					left_pos),
-				dal_fixed31_32_sub(
+				dc_fixpt_sub(
 					right_pos,
 					left_pos));
 		else if (hw_pos == HW_POINT_POSITION_LEFT)
-			point->coeff = dal_fixed31_32_zero;
+			point->coeff = dc_fixpt_zero;
 		else if (hw_pos == HW_POINT_POSITION_RIGHT)
-			point->coeff = dal_fixed31_32_from_int(2);
+			point->coeff = dc_fixpt_from_int(2);
 		else {
 			BREAK_TO_DEBUGGER();
 			return false;
@@ -502,45 +559,45 @@ static struct fixed31_32 calculate_mapped_value(
 
 	if ((point->left_index < 0) || (point->left_index > max_index)) {
 		BREAK_TO_DEBUGGER();
-		return dal_fixed31_32_zero;
+		return dc_fixpt_zero;
 	}
 
 	if ((point->right_index < 0) || (point->right_index > max_index)) {
 		BREAK_TO_DEBUGGER();
-		return dal_fixed31_32_zero;
+		return dc_fixpt_zero;
 	}
 
 	if (point->pos == HW_POINT_POSITION_MIDDLE)
 		if (channel == CHANNEL_NAME_RED)
-			result = dal_fixed31_32_add(
-				dal_fixed31_32_mul(
+			result = dc_fixpt_add(
+				dc_fixpt_mul(
 					point->coeff,
-					dal_fixed31_32_sub(
+					dc_fixpt_sub(
 						rgb[point->right_index].r,
 						rgb[point->left_index].r)),
 				rgb[point->left_index].r);
 		else if (channel == CHANNEL_NAME_GREEN)
-			result = dal_fixed31_32_add(
-				dal_fixed31_32_mul(
+			result = dc_fixpt_add(
+				dc_fixpt_mul(
 					point->coeff,
-					dal_fixed31_32_sub(
+					dc_fixpt_sub(
 						rgb[point->right_index].g,
 						rgb[point->left_index].g)),
 				rgb[point->left_index].g);
 		else
-			result = dal_fixed31_32_add(
-				dal_fixed31_32_mul(
+			result = dc_fixpt_add(
+				dc_fixpt_mul(
 					point->coeff,
-					dal_fixed31_32_sub(
+					dc_fixpt_sub(
 						rgb[point->right_index].b,
 						rgb[point->left_index].b)),
 				rgb[point->left_index].b);
 	else if (point->pos == HW_POINT_POSITION_LEFT) {
 		BREAK_TO_DEBUGGER();
-		result = dal_fixed31_32_zero;
+		result = dc_fixpt_zero;
 	} else {
 		BREAK_TO_DEBUGGER();
-		result = dal_fixed31_32_one;
+		result = dc_fixpt_one;
 	}
 
 	return result;
@@ -558,7 +615,7 @@ static void build_pq(struct pwl_float_data_ex *rgb_regamma,
 	struct fixed31_32 x;
 	struct fixed31_32 output;
 	struct fixed31_32 scaling_factor =
-			dal_fixed31_32_from_fraction(sdr_white_level, 10000);
+			dc_fixpt_from_fraction(sdr_white_level, 10000);
 
 	if (!pq_initialized && sdr_white_level == 80) {
 		precompute_pq();
@@ -579,15 +636,15 @@ static void build_pq(struct pwl_float_data_ex *rgb_regamma,
 		if (sdr_white_level == 80) {
 			output = pq_table[i];
 		} else {
-			x = dal_fixed31_32_mul(coord_x->x, scaling_factor);
+			x = dc_fixpt_mul(coord_x->x, scaling_factor);
 			compute_pq(x, &output);
 		}
 
 		/* should really not happen? */
-		if (dal_fixed31_32_lt(output, dal_fixed31_32_zero))
-			output = dal_fixed31_32_zero;
-		else if (dal_fixed31_32_lt(dal_fixed31_32_one, output))
-			output = dal_fixed31_32_one;
+		if (dc_fixpt_lt(output, dc_fixpt_zero))
+			output = dc_fixpt_zero;
+		else if (dc_fixpt_lt(dc_fixpt_one, output))
+			output = dc_fixpt_one;
 
 		rgb->r = output;
 		rgb->g = output;
@@ -605,7 +662,7 @@ static void build_de_pq(struct pwl_float_data_ex *de_pq,
 	uint32_t i;
 	struct fixed31_32 output;
 
-	struct fixed31_32 scaling_factor = dal_fixed31_32_from_int(125);
+	struct fixed31_32 scaling_factor = dc_fixpt_from_int(125);
 
 	if (!de_pq_initialized) {
 		precompute_de_pq();
@@ -616,9 +673,9 @@ static void build_de_pq(struct pwl_float_data_ex *de_pq,
 	for (i = 0; i <= hw_points_num; i++) {
 		output = de_pq_table[i];
 		/* should really not happen? */
-		if (dal_fixed31_32_lt(output, dal_fixed31_32_zero))
-			output = dal_fixed31_32_zero;
-		else if (dal_fixed31_32_lt(scaling_factor, output))
+		if (dc_fixpt_lt(output, dc_fixpt_zero))
+			output = dc_fixpt_zero;
+		else if (dc_fixpt_lt(scaling_factor, output))
 			output = scaling_factor;
 		de_pq[i].r = output;
 		de_pq[i].g = output;
@@ -670,9 +727,9 @@ static void build_degamma(struct pwl_float_data_ex *curve,
 	end_index = begin_index + 12 * NUM_PTS_IN_REGION;
 
 	while (i != begin_index) {
-		curve[i].r = dal_fixed31_32_zero;
-		curve[i].g = dal_fixed31_32_zero;
-		curve[i].b = dal_fixed31_32_zero;
+		curve[i].r = dc_fixpt_zero;
+		curve[i].g = dc_fixpt_zero;
+		curve[i].b = dc_fixpt_zero;
 		i++;
 	}
 
@@ -684,19 +741,61 @@ static void build_degamma(struct pwl_float_data_ex *curve,
 		i++;
 	}
 	while (i != hw_points_num + 1) {
-		curve[i].r = dal_fixed31_32_one;
-		curve[i].g = dal_fixed31_32_one;
-		curve[i].b = dal_fixed31_32_one;
+		curve[i].r = dc_fixpt_one;
+		curve[i].g = dc_fixpt_one;
+		curve[i].b = dc_fixpt_one;
 		i++;
 	}
 }
 
-static bool scale_gamma(struct pwl_float_data *pwl_rgb,
+static void build_hlg_degamma(struct pwl_float_data_ex *degamma,
+		uint32_t hw_points_num,
+		const struct hw_x_point *coordinate_x, bool is_light0_12)
+{
+	uint32_t i;
+
+	struct pwl_float_data_ex *rgb = degamma;
+	const struct hw_x_point *coord_x = coordinate_x;
+
+	i = 0;
+
+	while (i != hw_points_num + 1) {
+		compute_hlg_oetf(coord_x->x, is_light0_12, &rgb->r);
+		rgb->g = rgb->r;
+		rgb->b = rgb->r;
+		++coord_x;
+		++rgb;
+		++i;
+	}
+}
+
+static void build_hlg_regamma(struct pwl_float_data_ex *regamma,
+		uint32_t hw_points_num,
+		const struct hw_x_point *coordinate_x, bool is_light0_12)
+{
+	uint32_t i;
+
+	struct pwl_float_data_ex *rgb = regamma;
+	const struct hw_x_point *coord_x = coordinate_x;
+
+	i = 0;
+
+	while (i != hw_points_num + 1) {
+		compute_hlg_eotf(coord_x->x, is_light0_12, &rgb->r);
+		rgb->g = rgb->r;
+		rgb->b = rgb->r;
+		++coord_x;
+		++rgb;
+		++i;
+	}
+}
+
+static void scale_gamma(struct pwl_float_data *pwl_rgb,
 		const struct dc_gamma *ramp,
 		struct dividers dividers)
 {
-	const struct fixed31_32 max_driver = dal_fixed31_32_from_int(0xFFFF);
-	const struct fixed31_32 max_os = dal_fixed31_32_from_int(0xFF00);
+	const struct fixed31_32 max_driver = dc_fixpt_from_int(0xFFFF);
+	const struct fixed31_32 max_os = dc_fixpt_from_int(0xFF00);
 	struct fixed31_32 scaler = max_os;
 	uint32_t i;
 	struct pwl_float_data *rgb = pwl_rgb;
@@ -705,9 +804,9 @@ static bool scale_gamma(struct pwl_float_data *pwl_rgb,
 	i = 0;
 
 	do {
-		if (dal_fixed31_32_lt(max_os, ramp->entries.red[i]) ||
-			dal_fixed31_32_lt(max_os, ramp->entries.green[i]) ||
-			dal_fixed31_32_lt(max_os, ramp->entries.blue[i])) {
+		if (dc_fixpt_lt(max_os, ramp->entries.red[i]) ||
+			dc_fixpt_lt(max_os, ramp->entries.green[i]) ||
+			dc_fixpt_lt(max_os, ramp->entries.blue[i])) {
 			scaler = max_driver;
 			break;
 		}
@@ -717,109 +816,170 @@ static bool scale_gamma(struct pwl_float_data *pwl_rgb,
 	i = 0;
 
 	do {
-		rgb->r = dal_fixed31_32_div(
+		rgb->r = dc_fixpt_div(
 			ramp->entries.red[i], scaler);
-		rgb->g = dal_fixed31_32_div(
+		rgb->g = dc_fixpt_div(
 			ramp->entries.green[i], scaler);
-		rgb->b = dal_fixed31_32_div(
+		rgb->b = dc_fixpt_div(
 			ramp->entries.blue[i], scaler);
 
 		++rgb;
 		++i;
 	} while (i != ramp->num_entries);
 
-	rgb->r = dal_fixed31_32_mul(rgb_last->r,
+	rgb->r = dc_fixpt_mul(rgb_last->r,
 			dividers.divider1);
-	rgb->g = dal_fixed31_32_mul(rgb_last->g,
+	rgb->g = dc_fixpt_mul(rgb_last->g,
 			dividers.divider1);
-	rgb->b = dal_fixed31_32_mul(rgb_last->b,
+	rgb->b = dc_fixpt_mul(rgb_last->b,
 			dividers.divider1);
 
 	++rgb;
 
-	rgb->r = dal_fixed31_32_mul(rgb_last->r,
+	rgb->r = dc_fixpt_mul(rgb_last->r,
 			dividers.divider2);
-	rgb->g = dal_fixed31_32_mul(rgb_last->g,
+	rgb->g = dc_fixpt_mul(rgb_last->g,
 			dividers.divider2);
-	rgb->b = dal_fixed31_32_mul(rgb_last->b,
+	rgb->b = dc_fixpt_mul(rgb_last->b,
 			dividers.divider2);
 
 	++rgb;
 
-	rgb->r = dal_fixed31_32_mul(rgb_last->r,
+	rgb->r = dc_fixpt_mul(rgb_last->r,
 			dividers.divider3);
-	rgb->g = dal_fixed31_32_mul(rgb_last->g,
+	rgb->g = dc_fixpt_mul(rgb_last->g,
 			dividers.divider3);
-	rgb->b = dal_fixed31_32_mul(rgb_last->b,
+	rgb->b = dc_fixpt_mul(rgb_last->b,
 			dividers.divider3);
-
-	return true;
 }
 
-static bool scale_gamma_dx(struct pwl_float_data *pwl_rgb,
+static void scale_gamma_dx(struct pwl_float_data *pwl_rgb,
 		const struct dc_gamma *ramp,
 		struct dividers dividers)
 {
 	uint32_t i;
-	struct fixed31_32 min = dal_fixed31_32_zero;
-	struct fixed31_32 max = dal_fixed31_32_one;
+	struct fixed31_32 min = dc_fixpt_zero;
+	struct fixed31_32 max = dc_fixpt_one;
 
-	struct fixed31_32 delta = dal_fixed31_32_zero;
-	struct fixed31_32 offset = dal_fixed31_32_zero;
+	struct fixed31_32 delta = dc_fixpt_zero;
+	struct fixed31_32 offset = dc_fixpt_zero;
 
 	for (i = 0 ; i < ramp->num_entries; i++) {
-		if (dal_fixed31_32_lt(ramp->entries.red[i], min))
+		if (dc_fixpt_lt(ramp->entries.red[i], min))
 			min = ramp->entries.red[i];
 
-		if (dal_fixed31_32_lt(ramp->entries.green[i], min))
+		if (dc_fixpt_lt(ramp->entries.green[i], min))
 			min = ramp->entries.green[i];
 
-		if (dal_fixed31_32_lt(ramp->entries.blue[i], min))
+		if (dc_fixpt_lt(ramp->entries.blue[i], min))
 			min = ramp->entries.blue[i];
 
-		if (dal_fixed31_32_lt(max, ramp->entries.red[i]))
+		if (dc_fixpt_lt(max, ramp->entries.red[i]))
 			max = ramp->entries.red[i];
 
-		if (dal_fixed31_32_lt(max, ramp->entries.green[i]))
+		if (dc_fixpt_lt(max, ramp->entries.green[i]))
 			max = ramp->entries.green[i];
 
-		if (dal_fixed31_32_lt(max, ramp->entries.blue[i]))
+		if (dc_fixpt_lt(max, ramp->entries.blue[i]))
 			max = ramp->entries.blue[i];
 	}
 
-	if (dal_fixed31_32_lt(min, dal_fixed31_32_zero))
-		delta = dal_fixed31_32_neg(min);
+	if (dc_fixpt_lt(min, dc_fixpt_zero))
+		delta = dc_fixpt_neg(min);
 
-	offset = dal_fixed31_32_add(min, max);
+	offset = dc_fixpt_add(min, max);
 
 	for (i = 0 ; i < ramp->num_entries; i++) {
-		pwl_rgb[i].r = dal_fixed31_32_div(
-			dal_fixed31_32_add(
+		pwl_rgb[i].r = dc_fixpt_div(
+			dc_fixpt_add(
 				ramp->entries.red[i], delta), offset);
-		pwl_rgb[i].g = dal_fixed31_32_div(
-			dal_fixed31_32_add(
+		pwl_rgb[i].g = dc_fixpt_div(
+			dc_fixpt_add(
 				ramp->entries.green[i], delta), offset);
-		pwl_rgb[i].b = dal_fixed31_32_div(
-			dal_fixed31_32_add(
+		pwl_rgb[i].b = dc_fixpt_div(
+			dc_fixpt_add(
 				ramp->entries.blue[i], delta), offset);
 
 	}
 
-	pwl_rgb[i].r =  dal_fixed31_32_sub(dal_fixed31_32_mul_int(
+	pwl_rgb[i].r =  dc_fixpt_sub(dc_fixpt_mul_int(
 				pwl_rgb[i-1].r, 2), pwl_rgb[i-2].r);
-	pwl_rgb[i].g =  dal_fixed31_32_sub(dal_fixed31_32_mul_int(
+	pwl_rgb[i].g =  dc_fixpt_sub(dc_fixpt_mul_int(
 				pwl_rgb[i-1].g, 2), pwl_rgb[i-2].g);
-	pwl_rgb[i].b =  dal_fixed31_32_sub(dal_fixed31_32_mul_int(
+	pwl_rgb[i].b =  dc_fixpt_sub(dc_fixpt_mul_int(
 				pwl_rgb[i-1].b, 2), pwl_rgb[i-2].b);
 	++i;
-	pwl_rgb[i].r =  dal_fixed31_32_sub(dal_fixed31_32_mul_int(
+	pwl_rgb[i].r =  dc_fixpt_sub(dc_fixpt_mul_int(
 				pwl_rgb[i-1].r, 2), pwl_rgb[i-2].r);
-	pwl_rgb[i].g =  dal_fixed31_32_sub(dal_fixed31_32_mul_int(
+	pwl_rgb[i].g =  dc_fixpt_sub(dc_fixpt_mul_int(
 				pwl_rgb[i-1].g, 2), pwl_rgb[i-2].g);
-	pwl_rgb[i].b =  dal_fixed31_32_sub(dal_fixed31_32_mul_int(
+	pwl_rgb[i].b =  dc_fixpt_sub(dc_fixpt_mul_int(
 				pwl_rgb[i-1].b, 2), pwl_rgb[i-2].b);
+}
 
-	return true;
+/* todo: all these scale_gamma functions are inherently the same but
+ *  take different structures as params or different format for ramp
+ *  values. We could probably implement it in a more generic fashion
+ */
+static void scale_user_regamma_ramp(struct pwl_float_data *pwl_rgb,
+		const struct regamma_ramp *ramp,
+		struct dividers dividers)
+{
+	unsigned short max_driver = 0xFFFF;
+	unsigned short max_os = 0xFF00;
+	unsigned short scaler = max_os;
+	uint32_t i;
+	struct pwl_float_data *rgb = pwl_rgb;
+	struct pwl_float_data *rgb_last = rgb + GAMMA_RGB_256_ENTRIES - 1;
+
+	i = 0;
+	do {
+		if (ramp->gamma[i] > max_os ||
+				ramp->gamma[i + 256] > max_os ||
+				ramp->gamma[i + 512] > max_os) {
+			scaler = max_driver;
+			break;
+		}
+		i++;
+	} while (i != GAMMA_RGB_256_ENTRIES);
+
+	i = 0;
+	do {
+		rgb->r = dc_fixpt_from_fraction(
+				ramp->gamma[i], scaler);
+		rgb->g = dc_fixpt_from_fraction(
+				ramp->gamma[i + 256], scaler);
+		rgb->b = dc_fixpt_from_fraction(
+				ramp->gamma[i + 512], scaler);
+
+		++rgb;
+		++i;
+	} while (i != GAMMA_RGB_256_ENTRIES);
+
+	rgb->r = dc_fixpt_mul(rgb_last->r,
+			dividers.divider1);
+	rgb->g = dc_fixpt_mul(rgb_last->g,
+			dividers.divider1);
+	rgb->b = dc_fixpt_mul(rgb_last->b,
+			dividers.divider1);
+
+	++rgb;
+
+	rgb->r = dc_fixpt_mul(rgb_last->r,
+			dividers.divider2);
+	rgb->g = dc_fixpt_mul(rgb_last->g,
+			dividers.divider2);
+	rgb->b = dc_fixpt_mul(rgb_last->b,
+			dividers.divider2);
+
+	++rgb;
+
+	rgb->r = dc_fixpt_mul(rgb_last->r,
+			dividers.divider3);
+	rgb->g = dc_fixpt_mul(rgb_last->g,
+			dividers.divider3);
+	rgb->b = dc_fixpt_mul(rgb_last->b,
+			dividers.divider3);
 }
 
 /*
@@ -852,7 +1012,7 @@ static void apply_lut_1d(
 	struct fixed31_32 lut2;
 	const int max_lut_index = 4095;
 	const struct fixed31_32 max_lut_index_f =
-			dal_fixed31_32_from_int_nonconst(max_lut_index);
+			dc_fixpt_from_int(max_lut_index);
 	int32_t index = 0, index_next = 0;
 	struct fixed31_32 index_f;
 	struct fixed31_32 delta_lut;
@@ -870,10 +1030,10 @@ static void apply_lut_1d(
 			else
 				regamma_y = &tf_pts->blue[i];
 
-			norm_y = dal_fixed31_32_mul(max_lut_index_f,
+			norm_y = dc_fixpt_mul(max_lut_index_f,
 						   *regamma_y);
-			index = dal_fixed31_32_floor(norm_y);
-			index_f = dal_fixed31_32_from_int_nonconst(index);
+			index = dc_fixpt_floor(norm_y);
+			index_f = dc_fixpt_from_int(index);
 
 			if (index < 0 || index > max_lut_index)
 				continue;
@@ -892,11 +1052,11 @@ static void apply_lut_1d(
 			}
 
 			// we have everything now, so interpolate
-			delta_lut = dal_fixed31_32_sub(lut2, lut1);
-			delta_index = dal_fixed31_32_sub(norm_y, index_f);
+			delta_lut = dc_fixpt_sub(lut2, lut1);
+			delta_index = dc_fixpt_sub(norm_y, index_f);
 
-			*regamma_y = dal_fixed31_32_add(lut1,
-				dal_fixed31_32_mul(delta_index, delta_lut));
+			*regamma_y = dc_fixpt_add(lut1,
+				dc_fixpt_mul(delta_index, delta_lut));
 		}
 	}
 }
@@ -912,7 +1072,7 @@ static void build_evenly_distributed_points(
 	uint32_t i = 0;
 
 	do {
-		struct fixed31_32 value = dal_fixed31_32_from_fraction(i,
+		struct fixed31_32 value = dc_fixpt_from_fraction(i,
 			numberof_points - 1);
 
 		p->r = value;
@@ -923,21 +1083,21 @@ static void build_evenly_distributed_points(
 		++i;
 	} while (i != numberof_points);
 
-	p->r = dal_fixed31_32_div(p_last->r, dividers.divider1);
-	p->g = dal_fixed31_32_div(p_last->g, dividers.divider1);
-	p->b = dal_fixed31_32_div(p_last->b, dividers.divider1);
+	p->r = dc_fixpt_div(p_last->r, dividers.divider1);
+	p->g = dc_fixpt_div(p_last->g, dividers.divider1);
+	p->b = dc_fixpt_div(p_last->b, dividers.divider1);
 
 	++p;
 
-	p->r = dal_fixed31_32_div(p_last->r, dividers.divider2);
-	p->g = dal_fixed31_32_div(p_last->g, dividers.divider2);
-	p->b = dal_fixed31_32_div(p_last->b, dividers.divider2);
+	p->r = dc_fixpt_div(p_last->r, dividers.divider2);
+	p->g = dc_fixpt_div(p_last->g, dividers.divider2);
+	p->b = dc_fixpt_div(p_last->b, dividers.divider2);
 
 	++p;
 
-	p->r = dal_fixed31_32_div(p_last->r, dividers.divider3);
-	p->g = dal_fixed31_32_div(p_last->g, dividers.divider3);
-	p->b = dal_fixed31_32_div(p_last->b, dividers.divider3);
+	p->r = dc_fixpt_div(p_last->r, dividers.divider3);
+	p->g = dc_fixpt_div(p_last->g, dividers.divider3);
+	p->b = dc_fixpt_div(p_last->b, dividers.divider3);
 }
 
 static inline void copy_rgb_regamma_to_coordinates_x(
@@ -949,7 +1109,7 @@ static inline void copy_rgb_regamma_to_coordinates_x(
 	uint32_t i = 0;
 	const struct pwl_float_data_ex *rgb_regamma = rgb_ex;
 
-	while (i <= hw_points_num) {
+	while (i <= hw_points_num + 1) {
 		coords->regamma_y_red = rgb_regamma->r;
 		coords->regamma_y_green = rgb_regamma->g;
 		coords->regamma_y_blue = rgb_regamma->b;
@@ -1002,6 +1162,102 @@ static bool calculate_interpolated_hardware_curve(
 	return true;
 }
 
+/* The "old" interpolation uses a complicated scheme to build an array of
+ * coefficients while also using an array of 0-255 normalized to 0-1
+ * Then there's another loop using both of the above + new scaled user ramp
+ * and we concatenate them. It also searches for points of interpolation and
+ * uses enums for positions.
+ *
+ * This function uses a different approach:
+ * user ramp is always applied on X with 0/255, 1/255, 2/255, ..., 255/255
+ * To find index for hwX , we notice the following:
+ * i/255 <= hwX < (i+1)/255  <=> i <= 255*hwX < i+1
+ * See apply_lut_1d which is the same principle, but on 4K entry 1D LUT
+ *
+ * Once the index is known, combined Y is simply:
+ * user_ramp(index) + (hwX-index/255)*(user_ramp(index+1) - user_ramp(index)
+ *
+ * We should switch to this method in all cases, it's simpler and faster
+ * ToDo one day - for now this only applies to ADL regamma to avoid regression
+ * for regular use cases (sRGB and PQ)
+ */
+static void interpolate_user_regamma(uint32_t hw_points_num,
+		struct pwl_float_data *rgb_user,
+		bool apply_degamma,
+		struct dc_transfer_func_distributed_points *tf_pts)
+{
+	uint32_t i;
+	uint32_t color = 0;
+	int32_t index;
+	int32_t index_next;
+	struct fixed31_32 *tf_point;
+	struct fixed31_32 hw_x;
+	struct fixed31_32 norm_factor =
+			dc_fixpt_from_int(255);
+	struct fixed31_32 norm_x;
+	struct fixed31_32 index_f;
+	struct fixed31_32 lut1;
+	struct fixed31_32 lut2;
+	struct fixed31_32 delta_lut;
+	struct fixed31_32 delta_index;
+
+	i = 0;
+	/* fixed_pt library has problems handling too small values */
+	while (i != 32) {
+		tf_pts->red[i] = dc_fixpt_zero;
+		tf_pts->green[i] = dc_fixpt_zero;
+		tf_pts->blue[i] = dc_fixpt_zero;
+		++i;
+	}
+	while (i <= hw_points_num + 1) {
+		for (color = 0; color < 3; color++) {
+			if (color == 0)
+				tf_point = &tf_pts->red[i];
+			else if (color == 1)
+				tf_point = &tf_pts->green[i];
+			else
+				tf_point = &tf_pts->blue[i];
+
+			if (apply_degamma) {
+				if (color == 0)
+					hw_x = coordinates_x[i].regamma_y_red;
+				else if (color == 1)
+					hw_x = coordinates_x[i].regamma_y_green;
+				else
+					hw_x = coordinates_x[i].regamma_y_blue;
+			} else
+				hw_x = coordinates_x[i].x;
+
+			norm_x = dc_fixpt_mul(norm_factor, hw_x);
+			index = dc_fixpt_floor(norm_x);
+			if (index < 0 || index > 255)
+				continue;
+
+			index_f = dc_fixpt_from_int(index);
+			index_next = (index == 255) ? index : index + 1;
+
+			if (color == 0) {
+				lut1 = rgb_user[index].r;
+				lut2 = rgb_user[index_next].r;
+			} else if (color == 1) {
+				lut1 = rgb_user[index].g;
+				lut2 = rgb_user[index_next].g;
+			} else {
+				lut1 = rgb_user[index].b;
+				lut2 = rgb_user[index_next].b;
+			}
+
+			// we have everything now, so interpolate
+			delta_lut = dc_fixpt_sub(lut2, lut1);
+			delta_index = dc_fixpt_sub(norm_x, index_f);
+
+			*tf_point = dc_fixpt_add(lut1,
+				dc_fixpt_mul(delta_index, delta_lut));
+		}
+		++i;
+	}
+}
+
 static void build_new_custom_resulted_curve(
 	uint32_t hw_points_num,
 	struct dc_transfer_func_distributed_points *tf_pts)
@@ -1011,20 +1267,43 @@ static void build_new_custom_resulted_curve(
 	i = 0;
 
 	while (i != hw_points_num + 1) {
-		tf_pts->red[i] = dal_fixed31_32_clamp(
-			tf_pts->red[i], dal_fixed31_32_zero,
-			dal_fixed31_32_one);
-		tf_pts->green[i] = dal_fixed31_32_clamp(
-			tf_pts->green[i], dal_fixed31_32_zero,
-			dal_fixed31_32_one);
-		tf_pts->blue[i] = dal_fixed31_32_clamp(
-			tf_pts->blue[i], dal_fixed31_32_zero,
-			dal_fixed31_32_one);
+		tf_pts->red[i] = dc_fixpt_clamp(
+			tf_pts->red[i], dc_fixpt_zero,
+			dc_fixpt_one);
+		tf_pts->green[i] = dc_fixpt_clamp(
+			tf_pts->green[i], dc_fixpt_zero,
+			dc_fixpt_one);
+		tf_pts->blue[i] = dc_fixpt_clamp(
+			tf_pts->blue[i], dc_fixpt_zero,
+			dc_fixpt_one);
 
 		++i;
 	}
 }
 
+static void apply_degamma_for_user_regamma(struct pwl_float_data_ex *rgb_regamma,
+		uint32_t hw_points_num)
+{
+	uint32_t i;
+
+	struct gamma_coefficients coeff;
+	struct pwl_float_data_ex *rgb = rgb_regamma;
+	const struct hw_x_point *coord_x = coordinates_x;
+
+	build_coefficients(&coeff, true);
+
+	i = 0;
+	while (i != hw_points_num + 1) {
+		rgb->r = translate_from_linear_space_ex(
+				coord_x->x, &coeff, 0);
+		rgb->g = rgb->r;
+		rgb->b = rgb->r;
+		++coord_x;
+		++rgb;
+		++i;
+	}
+}
+
 static bool map_regamma_hw_to_x_user(
 	const struct dc_gamma *ramp,
 	struct pixel_gamma_point *coeff128,
@@ -1062,6 +1341,7 @@ static bool map_regamma_hw_to_x_user(
 		}
 	}
 
+	/* this should be named differently, all it does is clamp to 0-1 */
 	build_new_custom_resulted_curve(hw_points_num, tf_pts);
 
 	return true;
@@ -1093,25 +1373,28 @@ bool mod_color_calculate_regamma_params(struct dc_transfer_func *output_tf,
 
 	output_tf->type = TF_TYPE_DISTRIBUTED_POINTS;
 
-	rgb_user = kzalloc(sizeof(*rgb_user) * (ramp->num_entries + _EXTRA_POINTS),
-			   GFP_KERNEL);
+	rgb_user = kvcalloc(ramp->num_entries + _EXTRA_POINTS,
+			    sizeof(*rgb_user),
+			    GFP_KERNEL);
 	if (!rgb_user)
 		goto rgb_user_alloc_fail;
-	rgb_regamma = kzalloc(sizeof(*rgb_regamma) * (MAX_HW_POINTS + _EXTRA_POINTS),
-			GFP_KERNEL);
+	rgb_regamma = kvcalloc(MAX_HW_POINTS + _EXTRA_POINTS,
+			       sizeof(*rgb_regamma),
+			       GFP_KERNEL);
 	if (!rgb_regamma)
 		goto rgb_regamma_alloc_fail;
-	axix_x = kzalloc(sizeof(*axix_x) * (ramp->num_entries + 3),
-			 GFP_KERNEL);
+	axix_x = kvcalloc(ramp->num_entries + 3, sizeof(*axix_x),
+			  GFP_KERNEL);
 	if (!axix_x)
 		goto axix_x_alloc_fail;
-	coeff = kzalloc(sizeof(*coeff) * (MAX_HW_POINTS + _EXTRA_POINTS), GFP_KERNEL);
+	coeff = kvcalloc(MAX_HW_POINTS + _EXTRA_POINTS, sizeof(*coeff),
+			 GFP_KERNEL);
 	if (!coeff)
 		goto coeff_alloc_fail;
 
-	dividers.divider1 = dal_fixed31_32_from_fraction(3, 2);
-	dividers.divider2 = dal_fixed31_32_from_int(2);
-	dividers.divider3 = dal_fixed31_32_from_fraction(5, 2);
+	dividers.divider1 = dc_fixpt_from_fraction(3, 2);
+	dividers.divider2 = dc_fixpt_from_int(2);
+	dividers.divider3 = dc_fixpt_from_fraction(5, 2);
 
 	tf = output_tf->tf;
 
@@ -1157,13 +1440,122 @@ bool mod_color_calculate_regamma_params(struct dc_transfer_func *output_tf,
 
 	ret = true;
 
-	kfree(coeff);
+	kvfree(coeff);
 coeff_alloc_fail:
-	kfree(axix_x);
+	kvfree(axix_x);
 axix_x_alloc_fail:
+	kvfree(rgb_regamma);
+rgb_regamma_alloc_fail:
+	kvfree(rgb_user);
+rgb_user_alloc_fail:
+	return ret;
+}
+
+bool calculate_user_regamma_coeff(struct dc_transfer_func *output_tf,
+		const struct regamma_lut *regamma)
+{
+	struct gamma_coefficients coeff;
+	const struct hw_x_point *coord_x = coordinates_x;
+	uint32_t i = 0;
+
+	do {
+		coeff.a0[i] = dc_fixpt_from_fraction(
+				regamma->coeff.A0[i], 10000000);
+		coeff.a1[i] = dc_fixpt_from_fraction(
+				regamma->coeff.A1[i], 1000);
+		coeff.a2[i] = dc_fixpt_from_fraction(
+				regamma->coeff.A2[i], 1000);
+		coeff.a3[i] = dc_fixpt_from_fraction(
+				regamma->coeff.A3[i], 1000);
+		coeff.user_gamma[i] = dc_fixpt_from_fraction(
+				regamma->coeff.gamma[i], 1000);
+
+		++i;
+	} while (i != 3);
+
+	i = 0;
+	/* fixed_pt library has problems handling too small values */
+	while (i != 32) {
+		output_tf->tf_pts.red[i] = dc_fixpt_zero;
+		output_tf->tf_pts.green[i] = dc_fixpt_zero;
+		output_tf->tf_pts.blue[i] = dc_fixpt_zero;
+		++coord_x;
+		++i;
+	}
+	while (i != MAX_HW_POINTS + 1) {
+		output_tf->tf_pts.red[i] = translate_from_linear_space_ex(
+				coord_x->x, &coeff, 0);
+		output_tf->tf_pts.green[i] = translate_from_linear_space_ex(
+				coord_x->x, &coeff, 1);
+		output_tf->tf_pts.blue[i] = translate_from_linear_space_ex(
+				coord_x->x, &coeff, 2);
+		++coord_x;
+		++i;
+	}
+
+	// this function just clamps output to 0-1
+	build_new_custom_resulted_curve(MAX_HW_POINTS, &output_tf->tf_pts);
+	output_tf->type = TF_TYPE_DISTRIBUTED_POINTS;
+
+	return true;
+}
+
+bool calculate_user_regamma_ramp(struct dc_transfer_func *output_tf,
+		const struct regamma_lut *regamma)
+{
+	struct dc_transfer_func_distributed_points *tf_pts = &output_tf->tf_pts;
+	struct dividers dividers;
+
+	struct pwl_float_data *rgb_user = NULL;
+	struct pwl_float_data_ex *rgb_regamma = NULL;
+	bool ret = false;
+
+	if (regamma == NULL)
+		return false;
+
+	output_tf->type = TF_TYPE_DISTRIBUTED_POINTS;
+
+	rgb_user = kcalloc(GAMMA_RGB_256_ENTRIES + _EXTRA_POINTS,
+			   sizeof(*rgb_user),
+			   GFP_KERNEL);
+	if (!rgb_user)
+		goto rgb_user_alloc_fail;
+
+	rgb_regamma = kcalloc(MAX_HW_POINTS + _EXTRA_POINTS,
+			      sizeof(*rgb_regamma),
+			      GFP_KERNEL);
+	if (!rgb_regamma)
+		goto rgb_regamma_alloc_fail;
+
+	dividers.divider1 = dc_fixpt_from_fraction(3, 2);
+	dividers.divider2 = dc_fixpt_from_int(2);
+	dividers.divider3 = dc_fixpt_from_fraction(5, 2);
+
+	scale_user_regamma_ramp(rgb_user, &regamma->ramp, dividers);
+
+	if (regamma->flags.bits.applyDegamma == 1) {
+		apply_degamma_for_user_regamma(rgb_regamma, MAX_HW_POINTS);
+		copy_rgb_regamma_to_coordinates_x(coordinates_x,
+				MAX_HW_POINTS, rgb_regamma);
+	}
+
+	interpolate_user_regamma(MAX_HW_POINTS, rgb_user,
+			regamma->flags.bits.applyDegamma, tf_pts);
+
+	// no custom HDR curves!
+	tf_pts->end_exponent = 0;
+	tf_pts->x_point_at_y1_red = 1;
+	tf_pts->x_point_at_y1_green = 1;
+	tf_pts->x_point_at_y1_blue = 1;
+
+	// this function just clamps output to 0-1
+	build_new_custom_resulted_curve(MAX_HW_POINTS, tf_pts);
+
+	ret = true;
+
 	kfree(rgb_regamma);
 rgb_regamma_alloc_fail:
-	kfree(rgb_user);
+	kvfree(rgb_user);
 rgb_user_alloc_fail:
 	return ret;
 }
@@ -1192,25 +1584,27 @@ bool mod_color_calculate_degamma_params(struct dc_transfer_func *input_tf,
 
 	input_tf->type = TF_TYPE_DISTRIBUTED_POINTS;
 
-	rgb_user = kzalloc(sizeof(*rgb_user) * (ramp->num_entries + _EXTRA_POINTS),
-			   GFP_KERNEL);
+	rgb_user = kvcalloc(ramp->num_entries + _EXTRA_POINTS,
+			    sizeof(*rgb_user),
+			    GFP_KERNEL);
 	if (!rgb_user)
 		goto rgb_user_alloc_fail;
-	curve = kzalloc(sizeof(*curve) * (MAX_HW_POINTS + _EXTRA_POINTS),
-			GFP_KERNEL);
+	curve = kvcalloc(MAX_HW_POINTS + _EXTRA_POINTS, sizeof(*curve),
+			 GFP_KERNEL);
 	if (!curve)
 		goto curve_alloc_fail;
-	axix_x = kzalloc(sizeof(*axix_x) * (ramp->num_entries + _EXTRA_POINTS),
-			 GFP_KERNEL);
+	axix_x = kvcalloc(ramp->num_entries + _EXTRA_POINTS, sizeof(*axix_x),
+			  GFP_KERNEL);
 	if (!axix_x)
 		goto axix_x_alloc_fail;
-	coeff = kzalloc(sizeof(*coeff) * (MAX_HW_POINTS + _EXTRA_POINTS), GFP_KERNEL);
+	coeff = kvcalloc(MAX_HW_POINTS + _EXTRA_POINTS, sizeof(*coeff),
+			 GFP_KERNEL);
 	if (!coeff)
 		goto coeff_alloc_fail;
 
-	dividers.divider1 = dal_fixed31_32_from_fraction(3, 2);
-	dividers.divider2 = dal_fixed31_32_from_int(2);
-	dividers.divider3 = dal_fixed31_32_from_fraction(5, 2);
+	dividers.divider1 = dc_fixpt_from_fraction(3, 2);
+	dividers.divider2 = dc_fixpt_from_int(2);
+	dividers.divider3 = dc_fixpt_from_fraction(5, 2);
 
 	tf = input_tf->tf;
 
@@ -1246,13 +1640,13 @@ bool mod_color_calculate_degamma_params(struct dc_transfer_func *input_tf,
 
 	ret = true;
 
-	kfree(coeff);
+	kvfree(coeff);
 coeff_alloc_fail:
-	kfree(axix_x);
+	kvfree(axix_x);
 axix_x_alloc_fail:
-	kfree(curve);
+	kvfree(curve);
 curve_alloc_fail:
-	kfree(rgb_user);
+	kvfree(rgb_user);
 rgb_user_alloc_fail:
 
 	return ret;
@@ -1281,8 +1675,9 @@ bool  mod_color_calculate_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 	} else if (trans == TRANSFER_FUNCTION_PQ) {
-		rgb_regamma = kzalloc(sizeof(*rgb_regamma) * (MAX_HW_POINTS +
-						_EXTRA_POINTS), GFP_KERNEL);
+		rgb_regamma = kvcalloc(MAX_HW_POINTS + _EXTRA_POINTS,
+				       sizeof(*rgb_regamma),
+				       GFP_KERNEL);
 		if (!rgb_regamma)
 			goto rgb_regamma_alloc_fail;
 		points->end_exponent = 7;
@@ -1302,11 +1697,12 @@ bool  mod_color_calculate_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 
-		kfree(rgb_regamma);
+		kvfree(rgb_regamma);
 	} else if (trans == TRANSFER_FUNCTION_SRGB ||
 			  trans == TRANSFER_FUNCTION_BT709) {
-		rgb_regamma = kzalloc(sizeof(*rgb_regamma) * (MAX_HW_POINTS +
-						_EXTRA_POINTS), GFP_KERNEL);
+		rgb_regamma = kvcalloc(MAX_HW_POINTS + _EXTRA_POINTS,
+				       sizeof(*rgb_regamma),
+				       GFP_KERNEL);
 		if (!rgb_regamma)
 			goto rgb_regamma_alloc_fail;
 		points->end_exponent = 0;
@@ -1324,7 +1720,26 @@ bool  mod_color_calculate_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 
-		kfree(rgb_regamma);
+		kvfree(rgb_regamma);
+	} else if (trans == TRANSFER_FUNCTION_HLG ||
+		trans == TRANSFER_FUNCTION_HLG12) {
+		rgb_regamma = kvcalloc(MAX_HW_POINTS + _EXTRA_POINTS,
+				       sizeof(*rgb_regamma),
+				       GFP_KERNEL);
+		if (!rgb_regamma)
+			goto rgb_regamma_alloc_fail;
+
+		build_hlg_regamma(rgb_regamma,
+				MAX_HW_POINTS,
+				coordinates_x,
+				trans == TRANSFER_FUNCTION_HLG12 ? true:false);
+		for (i = 0; i <= MAX_HW_POINTS ; i++) {
+			points->red[i]    = rgb_regamma[i].r;
+			points->green[i]  = rgb_regamma[i].g;
+			points->blue[i]   = rgb_regamma[i].b;
+		}
+		ret = true;
+		kvfree(rgb_regamma);
 	}
 rgb_regamma_alloc_fail:
 	return ret;
@@ -1348,8 +1763,9 @@ bool  mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 	} else if (trans == TRANSFER_FUNCTION_PQ) {
-		rgb_degamma = kzalloc(sizeof(*rgb_degamma) * (MAX_HW_POINTS +
-						_EXTRA_POINTS), GFP_KERNEL);
+		rgb_degamma = kvcalloc(MAX_HW_POINTS + _EXTRA_POINTS,
+				       sizeof(*rgb_degamma),
+				       GFP_KERNEL);
 		if (!rgb_degamma)
 			goto rgb_degamma_alloc_fail;
 
@@ -1364,11 +1780,12 @@ bool  mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 
-		kfree(rgb_degamma);
+		kvfree(rgb_degamma);
 	} else if (trans == TRANSFER_FUNCTION_SRGB ||
 			  trans == TRANSFER_FUNCTION_BT709) {
-		rgb_degamma = kzalloc(sizeof(*rgb_degamma) * (MAX_HW_POINTS +
-						_EXTRA_POINTS), GFP_KERNEL);
+		rgb_degamma = kvcalloc(MAX_HW_POINTS + _EXTRA_POINTS,
+				       sizeof(*rgb_degamma),
+				       GFP_KERNEL);
 		if (!rgb_degamma)
 			goto rgb_degamma_alloc_fail;
 
@@ -1382,7 +1799,26 @@ bool  mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans,
 		}
 		ret = true;
 
-		kfree(rgb_degamma);
+		kvfree(rgb_degamma);
+	} else if (trans == TRANSFER_FUNCTION_HLG ||
+		trans == TRANSFER_FUNCTION_HLG12) {
+		rgb_degamma = kvcalloc(MAX_HW_POINTS + _EXTRA_POINTS,
+				       sizeof(*rgb_degamma),
+				       GFP_KERNEL);
+		if (!rgb_degamma)
+			goto rgb_degamma_alloc_fail;
+
+		build_hlg_degamma(rgb_degamma,
+				MAX_HW_POINTS,
+				coordinates_x,
+				trans == TRANSFER_FUNCTION_HLG12 ? true:false);
+		for (i = 0; i <= MAX_HW_POINTS ; i++) {
+			points->red[i]    = rgb_degamma[i].r;
+			points->green[i]  = rgb_degamma[i].g;
+			points->blue[i]   = rgb_degamma[i].b;
+		}
+		ret = true;
+		kvfree(rgb_degamma);
 	}
 	points->end_exponent = 0;
 	points->x_point_at_y1_red = 1;
diff --git a/drivers/gpu/drm/amd/display/modules/color/color_gamma.h b/drivers/gpu/drm/amd/display/modules/color/color_gamma.h
index b7f9bc27d101..b64048991a95 100644
--- a/drivers/gpu/drm/amd/display/modules/color/color_gamma.h
+++ b/drivers/gpu/drm/amd/display/modules/color/color_gamma.h
@@ -32,6 +32,47 @@ struct dc_transfer_func_distributed_points;
 struct dc_rgb_fixed;
 enum dc_transfer_func_predefined;
 
+/* For SetRegamma ADL interface support
+ * Must match escape type
+ */
+union regamma_flags {
+	unsigned int raw;
+	struct {
+		unsigned int gammaRampArray       :1;    // RegammaRamp is in use
+		unsigned int gammaFromEdid        :1;    //gamma from edid is in use
+		unsigned int gammaFromEdidEx      :1;    //gamma from edid is in use , but only for Display Id 1.2
+		unsigned int gammaFromUser        :1;    //user custom gamma is used
+		unsigned int coeffFromUser        :1;    //coeff. A0-A3 from user is in use
+		unsigned int coeffFromEdid        :1;    //coeff. A0-A3 from edid is in use
+		unsigned int applyDegamma         :1;    //flag for additional degamma correction in driver
+		unsigned int gammaPredefinedSRGB  :1;    //flag for SRGB gamma
+		unsigned int gammaPredefinedPQ    :1;    //flag for PQ gamma
+		unsigned int gammaPredefinedPQ2084Interim :1;    //flag for PQ gamma, lower max nits
+		unsigned int gammaPredefined36    :1;    //flag for 3.6 gamma
+		unsigned int gammaPredefinedReset :1;    //flag to return to previous gamma
+	} bits;
+};
+
+struct regamma_ramp {
+	unsigned short gamma[256*3];  // gamma ramp packed  in same way as OS windows ,r , g & b
+};
+
+struct regamma_coeff {
+	int    gamma[3];
+	int    A0[3];
+	int    A1[3];
+	int    A2[3];
+	int    A3[3];
+};
+
+struct regamma_lut {
+	union regamma_flags flags;
+	union {
+		struct regamma_ramp ramp;
+		struct regamma_coeff coeff;
+	};
+};
+
 void setup_x_points_distribution(void);
 void precompute_pq(void);
 void precompute_de_pq(void);
@@ -45,9 +86,14 @@ bool mod_color_calculate_degamma_params(struct dc_transfer_func *output_tf,
 bool mod_color_calculate_curve(enum dc_transfer_func_predefined  trans,
 		struct dc_transfer_func_distributed_points *points);
 
-bool  mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans,
+bool mod_color_calculate_degamma_curve(enum dc_transfer_func_predefined trans,
 				struct dc_transfer_func_distributed_points *points);
 
+bool calculate_user_regamma_coeff(struct dc_transfer_func *output_tf,
+		const struct regamma_lut *regamma);
+
+bool calculate_user_regamma_ramp(struct dc_transfer_func *output_tf,
+		const struct regamma_lut *regamma);
 
 
 #endif /* COLOR_MOD_COLOR_GAMMA_H_ */
diff --git a/drivers/gpu/drm/amd/display/modules/color/luts_1d.h b/drivers/gpu/drm/amd/display/modules/color/luts_1d.h
new file mode 100644
index 000000000000..66b1fad572ac
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/modules/color/luts_1d.h
@@ -0,0 +1,51 @@
+/*
+ * Copyright 2016 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+#ifndef LUTS_1D_H
+#define LUTS_1D_H
+
+#include "hw_shared.h"
+
+struct point_config {
+	uint32_t custom_float_x;
+	uint32_t custom_float_y;
+	uint32_t custom_float_slope;
+};
+
+struct lut_point {
+	uint32_t red;
+	uint32_t green;
+	uint32_t blue;
+	uint32_t delta_red;
+	uint32_t delta_green;
+	uint32_t delta_blue;
+};
+
+struct pwl_1dlut_parameter {
+	struct gamma_curve	arr_curve_points[34];
+	struct point_config	arr_points[2];
+	struct lut_point rgb_resulted[256];
+	uint32_t hw_points_num;
+};
+#endif // LUTS_1D_H
diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
index 27d4003aa2c7..fa344ceafc17 100644
--- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
+++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
@@ -155,7 +155,8 @@ struct mod_freesync *mod_freesync_create(struct dc *dc)
 	if (core_freesync == NULL)
 		goto fail_alloc_context;
 
-	core_freesync->map = kzalloc(sizeof(struct freesync_entity) * MOD_FREESYNC_MAX_CONCURRENT_STREAMS,
+	core_freesync->map = kcalloc(MOD_FREESYNC_MAX_CONCURRENT_STREAMS,
+					sizeof(struct freesync_entity),
 					GFP_KERNEL);
 
 	if (core_freesync->map == NULL)
diff --git a/drivers/gpu/drm/amd/display/modules/inc/mod_stats.h b/drivers/gpu/drm/amd/display/modules/inc/mod_stats.h
index 3230e2adb870..3812094b52e8 100644
--- a/drivers/gpu/drm/amd/display/modules/inc/mod_stats.h
+++ b/drivers/gpu/drm/amd/display/modules/inc/mod_stats.h
@@ -46,6 +46,10 @@ void mod_stats_dump(struct mod_stats *mod_stats);
 
 void mod_stats_reset_data(struct mod_stats *mod_stats);
 
+void mod_stats_update_event(struct mod_stats *mod_stats,
+		char *event_string,
+		unsigned int length);
+
 void mod_stats_update_flip(struct mod_stats *mod_stats,
 		unsigned long timestamp_in_ns);
 
diff --git a/drivers/gpu/drm/amd/display/modules/stats/stats.c b/drivers/gpu/drm/amd/display/modules/stats/stats.c
index 041f87b73d5f..3d4c1b1ab8c4 100644
--- a/drivers/gpu/drm/amd/display/modules/stats/stats.c
+++ b/drivers/gpu/drm/amd/display/modules/stats/stats.c
@@ -29,16 +29,21 @@
 #include "core_types.h"
 
 #define DAL_STATS_ENABLE_REGKEY			"DalStatsEnable"
-#define DAL_STATS_ENABLE_REGKEY_DEFAULT		0x00000001
+#define DAL_STATS_ENABLE_REGKEY_DEFAULT		0x00000000
 #define DAL_STATS_ENABLE_REGKEY_ENABLED		0x00000001
 
 #define DAL_STATS_ENTRIES_REGKEY		"DalStatsEntries"
 #define DAL_STATS_ENTRIES_REGKEY_DEFAULT	0x00350000
 #define DAL_STATS_ENTRIES_REGKEY_MAX		0x01000000
 
+#define DAL_STATS_EVENT_ENTRIES_DEFAULT		0x00000100
+
 #define MOD_STATS_NUM_VSYNCS			5
+#define MOD_STATS_EVENT_STRING_MAX		512
 
 struct stats_time_cache {
+	unsigned int entry_id;
+
 	unsigned long flip_timestamp_in_ns;
 	unsigned long vupdate_timestamp_in_ns;
 
@@ -63,15 +68,26 @@ struct stats_time_cache {
 	unsigned int flags;
 };
 
+struct stats_event_cache {
+	unsigned int entry_id;
+	char event_string[MOD_STATS_EVENT_STRING_MAX];
+};
+
 struct core_stats {
 	struct mod_stats public;
 	struct dc *dc;
 
+	bool enabled;
+	unsigned int entries;
+	unsigned int event_entries;
+	unsigned int entry_id;
+
 	struct stats_time_cache *time;
 	unsigned int index;
 
-	bool enabled;
-	unsigned int entries;
+	struct stats_event_cache *events;
+	unsigned int event_index;
+
 };
 
 #define MOD_STATS_TO_CORE(mod_stats)\
@@ -99,12 +115,12 @@ struct mod_stats *mod_stats_create(struct dc *dc)
 	unsigned int reg_data;
 	int i = 0;
 
+	if (dc == NULL)
+		goto fail_construct;
+
 	core_stats = kzalloc(sizeof(struct core_stats), GFP_KERNEL);
 
 	if (core_stats == NULL)
-		goto fail_alloc_context;
-
-	if (dc == NULL)
 		goto fail_construct;
 
 	core_stats->dc = dc;
@@ -115,33 +131,53 @@ struct mod_stats *mod_stats_create(struct dc *dc)
 			&reg_data, sizeof(unsigned int), &flag))
 		core_stats->enabled = reg_data;
 
-	core_stats->entries = DAL_STATS_ENTRIES_REGKEY_DEFAULT;
-	if (dm_read_persistent_data(dc->ctx, NULL, NULL,
-			DAL_STATS_ENTRIES_REGKEY,
-			&reg_data, sizeof(unsigned int), &flag)) {
-		if (reg_data > DAL_STATS_ENTRIES_REGKEY_MAX)
-			core_stats->entries = DAL_STATS_ENTRIES_REGKEY_MAX;
-		else
-			core_stats->entries = reg_data;
-	}
+	if (core_stats->enabled) {
+		core_stats->entries = DAL_STATS_ENTRIES_REGKEY_DEFAULT;
+		if (dm_read_persistent_data(dc->ctx, NULL, NULL,
+				DAL_STATS_ENTRIES_REGKEY,
+				&reg_data, sizeof(unsigned int), &flag)) {
+			if (reg_data > DAL_STATS_ENTRIES_REGKEY_MAX)
+				core_stats->entries = DAL_STATS_ENTRIES_REGKEY_MAX;
+			else
+				core_stats->entries = reg_data;
+		}
+		core_stats->time = kcalloc(core_stats->entries,
+						sizeof(struct stats_time_cache),
+						GFP_KERNEL);
 
-	core_stats->time = kzalloc(sizeof(struct stats_time_cache) * core_stats->entries,
-					GFP_KERNEL);
+		if (core_stats->time == NULL)
+			goto fail_construct_time;
 
-	if (core_stats->time == NULL)
-		goto fail_construct;
+		core_stats->event_entries = DAL_STATS_EVENT_ENTRIES_DEFAULT;
+		core_stats->events = kcalloc(core_stats->event_entries,
+					     sizeof(struct stats_event_cache),
+					     GFP_KERNEL);
+
+		if (core_stats->events == NULL)
+			goto fail_construct_events;
+
+	} else {
+		core_stats->entries = 0;
+	}
 
 	/* Purposely leave index 0 unused so we don't need special logic to
 	 * handle calculation cases that depend on previous flip data.
 	 */
 	core_stats->index = 1;
+	core_stats->event_index = 0;
+
+	// Keeps track of ordering within the different stats structures
+	core_stats->entry_id = 0;
 
 	return &core_stats->public;
 
-fail_construct:
+fail_construct_events:
+	kfree(core_stats->time);
+
+fail_construct_time:
 	kfree(core_stats);
 
-fail_alloc_context:
+fail_construct:
 	return NULL;
 }
 
@@ -153,6 +189,9 @@ void mod_stats_destroy(struct mod_stats *mod_stats)
 		if (core_stats->time != NULL)
 			kfree(core_stats->time);
 
+		if (core_stats->events != NULL)
+			kfree(core_stats->events);
+
 		kfree(core_stats);
 	}
 }
@@ -163,7 +202,11 @@ void mod_stats_dump(struct mod_stats *mod_stats)
 	struct dal_logger *logger = NULL;
 	struct core_stats *core_stats = NULL;
 	struct stats_time_cache *time = NULL;
+	struct stats_event_cache *events = NULL;
+	unsigned int time_index = 1;
+	unsigned int event_index = 0;
 	unsigned int index = 0;
+	struct log_entry log_entry;
 
 	if (mod_stats == NULL)
 		return;
@@ -172,45 +215,62 @@ void mod_stats_dump(struct mod_stats *mod_stats)
 	dc = core_stats->dc;
 	logger = dc->ctx->logger;
 	time = core_stats->time;
-
-	//LogEntry* pLog = GetLog()->Open(LogMajor_ISR, LogMinor_ISR_FreeSyncSW);
-
-	//if (!pLog->IsDummyEntry())
-	{
-		dm_logger_write(logger, LOG_PROFILING, "==Display Caps==\n");
-		dm_logger_write(logger, LOG_PROFILING, "\n");
-		dm_logger_write(logger, LOG_PROFILING, "\n");
-
-		dm_logger_write(logger, LOG_PROFILING, "==Stats==\n");
-		dm_logger_write(logger, LOG_PROFILING,
-			"render avgRender minWindow midPoint maxWindow vsyncToFlip flipToVsync #vsyncBetweenFlip #frame insertDuration vTotalMin vTotalMax eventTrigs vSyncTime1 vSyncTime2 vSyncTime3 vSyncTime4 vSyncTime5 flags\n");
-
-		for (int i = 0; i < core_stats->index && i < core_stats->entries; i++) {
-			dm_logger_write(logger, LOG_PROFILING,
-					"%u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u  %u\n",
-					time[i].render_time_in_us,
-					time[i].avg_render_time_in_us_last_ten,
-					time[i].min_window,
-					time[i].lfc_mid_point_in_us,
-					time[i].max_window,
-					time[i].vsync_to_flip_time_in_us,
-					time[i].flip_to_vsync_time_in_us,
-					time[i].num_vsync_between_flips,
-					time[i].num_frames_inserted,
-					time[i].inserted_duration_in_us,
-					time[i].v_total_min,
-					time[i].v_total_max,
-					time[i].event_triggers,
-					time[i].v_sync_time_in_us[0],
-					time[i].v_sync_time_in_us[1],
-					time[i].v_sync_time_in_us[2],
-					time[i].v_sync_time_in_us[3],
-					time[i].v_sync_time_in_us[4],
-					time[i].flags);
+	events = core_stats->events;
+
+	DISPLAY_STATS_BEGIN(log_entry);
+
+	DISPLAY_STATS("==Display Caps==\n");
+
+	DISPLAY_STATS("==Display Stats==\n");
+
+	DISPLAY_STATS("%10s %10s %10s %10s %10s"
+			" %11s %11s %17s %10s %14s"
+			" %10s %10s %10s %10s %10s"
+			" %10s %10s %10s %10s\n",
+		"render", "avgRender",
+		"minWindow", "midPoint", "maxWindow",
+		"vsyncToFlip", "flipToVsync", "vsyncsBetweenFlip",
+		"numFrame", "insertDuration",
+		"vTotalMin", "vTotalMax", "eventTrigs",
+		"vSyncTime1", "vSyncTime2", "vSyncTime3",
+		"vSyncTime4", "vSyncTime5", "flags");
+
+	for (int i = 0; i < core_stats->entry_id; i++) {
+		if (event_index < core_stats->event_index &&
+				i == events[event_index].entry_id) {
+			DISPLAY_STATS("==Event==%s\n", events[event_index].event_string);
+			event_index++;
+		} else if (time_index < core_stats->index &&
+				i == time[time_index].entry_id) {
+			DISPLAY_STATS("%10u %10u %10u %10u %10u"
+					" %11u %11u %17u %10u %14u"
+					" %10u %10u %10u %10u %10u"
+					" %10u %10u %10u %10u\n",
+				time[time_index].render_time_in_us,
+				time[time_index].avg_render_time_in_us_last_ten,
+				time[time_index].min_window,
+				time[time_index].lfc_mid_point_in_us,
+				time[time_index].max_window,
+				time[time_index].vsync_to_flip_time_in_us,
+				time[time_index].flip_to_vsync_time_in_us,
+				time[time_index].num_vsync_between_flips,
+				time[time_index].num_frames_inserted,
+				time[time_index].inserted_duration_in_us,
+				time[time_index].v_total_min,
+				time[time_index].v_total_max,
+				time[time_index].event_triggers,
+				time[time_index].v_sync_time_in_us[0],
+				time[time_index].v_sync_time_in_us[1],
+				time[time_index].v_sync_time_in_us[2],
+				time[time_index].v_sync_time_in_us[3],
+				time[time_index].v_sync_time_in_us[4],
+				time[time_index].flags);
+
+			time_index++;
 		}
 	}
-	//GetLog()->Close(pLog);
-	//GetLog()->UnSetLogMask(LogMajor_ISR, LogMinor_ISR_FreeSyncSW);
+
+	DISPLAY_STATS_END(log_entry);
 }
 
 void mod_stats_reset_data(struct mod_stats *mod_stats)
@@ -227,7 +287,46 @@ void mod_stats_reset_data(struct mod_stats *mod_stats)
 	memset(core_stats->time, 0,
 		sizeof(struct stats_time_cache) * core_stats->entries);
 
-	core_stats->index = 0;
+	memset(core_stats->events, 0,
+		sizeof(struct stats_event_cache) * core_stats->event_entries);
+
+	core_stats->index = 1;
+	core_stats->event_index = 0;
+
+	// Keeps track of ordering within the different stats structures
+	core_stats->entry_id = 0;
+}
+
+void mod_stats_update_event(struct mod_stats *mod_stats,
+		char *event_string,
+		unsigned int length)
+{
+	struct core_stats *core_stats = NULL;
+	struct stats_event_cache *events = NULL;
+	unsigned int index = 0;
+	unsigned int copy_length = 0;
+
+	if (mod_stats == NULL)
+		return;
+
+	core_stats = MOD_STATS_TO_CORE(mod_stats);
+
+	if (core_stats->event_index >= core_stats->event_entries)
+		return;
+
+	events = core_stats->events;
+	index = core_stats->event_index;
+
+	copy_length = length;
+	if (length > MOD_STATS_EVENT_STRING_MAX)
+		copy_length = MOD_STATS_EVENT_STRING_MAX;
+
+	memcpy(&events[index].event_string, event_string, copy_length);
+	events[index].event_string[copy_length - 1] = '\0';
+
+	events[index].entry_id = core_stats->entry_id;
+	core_stats->event_index++;
+	core_stats->entry_id++;
 }
 
 void mod_stats_update_flip(struct mod_stats *mod_stats,
@@ -250,7 +349,7 @@ void mod_stats_update_flip(struct mod_stats *mod_stats,
 
 	time[index].flip_timestamp_in_ns = timestamp_in_ns;
 	time[index].render_time_in_us =
-		timestamp_in_ns - time[index - 1].flip_timestamp_in_ns;
+		(timestamp_in_ns - time[index - 1].flip_timestamp_in_ns) / 1000;
 
 	if (index >= 10) {
 		for (unsigned int i = 0; i < 10; i++)
@@ -261,12 +360,16 @@ void mod_stats_update_flip(struct mod_stats *mod_stats,
 
 	if (time[index].num_vsync_between_flips > 0)
 		time[index].vsync_to_flip_time_in_us =
-			timestamp_in_ns - time[index].vupdate_timestamp_in_ns;
+			(timestamp_in_ns -
+				time[index].vupdate_timestamp_in_ns) / 1000;
 	else
 		time[index].vsync_to_flip_time_in_us =
-			timestamp_in_ns - time[index - 1].vupdate_timestamp_in_ns;
+			(timestamp_in_ns -
+				time[index - 1].vupdate_timestamp_in_ns) / 1000;
 
+	time[index].entry_id = core_stats->entry_id;
 	core_stats->index++;
+	core_stats->entry_id++;
 }
 
 void mod_stats_update_vupdate(struct mod_stats *mod_stats,
@@ -275,6 +378,8 @@ void mod_stats_update_vupdate(struct mod_stats *mod_stats,
 	struct core_stats *core_stats = NULL;
 	struct stats_time_cache *time = NULL;
 	unsigned int index = 0;
+	unsigned int num_vsyncs = 0;
+	unsigned int prev_vsync_in_ns = 0;
 
 	if (mod_stats == NULL)
 		return;
@@ -286,14 +391,27 @@ void mod_stats_update_vupdate(struct mod_stats *mod_stats,
 
 	time = core_stats->time;
 	index = core_stats->index;
+	num_vsyncs = time[index].num_vsync_between_flips;
+
+	if (num_vsyncs < MOD_STATS_NUM_VSYNCS) {
+		if (num_vsyncs == 0) {
+			prev_vsync_in_ns =
+				time[index - 1].vupdate_timestamp_in_ns;
+
+			time[index].flip_to_vsync_time_in_us =
+				(timestamp_in_ns -
+					time[index - 1].flip_timestamp_in_ns) /
+					1000;
+		} else {
+			prev_vsync_in_ns =
+				time[index].vupdate_timestamp_in_ns;
+		}
 
-	time[index].vupdate_timestamp_in_ns = timestamp_in_ns;
-	if (time[index].num_vsync_between_flips < MOD_STATS_NUM_VSYNCS)
-		time[index].v_sync_time_in_us[time[index].num_vsync_between_flips] =
-			timestamp_in_ns - time[index - 1].vupdate_timestamp_in_ns;
-	time[index].flip_to_vsync_time_in_us =
-		timestamp_in_ns - time[index - 1].flip_timestamp_in_ns;
+		time[index].v_sync_time_in_us[num_vsyncs] =
+			(timestamp_in_ns - prev_vsync_in_ns) / 1000;
+	}
 
+	time[index].vupdate_timestamp_in_ns = timestamp_in_ns;
 	time[index].num_vsync_between_flips++;
 }
 
diff --git a/drivers/gpu/drm/amd/include/amd_pcie.h b/drivers/gpu/drm/amd/include/amd_pcie.h
index 5eb895fd98bf..9cb9ceb4d74d 100644
--- a/drivers/gpu/drm/amd/include/amd_pcie.h
+++ b/drivers/gpu/drm/amd/include/amd_pcie.h
@@ -27,6 +27,7 @@
 #define CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1        0x00010000
 #define CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2        0x00020000
 #define CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3        0x00040000
+#define CAIL_PCIE_LINK_SPEED_SUPPORT_GEN4        0x00080000
 #define CAIL_PCIE_LINK_SPEED_SUPPORT_MASK        0xFFFF0000
 #define CAIL_PCIE_LINK_SPEED_SUPPORT_SHIFT       16
 
@@ -34,6 +35,7 @@
 #define CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN1   0x00000001
 #define CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN2   0x00000002
 #define CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN3   0x00000004
+#define CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_GEN4   0x00000008
 #define CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_MASK   0x0000FFFF
 #define CAIL_ASIC_PCIE_LINK_SPEED_SUPPORT_SHIFT  0
 
diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h
index 9fa3aaef3f33..265621d8945c 100644
--- a/drivers/gpu/drm/amd/include/amd_shared.h
+++ b/drivers/gpu/drm/amd/include/amd_shared.h
@@ -92,7 +92,7 @@ enum amd_powergating_state {
 #define AMD_CG_SUPPORT_GFX_3D_CGLS		(1 << 21)
 #define AMD_CG_SUPPORT_DRM_MGCG			(1 << 22)
 #define AMD_CG_SUPPORT_DF_MGCG			(1 << 23)
-
+#define AMD_CG_SUPPORT_VCN_MGCG			(1 << 24)
 /* PG flags */
 #define AMD_PG_SUPPORT_GFX_PG			(1 << 0)
 #define AMD_PG_SUPPORT_GFX_SMG			(1 << 1)
@@ -108,46 +108,77 @@ enum amd_powergating_state {
 #define AMD_PG_SUPPORT_GFX_QUICK_MG		(1 << 11)
 #define AMD_PG_SUPPORT_GFX_PIPELINE		(1 << 12)
 #define AMD_PG_SUPPORT_MMHUB			(1 << 13)
+#define AMD_PG_SUPPORT_VCN			(1 << 14)
+
+enum PP_FEATURE_MASK {
+	PP_SCLK_DPM_MASK = 0x1,
+	PP_MCLK_DPM_MASK = 0x2,
+	PP_PCIE_DPM_MASK = 0x4,
+	PP_SCLK_DEEP_SLEEP_MASK = 0x8,
+	PP_POWER_CONTAINMENT_MASK = 0x10,
+	PP_UVD_HANDSHAKE_MASK = 0x20,
+	PP_SMC_VOLTAGE_CONTROL_MASK = 0x40,
+	PP_VBI_TIME_SUPPORT_MASK = 0x80,
+	PP_ULV_MASK = 0x100,
+	PP_ENABLE_GFX_CG_THRU_SMU = 0x200,
+	PP_CLOCK_STRETCH_MASK = 0x400,
+	PP_OD_FUZZY_FAN_CONTROL_MASK = 0x800,
+	PP_SOCCLK_DPM_MASK = 0x1000,
+	PP_DCEFCLK_DPM_MASK = 0x2000,
+	PP_OVERDRIVE_MASK = 0x4000,
+	PP_GFXOFF_MASK = 0x8000,
+	PP_ACG_MASK = 0x10000,
+	PP_STUTTER_MODE = 0x20000,
+};
 
+/**
+ * struct amd_ip_funcs - general hooks for managing amdgpu IP Blocks
+ */
 struct amd_ip_funcs {
-	/* Name of IP block */
+	/** @name: Name of IP block */
 	char *name;
-	/* sets up early driver state (pre sw_init), does not configure hw - Optional */
+	/**
+	 * @early_init:
+	 *
+	 * sets up early driver state (pre sw_init),
+	 * does not configure hw - Optional
+	 */
 	int (*early_init)(void *handle);
-	/* sets up late driver/hw state (post hw_init) - Optional */
+	/** @late_init: sets up late driver/hw state (post hw_init) - Optional */
 	int (*late_init)(void *handle);
-	/* sets up driver state, does not configure hw */
+	/** @sw_init: sets up driver state, does not configure hw */
 	int (*sw_init)(void *handle);
-	/* tears down driver state, does not configure hw */
+	/** @sw_fini: tears down driver state, does not configure hw */
 	int (*sw_fini)(void *handle);
-	/* sets up the hw state */
+	/** @hw_init: sets up the hw state */
 	int (*hw_init)(void *handle);
-	/* tears down the hw state */
+	/** @hw_fini: tears down the hw state */
 	int (*hw_fini)(void *handle);
+	/** @late_fini: final cleanup */
 	void (*late_fini)(void *handle);
-	/* handles IP specific hw/sw changes for suspend */
+	/** @suspend: handles IP specific hw/sw changes for suspend */
 	int (*suspend)(void *handle);
-	/* handles IP specific hw/sw changes for resume */
+	/** @resume: handles IP specific hw/sw changes for resume */
 	int (*resume)(void *handle);
-	/* returns current IP block idle status */
+	/** @is_idle: returns current IP block idle status */
 	bool (*is_idle)(void *handle);
-	/* poll for idle */
+	/** @wait_for_idle: poll for idle */
 	int (*wait_for_idle)(void *handle);
-	/* check soft reset the IP block */
+	/** @check_soft_reset: check soft reset the IP block */
 	bool (*check_soft_reset)(void *handle);
-	/* pre soft reset the IP block */
+	/** @pre_soft_reset: pre soft reset the IP block */
 	int (*pre_soft_reset)(void *handle);
-	/* soft reset the IP block */
+	/** @soft_reset: soft reset the IP block */
 	int (*soft_reset)(void *handle);
-	/* post soft reset the IP block */
+	/** @post_soft_reset: post soft reset the IP block */
 	int (*post_soft_reset)(void *handle);
-	/* enable/disable cg for the IP block */
+	/** @set_clockgating_state: enable/disable cg for the IP block */
 	int (*set_clockgating_state)(void *handle,
 				     enum amd_clockgating_state state);
-	/* enable/disable pg for the IP block */
+	/** @set_powergating_state: enable/disable pg for the IP block */
 	int (*set_powergating_state)(void *handle,
 				     enum amd_powergating_state state);
-	/* get current clockgating status */
+	/** @get_clockgating_state: get current clockgating status */
 	void (*get_clockgating_state)(void *handle, u32 *flags);
 };
 
diff --git a/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_offset.h
index f730d0629020..b6f74bf4af02 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_offset.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_offset.h
@@ -2095,6 +2095,18 @@
 #define mmDC_GPIO_AUX_CTRL_2_BASE_IDX                                                                  2
 #define mmDC_GPIO_RXEN                                                                                 0x212f
 #define mmDC_GPIO_RXEN_BASE_IDX                                                                        2
+#define mmDC_GPIO_AUX_CTRL_3                                                                           0x2130
+#define mmDC_GPIO_AUX_CTRL_3_BASE_IDX                                                                  2
+#define mmDC_GPIO_AUX_CTRL_4                                                                           0x2131
+#define mmDC_GPIO_AUX_CTRL_4_BASE_IDX                                                                  2
+#define mmDC_GPIO_AUX_CTRL_5                                                                           0x2132
+#define mmDC_GPIO_AUX_CTRL_5_BASE_IDX                                                                  2
+#define mmAUXI2C_PAD_ALL_PWR_OK                                                                        0x2133
+#define mmAUXI2C_PAD_ALL_PWR_OK_BASE_IDX                                                               2
+#define mmDC_GPIO_PULLUPEN                                                                             0x2134
+#define mmDC_GPIO_PULLUPEN_BASE_IDX                                                                    2
+#define mmDC_GPIO_AUX_CTRL_6                                                                           0x2135
+#define mmDC_GPIO_AUX_CTRL_6_BASE_IDX                                                                  2
 #define mmBPHYC_DAC_MACRO_CNTL                                                                         0x2136
 #define mmBPHYC_DAC_MACRO_CNTL_BASE_IDX                                                                2
 #define mmDAC_MACRO_CNTL_RESERVED0                                                                     0x2136
diff --git a/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_sh_mask.h
index 6d3162c42957..bcd190a3fcdd 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/dce/dce_12_0_sh_mask.h
@@ -10971,6 +10971,158 @@
 #define DC_GPIO_RXEN__DC_GPIO_BLON_RXEN_MASK                                                                  0x00100000L
 #define DC_GPIO_RXEN__DC_GPIO_DIGON_RXEN_MASK                                                                 0x00200000L
 #define DC_GPIO_RXEN__DC_GPIO_ENA_BL_RXEN_MASK                                                                0x00400000L
+//DC_GPIO_AUX_CTRL_3
+#define DC_GPIO_AUX_CTRL_3__AUX1_NEN_RTERM__SHIFT                                                             0x0
+#define DC_GPIO_AUX_CTRL_3__AUX2_NEN_RTERM__SHIFT                                                             0x1
+#define DC_GPIO_AUX_CTRL_3__AUX3_NEN_RTERM__SHIFT                                                             0x2
+#define DC_GPIO_AUX_CTRL_3__AUX4_NEN_RTERM__SHIFT                                                             0x3
+#define DC_GPIO_AUX_CTRL_3__AUX5_NEN_RTERM__SHIFT                                                             0x4
+#define DC_GPIO_AUX_CTRL_3__AUX6_NEN_RTERM__SHIFT                                                             0x5
+#define DC_GPIO_AUX_CTRL_3__AUX1_DP_DN_SWAP__SHIFT                                                            0x8
+#define DC_GPIO_AUX_CTRL_3__AUX2_DP_DN_SWAP__SHIFT                                                            0x9
+#define DC_GPIO_AUX_CTRL_3__AUX3_DP_DN_SWAP__SHIFT                                                            0xa
+#define DC_GPIO_AUX_CTRL_3__AUX4_DP_DN_SWAP__SHIFT                                                            0xb
+#define DC_GPIO_AUX_CTRL_3__AUX5_DP_DN_SWAP__SHIFT                                                            0xc
+#define DC_GPIO_AUX_CTRL_3__AUX6_DP_DN_SWAP__SHIFT                                                            0xd
+#define DC_GPIO_AUX_CTRL_3__AUX1_HYS_TUNE__SHIFT                                                              0x10
+#define DC_GPIO_AUX_CTRL_3__AUX2_HYS_TUNE__SHIFT                                                              0x12
+#define DC_GPIO_AUX_CTRL_3__AUX3_HYS_TUNE__SHIFT                                                              0x14
+#define DC_GPIO_AUX_CTRL_3__AUX4_HYS_TUNE__SHIFT                                                              0x16
+#define DC_GPIO_AUX_CTRL_3__AUX5_HYS_TUNE__SHIFT                                                              0x18
+#define DC_GPIO_AUX_CTRL_3__AUX6_HYS_TUNE__SHIFT                                                              0x1a
+#define DC_GPIO_AUX_CTRL_3__AUX1_NEN_RTERM_MASK                                                               0x00000001L
+#define DC_GPIO_AUX_CTRL_3__AUX2_NEN_RTERM_MASK                                                               0x00000002L
+#define DC_GPIO_AUX_CTRL_3__AUX3_NEN_RTERM_MASK                                                               0x00000004L
+#define DC_GPIO_AUX_CTRL_3__AUX4_NEN_RTERM_MASK                                                               0x00000008L
+#define DC_GPIO_AUX_CTRL_3__AUX5_NEN_RTERM_MASK                                                               0x00000010L
+#define DC_GPIO_AUX_CTRL_3__AUX6_NEN_RTERM_MASK                                                               0x00000020L
+#define DC_GPIO_AUX_CTRL_3__AUX1_DP_DN_SWAP_MASK                                                              0x00000100L
+#define DC_GPIO_AUX_CTRL_3__AUX2_DP_DN_SWAP_MASK                                                              0x00000200L
+#define DC_GPIO_AUX_CTRL_3__AUX3_DP_DN_SWAP_MASK                                                              0x00000400L
+#define DC_GPIO_AUX_CTRL_3__AUX4_DP_DN_SWAP_MASK                                                              0x00000800L
+#define DC_GPIO_AUX_CTRL_3__AUX5_DP_DN_SWAP_MASK                                                              0x00001000L
+#define DC_GPIO_AUX_CTRL_3__AUX6_DP_DN_SWAP_MASK                                                              0x00002000L
+#define DC_GPIO_AUX_CTRL_3__AUX1_HYS_TUNE_MASK                                                                0x00030000L
+#define DC_GPIO_AUX_CTRL_3__AUX2_HYS_TUNE_MASK                                                                0x000C0000L
+#define DC_GPIO_AUX_CTRL_3__AUX3_HYS_TUNE_MASK                                                                0x00300000L
+#define DC_GPIO_AUX_CTRL_3__AUX4_HYS_TUNE_MASK                                                                0x00C00000L
+#define DC_GPIO_AUX_CTRL_3__AUX5_HYS_TUNE_MASK                                                                0x03000000L
+#define DC_GPIO_AUX_CTRL_3__AUX6_HYS_TUNE_MASK                                                                0x0C000000L
+//DC_GPIO_AUX_CTRL_4
+#define DC_GPIO_AUX_CTRL_4__AUX1_AUX_CTRL__SHIFT                                                              0x0
+#define DC_GPIO_AUX_CTRL_4__AUX2_AUX_CTRL__SHIFT                                                              0x4
+#define DC_GPIO_AUX_CTRL_4__AUX3_AUX_CTRL__SHIFT                                                              0x8
+#define DC_GPIO_AUX_CTRL_4__AUX4_AUX_CTRL__SHIFT                                                              0xc
+#define DC_GPIO_AUX_CTRL_4__AUX5_AUX_CTRL__SHIFT                                                              0x10
+#define DC_GPIO_AUX_CTRL_4__AUX6_AUX_CTRL__SHIFT                                                              0x14
+#define DC_GPIO_AUX_CTRL_4__AUX1_AUX_CTRL_MASK                                                                0x0000000FL
+#define DC_GPIO_AUX_CTRL_4__AUX2_AUX_CTRL_MASK                                                                0x000000F0L
+#define DC_GPIO_AUX_CTRL_4__AUX3_AUX_CTRL_MASK                                                                0x00000F00L
+#define DC_GPIO_AUX_CTRL_4__AUX4_AUX_CTRL_MASK                                                                0x0000F000L
+#define DC_GPIO_AUX_CTRL_4__AUX5_AUX_CTRL_MASK                                                                0x000F0000L
+#define DC_GPIO_AUX_CTRL_4__AUX6_AUX_CTRL_MASK                                                                0x00F00000L
+//DC_GPIO_AUX_CTRL_5
+#define DC_GPIO_AUX_CTRL_5__AUX1_VOD_TUNE__SHIFT                                                              0x0
+#define DC_GPIO_AUX_CTRL_5__AUX2_VOD_TUNE__SHIFT                                                              0x2
+#define DC_GPIO_AUX_CTRL_5__AUX3_VOD_TUNE__SHIFT                                                              0x4
+#define DC_GPIO_AUX_CTRL_5__AUX4_VOD_TUNE__SHIFT                                                              0x6
+#define DC_GPIO_AUX_CTRL_5__AUX5_VOD_TUNE__SHIFT                                                              0x8
+#define DC_GPIO_AUX_CTRL_5__AUX6_VOD_TUNE__SHIFT                                                              0xa
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD1_I2CMODE__SHIFT                                                           0xc
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD2_I2CMODE__SHIFT                                                           0xd
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD3_I2CMODE__SHIFT                                                           0xe
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD4_I2CMODE__SHIFT                                                           0xf
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD5_I2CMODE__SHIFT                                                           0x10
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD6_I2CMODE__SHIFT                                                           0x11
+#define DC_GPIO_AUX_CTRL_5__DDC1_I2C_VPH_1V2_EN__SHIFT                                                        0x12
+#define DC_GPIO_AUX_CTRL_5__DDC2_I2C_VPH_1V2_EN__SHIFT                                                        0x13
+#define DC_GPIO_AUX_CTRL_5__DDC3_I2C_VPH_1V2_EN__SHIFT                                                        0x14
+#define DC_GPIO_AUX_CTRL_5__DDC4_I2C_VPH_1V2_EN__SHIFT                                                        0x15
+#define DC_GPIO_AUX_CTRL_5__DDC5_I2C_VPH_1V2_EN__SHIFT                                                        0x16
+#define DC_GPIO_AUX_CTRL_5__DDC6_I2C_VPH_1V2_EN__SHIFT                                                        0x17
+#define DC_GPIO_AUX_CTRL_5__DDC1_PAD_I2C_CTRL__SHIFT                                                          0x18
+#define DC_GPIO_AUX_CTRL_5__DDC2_PAD_I2C_CTRL__SHIFT                                                          0x19
+#define DC_GPIO_AUX_CTRL_5__DDC3_PAD_I2C_CTRL__SHIFT                                                          0x1a
+#define DC_GPIO_AUX_CTRL_5__DDC4_PAD_I2C_CTRL__SHIFT                                                          0x1b
+#define DC_GPIO_AUX_CTRL_5__DDC5_PAD_I2C_CTRL__SHIFT                                                          0x1c
+#define DC_GPIO_AUX_CTRL_5__DDC6_PAD_I2C_CTRL__SHIFT                                                          0x1d
+#define DC_GPIO_AUX_CTRL_5__AUX1_VOD_TUNE_MASK                                                                0x00000003L
+#define DC_GPIO_AUX_CTRL_5__AUX2_VOD_TUNE_MASK                                                                0x0000000CL
+#define DC_GPIO_AUX_CTRL_5__AUX3_VOD_TUNE_MASK                                                                0x00000030L
+#define DC_GPIO_AUX_CTRL_5__AUX4_VOD_TUNE_MASK                                                                0x000000C0L
+#define DC_GPIO_AUX_CTRL_5__AUX5_VOD_TUNE_MASK                                                                0x00000300L
+#define DC_GPIO_AUX_CTRL_5__AUX6_VOD_TUNE_MASK                                                                0x00000C00L
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD1_I2CMODE_MASK                                                             0x00001000L
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD2_I2CMODE_MASK                                                             0x00002000L
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD3_I2CMODE_MASK                                                             0x00004000L
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD4_I2CMODE_MASK                                                             0x00008000L
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD5_I2CMODE_MASK                                                             0x00010000L
+#define DC_GPIO_AUX_CTRL_5__DDC_PAD6_I2CMODE_MASK                                                             0x00020000L
+#define DC_GPIO_AUX_CTRL_5__DDC1_I2C_VPH_1V2_EN_MASK                                                          0x00040000L
+#define DC_GPIO_AUX_CTRL_5__DDC2_I2C_VPH_1V2_EN_MASK                                                          0x00080000L
+#define DC_GPIO_AUX_CTRL_5__DDC3_I2C_VPH_1V2_EN_MASK                                                          0x00100000L
+#define DC_GPIO_AUX_CTRL_5__DDC4_I2C_VPH_1V2_EN_MASK                                                          0x00200000L
+#define DC_GPIO_AUX_CTRL_5__DDC5_I2C_VPH_1V2_EN_MASK                                                          0x00400000L
+#define DC_GPIO_AUX_CTRL_5__DDC6_I2C_VPH_1V2_EN_MASK                                                          0x00800000L
+#define DC_GPIO_AUX_CTRL_5__DDC1_PAD_I2C_CTRL_MASK                                                            0x01000000L
+#define DC_GPIO_AUX_CTRL_5__DDC2_PAD_I2C_CTRL_MASK                                                            0x02000000L
+#define DC_GPIO_AUX_CTRL_5__DDC3_PAD_I2C_CTRL_MASK                                                            0x04000000L
+#define DC_GPIO_AUX_CTRL_5__DDC4_PAD_I2C_CTRL_MASK                                                            0x08000000L
+#define DC_GPIO_AUX_CTRL_5__DDC5_PAD_I2C_CTRL_MASK                                                            0x10000000L
+#define DC_GPIO_AUX_CTRL_5__DDC6_PAD_I2C_CTRL_MASK                                                            0x20000000L
+//AUXI2C_PAD_ALL_PWR_OK
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY1_ALL_PWR_OK__SHIFT                                                  0x0
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY2_ALL_PWR_OK__SHIFT                                                  0x1
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY3_ALL_PWR_OK__SHIFT                                                  0x2
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY4_ALL_PWR_OK__SHIFT                                                  0x3
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY5_ALL_PWR_OK__SHIFT                                                  0x4
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY6_ALL_PWR_OK__SHIFT                                                  0x5
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY1_ALL_PWR_OK_MASK                                                    0x00000001L
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY2_ALL_PWR_OK_MASK                                                    0x00000002L
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY3_ALL_PWR_OK_MASK                                                    0x00000004L
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY4_ALL_PWR_OK_MASK                                                    0x00000008L
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY5_ALL_PWR_OK_MASK                                                    0x00000010L
+#define AUXI2C_PAD_ALL_PWR_OK__AUXI2C_PHY6_ALL_PWR_OK_MASK                                                    0x00000020L
+//DC_GPIO_PULLUPEN
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICA_PU_EN__SHIFT                                                       0x0
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICB_PU_EN__SHIFT                                                       0x1
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICC_PU_EN__SHIFT                                                       0x2
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICD_PU_EN__SHIFT                                                       0x3
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICE_PU_EN__SHIFT                                                       0x4
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICF_PU_EN__SHIFT                                                       0x5
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICG_PU_EN__SHIFT                                                       0x6
+#define DC_GPIO_PULLUPEN__DC_GPIO_HSYNCA_PU_EN__SHIFT                                                         0x8
+#define DC_GPIO_PULLUPEN__DC_GPIO_VSYNCA_PU_EN__SHIFT                                                         0x9
+#define DC_GPIO_PULLUPEN__DC_GPIO_HPD1_PU_EN__SHIFT                                                           0xe
+#define DC_GPIO_PULLUPEN__DC_GPIO_BLON_PU_EN__SHIFT                                                           0x14
+#define DC_GPIO_PULLUPEN__DC_GPIO_DIGON_PU_EN__SHIFT                                                          0x15
+#define DC_GPIO_PULLUPEN__DC_GPIO_ENA_BL_PU_EN__SHIFT                                                         0x16
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICA_PU_EN_MASK                                                         0x00000001L
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICB_PU_EN_MASK                                                         0x00000002L
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICC_PU_EN_MASK                                                         0x00000004L
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICD_PU_EN_MASK                                                         0x00000008L
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICE_PU_EN_MASK                                                         0x00000010L
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICF_PU_EN_MASK                                                         0x00000020L
+#define DC_GPIO_PULLUPEN__DC_GPIO_GENERICG_PU_EN_MASK                                                         0x00000040L
+#define DC_GPIO_PULLUPEN__DC_GPIO_HSYNCA_PU_EN_MASK                                                           0x00000100L
+#define DC_GPIO_PULLUPEN__DC_GPIO_VSYNCA_PU_EN_MASK                                                           0x00000200L
+#define DC_GPIO_PULLUPEN__DC_GPIO_HPD1_PU_EN_MASK                                                             0x00004000L
+#define DC_GPIO_PULLUPEN__DC_GPIO_BLON_PU_EN_MASK                                                             0x00100000L
+#define DC_GPIO_PULLUPEN__DC_GPIO_DIGON_PU_EN_MASK                                                            0x00200000L
+#define DC_GPIO_PULLUPEN__DC_GPIO_ENA_BL_PU_EN_MASK                                                           0x00400000L
+//DC_GPIO_AUX_CTRL_6
+#define DC_GPIO_AUX_CTRL_6__AUX1_PAD_RXSEL__SHIFT                                                             0x0
+#define DC_GPIO_AUX_CTRL_6__AUX2_PAD_RXSEL__SHIFT                                                             0x2
+#define DC_GPIO_AUX_CTRL_6__AUX3_PAD_RXSEL__SHIFT                                                             0x4
+#define DC_GPIO_AUX_CTRL_6__AUX4_PAD_RXSEL__SHIFT                                                             0x6
+#define DC_GPIO_AUX_CTRL_6__AUX5_PAD_RXSEL__SHIFT                                                             0x8
+#define DC_GPIO_AUX_CTRL_6__AUX6_PAD_RXSEL__SHIFT                                                             0xa
+#define DC_GPIO_AUX_CTRL_6__AUX1_PAD_RXSEL_MASK                                                               0x00000003L
+#define DC_GPIO_AUX_CTRL_6__AUX2_PAD_RXSEL_MASK                                                               0x0000000CL
+#define DC_GPIO_AUX_CTRL_6__AUX3_PAD_RXSEL_MASK                                                               0x00000030L
+#define DC_GPIO_AUX_CTRL_6__AUX4_PAD_RXSEL_MASK                                                               0x000000C0L
+#define DC_GPIO_AUX_CTRL_6__AUX5_PAD_RXSEL_MASK                                                               0x00000300L
+#define DC_GPIO_AUX_CTRL_6__AUX6_PAD_RXSEL_MASK                                                               0x00000C00L
 //BPHYC_DAC_MACRO_CNTL
 #define BPHYC_DAC_MACRO_CNTL__BPHYC_DAC_WHITE_LEVEL__SHIFT                                                    0x0
 #define BPHYC_DAC_MACRO_CNTL__BPHYC_DAC_WHITE_FINE_CONTROL__SHIFT                                             0x8
diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_offset.h
index 4ccf9681c45d..721c61171045 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_offset.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_offset.h
@@ -3895,6 +3895,10 @@
 #define mmCM0_CM_MEM_PWR_CTRL_BASE_IDX                                                                 2
 #define mmCM0_CM_MEM_PWR_STATUS                                                                        0x0d33
 #define mmCM0_CM_MEM_PWR_STATUS_BASE_IDX                                                               2
+#define mmCM0_CM_TEST_DEBUG_INDEX                                                                      0x0d35
+#define mmCM0_CM_TEST_DEBUG_INDEX_BASE_IDX                                                             2
+#define mmCM0_CM_TEST_DEBUG_DATA                                                                       0x0d36
+#define mmCM0_CM_TEST_DEBUG_DATA_BASE_IDX                                                              2
 
 
 // addressBlock: dce_dc_dpp0_dispdec_dpp_dcperfmon_dc_perfmon_dispdec
@@ -4367,7 +4371,10 @@
 #define mmCM1_CM_MEM_PWR_CTRL_BASE_IDX                                                                 2
 #define mmCM1_CM_MEM_PWR_STATUS                                                                        0x0e4e
 #define mmCM1_CM_MEM_PWR_STATUS_BASE_IDX                                                               2
-
+#define mmCM1_CM_TEST_DEBUG_INDEX                                                                      0x0e50
+#define mmCM1_CM_TEST_DEBUG_INDEX_BASE_IDX                                                             2
+#define mmCM1_CM_TEST_DEBUG_DATA                                                                       0x0e51
+#define mmCM1_CM_TEST_DEBUG_DATA_BASE_IDX                                                              2
 
 // addressBlock: dce_dc_dpp1_dispdec_dpp_dcperfmon_dc_perfmon_dispdec
 // base address: 0x399c
@@ -4839,7 +4846,10 @@
 #define mmCM2_CM_MEM_PWR_CTRL_BASE_IDX                                                                 2
 #define mmCM2_CM_MEM_PWR_STATUS                                                                        0x0f69
 #define mmCM2_CM_MEM_PWR_STATUS_BASE_IDX                                                               2
-
+#define mmCM2_CM_TEST_DEBUG_INDEX                                                                      0x0f6b
+#define mmCM2_CM_TEST_DEBUG_INDEX_BASE_IDX                                                             2
+#define mmCM2_CM_TEST_DEBUG_DATA                                                                       0x0f6c
+#define mmCM2_CM_TEST_DEBUG_DATA_BASE_IDX                                                              2
 
 // addressBlock: dce_dc_dpp2_dispdec_dpp_dcperfmon_dc_perfmon_dispdec
 // base address: 0x3e08
@@ -5311,7 +5321,10 @@
 #define mmCM3_CM_MEM_PWR_CTRL_BASE_IDX                                                                 2
 #define mmCM3_CM_MEM_PWR_STATUS                                                                        0x1084
 #define mmCM3_CM_MEM_PWR_STATUS_BASE_IDX                                                               2
-
+#define mmCM3_CM_TEST_DEBUG_INDEX                                                                      0x1086
+#define mmCM3_CM_TEST_DEBUG_INDEX_BASE_IDX                                                             2
+#define mmCM3_CM_TEST_DEBUG_DATA                                                                       0x1087
+#define mmCM3_CM_TEST_DEBUG_DATA_BASE_IDX                                                              2
 
 // addressBlock: dce_dc_dpp3_dispdec_dpp_dcperfmon_dc_perfmon_dispdec
 // base address: 0x4274
diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_sh_mask.h
index e2a2f114bd8e..e7c0cad41081 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_sh_mask.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_1_0_sh_mask.h
@@ -14049,6 +14049,14 @@
 #define CM0_CM_MEM_PWR_STATUS__RGAM_MEM_PWR_STATE__SHIFT                                                      0x2
 #define CM0_CM_MEM_PWR_STATUS__SHARED_MEM_PWR_STATE_MASK                                                      0x00000003L
 #define CM0_CM_MEM_PWR_STATUS__RGAM_MEM_PWR_STATE_MASK                                                        0x0000000CL
+//CM0_CM_TEST_DEBUG_INDEX
+#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_INDEX__SHIFT                                                   0x0
+#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_WRITE_EN__SHIFT                                                0x8
+#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_INDEX_MASK                                                     0x000000FFL
+#define CM0_CM_TEST_DEBUG_INDEX__CM_TEST_DEBUG_WRITE_EN_MASK                                                  0x00000100L
+//CM0_CM_TEST_DEBUG_DATA
+#define CM0_CM_TEST_DEBUG_DATA__CM_TEST_DEBUG_DATA__SHIFT                                                     0x0
+#define CM0_CM_TEST_DEBUG_DATA__CM_TEST_DEBUG_DATA_MASK                                                       0xFFFFFFFFL
 
 
 // addressBlock: dce_dc_dpp0_dispdec_dpp_dcperfmon_dc_perfmon_dispdec
diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_default.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_default.h
new file mode 100644
index 000000000000..9e19e723081b
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_default.h
@@ -0,0 +1,26 @@
+/*
+ * Copyright (C) 2018  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+#ifndef _df_1_7_DEFAULT_HEADER
+#define _df_1_7_DEFAULT_HEADER
+
+#define mmFabricConfigAccessControl_DEFAULT						0x00000000
+
+#endif
diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_offset.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_offset.h
new file mode 100644
index 000000000000..e6044e27a913
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_offset.h
@@ -0,0 +1,37 @@
+/*
+ * Copyright (C) 2018  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+#ifndef _df_1_7_OFFSET_HEADER
+#define _df_1_7_OFFSET_HEADER
+
+#define mmFabricConfigAccessControl									0x0410
+#define mmFabricConfigAccessControl_BASE_IDX								0
+
+#define mmDF_PIE_AON0_DfGlobalClkGater									0x00fc
+#define mmDF_PIE_AON0_DfGlobalClkGater_BASE_IDX								0
+
+#define mmDF_CS_AON0_DramBaseAddress0									0x0044
+#define mmDF_CS_AON0_DramBaseAddress0_BASE_IDX								0
+
+#define mmDF_CS_AON0_CoherentSlaveModeCtrlA0								0x0214
+#define mmDF_CS_AON0_CoherentSlaveModeCtrlA0_BASE_IDX							0
+
+
+#endif
diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_sh_mask.h
new file mode 100644
index 000000000000..a78c99480e2d
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_1_7_sh_mask.h
@@ -0,0 +1,52 @@
+/*
+ * Copyright (C) 2018  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+#ifndef _df_1_7_SH_MASK_HEADER
+#define _df_1_7_SH_MASK_HEADER
+
+/* FabricConfigAccessControl */
+#define FabricConfigAccessControl__CfgRegInstAccEn__SHIFT						0x0
+#define FabricConfigAccessControl__CfgRegInstAccRegLock__SHIFT						0x1
+#define FabricConfigAccessControl__CfgRegInstID__SHIFT							0x10
+#define FabricConfigAccessControl__CfgRegInstAccEn_MASK							0x00000001L
+#define FabricConfigAccessControl__CfgRegInstAccRegLock_MASK						0x00000002L
+#define FabricConfigAccessControl__CfgRegInstID_MASK							0x00FF0000L
+
+/* DF_PIE_AON0_DfGlobalClkGater */
+#define DF_PIE_AON0_DfGlobalClkGater__MGCGMode__SHIFT							0x0
+#define DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK							0x0000000FL
+
+/* DF_CS_AON0_DramBaseAddress0 */
+#define DF_CS_AON0_DramBaseAddress0__AddrRngVal__SHIFT							0x0
+#define DF_CS_AON0_DramBaseAddress0__LgcyMmioHoleEn__SHIFT						0x1
+#define DF_CS_AON0_DramBaseAddress0__IntLvNumChan__SHIFT						0x4
+#define DF_CS_AON0_DramBaseAddress0__IntLvAddrSel__SHIFT						0x8
+#define DF_CS_AON0_DramBaseAddress0__DramBaseAddr__SHIFT						0xc
+#define DF_CS_AON0_DramBaseAddress0__AddrRngVal_MASK							0x00000001L
+#define DF_CS_AON0_DramBaseAddress0__LgcyMmioHoleEn_MASK						0x00000002L
+#define DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK							0x000000F0L
+#define DF_CS_AON0_DramBaseAddress0__IntLvAddrSel_MASK							0x00000700L
+#define DF_CS_AON0_DramBaseAddress0__DramBaseAddr_MASK							0xFFFFF000L
+
+//DF_CS_AON0_CoherentSlaveModeCtrlA0
+#define DF_CS_AON0_CoherentSlaveModeCtrlA0__ForceParWrRMW__SHIFT					0x3
+#define DF_CS_AON0_CoherentSlaveModeCtrlA0__ForceParWrRMW_MASK						0x00000008L
+
+#endif
diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_default.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_default.h
new file mode 100644
index 000000000000..e58c207ac980
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_default.h
@@ -0,0 +1,26 @@
+/*
+ * Copyright (C) 2018  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+#ifndef _df_3_6_DEFAULT_HEADER
+#define _df_3_6_DEFAULT_HEADER
+
+#define mmFabricConfigAccessControl_DEFAULT						0x00000000
+
+#endif
diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_offset.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_offset.h
new file mode 100644
index 000000000000..a9575db8d7aa
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_offset.h
@@ -0,0 +1,33 @@
+/*
+ * Copyright (C) 2018  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+#ifndef _df_3_6_OFFSET_HEADER
+#define _df_3_6_OFFSET_HEADER
+
+#define mmFabricConfigAccessControl									0x0410
+#define mmFabricConfigAccessControl_BASE_IDX								0
+
+#define mmDF_PIE_AON0_DfGlobalClkGater									0x00fc
+#define mmDF_PIE_AON0_DfGlobalClkGater_BASE_IDX								0
+
+#define mmDF_CS_UMC_AON0_DramBaseAddress0								0x0044
+#define mmDF_CS_UMC_AON0_DramBaseAddress0_BASE_IDX							0
+
+#endif
diff --git a/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_sh_mask.h
new file mode 100644
index 000000000000..06fac509e987
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/asic_reg/df/df_3_6_sh_mask.h
@@ -0,0 +1,48 @@
+/*
+ * Copyright (C) 2018  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+#ifndef _df_3_6_SH_MASK_HEADER
+#define _df_3_6_SH_MASK_HEADER
+
+/* FabricConfigAccessControl */
+#define FabricConfigAccessControl__CfgRegInstAccEn__SHIFT						0x0
+#define FabricConfigAccessControl__CfgRegInstAccRegLock__SHIFT						0x1
+#define FabricConfigAccessControl__CfgRegInstID__SHIFT							0x10
+#define FabricConfigAccessControl__CfgRegInstAccEn_MASK							0x00000001L
+#define FabricConfigAccessControl__CfgRegInstAccRegLock_MASK						0x00000002L
+#define FabricConfigAccessControl__CfgRegInstID_MASK							0x00FF0000L
+
+/* DF_PIE_AON0_DfGlobalClkGater */
+#define DF_PIE_AON0_DfGlobalClkGater__MGCGMode__SHIFT							0x0
+#define DF_PIE_AON0_DfGlobalClkGater__MGCGMode_MASK							0x0000000FL
+
+/* DF_CS_AON0_DramBaseAddress0 */
+#define DF_CS_UMC_AON0_DramBaseAddress0__AddrRngVal__SHIFT						0x0
+#define DF_CS_UMC_AON0_DramBaseAddress0__LgcyMmioHoleEn__SHIFT						0x1
+#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvNumChan__SHIFT						0x2
+#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvAddrSel__SHIFT						0x9
+#define DF_CS_UMC_AON0_DramBaseAddress0__DramBaseAddr__SHIFT						0xc
+#define DF_CS_UMC_AON0_DramBaseAddress0__AddrRngVal_MASK						0x00000001L
+#define DF_CS_UMC_AON0_DramBaseAddress0__LgcyMmioHoleEn_MASK						0x00000002L
+#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvNumChan_MASK						0x0000003CL
+#define DF_CS_UMC_AON0_DramBaseAddress0__IntLvAddrSel_MASK						0x00000E00L
+#define DF_CS_UMC_AON0_DramBaseAddress0__DramBaseAddr_MASK						0xFFFFF000L
+
+#endif
diff --git a/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_offset.h b/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_offset.h
index 18a32477ed1d..fe0cbaade3c3 100644
--- a/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_offset.h
+++ b/drivers/gpu/drm/amd/include/asic_reg/vcn/vcn_1_0_offset.h
@@ -89,6 +89,8 @@
 #define mmUVD_JPEG_RB_SIZE_BASE_IDX                                                                    1
 #define mmUVD_JPEG_ADDR_CONFIG                                                                         0x021f
 #define mmUVD_JPEG_ADDR_CONFIG_BASE_IDX                                                                1
+#define mmUVD_JPEG_PITCH                                                                               0x0222
+#define mmUVD_JPEG_PITCH_BASE_IDX                                                                      1
 #define mmUVD_JPEG_GPCOM_CMD                                                                           0x022c
 #define mmUVD_JPEG_GPCOM_CMD_BASE_IDX                                                                  1
 #define mmUVD_JPEG_GPCOM_DATA0                                                                         0x022d
@@ -203,6 +205,8 @@
 #define mmUVD_RB_WPTR4_BASE_IDX                                                                        1
 #define mmUVD_JRBC_RB_RPTR                                                                             0x0457
 #define mmUVD_JRBC_RB_RPTR_BASE_IDX                                                                    1
+#define mmUVD_LMI_JPEG_VMID                                                                            0x045d
+#define mmUVD_LMI_JPEG_VMID_BASE_IDX                                                                   1
 #define mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH                                                            0x045e
 #define mmUVD_LMI_VCPU_CACHE_64BIT_BAR_HIGH_BASE_IDX                                                   1
 #define mmUVD_LMI_VCPU_CACHE_64BIT_BAR_LOW                                                             0x045f
@@ -231,6 +235,8 @@
 #define mmUVD_LMI_JRBC_IB_64BIT_BAR_HIGH_BASE_IDX                                                      1
 #define mmUVD_LMI_JRBC_IB_VMID                                                                         0x0507
 #define mmUVD_LMI_JRBC_IB_VMID_BASE_IDX                                                                1
+#define mmUVD_LMI_JRBC_RB_VMID                                                                         0x0508
+#define mmUVD_LMI_JRBC_RB_VMID_BASE_IDX                                                                1
 #define mmUVD_JRBC_RB_WPTR                                                                             0x0509
 #define mmUVD_JRBC_RB_WPTR_BASE_IDX                                                                    1
 #define mmUVD_JRBC_RB_CNTL                                                                             0x050a
@@ -239,6 +245,20 @@
 #define mmUVD_JRBC_IB_SIZE_BASE_IDX                                                                    1
 #define mmUVD_JRBC_LMI_SWAP_CNTL                                                                       0x050d
 #define mmUVD_JRBC_LMI_SWAP_CNTL_BASE_IDX                                                              1
+#define mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_LOW                                                         0x050e
+#define mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_LOW_BASE_IDX                                                1
+#define mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_HIGH                                                        0x050f
+#define mmUVD_LMI_JRBC_RB_MEM_WR_64BIT_BAR_HIGH_BASE_IDX                                               1
+#define mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_LOW                                                         0x0510
+#define mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_LOW_BASE_IDX                                                1
+#define mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_HIGH                                                        0x0511
+#define mmUVD_LMI_JRBC_RB_MEM_RD_64BIT_BAR_HIGH_BASE_IDX                                               1
+#define mmUVD_JRBC_RB_REF_DATA                                                                         0x0512
+#define mmUVD_JRBC_RB_REF_DATA_BASE_IDX                                                                1
+#define mmUVD_JRBC_RB_COND_RD_TIMER                                                                    0x0513
+#define mmUVD_JRBC_RB_COND_RD_TIMER_BASE_IDX                                                           1
+#define mmUVD_JRBC_EXTERNAL_REG_BASE                                                                   0x0517
+#define mmUVD_JRBC_EXTERNAL_REG_BASE_BASE_IDX                                                          1
 #define mmUVD_JRBC_SOFT_RESET                                                                          0x0519
 #define mmUVD_JRBC_SOFT_RESET_BASE_IDX                                                                 1
 #define mmUVD_JRBC_STATUS                                                                              0x051a
diff --git a/drivers/gpu/drm/amd/include/atombios.h b/drivers/gpu/drm/amd/include/atombios.h
index f696bbb643ef..7931502fa54f 100644
--- a/drivers/gpu/drm/amd/include/atombios.h
+++ b/drivers/gpu/drm/amd/include/atombios.h
@@ -632,6 +632,13 @@ typedef struct _COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_2
   ULONG ulReserved;
 }COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_2;
 
+typedef struct _COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_3
+{
+  COMPUTE_MEMORY_ENGINE_PLL_PARAMETERS_V4 ulClock;
+  USHORT  usMclk_fcw_frac;                  //fractional divider of fcw = usSclk_fcw_frac/65536
+  USHORT  usMclk_fcw_int;                   //integer divider of fcwc
+}COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_3;
+
 //Input parameter of DynamicMemorySettingsTable
 //when ATOM_COMPUTE_CLOCK_FREQ.ulComputeClockFlag = COMPUTE_MEMORY_PLL_PARAM
 typedef struct _DYNAMICE_MEMORY_SETTINGS_PARAMETER
diff --git a/drivers/gpu/drm/amd/include/atomfirmware.h b/drivers/gpu/drm/amd/include/atomfirmware.h
index 0f5ad54d3fd3..33b4de4ad66e 100644
--- a/drivers/gpu/drm/amd/include/atomfirmware.h
+++ b/drivers/gpu/drm/amd/include/atomfirmware.h
@@ -501,6 +501,32 @@ enum atom_cooling_solution_id{
   LIQUID_COOLING = 0x01
 };
 
+struct atom_firmware_info_v3_2 {
+  struct atom_common_table_header table_header;
+  uint32_t firmware_revision;
+  uint32_t bootup_sclk_in10khz;
+  uint32_t bootup_mclk_in10khz;
+  uint32_t firmware_capability;             // enum atombios_firmware_capability
+  uint32_t main_call_parser_entry;          /* direct address of main parser call in VBIOS binary. */
+  uint32_t bios_scratch_reg_startaddr;      // 1st bios scratch register dword address
+  uint16_t bootup_vddc_mv;
+  uint16_t bootup_vddci_mv;
+  uint16_t bootup_mvddc_mv;
+  uint16_t bootup_vddgfx_mv;
+  uint8_t  mem_module_id;
+  uint8_t  coolingsolution_id;              /*0: Air cooling; 1: Liquid cooling ... */
+  uint8_t  reserved1[2];
+  uint32_t mc_baseaddr_high;
+  uint32_t mc_baseaddr_low;
+  uint8_t  board_i2c_feature_id;            // enum of atom_board_i2c_feature_id_def
+  uint8_t  board_i2c_feature_gpio_id;       // i2c id find in gpio_lut data table gpio_id
+  uint8_t  board_i2c_feature_slave_addr;
+  uint8_t  reserved3;
+  uint16_t bootup_mvddq_mv;
+  uint16_t bootup_mvpp_mv;
+  uint32_t zfbstartaddrin16mb;
+  uint32_t reserved2[3];
+};
 
 /* 
   ***************************************************************************
@@ -1169,7 +1195,64 @@ struct  atom_gfx_info_v2_2
   uint32_t rlc_gpu_timer_refclk; 
 };
 
+struct  atom_gfx_info_v2_3 {
+  struct  atom_common_table_header  table_header;
+  uint8_t gfxip_min_ver;
+  uint8_t gfxip_max_ver;
+  uint8_t max_shader_engines;
+  uint8_t max_tile_pipes;
+  uint8_t max_cu_per_sh;
+  uint8_t max_sh_per_se;
+  uint8_t max_backends_per_se;
+  uint8_t max_texture_channel_caches;
+  uint32_t regaddr_cp_dma_src_addr;
+  uint32_t regaddr_cp_dma_src_addr_hi;
+  uint32_t regaddr_cp_dma_dst_addr;
+  uint32_t regaddr_cp_dma_dst_addr_hi;
+  uint32_t regaddr_cp_dma_command;
+  uint32_t regaddr_cp_status;
+  uint32_t regaddr_rlc_gpu_clock_32;
+  uint32_t rlc_gpu_timer_refclk;
+  uint8_t active_cu_per_sh;
+  uint8_t active_rb_per_se;
+  uint16_t gcgoldenoffset;
+  uint32_t rm21_sram_vmin_value;
+};
 
+struct  atom_gfx_info_v2_4 {
+  struct  atom_common_table_header  table_header;
+  uint8_t gfxip_min_ver;
+  uint8_t gfxip_max_ver;
+  uint8_t gc_num_se;
+  uint8_t max_tile_pipes;
+  uint8_t gc_num_cu_per_sh;
+  uint8_t gc_num_sh_per_se;
+  uint8_t gc_num_rb_per_se;
+  uint8_t gc_num_tccs;
+  uint32_t regaddr_cp_dma_src_addr;
+  uint32_t regaddr_cp_dma_src_addr_hi;
+  uint32_t regaddr_cp_dma_dst_addr;
+  uint32_t regaddr_cp_dma_dst_addr_hi;
+  uint32_t regaddr_cp_dma_command;
+  uint32_t regaddr_cp_status;
+  uint32_t regaddr_rlc_gpu_clock_32;
+  uint32_t rlc_gpu_timer_refclk;
+  uint8_t active_cu_per_sh;
+  uint8_t active_rb_per_se;
+  uint16_t gcgoldenoffset;
+  uint16_t gc_num_gprs;
+  uint16_t gc_gsprim_buff_depth;
+  uint16_t gc_parameter_cache_depth;
+  uint16_t gc_wave_size;
+  uint16_t gc_max_waves_per_simd;
+  uint16_t gc_lds_size;
+  uint8_t gc_num_max_gs_thds;
+  uint8_t gc_gs_table_depth;
+  uint8_t gc_double_offchip_lds_buffer;
+  uint8_t gc_max_scratch_slots_per_cu;
+  uint32_t sram_rm_fuses_val;
+  uint32_t sram_custom_rm_fuses_val;
+};
 
 /* 
   ***************************************************************************
@@ -1198,6 +1281,76 @@ struct atom_smu_info_v3_1
   uint8_t  fw_ctf_polarity;         // GPIO polarity for CTF
 };
 
+struct atom_smu_info_v3_2 {
+  struct   atom_common_table_header  table_header;
+  uint8_t  smuip_min_ver;
+  uint8_t  smuip_max_ver;
+  uint8_t  smu_rsd1;
+  uint8_t  gpuclk_ss_mode;
+  uint16_t sclk_ss_percentage;
+  uint16_t sclk_ss_rate_10hz;
+  uint16_t gpuclk_ss_percentage;    // in unit of 0.001%
+  uint16_t gpuclk_ss_rate_10hz;
+  uint32_t core_refclk_10khz;
+  uint8_t  ac_dc_gpio_bit;          // GPIO bit shift in SMU_GPIOPAD_A  configured for AC/DC switching, =0xff means invalid
+  uint8_t  ac_dc_polarity;          // GPIO polarity for AC/DC switching
+  uint8_t  vr0hot_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A  configured for VR0 HOT event, =0xff means invalid
+  uint8_t  vr0hot_polarity;         // GPIO polarity for VR0 HOT event
+  uint8_t  vr1hot_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A configured for VR1 HOT event , =0xff means invalid
+  uint8_t  vr1hot_polarity;         // GPIO polarity for VR1 HOT event
+  uint8_t  fw_ctf_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A configured for CTF, =0xff means invalid
+  uint8_t  fw_ctf_polarity;         // GPIO polarity for CTF
+  uint8_t  pcc_gpio_bit;            // GPIO bit shift in SMU_GPIOPAD_A configured for PCC, =0xff means invalid
+  uint8_t  pcc_gpio_polarity;       // GPIO polarity for CTF
+  uint16_t smugoldenoffset;
+  uint32_t gpupll_vco_freq_10khz;
+  uint32_t bootup_smnclk_10khz;
+  uint32_t bootup_socclk_10khz;
+  uint32_t bootup_mp0clk_10khz;
+  uint32_t bootup_mp1clk_10khz;
+  uint32_t bootup_lclk_10khz;
+  uint32_t bootup_dcefclk_10khz;
+  uint32_t ctf_threshold_override_value;
+  uint32_t reserved[5];
+};
+
+struct atom_smu_info_v3_3 {
+  struct   atom_common_table_header  table_header;
+  uint8_t  smuip_min_ver;
+  uint8_t  smuip_max_ver;
+  uint8_t  smu_rsd1;
+  uint8_t  gpuclk_ss_mode;
+  uint16_t sclk_ss_percentage;
+  uint16_t sclk_ss_rate_10hz;
+  uint16_t gpuclk_ss_percentage;    // in unit of 0.001%
+  uint16_t gpuclk_ss_rate_10hz;
+  uint32_t core_refclk_10khz;
+  uint8_t  ac_dc_gpio_bit;          // GPIO bit shift in SMU_GPIOPAD_A  configured for AC/DC switching, =0xff means invalid
+  uint8_t  ac_dc_polarity;          // GPIO polarity for AC/DC switching
+  uint8_t  vr0hot_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A  configured for VR0 HOT event, =0xff means invalid
+  uint8_t  vr0hot_polarity;         // GPIO polarity for VR0 HOT event
+  uint8_t  vr1hot_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A configured for VR1 HOT event , =0xff means invalid
+  uint8_t  vr1hot_polarity;         // GPIO polarity for VR1 HOT event
+  uint8_t  fw_ctf_gpio_bit;         // GPIO bit shift in SMU_GPIOPAD_A configured for CTF, =0xff means invalid
+  uint8_t  fw_ctf_polarity;         // GPIO polarity for CTF
+  uint8_t  pcc_gpio_bit;            // GPIO bit shift in SMU_GPIOPAD_A configured for PCC, =0xff means invalid
+  uint8_t  pcc_gpio_polarity;       // GPIO polarity for CTF
+  uint16_t smugoldenoffset;
+  uint32_t gpupll_vco_freq_10khz;
+  uint32_t bootup_smnclk_10khz;
+  uint32_t bootup_socclk_10khz;
+  uint32_t bootup_mp0clk_10khz;
+  uint32_t bootup_mp1clk_10khz;
+  uint32_t bootup_lclk_10khz;
+  uint32_t bootup_dcefclk_10khz;
+  uint32_t ctf_threshold_override_value;
+  uint32_t syspll3_0_vco_freq_10khz;
+  uint32_t syspll3_1_vco_freq_10khz;
+  uint32_t bootup_fclk_10khz;
+  uint32_t bootup_waflclk_10khz;
+  uint32_t reserved[3];
+};
+
 /*
  ***************************************************************************
    Data Table smc_dpm_info  structure
@@ -1280,9 +1433,11 @@ struct atom_smc_dpm_info_v4_1
 	uint8_t  acggfxclkspreadpercent;
 	uint16_t acggfxclkspreadfreq;
 
-	uint32_t boardreserved[10];
-};
+	uint8_t Vr2_I2C_address;
+	uint8_t padding_vr2[3];
 
+	uint32_t boardreserved[9];
+};
 
 /* 
   ***************************************************************************
@@ -1864,6 +2019,53 @@ enum atom_smu9_syspll0_clock_id
   SMU9_SYSPLL0_DISPCLK_ID  = 11,      //	DISPCLK
 };
 
+enum atom_smu11_syspll_id {
+  SMU11_SYSPLL0_ID            = 0,
+  SMU11_SYSPLL1_0_ID          = 1,
+  SMU11_SYSPLL1_1_ID          = 2,
+  SMU11_SYSPLL1_2_ID          = 3,
+  SMU11_SYSPLL2_ID            = 4,
+  SMU11_SYSPLL3_0_ID          = 5,
+  SMU11_SYSPLL3_1_ID          = 6,
+};
+
+enum atom_smu11_syspll0_clock_id {
+  SMU11_SYSPLL0_ECLK_ID     = 0,       //	ECLK
+  SMU11_SYSPLL0_SOCCLK_ID   = 1,       //	SOCCLK
+  SMU11_SYSPLL0_MP0CLK_ID   = 2,       //	MP0CLK
+  SMU11_SYSPLL0_DCLK_ID     = 3,       //	DCLK
+  SMU11_SYSPLL0_VCLK_ID     = 4,       //	VCLK
+  SMU11_SYSPLL0_DCEFCLK_ID  = 5,       //	DCEFCLK
+};
+
+enum atom_smu11_syspll1_0_clock_id {
+  SMU11_SYSPLL1_0_UCLKA_ID   = 0,       // UCLK_a
+};
+
+enum atom_smu11_syspll1_1_clock_id {
+  SMU11_SYSPLL1_0_UCLKB_ID   = 0,       // UCLK_b
+};
+
+enum atom_smu11_syspll1_2_clock_id {
+  SMU11_SYSPLL1_0_FCLK_ID   = 0,        // FCLK
+};
+
+enum atom_smu11_syspll2_clock_id {
+  SMU11_SYSPLL2_GFXCLK_ID   = 0,        // GFXCLK
+};
+
+enum atom_smu11_syspll3_0_clock_id {
+  SMU11_SYSPLL3_0_WAFCLK_ID = 0,       //	WAFCLK
+  SMU11_SYSPLL3_0_DISPCLK_ID = 1,      //	DISPCLK
+  SMU11_SYSPLL3_0_DPREFCLK_ID = 2,     //	DPREFCLK
+};
+
+enum atom_smu11_syspll3_1_clock_id {
+  SMU11_SYSPLL3_1_MP1CLK_ID = 0,       //	MP1CLK
+  SMU11_SYSPLL3_1_SMNCLK_ID = 1,       //	SMNCLK
+  SMU11_SYSPLL3_1_LCLK_ID = 2,         //	LCLK
+};
+
 struct  atom_get_smu_clock_info_output_parameters_v3_1
 {
   union {
diff --git a/drivers/gpu/drm/amd/include/cgs_common.h b/drivers/gpu/drm/amd/include/cgs_common.h
index f2814ae7ecdd..a69deb3a2ac0 100644
--- a/drivers/gpu/drm/amd/include/cgs_common.h
+++ b/drivers/gpu/drm/amd/include/cgs_common.h
@@ -42,20 +42,6 @@ enum cgs_ind_reg {
 	CGS_IND_REG__AUDIO_ENDPT
 };
 
-/**
- * enum cgs_engine - Engines that can be statically power-gated
- */
-enum cgs_engine {
-	CGS_ENGINE__UVD,
-	CGS_ENGINE__VCE,
-	CGS_ENGINE__VP8,
-	CGS_ENGINE__ACP_DMA,
-	CGS_ENGINE__ACP_DSP0,
-	CGS_ENGINE__ACP_DSP1,
-	CGS_ENGINE__ISP,
-	/* ... */
-};
-
 /*
  * enum cgs_ucode_id - Firmware types for different IPs
  */
@@ -76,17 +62,6 @@ enum cgs_ucode_id {
 	CGS_UCODE_ID_MAXIMUM,
 };
 
-/*
- * enum cgs_resource_type - GPU resource type
- */
-enum cgs_resource_type {
-	CGS_RESOURCE_TYPE_MMIO = 0,
-	CGS_RESOURCE_TYPE_FB,
-	CGS_RESOURCE_TYPE_IO,
-	CGS_RESOURCE_TYPE_DOORBELL,
-	CGS_RESOURCE_TYPE_ROM,
-};
-
 /**
  * struct cgs_firmware_info - Firmware information
  */
@@ -104,17 +79,6 @@ struct cgs_firmware_info {
 	bool			is_kicker;
 };
 
-struct cgs_mode_info {
-	uint32_t		refresh_rate;
-	uint32_t		vblank_time_us;
-};
-
-struct cgs_display_info {
-	uint32_t		display_count;
-	uint32_t		active_display_mask;
-	struct cgs_mode_info *mode_info;
-};
-
 typedef unsigned long cgs_handle_t;
 
 /**
@@ -170,119 +134,18 @@ typedef void (*cgs_write_ind_register_t)(struct cgs_device *cgs_device, enum cgs
 #define CGS_WREG32_FIELD_IND(device, space, reg, field, val)	\
 	cgs_write_ind_register(device, space, ix##reg, (cgs_read_ind_register(device, space, ix##reg) & ~CGS_REG_FIELD_MASK(reg, field)) | (val) << CGS_REG_FIELD_SHIFT(reg, field))
 
-/**
- * cgs_get_pci_resource() - provide access to a device resource (PCI BAR)
- * @cgs_device:	opaque device handle
- * @resource_type:	Type of Resource (MMIO, IO, ROM, FB, DOORBELL)
- * @size:	size of the region
- * @offset:	offset from the start of the region
- * @resource_base:	base address (not including offset) returned
- *
- * Return: 0 on success, -errno otherwise
- */
-typedef int (*cgs_get_pci_resource_t)(struct cgs_device *cgs_device,
-				      enum cgs_resource_type resource_type,
-				      uint64_t size,
-				      uint64_t offset,
-				      uint64_t *resource_base);
-
-/**
- * cgs_atom_get_data_table() - Get a pointer to an ATOM BIOS data table
- * @cgs_device:	opaque device handle
- * @table:	data table index
- * @size:	size of the table (output, may be NULL)
- * @frev:	table format revision (output, may be NULL)
- * @crev:	table content revision (output, may be NULL)
- *
- * Return: Pointer to start of the table, or NULL on failure
- */
-typedef const void *(*cgs_atom_get_data_table_t)(
-	struct cgs_device *cgs_device, unsigned table,
-	uint16_t *size, uint8_t *frev, uint8_t *crev);
-
-/**
- * cgs_atom_get_cmd_table_revs() - Get ATOM BIOS command table revisions
- * @cgs_device:	opaque device handle
- * @table:	data table index
- * @frev:	table format revision (output, may be NULL)
- * @crev:	table content revision (output, may be NULL)
- *
- * Return: 0 on success, -errno otherwise
- */
-typedef int (*cgs_atom_get_cmd_table_revs_t)(struct cgs_device *cgs_device, unsigned table,
-					     uint8_t *frev, uint8_t *crev);
-
-/**
- * cgs_atom_exec_cmd_table() - Execute an ATOM BIOS command table
- * @cgs_device: opaque device handle
- * @table:	command table index
- * @args:	arguments
- *
- * Return: 0 on success, -errno otherwise
- */
-typedef int (*cgs_atom_exec_cmd_table_t)(struct cgs_device *cgs_device,
-					 unsigned table, void *args);
-
-/**
- * cgs_get_firmware_info - Get the firmware information from core driver
- * @cgs_device: opaque device handle
- * @type: the firmware type
- * @info: returend firmware information
- *
- * Return: 0 on success, -errno otherwise
- */
 typedef int (*cgs_get_firmware_info)(struct cgs_device *cgs_device,
 				     enum cgs_ucode_id type,
 				     struct cgs_firmware_info *info);
 
-typedef int (*cgs_rel_firmware)(struct cgs_device *cgs_device,
-					 enum cgs_ucode_id type);
-
-typedef int(*cgs_set_powergating_state)(struct cgs_device *cgs_device,
-				  enum amd_ip_block_type block_type,
-				  enum amd_powergating_state state);
-
-typedef int(*cgs_set_clockgating_state)(struct cgs_device *cgs_device,
-				  enum amd_ip_block_type block_type,
-				  enum amd_clockgating_state state);
-
-typedef int(*cgs_get_active_displays_info)(
-					struct cgs_device *cgs_device,
-					struct cgs_display_info *info);
-
-typedef int (*cgs_notify_dpm_enabled)(struct cgs_device *cgs_device, bool enabled);
-
-typedef int (*cgs_is_virtualization_enabled_t)(void *cgs_device);
-
-typedef int (*cgs_enter_safe_mode)(struct cgs_device *cgs_device, bool en);
-
-typedef void (*cgs_lock_grbm_idx)(struct cgs_device *cgs_device, bool lock);
-
 struct cgs_ops {
 	/* MMIO access */
 	cgs_read_register_t read_register;
 	cgs_write_register_t write_register;
 	cgs_read_ind_register_t read_ind_register;
 	cgs_write_ind_register_t write_ind_register;
-	/* PCI resources */
-	cgs_get_pci_resource_t get_pci_resource;
-	/* ATOM BIOS */
-	cgs_atom_get_data_table_t atom_get_data_table;
-	cgs_atom_get_cmd_table_revs_t atom_get_cmd_table_revs;
-	cgs_atom_exec_cmd_table_t atom_exec_cmd_table;
 	/* Firmware Info */
 	cgs_get_firmware_info get_firmware_info;
-	cgs_rel_firmware rel_firmware;
-	/* cg pg interface*/
-	cgs_set_powergating_state set_powergating_state;
-	cgs_set_clockgating_state set_clockgating_state;
-	/* display manager */
-	cgs_get_active_displays_info get_active_displays_info;
-	/* notify dpm enabled */
-	cgs_notify_dpm_enabled notify_dpm_enabled;
-	cgs_is_virtualization_enabled_t is_virtualization_enabled;
-	cgs_enter_safe_mode enter_safe_mode;
-	cgs_lock_grbm_idx lock_grbm_idx;
 };
 
 struct cgs_os_ops; /* To be define in OS-specific CGS header */
@@ -309,40 +172,7 @@ struct cgs_device
 #define cgs_write_ind_register(dev,space,index,value)		\
 	CGS_CALL(write_ind_register,dev,space,index,value)
 
-#define cgs_atom_get_data_table(dev,table,size,frev,crev)	\
-	CGS_CALL(atom_get_data_table,dev,table,size,frev,crev)
-#define cgs_atom_get_cmd_table_revs(dev,table,frev,crev)	\
-	CGS_CALL(atom_get_cmd_table_revs,dev,table,frev,crev)
-#define cgs_atom_exec_cmd_table(dev,table,args)		\
-	CGS_CALL(atom_exec_cmd_table,dev,table,args)
-
 #define cgs_get_firmware_info(dev, type, info)	\
 	CGS_CALL(get_firmware_info, dev, type, info)
-#define cgs_rel_firmware(dev, type)	\
-	CGS_CALL(rel_firmware, dev, type)
-#define cgs_set_powergating_state(dev, block_type, state)	\
-	CGS_CALL(set_powergating_state, dev, block_type, state)
-#define cgs_set_clockgating_state(dev, block_type, state)	\
-	CGS_CALL(set_clockgating_state, dev, block_type, state)
-#define cgs_notify_dpm_enabled(dev, enabled)	\
-	CGS_CALL(notify_dpm_enabled, dev, enabled)
-
-#define cgs_get_active_displays_info(dev, info)	\
-	CGS_CALL(get_active_displays_info, dev, info)
-
-#define cgs_get_pci_resource(cgs_device, resource_type, size, offset, \
-	resource_base) \
-	CGS_CALL(get_pci_resource, cgs_device, resource_type, size, offset, \
-	resource_base)
-
-#define cgs_is_virtualization_enabled(cgs_device) \
-		CGS_CALL(is_virtualization_enabled, cgs_device)
-
-#define cgs_enter_safe_mode(cgs_device, en) \
-		CGS_CALL(enter_safe_mode, cgs_device, en)
-
-#define cgs_lock_grbm_idx(cgs_device, lock) \
-		CGS_CALL(lock_grbm_idx, cgs_device, lock)
-
 
 #endif /* _CGS_COMMON_H */
diff --git a/drivers/gpu/drm/amd/include/dm_pp_interface.h b/drivers/gpu/drm/amd/include/dm_pp_interface.h
index 7852952d1fde..1d93a0c574c9 100644
--- a/drivers/gpu/drm/amd/include/dm_pp_interface.h
+++ b/drivers/gpu/drm/amd/include/dm_pp_interface.h
@@ -23,6 +23,8 @@
 #ifndef _DM_PP_INTERFACE_
 #define _DM_PP_INTERFACE_
 
+#include "dm_services_types.h"
+
 #define PP_MAX_CLOCK_LEVELS 16
 
 enum amd_pp_display_config_type{
@@ -189,39 +191,4 @@ struct pp_display_clock_request {
 	uint32_t clock_freq_in_khz;
 };
 
-#define PP_MAX_WM_SETS 4
-
-enum pp_wm_set_id {
-	DC_WM_SET_A = 0,
-	DC_WM_SET_B,
-	DC_WM_SET_C,
-	DC_WM_SET_D,
-	DC_WM_SET_INVALID = 0xffff,
-};
-
-struct pp_wm_set_with_dmif_clock_range_soc15 {
-	enum pp_wm_set_id wm_set_id;
-	uint32_t wm_min_dcefclk_in_khz;
-	uint32_t wm_max_dcefclk_in_khz;
-	uint32_t wm_min_memclk_in_khz;
-	uint32_t wm_max_memclk_in_khz;
-};
-
-struct pp_wm_set_with_mcif_clock_range_soc15 {
-	enum pp_wm_set_id wm_set_id;
-	uint32_t wm_min_socclk_in_khz;
-	uint32_t wm_max_socclk_in_khz;
-	uint32_t wm_min_memclk_in_khz;
-	uint32_t wm_max_memclk_in_khz;
-};
-
-struct pp_wm_sets_with_clock_ranges_soc15 {
-	uint32_t num_wm_sets_dmif;
-	uint32_t num_wm_sets_mcif;
-	struct pp_wm_set_with_dmif_clock_range_soc15
-		wm_sets_dmif[PP_MAX_WM_SETS];
-	struct pp_wm_set_with_mcif_clock_range_soc15
-		wm_sets_mcif[PP_MAX_WM_SETS];
-};
-
 #endif /* _DM_PP_INTERFACE_ */
diff --git a/drivers/gpu/drm/amd/include/ivsrcid/gfx/irqsrcs_gfx_9_0.h b/drivers/gpu/drm/amd/include/ivsrcid/gfx/irqsrcs_gfx_9_0.h
new file mode 100644
index 000000000000..36306c57a2b4
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/ivsrcid/gfx/irqsrcs_gfx_9_0.h
@@ -0,0 +1,55 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __IRQSRCS_GFX_9_0_H__
+#define __IRQSRCS_GFX_9_0_H__
+
+
+#define GFX_9_0__SRCID__CP_RB_INTERRUPT_PKT					176		/* B0 CP_INTERRUPT pkt in RB */
+#define GFX_9_0__SRCID__CP_IB1_INTERRUPT_PKT				177		/* B1 CP_INTERRUPT pkt in IB1 */
+#define GFX_9_0__SRCID__CP_IB2_INTERRUPT_PKT				178		/* B2 CP_INTERRUPT pkt in IB2 */
+#define GFX_9_0__SRCID__CP_PM4_PKT_RSVD_BIT_ERROR			180		/* B4 PM4 Pkt Rsvd Bits Error */
+#define GFX_9_0__SRCID__CP_EOP_INTERRUPT					181		/* B5 End-of-Pipe Interrupt */
+#define GFX_9_0__SRCID__CP_BAD_OPCODE_ERROR					183		/* B7 Bad Opcode Error */
+#define GFX_9_0__SRCID__CP_PRIV_REG_FAULT					184		/* B8 Privileged Register Fault */
+#define GFX_9_0__SRCID__CP_PRIV_INSTR_FAULT					185		/* B9 Privileged Instr Fault */
+#define GFX_9_0__SRCID__CP_WAIT_MEM_SEM_FAULT				186		/* BA Wait Memory Semaphore Fault (Synchronization Object Fault) */
+#define GFX_9_0__SRCID__CP_CTX_EMPTY_INTERRUPT				187		/* BB Context Empty Interrupt */
+#define GFX_9_0__SRCID__CP_CTX_BUSY_INTERRUPT				188		/* BC Context Busy Interrupt */
+#define GFX_9_0__SRCID__CP_ME_WAIT_REG_MEM_POLL_TIMEOUT		192		/* C0 CP.ME Wait_Reg_Mem Poll Timeout */
+#define GFX_9_0__SRCID__CP_SIG_INCOMPLETE					193		/* C1 "Surface Probe Fault Signal Incomplete" */
+#define GFX_9_0__SRCID__CP_PREEMPT_ACK					    194		/* C2 Preemption Ack-wledge */
+#define GFX_9_0__SRCID__CP_GPF					            195		/* C3 General Protection Fault (GPF) */
+#define GFX_9_0__SRCID__CP_GDS_ALLOC_ERROR					196		/* C4 GDS Alloc Error */
+#define GFX_9_0__SRCID__CP_ECC_ERROR					    197		/* C5 ECC  Error */
+#define GFX_9_0__SRCID__CP_COMPUTE_QUERY_STATUS             199     /* C7 Compute query status */
+#define GFX_9_0__SRCID__CP_VM_DOORBELL					    200		/* C8 Unattached VM Doorbell Received */
+#define GFX_9_0__SRCID__CP_FUE_ERROR					    201		/* C9 ECC FUE Error */
+#define GFX_9_0__SRCID__RLC_STRM_PERF_MONITOR_INTERRUPT		202		/* CA Streaming Perf Monitor Interrupt */
+#define GFX_9_0__SRCID__GRBM_RD_TIMEOUT_ERROR				232		/* E8 CRead timeout error */
+#define GFX_9_0__SRCID__GRBM_REG_GUI_IDLE					233		/* E9 Register GUI Idle */
+#define GFX_9_0__SRCID__SQ_INTERRUPT_ID					    239		/* EF SQ Interrupt (ttrace wrap, errors) */
+
+#endif /* __IRQSRCS_GFX_9_0_H__ */
diff --git a/drivers/gpu/drm/amd/include/ivsrcid/ivsrcid_vislands30.h b/drivers/gpu/drm/amd/include/ivsrcid/ivsrcid_vislands30.h
index c6b6f97de9de..aaed7f59e0e2 100644
--- a/drivers/gpu/drm/amd/include/ivsrcid/ivsrcid_vislands30.h
+++ b/drivers/gpu/drm/amd/include/ivsrcid/ivsrcid_vislands30.h
@@ -198,4 +198,102 @@
 #define VISLANDS30_IV_SRCID_HPD_RX_F		                    42	    // 0x2a		
 #define VISLANDS30_IV_EXTID_HPD_RX_F                         11
 
+#define VISLANDS30_IV_SRCID_GPIO_19                            0x00000053  /* 83 */
+
+#define VISLANDS30_IV_SRCID_SRBM_READ_TIMEOUT_ERR              0x00000060  /* 96 */
+#define VISLANDS30_IV_SRCID_SRBM_CTX_SWITCH                    0x00000061  /* 97 */
+
+#define VISLANDS30_IV_SRBM_REG_ACCESS_ERROR                    0x00000062  /* 98 */
+
+
+#define VISLANDS30_IV_SRCID_UVD_ENC_GEN_PURP                   0x00000077  /* 119 */
+#define VISLANDS30_IV_SRCID_UVD_SYSTEM_MESSAGE                 0x0000007c  /* 124 */
+
+#define VISLANDS30_IV_SRCID_BIF_PF_VF_MSGBUF_VALID             0x00000087  /* 135 */
+
+#define VISLANDS30_IV_SRCID_BIF_VF_PF_MSGBUF_ACK               0x0000008a  /* 138 */
+
+#define VISLANDS30_IV_SRCID_SYS_PAGE_INV_FAULT                 0x0000008c  /* 140 */
+#define VISLANDS30_IV_SRCID_SYS_MEM_PROT_FAULT                 0x0000008d  /* 141 */
+
+#define VISLANDS30_IV_SRCID_SEM_PAGE_INV_FAULT                 0x00000090  /* 144 */
+#define VISLANDS30_IV_SRCID_SEM_MEM_PROT_FAULT                 0x00000091  /* 145 */
+
+#define VISLANDS30_IV_SRCID_GFX_PAGE_INV_FAULT                 0x00000092  /* 146 */
+#define VISLANDS30_IV_SRCID_GFX_MEM_PROT_FAULT                 0x00000093  /* 147 */
+
+#define VISLANDS30_IV_SRCID_ACP                                0x000000a2  /* 162 */
+
+#define VISLANDS30_IV_SRCID_VCE_TRAP                           0x000000a7  /* 167 */
+#define VISLANDS30_IV_EXTID_VCE_TRAP_GENERAL_PURPOSE           0
+#define VISLANDS30_IV_EXTID_VCE_TRAP_LOW_LATENCY               1
+#define VISLANDS30_IV_EXTID_VCE_TRAP_REAL_TIME                 2
+
+#define VISLANDS30_IV_SRCID_CP_INT_RB                          0x000000b0  /* 176 */
+#define VISLANDS30_IV_SRCID_CP_INT_IB1                         0x000000b1  /* 177 */
+#define VISLANDS30_IV_SRCID_CP_INT_IB2                         0x000000b2  /* 178 */
+#define VISLANDS30_IV_SRCID_CP_PM4_RES_BITS_ERR                0x000000b4  /* 180 */
+#define VISLANDS30_IV_SRCID_CP_END_OF_PIPE                     0x000000b5  /* 181 */
+#define VISLANDS30_IV_SRCID_CP_BAD_OPCODE                      0x000000b7  /* 183 */
+#define VISLANDS30_IV_SRCID_CP_PRIV_REG_FAULT                  0x000000b8  /* 184 */
+#define VISLANDS30_IV_SRCID_CP_PRIV_INSTR_FAULT                0x000000b9  /* 185 */
+#define VISLANDS30_IV_SRCID_CP_WAIT_MEM_SEM_FAULT              0x000000ba  /* 186 */
+#define VISLANDS30_IV_SRCID_CP_GUI_IDLE                        0x000000bb  /* 187 */
+#define VISLANDS30_IV_SRCID_CP_GUI_BUSY                        0x000000bc  /* 188 */
+
+#define VISLANDS30_IV_SRCID_CP_COMPUTE_QUERY_STATUS            0x000000bf  /* 191 */
+#define VISLANDS30_IV_SRCID_CP_ECC_ERROR                       0x000000c5  /* 197 */
+
+#define CARRIZO_IV_SRCID_CP_COMPUTE_QUERY_STATUS               0x000000c7  /* 199 */
+
+#define VISLANDS30_IV_SRCID_CP_WAIT_REG_MEM_POLL_TIMEOUT       0x000000c0  /* 192 */
+#define VISLANDS30_IV_SRCID_CP_SEM_SIG_INCOMPL                 0x000000c1  /* 193 */
+#define VISLANDS30_IV_SRCID_CP_PREEMPT_ACK                     0x000000c2  /* 194 */
+#define VISLANDS30_IV_SRCID_CP_GENERAL_PROT_FAULT              0x000000c3  /* 195 */
+#define VISLANDS30_IV_SRCID_CP_GDS_ALLOC_ERROR                 0x000000c4  /* 196 */
+#define VISLANDS30_IV_SRCID_CP_ECC_ERROR                       0x000000c5  /* 197 */
+
+#define VISLANDS30_IV_SRCID_RLC_STRM_PERF_MONITOR              0x000000ca  /* 202 */
+
+#define VISLANDS30_IV_SDMA_ATOMIC_SRC_ID                       0x000000da  /* 218 */
+
+#define VISLANDS30_IV_SRCID_SDMA_ECC_ERROR                     0x000000dc  /* 220 */
+
+#define VISLANDS30_IV_SRCID_SDMA_TRAP          	               0x000000e0  /* 224 */
+#define VISLANDS30_IV_SRCID_SDMA_SEM_INCOMPLETE                0x000000e1  /* 225 */
+#define VISLANDS30_IV_SRCID_SDMA_SEM_WAIT                      0x000000e2  /* 226 */
+
+
+#define VISLANDS30_IV_SRCID_SMU_DISP_TIMER2_TRIGGER            0x000000e5  /* 229 */
+
+#define VISLANDS30_IV_SRCID_CG_TSS_THERMAL_LOW_TO_HIGH         0x000000e6  /* 230 */
+#define VISLANDS30_IV_SRCID_CG_TSS_THERMAL_HIGH_TO_LOW         0x000000e7  /* 231 */
+
+#define VISLANDS30_IV_SRCID_GRBM_READ_TIMEOUT_ERR              0x000000e8  /* 232 */
+#define VISLANDS30_IV_SRCID_GRBM_REG_GUI_IDLE                  0x000000e9  /* 233 */
+
+#define VISLANDS30_IV_SRCID_SQ_INTERRUPT_MSG                   0x000000ef  /* 239 */
+
+#define VISLANDS30_IV_SRCID_SDMA_PREEMPT                       0x000000f0  /* 240 */
+#define VISLANDS30_IV_SRCID_SDMA_VM_HOLE                       0x000000f2  /* 242 */
+#define VISLANDS30_IV_SRCID_SDMA_CTXEMPTY                      0x000000f3  /* 243 */
+#define VISLANDS30_IV_SRCID_SDMA_DOORBELL_INVALID              0x000000f4  /* 244 */
+#define VISLANDS30_IV_SRCID_SDMA_FROZEN                        0x000000f5  /* 245 */
+#define VISLANDS30_IV_SRCID_SDMA_POLL_TIMEOUT                  0x000000f6  /* 246 */
+#define VISLANDS30_IV_SRCID_SDMA_SRBM_WRITE                    0x000000f7  /* 247 */
+
+#define VISLANDS30_IV_SRCID_CG_THERMAL_TRIG                    0x000000f8  /* 248 */
+
+#define VISLANDS30_IV_SRCID_SMU_DISP_TIMER_TRIGGER             0x000000fd  /* 253 */
+
+/* These are not "real" source ids defined by HW */
+#define VISLANDS30_IV_SRCID_VM_CONTEXT_ALL                     0x00000100  /* 256 */
+#define VISLANDS30_IV_EXTID_VM_CONTEXT0_ALL                    0
+#define VISLANDS30_IV_EXTID_VM_CONTEXT1_ALL                    1
+
+
+/* IV Extended IDs */
+#define VISLANDS30_IV_EXTID_NONE                               0x00000000
+#define VISLANDS30_IV_EXTID_INVALID                            0xffffffff
+
 #endif // _IVSRCID_VISLANDS30_H_
diff --git a/drivers/gpu/drm/amd/include/ivsrcid/sdma0/irqsrcs_sdma0_4_0.h b/drivers/gpu/drm/amd/include/ivsrcid/sdma0/irqsrcs_sdma0_4_0.h
new file mode 100644
index 000000000000..802413832fe8
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/ivsrcid/sdma0/irqsrcs_sdma0_4_0.h
@@ -0,0 +1,50 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __IRQSRCS_SDMA0_4_0_H__
+#define __IRQSRCS_SDMA0_4_0_H__
+
+#define SDMA0_4_0__SRCID__SDMA_ATOMIC_RTN_DONE                         217             /* 0xD9 SDMA atomic*_rtn ops complete  */
+#define SDMA0_4_0__SRCID__SDMA_ATOMIC_TIMEOUT                          218             /* 0xDA SDMA atomic CMPSWAP loop timeout  */
+#define SDMA0_4_0__SRCID__SDMA_IB_PREEMPT                                      219             /* 0xDB sdma mid-command buffer preempt interrupt  */
+#define SDMA0_4_0__SRCID__SDMA_ECC                                             220             /* 0xDC ECC  Error  */
+#define SDMA0_4_0__SRCID__SDMA_PAGE_FAULT                                      221             /* 0xDD Page Fault Error from UTCL2 when nack=3  */
+#define SDMA0_4_0__SRCID__SDMA_PAGE_NULL                                       222             /* 0xDE Page Null from UTCL2 when nack=2  */
+#define SDMA0_4_0__SRCID__SDMA_XNACK                                       223         /* 0xDF Page retry  timeout after UTCL2 return nack=1  */
+#define SDMA0_4_0__SRCID__SDMA_TRAP                                            224             /* 0xE0 Trap  */
+#define SDMA0_4_0__SRCID__SDMA_SEM_INCOMPLETE_TIMEOUT          225             /* 0xE1 0xDAGPF (Sem incomplete timeout)  */
+#define SDMA0_4_0__SRCID__SDMA_SEM_WAIT_FAIL_TIMEOUT           226             /* 0xE2 Semaphore wait fail timeout  */
+#define SDMA0_4_0__SRCID__SDMA_SRAM_ECC                                            228         /* 0xE4 SRAM ECC Error  */
+#define SDMA0_4_0__SRCID__SDMA_PREEMPT                                     240         /* 0xF0 SDMA New Run List  */
+#define SDMA0_4_0__SRCID__SDMA_VM_HOLE                                     242         /* 0xF2 MC or SEM address in VM hole  */
+#define SDMA0_4_0__SRCID__SDMA_CTXEMPTY                                            243         /* 0xF3 Context Empty  */
+#define SDMA0_4_0__SRCID__SDMA_DOORBELL_INVALID                                244             /* 0xF4 Doorbell BE invalid  */
+#define SDMA0_4_0__SRCID__SDMA_FROZEN                                      245         /* 0xF5 SDMA Frozen  */
+#define SDMA0_4_0__SRCID__SDMA_POLL_TIMEOUT                                    246             /* 0xF6 SRBM read poll timeout  */
+#define SDMA0_4_0__SRCID__SDMA_SRBMWRITE                                       247             /* 0xF7 SRBM write Protection  */
+
+#endif /* __IRQSRCS_SDMA_4_0_H__ */
+
+
diff --git a/drivers/gpu/drm/amd/include/ivsrcid/sdma1/irqsrcs_sdma1_4_0.h b/drivers/gpu/drm/amd/include/ivsrcid/sdma1/irqsrcs_sdma1_4_0.h
new file mode 100644
index 000000000000..d12a35619f9a
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/ivsrcid/sdma1/irqsrcs_sdma1_4_0.h
@@ -0,0 +1,50 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __IRQSRCS_SDMA1_4_0_H__
+#define __IRQSRCS_SDMA1_4_0_H__
+
+#define SDMA1_4_0__SRCID__SDMA_ATOMIC_RTN_DONE                         217             /* 0xD9 SDMA atomic*_rtn ops complete  */
+#define SDMA1_4_0__SRCID__SDMA_ATOMIC_TIMEOUT                          218             /* 0xDA SDMA atomic CMPSWAP loop timeout  */
+#define SDMA1_4_0__SRCID__SDMA_IB_PREEMPT                                      219             /* 0xDB sdma mid-command buffer preempt interrupt  */
+#define SDMA1_4_0__SRCID__SDMA_ECC                                             220             /* 0xDC ECC  Error  */
+#define SDMA1_4_0__SRCID__SDMA_PAGE_FAULT                                      221             /* 0xDD Page Fault Error from UTCL2 when nack=3  */
+#define SDMA1_4_0__SRCID__SDMA_PAGE_NULL                                       222             /* 0xDE Page Null from UTCL2 when nack=2  */
+#define SDMA1_4_0__SRCID__SDMA_XNACK                                       223         /* 0xDF Page retry  timeout after UTCL2 return nack=1  */
+#define SDMA1_4_0__SRCID__SDMA_TRAP                                            224             /* 0xE0 Trap  */
+#define SDMA1_4_0__SRCID__SDMA_SEM_INCOMPLETE_TIMEOUT          225             /* 0xE1 0xDAGPF (Sem incomplete timeout)  */
+#define SDMA1_4_0__SRCID__SDMA_SEM_WAIT_FAIL_TIMEOUT           226             /* 0xE2 Semaphore wait fail timeout  */
+#define SDMA1_4_0__SRCID__SDMA_SRAM_ECC                                            228         /* 0xE4 SRAM ECC Error  */
+#define SDMA1_4_0__SRCID__SDMA_PREEMPT                                     240         /* 0xF0 SDMA New Run List  */
+#define SDMA1_4_0__SRCID__SDMA_VM_HOLE                                     242         /* 0xF2 MC or SEM address in VM hole  */
+#define SDMA1_4_0__SRCID__SDMA_CTXEMPTY                                            243         /* 0xF3 Context Empty  */
+#define SDMA1_4_0__SRCID__SDMA_DOORBELL_INVALID                                244             /* 0xF4 Doorbell BE invalid  */
+#define SDMA1_4_0__SRCID__SDMA_FROZEN                                      245         /* 0xF5 SDMA Frozen  */
+#define SDMA1_4_0__SRCID__SDMA_POLL_TIMEOUT                                    246             /* 0xF6 SRBM read poll timeout  */
+#define SDMA1_4_0__SRCID__SDMA_SRBMWRITE                                       247             /* 0xF7 SRBM write Protection  */
+
+#endif /* __IRQSRCS_SDMA1_4_0_H__ */
+
+
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk110.c b/drivers/gpu/drm/amd/include/ivsrcid/smuio/irqsrcs_smuio_9_0.h
index a9aa69c82e8e..02bab4673cd4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk110.c
+++ b/drivers/gpu/drm/amd/include/ivsrcid/smuio/irqsrcs_smuio_9_0.h
@@ -1,5 +1,5 @@
 /*
- * Copyright 2016 Red Hat Inc.
+ * Copyright 2017 Advanced Micro Devices, Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -19,16 +19,14 @@
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
  *
- * Authors: Ben Skeggs
+ * Authors: AMD
+ *
  */
-#include "changk104.h"
 
-#include <nvif/class.h>
+#ifndef __IRQSRCS_SMUIO_9_0_H__
+#define __IRQSRCS_SMUIO_9_0_H__
+
+#define SMUIO_9_0__SRCID__SMUIO_GPIO19			83		/* GPIO19 interrupt  */
+
+#endif /* __IRQSRCS_SMUIO_9_0_H__ */
 
-const struct nvkm_fifo_chan_oclass
-gk110_fifo_gpfifo_oclass = {
-	.base.oclass = KEPLER_CHANNEL_GPFIFO_B,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = gk104_fifo_gpfifo_new,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgk104.c b/drivers/gpu/drm/amd/include/ivsrcid/thm/irqsrcs_thm_9_0.h
index 0c09fe85e952..5218bc53fb2d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgk104.c
+++ b/drivers/gpu/drm/amd/include/ivsrcid/thm/irqsrcs_thm_9_0.h
@@ -1,5 +1,5 @@
 /*
- * Copyright 2012 Red Hat Inc.
+ * Copyright 2017 Advanced Micro Devices, Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -19,19 +19,15 @@
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
  *
- * Authors: Ben Skeggs
+ * Authors: AMD
+ *
  */
-#include "channv50.h"
-#include "rootnv50.h"
 
-#include <nvif/class.h>
+#ifndef __IRQSRCS_THM_9_0_H__
+#define __IRQSRCS_THM_9_0_H__
+
+#define THM_9_0__SRCID__THM_DIG_THERM_L2H		0		/* ASIC_TEMP > CG_THERMAL_INT.DIG_THERM_INTH  */
+#define THM_9_0__SRCID__THM_DIG_THERM_H2L		1		/* ASIC_TEMP < CG_THERMAL_INT.DIG_THERM_INTL  */
+
+#endif /* __IRQSRCS_THM_9_0_H__ */
 
-const struct nv50_disp_pioc_oclass
-gk104_disp_oimm_oclass = {
-	.base.oclass = GK104_DISP_OVERLAY,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_oimm_new,
-	.func = &gf119_disp_pioc_func,
-	.chid = { 9, 9 },
-};
diff --git a/drivers/gpu/drm/amd/display/dc/dml/soc_bounding_box.h b/drivers/gpu/drm/amd/include/ivsrcid/uvd/irqsrcs_uvd_7_0.h
index 7a65206a6d21..fb041aee6c66 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/soc_bounding_box.h
+++ b/drivers/gpu/drm/amd/include/ivsrcid/uvd/irqsrcs_uvd_7_0.h
@@ -23,13 +23,12 @@
  *
  */
 
-#ifndef __SOC_BOUNDING_BOX_H__
-#define __SOC_BOUNDING_BOX_H__
+#ifndef __IRQSRCS_UVD_7_0_H__
+#define __IRQSRCS_UVD_7_0_H__
 
-#include "dml_common_defs.h"
+#define UVD_7_0__SRCID__UVD_ENC_GEN_PURP                119
+#define UVD_7_0__SRCID__UVD_ENC_LOW_LATENCY             120
+#define UVD_7_0__SRCID__UVD_SYSTEM_MESSAGE_INTERRUPT	124		/* UVD system message interrupt  */
 
-void dml_socbb_set_latencies(soc_bounding_box_st *to_box, soc_bounding_box_st *from_box);
-voltage_scaling_st dml_socbb_voltage_scaling(const soc_bounding_box_st *box, enum voltage_state voltage);
-double dml_socbb_return_bw_mhz(soc_bounding_box_st *box, enum voltage_state voltage);
+#endif /* __IRQSRCS_UVD_7_0_H__ */
 
-#endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursg84.c b/drivers/gpu/drm/amd/include/ivsrcid/vce/irqsrcs_vce_4_0.h
index fa781b5a7e07..3440bab565af 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursg84.c
+++ b/drivers/gpu/drm/amd/include/ivsrcid/vce/irqsrcs_vce_4_0.h
@@ -1,5 +1,5 @@
 /*
- * Copyright 2015 Red Hat Inc.
+ * Copyright 2017 Advanced Micro Devices, Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -19,19 +19,16 @@
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
  *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
+ * Authors: AMD
+ *
  */
-#include "channv50.h"
-#include "rootnv50.h"
 
-#include <nvif/class.h>
+#ifndef __IRQSRCS_VCE_4_0_H__
+#define __IRQSRCS_VCE_4_0_H__
+
+#define VCE_4_0__CTXID__VCE_TRAP_GENERAL_PURPOSE               0
+#define VCE_4_0__CTXID__VCE_TRAP_LOW_LATENCY                   1
+#define VCE_4_0__CTXID__VCE_TRAP_REAL_TIME                     2
+
+#endif /* __IRQSRCS_VCE_4_0_H__ */
 
-const struct nv50_disp_pioc_oclass
-g84_disp_curs_oclass = {
-	.base.oclass = G82_DISP_CURSOR,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_curs_new,
-	.func = &nv50_disp_pioc_func,
-	.chid = { 7, 7 },
-};
diff --git a/drivers/gpu/drm/amd/include/ivsrcid/vcn/irqsrcs_vcn_1_0.h b/drivers/gpu/drm/amd/include/ivsrcid/vcn/irqsrcs_vcn_1_0.h
new file mode 100644
index 000000000000..e5951709bfc3
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/ivsrcid/vcn/irqsrcs_vcn_1_0.h
@@ -0,0 +1,34 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __IRQSRCS_VCN_1_0_H__
+#define __IRQSRCS_VCN_1_0_H__
+
+#define VCN_1_0__SRCID__UVD_ENC_GENERAL_PURPOSE                 119     /* 0x77 Encoder General Purpose  */
+#define VCN_1_0__SRCID__UVD_ENC_LOW_LATENCY                     120     /* 0x78 Encoder Low Latency  */
+#define VCN_1_0__SRCID__UVD_SYSTEM_MESSAGE_INTERRUPT	        124		/* 0x7c UVD system message interrupt  */
+
+#endif /* __IRQSRCS_VCN_1_0_H__ */
+
diff --git a/drivers/gpu/drm/amd/include/ivsrcid/vmc/irqsrcs_vmc_1_0.h b/drivers/gpu/drm/amd/include/ivsrcid/vmc/irqsrcs_vmc_1_0.h
new file mode 100644
index 000000000000..d130936c9989
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/ivsrcid/vmc/irqsrcs_vmc_1_0.h
@@ -0,0 +1,37 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __IRQSRCS_VMC_1_0_H__
+#define __IRQSRCS_VMC_1_0_H__
+
+
+#define VMC_1_0__SRCID__VM_FAULT                            0
+#define VMC_1_0__SRCID__VM_CONTEXT0_ALL                     256
+#define VMC_1_0__SRCID__VM_CONTEXT1_ALL                     257
+
+#define UTCL2_1_0__SRCID__FAULT                             0       /* UTC L2 has encountered a fault or retry scenario */
+
+
+#endif /* __IRQSRCS_VMC_1_0_H__ */
diff --git a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
index 237289a72bb7..5733fbee07f7 100644
--- a/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_kfd_interface.h
@@ -100,6 +100,21 @@ struct kgd2kfd_shared_resources {
 	/* Bit n == 1 means Queue n is available for KFD */
 	DECLARE_BITMAP(queue_bitmap, KGD_MAX_QUEUES);
 
+	/* Doorbell assignments (SOC15 and later chips only). Only
+	 * specific doorbells are routed to each SDMA engine. Others
+	 * are routed to IH and VCN. They are not usable by the CP.
+	 *
+	 * Any doorbell number D that satisfies the following condition
+	 * is reserved: (D & reserved_doorbell_mask) == reserved_doorbell_val
+	 *
+	 * KFD currently uses 1024 (= 0x3ff) doorbells per process. If
+	 * doorbells 0x0f0-0x0f7 and 0x2f-0x2f7 are reserved, that means
+	 * mask would be set to 0x1f8 and val set to 0x0f0.
+	 */
+	unsigned int sdma_doorbell[2][2];
+	unsigned int reserved_doorbell_mask;
+	unsigned int reserved_doorbell_val;
+
 	/* Base address of doorbell aperture. */
 	phys_addr_t doorbell_physical_address;
 
@@ -173,8 +188,6 @@ struct tile_config {
  * @set_pasid_vmid_mapping: Exposes pasid/vmid pair to the H/W for no cp
  * scheduling mode. Only used for no cp scheduling mode.
  *
- * @init_pipeline: Initialized the compute pipelines.
- *
  * @hqd_load: Loads the mqd structure to a H/W hqd slot. used only for no cp
  * sceduling mode.
  *
@@ -274,9 +287,6 @@ struct kfd2kgd_calls {
 	int (*set_pasid_vmid_mapping)(struct kgd_dev *kgd, unsigned int pasid,
 					unsigned int vmid);
 
-	int (*init_pipeline)(struct kgd_dev *kgd, uint32_t pipe_id,
-				uint32_t hpd_size, uint64_t hpd_gpu_addr);
-
 	int (*init_interrupts)(struct kgd_dev *kgd, uint32_t pipe_id);
 
 	int (*hqd_load)(struct kgd_dev *kgd, void *mqd, uint32_t pipe_id,
@@ -382,6 +392,10 @@ struct kfd2kgd_calls {
  *
  * @resume: Notifies amdkfd about a resume action done to a kgd device
  *
+ * @quiesce_mm: Quiesce all user queue access to specified MM address space
+ *
+ * @resume_mm: Resume user queue access to specified MM address space
+ *
  * @schedule_evict_and_restore_process: Schedules work queue that will prepare
  * for safe eviction of KFD BOs that belong to the specified process.
  *
@@ -399,6 +413,8 @@ struct kgd2kfd_calls {
 	void (*interrupt)(struct kfd_dev *kfd, const void *ih_ring_entry);
 	void (*suspend)(struct kfd_dev *kfd);
 	int (*resume)(struct kfd_dev *kfd);
+	int (*quiesce_mm)(struct mm_struct *mm);
+	int (*resume_mm)(struct mm_struct *mm);
 	int (*schedule_evict_and_restore_process)(struct mm_struct *mm,
 			struct dma_fence *fence);
 };
diff --git a/drivers/gpu/drm/amd/include/kgd_pp_interface.h b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
index 5c840c022b52..6a41b81c7325 100644
--- a/drivers/gpu/drm/amd/include/kgd_pp_interface.h
+++ b/drivers/gpu/drm/amd/include/kgd_pp_interface.h
@@ -94,6 +94,7 @@ enum pp_clock_type {
 	PP_PCIE,
 	OD_SCLK,
 	OD_MCLK,
+	OD_RANGE,
 };
 
 enum amd_pp_sensors {
@@ -149,13 +150,6 @@ struct pp_states_info {
 	uint32_t states[16];
 };
 
-struct pp_gpu_power {
-	uint32_t vddc_power;
-	uint32_t vddci_power;
-	uint32_t max_gpu_power;
-	uint32_t average_gpu_power;
-};
-
 #define PP_GROUP_MASK        0xF0000000
 #define PP_GROUP_SHIFT       28
 
@@ -198,7 +192,6 @@ struct amd_pp_simple_clock_info;
 struct amd_pp_display_configuration;
 struct amd_pp_clock_info;
 struct pp_display_clock_request;
-struct pp_wm_sets_with_clock_ranges_soc15;
 struct pp_clock_levels_with_voltage;
 struct pp_clock_levels_with_latency;
 struct amd_pp_clocks;
@@ -238,21 +231,19 @@ struct amd_pm_funcs {
 	void (*debugfs_print_current_performance_level)(void *handle, struct seq_file *m);
 	int (*switch_power_profile)(void *handle, enum PP_SMC_POWER_PROFILE type, bool en);
 /* export to amdgpu */
-	void (*powergate_uvd)(void *handle, bool gate);
-	void (*powergate_vce)(void *handle, bool gate);
 	struct amd_vce_state *(*get_vce_clock_state)(void *handle, u32 idx);
 	int (*dispatch_tasks)(void *handle, enum amd_pp_task task_id,
 			enum amd_pm_state_type *user_state);
 	int (*load_firmware)(void *handle);
 	int (*wait_for_fw_loading_complete)(void *handle);
+	int (*set_powergating_by_smu)(void *handle,
+				uint32_t block_type, bool gate);
 	int (*set_clockgating_by_smu)(void *handle, uint32_t msg_id);
-	int (*notify_smu_memory_info)(void *handle, uint32_t virtual_addr_low,
-					uint32_t virtual_addr_hi,
-					uint32_t mc_addr_low,
-					uint32_t mc_addr_hi,
-					uint32_t size);
 	int (*set_power_limit)(void *handle, uint32_t n);
 	int (*get_power_limit)(void *handle, uint32_t *limit, bool default_limit);
+	int (*get_power_profile_mode)(void *handle, char *buf);
+	int (*set_power_profile_mode)(void *handle, long *input, uint32_t size);
+	int (*odn_edit_dpm_table)(void *handle, uint32_t type, long *input, uint32_t size);
 /* export to DC */
 	u32 (*get_sclk)(void *handle, bool low);
 	u32 (*get_mclk)(void *handle, bool low);
@@ -272,15 +263,12 @@ struct amd_pm_funcs {
 		enum amd_pp_clock_type type,
 		struct pp_clock_levels_with_voltage *clocks);
 	int (*set_watermarks_for_clocks_ranges)(void *handle,
-		struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges);
+						void *clock_ranges);
 	int (*display_clock_voltage_request)(void *handle,
 				struct pp_display_clock_request *clock);
 	int (*get_display_mode_validation_clocks)(void *handle,
 		struct amd_pp_simple_clock_info *clocks);
-	int (*get_power_profile_mode)(void *handle, char *buf);
-	int (*set_power_profile_mode)(void *handle, long *input, uint32_t size);
-	int (*odn_edit_dpm_table)(void *handle, uint32_t type, long *input, uint32_t size);
-	int (*set_mmhub_powergating_by_smu)(void *handle);
+	int (*notify_smu_enable_pwe)(void *handle);
 };
 
 #endif
diff --git a/drivers/gpu/drm/amd/include/soc15_ih_clientid.h b/drivers/gpu/drm/amd/include/soc15_ih_clientid.h
index a12d4f27cfa4..12e196c15bbe 100644
--- a/drivers/gpu/drm/amd/include/soc15_ih_clientid.h
+++ b/drivers/gpu/drm/amd/include/soc15_ih_clientid.h
@@ -43,6 +43,7 @@ enum soc15_ih_clientid {
 	SOC15_IH_CLIENTID_SE2SH		= 0x0c,
 	SOC15_IH_CLIENTID_SE3SH		= 0x0d,
 	SOC15_IH_CLIENTID_SYSHUB	= 0x0e,
+	SOC15_IH_CLIENTID_UVD1		= 0x0e,
 	SOC15_IH_CLIENTID_THM		= 0x0f,
 	SOC15_IH_CLIENTID_UVD		= 0x10,
 	SOC15_IH_CLIENTID_VCE0		= 0x11,
diff --git a/drivers/gpu/drm/amd/include/v9_structs.h b/drivers/gpu/drm/amd/include/v9_structs.h
index 2fb25abaf7c8..ceaf4932258d 100644
--- a/drivers/gpu/drm/amd/include/v9_structs.h
+++ b/drivers/gpu/drm/amd/include/v9_structs.h
@@ -29,10 +29,10 @@ struct v9_sdma_mqd {
 	uint32_t sdmax_rlcx_rb_base;
 	uint32_t sdmax_rlcx_rb_base_hi;
 	uint32_t sdmax_rlcx_rb_rptr;
+	uint32_t sdmax_rlcx_rb_rptr_hi;
 	uint32_t sdmax_rlcx_rb_wptr;
+	uint32_t sdmax_rlcx_rb_wptr_hi;
 	uint32_t sdmax_rlcx_rb_wptr_poll_cntl;
-	uint32_t sdmax_rlcx_rb_wptr_poll_addr_hi;
-	uint32_t sdmax_rlcx_rb_wptr_poll_addr_lo;
 	uint32_t sdmax_rlcx_rb_rptr_addr_hi;
 	uint32_t sdmax_rlcx_rb_rptr_addr_lo;
 	uint32_t sdmax_rlcx_ib_cntl;
@@ -44,29 +44,29 @@ struct v9_sdma_mqd {
 	uint32_t sdmax_rlcx_skip_cntl;
 	uint32_t sdmax_rlcx_context_status;
 	uint32_t sdmax_rlcx_doorbell;
-	uint32_t sdmax_rlcx_virtual_addr;
-	uint32_t sdmax_rlcx_ape1_cntl;
+	uint32_t sdmax_rlcx_status;
 	uint32_t sdmax_rlcx_doorbell_log;
-	uint32_t reserved_22;
-	uint32_t reserved_23;
-	uint32_t reserved_24;
-	uint32_t reserved_25;
-	uint32_t reserved_26;
-	uint32_t reserved_27;
-	uint32_t reserved_28;
-	uint32_t reserved_29;
-	uint32_t reserved_30;
-	uint32_t reserved_31;
-	uint32_t reserved_32;
-	uint32_t reserved_33;
-	uint32_t reserved_34;
-	uint32_t reserved_35;
-	uint32_t reserved_36;
-	uint32_t reserved_37;
-	uint32_t reserved_38;
-	uint32_t reserved_39;
-	uint32_t reserved_40;
-	uint32_t reserved_41;
+	uint32_t sdmax_rlcx_watermark;
+	uint32_t sdmax_rlcx_doorbell_offset;
+	uint32_t sdmax_rlcx_csa_addr_lo;
+	uint32_t sdmax_rlcx_csa_addr_hi;
+	uint32_t sdmax_rlcx_ib_sub_remain;
+	uint32_t sdmax_rlcx_preempt;
+	uint32_t sdmax_rlcx_dummy_reg;
+	uint32_t sdmax_rlcx_rb_wptr_poll_addr_hi;
+	uint32_t sdmax_rlcx_rb_wptr_poll_addr_lo;
+	uint32_t sdmax_rlcx_rb_aql_cntl;
+	uint32_t sdmax_rlcx_minor_ptr_update;
+	uint32_t sdmax_rlcx_midcmd_data0;
+	uint32_t sdmax_rlcx_midcmd_data1;
+	uint32_t sdmax_rlcx_midcmd_data2;
+	uint32_t sdmax_rlcx_midcmd_data3;
+	uint32_t sdmax_rlcx_midcmd_data4;
+	uint32_t sdmax_rlcx_midcmd_data5;
+	uint32_t sdmax_rlcx_midcmd_data6;
+	uint32_t sdmax_rlcx_midcmd_data7;
+	uint32_t sdmax_rlcx_midcmd_data8;
+	uint32_t sdmax_rlcx_midcmd_cntl;
 	uint32_t reserved_42;
 	uint32_t reserved_43;
 	uint32_t reserved_44;
diff --git a/drivers/gpu/drm/amd/include/vega20_ip_offset.h b/drivers/gpu/drm/amd/include/vega20_ip_offset.h
new file mode 100644
index 000000000000..2a2a9cc8bedb
--- /dev/null
+++ b/drivers/gpu/drm/amd/include/vega20_ip_offset.h
@@ -0,0 +1,1051 @@
+/*
+ * Copyright (C) 2018  Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included
+ * in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
+ * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN
+ * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ */
+#ifndef _vega20_ip_offset_HEADER
+#define _vega20_ip_offset_HEADER
+
+#define MAX_INSTANCE                                       6
+#define MAX_SEGMENT                                        6
+
+
+struct IP_BASE_INSTANCE
+{
+    unsigned int segment[MAX_SEGMENT];
+};
+
+struct IP_BASE
+{
+    struct IP_BASE_INSTANCE instance[MAX_INSTANCE];
+};
+
+
+static const struct IP_BASE ATHUB_BASE            ={ { { { 0x00000C20, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE CLK_BASE            ={ { { { 0x00016C00, 0x00016E00, 0x00017000, 0x00017200, 0x0001B000, 0x0001B200 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE DCE_BASE            ={ { { { 0x00000012, 0x000000C0, 0x000034C0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE DF_BASE            ={ { { { 0x00007000, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE FUSE_BASE            ={ { { { 0x00017400, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE GC_BASE            ={ { { { 0x00002000, 0x0000A000, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE HDP_BASE            ={ { { { 0x00000F20, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE MMHUB_BASE            ={ { { { 0x0001A000, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE MP0_BASE            ={ { { { 0x00016000, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE MP1_BASE            ={ { { { 0x00016000, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE NBIO_BASE            ={ { { { 0x00000000, 0x00000014, 0x00000D20, 0x00010400, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE OSSSYS_BASE            ={ { { { 0x000010A0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE SDMA0_BASE            ={ { { { 0x00001260, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE SDMA1_BASE            ={ { { { 0x00001860, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE SMUIO_BASE            ={ { { { 0x00016800, 0x00016A00, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE THM_BASE            ={ { { { 0x00016600, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE UMC_BASE            ={ { { { 0x00014000, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE UVD_BASE            ={ { { { 0x00007800, 0x00007E00, 0, 0, 0, 0 } },
+                                        { { 0, 0x00009000, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+/* Adjust VCE_BASE to make vce_4_1 use vce_4_0 offset header files*/
+static const struct IP_BASE VCE_BASE            ={ { { { 0x00007E00/* 0x00008800 */, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE XDMA_BASE            ={ { { { 0x00003400, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+static const struct IP_BASE RSMU_BASE            ={ { { { 0x00012000, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } },
+                                        { { 0, 0, 0, 0, 0, 0 } } } };
+
+
+#define ATHUB_BASE__INST0_SEG0                     0x00000C20
+#define ATHUB_BASE__INST0_SEG1                     0
+#define ATHUB_BASE__INST0_SEG2                     0
+#define ATHUB_BASE__INST0_SEG3                     0
+#define ATHUB_BASE__INST0_SEG4                     0
+#define ATHUB_BASE__INST0_SEG5                     0
+
+#define ATHUB_BASE__INST1_SEG0                     0
+#define ATHUB_BASE__INST1_SEG1                     0
+#define ATHUB_BASE__INST1_SEG2                     0
+#define ATHUB_BASE__INST1_SEG3                     0
+#define ATHUB_BASE__INST1_SEG4                     0
+#define ATHUB_BASE__INST1_SEG5                     0
+
+#define ATHUB_BASE__INST2_SEG0                     0
+#define ATHUB_BASE__INST2_SEG1                     0
+#define ATHUB_BASE__INST2_SEG2                     0
+#define ATHUB_BASE__INST2_SEG3                     0
+#define ATHUB_BASE__INST2_SEG4                     0
+#define ATHUB_BASE__INST2_SEG5                     0
+
+#define ATHUB_BASE__INST3_SEG0                     0
+#define ATHUB_BASE__INST3_SEG1                     0
+#define ATHUB_BASE__INST3_SEG2                     0
+#define ATHUB_BASE__INST3_SEG3                     0
+#define ATHUB_BASE__INST3_SEG4                     0
+#define ATHUB_BASE__INST3_SEG5                     0
+
+#define ATHUB_BASE__INST4_SEG0                     0
+#define ATHUB_BASE__INST4_SEG1                     0
+#define ATHUB_BASE__INST4_SEG2                     0
+#define ATHUB_BASE__INST4_SEG3                     0
+#define ATHUB_BASE__INST4_SEG4                     0
+#define ATHUB_BASE__INST4_SEG5                     0
+
+#define ATHUB_BASE__INST5_SEG0                     0
+#define ATHUB_BASE__INST5_SEG1                     0
+#define ATHUB_BASE__INST5_SEG2                     0
+#define ATHUB_BASE__INST5_SEG3                     0
+#define ATHUB_BASE__INST5_SEG4                     0
+#define ATHUB_BASE__INST5_SEG5                     0
+
+#define CLK_BASE__INST0_SEG0                       0x00016C00
+#define CLK_BASE__INST0_SEG1                       0x00016E00
+#define CLK_BASE__INST0_SEG2                       0x00017000
+#define CLK_BASE__INST0_SEG3                       0x00017200
+#define CLK_BASE__INST0_SEG4                       0x0001B000
+#define CLK_BASE__INST0_SEG5                       0x0001B200
+
+#define CLK_BASE__INST1_SEG0                       0
+#define CLK_BASE__INST1_SEG1                       0
+#define CLK_BASE__INST1_SEG2                       0
+#define CLK_BASE__INST1_SEG3                       0
+#define CLK_BASE__INST1_SEG4                       0
+#define CLK_BASE__INST1_SEG5                       0
+
+#define CLK_BASE__INST2_SEG0                       0
+#define CLK_BASE__INST2_SEG1                       0
+#define CLK_BASE__INST2_SEG2                       0
+#define CLK_BASE__INST2_SEG3                       0
+#define CLK_BASE__INST2_SEG4                       0
+#define CLK_BASE__INST2_SEG5                       0
+
+#define CLK_BASE__INST3_SEG0                       0
+#define CLK_BASE__INST3_SEG1                       0
+#define CLK_BASE__INST3_SEG2                       0
+#define CLK_BASE__INST3_SEG3                       0
+#define CLK_BASE__INST3_SEG4                       0
+#define CLK_BASE__INST3_SEG5                       0
+
+#define CLK_BASE__INST4_SEG0                       0
+#define CLK_BASE__INST4_SEG1                       0
+#define CLK_BASE__INST4_SEG2                       0
+#define CLK_BASE__INST4_SEG3                       0
+#define CLK_BASE__INST4_SEG4                       0
+#define CLK_BASE__INST4_SEG5                       0
+
+#define CLK_BASE__INST5_SEG0                       0
+#define CLK_BASE__INST5_SEG1                       0
+#define CLK_BASE__INST5_SEG2                       0
+#define CLK_BASE__INST5_SEG3                       0
+#define CLK_BASE__INST5_SEG4                       0
+#define CLK_BASE__INST5_SEG5                       0
+
+#define DCE_BASE__INST0_SEG0                       0x00000012
+#define DCE_BASE__INST0_SEG1                       0x000000C0
+#define DCE_BASE__INST0_SEG2                       0x000034C0
+#define DCE_BASE__INST0_SEG3                       0
+#define DCE_BASE__INST0_SEG4                       0
+#define DCE_BASE__INST0_SEG5                       0
+
+#define DCE_BASE__INST1_SEG0                       0
+#define DCE_BASE__INST1_SEG1                       0
+#define DCE_BASE__INST1_SEG2                       0
+#define DCE_BASE__INST1_SEG3                       0
+#define DCE_BASE__INST1_SEG4                       0
+#define DCE_BASE__INST1_SEG5                       0
+
+#define DCE_BASE__INST2_SEG0                       0
+#define DCE_BASE__INST2_SEG1                       0
+#define DCE_BASE__INST2_SEG2                       0
+#define DCE_BASE__INST2_SEG3                       0
+#define DCE_BASE__INST2_SEG4                       0
+#define DCE_BASE__INST2_SEG5                       0
+
+#define DCE_BASE__INST3_SEG0                       0
+#define DCE_BASE__INST3_SEG1                       0
+#define DCE_BASE__INST3_SEG2                       0
+#define DCE_BASE__INST3_SEG3                       0
+#define DCE_BASE__INST3_SEG4                       0
+#define DCE_BASE__INST3_SEG5                       0
+
+#define DCE_BASE__INST4_SEG0                       0
+#define DCE_BASE__INST4_SEG1                       0
+#define DCE_BASE__INST4_SEG2                       0
+#define DCE_BASE__INST4_SEG3                       0
+#define DCE_BASE__INST4_SEG4                       0
+#define DCE_BASE__INST4_SEG5                       0
+
+#define DCE_BASE__INST5_SEG0                       0
+#define DCE_BASE__INST5_SEG1                       0
+#define DCE_BASE__INST5_SEG2                       0
+#define DCE_BASE__INST5_SEG3                       0
+#define DCE_BASE__INST5_SEG4                       0
+#define DCE_BASE__INST5_SEG5                       0
+
+#define DF_BASE__INST0_SEG0                        0x00007000
+#define DF_BASE__INST0_SEG1                        0
+#define DF_BASE__INST0_SEG2                        0
+#define DF_BASE__INST0_SEG3                        0
+#define DF_BASE__INST0_SEG4                        0
+#define DF_BASE__INST0_SEG5                        0
+
+#define DF_BASE__INST1_SEG0                        0
+#define DF_BASE__INST1_SEG1                        0
+#define DF_BASE__INST1_SEG2                        0
+#define DF_BASE__INST1_SEG3                        0
+#define DF_BASE__INST1_SEG4                        0
+#define DF_BASE__INST1_SEG5                        0
+
+#define DF_BASE__INST2_SEG0                        0
+#define DF_BASE__INST2_SEG1                        0
+#define DF_BASE__INST2_SEG2                        0
+#define DF_BASE__INST2_SEG3                        0
+#define DF_BASE__INST2_SEG4                        0
+#define DF_BASE__INST2_SEG5                        0
+
+#define DF_BASE__INST3_SEG0                        0
+#define DF_BASE__INST3_SEG1                        0
+#define DF_BASE__INST3_SEG2                        0
+#define DF_BASE__INST3_SEG3                        0
+#define DF_BASE__INST3_SEG4                        0
+#define DF_BASE__INST3_SEG5                        0
+
+#define DF_BASE__INST4_SEG0                        0
+#define DF_BASE__INST4_SEG1                        0
+#define DF_BASE__INST4_SEG2                        0
+#define DF_BASE__INST4_SEG3                        0
+#define DF_BASE__INST4_SEG4                        0
+#define DF_BASE__INST4_SEG5                        0
+
+#define DF_BASE__INST5_SEG0                        0
+#define DF_BASE__INST5_SEG1                        0
+#define DF_BASE__INST5_SEG2                        0
+#define DF_BASE__INST5_SEG3                        0
+#define DF_BASE__INST5_SEG4                        0
+#define DF_BASE__INST5_SEG5                        0
+
+#define FUSE_BASE__INST0_SEG0                      0x00017400
+#define FUSE_BASE__INST0_SEG1                      0
+#define FUSE_BASE__INST0_SEG2                      0
+#define FUSE_BASE__INST0_SEG3                      0
+#define FUSE_BASE__INST0_SEG4                      0
+#define FUSE_BASE__INST0_SEG5                      0
+
+#define FUSE_BASE__INST1_SEG0                      0
+#define FUSE_BASE__INST1_SEG1                      0
+#define FUSE_BASE__INST1_SEG2                      0
+#define FUSE_BASE__INST1_SEG3                      0
+#define FUSE_BASE__INST1_SEG4                      0
+#define FUSE_BASE__INST1_SEG5                      0
+
+#define FUSE_BASE__INST2_SEG0                      0
+#define FUSE_BASE__INST2_SEG1                      0
+#define FUSE_BASE__INST2_SEG2                      0
+#define FUSE_BASE__INST2_SEG3                      0
+#define FUSE_BASE__INST2_SEG4                      0
+#define FUSE_BASE__INST2_SEG5                      0
+
+#define FUSE_BASE__INST3_SEG0                      0
+#define FUSE_BASE__INST3_SEG1                      0
+#define FUSE_BASE__INST3_SEG2                      0
+#define FUSE_BASE__INST3_SEG3                      0
+#define FUSE_BASE__INST3_SEG4                      0
+#define FUSE_BASE__INST3_SEG5                      0
+
+#define FUSE_BASE__INST4_SEG0                      0
+#define FUSE_BASE__INST4_SEG1                      0
+#define FUSE_BASE__INST4_SEG2                      0
+#define FUSE_BASE__INST4_SEG3                      0
+#define FUSE_BASE__INST4_SEG4                      0
+#define FUSE_BASE__INST4_SEG5                      0
+
+#define FUSE_BASE__INST5_SEG0                      0
+#define FUSE_BASE__INST5_SEG1                      0
+#define FUSE_BASE__INST5_SEG2                      0
+#define FUSE_BASE__INST5_SEG3                      0
+#define FUSE_BASE__INST5_SEG4                      0
+#define FUSE_BASE__INST5_SEG5                      0
+
+#define GC_BASE__INST0_SEG0                        0x00002000
+#define GC_BASE__INST0_SEG1                        0x0000A000
+#define GC_BASE__INST0_SEG2                        0
+#define GC_BASE__INST0_SEG3                        0
+#define GC_BASE__INST0_SEG4                        0
+#define GC_BASE__INST0_SEG5                        0
+
+#define GC_BASE__INST1_SEG0                        0
+#define GC_BASE__INST1_SEG1                        0
+#define GC_BASE__INST1_SEG2                        0
+#define GC_BASE__INST1_SEG3                        0
+#define GC_BASE__INST1_SEG4                        0
+#define GC_BASE__INST1_SEG5                        0
+
+#define GC_BASE__INST2_SEG0                        0
+#define GC_BASE__INST2_SEG1                        0
+#define GC_BASE__INST2_SEG2                        0
+#define GC_BASE__INST2_SEG3                        0
+#define GC_BASE__INST2_SEG4                        0
+#define GC_BASE__INST2_SEG5                        0
+
+#define GC_BASE__INST3_SEG0                        0
+#define GC_BASE__INST3_SEG1                        0
+#define GC_BASE__INST3_SEG2                        0
+#define GC_BASE__INST3_SEG3                        0
+#define GC_BASE__INST3_SEG4                        0
+#define GC_BASE__INST3_SEG5                        0
+
+#define GC_BASE__INST4_SEG0                        0
+#define GC_BASE__INST4_SEG1                        0
+#define GC_BASE__INST4_SEG2                        0
+#define GC_BASE__INST4_SEG3                        0
+#define GC_BASE__INST4_SEG4                        0
+#define GC_BASE__INST4_SEG5                        0
+
+#define GC_BASE__INST5_SEG0                        0
+#define GC_BASE__INST5_SEG1                        0
+#define GC_BASE__INST5_SEG2                        0
+#define GC_BASE__INST5_SEG3                        0
+#define GC_BASE__INST5_SEG4                        0
+#define GC_BASE__INST5_SEG5                        0
+
+#define HDP_BASE__INST0_SEG0                       0x00000F20
+#define HDP_BASE__INST0_SEG1                       0
+#define HDP_BASE__INST0_SEG2                       0
+#define HDP_BASE__INST0_SEG3                       0
+#define HDP_BASE__INST0_SEG4                       0
+#define HDP_BASE__INST0_SEG5                       0
+
+#define HDP_BASE__INST1_SEG0                       0
+#define HDP_BASE__INST1_SEG1                       0
+#define HDP_BASE__INST1_SEG2                       0
+#define HDP_BASE__INST1_SEG3                       0
+#define HDP_BASE__INST1_SEG4                       0
+#define HDP_BASE__INST1_SEG5                       0
+
+#define HDP_BASE__INST2_SEG0                       0
+#define HDP_BASE__INST2_SEG1                       0
+#define HDP_BASE__INST2_SEG2                       0
+#define HDP_BASE__INST2_SEG3                       0
+#define HDP_BASE__INST2_SEG4                       0
+#define HDP_BASE__INST2_SEG5                       0
+
+#define HDP_BASE__INST3_SEG0                       0
+#define HDP_BASE__INST3_SEG1                       0
+#define HDP_BASE__INST3_SEG2                       0
+#define HDP_BASE__INST3_SEG3                       0
+#define HDP_BASE__INST3_SEG4                       0
+#define HDP_BASE__INST3_SEG5                       0
+
+#define HDP_BASE__INST4_SEG0                       0
+#define HDP_BASE__INST4_SEG1                       0
+#define HDP_BASE__INST4_SEG2                       0
+#define HDP_BASE__INST4_SEG3                       0
+#define HDP_BASE__INST4_SEG4                       0
+#define HDP_BASE__INST4_SEG5                       0
+
+#define HDP_BASE__INST5_SEG0                       0
+#define HDP_BASE__INST5_SEG1                       0
+#define HDP_BASE__INST5_SEG2                       0
+#define HDP_BASE__INST5_SEG3                       0
+#define HDP_BASE__INST5_SEG4                       0
+#define HDP_BASE__INST5_SEG5                       0
+
+#define MMHUB_BASE__INST0_SEG0                     0x0001A000
+#define MMHUB_BASE__INST0_SEG1                     0
+#define MMHUB_BASE__INST0_SEG2                     0
+#define MMHUB_BASE__INST0_SEG3                     0
+#define MMHUB_BASE__INST0_SEG4                     0
+#define MMHUB_BASE__INST0_SEG5                     0
+
+#define MMHUB_BASE__INST1_SEG0                     0
+#define MMHUB_BASE__INST1_SEG1                     0
+#define MMHUB_BASE__INST1_SEG2                     0
+#define MMHUB_BASE__INST1_SEG3                     0
+#define MMHUB_BASE__INST1_SEG4                     0
+#define MMHUB_BASE__INST1_SEG5                     0
+
+#define MMHUB_BASE__INST2_SEG0                     0
+#define MMHUB_BASE__INST2_SEG1                     0
+#define MMHUB_BASE__INST2_SEG2                     0
+#define MMHUB_BASE__INST2_SEG3                     0
+#define MMHUB_BASE__INST2_SEG4                     0
+#define MMHUB_BASE__INST2_SEG5                     0
+
+#define MMHUB_BASE__INST3_SEG0                     0
+#define MMHUB_BASE__INST3_SEG1                     0
+#define MMHUB_BASE__INST3_SEG2                     0
+#define MMHUB_BASE__INST3_SEG3                     0
+#define MMHUB_BASE__INST3_SEG4                     0
+#define MMHUB_BASE__INST3_SEG5                     0
+
+#define MMHUB_BASE__INST4_SEG0                     0
+#define MMHUB_BASE__INST4_SEG1                     0
+#define MMHUB_BASE__INST4_SEG2                     0
+#define MMHUB_BASE__INST4_SEG3                     0
+#define MMHUB_BASE__INST4_SEG4                     0
+#define MMHUB_BASE__INST4_SEG5                     0
+
+#define MMHUB_BASE__INST5_SEG0                     0
+#define MMHUB_BASE__INST5_SEG1                     0
+#define MMHUB_BASE__INST5_SEG2                     0
+#define MMHUB_BASE__INST5_SEG3                     0
+#define MMHUB_BASE__INST5_SEG4                     0
+#define MMHUB_BASE__INST5_SEG5                     0
+
+#define MP0_BASE__INST0_SEG0                       0x00016000
+#define MP0_BASE__INST0_SEG1                       0
+#define MP0_BASE__INST0_SEG2                       0
+#define MP0_BASE__INST0_SEG3                       0
+#define MP0_BASE__INST0_SEG4                       0
+#define MP0_BASE__INST0_SEG5                       0
+
+#define MP0_BASE__INST1_SEG0                       0
+#define MP0_BASE__INST1_SEG1                       0
+#define MP0_BASE__INST1_SEG2                       0
+#define MP0_BASE__INST1_SEG3                       0
+#define MP0_BASE__INST1_SEG4                       0
+#define MP0_BASE__INST1_SEG5                       0
+
+#define MP0_BASE__INST2_SEG0                       0
+#define MP0_BASE__INST2_SEG1                       0
+#define MP0_BASE__INST2_SEG2                       0
+#define MP0_BASE__INST2_SEG3                       0
+#define MP0_BASE__INST2_SEG4                       0
+#define MP0_BASE__INST2_SEG5                       0
+
+#define MP0_BASE__INST3_SEG0                       0
+#define MP0_BASE__INST3_SEG1                       0
+#define MP0_BASE__INST3_SEG2                       0
+#define MP0_BASE__INST3_SEG3                       0
+#define MP0_BASE__INST3_SEG4                       0
+#define MP0_BASE__INST3_SEG5                       0
+
+#define MP0_BASE__INST4_SEG0                       0
+#define MP0_BASE__INST4_SEG1                       0
+#define MP0_BASE__INST4_SEG2                       0
+#define MP0_BASE__INST4_SEG3                       0
+#define MP0_BASE__INST4_SEG4                       0
+#define MP0_BASE__INST4_SEG5                       0
+
+#define MP0_BASE__INST5_SEG0                       0
+#define MP0_BASE__INST5_SEG1                       0
+#define MP0_BASE__INST5_SEG2                       0
+#define MP0_BASE__INST5_SEG3                       0
+#define MP0_BASE__INST5_SEG4                       0
+#define MP0_BASE__INST5_SEG5                       0
+
+#define MP1_BASE__INST0_SEG0                       0x00016000
+#define MP1_BASE__INST0_SEG1                       0
+#define MP1_BASE__INST0_SEG2                       0
+#define MP1_BASE__INST0_SEG3                       0
+#define MP1_BASE__INST0_SEG4                       0
+#define MP1_BASE__INST0_SEG5                       0
+
+#define MP1_BASE__INST1_SEG0                       0
+#define MP1_BASE__INST1_SEG1                       0
+#define MP1_BASE__INST1_SEG2                       0
+#define MP1_BASE__INST1_SEG3                       0
+#define MP1_BASE__INST1_SEG4                       0
+#define MP1_BASE__INST1_SEG5                       0
+
+#define MP1_BASE__INST2_SEG0                       0
+#define MP1_BASE__INST2_SEG1                       0
+#define MP1_BASE__INST2_SEG2                       0
+#define MP1_BASE__INST2_SEG3                       0
+#define MP1_BASE__INST2_SEG4                       0
+#define MP1_BASE__INST2_SEG5                       0
+
+#define MP1_BASE__INST3_SEG0                       0
+#define MP1_BASE__INST3_SEG1                       0
+#define MP1_BASE__INST3_SEG2                       0
+#define MP1_BASE__INST3_SEG3                       0
+#define MP1_BASE__INST3_SEG4                       0
+#define MP1_BASE__INST3_SEG5                       0
+
+#define MP1_BASE__INST4_SEG0                       0
+#define MP1_BASE__INST4_SEG1                       0
+#define MP1_BASE__INST4_SEG2                       0
+#define MP1_BASE__INST4_SEG3                       0
+#define MP1_BASE__INST4_SEG4                       0
+#define MP1_BASE__INST4_SEG5                       0
+
+#define MP1_BASE__INST5_SEG0                       0
+#define MP1_BASE__INST5_SEG1                       0
+#define MP1_BASE__INST5_SEG2                       0
+#define MP1_BASE__INST5_SEG3                       0
+#define MP1_BASE__INST5_SEG4                       0
+#define MP1_BASE__INST5_SEG5                       0
+
+#define NBIO_BASE__INST0_SEG0                      0x00000000
+#define NBIO_BASE__INST0_SEG1                      0x00000014
+#define NBIO_BASE__INST0_SEG2                      0x00000D20
+#define NBIO_BASE__INST0_SEG3                      0x00010400
+#define NBIO_BASE__INST0_SEG4                      0
+#define NBIO_BASE__INST0_SEG5                      0
+
+#define NBIO_BASE__INST1_SEG0                      0
+#define NBIO_BASE__INST1_SEG1                      0
+#define NBIO_BASE__INST1_SEG2                      0
+#define NBIO_BASE__INST1_SEG3                      0
+#define NBIO_BASE__INST1_SEG4                      0
+#define NBIO_BASE__INST1_SEG5                      0
+
+#define NBIO_BASE__INST2_SEG0                      0
+#define NBIO_BASE__INST2_SEG1                      0
+#define NBIO_BASE__INST2_SEG2                      0
+#define NBIO_BASE__INST2_SEG3                      0
+#define NBIO_BASE__INST2_SEG4                      0
+#define NBIO_BASE__INST2_SEG5                      0
+
+#define NBIO_BASE__INST3_SEG0                      0
+#define NBIO_BASE__INST3_SEG1                      0
+#define NBIO_BASE__INST3_SEG2                      0
+#define NBIO_BASE__INST3_SEG3                      0
+#define NBIO_BASE__INST3_SEG4                      0
+#define NBIO_BASE__INST3_SEG5                      0
+
+#define NBIO_BASE__INST4_SEG0                      0
+#define NBIO_BASE__INST4_SEG1                      0
+#define NBIO_BASE__INST4_SEG2                      0
+#define NBIO_BASE__INST4_SEG3                      0
+#define NBIO_BASE__INST4_SEG4                      0
+#define NBIO_BASE__INST4_SEG5                      0
+
+#define NBIO_BASE__INST5_SEG0                      0
+#define NBIO_BASE__INST5_SEG1                      0
+#define NBIO_BASE__INST5_SEG2                      0
+#define NBIO_BASE__INST5_SEG3                      0
+#define NBIO_BASE__INST5_SEG4                      0
+#define NBIO_BASE__INST5_SEG5                      0
+
+#define OSSSYS_BASE__INST0_SEG0                    0x000010A0
+#define OSSSYS_BASE__INST0_SEG1                    0
+#define OSSSYS_BASE__INST0_SEG2                    0
+#define OSSSYS_BASE__INST0_SEG3                    0
+#define OSSSYS_BASE__INST0_SEG4                    0
+#define OSSSYS_BASE__INST0_SEG5                    0
+
+#define OSSSYS_BASE__INST1_SEG0                    0
+#define OSSSYS_BASE__INST1_SEG1                    0
+#define OSSSYS_BASE__INST1_SEG2                    0
+#define OSSSYS_BASE__INST1_SEG3                    0
+#define OSSSYS_BASE__INST1_SEG4                    0
+#define OSSSYS_BASE__INST1_SEG5                    0
+
+#define OSSSYS_BASE__INST2_SEG0                    0
+#define OSSSYS_BASE__INST2_SEG1                    0
+#define OSSSYS_BASE__INST2_SEG2                    0
+#define OSSSYS_BASE__INST2_SEG3                    0
+#define OSSSYS_BASE__INST2_SEG4                    0
+#define OSSSYS_BASE__INST2_SEG5                    0
+
+#define OSSSYS_BASE__INST3_SEG0                    0
+#define OSSSYS_BASE__INST3_SEG1                    0
+#define OSSSYS_BASE__INST3_SEG2                    0
+#define OSSSYS_BASE__INST3_SEG3                    0
+#define OSSSYS_BASE__INST3_SEG4                    0
+#define OSSSYS_BASE__INST3_SEG5                    0
+
+#define OSSSYS_BASE__INST4_SEG0                    0
+#define OSSSYS_BASE__INST4_SEG1                    0
+#define OSSSYS_BASE__INST4_SEG2                    0
+#define OSSSYS_BASE__INST4_SEG3                    0
+#define OSSSYS_BASE__INST4_SEG4                    0
+#define OSSSYS_BASE__INST4_SEG5                    0
+
+#define OSSSYS_BASE__INST5_SEG0                    0
+#define OSSSYS_BASE__INST5_SEG1                    0
+#define OSSSYS_BASE__INST5_SEG2                    0
+#define OSSSYS_BASE__INST5_SEG3                    0
+#define OSSSYS_BASE__INST5_SEG4                    0
+#define OSSSYS_BASE__INST5_SEG5                    0
+
+#define SDMA0_BASE__INST0_SEG0                     0x00001260
+#define SDMA0_BASE__INST0_SEG1                     0
+#define SDMA0_BASE__INST0_SEG2                     0
+#define SDMA0_BASE__INST0_SEG3                     0
+#define SDMA0_BASE__INST0_SEG4                     0
+#define SDMA0_BASE__INST0_SEG5                     0
+
+#define SDMA0_BASE__INST1_SEG0                     0
+#define SDMA0_BASE__INST1_SEG1                     0
+#define SDMA0_BASE__INST1_SEG2                     0
+#define SDMA0_BASE__INST1_SEG3                     0
+#define SDMA0_BASE__INST1_SEG4                     0
+#define SDMA0_BASE__INST1_SEG5                     0
+
+#define SDMA0_BASE__INST2_SEG0                     0
+#define SDMA0_BASE__INST2_SEG1                     0
+#define SDMA0_BASE__INST2_SEG2                     0
+#define SDMA0_BASE__INST2_SEG3                     0
+#define SDMA0_BASE__INST2_SEG4                     0
+#define SDMA0_BASE__INST2_SEG5                     0
+
+#define SDMA0_BASE__INST3_SEG0                     0
+#define SDMA0_BASE__INST3_SEG1                     0
+#define SDMA0_BASE__INST3_SEG2                     0
+#define SDMA0_BASE__INST3_SEG3                     0
+#define SDMA0_BASE__INST3_SEG4                     0
+#define SDMA0_BASE__INST3_SEG5                     0
+
+#define SDMA0_BASE__INST4_SEG0                     0
+#define SDMA0_BASE__INST4_SEG1                     0
+#define SDMA0_BASE__INST4_SEG2                     0
+#define SDMA0_BASE__INST4_SEG3                     0
+#define SDMA0_BASE__INST4_SEG4                     0
+#define SDMA0_BASE__INST4_SEG5                     0
+
+#define SDMA0_BASE__INST5_SEG0                     0
+#define SDMA0_BASE__INST5_SEG1                     0
+#define SDMA0_BASE__INST5_SEG2                     0
+#define SDMA0_BASE__INST5_SEG3                     0
+#define SDMA0_BASE__INST5_SEG4                     0
+#define SDMA0_BASE__INST5_SEG5                     0
+
+#define SDMA1_BASE__INST0_SEG0                     0x00001860
+#define SDMA1_BASE__INST0_SEG1                     0
+#define SDMA1_BASE__INST0_SEG2                     0
+#define SDMA1_BASE__INST0_SEG3                     0
+#define SDMA1_BASE__INST0_SEG4                     0
+#define SDMA1_BASE__INST0_SEG5                     0
+
+#define SDMA1_BASE__INST1_SEG0                     0
+#define SDMA1_BASE__INST1_SEG1                     0
+#define SDMA1_BASE__INST1_SEG2                     0
+#define SDMA1_BASE__INST1_SEG3                     0
+#define SDMA1_BASE__INST1_SEG4                     0
+#define SDMA1_BASE__INST1_SEG5                     0
+
+#define SDMA1_BASE__INST2_SEG0                     0
+#define SDMA1_BASE__INST2_SEG1                     0
+#define SDMA1_BASE__INST2_SEG2                     0
+#define SDMA1_BASE__INST2_SEG3                     0
+#define SDMA1_BASE__INST2_SEG4                     0
+#define SDMA1_BASE__INST2_SEG5                     0
+
+#define SDMA1_BASE__INST3_SEG0                     0
+#define SDMA1_BASE__INST3_SEG1                     0
+#define SDMA1_BASE__INST3_SEG2                     0
+#define SDMA1_BASE__INST3_SEG3                     0
+#define SDMA1_BASE__INST3_SEG4                     0
+#define SDMA1_BASE__INST3_SEG5                     0
+
+#define SDMA1_BASE__INST4_SEG0                     0
+#define SDMA1_BASE__INST4_SEG1                     0
+#define SDMA1_BASE__INST4_SEG2                     0
+#define SDMA1_BASE__INST4_SEG3                     0
+#define SDMA1_BASE__INST4_SEG4                     0
+#define SDMA1_BASE__INST4_SEG5                     0
+
+#define SDMA1_BASE__INST5_SEG0                     0
+#define SDMA1_BASE__INST5_SEG1                     0
+#define SDMA1_BASE__INST5_SEG2                     0
+#define SDMA1_BASE__INST5_SEG3                     0
+#define SDMA1_BASE__INST5_SEG4                     0
+#define SDMA1_BASE__INST5_SEG5                     0
+
+#define SMUIO_BASE__INST0_SEG0                     0x00016800
+#define SMUIO_BASE__INST0_SEG1                     0x00016A00
+#define SMUIO_BASE__INST0_SEG2                     0
+#define SMUIO_BASE__INST0_SEG3                     0
+#define SMUIO_BASE__INST0_SEG4                     0
+#define SMUIO_BASE__INST0_SEG5                     0
+
+#define SMUIO_BASE__INST1_SEG0                     0
+#define SMUIO_BASE__INST1_SEG1                     0
+#define SMUIO_BASE__INST1_SEG2                     0
+#define SMUIO_BASE__INST1_SEG3                     0
+#define SMUIO_BASE__INST1_SEG4                     0
+#define SMUIO_BASE__INST1_SEG5                     0
+
+#define SMUIO_BASE__INST2_SEG0                     0
+#define SMUIO_BASE__INST2_SEG1                     0
+#define SMUIO_BASE__INST2_SEG2                     0
+#define SMUIO_BASE__INST2_SEG3                     0
+#define SMUIO_BASE__INST2_SEG4                     0
+#define SMUIO_BASE__INST2_SEG5                     0
+
+#define SMUIO_BASE__INST3_SEG0                     0
+#define SMUIO_BASE__INST3_SEG1                     0
+#define SMUIO_BASE__INST3_SEG2                     0
+#define SMUIO_BASE__INST3_SEG3                     0
+#define SMUIO_BASE__INST3_SEG4                     0
+#define SMUIO_BASE__INST3_SEG5                     0
+
+#define SMUIO_BASE__INST4_SEG0                     0
+#define SMUIO_BASE__INST4_SEG1                     0
+#define SMUIO_BASE__INST4_SEG2                     0
+#define SMUIO_BASE__INST4_SEG3                     0
+#define SMUIO_BASE__INST4_SEG4                     0
+#define SMUIO_BASE__INST4_SEG5                     0
+
+#define SMUIO_BASE__INST5_SEG0                     0
+#define SMUIO_BASE__INST5_SEG1                     0
+#define SMUIO_BASE__INST5_SEG2                     0
+#define SMUIO_BASE__INST5_SEG3                     0
+#define SMUIO_BASE__INST5_SEG4                     0
+#define SMUIO_BASE__INST5_SEG5                     0
+
+#define THM_BASE__INST0_SEG0                       0x00016600
+#define THM_BASE__INST0_SEG1                       0
+#define THM_BASE__INST0_SEG2                       0
+#define THM_BASE__INST0_SEG3                       0
+#define THM_BASE__INST0_SEG4                       0
+#define THM_BASE__INST0_SEG5                       0
+
+#define THM_BASE__INST1_SEG0                       0
+#define THM_BASE__INST1_SEG1                       0
+#define THM_BASE__INST1_SEG2                       0
+#define THM_BASE__INST1_SEG3                       0
+#define THM_BASE__INST1_SEG4                       0
+#define THM_BASE__INST1_SEG5                       0
+
+#define THM_BASE__INST2_SEG0                       0
+#define THM_BASE__INST2_SEG1                       0
+#define THM_BASE__INST2_SEG2                       0
+#define THM_BASE__INST2_SEG3                       0
+#define THM_BASE__INST2_SEG4                       0
+#define THM_BASE__INST2_SEG5                       0
+
+#define THM_BASE__INST3_SEG0                       0
+#define THM_BASE__INST3_SEG1                       0
+#define THM_BASE__INST3_SEG2                       0
+#define THM_BASE__INST3_SEG3                       0
+#define THM_BASE__INST3_SEG4                       0
+#define THM_BASE__INST3_SEG5                       0
+
+#define THM_BASE__INST4_SEG0                       0
+#define THM_BASE__INST4_SEG1                       0
+#define THM_BASE__INST4_SEG2                       0
+#define THM_BASE__INST4_SEG3                       0
+#define THM_BASE__INST4_SEG4                       0
+#define THM_BASE__INST4_SEG5                       0
+
+#define THM_BASE__INST5_SEG0                       0
+#define THM_BASE__INST5_SEG1                       0
+#define THM_BASE__INST5_SEG2                       0
+#define THM_BASE__INST5_SEG3                       0
+#define THM_BASE__INST5_SEG4                       0
+#define THM_BASE__INST5_SEG5                       0
+
+#define UMC_BASE__INST0_SEG0                       0x00014000
+#define UMC_BASE__INST0_SEG1                       0
+#define UMC_BASE__INST0_SEG2                       0
+#define UMC_BASE__INST0_SEG3                       0
+#define UMC_BASE__INST0_SEG4                       0
+#define UMC_BASE__INST0_SEG5                       0
+
+#define UMC_BASE__INST1_SEG0                       0
+#define UMC_BASE__INST1_SEG1                       0
+#define UMC_BASE__INST1_SEG2                       0
+#define UMC_BASE__INST1_SEG3                       0
+#define UMC_BASE__INST1_SEG4                       0
+#define UMC_BASE__INST1_SEG5                       0
+
+#define UMC_BASE__INST2_SEG0                       0
+#define UMC_BASE__INST2_SEG1                       0
+#define UMC_BASE__INST2_SEG2                       0
+#define UMC_BASE__INST2_SEG3                       0
+#define UMC_BASE__INST2_SEG4                       0
+#define UMC_BASE__INST2_SEG5                       0
+
+#define UMC_BASE__INST3_SEG0                       0
+#define UMC_BASE__INST3_SEG1                       0
+#define UMC_BASE__INST3_SEG2                       0
+#define UMC_BASE__INST3_SEG3                       0
+#define UMC_BASE__INST3_SEG4                       0
+#define UMC_BASE__INST3_SEG5                       0
+
+#define UMC_BASE__INST4_SEG0                       0
+#define UMC_BASE__INST4_SEG1                       0
+#define UMC_BASE__INST4_SEG2                       0
+#define UMC_BASE__INST4_SEG3                       0
+#define UMC_BASE__INST4_SEG4                       0
+#define UMC_BASE__INST4_SEG5                       0
+
+#define UMC_BASE__INST5_SEG0                       0
+#define UMC_BASE__INST5_SEG1                       0
+#define UMC_BASE__INST5_SEG2                       0
+#define UMC_BASE__INST5_SEG3                       0
+#define UMC_BASE__INST5_SEG4                       0
+#define UMC_BASE__INST5_SEG5                       0
+
+#define UVD_BASE__INST0_SEG0                       0x00007800
+#define UVD_BASE__INST0_SEG1                       0x00007E00
+#define UVD_BASE__INST0_SEG2                       0
+#define UVD_BASE__INST0_SEG3                       0
+#define UVD_BASE__INST0_SEG4                       0
+#define UVD_BASE__INST0_SEG5                       0
+
+#define UVD_BASE__INST1_SEG0                       0
+#define UVD_BASE__INST1_SEG1                       0x00009000
+#define UVD_BASE__INST1_SEG2                       0
+#define UVD_BASE__INST1_SEG3                       0
+#define UVD_BASE__INST1_SEG4                       0
+#define UVD_BASE__INST1_SEG5                       0
+
+#define UVD_BASE__INST2_SEG0                       0
+#define UVD_BASE__INST2_SEG1                       0
+#define UVD_BASE__INST2_SEG2                       0
+#define UVD_BASE__INST2_SEG3                       0
+#define UVD_BASE__INST2_SEG4                       0
+#define UVD_BASE__INST2_SEG5                       0
+
+#define UVD_BASE__INST3_SEG0                       0
+#define UVD_BASE__INST3_SEG1                       0
+#define UVD_BASE__INST3_SEG2                       0
+#define UVD_BASE__INST3_SEG3                       0
+#define UVD_BASE__INST3_SEG4                       0
+#define UVD_BASE__INST3_SEG5                       0
+
+#define UVD_BASE__INST4_SEG0                       0
+#define UVD_BASE__INST4_SEG1                       0
+#define UVD_BASE__INST4_SEG2                       0
+#define UVD_BASE__INST4_SEG3                       0
+#define UVD_BASE__INST4_SEG4                       0
+#define UVD_BASE__INST4_SEG5                       0
+
+#define UVD_BASE__INST5_SEG0                       0
+#define UVD_BASE__INST5_SEG1                       0
+#define UVD_BASE__INST5_SEG2                       0
+#define UVD_BASE__INST5_SEG3                       0
+#define UVD_BASE__INST5_SEG4                       0
+#define UVD_BASE__INST5_SEG5                       0
+
+#define VCE_BASE__INST0_SEG0                       0x00008800
+#define VCE_BASE__INST0_SEG1                       0
+#define VCE_BASE__INST0_SEG2                       0
+#define VCE_BASE__INST0_SEG3                       0
+#define VCE_BASE__INST0_SEG4                       0
+#define VCE_BASE__INST0_SEG5                       0
+
+#define VCE_BASE__INST1_SEG0                       0
+#define VCE_BASE__INST1_SEG1                       0
+#define VCE_BASE__INST1_SEG2                       0
+#define VCE_BASE__INST1_SEG3                       0
+#define VCE_BASE__INST1_SEG4                       0
+#define VCE_BASE__INST1_SEG5                       0
+
+#define VCE_BASE__INST2_SEG0                       0
+#define VCE_BASE__INST2_SEG1                       0
+#define VCE_BASE__INST2_SEG2                       0
+#define VCE_BASE__INST2_SEG3                       0
+#define VCE_BASE__INST2_SEG4                       0
+#define VCE_BASE__INST2_SEG5                       0
+
+#define VCE_BASE__INST3_SEG0                       0
+#define VCE_BASE__INST3_SEG1                       0
+#define VCE_BASE__INST3_SEG2                       0
+#define VCE_BASE__INST3_SEG3                       0
+#define VCE_BASE__INST3_SEG4                       0
+#define VCE_BASE__INST3_SEG5                       0
+
+#define VCE_BASE__INST4_SEG0                       0
+#define VCE_BASE__INST4_SEG1                       0
+#define VCE_BASE__INST4_SEG2                       0
+#define VCE_BASE__INST4_SEG3                       0
+#define VCE_BASE__INST4_SEG4                       0
+#define VCE_BASE__INST4_SEG5                       0
+
+#define VCE_BASE__INST5_SEG0                       0
+#define VCE_BASE__INST5_SEG1                       0
+#define VCE_BASE__INST5_SEG2                       0
+#define VCE_BASE__INST5_SEG3                       0
+#define VCE_BASE__INST5_SEG4                       0
+#define VCE_BASE__INST5_SEG5                       0
+
+#define XDMA_BASE__INST0_SEG0                      0x00003400
+#define XDMA_BASE__INST0_SEG1                      0
+#define XDMA_BASE__INST0_SEG2                      0
+#define XDMA_BASE__INST0_SEG3                      0
+#define XDMA_BASE__INST0_SEG4                      0
+#define XDMA_BASE__INST0_SEG5                      0
+
+#define XDMA_BASE__INST1_SEG0                      0
+#define XDMA_BASE__INST1_SEG1                      0
+#define XDMA_BASE__INST1_SEG2                      0
+#define XDMA_BASE__INST1_SEG3                      0
+#define XDMA_BASE__INST1_SEG4                      0
+#define XDMA_BASE__INST1_SEG5                      0
+
+#define XDMA_BASE__INST2_SEG0                      0
+#define XDMA_BASE__INST2_SEG1                      0
+#define XDMA_BASE__INST2_SEG2                      0
+#define XDMA_BASE__INST2_SEG3                      0
+#define XDMA_BASE__INST2_SEG4                      0
+#define XDMA_BASE__INST2_SEG5                      0
+
+#define XDMA_BASE__INST3_SEG0                      0
+#define XDMA_BASE__INST3_SEG1                      0
+#define XDMA_BASE__INST3_SEG2                      0
+#define XDMA_BASE__INST3_SEG3                      0
+#define XDMA_BASE__INST3_SEG4                      0
+#define XDMA_BASE__INST3_SEG5                      0
+
+#define XDMA_BASE__INST4_SEG0                      0
+#define XDMA_BASE__INST4_SEG1                      0
+#define XDMA_BASE__INST4_SEG2                      0
+#define XDMA_BASE__INST4_SEG3                      0
+#define XDMA_BASE__INST4_SEG4                      0
+#define XDMA_BASE__INST4_SEG5                      0
+
+#define XDMA_BASE__INST5_SEG0                      0
+#define XDMA_BASE__INST5_SEG1                      0
+#define XDMA_BASE__INST5_SEG2                      0
+#define XDMA_BASE__INST5_SEG3                      0
+#define XDMA_BASE__INST5_SEG4                      0
+#define XDMA_BASE__INST5_SEG5                      0
+
+#define RSMU_BASE__INST0_SEG0                      0x00012000
+#define RSMU_BASE__INST0_SEG1                      0
+#define RSMU_BASE__INST0_SEG2                      0
+#define RSMU_BASE__INST0_SEG3                      0
+#define RSMU_BASE__INST0_SEG4                      0
+#define RSMU_BASE__INST0_SEG5                      0
+
+#define RSMU_BASE__INST1_SEG0                      0
+#define RSMU_BASE__INST1_SEG1                      0
+#define RSMU_BASE__INST1_SEG2                      0
+#define RSMU_BASE__INST1_SEG3                      0
+#define RSMU_BASE__INST1_SEG4                      0
+#define RSMU_BASE__INST1_SEG5                      0
+
+#define RSMU_BASE__INST2_SEG0                      0
+#define RSMU_BASE__INST2_SEG1                      0
+#define RSMU_BASE__INST2_SEG2                      0
+#define RSMU_BASE__INST2_SEG3                      0
+#define RSMU_BASE__INST2_SEG4                      0
+#define RSMU_BASE__INST2_SEG5                      0
+
+#define RSMU_BASE__INST3_SEG0                      0
+#define RSMU_BASE__INST3_SEG1                      0
+#define RSMU_BASE__INST3_SEG2                      0
+#define RSMU_BASE__INST3_SEG3                      0
+#define RSMU_BASE__INST3_SEG4                      0
+#define RSMU_BASE__INST3_SEG5                      0
+
+#define RSMU_BASE__INST4_SEG0                      0
+#define RSMU_BASE__INST4_SEG1                      0
+#define RSMU_BASE__INST4_SEG2                      0
+#define RSMU_BASE__INST4_SEG3                      0
+#define RSMU_BASE__INST4_SEG4                      0
+#define RSMU_BASE__INST4_SEG5                      0
+
+#define RSMU_BASE__INST5_SEG0                      0
+#define RSMU_BASE__INST5_SEG1                      0
+#define RSMU_BASE__INST5_SEG2                      0
+#define RSMU_BASE__INST5_SEG3                      0
+#define RSMU_BASE__INST5_SEG4                      0
+#define RSMU_BASE__INST5_SEG5                      0
+
+#endif
+
diff --git a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
index 7e8ad30d98e2..75c208283e5f 100644
--- a/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
+++ b/drivers/gpu/drm/amd/powerplay/amd_powerplay.c
@@ -25,30 +25,16 @@
 #include <linux/kernel.h>
 #include <linux/gfp.h>
 #include <linux/slab.h>
+#include <linux/firmware.h>
 #include "amd_shared.h"
 #include "amd_powerplay.h"
 #include "power_state.h"
 #include "amdgpu.h"
 #include "hwmgr.h"
 
-#define PP_DPM_DISABLED 0xCCCC
-
-static int pp_dpm_dispatch_tasks(void *handle, enum amd_pp_task task_id,
-		enum amd_pm_state_type *user_state);
 
 static const struct amd_pm_funcs pp_dpm_funcs;
 
-static inline int pp_check(struct pp_hwmgr *hwmgr)
-{
-	if (hwmgr == NULL || hwmgr->smumgr_funcs == NULL)
-		return -EINVAL;
-
-	if (hwmgr->pm_en == 0 || hwmgr->hwmgr_func == NULL)
-		return PP_DPM_DISABLED;
-
-	return 0;
-}
-
 static int amd_powerplay_create(struct amdgpu_device *adev)
 {
 	struct pp_hwmgr *hwmgr;
@@ -61,19 +47,21 @@ static int amd_powerplay_create(struct amdgpu_device *adev)
 		return -ENOMEM;
 
 	hwmgr->adev = adev;
-	hwmgr->pm_en = (amdgpu_dpm != 0 && !amdgpu_sriov_vf(adev)) ? true : false;
+	hwmgr->not_vf = !amdgpu_sriov_vf(adev);
+	hwmgr->pm_en = (amdgpu_dpm && hwmgr->not_vf) ? true : false;
 	hwmgr->device = amdgpu_cgs_create_device(adev);
 	mutex_init(&hwmgr->smu_lock);
 	hwmgr->chip_family = adev->family;
 	hwmgr->chip_id = adev->asic_type;
-	hwmgr->feature_mask = amdgpu_pp_feature_mask;
+	hwmgr->feature_mask = adev->powerplay.pp_feature;
+	hwmgr->display_config = &adev->pm.pm_display_cfg;
 	adev->powerplay.pp_handle = hwmgr;
 	adev->powerplay.pp_funcs = &pp_dpm_funcs;
 	return 0;
 }
 
 
-static int amd_powerplay_destroy(struct amdgpu_device *adev)
+static void amd_powerplay_destroy(struct amdgpu_device *adev)
 {
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
 
@@ -82,8 +70,6 @@ static int amd_powerplay_destroy(struct amdgpu_device *adev)
 
 	kfree(hwmgr);
 	hwmgr = NULL;
-
-	return 0;
 }
 
 static int pp_early_init(void *handle)
@@ -109,18 +95,9 @@ static int pp_sw_init(void *handle)
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret >= 0) {
-		if (hwmgr->smumgr_funcs->smu_init == NULL)
-			return -EINVAL;
-
-		ret = hwmgr->smumgr_funcs->smu_init(hwmgr);
+	ret = hwmgr_sw_init(hwmgr);
 
-		phm_register_irq_handlers(hwmgr);
-
-		pr_debug("amdgpu: powerplay sw initialized\n");
-	}
+	pr_debug("powerplay sw init %s\n", ret ? "failed" : "successfully");
 
 	return ret;
 }
@@ -129,16 +106,14 @@ static int pp_sw_fini(void *handle)
 {
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
-	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret >= 0) {
-		if (hwmgr->smumgr_funcs->smu_fini != NULL)
-			hwmgr->smumgr_funcs->smu_fini(hwmgr);
-	}
+	hwmgr_sw_fini(hwmgr);
 
-	if (adev->firmware.load_type == AMDGPU_FW_LOAD_SMU)
+	if (adev->firmware.load_type == AMDGPU_FW_LOAD_SMU) {
+		release_firmware(adev->pm.fw);
+		adev->pm.fw = NULL;
 		amdgpu_ucode_fini_bo(adev);
+	}
 
 	return 0;
 }
@@ -152,55 +127,68 @@ static int pp_hw_init(void *handle)
 	if (adev->firmware.load_type == AMDGPU_FW_LOAD_SMU)
 		amdgpu_ucode_init_bo(adev);
 
-	ret = pp_check(hwmgr);
+	ret = hwmgr_hw_init(hwmgr);
 
-	if (ret >= 0) {
-		if (hwmgr->smumgr_funcs->start_smu == NULL)
-			return -EINVAL;
+	if (ret)
+		pr_err("powerplay hw init failed\n");
 
-		if (hwmgr->smumgr_funcs->start_smu(hwmgr)) {
-			pr_err("smc start failed\n");
-			hwmgr->smumgr_funcs->smu_fini(hwmgr);
-			return -EINVAL;
-		}
-		if (ret == PP_DPM_DISABLED)
-			goto exit;
-		ret = hwmgr_hw_init(hwmgr);
-		if (ret)
-			goto exit;
-	}
 	return ret;
-exit:
-	hwmgr->pm_en = 0;
-	cgs_notify_dpm_enabled(hwmgr->device, false);
-	return 0;
-
 }
 
 static int pp_hw_fini(void *handle)
 {
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
-	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret == 0)
-		hwmgr_hw_fini(hwmgr);
+	hwmgr_hw_fini(hwmgr);
 
 	return 0;
 }
 
+static void pp_reserve_vram_for_smu(struct amdgpu_device *adev)
+{
+	int r = -EINVAL;
+	void *cpu_ptr = NULL;
+	uint64_t gpu_addr;
+	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
+
+	if (amdgpu_bo_create_kernel(adev, adev->pm.smu_prv_buffer_size,
+						PAGE_SIZE, AMDGPU_GEM_DOMAIN_GTT,
+						&adev->pm.smu_prv_buffer,
+						&gpu_addr,
+						&cpu_ptr)) {
+		DRM_ERROR("amdgpu: failed to create smu prv buffer\n");
+		return;
+	}
+
+	if (hwmgr->hwmgr_func->notify_cac_buffer_info)
+		r = hwmgr->hwmgr_func->notify_cac_buffer_info(hwmgr,
+					lower_32_bits((unsigned long)cpu_ptr),
+					upper_32_bits((unsigned long)cpu_ptr),
+					lower_32_bits(gpu_addr),
+					upper_32_bits(gpu_addr),
+					adev->pm.smu_prv_buffer_size);
+
+	if (r) {
+		amdgpu_bo_free_kernel(&adev->pm.smu_prv_buffer, NULL, NULL);
+		adev->pm.smu_prv_buffer = NULL;
+		DRM_ERROR("amdgpu: failed to notify SMU buffer address\n");
+	}
+}
+
 static int pp_late_init(void *handle)
 {
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
-	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret == 0)
-		pp_dpm_dispatch_tasks(hwmgr,
+	if (hwmgr && hwmgr->pm_en) {
+		mutex_lock(&hwmgr->smu_lock);
+		hwmgr_handle_task(hwmgr,
 					AMD_PP_TASK_COMPLETE_INIT, NULL);
+		mutex_unlock(&hwmgr->smu_lock);
+	}
+	if (adev->pm.smu_prv_buffer_size != 0)
+		pp_reserve_vram_for_smu(adev);
 
 	return 0;
 }
@@ -209,6 +197,8 @@ static void pp_late_fini(void *handle)
 {
 	struct amdgpu_device *adev = handle;
 
+	if (adev->pm.smu_prv_buffer)
+		amdgpu_bo_free_kernel(&adev->pm.smu_prv_buffer, NULL, NULL);
 	amd_powerplay_destroy(adev);
 }
 
@@ -231,61 +221,23 @@ static int pp_sw_reset(void *handle)
 static int pp_set_powergating_state(void *handle,
 				    enum amd_powergating_state state)
 {
-	struct amdgpu_device *adev = handle;
-	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
-
-	if (hwmgr->hwmgr_func->enable_per_cu_power_gating == NULL) {
-		pr_info("%s was not implemented.\n", __func__);
-		return 0;
-	}
-
-	/* Enable/disable GFX per cu powergating through SMU */
-	return hwmgr->hwmgr_func->enable_per_cu_power_gating(hwmgr,
-			state == AMD_PG_STATE_GATE);
+	return 0;
 }
 
 static int pp_suspend(void *handle)
 {
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
-	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret == 0)
-		hwmgr_hw_suspend(hwmgr);
-	return 0;
+	return hwmgr_suspend(hwmgr);
 }
 
 static int pp_resume(void *handle)
 {
 	struct amdgpu_device *adev = handle;
 	struct pp_hwmgr *hwmgr = adev->powerplay.pp_handle;
-	int ret;
-
-	ret = pp_check(hwmgr);
-
-	if (ret < 0)
-		return ret;
-
-	if (hwmgr->smumgr_funcs->start_smu == NULL)
-		return -EINVAL;
-
-	if (hwmgr->smumgr_funcs->start_smu(hwmgr)) {
-		pr_err("smc start failed\n");
-		hwmgr->smumgr_funcs->smu_fini(hwmgr);
-		return -EINVAL;
-	}
 
-	if (ret == PP_DPM_DISABLED)
-		return 0;
-
-	return hwmgr_hw_resume(hwmgr);
+	return hwmgr_resume(hwmgr);
 }
 
 static int pp_set_clockgating_state(void *handle,
@@ -334,12 +286,9 @@ static int pp_dpm_fw_loading_complete(void *handle)
 static int pp_set_clockgating_by_smu(void *handle, uint32_t msg_id)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->update_clock_gatings == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -362,10 +311,10 @@ static void pp_dpm_en_umd_pstate(struct pp_hwmgr  *hwmgr,
 		if (*level & profile_mode_mask) {
 			hwmgr->saved_dpm_level = hwmgr->dpm_level;
 			hwmgr->en_umd_pstate = true;
-			cgs_set_clockgating_state(hwmgr->device,
+			amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_GFX,
 						AMD_CG_STATE_UNGATE);
-			cgs_set_powergating_state(hwmgr->device,
+			amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 					AMD_IP_BLOCK_TYPE_GFX,
 					AMD_PG_STATE_UNGATE);
 		}
@@ -375,10 +324,10 @@ static void pp_dpm_en_umd_pstate(struct pp_hwmgr  *hwmgr,
 			if (*level == AMD_DPM_FORCED_LEVEL_PROFILE_EXIT)
 				*level = hwmgr->saved_dpm_level;
 			hwmgr->en_umd_pstate = false;
-			cgs_set_clockgating_state(hwmgr->device,
+			amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 					AMD_IP_BLOCK_TYPE_GFX,
 					AMD_CG_STATE_GATE);
-			cgs_set_powergating_state(hwmgr->device,
+			amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 					AMD_IP_BLOCK_TYPE_GFX,
 					AMD_PG_STATE_GATE);
 		}
@@ -389,12 +338,9 @@ static int pp_dpm_force_performance_level(void *handle,
 					enum amd_dpm_forced_level level)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
 
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (level == hwmgr->dpm_level)
 		return 0;
@@ -412,13 +358,10 @@ static enum amd_dpm_forced_level pp_dpm_get_performance_level(
 								void *handle)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 	enum amd_dpm_forced_level level;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
 	level = hwmgr->dpm_level;
@@ -429,13 +372,10 @@ static enum amd_dpm_forced_level pp_dpm_get_performance_level(
 static uint32_t pp_dpm_get_sclk(void *handle, bool low)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 	uint32_t clk = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return 0;
 
 	if (hwmgr->hwmgr_func->get_sclk == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -450,13 +390,10 @@ static uint32_t pp_dpm_get_sclk(void *handle, bool low)
 static uint32_t pp_dpm_get_mclk(void *handle, bool low)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 	uint32_t clk = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return 0;
 
 	if (hwmgr->hwmgr_func->get_mclk == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -471,11 +408,8 @@ static uint32_t pp_dpm_get_mclk(void *handle, bool low)
 static void pp_dpm_powergate_vce(void *handle, bool gate)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
+	if (!hwmgr || !hwmgr->pm_en)
 		return;
 
 	if (hwmgr->hwmgr_func->powergate_vce == NULL) {
@@ -490,11 +424,8 @@ static void pp_dpm_powergate_vce(void *handle, bool gate)
 static void pp_dpm_powergate_uvd(void *handle, bool gate)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
+	if (!hwmgr || !hwmgr->pm_en)
 		return;
 
 	if (hwmgr->hwmgr_func->powergate_uvd == NULL) {
@@ -512,10 +443,8 @@ static int pp_dpm_dispatch_tasks(void *handle, enum amd_pp_task task_id,
 	int ret = 0;
 	struct pp_hwmgr *hwmgr = handle;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
 	ret = hwmgr_handle_task(hwmgr, task_id, user_state);
@@ -528,15 +457,9 @@ static enum amd_pm_state_type pp_dpm_get_current_power_state(void *handle)
 {
 	struct pp_hwmgr *hwmgr = handle;
 	struct pp_power_state *state;
-	int ret = 0;
 	enum amd_pm_state_type pm_type;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
-
-	if (hwmgr->current_ps == NULL)
+	if (!hwmgr || !hwmgr->pm_en || !hwmgr->current_ps)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -568,11 +491,8 @@ static enum amd_pm_state_type pp_dpm_get_current_power_state(void *handle)
 static void pp_dpm_set_fan_control_mode(void *handle, uint32_t mode)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
 
-	if (ret)
+	if (!hwmgr || !hwmgr->pm_en)
 		return;
 
 	if (hwmgr->hwmgr_func->set_fan_control_mode == NULL) {
@@ -587,13 +507,10 @@ static void pp_dpm_set_fan_control_mode(void *handle, uint32_t mode)
 static uint32_t pp_dpm_get_fan_control_mode(void *handle)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 	uint32_t mode = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return 0;
 
 	if (hwmgr->hwmgr_func->get_fan_control_mode == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -610,10 +527,8 @@ static int pp_dpm_set_fan_speed_percent(void *handle, uint32_t percent)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->set_fan_speed_percent == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -630,10 +545,8 @@ static int pp_dpm_get_fan_speed_percent(void *handle, uint32_t *speed)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->get_fan_speed_percent == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -651,10 +564,8 @@ static int pp_dpm_get_fan_speed_rpm(void *handle, uint32_t *rpm)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->get_fan_speed_rpm == NULL)
 		return -EINVAL;
@@ -670,16 +581,10 @@ static int pp_dpm_get_pp_num_states(void *handle,
 {
 	struct pp_hwmgr *hwmgr = handle;
 	int i;
-	int ret = 0;
 
 	memset(data, 0, sizeof(*data));
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
-
-	if (hwmgr->ps == NULL)
+	if (!hwmgr || !hwmgr->pm_en ||!hwmgr->ps)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -713,15 +618,9 @@ static int pp_dpm_get_pp_num_states(void *handle,
 static int pp_dpm_get_pp_table(void *handle, char **table)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 	int size = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
-
-	if (!hwmgr->soft_pp_table)
+	if (!hwmgr || !hwmgr->pm_en ||!hwmgr->soft_pp_table)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -736,10 +635,6 @@ static int amd_powerplay_reset(void *handle)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret;
 
-	ret = pp_check(hwmgr);
-	if (ret)
-		return ret;
-
 	ret = hwmgr_hw_fini(hwmgr);
 	if (ret)
 		return ret;
@@ -754,40 +649,38 @@ static int amd_powerplay_reset(void *handle)
 static int pp_dpm_set_pp_table(void *handle, const char *buf, size_t size)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
+	int ret = -ENOMEM;
 
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
 	if (!hwmgr->hardcode_pp_table) {
 		hwmgr->hardcode_pp_table = kmemdup(hwmgr->soft_pp_table,
 						   hwmgr->soft_pp_table_size,
 						   GFP_KERNEL);
-		if (!hwmgr->hardcode_pp_table) {
-			mutex_unlock(&hwmgr->smu_lock);
-			return -ENOMEM;
-		}
+		if (!hwmgr->hardcode_pp_table)
+			goto err;
 	}
 
 	memcpy(hwmgr->hardcode_pp_table, buf, size);
 
 	hwmgr->soft_pp_table = hwmgr->hardcode_pp_table;
-	mutex_unlock(&hwmgr->smu_lock);
 
 	ret = amd_powerplay_reset(handle);
 	if (ret)
-		return ret;
+		goto err;
 
 	if (hwmgr->hwmgr_func->avfs_control) {
 		ret = hwmgr->hwmgr_func->avfs_control(hwmgr, false);
 		if (ret)
-			return ret;
+			goto err;
 	}
-
+	mutex_unlock(&hwmgr->smu_lock);
 	return 0;
+err:
+	mutex_unlock(&hwmgr->smu_lock);
+	return ret;
 }
 
 static int pp_dpm_force_clock_level(void *handle,
@@ -796,10 +689,8 @@ static int pp_dpm_force_clock_level(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->force_clock_level == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -820,10 +711,8 @@ static int pp_dpm_print_clock_levels(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->print_clock_levels == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -840,10 +729,8 @@ static int pp_dpm_get_sclk_od(void *handle)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->get_sclk_od == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -860,10 +747,8 @@ static int pp_dpm_set_sclk_od(void *handle, uint32_t value)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->set_sclk_od == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -881,10 +766,8 @@ static int pp_dpm_get_mclk_od(void *handle)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->get_mclk_od == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -901,10 +784,8 @@ static int pp_dpm_set_mclk_od(void *handle, uint32_t value)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->set_mclk_od == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -922,11 +803,7 @@ static int pp_dpm_read_sensor(void *handle, int idx,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret)
-		return ret;
-
-	if (value == NULL)
+	if (!hwmgr || !hwmgr->pm_en || !value)
 		return -EINVAL;
 
 	switch (idx) {
@@ -948,14 +825,11 @@ static struct amd_vce_state*
 pp_dpm_get_vce_clock_state(void *handle, unsigned idx)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
 
-	if (ret)
+	if (!hwmgr || !hwmgr->pm_en)
 		return NULL;
 
-	if (hwmgr && idx < hwmgr->num_vce_state_tables)
+	if (idx < hwmgr->num_vce_state_tables)
 		return &hwmgr->vce_states[idx];
 	return NULL;
 }
@@ -964,7 +838,7 @@ static int pp_get_power_profile_mode(void *handle, char *buf)
 {
 	struct pp_hwmgr *hwmgr = handle;
 
-	if (!buf || pp_check(hwmgr))
+	if (!hwmgr || !hwmgr->pm_en || !buf)
 		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->get_power_profile_mode == NULL) {
@@ -980,12 +854,12 @@ static int pp_set_power_profile_mode(void *handle, long *input, uint32_t size)
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = -EINVAL;
 
-	if (pp_check(hwmgr))
-		return -EINVAL;
+	if (!hwmgr || !hwmgr->pm_en)
+		return ret;
 
 	if (hwmgr->hwmgr_func->set_power_profile_mode == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
-		return -EINVAL;
+		return ret;
 	}
 	mutex_lock(&hwmgr->smu_lock);
 	if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_MANUAL)
@@ -998,7 +872,7 @@ static int pp_odn_edit_dpm_table(void *handle, uint32_t type, long *input, uint3
 {
 	struct pp_hwmgr *hwmgr = handle;
 
-	if (pp_check(hwmgr))
+	if (!hwmgr || !hwmgr->pm_en)
 		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->odn_edit_dpm_table == NULL) {
@@ -1016,7 +890,7 @@ static int pp_dpm_switch_power_profile(void *handle,
 	long workload;
 	uint32_t index;
 
-	if (pp_check(hwmgr))
+	if (!hwmgr || !hwmgr->pm_en)
 		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->set_power_profile_mode == NULL) {
@@ -1048,46 +922,12 @@ static int pp_dpm_switch_power_profile(void *handle,
 	return 0;
 }
 
-static int pp_dpm_notify_smu_memory_info(void *handle,
-					uint32_t virtual_addr_low,
-					uint32_t virtual_addr_hi,
-					uint32_t mc_addr_low,
-					uint32_t mc_addr_hi,
-					uint32_t size)
-{
-	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
-
-	if (hwmgr->hwmgr_func->notify_cac_buffer_info == NULL) {
-		pr_info("%s was not implemented.\n", __func__);
-		return -EINVAL;
-	}
-
-	mutex_lock(&hwmgr->smu_lock);
-
-	ret = hwmgr->hwmgr_func->notify_cac_buffer_info(hwmgr, virtual_addr_low,
-					virtual_addr_hi, mc_addr_low, mc_addr_hi,
-					size);
-
-	mutex_unlock(&hwmgr->smu_lock);
-
-	return ret;
-}
-
 static int pp_set_power_limit(void *handle, uint32_t limit)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (hwmgr->hwmgr_func->set_power_limit == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
@@ -1104,20 +944,14 @@ static int pp_set_power_limit(void *handle, uint32_t limit)
 	hwmgr->hwmgr_func->set_power_limit(hwmgr, limit);
 	hwmgr->power_limit = limit;
 	mutex_unlock(&hwmgr->smu_lock);
-	return ret;
+	return 0;
 }
 
 static int pp_get_power_limit(void *handle, uint32_t *limit, bool default_limit)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
 
-	if (limit == NULL)
+	if (!hwmgr || !hwmgr->pm_en ||!limit)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1129,19 +963,16 @@ static int pp_get_power_limit(void *handle, uint32_t *limit, bool default_limit)
 
 	mutex_unlock(&hwmgr->smu_lock);
 
-	return ret;
+	return 0;
 }
 
 static int pp_display_configuration_change(void *handle,
 	const struct amd_pp_display_configuration *display_config)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
-
-	ret = pp_check(hwmgr);
 
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
 	phm_store_dal_configuration_data(hwmgr, display_config);
@@ -1155,12 +986,7 @@ static int pp_get_display_power_level(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
-
-	if (output == NULL)
+	if (!hwmgr || !hwmgr->pm_en ||!output)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1177,10 +1003,8 @@ static int pp_get_current_clocks(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
 
@@ -1225,10 +1049,8 @@ static int pp_get_clock_by_type(void *handle, enum amd_pp_clock_type type, struc
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
 	if (clocks == NULL)
 		return -EINVAL;
@@ -1246,11 +1068,7 @@ static int pp_get_clock_by_type_with_latency(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret)
-		return ret;
-
-	if (!clocks)
+	if (!hwmgr || !hwmgr->pm_en ||!clocks)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1266,11 +1084,7 @@ static int pp_get_clock_by_type_with_voltage(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret)
-		return ret;
-
-	if (!clocks)
+	if (!hwmgr || !hwmgr->pm_en ||!clocks)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1282,21 +1096,17 @@ static int pp_get_clock_by_type_with_voltage(void *handle,
 }
 
 static int pp_set_watermarks_for_clocks_ranges(void *handle,
-		struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges)
+		void *clock_ranges)
 {
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret)
-		return ret;
-
-	if (!wm_with_clock_ranges)
+	if (!hwmgr || !hwmgr->pm_en || !clock_ranges)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
 	ret = phm_set_watermarks_for_clocks_ranges(hwmgr,
-			wm_with_clock_ranges);
+			clock_ranges);
 	mutex_unlock(&hwmgr->smu_lock);
 
 	return ret;
@@ -1308,11 +1118,7 @@ static int pp_display_clock_voltage_request(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-	if (ret)
-		return ret;
-
-	if (!clock)
+	if (!hwmgr || !hwmgr->pm_en ||!clock)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1328,12 +1134,7 @@ static int pp_get_display_mode_validation_clocks(void *handle,
 	struct pp_hwmgr *hwmgr = handle;
 	int ret = 0;
 
-	ret = pp_check(hwmgr);
-
-	if (ret)
-		return ret;
-
-	if (clocks == NULL)
+	if (!hwmgr || !hwmgr->pm_en ||!clocks)
 		return -EINVAL;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1345,22 +1146,78 @@ static int pp_get_display_mode_validation_clocks(void *handle,
 	return ret;
 }
 
-static int pp_set_mmhub_powergating_by_smu(void *handle)
+static int pp_dpm_powergate_mmhub(void *handle)
 {
 	struct pp_hwmgr *hwmgr = handle;
-	int ret = 0;
 
-	ret = pp_check(hwmgr);
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
 
-	if (ret)
-		return ret;
+	if (hwmgr->hwmgr_func->powergate_mmhub == NULL) {
+		pr_info("%s was not implemented.\n", __func__);
+		return 0;
+	}
+
+	return hwmgr->hwmgr_func->powergate_mmhub(hwmgr);
+}
+
+static int pp_dpm_powergate_gfx(void *handle, bool gate)
+{
+	struct pp_hwmgr *hwmgr = handle;
+
+	if (!hwmgr || !hwmgr->pm_en)
+		return 0;
 
-	if (hwmgr->hwmgr_func->set_mmhub_powergating_by_smu == NULL) {
+	if (hwmgr->hwmgr_func->powergate_gfx == NULL) {
 		pr_info("%s was not implemented.\n", __func__);
 		return 0;
 	}
 
-	return hwmgr->hwmgr_func->set_mmhub_powergating_by_smu(hwmgr);
+	return hwmgr->hwmgr_func->powergate_gfx(hwmgr, gate);
+}
+
+static int pp_set_powergating_by_smu(void *handle,
+				uint32_t block_type, bool gate)
+{
+	int ret = 0;
+
+	switch (block_type) {
+	case AMD_IP_BLOCK_TYPE_UVD:
+	case AMD_IP_BLOCK_TYPE_VCN:
+		pp_dpm_powergate_uvd(handle, gate);
+		break;
+	case AMD_IP_BLOCK_TYPE_VCE:
+		pp_dpm_powergate_vce(handle, gate);
+		break;
+	case AMD_IP_BLOCK_TYPE_GMC:
+		pp_dpm_powergate_mmhub(handle);
+		break;
+	case AMD_IP_BLOCK_TYPE_GFX:
+		ret = pp_dpm_powergate_gfx(handle, gate);
+		break;
+	default:
+		break;
+	}
+	return ret;
+}
+
+static int pp_notify_smu_enable_pwe(void *handle)
+{
+	struct pp_hwmgr *hwmgr = handle;
+
+	if (!hwmgr || !hwmgr->pm_en)
+		return -EINVAL;
+
+	if (hwmgr->hwmgr_func->smus_notify_pwe == NULL) {
+		pr_info("%s was not implemented.\n", __func__);
+		return -EINVAL;;
+	}
+
+	mutex_lock(&hwmgr->smu_lock);
+	hwmgr->hwmgr_func->smus_notify_pwe(hwmgr);
+	mutex_unlock(&hwmgr->smu_lock);
+
+	return 0;
 }
 
 static const struct amd_pm_funcs pp_dpm_funcs = {
@@ -1369,8 +1226,6 @@ static const struct amd_pm_funcs pp_dpm_funcs = {
 	.force_performance_level = pp_dpm_force_performance_level,
 	.get_performance_level = pp_dpm_get_performance_level,
 	.get_current_power_state = pp_dpm_get_current_power_state,
-	.powergate_vce = pp_dpm_powergate_vce,
-	.powergate_uvd = pp_dpm_powergate_uvd,
 	.dispatch_tasks = pp_dpm_dispatch_tasks,
 	.set_fan_control_mode = pp_dpm_set_fan_control_mode,
 	.get_fan_control_mode = pp_dpm_get_fan_control_mode,
@@ -1390,7 +1245,7 @@ static const struct amd_pm_funcs pp_dpm_funcs = {
 	.get_vce_clock_state = pp_dpm_get_vce_clock_state,
 	.switch_power_profile = pp_dpm_switch_power_profile,
 	.set_clockgating_by_smu = pp_set_clockgating_by_smu,
-	.notify_smu_memory_info = pp_dpm_notify_smu_memory_info,
+	.set_powergating_by_smu = pp_set_powergating_by_smu,
 	.get_power_profile_mode = pp_get_power_profile_mode,
 	.set_power_profile_mode = pp_set_power_profile_mode,
 	.odn_edit_dpm_table = pp_odn_edit_dpm_table,
@@ -1408,5 +1263,5 @@ static const struct amd_pm_funcs pp_dpm_funcs = {
 	.set_watermarks_for_clocks_ranges = pp_set_watermarks_for_clocks_ranges,
 	.display_clock_voltage_request = pp_display_clock_voltage_request,
 	.get_display_mode_validation_clocks = pp_get_display_mode_validation_clocks,
-	.set_mmhub_powergating_by_smu = pp_set_mmhub_powergating_by_smu,
+	.notify_smu_enable_pwe = pp_notify_smu_enable_pwe,
 };
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
index ae2e9339dd6b..53207e76b0f3 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hardwaremanager.c
@@ -75,8 +75,7 @@ int phm_set_power_state(struct pp_hwmgr *hwmgr,
 
 int phm_enable_dynamic_state_management(struct pp_hwmgr *hwmgr)
 {
-	int ret = 1;
-	bool enabled;
+	int ret = -EINVAL;;
 	PHM_FUNC_CHECK(hwmgr);
 
 	if (smum_is_dpm_running(hwmgr)) {
@@ -87,17 +86,12 @@ int phm_enable_dynamic_state_management(struct pp_hwmgr *hwmgr)
 	if (NULL != hwmgr->hwmgr_func->dynamic_state_management_enable)
 		ret = hwmgr->hwmgr_func->dynamic_state_management_enable(hwmgr);
 
-	enabled = ret == 0;
-
-	cgs_notify_dpm_enabled(hwmgr->device, enabled);
-
 	return ret;
 }
 
 int phm_disable_dynamic_state_management(struct pp_hwmgr *hwmgr)
 {
-	int ret = -1;
-	bool enabled;
+	int ret = -EINVAL;
 
 	PHM_FUNC_CHECK(hwmgr);
 
@@ -109,10 +103,6 @@ int phm_disable_dynamic_state_management(struct pp_hwmgr *hwmgr)
 	if (hwmgr->hwmgr_func->dynamic_state_management_disable)
 		ret = hwmgr->hwmgr_func->dynamic_state_management_disable(hwmgr);
 
-	enabled = ret == 0 ? false : true;
-
-	cgs_notify_dpm_enabled(hwmgr->device, enabled);
-
 	return ret;
 }
 
@@ -142,6 +132,15 @@ int phm_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 	return 0;
 }
 
+int phm_apply_clock_adjust_rules(struct pp_hwmgr *hwmgr)
+{
+	PHM_FUNC_CHECK(hwmgr);
+
+	if (hwmgr->hwmgr_func->apply_clocks_adjust_rules != NULL)
+		return hwmgr->hwmgr_func->apply_clocks_adjust_rules(hwmgr);
+	return 0;
+}
+
 int phm_powerdown_uvd(struct pp_hwmgr *hwmgr)
 {
 	PHM_FUNC_CHECK(hwmgr);
@@ -171,6 +170,16 @@ int phm_disable_clock_power_gatings(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
+int phm_pre_display_configuration_changed(struct pp_hwmgr *hwmgr)
+{
+	PHM_FUNC_CHECK(hwmgr);
+
+	if (NULL != hwmgr->hwmgr_func->pre_display_config_changed)
+		hwmgr->hwmgr_func->pre_display_config_changed(hwmgr);
+
+	return 0;
+
+}
 
 int phm_display_configuration_changed(struct pp_hwmgr *hwmgr)
 {
@@ -275,13 +284,11 @@ int phm_store_dal_configuration_data(struct pp_hwmgr *hwmgr,
 	if (display_config == NULL)
 		return -EINVAL;
 
-	hwmgr->display_config = *display_config;
-
 	if (NULL != hwmgr->hwmgr_func->set_deep_sleep_dcefclk)
-		hwmgr->hwmgr_func->set_deep_sleep_dcefclk(hwmgr, hwmgr->display_config.min_dcef_deep_sleep_set_clk);
+		hwmgr->hwmgr_func->set_deep_sleep_dcefclk(hwmgr, display_config->min_dcef_deep_sleep_set_clk);
 
-	for (index = 0; index < hwmgr->display_config.num_path_including_non_display; index++) {
-		if (hwmgr->display_config.displays[index].controller_id != 0)
+	for (index = 0; index < display_config->num_path_including_non_display; index++) {
+		if (display_config->displays[index].controller_id != 0)
 			number_of_active_display++;
 	}
 
@@ -428,7 +435,7 @@ int phm_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr,
 }
 
 int phm_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr,
-		struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges)
+					void *clock_ranges)
 {
 	PHM_FUNC_CHECK(hwmgr);
 
@@ -436,7 +443,7 @@ int phm_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr,
 		return -EINVAL;
 
 	return hwmgr->hwmgr_func->set_watermarks_for_clocks_ranges(hwmgr,
-			wm_with_clock_ranges);
+								clock_ranges);
 }
 
 int phm_display_clock_voltage_request(struct pp_hwmgr *hwmgr,
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
index 42982055b161..8994aa5c8cf8 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c
@@ -40,6 +40,7 @@ extern const struct pp_smumgr_func iceland_smu_funcs;
 extern const struct pp_smumgr_func tonga_smu_funcs;
 extern const struct pp_smumgr_func fiji_smu_funcs;
 extern const struct pp_smumgr_func polaris10_smu_funcs;
+extern const struct pp_smumgr_func vegam_smu_funcs;
 extern const struct pp_smumgr_func vega10_smu_funcs;
 extern const struct pp_smumgr_func vega12_smu_funcs;
 extern const struct pp_smumgr_func smu10_smu_funcs;
@@ -76,11 +77,10 @@ static void hwmgr_init_workload_prority(struct pp_hwmgr *hwmgr)
 
 int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 {
-	if (hwmgr == NULL)
+	if (!hwmgr)
 		return -EINVAL;
 
 	hwmgr->usec_timeout = AMD_MAX_USEC_TIMEOUT;
-	hwmgr->power_source = PP_PowerSource_AC;
 	hwmgr->pp_table_version = PP_TABLE_V1;
 	hwmgr->dpm_level = AMD_DPM_FORCED_LEVEL_AUTO;
 	hwmgr->request_dpm_level = AMD_DPM_FORCED_LEVEL_AUTO;
@@ -95,7 +95,8 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 		hwmgr->smumgr_funcs = &ci_smu_funcs;
 		ci_set_asic_special_caps(hwmgr);
 		hwmgr->feature_mask &= ~(PP_VBI_TIME_SUPPORT_MASK |
-					PP_ENABLE_GFX_CG_THRU_SMU);
+					 PP_ENABLE_GFX_CG_THRU_SMU |
+					 PP_GFXOFF_MASK);
 		hwmgr->pp_table_version = PP_TABLE_V0;
 		hwmgr->od_enabled = false;
 		smu7_init_function_pointers(hwmgr);
@@ -103,9 +104,11 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 	case AMDGPU_FAMILY_CZ:
 		hwmgr->od_enabled = false;
 		hwmgr->smumgr_funcs = &smu8_smu_funcs;
+		hwmgr->feature_mask &= ~PP_GFXOFF_MASK;
 		smu8_init_function_pointers(hwmgr);
 		break;
 	case AMDGPU_FAMILY_VI:
+		hwmgr->feature_mask &= ~PP_GFXOFF_MASK;
 		switch (hwmgr->chip_id) {
 		case CHIP_TOPAZ:
 			hwmgr->smumgr_funcs = &iceland_smu_funcs;
@@ -133,6 +136,11 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 			polaris_set_asic_special_caps(hwmgr);
 			hwmgr->feature_mask &= ~(PP_UVD_HANDSHAKE_MASK);
 			break;
+		case CHIP_VEGAM:
+			hwmgr->smumgr_funcs = &vegam_smu_funcs;
+			polaris_set_asic_special_caps(hwmgr);
+			hwmgr->feature_mask &= ~(PP_UVD_HANDSHAKE_MASK);
+			break;
 		default:
 			return -EINVAL;
 		}
@@ -141,6 +149,8 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 	case AMDGPU_FAMILY_AI:
 		switch (hwmgr->chip_id) {
 		case CHIP_VEGA10:
+		case CHIP_VEGA20:
+			hwmgr->feature_mask &= ~PP_GFXOFF_MASK;
 			hwmgr->smumgr_funcs = &vega10_smu_funcs;
 			vega10_hwmgr_init(hwmgr);
 			break;
@@ -170,25 +180,66 @@ int hwmgr_early_init(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
+int hwmgr_sw_init(struct pp_hwmgr *hwmgr)
+{
+	if (!hwmgr|| !hwmgr->smumgr_funcs || !hwmgr->smumgr_funcs->smu_init)
+		return -EINVAL;
+
+	phm_register_irq_handlers(hwmgr);
+
+	return hwmgr->smumgr_funcs->smu_init(hwmgr);
+}
+
+
+int hwmgr_sw_fini(struct pp_hwmgr *hwmgr)
+{
+	if (hwmgr && hwmgr->smumgr_funcs && hwmgr->smumgr_funcs->smu_fini)
+		hwmgr->smumgr_funcs->smu_fini(hwmgr);
+
+	return 0;
+}
+
 int hwmgr_hw_init(struct pp_hwmgr *hwmgr)
 {
 	int ret = 0;
 
-	if (hwmgr == NULL)
+	if (!hwmgr || !hwmgr->smumgr_funcs)
 		return -EINVAL;
 
-	if (hwmgr->pptable_func == NULL ||
-	    hwmgr->pptable_func->pptable_init == NULL ||
-	    hwmgr->hwmgr_func->backend_init == NULL)
-		return -EINVAL;
+	if (hwmgr->smumgr_funcs->start_smu) {
+		ret = hwmgr->smumgr_funcs->start_smu(hwmgr);
+		if (ret) {
+			pr_err("smc start failed\n");
+			return -EINVAL;
+		}
+	}
+
+	if (!hwmgr->pm_en)
+		return 0;
+
+	if (!hwmgr->pptable_func ||
+	    !hwmgr->pptable_func->pptable_init ||
+	    !hwmgr->hwmgr_func->backend_init) {
+		hwmgr->pm_en = false;
+		pr_info("dpm not supported \n");
+		return 0;
+	}
 
 	ret = hwmgr->pptable_func->pptable_init(hwmgr);
 	if (ret)
 		goto err;
 
+	((struct amdgpu_device *)hwmgr->adev)->pm.no_fan =
+				hwmgr->thermal_controller.fanInfo.bNoFan;
+
 	ret = hwmgr->hwmgr_func->backend_init(hwmgr);
 	if (ret)
 		goto err1;
+ /* make sure dc limits are valid */
+	if ((hwmgr->dyn_state.max_clock_voltage_on_dc.sclk == 0) ||
+			(hwmgr->dyn_state.max_clock_voltage_on_dc.mclk == 0))
+			hwmgr->dyn_state.max_clock_voltage_on_dc =
+					hwmgr->dyn_state.max_clock_voltage_on_ac;
 
 	ret = psm_init_power_state_table(hwmgr);
 	if (ret)
@@ -206,6 +257,8 @@ int hwmgr_hw_init(struct pp_hwmgr *hwmgr)
 	if (ret)
 		goto err2;
 
+	((struct amdgpu_device *)hwmgr->adev)->pm.dpm_enabled = true;
+
 	return 0;
 err2:
 	if (hwmgr->hwmgr_func->backend_fini)
@@ -214,14 +267,13 @@ err1:
 	if (hwmgr->pptable_func->pptable_fini)
 		hwmgr->pptable_func->pptable_fini(hwmgr);
 err:
-	pr_err("amdgpu: powerplay initialization failed\n");
 	return ret;
 }
 
 int hwmgr_hw_fini(struct pp_hwmgr *hwmgr)
 {
-	if (hwmgr == NULL)
-		return -EINVAL;
+	if (!hwmgr || !hwmgr->pm_en)
+		return 0;
 
 	phm_stop_thermal_controller(hwmgr);
 	psm_set_boot_states(hwmgr);
@@ -236,12 +288,12 @@ int hwmgr_hw_fini(struct pp_hwmgr *hwmgr)
 	return psm_fini_power_state_table(hwmgr);
 }
 
-int hwmgr_hw_suspend(struct pp_hwmgr *hwmgr)
+int hwmgr_suspend(struct pp_hwmgr *hwmgr)
 {
 	int ret = 0;
 
-	if (hwmgr == NULL)
-		return -EINVAL;
+	if (!hwmgr || !hwmgr->pm_en)
+		return 0;
 
 	phm_disable_smc_firmware_ctf(hwmgr);
 	ret = psm_set_boot_states(hwmgr);
@@ -255,13 +307,23 @@ int hwmgr_hw_suspend(struct pp_hwmgr *hwmgr)
 	return ret;
 }
 
-int hwmgr_hw_resume(struct pp_hwmgr *hwmgr)
+int hwmgr_resume(struct pp_hwmgr *hwmgr)
 {
 	int ret = 0;
 
-	if (hwmgr == NULL)
+	if (!hwmgr)
 		return -EINVAL;
 
+	if (hwmgr->smumgr_funcs && hwmgr->smumgr_funcs->start_smu) {
+		if (hwmgr->smumgr_funcs->start_smu(hwmgr)) {
+			pr_err("smc start failed\n");
+			return -EINVAL;
+		}
+	}
+
+	if (!hwmgr->pm_en)
+		return 0;
+
 	ret = phm_setup_asic(hwmgr);
 	if (ret)
 		return ret;
@@ -270,9 +332,6 @@ int hwmgr_hw_resume(struct pp_hwmgr *hwmgr)
 	if (ret)
 		return ret;
 	ret = phm_start_thermal_controller(hwmgr);
-	if (ret)
-		return ret;
-
 	ret |= psm_set_performance_states(hwmgr);
 	if (ret)
 		return ret;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
index 0f2851b5b368..91ffb7bc4ee7 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/pp_psm.c
@@ -46,11 +46,11 @@ int psm_init_power_state_table(struct pp_hwmgr *hwmgr)
 					  sizeof(struct pp_power_state);
 
 	if (table_entries == 0 || size == 0) {
-		pr_warn("Please check whether power state management is suppported on this asic\n");
+		pr_warn("Please check whether power state management is supported on this asic\n");
 		return 0;
 	}
 
-	hwmgr->ps = kzalloc(size * table_entries, GFP_KERNEL);
+	hwmgr->ps = kcalloc(table_entries, size, GFP_KERNEL);
 	if (hwmgr->ps == NULL)
 		return -ENOMEM;
 
@@ -265,10 +265,18 @@ int psm_adjust_power_state_dynamic(struct pp_hwmgr *hwmgr, bool skip,
 	if (skip)
 		return 0;
 
+	phm_pre_display_configuration_changed(hwmgr);
+
 	phm_display_configuration_changed(hwmgr);
 
 	if (hwmgr->ps)
 		power_state_management(hwmgr, new_ps);
+	else
+		/*
+		 * for vega12/vega20 which does not support power state manager
+		 * DAL clock limits should also be honoured
+		 */
+		phm_apply_clock_adjust_rules(hwmgr);
 
 	phm_notify_smc_display_config_after_ps_adjustment(hwmgr);
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
index c6febbf0bf69..01dc46dc9c8a 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.c
@@ -23,7 +23,8 @@
 #include "pp_debug.h"
 #include <linux/module.h>
 #include <linux/slab.h>
-
+#include <linux/delay.h>
+#include "atom.h"
 #include "ppatomctrl.h"
 #include "atombios.h"
 #include "cgs_common.h"
@@ -128,7 +129,6 @@ static int atomctrl_set_mc_reg_address_table(
 	return 0;
 }
 
-
 int atomctrl_initialize_mc_reg_table(
 		struct pp_hwmgr *hwmgr,
 		uint8_t module_index,
@@ -141,7 +141,7 @@ int atomctrl_initialize_mc_reg_table(
 	u16 size;
 
 	vram_info = (ATOM_VRAM_INFO_HEADER_V2_1 *)
-		cgs_atom_get_data_table(hwmgr->device,
+		smu_atom_get_data_table(hwmgr->adev,
 				GetIndexIntoMasterTable(DATA, VRAM_Info), &size, &frev, &crev);
 
 	if (module_index >= vram_info->ucNumOfVRAMModule) {
@@ -174,6 +174,8 @@ int atomctrl_set_engine_dram_timings_rv770(
 		uint32_t engine_clock,
 		uint32_t memory_clock)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
+
 	SET_ENGINE_CLOCK_PS_ALLOCATION engine_clock_parameters;
 
 	/* They are both in 10KHz Units. */
@@ -184,9 +186,10 @@ int atomctrl_set_engine_dram_timings_rv770(
 	/* in 10 khz units.*/
 	engine_clock_parameters.sReserved.ulClock =
 		cpu_to_le32(memory_clock & SET_CLOCK_FREQ_MASK);
-	return cgs_atom_exec_cmd_table(hwmgr->device,
+
+	return amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, DynamicMemorySettings),
-			&engine_clock_parameters);
+			(uint32_t *)&engine_clock_parameters);
 }
 
 /**
@@ -203,7 +206,7 @@ static ATOM_VOLTAGE_OBJECT_INFO *get_voltage_info_table(void *device)
 	union voltage_object_info *voltage_info;
 
 	voltage_info = (union voltage_object_info *)
-		cgs_atom_get_data_table(device, index,
+		smu_atom_get_data_table(device, index,
 			&size, &frev, &crev);
 
 	if (voltage_info != NULL)
@@ -247,16 +250,16 @@ int atomctrl_get_memory_pll_dividers_si(
 		pp_atomctrl_memory_clock_param *mpll_param,
 		bool strobe_mode)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_1 mpll_parameters;
 	int result;
 
 	mpll_parameters.ulClock = cpu_to_le32(clock_value);
 	mpll_parameters.ucInputFlag = (uint8_t)((strobe_mode) ? 1 : 0);
 
-	result = cgs_atom_exec_cmd_table
-		(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 		 GetIndexIntoMasterTable(COMMAND, ComputeMemoryClockParam),
-		 &mpll_parameters);
+		(uint32_t *)&mpll_parameters);
 
 	if (0 == result) {
 		mpll_param->mpll_fb_divider.clk_frac =
@@ -295,14 +298,15 @@ int atomctrl_get_memory_pll_dividers_si(
 int atomctrl_get_memory_pll_dividers_vi(struct pp_hwmgr *hwmgr,
 		uint32_t clock_value, pp_atomctrl_memory_clock_param *mpll_param)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_2 mpll_parameters;
 	int result;
 
 	mpll_parameters.ulClock.ulClock = cpu_to_le32(clock_value);
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ComputeMemoryClockParam),
-			&mpll_parameters);
+			(uint32_t *)&mpll_parameters);
 
 	if (!result)
 		mpll_param->mpll_post_divider =
@@ -311,19 +315,49 @@ int atomctrl_get_memory_pll_dividers_vi(struct pp_hwmgr *hwmgr,
 	return result;
 }
 
+int atomctrl_get_memory_pll_dividers_ai(struct pp_hwmgr *hwmgr,
+					uint32_t clock_value,
+					pp_atomctrl_memory_clock_param_ai *mpll_param)
+{
+	struct amdgpu_device *adev = hwmgr->adev;
+	COMPUTE_MEMORY_CLOCK_PARAM_PARAMETERS_V2_3 mpll_parameters = {{0}, 0, 0};
+	int result;
+
+	mpll_parameters.ulClock.ulClock = cpu_to_le32(clock_value);
+
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
+			GetIndexIntoMasterTable(COMMAND, ComputeMemoryClockParam),
+			(uint32_t *)&mpll_parameters);
+
+	/* VEGAM's mpll takes sometime to finish computing */
+	udelay(10);
+
+	if (!result) {
+		mpll_param->ulMclk_fcw_int =
+			le16_to_cpu(mpll_parameters.usMclk_fcw_int);
+		mpll_param->ulMclk_fcw_frac =
+			le16_to_cpu(mpll_parameters.usMclk_fcw_frac);
+		mpll_param->ulClock =
+			le32_to_cpu(mpll_parameters.ulClock.ulClock);
+		mpll_param->ulPostDiv = mpll_parameters.ulClock.ucPostDiv;
+	}
+
+	return result;
+}
+
 int atomctrl_get_engine_pll_dividers_kong(struct pp_hwmgr *hwmgr,
 					  uint32_t clock_value,
 					  pp_atomctrl_clock_dividers_kong *dividers)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	COMPUTE_MEMORY_ENGINE_PLL_PARAMETERS_V4 pll_parameters;
 	int result;
 
 	pll_parameters.ulClock = cpu_to_le32(clock_value);
 
-	result = cgs_atom_exec_cmd_table
-		(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 		 GetIndexIntoMasterTable(COMMAND, ComputeMemoryEnginePLL),
-		 &pll_parameters);
+		(uint32_t *)&pll_parameters);
 
 	if (0 == result) {
 		dividers->pll_post_divider = pll_parameters.ucPostDiv;
@@ -338,16 +372,16 @@ int atomctrl_get_engine_pll_dividers_vi(
 		uint32_t clock_value,
 		pp_atomctrl_clock_dividers_vi *dividers)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	COMPUTE_GPU_CLOCK_OUTPUT_PARAMETERS_V1_6 pll_patameters;
 	int result;
 
 	pll_patameters.ulClock.ulClock = cpu_to_le32(clock_value);
 	pll_patameters.ulClock.ucPostDiv = COMPUTE_GPUCLK_INPUT_FLAG_SCLK;
 
-	result = cgs_atom_exec_cmd_table
-		(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 		 GetIndexIntoMasterTable(COMMAND, ComputeMemoryEnginePLL),
-		 &pll_patameters);
+		(uint32_t *)&pll_patameters);
 
 	if (0 == result) {
 		dividers->pll_post_divider =
@@ -375,16 +409,16 @@ int atomctrl_get_engine_pll_dividers_ai(struct pp_hwmgr *hwmgr,
 		uint32_t clock_value,
 		pp_atomctrl_clock_dividers_ai *dividers)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	COMPUTE_GPU_CLOCK_OUTPUT_PARAMETERS_V1_7 pll_patameters;
 	int result;
 
 	pll_patameters.ulClock.ulClock = cpu_to_le32(clock_value);
 	pll_patameters.ulClock.ucPostDiv = COMPUTE_GPUCLK_INPUT_FLAG_SCLK;
 
-	result = cgs_atom_exec_cmd_table
-		(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 		 GetIndexIntoMasterTable(COMMAND, ComputeMemoryEnginePLL),
-		 &pll_patameters);
+		(uint32_t *)&pll_patameters);
 
 	if (0 == result) {
 		dividers->usSclk_fcw_frac     = le16_to_cpu(pll_patameters.usSclk_fcw_frac);
@@ -407,6 +441,7 @@ int atomctrl_get_dfs_pll_dividers_vi(
 		uint32_t clock_value,
 		pp_atomctrl_clock_dividers_vi *dividers)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	COMPUTE_GPU_CLOCK_OUTPUT_PARAMETERS_V1_6 pll_patameters;
 	int result;
 
@@ -414,10 +449,9 @@ int atomctrl_get_dfs_pll_dividers_vi(
 	pll_patameters.ulClock.ucPostDiv =
 		COMPUTE_GPUCLK_INPUT_FLAG_DEFAULT_GPUCLK;
 
-	result = cgs_atom_exec_cmd_table
-		(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 		 GetIndexIntoMasterTable(COMMAND, ComputeMemoryEnginePLL),
-		 &pll_patameters);
+		(uint32_t *)&pll_patameters);
 
 	if (0 == result) {
 		dividers->pll_post_divider =
@@ -452,7 +486,7 @@ uint32_t atomctrl_get_reference_clock(struct pp_hwmgr *hwmgr)
 	uint32_t clock;
 
 	fw_info = (ATOM_FIRMWARE_INFO *)
-		cgs_atom_get_data_table(hwmgr->device,
+		smu_atom_get_data_table(hwmgr->adev,
 			GetIndexIntoMasterTable(DATA, FirmwareInfo),
 			&size, &frev, &crev);
 
@@ -476,7 +510,7 @@ bool atomctrl_is_voltage_controlled_by_gpio_v3(
 		uint8_t voltage_mode)
 {
 	ATOM_VOLTAGE_OBJECT_INFO_V3_1 *voltage_info =
-		(ATOM_VOLTAGE_OBJECT_INFO_V3_1 *)get_voltage_info_table(hwmgr->device);
+		(ATOM_VOLTAGE_OBJECT_INFO_V3_1 *)get_voltage_info_table(hwmgr->adev);
 	bool ret;
 
 	PP_ASSERT_WITH_CODE((NULL != voltage_info),
@@ -495,7 +529,7 @@ int atomctrl_get_voltage_table_v3(
 		pp_atomctrl_voltage_table *voltage_table)
 {
 	ATOM_VOLTAGE_OBJECT_INFO_V3_1 *voltage_info =
-		(ATOM_VOLTAGE_OBJECT_INFO_V3_1 *)get_voltage_info_table(hwmgr->device);
+		(ATOM_VOLTAGE_OBJECT_INFO_V3_1 *)get_voltage_info_table(hwmgr->adev);
 	const ATOM_VOLTAGE_OBJECT_V3 *voltage_object;
 	unsigned int i;
 
@@ -572,7 +606,7 @@ static ATOM_GPIO_PIN_LUT *get_gpio_lookup_table(void *device)
 	void *table_address;
 
 	table_address = (ATOM_GPIO_PIN_LUT *)
-		cgs_atom_get_data_table(device,
+		smu_atom_get_data_table(device,
 				GetIndexIntoMasterTable(DATA, GPIO_Pin_LUT),
 				&size, &frev, &crev);
 
@@ -592,7 +626,7 @@ bool atomctrl_get_pp_assign_pin(
 {
 	bool bRet = false;
 	ATOM_GPIO_PIN_LUT *gpio_lookup_table =
-		get_gpio_lookup_table(hwmgr->device);
+		get_gpio_lookup_table(hwmgr->adev);
 
 	PP_ASSERT_WITH_CODE((NULL != gpio_lookup_table),
 			"Could not find GPIO lookup Table in BIOS.", return false);
@@ -613,7 +647,7 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 		bool debug)
 {
 	ATOM_ASIC_PROFILING_INFO_V3_4 *getASICProfilingInfo;
-
+	struct amdgpu_device *adev = hwmgr->adev;
 	EFUSE_LINEAR_FUNC_PARAM sRO_fuse;
 	EFUSE_LINEAR_FUNC_PARAM sCACm_fuse;
 	EFUSE_LINEAR_FUNC_PARAM sCACb_fuse;
@@ -640,7 +674,7 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	int result;
 
 	getASICProfilingInfo = (ATOM_ASIC_PROFILING_INFO_V3_4 *)
-			cgs_atom_get_data_table(hwmgr->device,
+			smu_atom_get_data_table(hwmgr->adev,
 					GetIndexIntoMasterTable(DATA, ASIC_ProfilingInfo),
 					NULL, NULL, NULL);
 
@@ -706,9 +740,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 
 	sOutput_FuseValues.sEfuse = sInput_FuseValues;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&sOutput_FuseValues);
+			(uint32_t *)&sOutput_FuseValues);
 
 	if (result)
 		return result;
@@ -727,9 +761,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 
 	sOutput_FuseValues.sEfuse = sInput_FuseValues;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&sOutput_FuseValues);
+			(uint32_t *)&sOutput_FuseValues);
 
 	if (result)
 		return result;
@@ -747,9 +781,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	sInput_FuseValues.ucBitLength = sCACb_fuse.ucEfuseLength;
 	sOutput_FuseValues.sEfuse = sInput_FuseValues;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&sOutput_FuseValues);
+			(uint32_t *)&sOutput_FuseValues);
 
 	if (result)
 		return result;
@@ -768,9 +802,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 
 	sOutput_FuseValues.sEfuse = sInput_FuseValues;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&sOutput_FuseValues);
+			(uint32_t *)&sOutput_FuseValues);
 
 	if (result)
 		return result;
@@ -790,9 +824,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 
 	sOutput_FuseValues.sEfuse = sInput_FuseValues;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&sOutput_FuseValues);
+			(uint32_t *)&sOutput_FuseValues);
 	if (result)
 		return result;
 
@@ -811,9 +845,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 	sInput_FuseValues.ucBitLength = sKv_b_fuse.ucEfuseLength;
 	sOutput_FuseValues.sEfuse = sInput_FuseValues;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&sOutput_FuseValues);
+			(uint32_t *)&sOutput_FuseValues);
 
 	if (result)
 		return result;
@@ -842,9 +876,9 @@ int atomctrl_calculate_voltage_evv_on_sclk(
 
 	sOutput_FuseValues.sEfuse = sInput_FuseValues;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&sOutput_FuseValues);
+			(uint32_t *)&sOutput_FuseValues);
 
 	if (result)
 		return result;
@@ -1053,8 +1087,9 @@ int atomctrl_get_voltage_evv_on_sclk(
 		uint32_t sclk, uint16_t virtual_voltage_Id,
 		uint16_t *voltage)
 {
-	int result;
+	struct amdgpu_device *adev = hwmgr->adev;
 	GET_VOLTAGE_INFO_INPUT_PARAMETER_V1_2 get_voltage_info_param_space;
+	int result;
 
 	get_voltage_info_param_space.ucVoltageType   =
 		voltage_type;
@@ -1065,14 +1100,12 @@ int atomctrl_get_voltage_evv_on_sclk(
 	get_voltage_info_param_space.ulSCLKFreq      =
 		cpu_to_le32(sclk);
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, GetVoltageInfo),
-			&get_voltage_info_param_space);
-
-	if (0 != result)
-		return result;
+			(uint32_t *)&get_voltage_info_param_space);
 
-	*voltage = le16_to_cpu(((GET_EVV_VOLTAGE_INFO_OUTPUT_PARAMETER_V1_2 *)
+	*voltage = result ? 0 :
+			le16_to_cpu(((GET_EVV_VOLTAGE_INFO_OUTPUT_PARAMETER_V1_2 *)
 				(&get_voltage_info_param_space))->usVoltageLevel);
 
 	return result;
@@ -1088,9 +1121,10 @@ int atomctrl_get_voltage_evv(struct pp_hwmgr *hwmgr,
 			     uint16_t virtual_voltage_id,
 			     uint16_t *voltage)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
+	GET_VOLTAGE_INFO_INPUT_PARAMETER_V1_2 get_voltage_info_param_space;
 	int result;
 	int entry_id;
-	GET_VOLTAGE_INFO_INPUT_PARAMETER_V1_2 get_voltage_info_param_space;
 
 	/* search for leakage voltage ID 0xff01 ~ 0xff08 and sckl */
 	for (entry_id = 0; entry_id < hwmgr->dyn_state.vddc_dependency_on_sclk->count; entry_id++) {
@@ -1111,9 +1145,9 @@ int atomctrl_get_voltage_evv(struct pp_hwmgr *hwmgr,
 	get_voltage_info_param_space.ulSCLKFreq =
 		cpu_to_le32(hwmgr->dyn_state.vddc_dependency_on_sclk->entries[entry_id].clk);
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, GetVoltageInfo),
-			&get_voltage_info_param_space);
+			(uint32_t *)&get_voltage_info_param_space);
 
 	if (0 != result)
 		return result;
@@ -1135,7 +1169,7 @@ uint32_t atomctrl_get_mpll_reference_clock(struct pp_hwmgr *hwmgr)
 	u16 size;
 
 	fw_info = (ATOM_COMMON_TABLE_HEADER *)
-		cgs_atom_get_data_table(hwmgr->device,
+		smu_atom_get_data_table(hwmgr->adev,
 				GetIndexIntoMasterTable(DATA, FirmwareInfo),
 				&size, &frev, &crev);
 
@@ -1167,7 +1201,7 @@ static ATOM_ASIC_INTERNAL_SS_INFO *asic_internal_ss_get_ss_table(void *device)
 	u16 size;
 
 	table = (ATOM_ASIC_INTERNAL_SS_INFO *)
-		cgs_atom_get_data_table(device,
+		smu_atom_get_data_table(device,
 			GetIndexIntoMasterTable(DATA, ASIC_InternalSS_Info),
 			&size, &frev, &crev);
 
@@ -1188,7 +1222,7 @@ static int asic_internal_ss_get_ss_asignment(struct pp_hwmgr *hwmgr,
 
 	memset(ssEntry, 0x00, sizeof(pp_atomctrl_internal_ss_info));
 
-	table = asic_internal_ss_get_ss_table(hwmgr->device);
+	table = asic_internal_ss_get_ss_table(hwmgr->adev);
 
 	if (NULL == table)
 		return -1;
@@ -1260,9 +1294,10 @@ int atomctrl_get_engine_clock_spread_spectrum(
 			ASIC_INTERNAL_ENGINE_SS, engine_clock, ssInfo);
 }
 
-int atomctrl_read_efuse(void *device, uint16_t start_index,
+int atomctrl_read_efuse(struct pp_hwmgr *hwmgr, uint16_t start_index,
 		uint16_t end_index, uint32_t mask, uint32_t *efuse)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 	READ_EFUSE_VALUE_PARAMETER efuse_param;
 
@@ -1272,11 +1307,10 @@ int atomctrl_read_efuse(void *device, uint16_t start_index,
 	efuse_param.sEfuse.ucBitLength  = (uint8_t)
 			((end_index - start_index) + 1);
 
-	result = cgs_atom_exec_cmd_table(device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, ReadEfuseValue),
-			&efuse_param);
-	if (!result)
-		*efuse = le32_to_cpu(efuse_param.ulEfuseValue) & mask;
+			(uint32_t *)&efuse_param);
+	*efuse = result ? 0 : le32_to_cpu(efuse_param.ulEfuseValue) & mask;
 
 	return result;
 }
@@ -1284,6 +1318,7 @@ int atomctrl_read_efuse(void *device, uint16_t start_index,
 int atomctrl_set_ac_timing_ai(struct pp_hwmgr *hwmgr, uint32_t memory_clock,
 			      uint8_t level)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	DYNAMICE_MEMORY_SETTINGS_PARAMETER_V2_1 memory_clock_parameters;
 	int result;
 
@@ -1293,10 +1328,9 @@ int atomctrl_set_ac_timing_ai(struct pp_hwmgr *hwmgr, uint32_t memory_clock,
 		ADJUST_MC_SETTING_PARAM;
 	memory_clock_parameters.asDPMMCReg.ucMclkDPMState = level;
 
-	result = cgs_atom_exec_cmd_table
-		(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 		 GetIndexIntoMasterTable(COMMAND, DynamicMemorySettings),
-		 &memory_clock_parameters);
+		(uint32_t *)&memory_clock_parameters);
 
 	return result;
 }
@@ -1304,7 +1338,7 @@ int atomctrl_set_ac_timing_ai(struct pp_hwmgr *hwmgr, uint32_t memory_clock,
 int atomctrl_get_voltage_evv_on_sclk_ai(struct pp_hwmgr *hwmgr, uint8_t voltage_type,
 				uint32_t sclk, uint16_t virtual_voltage_Id, uint32_t *voltage)
 {
-
+	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 	GET_VOLTAGE_INFO_INPUT_PARAMETER_V1_3 get_voltage_info_param_space;
 
@@ -1313,15 +1347,12 @@ int atomctrl_get_voltage_evv_on_sclk_ai(struct pp_hwmgr *hwmgr, uint8_t voltage_
 	get_voltage_info_param_space.usVoltageLevel = cpu_to_le16(virtual_voltage_Id);
 	get_voltage_info_param_space.ulSCLKFreq = cpu_to_le32(sclk);
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, GetVoltageInfo),
-			&get_voltage_info_param_space);
+			(uint32_t *)&get_voltage_info_param_space);
 
-	if (0 != result)
-		return result;
-
-	*voltage = le32_to_cpu(((GET_EVV_VOLTAGE_INFO_OUTPUT_PARAMETER_V1_3 *)
-				(&get_voltage_info_param_space))->ulVoltageLevel);
+	*voltage = result ? 0 :
+		le32_to_cpu(((GET_EVV_VOLTAGE_INFO_OUTPUT_PARAMETER_V1_3 *)(&get_voltage_info_param_space))->ulVoltageLevel);
 
 	return result;
 }
@@ -1334,7 +1365,7 @@ int atomctrl_get_smc_sclk_range_table(struct pp_hwmgr *hwmgr, struct pp_atom_ctr
 	u16 size;
 
 	ATOM_SMU_INFO_V2_1 *psmu_info =
-		(ATOM_SMU_INFO_V2_1 *)cgs_atom_get_data_table(hwmgr->device,
+		(ATOM_SMU_INFO_V2_1 *)smu_atom_get_data_table(hwmgr->adev,
 			GetIndexIntoMasterTable(DATA, SMU_Info),
 			&size, &frev, &crev);
 
@@ -1362,7 +1393,7 @@ int atomctrl_get_avfs_information(struct pp_hwmgr *hwmgr,
 		return -EINVAL;
 
 	profile = (ATOM_ASIC_PROFILING_INFO_V3_6 *)
-			cgs_atom_get_data_table(hwmgr->device,
+			smu_atom_get_data_table(hwmgr->adev,
 					GetIndexIntoMasterTable(DATA, ASIC_ProfilingInfo),
 					NULL, NULL, NULL);
 	if (!profile)
@@ -1402,7 +1433,7 @@ int  atomctrl_get_svi2_info(struct pp_hwmgr *hwmgr, uint8_t voltage_type,
 				uint16_t *load_line)
 {
 	ATOM_VOLTAGE_OBJECT_INFO_V3_1 *voltage_info =
-		(ATOM_VOLTAGE_OBJECT_INFO_V3_1 *)get_voltage_info_table(hwmgr->device);
+		(ATOM_VOLTAGE_OBJECT_INFO_V3_1 *)get_voltage_info_table(hwmgr->adev);
 
 	const ATOM_VOLTAGE_OBJECT_V3 *voltage_object;
 
@@ -1421,16 +1452,17 @@ int  atomctrl_get_svi2_info(struct pp_hwmgr *hwmgr, uint8_t voltage_type,
 
 int atomctrl_get_leakage_id_from_efuse(struct pp_hwmgr *hwmgr, uint16_t *virtual_voltage_id)
 {
-	int result;
+	struct amdgpu_device *adev = hwmgr->adev;
 	SET_VOLTAGE_PS_ALLOCATION allocation;
 	SET_VOLTAGE_PARAMETERS_V1_3 *voltage_parameters =
 			(SET_VOLTAGE_PARAMETERS_V1_3 *)&allocation.sASICSetVoltage;
+	int result;
 
 	voltage_parameters->ucVoltageMode = ATOM_GET_LEAKAGE_ID;
 
-	result = cgs_atom_exec_cmd_table(hwmgr->device,
+	result = amdgpu_atom_execute_table(adev->mode_info.atom_context,
 			GetIndexIntoMasterTable(COMMAND, SetVoltage),
-			voltage_parameters);
+			(uint32_t *)voltage_parameters);
 
 	*virtual_voltage_id = voltage_parameters->usVoltageLevel;
 
@@ -1453,7 +1485,7 @@ int atomctrl_get_leakage_vddc_base_on_leakage(struct pp_hwmgr *hwmgr,
 	ix = GetIndexIntoMasterTable(DATA, ASIC_ProfilingInfo);
 
 	profile = (ATOM_ASIC_PROFILING_INFO_V2_1 *)
-			cgs_atom_get_data_table(hwmgr->device,
+			smu_atom_get_data_table(hwmgr->adev,
 					ix,
 					NULL, NULL, NULL);
 	if (!profile)
@@ -1498,3 +1530,33 @@ int atomctrl_get_leakage_vddc_base_on_leakage(struct pp_hwmgr *hwmgr,
 
 	return 0;
 }
+
+void atomctrl_get_voltage_range(struct pp_hwmgr *hwmgr, uint32_t *max_vddc,
+							uint32_t *min_vddc)
+{
+	void *profile;
+
+	profile = smu_atom_get_data_table(hwmgr->adev,
+					GetIndexIntoMasterTable(DATA, ASIC_ProfilingInfo),
+					NULL, NULL, NULL);
+
+	if (profile) {
+		switch (hwmgr->chip_id) {
+		case CHIP_TONGA:
+		case CHIP_FIJI:
+			*max_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_3 *)profile)->ulMaxVddc) / 4;
+			*min_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_3 *)profile)->ulMinVddc) / 4;
+			return;
+		case CHIP_POLARIS11:
+		case CHIP_POLARIS10:
+		case CHIP_POLARIS12:
+			*max_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_6 *)profile)->ulMaxVddc) / 100;
+			*min_vddc = le32_to_cpu(((ATOM_ASIC_PROFILING_INFO_V3_6 *)profile)->ulMinVddc) / 100;
+			return;
+		default:
+			break;
+		}
+	}
+	*max_vddc = 0;
+	*min_vddc = 0;
+}
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
index c44a92064cf1..3ee54f182943 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomctrl.h
@@ -146,6 +146,14 @@ struct pp_atomctrl_memory_clock_param {
 };
 typedef struct pp_atomctrl_memory_clock_param pp_atomctrl_memory_clock_param;
 
+struct pp_atomctrl_memory_clock_param_ai {
+	uint32_t ulClock;
+	uint32_t ulPostDiv;
+	uint16_t ulMclk_fcw_frac;
+	uint16_t ulMclk_fcw_int;
+};
+typedef struct pp_atomctrl_memory_clock_param_ai pp_atomctrl_memory_clock_param_ai;
+
 struct pp_atomctrl_internal_ss_info {
 	uint32_t speed_spectrum_percentage;                      /* in 1/100 percentage */
 	uint32_t speed_spectrum_rate;                            /* in KHz */
@@ -295,10 +303,12 @@ extern bool atomctrl_is_voltage_controlled_by_gpio_v3(struct pp_hwmgr *hwmgr, ui
 extern int atomctrl_get_voltage_table_v3(struct pp_hwmgr *hwmgr, uint8_t voltage_type, uint8_t voltage_mode, pp_atomctrl_voltage_table *voltage_table);
 extern int atomctrl_get_memory_pll_dividers_vi(struct pp_hwmgr *hwmgr,
 		uint32_t clock_value, pp_atomctrl_memory_clock_param *mpll_param);
+extern int atomctrl_get_memory_pll_dividers_ai(struct pp_hwmgr *hwmgr,
+		uint32_t clock_value, pp_atomctrl_memory_clock_param_ai *mpll_param);
 extern int atomctrl_get_engine_pll_dividers_kong(struct pp_hwmgr *hwmgr,
 						 uint32_t clock_value,
 						 pp_atomctrl_clock_dividers_kong *dividers);
-extern int atomctrl_read_efuse(void *device, uint16_t start_index,
+extern int atomctrl_read_efuse(struct pp_hwmgr *hwmgr, uint16_t start_index,
 		uint16_t end_index, uint32_t mask, uint32_t *efuse);
 extern int atomctrl_calculate_voltage_evv_on_sclk(struct pp_hwmgr *hwmgr, uint8_t voltage_type,
 		uint32_t sclk, uint16_t virtual_voltage_Id, uint16_t *voltage, uint16_t dpm_level, bool debug);
@@ -320,5 +330,8 @@ extern int atomctrl_get_leakage_vddc_base_on_leakage(struct pp_hwmgr *hwmgr,
 					uint16_t virtual_voltage_id,
 					uint16_t efuse_voltage_id);
 extern int atomctrl_get_leakage_id_from_efuse(struct pp_hwmgr *hwmgr, uint16_t *virtual_voltage_id);
+
+extern void atomctrl_get_voltage_range(struct pp_hwmgr *hwmgr, uint32_t *max_vddc,
+							uint32_t *min_vddc);
 #endif
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c
index ad42caac033e..d27c1c9df286 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.c
@@ -23,9 +23,9 @@
 
 #include "ppatomfwctrl.h"
 #include "atomfirmware.h"
+#include "atom.h"
 #include "pp_debug.h"
 
-
 static const union atom_voltage_object_v4 *pp_atomfwctrl_lookup_voltage_type_v4(
 		const struct atom_voltage_objects_info_v4_1 *voltage_object_info_table,
 		uint8_t voltage_type, uint8_t voltage_mode)
@@ -38,35 +38,34 @@ static const union atom_voltage_object_v4 *pp_atomfwctrl_lookup_voltage_type_v4(
 
 	while (offset < size) {
 		const union atom_voltage_object_v4 *voltage_object =
-				(const union atom_voltage_object_v4 *)(start + offset);
+			(const union atom_voltage_object_v4 *)(start + offset);
 
-        if (voltage_type == voltage_object->gpio_voltage_obj.header.voltage_type &&
-            voltage_mode == voltage_object->gpio_voltage_obj.header.voltage_mode)
-            return voltage_object;
+		if (voltage_type == voltage_object->gpio_voltage_obj.header.voltage_type &&
+		    voltage_mode == voltage_object->gpio_voltage_obj.header.voltage_mode)
+			return voltage_object;
 
-        offset += le16_to_cpu(voltage_object->gpio_voltage_obj.header.object_size);
+		offset += le16_to_cpu(voltage_object->gpio_voltage_obj.header.object_size);
 
-    }
+	}
 
-    return NULL;
+	return NULL;
 }
 
 static struct atom_voltage_objects_info_v4_1 *pp_atomfwctrl_get_voltage_info_table(
 		struct pp_hwmgr *hwmgr)
 {
-    const void *table_address;
-    uint16_t idx;
+	const void *table_address;
+	uint16_t idx;
 
-    idx = GetIndexIntoMasterDataTable(voltageobject_info);
-    table_address =	cgs_atom_get_data_table(hwmgr->device,
-    		idx, NULL, NULL, NULL);
+	idx = GetIndexIntoMasterDataTable(voltageobject_info);
+	table_address = smu_atom_get_data_table(hwmgr->adev,
+						idx, NULL, NULL, NULL);
 
-    PP_ASSERT_WITH_CODE( 
-        table_address,
-        "Error retrieving BIOS Table Address!",
-        return NULL);
+	PP_ASSERT_WITH_CODE(table_address,
+			"Error retrieving BIOS Table Address!",
+			return NULL);
 
-    return (struct atom_voltage_objects_info_v4_1 *)table_address;
+	return (struct atom_voltage_objects_info_v4_1 *)table_address;
 }
 
 /**
@@ -167,7 +166,7 @@ static struct atom_gpio_pin_lut_v2_1 *pp_atomfwctrl_get_gpio_lookup_table(
 	uint16_t idx;
 
 	idx = GetIndexIntoMasterDataTable(gpio_pin_lut);
-	table_address =	cgs_atom_get_data_table(hwmgr->device,
+	table_address =	smu_atom_get_data_table(hwmgr->adev,
 			idx, NULL, NULL, NULL);
 	PP_ASSERT_WITH_CODE(table_address,
 			"Error retrieving BIOS Table Address!",
@@ -248,28 +247,30 @@ int pp_atomfwctrl_get_gpu_pll_dividers_vega10(struct pp_hwmgr *hwmgr,
 		uint32_t clock_type, uint32_t clock_value,
 		struct pp_atomfwctrl_clock_dividers_soc15 *dividers)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	struct compute_gpu_clock_input_parameter_v1_8 pll_parameters;
 	struct compute_gpu_clock_output_parameter_v1_8 *pll_output;
-	int result;
 	uint32_t idx;
 
 	pll_parameters.gpuclock_10khz = (uint32_t)clock_value;
 	pll_parameters.gpu_clock_type = clock_type;
 
 	idx = GetIndexIntoMasterCmdTable(computegpuclockparam);
-	result = cgs_atom_exec_cmd_table(hwmgr->device, idx, &pll_parameters);
-
-	if (!result) {
-		pll_output = (struct compute_gpu_clock_output_parameter_v1_8 *)
-				&pll_parameters;
-		dividers->ulClock = le32_to_cpu(pll_output->gpuclock_10khz);
-		dividers->ulDid = le32_to_cpu(pll_output->dfs_did);
-		dividers->ulPll_fb_mult = le32_to_cpu(pll_output->pll_fb_mult);
-		dividers->ulPll_ss_fbsmult = le32_to_cpu(pll_output->pll_ss_fbsmult);
-		dividers->usPll_ss_slew_frac = le16_to_cpu(pll_output->pll_ss_slew_frac);
-		dividers->ucPll_ss_enable = pll_output->pll_ss_enable;
-	}
-	return result;
+
+	if (amdgpu_atom_execute_table(
+		adev->mode_info.atom_context, idx, (uint32_t *)&pll_parameters))
+		return -EINVAL;
+
+	pll_output = (struct compute_gpu_clock_output_parameter_v1_8 *)
+			&pll_parameters;
+	dividers->ulClock = le32_to_cpu(pll_output->gpuclock_10khz);
+	dividers->ulDid = le32_to_cpu(pll_output->dfs_did);
+	dividers->ulPll_fb_mult = le32_to_cpu(pll_output->pll_fb_mult);
+	dividers->ulPll_ss_fbsmult = le32_to_cpu(pll_output->pll_ss_fbsmult);
+	dividers->usPll_ss_slew_frac = le16_to_cpu(pll_output->pll_ss_slew_frac);
+	dividers->ucPll_ss_enable = pll_output->pll_ss_enable;
+
+	return 0;
 }
 
 int pp_atomfwctrl_get_avfs_information(struct pp_hwmgr *hwmgr,
@@ -283,7 +284,7 @@ int pp_atomfwctrl_get_avfs_information(struct pp_hwmgr *hwmgr,
 
 	idx = GetIndexIntoMasterDataTable(asic_profiling_info);
 	profile = (struct atom_asic_profiling_info_v4_1 *)
-			cgs_atom_get_data_table(hwmgr->device,
+			smu_atom_get_data_table(hwmgr->adev,
 					idx, NULL, NULL, NULL);
 
 	if (!profile)
@@ -467,7 +468,7 @@ int pp_atomfwctrl_get_gpio_information(struct pp_hwmgr *hwmgr,
 
 	idx = GetIndexIntoMasterDataTable(smu_info);
 	info = (struct atom_smu_info_v3_1 *)
-		cgs_atom_get_data_table(hwmgr->device,
+		smu_atom_get_data_table(hwmgr->adev,
 				idx, NULL, NULL, NULL);
 
 	if (!info) {
@@ -487,37 +488,107 @@ int pp_atomfwctrl_get_gpio_information(struct pp_hwmgr *hwmgr,
 	return 0;
 }
 
-int pp_atomfwctrl__get_clk_information_by_clkid(struct pp_hwmgr *hwmgr, BIOS_CLKID id, uint32_t *frequency)
+int pp_atomfwctrl_get_clk_information_by_clkid(struct pp_hwmgr *hwmgr, BIOS_CLKID id, uint32_t *frequency)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	struct atom_get_smu_clock_info_parameters_v3_1   parameters;
 	struct atom_get_smu_clock_info_output_parameters_v3_1 *output;
 	uint32_t ix;
 
 	parameters.clk_id = id;
+	parameters.syspll_id = 0;
 	parameters.command = GET_SMU_CLOCK_INFO_V3_1_GET_CLOCK_FREQ;
+	parameters.dfsdid = 0;
 
 	ix = GetIndexIntoMasterCmdTable(getsmuclockinfo);
-	if (!cgs_atom_exec_cmd_table(hwmgr->device, ix, &parameters)) {
-		output = (struct atom_get_smu_clock_info_output_parameters_v3_1 *)&parameters;
-		*frequency = output->atom_smu_outputclkfreq.smu_clock_freq_hz / 10000;
-	} else {
-		pr_info("Error execute_table getsmuclockinfo!");
-		return -1;
-	}
+
+	if (amdgpu_atom_execute_table(
+		adev->mode_info.atom_context, ix, (uint32_t *)&parameters))
+		return -EINVAL;
+
+	output = (struct atom_get_smu_clock_info_output_parameters_v3_1 *)&parameters;
+	*frequency = le32_to_cpu(output->atom_smu_outputclkfreq.smu_clock_freq_hz) / 10000;
 
 	return 0;
 }
 
+static void pp_atomfwctrl_copy_vbios_bootup_values_3_2(struct pp_hwmgr *hwmgr,
+			struct pp_atomfwctrl_bios_boot_up_values *boot_values,
+			struct atom_firmware_info_v3_2 *fw_info)
+{
+	uint32_t frequency = 0;
+
+	boot_values->ulRevision = fw_info->firmware_revision;
+	boot_values->ulGfxClk   = fw_info->bootup_sclk_in10khz;
+	boot_values->ulUClk     = fw_info->bootup_mclk_in10khz;
+	boot_values->usVddc     = fw_info->bootup_vddc_mv;
+	boot_values->usVddci    = fw_info->bootup_vddci_mv;
+	boot_values->usMvddc    = fw_info->bootup_mvddc_mv;
+	boot_values->usVddGfx   = fw_info->bootup_vddgfx_mv;
+	boot_values->ucCoolingID = fw_info->coolingsolution_id;
+	boot_values->ulSocClk   = 0;
+	boot_values->ulDCEFClk   = 0;
+
+	if (!pp_atomfwctrl_get_clk_information_by_clkid(hwmgr, SMU11_SYSPLL0_SOCCLK_ID, &frequency))
+		boot_values->ulSocClk   = frequency;
+
+	if (!pp_atomfwctrl_get_clk_information_by_clkid(hwmgr, SMU11_SYSPLL0_DCEFCLK_ID, &frequency))
+		boot_values->ulDCEFClk  = frequency;
+
+	if (!pp_atomfwctrl_get_clk_information_by_clkid(hwmgr, SMU11_SYSPLL0_ECLK_ID, &frequency))
+		boot_values->ulEClk     = frequency;
+
+	if (!pp_atomfwctrl_get_clk_information_by_clkid(hwmgr, SMU11_SYSPLL0_VCLK_ID, &frequency))
+		boot_values->ulVClk     = frequency;
+
+	if (!pp_atomfwctrl_get_clk_information_by_clkid(hwmgr, SMU11_SYSPLL0_DCLK_ID, &frequency))
+		boot_values->ulDClk     = frequency;
+}
+
+static void pp_atomfwctrl_copy_vbios_bootup_values_3_1(struct pp_hwmgr *hwmgr,
+			struct pp_atomfwctrl_bios_boot_up_values *boot_values,
+			struct atom_firmware_info_v3_1 *fw_info)
+{
+	uint32_t frequency = 0;
+
+	boot_values->ulRevision = fw_info->firmware_revision;
+	boot_values->ulGfxClk   = fw_info->bootup_sclk_in10khz;
+	boot_values->ulUClk     = fw_info->bootup_mclk_in10khz;
+	boot_values->usVddc     = fw_info->bootup_vddc_mv;
+	boot_values->usVddci    = fw_info->bootup_vddci_mv;
+	boot_values->usMvddc    = fw_info->bootup_mvddc_mv;
+	boot_values->usVddGfx   = fw_info->bootup_vddgfx_mv;
+	boot_values->ucCoolingID = fw_info->coolingsolution_id;
+	boot_values->ulSocClk   = 0;
+	boot_values->ulDCEFClk   = 0;
+
+	if (!pp_atomfwctrl_get_clk_information_by_clkid(hwmgr, SMU9_SYSPLL0_SOCCLK_ID, &frequency))
+		boot_values->ulSocClk   = frequency;
+
+	if (!pp_atomfwctrl_get_clk_information_by_clkid(hwmgr, SMU9_SYSPLL0_DCEFCLK_ID, &frequency))
+		boot_values->ulDCEFClk  = frequency;
+
+	if (!pp_atomfwctrl_get_clk_information_by_clkid(hwmgr, SMU9_SYSPLL0_ECLK_ID, &frequency))
+		boot_values->ulEClk     = frequency;
+
+	if (!pp_atomfwctrl_get_clk_information_by_clkid(hwmgr, SMU9_SYSPLL0_VCLK_ID, &frequency))
+		boot_values->ulVClk     = frequency;
+
+	if (!pp_atomfwctrl_get_clk_information_by_clkid(hwmgr, SMU9_SYSPLL0_DCLK_ID, &frequency))
+		boot_values->ulDClk     = frequency;
+}
+
 int pp_atomfwctrl_get_vbios_bootup_values(struct pp_hwmgr *hwmgr,
 			struct pp_atomfwctrl_bios_boot_up_values *boot_values)
 {
-	struct atom_firmware_info_v3_1 *info = NULL;
+	struct atom_firmware_info_v3_2 *fwinfo_3_2;
+	struct atom_firmware_info_v3_1 *fwinfo_3_1;
+	struct atom_common_table_header *info = NULL;
 	uint16_t ix;
-	uint32_t frequency = 0;
 
 	ix = GetIndexIntoMasterDataTable(firmwareinfo);
-	info = (struct atom_firmware_info_v3_1 *)
-		cgs_atom_get_data_table(hwmgr->device,
+	info = (struct atom_common_table_header *)
+		smu_atom_get_data_table(hwmgr->adev,
 				ix, NULL, NULL, NULL);
 
 	if (!info) {
@@ -525,22 +596,18 @@ int pp_atomfwctrl_get_vbios_bootup_values(struct pp_hwmgr *hwmgr,
 		return -EINVAL;
 	}
 
-	boot_values->ulRevision = info->firmware_revision;
-	boot_values->ulGfxClk   = info->bootup_sclk_in10khz;
-	boot_values->ulUClk     = info->bootup_mclk_in10khz;
-	boot_values->usVddc     = info->bootup_vddc_mv;
-	boot_values->usVddci    = info->bootup_vddci_mv;
-	boot_values->usMvddc    = info->bootup_mvddc_mv;
-	boot_values->usVddGfx   = info->bootup_vddgfx_mv;
-	boot_values->ucCoolingID = info->coolingsolution_id;
-	boot_values->ulSocClk   = 0;
-	boot_values->ulDCEFClk   = 0;
-
-	if (!pp_atomfwctrl__get_clk_information_by_clkid(hwmgr, SMU9_SYSPLL0_SOCCLK_ID, &frequency))
-		boot_values->ulSocClk   = frequency;
-
-	if (!pp_atomfwctrl__get_clk_information_by_clkid(hwmgr, SMU9_SYSPLL0_DCEFCLK_ID, &frequency))
-		boot_values->ulDCEFClk   = frequency;
+	if ((info->format_revision == 3) && (info->content_revision == 2)) {
+		fwinfo_3_2 = (struct atom_firmware_info_v3_2 *)info;
+		pp_atomfwctrl_copy_vbios_bootup_values_3_2(hwmgr,
+				boot_values, fwinfo_3_2);
+	} else if ((info->format_revision == 3) && (info->content_revision == 1)) {
+		fwinfo_3_1 = (struct atom_firmware_info_v3_1 *)info;
+		pp_atomfwctrl_copy_vbios_bootup_values_3_1(hwmgr,
+				boot_values, fwinfo_3_1);
+	} else {
+		pr_info("Fw info table revision does not match!");
+		return -EINVAL;
+	}
 
 	return 0;
 }
@@ -553,7 +620,7 @@ int pp_atomfwctrl_get_smc_dpm_information(struct pp_hwmgr *hwmgr,
 
 	ix = GetIndexIntoMasterDataTable(smc_dpm_info);
 	info = (struct atom_smc_dpm_info_v4_1 *)
-		cgs_atom_get_data_table(hwmgr->device,
+		smu_atom_get_data_table(hwmgr->adev,
 				ix, NULL, NULL, NULL);
 	if (!info) {
 		pr_info("Error retrieving BIOS Table Address!");
@@ -632,5 +699,7 @@ int pp_atomfwctrl_get_smc_dpm_information(struct pp_hwmgr *hwmgr,
 	param->acggfxclkspreadpercent = info->acggfxclkspreadpercent;
 	param->acggfxclkspreadfreq = info->acggfxclkspreadfreq;
 
+	param->Vr2_I2C_address = info->Vr2_I2C_address;
+
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h
index 8df1e84f27c9..22e21668c93a 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/ppatomfwctrl.h
@@ -136,6 +136,9 @@ struct pp_atomfwctrl_bios_boot_up_values {
 	uint32_t   ulUClk;
 	uint32_t   ulSocClk;
 	uint32_t   ulDCEFClk;
+	uint32_t   ulEClk;
+	uint32_t   ulVClk;
+	uint32_t   ulDClk;
 	uint16_t   usVddc;
 	uint16_t   usVddci;
 	uint16_t   usMvddc;
@@ -207,6 +210,8 @@ struct pp_atomfwctrl_smc_dpm_parameters
 	uint8_t  acggfxclkspreadenabled;
 	uint8_t  acggfxclkspreadpercent;
 	uint16_t acggfxclkspreadfreq;
+
+	uint8_t Vr2_I2C_address;
 };
 
 int pp_atomfwctrl_get_gpu_pll_dividers_vega10(struct pp_hwmgr *hwmgr,
@@ -230,6 +235,8 @@ int pp_atomfwctrl_get_vbios_bootup_values(struct pp_hwmgr *hwmgr,
 			struct pp_atomfwctrl_bios_boot_up_values *boot_values);
 int pp_atomfwctrl_get_smc_dpm_information(struct pp_hwmgr *hwmgr,
 			struct pp_atomfwctrl_smc_dpm_parameters *param);
+int pp_atomfwctrl_get_clk_information_by_clkid(struct pp_hwmgr *hwmgr,
+					BIOS_CLKID id, uint32_t *frequency);
 
 #endif
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c b/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c
index c9eecce5683f..4e1fd5393845 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/process_pptables_v1_0.c
@@ -141,7 +141,7 @@ static const void *get_powerplay_table(struct pp_hwmgr *hwmgr)
 
 	if (!table_address) {
 		table_address = (ATOM_Tonga_POWERPLAYTABLE *)
-				cgs_atom_get_data_table(hwmgr->device,
+				smu_atom_get_data_table(hwmgr->adev,
 						index, &size, &frev, &crev);
 		hwmgr->soft_pp_table = table_address;	/*Cache the result in RAM.*/
 		hwmgr->soft_pp_table_size = size;
@@ -183,10 +183,10 @@ static int get_vddc_lookup_table(
 					ATOM_Tonga_Voltage_Lookup_Record,
 					entries, vddc_lookup_pp_tables, i);
 		record->us_calculated = 0;
-		record->us_vdd = atom_record->usVdd;
-		record->us_cac_low = atom_record->usCACLow;
-		record->us_cac_mid = atom_record->usCACMid;
-		record->us_cac_high = atom_record->usCACHigh;
+		record->us_vdd = le16_to_cpu(atom_record->usVdd);
+		record->us_cac_low = le16_to_cpu(atom_record->usCACLow);
+		record->us_cac_mid = le16_to_cpu(atom_record->usCACMid);
+		record->us_cac_high = le16_to_cpu(atom_record->usCACHigh);
 	}
 
 	*lookup_table = table;
@@ -728,6 +728,32 @@ static int get_mm_clock_voltage_table(
 	return 0;
 }
 
+static int get_gpio_table(struct pp_hwmgr *hwmgr,
+		struct phm_ppt_v1_gpio_table **pp_tonga_gpio_table,
+		const ATOM_Tonga_GPIO_Table *atom_gpio_table)
+{
+	uint32_t table_size;
+	struct phm_ppt_v1_gpio_table *pp_gpio_table;
+	struct phm_ppt_v1_information *pp_table_information =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+
+	table_size = sizeof(struct phm_ppt_v1_gpio_table);
+	pp_gpio_table = kzalloc(table_size, GFP_KERNEL);
+	if (!pp_gpio_table)
+		return -ENOMEM;
+
+	if (pp_table_information->vdd_dep_on_sclk->count <
+			atom_gpio_table->ucVRHotTriggeredSclkDpmIndex)
+		PP_ASSERT_WITH_CODE(false,
+				"SCLK DPM index for VRHot cannot exceed the total sclk level count!",);
+	else
+		pp_gpio_table->vrhot_triggered_sclk_dpm_index =
+				atom_gpio_table->ucVRHotTriggeredSclkDpmIndex;
+
+	*pp_tonga_gpio_table = pp_gpio_table;
+
+	return 0;
+}
 /**
  * Private Function used during initialization.
  * Initialize clock voltage dependency
@@ -761,11 +787,15 @@ static int init_clock_voltage_dependency(
 	const PPTable_Generic_SubTable_Header *pcie_table =
 		(const PPTable_Generic_SubTable_Header *)(((unsigned long) powerplay_table) +
 		le16_to_cpu(powerplay_table->usPCIETableOffset));
+	const ATOM_Tonga_GPIO_Table *gpio_table =
+		(const ATOM_Tonga_GPIO_Table *)(((unsigned long) powerplay_table) +
+		le16_to_cpu(powerplay_table->usGPIOTableOffset));
 
 	pp_table_information->vdd_dep_on_sclk = NULL;
 	pp_table_information->vdd_dep_on_mclk = NULL;
 	pp_table_information->mm_dep_table = NULL;
 	pp_table_information->pcie_table = NULL;
+	pp_table_information->gpio_table = NULL;
 
 	if (powerplay_table->usMMDependencyTableOffset != 0)
 		result = get_mm_clock_voltage_table(hwmgr,
@@ -810,6 +840,10 @@ static int init_clock_voltage_dependency(
 		result = get_valid_clk(hwmgr, &pp_table_information->valid_sclk_values,
 		pp_table_information->vdd_dep_on_sclk);
 
+	if (!result && gpio_table)
+		result = get_gpio_table(hwmgr, &pp_table_information->gpio_table,
+				gpio_table);
+
 	return result;
 }
 
@@ -836,12 +870,6 @@ static int init_over_drive_limits(
 	hwmgr->platform_descriptor.maxOverdriveVDDC = 0;
 	hwmgr->platform_descriptor.overdriveVDDCStep = 0;
 
-	if (hwmgr->platform_descriptor.overdriveLimit.engineClock == 0 \
-		|| hwmgr->platform_descriptor.overdriveLimit.memoryClock == 0) {
-		hwmgr->od_enabled = false;
-		pr_debug("OverDrive feature not support by VBIOS\n");
-	}
-
 	return 0;
 }
 
@@ -1116,6 +1144,9 @@ static int pp_tables_v1_0_uninitialize(struct pp_hwmgr *hwmgr)
 	kfree(pp_table_information->pcie_table);
 	pp_table_information->pcie_table = NULL;
 
+	kfree(pp_table_information->gpio_table);
+	pp_table_information->gpio_table = NULL;
+
 	kfree(hwmgr->pptable);
 	hwmgr->pptable = NULL;
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c
index 36ca7c419c90..925e17104f90 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/processpptables.c
@@ -837,7 +837,7 @@ static const ATOM_PPLIB_POWERPLAYTABLE *get_powerplay_table(
 			hwmgr->soft_pp_table = &soft_dummy_pp_table[0];
 			hwmgr->soft_pp_table_size = sizeof(soft_dummy_pp_table);
 		} else {
-			table_addr = cgs_atom_get_data_table(hwmgr->device,
+			table_addr = smu_atom_get_data_table(hwmgr->adev,
 					GetIndexIntoMasterTable(DATA, PowerPlayInfo),
 					&size, &frev, &crev);
 			hwmgr->soft_pp_table = table_addr;
@@ -1058,7 +1058,7 @@ static int init_overdrive_limits(struct pp_hwmgr *hwmgr,
 		return 0;
 
 	/* We assume here that fw_info is unchanged if this call fails.*/
-	fw_info = cgs_atom_get_data_table(hwmgr->device,
+	fw_info = smu_atom_get_data_table(hwmgr->adev,
 			 GetIndexIntoMasterTable(DATA, FirmwareInfo),
 			 &size, &frev, &crev);
 
@@ -1074,12 +1074,6 @@ static int init_overdrive_limits(struct pp_hwmgr *hwmgr,
 				powerplay_table,
 				(const ATOM_FIRMWARE_INFO_V2_1 *)fw_info);
 
-	if (hwmgr->platform_descriptor.overdriveLimit.engineClock == 0
-		&& hwmgr->platform_descriptor.overdriveLimit.memoryClock == 0) {
-		hwmgr->od_enabled = false;
-		pr_debug("OverDrive feature not support by VBIOS\n");
-	}
-
 	return result;
 }
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
index 10253b89b3d8..a63e00653324 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.c
@@ -34,7 +34,7 @@
 #include "rv_ppsmc.h"
 #include "smu10_hwmgr.h"
 #include "power_state.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 
 #define SMU10_MAX_DEEPSLEEP_DIVIDER_ID     5
 #define SMU10_MINIMUM_ENGINE_CLOCK         800   /* 8Mhz, the low boundary of engine clock allowed on this chip */
@@ -42,12 +42,48 @@
 #define SMU10_DISPCLK_BYPASS_THRESHOLD     10000 /* 100Mhz */
 #define SMC_RAM_END                     0x40000
 
+#define mmPWR_MISC_CNTL_STATUS					0x0183
+#define mmPWR_MISC_CNTL_STATUS_BASE_IDX				0
+#define PWR_MISC_CNTL_STATUS__PWR_GFX_RLC_CGPG_EN__SHIFT	0x0
+#define PWR_MISC_CNTL_STATUS__PWR_GFXOFF_STATUS__SHIFT		0x1
+#define PWR_MISC_CNTL_STATUS__PWR_GFX_RLC_CGPG_EN_MASK		0x00000001L
+#define PWR_MISC_CNTL_STATUS__PWR_GFXOFF_STATUS_MASK		0x00000006L
+
 static const unsigned long SMU10_Magic = (unsigned long) PHM_Rv_Magic;
 
 
 static int smu10_display_clock_voltage_request(struct pp_hwmgr *hwmgr,
-		struct pp_display_clock_request *clock_req);
+		struct pp_display_clock_request *clock_req)
+{
+	struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend);
+	enum amd_pp_clock_type clk_type = clock_req->clock_type;
+	uint32_t clk_freq = clock_req->clock_freq_in_khz / 1000;
+	PPSMC_Msg        msg;
 
+	switch (clk_type) {
+	case amd_pp_dcf_clock:
+		if (clk_freq == smu10_data->dcf_actual_hard_min_freq)
+			return 0;
+		msg =  PPSMC_MSG_SetHardMinDcefclkByFreq;
+		smu10_data->dcf_actual_hard_min_freq = clk_freq;
+		break;
+	case amd_pp_soc_clock:
+		 msg = PPSMC_MSG_SetHardMinSocclkByFreq;
+		break;
+	case amd_pp_f_clock:
+		if (clk_freq == smu10_data->f_actual_hard_min_freq)
+			return 0;
+		smu10_data->f_actual_hard_min_freq = clk_freq;
+		msg = PPSMC_MSG_SetHardMinFclkByFreq;
+		break;
+	default:
+		pr_info("[DisplayClockVoltageRequest]Invalid Clock Type!");
+		return -EINVAL;
+	}
+	smum_send_msg_to_smc_with_parameter(hwmgr, msg, clk_freq);
+
+	return 0;
+}
 
 static struct smu10_power_state *cast_smu10_ps(struct pp_hw_power_state *hw_ps)
 {
@@ -74,11 +110,15 @@ static int smu10_initialize_dpm_defaults(struct pp_hwmgr *hwmgr)
 	smu10_data->thermal_auto_throttling_treshold = 0;
 	smu10_data->is_nb_dpm_enabled = 1;
 	smu10_data->dpm_flags = 1;
-	smu10_data->gfx_off_controled_by_driver = false;
 	smu10_data->need_min_deep_sleep_dcefclk = true;
 	smu10_data->num_active_display = 0;
 	smu10_data->deep_sleep_dcefclk = 0;
 
+	if (hwmgr->feature_mask & PP_GFXOFF_MASK)
+		smu10_data->gfx_off_controled_by_driver = true;
+	else
+		smu10_data->gfx_off_controled_by_driver = false;
+
 	phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
 					PHM_PlatformCaps_SclkDeepSleep);
 
@@ -161,7 +201,7 @@ static int smu10_set_clock_limit(struct pp_hwmgr *hwmgr, const void *input)
 	struct PP_Clocks clocks = {0};
 	struct pp_display_clock_request clock_req;
 
-	clocks.dcefClock = hwmgr->display_config.min_dcef_set_clk;
+	clocks.dcefClock = hwmgr->display_config->min_dcef_set_clk;
 	clock_req.clock_type = amd_pp_dcf_clock;
 	clock_req.clock_freq_in_khz = clocks.dcefClock * 10;
 
@@ -206,12 +246,18 @@ static int smu10_set_power_state_tasks(struct pp_hwmgr *hwmgr, const void *input
 static int smu10_init_power_gate_state(struct pp_hwmgr *hwmgr)
 {
 	struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend);
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	smu10_data->vcn_power_gated = true;
 	smu10_data->isp_tileA_power_gated = true;
 	smu10_data->isp_tileB_power_gated = true;
 
-	return 0;
+	if (adev->pg_flags & AMD_PG_SUPPORT_GFX_PG)
+		return smum_send_msg_to_smc_with_parameter(hwmgr,
+							   PPSMC_MSG_SetGfxCGPG,
+							   true);
+	else
+		return 0;
 }
 
 
@@ -237,19 +283,37 @@ static int smu10_power_off_asic(struct pp_hwmgr *hwmgr)
 	return smu10_reset_cc6_data(hwmgr);
 }
 
+static bool smu10_is_gfx_on(struct pp_hwmgr *hwmgr)
+{
+	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
+
+	reg = RREG32_SOC15(PWR, 0, mmPWR_MISC_CNTL_STATUS);
+	if ((reg & PWR_MISC_CNTL_STATUS__PWR_GFXOFF_STATUS_MASK) ==
+	    (0x2 << PWR_MISC_CNTL_STATUS__PWR_GFXOFF_STATUS__SHIFT))
+		return true;
+
+	return false;
+}
+
 static int smu10_disable_gfx_off(struct pp_hwmgr *hwmgr)
 {
 	struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend);
 
-	if (smu10_data->gfx_off_controled_by_driver)
+	if (smu10_data->gfx_off_controled_by_driver) {
 		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_DisableGfxOff);
 
+		/* confirm gfx is back to "on" state */
+		while (!smu10_is_gfx_on(hwmgr))
+			msleep(1);
+	}
+
 	return 0;
 }
 
 static int smu10_disable_dpm_tasks(struct pp_hwmgr *hwmgr)
 {
-	return smu10_disable_gfx_off(hwmgr);
+	return 0;
 }
 
 static int smu10_enable_gfx_off(struct pp_hwmgr *hwmgr)
@@ -264,7 +328,15 @@ static int smu10_enable_gfx_off(struct pp_hwmgr *hwmgr)
 
 static int smu10_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
 {
-	return smu10_enable_gfx_off(hwmgr);
+	return 0;
+}
+
+static int smu10_gfx_off_control(struct pp_hwmgr *hwmgr, bool enable)
+{
+	if (enable)
+		return smu10_enable_gfx_off(hwmgr);
+	else
+		return smu10_disable_gfx_off(hwmgr);
 }
 
 static int smu10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
@@ -340,7 +412,7 @@ static int smu10_get_clock_voltage_dependency_table(struct pp_hwmgr *hwmgr,
 
 static int smu10_populate_clock_table(struct pp_hwmgr *hwmgr)
 {
-	int result;
+	uint32_t result;
 
 	struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend);
 	DpmClocks_t  *table = &(smu10_data->clock_table);
@@ -386,11 +458,11 @@ static int smu10_populate_clock_table(struct pp_hwmgr *hwmgr)
 
 	smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetMinGfxclkFrequency);
 	result = smum_get_argument(hwmgr);
-	smu10_data->gfx_min_freq_limit = result * 100;
+	smu10_data->gfx_min_freq_limit = result / 10 * 1000;
 
 	smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetMaxGfxclkFrequency);
 	result = smum_get_argument(hwmgr);
-	smu10_data->gfx_max_freq_limit = result * 100;
+	smu10_data->gfx_max_freq_limit = result / 10 * 1000;
 
 	return 0;
 }
@@ -436,8 +508,8 @@ static int smu10_hwmgr_backend_init(struct pp_hwmgr *hwmgr)
 
 	hwmgr->platform_descriptor.minimumClocksReductionPercentage = 50;
 
-	hwmgr->pstate_sclk = SMU10_UMD_PSTATE_GFXCLK;
-	hwmgr->pstate_mclk = SMU10_UMD_PSTATE_FCLK;
+	hwmgr->pstate_sclk = SMU10_UMD_PSTATE_GFXCLK * 100;
+	hwmgr->pstate_mclk = SMU10_UMD_PSTATE_FCLK * 100;
 
 	return result;
 }
@@ -472,6 +544,8 @@ static int smu10_hwmgr_backend_fini(struct pp_hwmgr *hwmgr)
 static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 				enum amd_dpm_forced_level level)
 {
+	struct smu10_hwmgr *data = hwmgr->backend;
+
 	if (hwmgr->smu_version < 0x1E3700) {
 		pr_info("smu firmware version too old, can not set dpm level\n");
 		return 0;
@@ -482,7 +556,7 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 	case AMD_DPM_FORCED_LEVEL_PROFILE_PEAK:
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinGfxClk,
-						SMU10_UMD_PSTATE_PEAK_GFXCLK);
+						data->gfx_max_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinFclkByFreq,
 						SMU10_UMD_PSTATE_PEAK_FCLK);
@@ -495,7 +569,7 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetSoftMaxGfxClk,
-						SMU10_UMD_PSTATE_PEAK_GFXCLK);
+						data->gfx_max_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetSoftMaxFclkByFreq,
 						SMU10_UMD_PSTATE_PEAK_FCLK);
@@ -509,10 +583,10 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 	case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK:
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinGfxClk,
-						SMU10_UMD_PSTATE_MIN_GFXCLK);
+						data->gfx_min_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetSoftMaxGfxClk,
-						SMU10_UMD_PSTATE_MIN_GFXCLK);
+						data->gfx_min_freq_limit/100);
 		break;
 	case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK:
 		smum_send_msg_to_smc_with_parameter(hwmgr,
@@ -552,10 +626,13 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 	case AMD_DPM_FORCED_LEVEL_AUTO:
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinGfxClk,
-						SMU10_UMD_PSTATE_MIN_GFXCLK);
+						data->gfx_min_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinFclkByFreq,
+						hwmgr->display_config->num_display > 3 ?
+						SMU10_UMD_PSTATE_PEAK_FCLK :
 						SMU10_UMD_PSTATE_MIN_FCLK);
+
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinSocclkByFreq,
 						SMU10_UMD_PSTATE_MIN_SOCCLK);
@@ -565,7 +642,7 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetSoftMaxGfxClk,
-						SMU10_UMD_PSTATE_PEAK_GFXCLK);
+						data->gfx_max_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetSoftMaxFclkByFreq,
 						SMU10_UMD_PSTATE_PEAK_FCLK);
@@ -579,10 +656,10 @@ static int smu10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 	case AMD_DPM_FORCED_LEVEL_LOW:
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinGfxClk,
-						SMU10_UMD_PSTATE_MIN_GFXCLK);
+						data->gfx_min_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetSoftMaxGfxClk,
-						SMU10_UMD_PSTATE_MIN_GFXCLK);
+						data->gfx_min_freq_limit/100);
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SetHardMinFclkByFreq,
 						SMU10_UMD_PSTATE_MIN_FCLK);
@@ -699,6 +776,16 @@ static int smu10_set_cpu_power_state(struct pp_hwmgr *hwmgr)
 static int smu10_store_cc6_data(struct pp_hwmgr *hwmgr, uint32_t separation_time,
 			bool cc6_disable, bool pstate_disable, bool pstate_switch_disable)
 {
+	struct smu10_hwmgr *data = (struct smu10_hwmgr *)(hwmgr->backend);
+
+	if (separation_time != data->separation_time ||
+			cc6_disable != data->cc6_disable ||
+			pstate_disable != data->pstate_disable) {
+		data->separation_time = separation_time;
+		data->cc6_disable = cc6_disable;
+		data->pstate_disable = pstate_disable;
+		data->cc6_setting_changed = true;
+	}
 	return 0;
 }
 
@@ -711,6 +798,51 @@ static int smu10_get_dal_power_level(struct pp_hwmgr *hwmgr,
 static int smu10_force_clock_level(struct pp_hwmgr *hwmgr,
 		enum pp_clock_type type, uint32_t mask)
 {
+	struct smu10_hwmgr *data = hwmgr->backend;
+	struct smu10_voltage_dependency_table *mclk_table =
+					data->clock_vol_info.vdd_dep_on_fclk;
+	uint32_t low, high;
+
+	low = mask ? (ffs(mask) - 1) : 0;
+	high = mask ? (fls(mask) - 1) : 0;
+
+	switch (type) {
+	case PP_SCLK:
+		if (low > 2 || high > 2) {
+			pr_info("Currently sclk only support 3 levels on RV\n");
+			return -EINVAL;
+		}
+
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+						PPSMC_MSG_SetHardMinGfxClk,
+						low == 2 ? data->gfx_max_freq_limit/100 :
+						low == 1 ? SMU10_UMD_PSTATE_GFXCLK :
+						data->gfx_min_freq_limit/100);
+
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+						PPSMC_MSG_SetSoftMaxGfxClk,
+						high == 0 ? data->gfx_min_freq_limit/100 :
+						high == 1 ? SMU10_UMD_PSTATE_GFXCLK :
+						data->gfx_max_freq_limit/100);
+		break;
+
+	case PP_MCLK:
+		if (low > mclk_table->count - 1 || high > mclk_table->count - 1)
+			return -EINVAL;
+
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+						PPSMC_MSG_SetHardMinFclkByFreq,
+						mclk_table->entries[low].clk/100);
+
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+						PPSMC_MSG_SetSoftMaxFclkByFreq,
+						mclk_table->entries[high].clk/100);
+		break;
+
+	case PP_PCIE:
+	default:
+		break;
+	}
 	return 0;
 }
 
@@ -720,21 +852,30 @@ static int smu10_print_clock_levels(struct pp_hwmgr *hwmgr,
 	struct smu10_hwmgr *data = (struct smu10_hwmgr *)(hwmgr->backend);
 	struct smu10_voltage_dependency_table *mclk_table =
 			data->clock_vol_info.vdd_dep_on_fclk;
-	int i, now, size = 0;
+	uint32_t i, now, size = 0;
 
 	switch (type) {
 	case PP_SCLK:
 		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetGfxclkFrequency);
 		now = smum_get_argument(hwmgr);
 
+	/* driver only know min/max gfx_clk, Add level 1 for all other gfx clks */
+		if (now == data->gfx_max_freq_limit/100)
+			i = 2;
+		else if (now == data->gfx_min_freq_limit/100)
+			i = 0;
+		else
+			i = 1;
+
 		size += sprintf(buf + size, "0: %uMhz %s\n",
-				data->gfx_min_freq_limit / 100,
-				((data->gfx_min_freq_limit / 100)
-				 == now) ? "*" : "");
+					data->gfx_min_freq_limit/100,
+					i == 0 ? "*" : "");
 		size += sprintf(buf + size, "1: %uMhz %s\n",
-				data->gfx_max_freq_limit / 100,
-				((data->gfx_max_freq_limit / 100)
-				 == now) ? "*" : "");
+					i == 1 ? now : SMU10_UMD_PSTATE_GFXCLK,
+					i == 1 ? "*" : "");
+		size += sprintf(buf + size, "2: %uMhz %s\n",
+					data->gfx_max_freq_limit/100,
+					i == 2 ? "*" : "");
 		break;
 	case PP_MCLK:
 		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetFclkFrequency);
@@ -852,7 +993,7 @@ static int smu10_get_clock_by_type_with_latency(struct pp_hwmgr *hwmgr,
 
 	clocks->num_levels = 0;
 	for (i = 0; i < pclk_vol_table->count; i++) {
-		clocks->data[i].clocks_in_khz = pclk_vol_table->entries[i].clk;
+		clocks->data[i].clocks_in_khz = pclk_vol_table->entries[i].clk * 10;
 		clocks->data[i].latency_in_us = latency_required ?
 						smu10_get_mem_latency(hwmgr,
 						pclk_vol_table->entries[i].clk) :
@@ -888,6 +1029,12 @@ static int smu10_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr,
 	case amd_pp_soc_clock:
 		pclk_vol_table = pinfo->vdd_dep_on_socclk;
 		break;
+	case amd_pp_disp_clock:
+		pclk_vol_table = pinfo->vdd_dep_on_dispclk;
+		break;
+	case amd_pp_phy_clock:
+		pclk_vol_table = pinfo->vdd_dep_on_phyclk;
+		break;
 	default:
 		return -EINVAL;
 	}
@@ -897,7 +1044,7 @@ static int smu10_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr,
 
 	clocks->num_levels = 0;
 	for (i = 0; i < pclk_vol_table->count; i++) {
-		clocks->data[i].clocks_in_khz = pclk_vol_table->entries[i].clk;
+		clocks->data[i].clocks_in_khz = pclk_vol_table->entries[i].clk  * 10;
 		clocks->data[i].voltage_in_mv = pclk_vol_table->entries[i].vol;
 		clocks->num_levels++;
 	}
@@ -905,39 +1052,7 @@ static int smu10_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr,
 	return 0;
 }
 
-static int smu10_display_clock_voltage_request(struct pp_hwmgr *hwmgr,
-		struct pp_display_clock_request *clock_req)
-{
-	struct smu10_hwmgr *smu10_data = (struct smu10_hwmgr *)(hwmgr->backend);
-	enum amd_pp_clock_type clk_type = clock_req->clock_type;
-	uint32_t clk_freq = clock_req->clock_freq_in_khz / 1000;
-	PPSMC_Msg        msg;
-
-	switch (clk_type) {
-	case amd_pp_dcf_clock:
-		if (clk_freq == smu10_data->dcf_actual_hard_min_freq)
-			return 0;
-		msg =  PPSMC_MSG_SetHardMinDcefclkByFreq;
-		smu10_data->dcf_actual_hard_min_freq = clk_freq;
-		break;
-	case amd_pp_soc_clock:
-		 msg = PPSMC_MSG_SetHardMinSocclkByFreq;
-		break;
-	case amd_pp_f_clock:
-		if (clk_freq == smu10_data->f_actual_hard_min_freq)
-			return 0;
-		smu10_data->f_actual_hard_min_freq = clk_freq;
-		msg = PPSMC_MSG_SetHardMinFclkByFreq;
-		break;
-	default:
-		pr_info("[DisplayClockVoltageRequest]Invalid Clock Type!");
-		return -EINVAL;
-	}
-
-	smum_send_msg_to_smc_with_parameter(hwmgr, msg, clk_freq);
 
-	return 0;
-}
 
 static int smu10_get_max_high_clocks(struct pp_hwmgr *hwmgr, struct amd_pp_simple_clock_info *clocks)
 {
@@ -947,9 +1062,8 @@ static int smu10_get_max_high_clocks(struct pp_hwmgr *hwmgr, struct amd_pp_simpl
 
 static int smu10_thermal_get_temperature(struct pp_hwmgr *hwmgr)
 {
-	uint32_t reg_offset = soc15_get_register_offset(THM_HWID, 0,
-			mmTHM_TCON_CUR_TMP_BASE_IDX, mmTHM_TCON_CUR_TMP);
-	uint32_t reg_value = cgs_read_register(hwmgr->device, reg_offset);
+	struct amdgpu_device *adev = hwmgr->adev;
+	uint32_t reg_value = RREG32_SOC15(THM, 0, mmTHM_TCON_CUR_TMP);
 	int cur_temp =
 		(reg_value & THM_TCON_CUR_TMP__CUR_TEMP_MASK) >> THM_TCON_CUR_TMP__CUR_TEMP__SHIFT;
 
@@ -993,11 +1107,48 @@ static int smu10_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 	return ret;
 }
 
-static int smu10_set_mmhub_powergating_by_smu(struct pp_hwmgr *hwmgr)
+static int smu10_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr,
+		void *clock_ranges)
+{
+	struct smu10_hwmgr *data = hwmgr->backend;
+	struct dm_pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges = clock_ranges;
+	Watermarks_t *table = &(data->water_marks_table);
+	int result = 0;
+
+	smu_set_watermarks_for_clocks_ranges(table,wm_with_clock_ranges);
+	smum_smc_table_manager(hwmgr, (uint8_t *)table, (uint16_t)SMU10_WMTABLE, false);
+	data->water_marks_exist = true;
+	return result;
+}
+
+static int smu10_smus_notify_pwe(struct pp_hwmgr *hwmgr)
+{
+
+	return smum_send_msg_to_smc(hwmgr, PPSMC_MSG_SetRccPfcPmeRestoreRegister);
+}
+
+static int smu10_powergate_mmhub(struct pp_hwmgr *hwmgr)
 {
 	return smum_send_msg_to_smc(hwmgr, PPSMC_MSG_PowerGateMmHub);
 }
 
+static void smu10_powergate_vcn(struct pp_hwmgr *hwmgr, bool bgate)
+{
+	if (bgate) {
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
+						AMD_IP_BLOCK_TYPE_VCN,
+						AMD_PG_STATE_GATE);
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+					PPSMC_MSG_PowerDownVcn, 0);
+	} else {
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+						PPSMC_MSG_PowerUpVcn, 0);
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
+						AMD_IP_BLOCK_TYPE_VCN,
+						AMD_PG_STATE_UNGATE);
+	}
+}
+
 static const struct pp_hwmgr_func smu10_hwmgr_funcs = {
 	.backend_init = smu10_hwmgr_backend_init,
 	.backend_fini = smu10_hwmgr_backend_fini,
@@ -1006,7 +1157,7 @@ static const struct pp_hwmgr_func smu10_hwmgr_funcs = {
 	.force_dpm_level = smu10_dpm_force_dpm_level,
 	.get_power_state_size = smu10_get_power_state_size,
 	.powerdown_uvd = NULL,
-	.powergate_uvd = NULL,
+	.powergate_uvd = smu10_powergate_vcn,
 	.powergate_vce = NULL,
 	.get_mclk = smu10_dpm_get_mclk,
 	.get_sclk = smu10_dpm_get_sclk,
@@ -1022,6 +1173,7 @@ static const struct pp_hwmgr_func smu10_hwmgr_funcs = {
 	.get_current_shallow_sleep_clocks = smu10_get_current_shallow_sleep_clocks,
 	.get_clock_by_type_with_latency = smu10_get_clock_by_type_with_latency,
 	.get_clock_by_type_with_voltage = smu10_get_clock_by_type_with_voltage,
+	.set_watermarks_for_clocks_ranges = smu10_set_watermarks_for_clocks_ranges,
 	.get_max_high_clocks = smu10_get_max_high_clocks,
 	.read_sensor = smu10_read_sensor,
 	.set_active_display_count = smu10_set_active_display_count,
@@ -1031,7 +1183,11 @@ static const struct pp_hwmgr_func smu10_hwmgr_funcs = {
 	.asic_setup = smu10_setup_asic_task,
 	.power_state_set = smu10_set_power_state_tasks,
 	.dynamic_state_management_disable = smu10_disable_dpm_tasks,
-	.set_mmhub_powergating_by_smu = smu10_set_mmhub_powergating_by_smu,
+	.powergate_mmhub = smu10_powergate_mmhub,
+	.smus_notify_pwe = smu10_smus_notify_pwe,
+	.gfx_off_control = smu10_gfx_off_control,
+	.display_clock_voltage_request = smu10_display_clock_voltage_request,
+	.powergate_gfx = smu10_gfx_off_control,
 };
 
 int smu10_init_function_pointers(struct pp_hwmgr *hwmgr)
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.h
index 175c3a592b6c..1fb296a996f3 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu10_hwmgr.h
@@ -290,6 +290,7 @@ struct smu10_hwmgr {
 	bool                           vcn_dpg_mode;
 
 	bool                           gfx_off_controled_by_driver;
+	bool                           water_marks_exist;
 	Watermarks_t                      water_marks_table;
 	struct smu10_clock_voltage_information   clock_vol_info;
 	DpmClocks_t                       clock_table;
@@ -310,11 +311,9 @@ int smu10_init_function_pointers(struct pp_hwmgr *hwmgr);
 #define SMU10_UMD_PSTATE_FCLK                   933
 #define SMU10_UMD_PSTATE_VCE                    0x03C00320
 
-#define SMU10_UMD_PSTATE_PEAK_GFXCLK            1100
 #define SMU10_UMD_PSTATE_PEAK_SOCCLK            757
 #define SMU10_UMD_PSTATE_PEAK_FCLK              1200
 
-#define SMU10_UMD_PSTATE_MIN_GFXCLK             200
 #define SMU10_UMD_PSTATE_MIN_FCLK               400
 #define SMU10_UMD_PSTATE_MIN_SOCCLK             200
 #define SMU10_UMD_PSTATE_MIN_VCE                0x0190012C
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.c
index f4cbaee4e2ca..683b29a99366 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.c
@@ -39,13 +39,6 @@ static int smu7_enable_disable_vce_dpm(struct pp_hwmgr *hwmgr, bool enable)
 			PPSMC_MSG_VCEDPM_Disable);
 }
 
-static int smu7_enable_disable_samu_dpm(struct pp_hwmgr *hwmgr, bool enable)
-{
-	return smum_send_msg_to_smc(hwmgr, enable ?
-			PPSMC_MSG_SAMUDPM_Enable :
-			PPSMC_MSG_SAMUDPM_Disable);
-}
-
 static int smu7_update_uvd_dpm(struct pp_hwmgr *hwmgr, bool bgate)
 {
 	if (!bgate)
@@ -60,13 +53,6 @@ static int smu7_update_vce_dpm(struct pp_hwmgr *hwmgr, bool bgate)
 	return smu7_enable_disable_vce_dpm(hwmgr, !bgate);
 }
 
-static int smu7_update_samu_dpm(struct pp_hwmgr *hwmgr, bool bgate)
-{
-	if (!bgate)
-		smum_update_smc_table(hwmgr, SMU_SAMU_TABLE);
-	return smu7_enable_disable_samu_dpm(hwmgr, !bgate);
-}
-
 int smu7_powerdown_uvd(struct pp_hwmgr *hwmgr)
 {
 	if (phm_cf_want_uvd_power_gating(hwmgr))
@@ -107,35 +93,15 @@ static int smu7_powerup_vce(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
-static int smu7_powerdown_samu(struct pp_hwmgr *hwmgr)
-{
-	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
-			PHM_PlatformCaps_SamuPowerGating))
-		return smum_send_msg_to_smc(hwmgr,
-				PPSMC_MSG_SAMPowerOFF);
-	return 0;
-}
-
-static int smu7_powerup_samu(struct pp_hwmgr *hwmgr)
-{
-	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
-			PHM_PlatformCaps_SamuPowerGating))
-		return smum_send_msg_to_smc(hwmgr,
-				PPSMC_MSG_SAMPowerON);
-	return 0;
-}
-
 int smu7_disable_clock_power_gating(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 
 	data->uvd_power_gated = false;
 	data->vce_power_gated = false;
-	data->samu_power_gated = false;
 
 	smu7_powerup_uvd(hwmgr);
 	smu7_powerup_vce(hwmgr);
-	smu7_powerup_samu(hwmgr);
 
 	return 0;
 }
@@ -147,20 +113,20 @@ void smu7_powergate_uvd(struct pp_hwmgr *hwmgr, bool bgate)
 	data->uvd_power_gated = bgate;
 
 	if (bgate) {
-		cgs_set_powergating_state(hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_UVD,
 						AMD_PG_STATE_GATE);
-		cgs_set_clockgating_state(hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 				AMD_IP_BLOCK_TYPE_UVD,
 				AMD_CG_STATE_GATE);
 		smu7_update_uvd_dpm(hwmgr, true);
 		smu7_powerdown_uvd(hwmgr);
 	} else {
 		smu7_powerup_uvd(hwmgr);
-		cgs_set_clockgating_state(hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 				AMD_IP_BLOCK_TYPE_UVD,
 				AMD_CG_STATE_UNGATE);
-		cgs_set_powergating_state(hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_UVD,
 						AMD_PG_STATE_UNGATE);
 		smu7_update_uvd_dpm(hwmgr, false);
@@ -175,46 +141,26 @@ void smu7_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate)
 	data->vce_power_gated = bgate;
 
 	if (bgate) {
-		cgs_set_powergating_state(hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_VCE,
 						AMD_PG_STATE_GATE);
-		cgs_set_clockgating_state(hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 				AMD_IP_BLOCK_TYPE_VCE,
 				AMD_CG_STATE_GATE);
 		smu7_update_vce_dpm(hwmgr, true);
 		smu7_powerdown_vce(hwmgr);
 	} else {
 		smu7_powerup_vce(hwmgr);
-		cgs_set_clockgating_state(hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 				AMD_IP_BLOCK_TYPE_VCE,
 				AMD_CG_STATE_UNGATE);
-		cgs_set_powergating_state(hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_VCE,
 						AMD_PG_STATE_UNGATE);
 		smu7_update_vce_dpm(hwmgr, false);
 	}
 }
 
-int smu7_powergate_samu(struct pp_hwmgr *hwmgr, bool bgate)
-{
-	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
-
-	if (data->samu_power_gated == bgate)
-		return 0;
-
-	data->samu_power_gated = bgate;
-
-	if (bgate) {
-		smu7_update_samu_dpm(hwmgr, true);
-		smu7_powerdown_samu(hwmgr);
-	} else {
-		smu7_powerup_samu(hwmgr);
-		smu7_update_samu_dpm(hwmgr, false);
-	}
-
-	return 0;
-}
-
 int smu7_update_clock_gatings(struct pp_hwmgr *hwmgr,
 					const uint32_t *msg_id)
 {
@@ -470,7 +416,7 @@ int smu7_update_clock_gatings(struct pp_hwmgr *hwmgr,
  * Powerplay will only control the static per CU Power Gating.
  * Dynamic per CU Power Gating will be done in gfx.
  */
-int smu7_enable_per_cu_power_gating(struct pp_hwmgr *hwmgr, bool enable)
+int smu7_powergate_gfx(struct pp_hwmgr *hwmgr, bool enable)
 {
 	struct amdgpu_device *adev = hwmgr->adev;
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.h
index 1ddce023218a..fc8f8a6acc72 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_clockpowergating.h
@@ -29,11 +29,10 @@
 void smu7_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate);
 void smu7_powergate_uvd(struct pp_hwmgr *hwmgr, bool bgate);
 int smu7_powerdown_uvd(struct pp_hwmgr *hwmgr);
-int smu7_powergate_samu(struct pp_hwmgr *hwmgr, bool bgate);
 int smu7_powergate_acp(struct pp_hwmgr *hwmgr, bool bgate);
 int smu7_disable_clock_power_gating(struct pp_hwmgr *hwmgr);
 int smu7_update_clock_gatings(struct pp_hwmgr *hwmgr,
 					const uint32_t *msg_id);
-int smu7_enable_per_cu_power_gating(struct pp_hwmgr *hwmgr, bool enable);
+int smu7_powergate_gfx(struct pp_hwmgr *hwmgr, bool enable);
 
 #endif
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
index 26fbeafc3c96..052e60dfaf9f 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c
@@ -48,6 +48,8 @@
 #include "processpptables.h"
 #include "pp_thermal.h"
 
+#include "ivsrcid/ivsrcid_vislands30.h"
+
 #define MC_CG_ARB_FREQ_F0           0x0a
 #define MC_CG_ARB_FREQ_F1           0x0b
 #define MC_CG_ARB_FREQ_F2           0x0c
@@ -61,10 +63,6 @@
 #define SMC_CG_IND_START            0xc0030000
 #define SMC_CG_IND_END              0xc0040000
 
-#define VOLTAGE_SCALE               4
-#define VOLTAGE_VID_OFFSET_SCALE1   625
-#define VOLTAGE_VID_OFFSET_SCALE2   100
-
 #define MEM_FREQ_LOW_LATENCY        25000
 #define MEM_FREQ_HIGH_LATENCY       80000
 
@@ -79,14 +77,23 @@
 #define PCIE_BUS_CLK                10000
 #define TCLK                        (PCIE_BUS_CLK / 10)
 
-static const struct profile_mode_setting smu7_profiling[5] =
+static const struct profile_mode_setting smu7_profiling[6] =
 					{{1, 0, 100, 30, 1, 0, 100, 10},
 					 {1, 10, 0, 30, 0, 0, 0, 0},
 					 {0, 0, 0, 0, 1, 10, 16, 31},
 					 {1, 0, 11, 50, 1, 0, 100, 10},
 					 {1, 0, 5, 30, 0, 0, 0, 0},
+					 {0, 0, 0, 0, 0, 0, 0, 0},
 					};
 
+#define PPSMC_MSG_SetVBITimeout_VEGAM    ((uint16_t) 0x310)
+
+#define ixPWR_SVI2_PLANE1_LOAD                     0xC0200280
+#define PWR_SVI2_PLANE1_LOAD__PSI1_MASK                    0x00000020L
+#define PWR_SVI2_PLANE1_LOAD__PSI0_EN_MASK                 0x00000040L
+#define PWR_SVI2_PLANE1_LOAD__PSI1__SHIFT                  0x00000005
+#define PWR_SVI2_PLANE1_LOAD__PSI0_EN__SHIFT               0x00000006
+
 /** Values for the CG_THERMAL_CTRL::DPM_EVENT_SRC field. */
 enum DPM_EVENT_SRC {
 	DPM_EVENT_SRC_ANALOG = 0,
@@ -168,6 +175,13 @@ static int smu7_get_current_pcie_lane_number(struct pp_hwmgr *hwmgr)
 */
 static int smu7_enable_smc_voltage_controller(struct pp_hwmgr *hwmgr)
 {
+	if (hwmgr->chip_id == CHIP_VEGAM) {
+		PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device,
+				CGS_IND_REG__SMC, PWR_SVI2_PLANE1_LOAD, PSI1, 0);
+		PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device,
+				CGS_IND_REG__SMC, PWR_SVI2_PLANE1_LOAD, PSI0_EN, 0);
+	}
+
 	if (hwmgr->feature_mask & PP_SMC_VOLTAGE_CONTROL_MASK)
 		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_Voltage_Cntl_Enable);
 
@@ -779,7 +793,8 @@ static int smu7_setup_dpm_tables_v1(struct pp_hwmgr *hwmgr)
 			data->dpm_table.sclk_table.count++;
 		}
 	}
-
+	if (hwmgr->platform_descriptor.overdriveLimit.engineClock == 0)
+		hwmgr->platform_descriptor.overdriveLimit.engineClock = dep_sclk_table->entries[i-1].clk;
 	/* Initialize Mclk DPM table based on allow Mclk values */
 	data->dpm_table.mclk_table.count = 0;
 	for (i = 0; i < dep_mclk_table->count; i++) {
@@ -794,32 +809,8 @@ static int smu7_setup_dpm_tables_v1(struct pp_hwmgr *hwmgr)
 		}
 	}
 
-	return 0;
-}
-
-static int smu7_get_voltage_dependency_table(
-			const struct phm_ppt_v1_clock_voltage_dependency_table *allowed_dep_table,
-			struct phm_ppt_v1_clock_voltage_dependency_table *dep_table)
-{
-	uint8_t i = 0;
-	PP_ASSERT_WITH_CODE((0 != allowed_dep_table->count),
-				"Voltage Lookup Table empty",
-				return -EINVAL);
-
-	dep_table->count = allowed_dep_table->count;
-	for (i=0; i<dep_table->count; i++) {
-		dep_table->entries[i].clk = allowed_dep_table->entries[i].clk;
-		dep_table->entries[i].vddInd = allowed_dep_table->entries[i].vddInd;
-		dep_table->entries[i].vdd_offset = allowed_dep_table->entries[i].vdd_offset;
-		dep_table->entries[i].vddc = allowed_dep_table->entries[i].vddc;
-		dep_table->entries[i].vddgfx = allowed_dep_table->entries[i].vddgfx;
-		dep_table->entries[i].vddci = allowed_dep_table->entries[i].vddci;
-		dep_table->entries[i].mvdd = allowed_dep_table->entries[i].mvdd;
-		dep_table->entries[i].phases = allowed_dep_table->entries[i].phases;
-		dep_table->entries[i].cks_enable = allowed_dep_table->entries[i].cks_enable;
-		dep_table->entries[i].cks_voffset = allowed_dep_table->entries[i].cks_voffset;
-	}
-
+	if (hwmgr->platform_descriptor.overdriveLimit.memoryClock == 0)
+		hwmgr->platform_descriptor.overdriveLimit.memoryClock = dep_mclk_table->entries[i-1].clk;
 	return 0;
 }
 
@@ -850,7 +841,7 @@ static int smu7_odn_initial_default_setting(struct pp_hwmgr *hwmgr)
 		entries[i].vddc = dep_sclk_table->entries[i].vddc;
 	}
 
-	smu7_get_voltage_dependency_table(dep_sclk_table,
+	smu_get_voltage_dependency_table_ppt_v1(dep_sclk_table,
 		(struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dependency_on_sclk));
 
 	odn_table->odn_memory_clock_dpm_levels.num_of_pl =
@@ -862,12 +853,94 @@ static int smu7_odn_initial_default_setting(struct pp_hwmgr *hwmgr)
 		entries[i].vddc = dep_mclk_table->entries[i].vddc;
 	}
 
-	smu7_get_voltage_dependency_table(dep_mclk_table,
+	smu_get_voltage_dependency_table_ppt_v1(dep_mclk_table,
 		(struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dependency_on_mclk));
 
 	return 0;
 }
 
+static void smu7_setup_voltage_range_from_vbios(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_sclk_table;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	uint32_t min_vddc = 0;
+	uint32_t max_vddc = 0;
+
+	if (!table_info)
+		return;
+
+	dep_sclk_table = table_info->vdd_dep_on_sclk;
+
+	atomctrl_get_voltage_range(hwmgr, &max_vddc, &min_vddc);
+
+	if (min_vddc == 0 || min_vddc > 2000
+		|| min_vddc > dep_sclk_table->entries[0].vddc)
+		min_vddc = dep_sclk_table->entries[0].vddc;
+
+	if (max_vddc == 0 || max_vddc > 2000
+		|| max_vddc < dep_sclk_table->entries[dep_sclk_table->count-1].vddc)
+		max_vddc = dep_sclk_table->entries[dep_sclk_table->count-1].vddc;
+
+	data->odn_dpm_table.min_vddc = min_vddc;
+	data->odn_dpm_table.max_vddc = max_vddc;
+}
+
+static void smu7_check_dpm_table_updated(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct smu7_odn_dpm_table *odn_table = &(data->odn_dpm_table);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	uint32_t i;
+
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_table;
+	struct phm_ppt_v1_clock_voltage_dependency_table *odn_dep_table;
+
+	if (table_info == NULL)
+		return;
+
+	for (i = 0; i < data->dpm_table.sclk_table.count; i++) {
+		if (odn_table->odn_core_clock_dpm_levels.entries[i].clock !=
+					data->dpm_table.sclk_table.dpm_levels[i].value) {
+			data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_SCLK;
+			break;
+		}
+	}
+
+	for (i = 0; i < data->dpm_table.mclk_table.count; i++) {
+		if (odn_table->odn_memory_clock_dpm_levels.entries[i].clock !=
+					data->dpm_table.mclk_table.dpm_levels[i].value) {
+			data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_MCLK;
+			break;
+		}
+	}
+
+	dep_table = table_info->vdd_dep_on_mclk;
+	odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dependency_on_mclk);
+
+	for (i = 0; i < dep_table->count; i++) {
+		if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) {
+			data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_MCLK;
+			return;
+		}
+	}
+
+	dep_table = table_info->vdd_dep_on_sclk;
+	odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dependency_on_sclk);
+	for (i = 0; i < dep_table->count; i++) {
+		if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) {
+			data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_SCLK;
+			return;
+		}
+	}
+	if (data->need_update_smu7_dpm_table & DPMTABLE_OD_UPDATE_VDDC) {
+		data->need_update_smu7_dpm_table &= ~DPMTABLE_OD_UPDATE_VDDC;
+		data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_SCLK | DPMTABLE_OD_UPDATE_MCLK;
+	}
+}
+
 static int smu7_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
@@ -886,9 +959,14 @@ static int smu7_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 			sizeof(struct smu7_dpm_table));
 
 	/* initialize ODN table */
-	if (hwmgr->od_enabled)
-		smu7_odn_initial_default_setting(hwmgr);
-
+	if (hwmgr->od_enabled) {
+		if (data->odn_dpm_table.max_vddc) {
+			smu7_check_dpm_table_updated(hwmgr);
+		} else {
+			smu7_setup_voltage_range_from_vbios(hwmgr);
+			smu7_odn_initial_default_setting(hwmgr);
+		}
+	}
 	return 0;
 }
 
@@ -965,6 +1043,22 @@ static int smu7_disable_deep_sleep_master_switch(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
+static int smu7_disable_sclk_vce_handshake(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	uint32_t soft_register_value = 0;
+	uint32_t handshake_disables_offset = data->soft_regs_start
+				+ smum_get_offsetof(hwmgr,
+					SMU_SoftRegisters, HandshakeDisables);
+
+	soft_register_value = cgs_read_ind_register(hwmgr->device,
+				CGS_IND_REG__SMC, handshake_disables_offset);
+	soft_register_value |= SMU7_VCE_SCLK_HANDSHAKE_DISABLE;
+	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+			handshake_disables_offset, soft_register_value);
+	return 0;
+}
+
 static int smu7_disable_handshake_uvd(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
@@ -987,23 +1081,29 @@ static int smu7_enable_sclk_mclk_dpm(struct pp_hwmgr *hwmgr)
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 
 	/* enable SCLK dpm */
-	if (!data->sclk_dpm_key_disabled)
+	if (!data->sclk_dpm_key_disabled) {
+		if (hwmgr->chip_id == CHIP_VEGAM)
+			smu7_disable_sclk_vce_handshake(hwmgr);
+
 		PP_ASSERT_WITH_CODE(
 		(0 == smum_send_msg_to_smc(hwmgr, PPSMC_MSG_DPM_Enable)),
 		"Failed to enable SCLK DPM during DPM Start Function!",
 		return -EINVAL);
+	}
 
 	/* enable MCLK dpm */
 	if (0 == data->mclk_dpm_key_disabled) {
 		if (!(hwmgr->feature_mask & PP_UVD_HANDSHAKE_MASK))
 			smu7_disable_handshake_uvd(hwmgr);
+
 		PP_ASSERT_WITH_CODE(
 				(0 == smum_send_msg_to_smc(hwmgr,
 						PPSMC_MSG_MCLKDPM_Enable)),
 				"Failed to enable MCLK DPM during DPM Start Function!",
 				return -EINVAL);
 
-		PHM_WRITE_FIELD(hwmgr->device, MC_SEQ_CNTL_3, CAC_EN, 0x1);
+		if (hwmgr->chip_family != CHIP_VEGAM)
+			PHM_WRITE_FIELD(hwmgr->device, MC_SEQ_CNTL_3, CAC_EN, 0x1);
 
 
 		if (hwmgr->chip_family == AMDGPU_FAMILY_CI) {
@@ -1019,8 +1119,13 @@ static int smu7_enable_sclk_mclk_dpm(struct pp_hwmgr *hwmgr)
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC1_CNTL, 0x5);
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_CPL_CNTL, 0x100005);
 			udelay(10);
-			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC0_CNTL, 0x400005);
-			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC1_CNTL, 0x400005);
+			if (hwmgr->chip_id == CHIP_VEGAM) {
+				cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC0_CNTL, 0x400009);
+				cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC1_CNTL, 0x400009);
+			} else {
+				cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC0_CNTL, 0x400005);
+				cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_MC1_CNTL, 0x400005);
+			}
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixLCAC_CPL_CNTL, 0x500005);
 		}
 	}
@@ -1229,7 +1334,7 @@ static int smu7_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
 
 		tmp_result = smu7_construct_voltage_tables(hwmgr);
 		PP_ASSERT_WITH_CODE((0 == tmp_result),
-				"Failed to contruct voltage tables!",
+				"Failed to construct voltage tables!",
 				result = tmp_result);
 	}
 	smum_initialize_mc_reg_table(hwmgr);
@@ -1261,10 +1366,12 @@ static int smu7_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
 	PP_ASSERT_WITH_CODE((0 == tmp_result),
 			"Failed to process firmware header!", result = tmp_result);
 
-	tmp_result = smu7_initial_switch_from_arbf0_to_f1(hwmgr);
-	PP_ASSERT_WITH_CODE((0 == tmp_result),
-			"Failed to initialize switch from ArbF0 to F1!",
-			result = tmp_result);
+	if (hwmgr->chip_id != CHIP_VEGAM) {
+		tmp_result = smu7_initial_switch_from_arbf0_to_f1(hwmgr);
+		PP_ASSERT_WITH_CODE((0 == tmp_result),
+				"Failed to initialize switch from ArbF0 to F1!",
+				result = tmp_result);
+	}
 
 	result = smu7_setup_default_dpm_tables(hwmgr);
 	PP_ASSERT_WITH_CODE(0 == result,
@@ -1473,7 +1580,7 @@ static void smu7_init_dpm_defaults(struct pp_hwmgr *hwmgr)
 	data->current_profile_setting.sclk_up_hyst = 0;
 	data->current_profile_setting.sclk_down_hyst = 100;
 	data->current_profile_setting.sclk_activity = SMU7_SCLK_TARGETACTIVITY_DFLT;
-	data->current_profile_setting.bupdate_sclk = 1;
+	data->current_profile_setting.bupdate_mclk = 1;
 	data->current_profile_setting.mclk_up_hyst = 0;
 	data->current_profile_setting.mclk_down_hyst = 100;
 	data->current_profile_setting.mclk_activity = SMU7_MCLK_TARGETACTIVITY_DFLT;
@@ -2754,6 +2861,9 @@ static int smu7_vblank_too_short(struct pp_hwmgr *hwmgr,
 	case CHIP_POLARIS12:
 		switch_limit_us = data->is_memory_gddr5 ? 190 : 150;
 		break;
+	case CHIP_VEGAM:
+		switch_limit_us = 30;
+		break;
 	default:
 		switch_limit_us = data->is_memory_gddr5 ? 450 : 150;
 		break;
@@ -2769,7 +2879,7 @@ static int smu7_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 				struct pp_power_state *request_ps,
 			const struct pp_power_state *current_ps)
 {
-
+	struct amdgpu_device *adev = hwmgr->adev;
 	struct smu7_power_state *smu7_ps =
 				cast_phw_smu7_power_state(&request_ps->hardware);
 	uint32_t sclk;
@@ -2777,8 +2887,6 @@ static int smu7_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 	struct PP_Clocks minimum_clocks = {0};
 	bool disable_mclk_switching;
 	bool disable_mclk_switching_for_frame_lock;
-	struct cgs_display_info info = {0};
-	struct cgs_mode_info mode_info = {0};
 	const struct phm_clock_and_voltage_limits *max_limits;
 	uint32_t i;
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
@@ -2787,7 +2895,6 @@ static int smu7_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 	int32_t count;
 	int32_t stable_pstate_sclk = 0, stable_pstate_mclk = 0;
 
-	info.mode_info = &mode_info;
 	data->battery_state = (PP_StateUILabel_Battery ==
 			request_ps->classification.ui_label);
 
@@ -2795,12 +2902,12 @@ static int smu7_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 				 "VI should always have 2 performance levels",
 				);
 
-	max_limits = (PP_PowerSource_AC == hwmgr->power_source) ?
+	max_limits = adev->pm.ac_power ?
 			&(hwmgr->dyn_state.max_clock_voltage_on_ac) :
 			&(hwmgr->dyn_state.max_clock_voltage_on_dc);
 
 	/* Cap clock DPM tables at DC MAX if it is in DC. */
-	if (PP_PowerSource_DC == hwmgr->power_source) {
+	if (!adev->pm.ac_power) {
 		for (i = 0; i < smu7_ps->performance_level_count; i++) {
 			if (smu7_ps->performance_levels[i].memory_clock > max_limits->mclk)
 				smu7_ps->performance_levels[i].memory_clock = max_limits->mclk;
@@ -2809,10 +2916,8 @@ static int smu7_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 		}
 	}
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-
-	minimum_clocks.engineClock = hwmgr->display_config.min_core_set_clock;
-	minimum_clocks.memoryClock = hwmgr->display_config.min_mem_set_clock;
+	minimum_clocks.engineClock = hwmgr->display_config->min_core_set_clock;
+	minimum_clocks.memoryClock = hwmgr->display_config->min_mem_set_clock;
 
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
 			PHM_PlatformCaps_StablePState)) {
@@ -2843,12 +2948,12 @@ static int smu7_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 				    PHM_PlatformCaps_DisableMclkSwitchingForFrameLock);
 
 
-	if (info.display_count == 0)
+	if (hwmgr->display_config->num_display == 0)
 		disable_mclk_switching = false;
 	else
-		disable_mclk_switching = ((1 < info.display_count) ||
+		disable_mclk_switching = ((1 < hwmgr->display_config->num_display) ||
 					  disable_mclk_switching_for_frame_lock ||
-					  smu7_vblank_too_short(hwmgr, mode_info.vblank_time_us));
+					  smu7_vblank_too_short(hwmgr, hwmgr->display_config->min_vblank_time));
 
 	sclk = smu7_ps->performance_levels[0].engine_clock;
 	mclk = smu7_ps->performance_levels[0].memory_clock;
@@ -2957,8 +3062,7 @@ static int smu7_dpm_patch_boot_state(struct pp_hwmgr *hwmgr,
 	/* First retrieve the Boot clocks and VDDC from the firmware info table.
 	 * We assume here that fw_info is unchanged if this call fails.
 	 */
-	fw_info = (ATOM_FIRMWARE_INFO_V2_2 *)cgs_atom_get_data_table(
-			hwmgr->device, index,
+	fw_info = (ATOM_FIRMWARE_INFO_V2_2 *)smu_atom_get_data_table(hwmgr->adev, index,
 			&size, &frev, &crev);
 	if (!fw_info)
 		/* During a test, there is no firmware info table. */
@@ -3081,7 +3185,7 @@ static int smu7_get_pp_table_entry_callback_func_v1(struct pp_hwmgr *hwmgr,
 	performance_level->pcie_gen = get_pcie_gen_support(data->pcie_gen_cap,
 			state_entry->ucPCIEGenLow);
 	performance_level->pcie_lane = get_pcie_lane_support(data->pcie_lane_cap,
-			state_entry->ucPCIELaneHigh);
+			state_entry->ucPCIELaneLow);
 
 	performance_level = &(smu7_power_state->performance_levels
 			[smu7_power_state->performance_level_count++]);
@@ -3366,34 +3470,35 @@ static int smu7_get_pp_table_entry(struct pp_hwmgr *hwmgr,
 	return 0;
 }
 
-static int smu7_get_gpu_power(struct pp_hwmgr *hwmgr,
-		struct pp_gpu_power *query)
+static int smu7_get_gpu_power(struct pp_hwmgr *hwmgr, u32 *query)
 {
-	PP_ASSERT_WITH_CODE(!smum_send_msg_to_smc(hwmgr,
-			PPSMC_MSG_PmStatusLogStart),
-			"Failed to start pm status log!",
-			return -1);
+	int i;
+	u32 tmp = 0;
+
+	if (!query)
+		return -EINVAL;
 
-	/* Sampling period from 50ms to 4sec */
-	msleep_interruptible(200);
+	smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetCurrPkgPwr, 0);
+	tmp = cgs_read_register(hwmgr->device, mmSMC_MSG_ARG_0);
+	*query = tmp;
 
-	PP_ASSERT_WITH_CODE(!smum_send_msg_to_smc(hwmgr,
-			PPSMC_MSG_PmStatusLogSample),
-			"Failed to sample pm status log!",
-			return -1);
+	if (tmp != 0)
+		return 0;
 
-	query->vddc_power = cgs_read_ind_register(hwmgr->device,
-			CGS_IND_REG__SMC,
-			ixSMU_PM_STATUS_40);
-	query->vddci_power = cgs_read_ind_register(hwmgr->device,
-			CGS_IND_REG__SMC,
-			ixSMU_PM_STATUS_49);
-	query->max_gpu_power = cgs_read_ind_register(hwmgr->device,
-			CGS_IND_REG__SMC,
-			ixSMU_PM_STATUS_94);
-	query->average_gpu_power = cgs_read_ind_register(hwmgr->device,
-			CGS_IND_REG__SMC,
-			ixSMU_PM_STATUS_95);
+	smum_send_msg_to_smc(hwmgr, PPSMC_MSG_PmStatusLogStart);
+	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+							ixSMU_PM_STATUS_94, 0);
+
+	for (i = 0; i < 10; i++) {
+		mdelay(1);
+		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_PmStatusLogSample);
+		tmp = cgs_read_ind_register(hwmgr->device,
+						CGS_IND_REG__SMC,
+						ixSMU_PM_STATUS_94);
+		if (tmp != 0)
+			break;
+	}
+	*query = tmp;
 
 	return 0;
 }
@@ -3446,10 +3551,7 @@ static int smu7_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 		*size = 4;
 		return 0;
 	case AMDGPU_PP_SENSOR_GPU_POWER:
-		if (*size < sizeof(struct pp_gpu_power))
-			return -EINVAL;
-		*size = sizeof(struct pp_gpu_power);
-		return smu7_get_gpu_power(hwmgr, (struct pp_gpu_power *)value);
+		return smu7_get_gpu_power(hwmgr, (uint32_t *)value);
 	case AMDGPU_PP_SENSOR_VDDGFX:
 		if ((data->vr_config & 0xff) == 0x2)
 			val_vid = PHM_READ_INDIRECT_FIELD(hwmgr->device,
@@ -3480,7 +3582,6 @@ static int smu7_find_dpm_states_clocks_in_dpm_table(struct pp_hwmgr *hwmgr, cons
 			[smu7_ps->performance_level_count - 1].memory_clock;
 	struct PP_Clocks min_clocks = {0};
 	uint32_t i;
-	struct cgs_display_info info = {0};
 
 	for (i = 0; i < sclk_table->count; i++) {
 		if (sclk == sclk_table->dpm_levels[i].value)
@@ -3507,9 +3608,8 @@ static int smu7_find_dpm_states_clocks_in_dpm_table(struct pp_hwmgr *hwmgr, cons
 	if (i >= mclk_table->count)
 		data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_MCLK;
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
 
-	if (data->display_timing.num_existing_displays != info.display_count)
+	if (data->display_timing.num_existing_displays != hwmgr->display_config->num_display)
 		data->need_update_smu7_dpm_table |= DPMTABLE_UPDATE_MCLK;
 
 	return 0;
@@ -3676,8 +3776,9 @@ static int smu7_trim_single_dpm_states(struct pp_hwmgr *hwmgr,
 	uint32_t i;
 
 	for (i = 0; i < dpm_table->count; i++) {
-		if ((dpm_table->dpm_levels[i].value < low_limit)
-		|| (dpm_table->dpm_levels[i].value > high_limit))
+	/*skip the trim if od is enabled*/
+		if (!hwmgr->od_enabled && (dpm_table->dpm_levels[i].value < low_limit
+			|| dpm_table->dpm_levels[i].value > high_limit))
 			dpm_table->dpm_levels[i].enabled = false;
 		else
 			dpm_table->dpm_levels[i].enabled = true;
@@ -3714,13 +3815,14 @@ static int smu7_trim_dpm_states(struct pp_hwmgr *hwmgr,
 static int smu7_generate_dpm_level_enable_mask(
 		struct pp_hwmgr *hwmgr, const void *input)
 {
-	int result;
+	int result = 0;
 	const struct phm_set_power_state_input *states =
 			(const struct phm_set_power_state_input *)input;
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 	const struct smu7_power_state *smu7_ps =
 			cast_const_phw_smu7_power_state(states->pnew_state);
 
+
 	result = smu7_trim_dpm_states(hwmgr, smu7_ps);
 	if (result)
 		return result;
@@ -3812,9 +3914,14 @@ static int smu7_notify_smc_display(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 
-	if (hwmgr->feature_mask & PP_VBI_TIME_SUPPORT_MASK)
-		smum_send_msg_to_smc_with_parameter(hwmgr,
-			(PPSMC_Msg)PPSMC_MSG_SetVBITimeout, data->frame_time_x2);
+	if (hwmgr->feature_mask & PP_VBI_TIME_SUPPORT_MASK) {
+		if (hwmgr->chip_id == CHIP_VEGAM)
+			smum_send_msg_to_smc_with_parameter(hwmgr,
+					(PPSMC_Msg)PPSMC_MSG_SetVBITimeout_VEGAM, data->frame_time_x2);
+		else
+			smum_send_msg_to_smc_with_parameter(hwmgr,
+					(PPSMC_Msg)PPSMC_MSG_SetVBITimeout, data->frame_time_x2);
+	}
 	return (smum_send_msg_to_smc(hwmgr, (PPSMC_Msg)PPSMC_HasDisplay) == 0) ?  0 : -EINVAL;
 }
 
@@ -3908,15 +4015,8 @@ smu7_notify_smc_display_change(struct pp_hwmgr *hwmgr, bool has_display)
 static int
 smu7_notify_smc_display_config_after_ps_adjustment(struct pp_hwmgr *hwmgr)
 {
-	uint32_t num_active_displays = 0;
-	struct cgs_display_info info = {0};
-
-	info.mode_info = NULL;
-	cgs_get_active_displays_info(hwmgr->device, &info);
-
-	num_active_displays = info.display_count;
-
-	if (num_active_displays > 1 && hwmgr->display_config.multi_monitor_in_sync != true)
+	if (hwmgr->display_config->num_display > 1 &&
+			!hwmgr->display_config->multi_monitor_in_sync)
 		smu7_notify_smc_display_change(hwmgr, false);
 
 	return 0;
@@ -3931,33 +4031,24 @@ smu7_notify_smc_display_config_after_ps_adjustment(struct pp_hwmgr *hwmgr)
 static int smu7_program_display_gap(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
-	uint32_t num_active_displays = 0;
 	uint32_t display_gap = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_DISPLAY_GAP_CNTL);
 	uint32_t display_gap2;
 	uint32_t pre_vbi_time_in_us;
 	uint32_t frame_time_in_us;
-	uint32_t ref_clock;
-	uint32_t refresh_rate = 0;
-	struct cgs_display_info info = {0};
-	struct cgs_mode_info mode_info = {0};
+	uint32_t ref_clock, refresh_rate;
 
-	info.mode_info = &mode_info;
-	cgs_get_active_displays_info(hwmgr->device, &info);
-	num_active_displays = info.display_count;
-
-	display_gap = PHM_SET_FIELD(display_gap, CG_DISPLAY_GAP_CNTL, DISP_GAP, (num_active_displays > 0) ? DISPLAY_GAP_VBLANK_OR_WM : DISPLAY_GAP_IGNORE);
+	display_gap = PHM_SET_FIELD(display_gap, CG_DISPLAY_GAP_CNTL, DISP_GAP, (hwmgr->display_config->num_display > 0) ? DISPLAY_GAP_VBLANK_OR_WM : DISPLAY_GAP_IGNORE);
 	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixCG_DISPLAY_GAP_CNTL, display_gap);
 
 	ref_clock =  amdgpu_asic_get_xclk((struct amdgpu_device *)hwmgr->adev);
-
-	refresh_rate = mode_info.refresh_rate;
+	refresh_rate = hwmgr->display_config->vrefresh;
 
 	if (0 == refresh_rate)
 		refresh_rate = 60;
 
 	frame_time_in_us = 1000000 / refresh_rate;
 
-	pre_vbi_time_in_us = frame_time_in_us - 200 - mode_info.vblank_time_us;
+	pre_vbi_time_in_us = frame_time_in_us - 200 - hwmgr->display_config->min_vblank_time;
 
 	data->frame_time_x2 = frame_time_in_us * 2 / 100;
 
@@ -4016,17 +4107,17 @@ static int smu7_register_irq_handlers(struct pp_hwmgr *hwmgr)
 
 	amdgpu_irq_add_id((struct amdgpu_device *)(hwmgr->adev),
 			AMDGPU_IH_CLIENTID_LEGACY,
-			230,
+			VISLANDS30_IV_SRCID_CG_TSS_THERMAL_LOW_TO_HIGH,
 			source);
 	amdgpu_irq_add_id((struct amdgpu_device *)(hwmgr->adev),
 			AMDGPU_IH_CLIENTID_LEGACY,
-			231,
+			VISLANDS30_IV_SRCID_CG_TSS_THERMAL_HIGH_TO_LOW,
 			source);
 
 	/* Register CTF(GPIO_19) interrupt */
 	amdgpu_irq_add_id((struct amdgpu_device *)(hwmgr->adev),
 			AMDGPU_IH_CLIENTID_LEGACY,
-			83,
+			VISLANDS30_IV_SRCID_GPIO_19,
 			source);
 
 	return 0;
@@ -4037,17 +4128,14 @@ smu7_check_smc_update_required_for_display_configuration(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 	bool is_update_required = false;
-	struct cgs_display_info info = {0, 0, NULL};
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-
-	if (data->display_timing.num_existing_displays != info.display_count)
+	if (data->display_timing.num_existing_displays != hwmgr->display_config->num_display)
 		is_update_required = true;
 
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SclkDeepSleep)) {
-		if (data->display_timing.min_clock_in_sr != hwmgr->display_config.min_core_set_clock_in_sr &&
+		if (data->display_timing.min_clock_in_sr != hwmgr->display_config->min_core_set_clock_in_sr &&
 			(data->display_timing.min_clock_in_sr >= SMU7_MINIMUM_ENGINE_CLOCK ||
-			hwmgr->display_config.min_core_set_clock_in_sr >= SMU7_MINIMUM_ENGINE_CLOCK))
+			hwmgr->display_config->min_core_set_clock_in_sr >= SMU7_MINIMUM_ENGINE_CLOCK))
 			is_update_required = true;
 	}
 	return is_update_required;
@@ -4102,7 +4190,7 @@ static int smu7_check_states_equal(struct pp_hwmgr *hwmgr,
 	return 0;
 }
 
-static int smu7_upload_mc_firmware(struct pp_hwmgr *hwmgr)
+static int smu7_check_mc_firmware(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 
@@ -4181,13 +4269,9 @@ static int smu7_read_clock_registers(struct pp_hwmgr *hwmgr)
 static int smu7_get_memory_type(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
-	uint32_t temp;
-
-	temp = cgs_read_register(hwmgr->device, mmMC_SEQ_MISC0);
+	struct amdgpu_device *adev = hwmgr->adev;
 
-	data->is_memory_gddr5 = (MC_SEQ_MISC0_GDDR5_VALUE ==
-			((temp & MC_SEQ_MISC0_GDDR5_MASK) >>
-			 MC_SEQ_MISC0_GDDR5_SHIFT));
+	data->is_memory_gddr5 = (adev->gmc.vram_type == AMDGPU_VRAM_TYPE_GDDR5);
 
 	return 0;
 }
@@ -4218,7 +4302,6 @@ static int smu7_init_power_gate_state(struct pp_hwmgr *hwmgr)
 
 	data->uvd_power_gated = false;
 	data->vce_power_gated = false;
-	data->samu_power_gated = false;
 
 	return 0;
 }
@@ -4235,7 +4318,7 @@ static int smu7_setup_asic_task(struct pp_hwmgr *hwmgr)
 {
 	int tmp_result, result = 0;
 
-	smu7_upload_mc_firmware(hwmgr);
+	smu7_check_mc_firmware(hwmgr);
 
 	tmp_result = smu7_read_clock_registers(hwmgr);
 	PP_ASSERT_WITH_CODE((0 == tmp_result),
@@ -4370,22 +4453,36 @@ static int smu7_print_clock_levels(struct pp_hwmgr *hwmgr,
 		break;
 	case OD_SCLK:
 		if (hwmgr->od_enabled) {
-			size = sprintf(buf, "%s: \n", "OD_SCLK");
+			size = sprintf(buf, "%s:\n", "OD_SCLK");
 			for (i = 0; i < odn_sclk_table->num_of_pl; i++)
-				size += sprintf(buf + size, "%d: %10uMhz %10u mV\n",
-					i, odn_sclk_table->entries[i].clock / 100,
+				size += sprintf(buf + size, "%d: %10uMHz %10umV\n",
+					i, odn_sclk_table->entries[i].clock/100,
 					odn_sclk_table->entries[i].vddc);
 		}
 		break;
 	case OD_MCLK:
 		if (hwmgr->od_enabled) {
-			size = sprintf(buf, "%s: \n", "OD_MCLK");
+			size = sprintf(buf, "%s:\n", "OD_MCLK");
 			for (i = 0; i < odn_mclk_table->num_of_pl; i++)
-				size += sprintf(buf + size, "%d: %10uMhz %10u mV\n",
-					i, odn_mclk_table->entries[i].clock / 100,
+				size += sprintf(buf + size, "%d: %10uMHz %10umV\n",
+					i, odn_mclk_table->entries[i].clock/100,
 					odn_mclk_table->entries[i].vddc);
 		}
 		break;
+	case OD_RANGE:
+		if (hwmgr->od_enabled) {
+			size = sprintf(buf, "%s:\n", "OD_RANGE");
+			size += sprintf(buf + size, "SCLK: %7uMHz %10uMHz\n",
+				data->golden_dpm_table.sclk_table.dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.engineClock/100);
+			size += sprintf(buf + size, "MCLK: %7uMHz %10uMHz\n",
+				data->golden_dpm_table.mclk_table.dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.memoryClock/100);
+			size += sprintf(buf + size, "VDDC: %7umV %11umV\n",
+				data->odn_dpm_table.min_vddc,
+				data->odn_dpm_table.max_vddc);
+		}
+		break;
 	default:
 		break;
 	}
@@ -4515,12 +4612,12 @@ static int smu7_get_sclks(struct pp_hwmgr *hwmgr, struct amd_pp_clocks *clocks)
 			return -EINVAL;
 		dep_sclk_table = table_info->vdd_dep_on_sclk;
 		for (i = 0; i < dep_sclk_table->count; i++)
-			clocks->clock[i] = dep_sclk_table->entries[i].clk;
+			clocks->clock[i] = dep_sclk_table->entries[i].clk * 10;
 		clocks->count = dep_sclk_table->count;
 	} else if (hwmgr->pp_table_version == PP_TABLE_V0) {
 		sclk_table = hwmgr->dyn_state.vddc_dependency_on_sclk;
 		for (i = 0; i < sclk_table->count; i++)
-			clocks->clock[i] = sclk_table->entries[i].clk;
+			clocks->clock[i] = sclk_table->entries[i].clk * 10;
 		clocks->count = sclk_table->count;
 	}
 
@@ -4552,7 +4649,7 @@ static int smu7_get_mclks(struct pp_hwmgr *hwmgr, struct amd_pp_clocks *clocks)
 			return -EINVAL;
 		dep_mclk_table = table_info->vdd_dep_on_mclk;
 		for (i = 0; i < dep_mclk_table->count; i++) {
-			clocks->clock[i] = dep_mclk_table->entries[i].clk;
+			clocks->clock[i] = dep_mclk_table->entries[i].clk * 10;
 			clocks->latency[i] = smu7_get_mem_latency(hwmgr,
 						dep_mclk_table->entries[i].clk);
 		}
@@ -4560,7 +4657,7 @@ static int smu7_get_mclks(struct pp_hwmgr *hwmgr, struct amd_pp_clocks *clocks)
 	} else if (hwmgr->pp_table_version == PP_TABLE_V0) {
 		mclk_table = hwmgr->dyn_state.vddc_dependency_on_mclk;
 		for (i = 0; i < mclk_table->count; i++)
-			clocks->clock[i] = mclk_table->entries[i].clk;
+			clocks->clock[i] = mclk_table->entries[i].clk * 10;
 		clocks->count = mclk_table->count;
 	}
 	return 0;
@@ -4669,36 +4766,27 @@ static bool smu7_check_clk_voltage_valid(struct pp_hwmgr *hwmgr,
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 
-	struct phm_ppt_v1_information *table_info =
-			(struct phm_ppt_v1_information *)(hwmgr->pptable);
-	uint32_t min_vddc;
-	struct phm_ppt_v1_clock_voltage_dependency_table *dep_sclk_table;
-
-	if (table_info == NULL)
-		return false;
-
-	dep_sclk_table = table_info->vdd_dep_on_sclk;
-	min_vddc = dep_sclk_table->entries[0].vddc;
-
-	if (voltage < min_vddc || voltage > 2000) {
-		pr_info("OD voltage is out of range [%d - 2000] mV\n", min_vddc);
+	if (voltage < data->odn_dpm_table.min_vddc || voltage > data->odn_dpm_table.max_vddc) {
+		pr_info("OD voltage is out of range [%d - %d] mV\n",
+						data->odn_dpm_table.min_vddc,
+						data->odn_dpm_table.max_vddc);
 		return false;
 	}
 
 	if (type == PP_OD_EDIT_SCLK_VDDC_TABLE) {
-		if (data->vbios_boot_state.sclk_bootup_value > clk ||
+		if (data->golden_dpm_table.sclk_table.dpm_levels[0].value > clk ||
 			hwmgr->platform_descriptor.overdriveLimit.engineClock < clk) {
 			pr_info("OD engine clock is out of range [%d - %d] MHz\n",
-				data->vbios_boot_state.sclk_bootup_value,
-				hwmgr->platform_descriptor.overdriveLimit.engineClock / 100);
+				data->golden_dpm_table.sclk_table.dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.engineClock/100);
 			return false;
 		}
 	} else if (type == PP_OD_EDIT_MCLK_VDDC_TABLE) {
-		if (data->vbios_boot_state.mclk_bootup_value > clk ||
+		if (data->golden_dpm_table.mclk_table.dpm_levels[0].value > clk ||
 			hwmgr->platform_descriptor.overdriveLimit.memoryClock < clk) {
 			pr_info("OD memory clock is out of range [%d - %d] MHz\n",
-				data->vbios_boot_state.mclk_bootup_value/100,
-				hwmgr->platform_descriptor.overdriveLimit.memoryClock / 100);
+				data->golden_dpm_table.mclk_table.dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.memoryClock/100);
 			return false;
 		}
 	} else {
@@ -4708,64 +4796,6 @@ static bool smu7_check_clk_voltage_valid(struct pp_hwmgr *hwmgr,
 	return true;
 }
 
-static void smu7_check_dpm_table_updated(struct pp_hwmgr *hwmgr)
-{
-	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
-	struct smu7_odn_dpm_table *odn_table = &(data->odn_dpm_table);
-	struct phm_ppt_v1_information *table_info =
-			(struct phm_ppt_v1_information *)(hwmgr->pptable);
-	uint32_t i;
-
-	struct phm_ppt_v1_clock_voltage_dependency_table *dep_table;
-	struct phm_ppt_v1_clock_voltage_dependency_table *odn_dep_table;
-
-	if (table_info == NULL)
-		return;
-
-	for (i=0; i<data->dpm_table.sclk_table.count; i++) {
-		if (odn_table->odn_core_clock_dpm_levels.entries[i].clock !=
-					data->dpm_table.sclk_table.dpm_levels[i].value) {
-			data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_SCLK;
-			break;
-		}
-	}
-
-	for (i=0; i<data->dpm_table.mclk_table.count; i++) {
-		if (odn_table->odn_memory_clock_dpm_levels.entries[i].clock !=
-					data->dpm_table.mclk_table.dpm_levels[i].value) {
-			data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_MCLK;
-			break;
-		}
-	}
-
-	dep_table = table_info->vdd_dep_on_mclk;
-	odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dependency_on_mclk);
-
-	for (i=0; i < dep_table->count; i++) {
-		if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) {
-			data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_MCLK;
-			return;
-		}
-	}
-	if (i == dep_table->count && data->need_update_smu7_dpm_table & DPMTABLE_OD_UPDATE_VDDC) {
-		data->need_update_smu7_dpm_table &= ~DPMTABLE_OD_UPDATE_VDDC;
-		data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_MCLK;
-	}
-
-	dep_table = table_info->vdd_dep_on_sclk;
-	odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dependency_on_sclk);
-	for (i=0; i < dep_table->count; i++) {
-		if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) {
-			data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_SCLK;
-			return;
-		}
-	}
-	if (i == dep_table->count && data->need_update_smu7_dpm_table & DPMTABLE_OD_UPDATE_VDDC) {
-		data->need_update_smu7_dpm_table &= ~DPMTABLE_OD_UPDATE_VDDC;
-		data->need_update_smu7_dpm_table |= DPMTABLE_OD_UPDATE_SCLK;
-	}
-}
-
 static int smu7_odn_edit_dpm_table(struct pp_hwmgr *hwmgr,
 					enum PP_OD_DPM_TABLE_COMMAND type,
 					long *input, uint32_t size)
@@ -4864,6 +4894,17 @@ static int smu7_get_power_profile_mode(struct pp_hwmgr *hwmgr, char *buf)
 	len = sizeof(smu7_profiling) / sizeof(struct profile_mode_setting);
 
 	for (i = 0; i < len; i++) {
+		if (i == hwmgr->power_profile_mode) {
+			size += sprintf(buf + size, "%3d %14s %s: %8d %16d %16d %16d %16d %16d\n",
+			i, profile_name[i], "*",
+			data->current_profile_setting.sclk_up_hyst,
+			data->current_profile_setting.sclk_down_hyst,
+			data->current_profile_setting.sclk_activity,
+			data->current_profile_setting.mclk_up_hyst,
+			data->current_profile_setting.mclk_down_hyst,
+			data->current_profile_setting.mclk_activity);
+			continue;
+		}
 		if (smu7_profiling[i].bupdate_sclk)
 			size += sprintf(buf + size, "%3d %16s: %8d %16d %16d ",
 			i, profile_name[i], smu7_profiling[i].sclk_up_hyst,
@@ -4883,24 +4924,6 @@ static int smu7_get_power_profile_mode(struct pp_hwmgr *hwmgr, char *buf)
 			"-", "-", "-");
 	}
 
-	size += sprintf(buf + size, "%3d %16s: %8d %16d %16d %16d %16d %16d\n",
-			i, profile_name[i],
-			data->custom_profile_setting.sclk_up_hyst,
-			data->custom_profile_setting.sclk_down_hyst,
-			data->custom_profile_setting.sclk_activity,
-			data->custom_profile_setting.mclk_up_hyst,
-			data->custom_profile_setting.mclk_down_hyst,
-			data->custom_profile_setting.mclk_activity);
-
-	size += sprintf(buf + size, "%3s %16s: %8d %16d %16d %16d %16d %16d\n",
-			"*", "CURRENT",
-			data->current_profile_setting.sclk_up_hyst,
-			data->current_profile_setting.sclk_down_hyst,
-			data->current_profile_setting.sclk_activity,
-			data->current_profile_setting.mclk_up_hyst,
-			data->current_profile_setting.mclk_down_hyst,
-			data->current_profile_setting.mclk_activity);
-
 	return size;
 }
 
@@ -4939,16 +4962,16 @@ static int smu7_set_power_profile_mode(struct pp_hwmgr *hwmgr, long *input, uint
 		if (size < 8)
 			return -EINVAL;
 
-		data->custom_profile_setting.bupdate_sclk = input[0];
-		data->custom_profile_setting.sclk_up_hyst = input[1];
-		data->custom_profile_setting.sclk_down_hyst = input[2];
-		data->custom_profile_setting.sclk_activity = input[3];
-		data->custom_profile_setting.bupdate_mclk = input[4];
-		data->custom_profile_setting.mclk_up_hyst = input[5];
-		data->custom_profile_setting.mclk_down_hyst = input[6];
-		data->custom_profile_setting.mclk_activity = input[7];
-		if (!smum_update_dpm_settings(hwmgr, &data->custom_profile_setting)) {
-			memcpy(&data->current_profile_setting, &data->custom_profile_setting, sizeof(struct profile_mode_setting));
+		tmp.bupdate_sclk = input[0];
+		tmp.sclk_up_hyst = input[1];
+		tmp.sclk_down_hyst = input[2];
+		tmp.sclk_activity = input[3];
+		tmp.bupdate_mclk = input[4];
+		tmp.mclk_up_hyst = input[5];
+		tmp.mclk_down_hyst = input[6];
+		tmp.mclk_activity = input[7];
+		if (!smum_update_dpm_settings(hwmgr, &tmp)) {
+			memcpy(&data->current_profile_setting, &tmp, sizeof(struct profile_mode_setting));
 			hwmgr->power_profile_mode = mode;
 		}
 		break;
@@ -5023,7 +5046,7 @@ static const struct pp_hwmgr_func smu7_hwmgr_funcs = {
 	.get_fan_control_mode = smu7_get_fan_control_mode,
 	.force_clock_level = smu7_force_clock_level,
 	.print_clock_levels = smu7_print_clock_levels,
-	.enable_per_cu_power_gating = smu7_enable_per_cu_power_gating,
+	.powergate_gfx = smu7_powergate_gfx,
 	.get_sclk_od = smu7_get_sclk_od,
 	.set_sclk_od = smu7_set_sclk_od,
 	.get_mclk_od = smu7_get_mclk_od,
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h
index f40179c9ca97..3784ce6e50ab 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.h
@@ -184,6 +184,8 @@ struct smu7_odn_dpm_table {
 	struct smu7_odn_clock_voltage_dependency_table	vdd_dependency_on_sclk;
 	struct smu7_odn_clock_voltage_dependency_table	vdd_dependency_on_mclk;
 	uint32_t					odn_mclk_min_limit;
+	uint32_t min_vddc;
+	uint32_t max_vddc;
 };
 
 struct profile_mode_setting {
@@ -308,7 +310,6 @@ struct smu7_hwmgr {
 	/* ---- Power Gating States ---- */
 	bool                           uvd_power_gated;
 	bool                           vce_power_gated;
-	bool                           samu_power_gated;
 	bool                           need_long_memory_training;
 
 	/* Application power optimization parameters */
@@ -325,7 +326,6 @@ struct smu7_hwmgr {
 	uint16_t                              mem_latency_high;
 	uint16_t                              mem_latency_low;
 	uint32_t                              vr_config;
-	struct profile_mode_setting           custom_profile_setting;
 	struct profile_mode_setting           current_profile_setting;
 };
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
index 03bc7453f3b1..c952845833d7 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_powertune.c
@@ -623,6 +623,190 @@ static const struct gpu_pt_config_reg DIDTConfig_Polaris11_Kicker[] =
 	{   0xFFFFFFFF  }  /* End of list */
 };
 
+static const struct gpu_pt_config_reg GCCACConfig_VegaM[] =
+{
+// ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
+//      Offset                             Mask                                                Shift                                               Value       Type
+// ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
+    // DIDT_SQ
+    //
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x00060013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x00860013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x01060013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x01860013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x02060013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x02860013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x03060013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x03860013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x04060013, GPU_CONFIGREG_GC_CAC_IND },
+
+    // DIDT_TD
+    //
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x000E0013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x008E0013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x010E0013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x018E0013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x020E0013, GPU_CONFIGREG_GC_CAC_IND },
+
+    // DIDT_TCP
+    //
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x00100013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x00900013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x01100013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x01900013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x02100013, GPU_CONFIGREG_GC_CAC_IND },
+    {   ixGC_CAC_CNTL,                     0xFFFFFFFF,                                         0,                                                  0x02900013, GPU_CONFIGREG_GC_CAC_IND },
+
+    {   0xFFFFFFFF  }  // End of list
+};
+
+static const struct gpu_pt_config_reg DIDTConfig_VegaM[] =
+{
+// ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
+//      Offset                             Mask                                                Shift                                               Value       Type
+// ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
+    // DIDT_SQ
+    //
+    {   ixDIDT_SQ_WEIGHT0_3,               DIDT_SQ_WEIGHT0_3__WEIGHT0_MASK,                    DIDT_SQ_WEIGHT0_3__WEIGHT0__SHIFT,                  0x0073,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT0_3,               DIDT_SQ_WEIGHT0_3__WEIGHT1_MASK,                    DIDT_SQ_WEIGHT0_3__WEIGHT1__SHIFT,                  0x00ab,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT0_3,               DIDT_SQ_WEIGHT0_3__WEIGHT2_MASK,                    DIDT_SQ_WEIGHT0_3__WEIGHT2__SHIFT,                  0x0084,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT0_3,               DIDT_SQ_WEIGHT0_3__WEIGHT3_MASK,                    DIDT_SQ_WEIGHT0_3__WEIGHT3__SHIFT,                  0x005a,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_WEIGHT4_7,               DIDT_SQ_WEIGHT4_7__WEIGHT4_MASK,                    DIDT_SQ_WEIGHT4_7__WEIGHT4__SHIFT,                  0x0067,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT4_7,               DIDT_SQ_WEIGHT4_7__WEIGHT5_MASK,                    DIDT_SQ_WEIGHT4_7__WEIGHT5__SHIFT,                  0x0084,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT4_7,               DIDT_SQ_WEIGHT4_7__WEIGHT6_MASK,                    DIDT_SQ_WEIGHT4_7__WEIGHT6__SHIFT,                  0x0027,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT4_7,               DIDT_SQ_WEIGHT4_7__WEIGHT7_MASK,                    DIDT_SQ_WEIGHT4_7__WEIGHT7__SHIFT,                  0x0046,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_WEIGHT8_11,              DIDT_SQ_WEIGHT8_11__WEIGHT8_MASK,                   DIDT_SQ_WEIGHT8_11__WEIGHT8__SHIFT,                 0x00aa,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT8_11,              DIDT_SQ_WEIGHT8_11__WEIGHT9_MASK,                   DIDT_SQ_WEIGHT8_11__WEIGHT9__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT8_11,              DIDT_SQ_WEIGHT8_11__WEIGHT10_MASK,                  DIDT_SQ_WEIGHT8_11__WEIGHT10__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_WEIGHT8_11,              DIDT_SQ_WEIGHT8_11__WEIGHT11_MASK,                  DIDT_SQ_WEIGHT8_11__WEIGHT11__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_CTRL1,                   DIDT_SQ_CTRL1__MIN_POWER_MASK,                      DIDT_SQ_CTRL1__MIN_POWER__SHIFT,                    0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL1,                   DIDT_SQ_CTRL1__MAX_POWER_MASK,                      DIDT_SQ_CTRL1__MAX_POWER__SHIFT,                    0xffff,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_CTRL_OCP,                DIDT_SQ_CTRL_OCP__UNUSED_0_MASK,                    DIDT_SQ_CTRL_OCP__UNUSED_0__SHIFT,                  0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL_OCP,                DIDT_SQ_CTRL_OCP__OCP_MAX_POWER_MASK,               DIDT_SQ_CTRL_OCP__OCP_MAX_POWER__SHIFT,             0xffff,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__MAX_POWER_DELTA_MASK,                DIDT_SQ_CTRL2__MAX_POWER_DELTA__SHIFT,              0x3853,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__UNUSED_0_MASK,                       DIDT_SQ_CTRL2__UNUSED_0__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK,       DIDT_SQ_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT,     0x005a,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__UNUSED_1_MASK,                       DIDT_SQ_CTRL2__UNUSED_1__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK,       DIDT_SQ_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT,     0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL2,                   DIDT_SQ_CTRL2__UNUSED_2_MASK,                       DIDT_SQ_CTRL2__UNUSED_2__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__DIDT_STALL_CTRL_ENABLE_MASK,    DIDT_SQ_STALL_CTRL__DIDT_STALL_CTRL_ENABLE__SHIFT,  0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK,       DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT,     0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK,       DIDT_SQ_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT,     0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__DIDT_HI_POWER_THRESHOLD_MASK,   DIDT_SQ_STALL_CTRL__DIDT_HI_POWER_THRESHOLD__SHIFT, 0x0ebb,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_STALL_CTRL,              DIDT_SQ_STALL_CTRL__UNUSED_0_MASK,                  DIDT_SQ_STALL_CTRL__UNUSED_0__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_TUNING_CTRL,             DIDT_SQ_TUNING_CTRL__DIDT_TUNING_ENABLE_MASK,       DIDT_SQ_TUNING_CTRL__DIDT_TUNING_ENABLE__SHIFT,     0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_TUNING_CTRL,             DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK,       DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT,     0x3853,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_TUNING_CTRL,             DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK,       DIDT_SQ_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT,     0x3153,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_TUNING_CTRL,             DIDT_SQ_TUNING_CTRL__UNUSED_0_MASK,                 DIDT_SQ_TUNING_CTRL__UNUSED_0__SHIFT,               0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_CTRL_EN_MASK,                   DIDT_SQ_CTRL0__DIDT_CTRL_EN__SHIFT,                 0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__USE_REF_CLOCK_MASK,                  DIDT_SQ_CTRL0__USE_REF_CLOCK__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__PHASE_OFFSET_MASK,                   DIDT_SQ_CTRL0__PHASE_OFFSET__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_CTRL_RST_MASK,                  DIDT_SQ_CTRL0__DIDT_CTRL_RST__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK,           DIDT_SQ_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT,         0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI_MASK,     DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT,   0x0010,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO_MASK,     DIDT_SQ_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT,   0x0010,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_SQ_CTRL0,                   DIDT_SQ_CTRL0__UNUSED_0_MASK,                       DIDT_SQ_CTRL0__UNUSED_0__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    // DIDT_TD
+    //
+    {   ixDIDT_TD_WEIGHT0_3,               DIDT_TD_WEIGHT0_3__WEIGHT0_MASK,                    DIDT_TD_WEIGHT0_3__WEIGHT0__SHIFT,                  0x000a,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_WEIGHT0_3,               DIDT_TD_WEIGHT0_3__WEIGHT1_MASK,                    DIDT_TD_WEIGHT0_3__WEIGHT1__SHIFT,                  0x0010,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_WEIGHT0_3,               DIDT_TD_WEIGHT0_3__WEIGHT2_MASK,                    DIDT_TD_WEIGHT0_3__WEIGHT2__SHIFT,                  0x0017,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_WEIGHT0_3,               DIDT_TD_WEIGHT0_3__WEIGHT3_MASK,                    DIDT_TD_WEIGHT0_3__WEIGHT3__SHIFT,                  0x002f,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TD_WEIGHT4_7,               DIDT_TD_WEIGHT4_7__WEIGHT4_MASK,                    DIDT_TD_WEIGHT4_7__WEIGHT4__SHIFT,                  0x0046,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_WEIGHT4_7,               DIDT_TD_WEIGHT4_7__WEIGHT5_MASK,                    DIDT_TD_WEIGHT4_7__WEIGHT5__SHIFT,                  0x005d,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_WEIGHT4_7,               DIDT_TD_WEIGHT4_7__WEIGHT6_MASK,                    DIDT_TD_WEIGHT4_7__WEIGHT6__SHIFT,                  0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_WEIGHT4_7,               DIDT_TD_WEIGHT4_7__WEIGHT7_MASK,                    DIDT_TD_WEIGHT4_7__WEIGHT7__SHIFT,                  0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TD_CTRL1,                   DIDT_TD_CTRL1__MIN_POWER_MASK,                      DIDT_TD_CTRL1__MIN_POWER__SHIFT,                    0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL1,                   DIDT_TD_CTRL1__MAX_POWER_MASK,                      DIDT_TD_CTRL1__MAX_POWER__SHIFT,                    0xffff,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TD_CTRL_OCP,                DIDT_TD_CTRL_OCP__UNUSED_0_MASK,                    DIDT_TD_CTRL_OCP__UNUSED_0__SHIFT,                  0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL_OCP,                DIDT_TD_CTRL_OCP__OCP_MAX_POWER_MASK,               DIDT_TD_CTRL_OCP__OCP_MAX_POWER__SHIFT,             0x00ff,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TD_CTRL2,                   DIDT_TD_CTRL2__MAX_POWER_DELTA_MASK,                DIDT_TD_CTRL2__MAX_POWER_DELTA__SHIFT,              0x3fff,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL2,                   DIDT_TD_CTRL2__UNUSED_0_MASK,                       DIDT_TD_CTRL2__UNUSED_0__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL2,                   DIDT_TD_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK,       DIDT_TD_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT,     0x000f,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL2,                   DIDT_TD_CTRL2__UNUSED_1_MASK,                       DIDT_TD_CTRL2__UNUSED_1__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL2,                   DIDT_TD_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK,       DIDT_TD_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT,     0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL2,                   DIDT_TD_CTRL2__UNUSED_2_MASK,                       DIDT_TD_CTRL2__UNUSED_2__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TD_STALL_CTRL,              DIDT_TD_STALL_CTRL__DIDT_STALL_CTRL_ENABLE_MASK,    DIDT_TD_STALL_CTRL__DIDT_STALL_CTRL_ENABLE__SHIFT,  0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_STALL_CTRL,              DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK,       DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT,     0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_STALL_CTRL,              DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK,       DIDT_TD_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT,     0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_STALL_CTRL,              DIDT_TD_STALL_CTRL__DIDT_HI_POWER_THRESHOLD_MASK,   DIDT_TD_STALL_CTRL__DIDT_HI_POWER_THRESHOLD__SHIFT, 0x01aa,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_STALL_CTRL,              DIDT_TD_STALL_CTRL__UNUSED_0_MASK,                  DIDT_TD_STALL_CTRL__UNUSED_0__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TD_TUNING_CTRL,             DIDT_TD_TUNING_CTRL__DIDT_TUNING_ENABLE_MASK,       DIDT_TD_TUNING_CTRL__DIDT_TUNING_ENABLE__SHIFT,     0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_TUNING_CTRL,             DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK,       DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT,     0x0dde,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_TUNING_CTRL,             DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK,       DIDT_TD_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT,     0x0dde,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_TUNING_CTRL,             DIDT_TD_TUNING_CTRL__UNUSED_0_MASK,                 DIDT_TD_TUNING_CTRL__UNUSED_0__SHIFT,               0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__DIDT_CTRL_EN_MASK,                   DIDT_TD_CTRL0__DIDT_CTRL_EN__SHIFT,                 0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__USE_REF_CLOCK_MASK,                  DIDT_TD_CTRL0__USE_REF_CLOCK__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__PHASE_OFFSET_MASK,                   DIDT_TD_CTRL0__PHASE_OFFSET__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__DIDT_CTRL_RST_MASK,                  DIDT_TD_CTRL0__DIDT_CTRL_RST__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK,           DIDT_TD_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT,         0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI_MASK,     DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT,   0x0009,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO_MASK,     DIDT_TD_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT,   0x0009,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TD_CTRL0,                   DIDT_TD_CTRL0__UNUSED_0_MASK,                       DIDT_TD_CTRL0__UNUSED_0__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    // DIDT_TCP
+    //
+    {   ixDIDT_TCP_WEIGHT0_3,              DIDT_TCP_WEIGHT0_3__WEIGHT0_MASK,                   DIDT_TCP_WEIGHT0_3__WEIGHT0__SHIFT,                 0x0004,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_WEIGHT0_3,              DIDT_TCP_WEIGHT0_3__WEIGHT1_MASK,                   DIDT_TCP_WEIGHT0_3__WEIGHT1__SHIFT,                 0x0037,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_WEIGHT0_3,              DIDT_TCP_WEIGHT0_3__WEIGHT2_MASK,                   DIDT_TCP_WEIGHT0_3__WEIGHT2__SHIFT,                 0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_WEIGHT0_3,              DIDT_TCP_WEIGHT0_3__WEIGHT3_MASK,                   DIDT_TCP_WEIGHT0_3__WEIGHT3__SHIFT,                 0x00ff,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TCP_WEIGHT4_7,              DIDT_TCP_WEIGHT4_7__WEIGHT4_MASK,                   DIDT_TCP_WEIGHT4_7__WEIGHT4__SHIFT,                 0x0054,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_WEIGHT4_7,              DIDT_TCP_WEIGHT4_7__WEIGHT5_MASK,                   DIDT_TCP_WEIGHT4_7__WEIGHT5__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_WEIGHT4_7,              DIDT_TCP_WEIGHT4_7__WEIGHT6_MASK,                   DIDT_TCP_WEIGHT4_7__WEIGHT6__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_WEIGHT4_7,              DIDT_TCP_WEIGHT4_7__WEIGHT7_MASK,                   DIDT_TCP_WEIGHT4_7__WEIGHT7__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TCP_CTRL1,                  DIDT_TCP_CTRL1__MIN_POWER_MASK,                     DIDT_TCP_CTRL1__MIN_POWER__SHIFT,                   0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL1,                  DIDT_TCP_CTRL1__MAX_POWER_MASK,                     DIDT_TCP_CTRL1__MAX_POWER__SHIFT,                   0xffff,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TCP_CTRL_OCP,               DIDT_TCP_CTRL_OCP__UNUSED_0_MASK,                   DIDT_TCP_CTRL_OCP__UNUSED_0__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL_OCP,               DIDT_TCP_CTRL_OCP__OCP_MAX_POWER_MASK,              DIDT_TCP_CTRL_OCP__OCP_MAX_POWER__SHIFT,            0xffff,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TCP_CTRL2,                  DIDT_TCP_CTRL2__MAX_POWER_DELTA_MASK,               DIDT_TCP_CTRL2__MAX_POWER_DELTA__SHIFT,             0x3dde,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL2,                  DIDT_TCP_CTRL2__UNUSED_0_MASK,                      DIDT_TCP_CTRL2__UNUSED_0__SHIFT,                    0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL2,                  DIDT_TCP_CTRL2__SHORT_TERM_INTERVAL_SIZE_MASK,      DIDT_TCP_CTRL2__SHORT_TERM_INTERVAL_SIZE__SHIFT,    0x0032,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL2,                  DIDT_TCP_CTRL2__UNUSED_1_MASK,                      DIDT_TCP_CTRL2__UNUSED_1__SHIFT,                    0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL2,                  DIDT_TCP_CTRL2__LONG_TERM_INTERVAL_RATIO_MASK,      DIDT_TCP_CTRL2__LONG_TERM_INTERVAL_RATIO__SHIFT,    0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL2,                  DIDT_TCP_CTRL2__UNUSED_2_MASK,                      DIDT_TCP_CTRL2__UNUSED_2__SHIFT,                    0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TCP_STALL_CTRL,             DIDT_TCP_STALL_CTRL__DIDT_STALL_CTRL_ENABLE_MASK,   DIDT_TCP_STALL_CTRL__DIDT_STALL_CTRL_ENABLE__SHIFT, 0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_STALL_CTRL,             DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_HI_MASK,      DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_HI__SHIFT,    0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_STALL_CTRL,             DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_LO_MASK,      DIDT_TCP_STALL_CTRL__DIDT_STALL_DELAY_LO__SHIFT,    0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_STALL_CTRL,             DIDT_TCP_STALL_CTRL__DIDT_HI_POWER_THRESHOLD_MASK,  DIDT_TCP_STALL_CTRL__DIDT_HI_POWER_THRESHOLD__SHIFT,0x01aa,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_STALL_CTRL,             DIDT_TCP_STALL_CTRL__UNUSED_0_MASK,                 DIDT_TCP_STALL_CTRL__UNUSED_0__SHIFT,               0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TCP_TUNING_CTRL,            DIDT_TCP_TUNING_CTRL__DIDT_TUNING_ENABLE_MASK,      DIDT_TCP_TUNING_CTRL__DIDT_TUNING_ENABLE__SHIFT,    0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_TUNING_CTRL,            DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_HI_MASK,      DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_HI__SHIFT,    0x3dde,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_TUNING_CTRL,            DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_LO_MASK,      DIDT_TCP_TUNING_CTRL__MAX_POWER_DELTA_LO__SHIFT,    0x3dde,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_TUNING_CTRL,            DIDT_TCP_TUNING_CTRL__UNUSED_0_MASK,                DIDT_TCP_TUNING_CTRL__UNUSED_0__SHIFT,              0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__DIDT_CTRL_EN_MASK,                   DIDT_TCP_CTRL0__DIDT_CTRL_EN__SHIFT,                 0x0001,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__USE_REF_CLOCK_MASK,                  DIDT_TCP_CTRL0__USE_REF_CLOCK__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__PHASE_OFFSET_MASK,                   DIDT_TCP_CTRL0__PHASE_OFFSET__SHIFT,                 0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__DIDT_CTRL_RST_MASK,                  DIDT_TCP_CTRL0__DIDT_CTRL_RST__SHIFT,                0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__DIDT_CLK_EN_OVERRIDE_MASK,           DIDT_TCP_CTRL0__DIDT_CLK_EN_OVERRIDE__SHIFT,         0x0000,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI_MASK,     DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_HI__SHIFT,   0x0010,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO_MASK,     DIDT_TCP_CTRL0__DIDT_MAX_STALLS_ALLOWED_LO__SHIFT,   0x0010,     GPU_CONFIGREG_DIDT_IND },
+    {   ixDIDT_TCP_CTRL0,                   DIDT_TCP_CTRL0__UNUSED_0_MASK,                       DIDT_TCP_CTRL0__UNUSED_0__SHIFT,                     0x0000,     GPU_CONFIGREG_DIDT_IND },
+
+    {   0xFFFFFFFF  }  // End of list
+};
 static int smu7_enable_didt(struct pp_hwmgr *hwmgr, const bool enable)
 {
 	uint32_t en = enable ? 1 : 0;
@@ -740,8 +924,8 @@ int smu7_enable_didt_config(struct pp_hwmgr *hwmgr)
 	    PP_CAP(PHM_PlatformCaps_TDRamping) ||
 	    PP_CAP(PHM_PlatformCaps_TCPRamping)) {
 
-		cgs_enter_safe_mode(hwmgr->device, true);
-		cgs_lock_grbm_idx(hwmgr->device, true);
+		adev->gfx.rlc.funcs->enter_safe_mode(adev);
+		mutex_lock(&adev->grbm_idx_mutex);
 		value = 0;
 		value2 = cgs_read_register(hwmgr->device, mmGRBM_GFX_INDEX);
 		for (count = 0; count < num_se; count++) {
@@ -752,67 +936,80 @@ int smu7_enable_didt_config(struct pp_hwmgr *hwmgr)
 
 			if (hwmgr->chip_id == CHIP_POLARIS10) {
 				result = smu7_program_pt_config_registers(hwmgr, GCCACConfig_Polaris10);
-				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
 				result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_Polaris10);
-				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
 			} else if (hwmgr->chip_id == CHIP_POLARIS11) {
 				result = smu7_program_pt_config_registers(hwmgr, GCCACConfig_Polaris11);
-				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
 				if (hwmgr->is_kicker)
 					result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_Polaris11_Kicker);
 				else
 					result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_Polaris11);
-				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
 			} else if (hwmgr->chip_id == CHIP_POLARIS12) {
 				result = smu7_program_pt_config_registers(hwmgr, GCCACConfig_Polaris11);
-				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
 				result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_Polaris12);
-				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", return result);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
+			} else if (hwmgr->chip_id == CHIP_VEGAM) {
+				result = smu7_program_pt_config_registers(hwmgr, GCCACConfig_VegaM);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
+				result = smu7_program_pt_config_registers(hwmgr, DIDTConfig_VegaM);
+				PP_ASSERT_WITH_CODE((result == 0), "DIDT Config failed.", goto error);
 			}
 		}
 		cgs_write_register(hwmgr->device, mmGRBM_GFX_INDEX, value2);
 
 		result = smu7_enable_didt(hwmgr, true);
-		PP_ASSERT_WITH_CODE((result == 0), "EnableDiDt failed.", return result);
+		PP_ASSERT_WITH_CODE((result == 0), "EnableDiDt failed.", goto error);
 
 		if (hwmgr->chip_id == CHIP_POLARIS11) {
 			result = smum_send_msg_to_smc(hwmgr,
 						(uint16_t)(PPSMC_MSG_EnableDpmDidt));
 			PP_ASSERT_WITH_CODE((0 == result),
-					"Failed to enable DPM DIDT.", return result);
+					"Failed to enable DPM DIDT.", goto error);
 		}
-		cgs_lock_grbm_idx(hwmgr->device, false);
-		cgs_enter_safe_mode(hwmgr->device, false);
+		mutex_unlock(&adev->grbm_idx_mutex);
+		adev->gfx.rlc.funcs->exit_safe_mode(adev);
 	}
 
 	return 0;
+error:
+	mutex_unlock(&adev->grbm_idx_mutex);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
+	return result;
 }
 
 int smu7_disable_didt_config(struct pp_hwmgr *hwmgr)
 {
 	int result;
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	if (PP_CAP(PHM_PlatformCaps_SQRamping) ||
 	    PP_CAP(PHM_PlatformCaps_DBRamping) ||
 	    PP_CAP(PHM_PlatformCaps_TDRamping) ||
 	    PP_CAP(PHM_PlatformCaps_TCPRamping)) {
 
-		cgs_enter_safe_mode(hwmgr->device, true);
+		adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 		result = smu7_enable_didt(hwmgr, false);
 		PP_ASSERT_WITH_CODE((result == 0),
 				"Post DIDT enable clock gating failed.",
-				return result);
+				goto error);
 		if (hwmgr->chip_id == CHIP_POLARIS11) {
 			result = smum_send_msg_to_smc(hwmgr,
 						(uint16_t)(PPSMC_MSG_DisableDpmDidt));
 			PP_ASSERT_WITH_CODE((0 == result),
-					"Failed to disable DPM DIDT.", return result);
+					"Failed to disable DPM DIDT.", goto error);
 		}
-		cgs_enter_safe_mode(hwmgr->device, false);
+		adev->gfx.rlc.funcs->exit_safe_mode(adev);
 	}
 
 	return 0;
+error:
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
+	return result;
 }
 
 int smu7_enable_smc_cac(struct pp_hwmgr *hwmgr)
@@ -852,12 +1049,10 @@ int smu7_set_power_limit(struct pp_hwmgr *hwmgr, uint32_t n)
 {
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 
-	n = (n & 0xff) << 8;
-
 	if (data->power_containment_features &
 			POWERCONTAINMENT_FEATURE_PkgPwrLimit)
 		return smum_send_msg_to_smc_with_parameter(hwmgr,
-				PPSMC_MSG_PkgPwrSetLimit, n);
+				PPSMC_MSG_PkgPwrSetLimit, n<<8);
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c
index 7b26607c646a..288802f209dd 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu8_hwmgr.c
@@ -314,8 +314,7 @@ static int smu8_get_system_info_data(struct pp_hwmgr *hwmgr)
 	uint8_t frev, crev;
 	uint16_t size;
 
-	info = (ATOM_INTEGRATED_SYSTEM_INFO_V1_9 *) cgs_atom_get_data_table(
-			hwmgr->device,
+	info = (ATOM_INTEGRATED_SYSTEM_INFO_V1_9 *)smu_atom_get_data_table(hwmgr->adev,
 			GetIndexIntoMasterTable(DATA, IntegratedSystemInfo),
 			&size, &frev, &crev);
 
@@ -694,7 +693,7 @@ static int smu8_update_sclk_limit(struct pp_hwmgr *hwmgr)
 	else
 		data->sclk_dpm.soft_max_clk  = table->entries[table->count - 1].clk;
 
-	clock = hwmgr->display_config.min_core_set_clock;
+	clock = hwmgr->display_config->min_core_set_clock;
 	if (clock == 0)
 		pr_debug("min_core_set_clock not set\n");
 
@@ -749,7 +748,7 @@ static int smu8_set_deep_sleep_sclk_threshold(struct pp_hwmgr *hwmgr)
 {
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
 				PHM_PlatformCaps_SclkDeepSleep)) {
-		uint32_t clks = hwmgr->display_config.min_core_set_clock_in_sr;
+		uint32_t clks = hwmgr->display_config->min_core_set_clock_in_sr;
 		if (clks == 0)
 			clks = SMU8_MIN_DEEP_SLEEP_SCLK;
 
@@ -1041,25 +1040,21 @@ static int smu8_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 	struct smu8_hwmgr *data = hwmgr->backend;
 	struct PP_Clocks clocks = {0, 0, 0, 0};
 	bool force_high;
-	uint32_t  num_of_active_displays = 0;
-	struct cgs_display_info info = {0};
 
 	smu8_ps->need_dfs_bypass = true;
 
 	data->battery_state = (PP_StateUILabel_Battery == prequest_ps->classification.ui_label);
 
-	clocks.memoryClock = hwmgr->display_config.min_mem_set_clock != 0 ?
-				hwmgr->display_config.min_mem_set_clock :
+	clocks.memoryClock = hwmgr->display_config->min_mem_set_clock != 0 ?
+				hwmgr->display_config->min_mem_set_clock :
 				data->sys_info.nbp_memory_clock[1];
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-	num_of_active_displays = info.display_count;
 
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_StablePState))
 		clocks.memoryClock = hwmgr->dyn_state.max_clock_voltage_on_ac.mclk;
 
 	force_high = (clocks.memoryClock > data->sys_info.nbp_memory_clock[SMU8_NUM_NBPMEMORYCLOCK - 1])
-			|| (num_of_active_displays >= 3);
+			|| (hwmgr->display_config->num_display >= 3);
 
 	smu8_ps->action = smu8_current_ps->action;
 
@@ -1609,17 +1604,17 @@ static int smu8_get_clock_by_type(struct pp_hwmgr *hwmgr, enum amd_pp_clock_type
 	switch (type) {
 	case amd_pp_disp_clock:
 		for (i = 0; i < clocks->count; i++)
-			clocks->clock[i] = data->sys_info.display_clock[i];
+			clocks->clock[i] = data->sys_info.display_clock[i] * 10;
 		break;
 	case amd_pp_sys_clock:
 		table = hwmgr->dyn_state.vddc_dependency_on_sclk;
 		for (i = 0; i < clocks->count; i++)
-			clocks->clock[i] = table->entries[i].clk;
+			clocks->clock[i] = table->entries[i].clk * 10;
 		break;
 	case amd_pp_mem_clock:
 		clocks->count = SMU8_NUM_NBPMEMORYCLOCK;
 		for (i = 0; i < clocks->count; i++)
-			clocks->clock[i] = data->sys_info.nbp_memory_clock[clocks->count - 1 - i];
+			clocks->clock[i] = data->sys_info.nbp_memory_clock[clocks->count - 1 - i] * 10;
 		break;
 	default:
 		return -1;
@@ -1897,20 +1892,20 @@ static void smu8_dpm_powergate_uvd(struct pp_hwmgr *hwmgr, bool bgate)
 	data->uvd_power_gated = bgate;
 
 	if (bgate) {
-		cgs_set_powergating_state(hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_UVD,
 						AMD_PG_STATE_GATE);
-		cgs_set_clockgating_state(hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_UVD,
 						AMD_CG_STATE_GATE);
 		smu8_dpm_update_uvd_dpm(hwmgr, true);
 		smu8_dpm_powerdown_uvd(hwmgr);
 	} else {
 		smu8_dpm_powerup_uvd(hwmgr);
-		cgs_set_clockgating_state(hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_UVD,
 						AMD_CG_STATE_UNGATE);
-		cgs_set_powergating_state(hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 						AMD_IP_BLOCK_TYPE_UVD,
 						AMD_PG_STATE_UNGATE);
 		smu8_dpm_update_uvd_dpm(hwmgr, false);
@@ -1923,12 +1918,10 @@ static void smu8_dpm_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate)
 	struct smu8_hwmgr *data = hwmgr->backend;
 
 	if (bgate) {
-		cgs_set_powergating_state(
-					hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 					AMD_IP_BLOCK_TYPE_VCE,
 					AMD_PG_STATE_GATE);
-		cgs_set_clockgating_state(
-					hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 					AMD_IP_BLOCK_TYPE_VCE,
 					AMD_CG_STATE_GATE);
 		smu8_enable_disable_vce_dpm(hwmgr, false);
@@ -1937,12 +1930,10 @@ static void smu8_dpm_powergate_vce(struct pp_hwmgr *hwmgr, bool bgate)
 	} else {
 		smu8_dpm_powerup_vce(hwmgr);
 		data->vce_power_gated = false;
-		cgs_set_clockgating_state(
-					hwmgr->device,
+		amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 					AMD_IP_BLOCK_TYPE_VCE,
 					AMD_CG_STATE_UNGATE);
-		cgs_set_powergating_state(
-					hwmgr->device,
+		amdgpu_device_ip_set_powergating_state(hwmgr->adev,
 					AMD_IP_BLOCK_TYPE_VCE,
 					AMD_PG_STATE_UNGATE);
 		smu8_dpm_update_vce_dpm(hwmgr);
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c
index 598122854ab5..2aab1b475945 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.c
@@ -24,6 +24,10 @@
 #include "pp_debug.h"
 #include "ppatomctrl.h"
 #include "ppsmc.h"
+#include "atom.h"
+#include "ivsrcid/thm/irqsrcs_thm_9_0.h"
+#include "ivsrcid/smuio/irqsrcs_smuio_9_0.h"
+#include "ivsrcid/ivsrcid_vislands30.h"
 
 uint8_t convert_to_vid(uint16_t vddc)
 {
@@ -542,17 +546,17 @@ int phm_irq_process(struct amdgpu_device *adev,
 	uint32_t src_id = entry->src_id;
 
 	if (client_id == AMDGPU_IH_CLIENTID_LEGACY) {
-		if (src_id == 230)
+		if (src_id == VISLANDS30_IV_SRCID_CG_TSS_THERMAL_LOW_TO_HIGH)
 			pr_warn("GPU over temperature range detected on PCIe %d:%d.%d!\n",
 						PCI_BUS_NUM(adev->pdev->devfn),
 						PCI_SLOT(adev->pdev->devfn),
 						PCI_FUNC(adev->pdev->devfn));
-		else if (src_id == 231)
+		else if (src_id == VISLANDS30_IV_SRCID_CG_TSS_THERMAL_HIGH_TO_LOW)
 			pr_warn("GPU under temperature range detected on PCIe %d:%d.%d!\n",
 					PCI_BUS_NUM(adev->pdev->devfn),
 					PCI_SLOT(adev->pdev->devfn),
 					PCI_FUNC(adev->pdev->devfn));
-		else if (src_id == 83)
+		else if (src_id == VISLANDS30_IV_SRCID_GPIO_19)
 			pr_warn("GPU Critical Temperature Fault detected on PCIe %d:%d.%d!\n",
 					PCI_BUS_NUM(adev->pdev->devfn),
 					PCI_SLOT(adev->pdev->devfn),
@@ -593,18 +597,115 @@ int smu9_register_irq_handlers(struct pp_hwmgr *hwmgr)
 
 	amdgpu_irq_add_id((struct amdgpu_device *)(hwmgr->adev),
 			SOC15_IH_CLIENTID_THM,
-			0,
+			THM_9_0__SRCID__THM_DIG_THERM_L2H,
 			source);
 	amdgpu_irq_add_id((struct amdgpu_device *)(hwmgr->adev),
 			SOC15_IH_CLIENTID_THM,
-			1,
+			THM_9_0__SRCID__THM_DIG_THERM_H2L,
 			source);
 
 	/* Register CTF(GPIO_19) interrupt */
 	amdgpu_irq_add_id((struct amdgpu_device *)(hwmgr->adev),
 			SOC15_IH_CLIENTID_ROM_SMUIO,
-			83,
+			SMUIO_9_0__SRCID__SMUIO_GPIO19,
 			source);
 
 	return 0;
 }
+
+void *smu_atom_get_data_table(void *dev, uint32_t table, uint16_t *size,
+						uint8_t *frev, uint8_t *crev)
+{
+	struct amdgpu_device *adev = dev;
+	uint16_t data_start;
+
+	if (amdgpu_atom_parse_data_header(
+		    adev->mode_info.atom_context, table, size,
+		    frev, crev, &data_start))
+		return (uint8_t *)adev->mode_info.atom_context->bios +
+			data_start;
+
+	return NULL;
+}
+
+int smu_get_voltage_dependency_table_ppt_v1(
+			const struct phm_ppt_v1_clock_voltage_dependency_table *allowed_dep_table,
+			struct phm_ppt_v1_clock_voltage_dependency_table *dep_table)
+{
+	uint8_t i = 0;
+	PP_ASSERT_WITH_CODE((0 != allowed_dep_table->count),
+				"Voltage Lookup Table empty",
+				return -EINVAL);
+
+	dep_table->count = allowed_dep_table->count;
+	for (i=0; i<dep_table->count; i++) {
+		dep_table->entries[i].clk = allowed_dep_table->entries[i].clk;
+		dep_table->entries[i].vddInd = allowed_dep_table->entries[i].vddInd;
+		dep_table->entries[i].vdd_offset = allowed_dep_table->entries[i].vdd_offset;
+		dep_table->entries[i].vddc = allowed_dep_table->entries[i].vddc;
+		dep_table->entries[i].vddgfx = allowed_dep_table->entries[i].vddgfx;
+		dep_table->entries[i].vddci = allowed_dep_table->entries[i].vddci;
+		dep_table->entries[i].mvdd = allowed_dep_table->entries[i].mvdd;
+		dep_table->entries[i].phases = allowed_dep_table->entries[i].phases;
+		dep_table->entries[i].cks_enable = allowed_dep_table->entries[i].cks_enable;
+		dep_table->entries[i].cks_voffset = allowed_dep_table->entries[i].cks_voffset;
+	}
+
+	return 0;
+}
+
+int smu_set_watermarks_for_clocks_ranges(void *wt_table,
+		struct dm_pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges)
+{
+	uint32_t i;
+	struct watermarks *table = wt_table;
+
+	if (!table || !wm_with_clock_ranges)
+		return -EINVAL;
+
+	if (wm_with_clock_ranges->num_wm_dmif_sets > 4 || wm_with_clock_ranges->num_wm_mcif_sets > 4)
+		return -EINVAL;
+
+	for (i = 0; i < wm_with_clock_ranges->num_wm_dmif_sets; i++) {
+		table->WatermarkRow[1][i].MinClock =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_dmif_clocks_ranges[i].wm_min_dcfclk_clk_in_khz) /
+			1000);
+		table->WatermarkRow[1][i].MaxClock =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_dmif_clocks_ranges[i].wm_max_dcfclk_clk_in_khz) /
+			100);
+		table->WatermarkRow[1][i].MinUclk =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_dmif_clocks_ranges[i].wm_min_mem_clk_in_khz) /
+			1000);
+		table->WatermarkRow[1][i].MaxUclk =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_dmif_clocks_ranges[i].wm_max_mem_clk_in_khz) /
+			1000);
+		table->WatermarkRow[1][i].WmSetting = (uint8_t)
+				wm_with_clock_ranges->wm_dmif_clocks_ranges[i].wm_set_id;
+	}
+
+	for (i = 0; i < wm_with_clock_ranges->num_wm_mcif_sets; i++) {
+		table->WatermarkRow[0][i].MinClock =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_mcif_clocks_ranges[i].wm_min_socclk_clk_in_khz) /
+			1000);
+		table->WatermarkRow[0][i].MaxClock =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_mcif_clocks_ranges[i].wm_max_socclk_clk_in_khz) /
+			1000);
+		table->WatermarkRow[0][i].MinUclk =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_mcif_clocks_ranges[i].wm_min_mem_clk_in_khz) /
+			1000);
+		table->WatermarkRow[0][i].MaxUclk =
+			cpu_to_le16((uint16_t)
+			(wm_with_clock_ranges->wm_mcif_clocks_ranges[i].wm_max_mem_clk_in_khz) /
+			1000);
+		table->WatermarkRow[0][i].WmSetting = (uint8_t)
+				wm_with_clock_ranges->wm_mcif_clocks_ranges[i].wm_set_id;
+	}
+	return 0;
+}
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h
index d37d16e4b613..5454289d5226 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu_helper.h
@@ -26,10 +26,27 @@
 struct pp_atomctrl_voltage_table;
 struct pp_hwmgr;
 struct phm_ppt_v1_voltage_lookup_table;
+struct Watermarks_t;
+struct pp_wm_sets_with_clock_ranges_soc15;
 
 uint8_t convert_to_vid(uint16_t vddc);
 uint16_t convert_to_vddc(uint8_t vid);
 
+struct watermark_row_generic_t {
+	uint16_t MinClock;
+	uint16_t MaxClock;
+	uint16_t MinUclk;
+	uint16_t MaxUclk;
+
+	uint8_t  WmSetting;
+	uint8_t  Padding[3];
+};
+
+struct watermarks {
+	struct watermark_row_generic_t WatermarkRow[2][4];
+	uint32_t     padding[7];
+};
+
 extern int phm_wait_for_register_unequal(struct pp_hwmgr *hwmgr,
 					uint32_t index,
 					uint32_t value, uint32_t mask);
@@ -82,6 +99,16 @@ int phm_irq_process(struct amdgpu_device *adev,
 
 int smu9_register_irq_handlers(struct pp_hwmgr *hwmgr);
 
+void *smu_atom_get_data_table(void *dev, uint32_t table, uint16_t *size,
+						uint8_t *frev, uint8_t *crev);
+
+int smu_get_voltage_dependency_table_ppt_v1(
+	const struct phm_ppt_v1_clock_voltage_dependency_table *allowed_dep_table,
+		struct phm_ppt_v1_clock_voltage_dependency_table *dep_table);
+
+int smu_set_watermarks_for_clocks_ranges(void *wt_table,
+		struct dm_pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges);
+
 #define PHM_FIELD_SHIFT(reg, field) reg##__##field##__SHIFT
 #define PHM_FIELD_MASK(reg, field) reg##__##field##_MASK
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
index 7cbb56ba6fab..1a0dccb3fac1 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c
@@ -36,7 +36,7 @@
 #include "smu9.h"
 #include "smu9_driver_if.h"
 #include "vega10_inc.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 #include "pppcielanes.h"
 #include "vega10_hwmgr.h"
 #include "vega10_processpptables.h"
@@ -51,20 +51,10 @@
 #include "smuio/smuio_9_0_offset.h"
 #include "smuio/smuio_9_0_sh_mask.h"
 
-#define VOLTAGE_SCALE  4
-#define VOLTAGE_VID_OFFSET_SCALE1   625
-#define VOLTAGE_VID_OFFSET_SCALE2   100
-
 #define HBM_MEMORY_CHANNEL_WIDTH    128
 
 static const uint32_t channel_number[] = {1, 2, 0, 4, 0, 8, 0, 16, 2};
 
-#define MEM_FREQ_LOW_LATENCY        25000
-#define MEM_FREQ_HIGH_LATENCY       80000
-#define MEM_LATENCY_HIGH            245
-#define MEM_LATENCY_LOW             35
-#define MEM_LATENCY_ERR             0xFFFF
-
 #define mmDF_CS_AON0_DramBaseAddress0                                                                  0x0044
 #define mmDF_CS_AON0_DramBaseAddress0_BASE_IDX                                                         0
 
@@ -79,8 +69,6 @@ static const uint32_t channel_number[] = {1, 2, 0, 4, 0, 8, 0, 16, 2};
 #define DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK                                                        0x000000F0L
 #define DF_CS_AON0_DramBaseAddress0__IntLvAddrSel_MASK                                                        0x00000700L
 #define DF_CS_AON0_DramBaseAddress0__DramBaseAddr_MASK                                                        0xFFFFF000L
-static int vega10_force_clock_level(struct pp_hwmgr *hwmgr,
-		enum pp_clock_type type, uint32_t mask);
 
 static const ULONG PhwVega10_Magic = (ULONG)(PHM_VIslands_Magic);
 
@@ -291,6 +279,52 @@ static int vega10_set_features_platform_caps(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
+static int vega10_odn_initial_default_setting(struct pp_hwmgr *hwmgr)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+	struct phm_ppt_v2_information *table_info =
+			(struct phm_ppt_v2_information *)(hwmgr->pptable);
+	struct vega10_odn_dpm_table *odn_table = &(data->odn_dpm_table);
+	struct vega10_odn_vddc_lookup_table *od_lookup_table;
+	struct phm_ppt_v1_voltage_lookup_table *vddc_lookup_table;
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_table[3];
+	struct phm_ppt_v1_clock_voltage_dependency_table *od_table[3];
+	uint32_t i;
+
+	od_lookup_table = &odn_table->vddc_lookup_table;
+	vddc_lookup_table = table_info->vddc_lookup_table;
+
+	for (i = 0; i < vddc_lookup_table->count; i++)
+		od_lookup_table->entries[i].us_vdd = vddc_lookup_table->entries[i].us_vdd;
+
+	od_lookup_table->count = vddc_lookup_table->count;
+
+	dep_table[0] = table_info->vdd_dep_on_sclk;
+	dep_table[1] = table_info->vdd_dep_on_mclk;
+	dep_table[2] = table_info->vdd_dep_on_socclk;
+	od_table[0] = (struct phm_ppt_v1_clock_voltage_dependency_table *)&odn_table->vdd_dep_on_sclk;
+	od_table[1] = (struct phm_ppt_v1_clock_voltage_dependency_table *)&odn_table->vdd_dep_on_mclk;
+	od_table[2] = (struct phm_ppt_v1_clock_voltage_dependency_table *)&odn_table->vdd_dep_on_socclk;
+
+	for (i = 0; i < 3; i++)
+		smu_get_voltage_dependency_table_ppt_v1(dep_table[i], od_table[i]);
+
+	if (odn_table->max_vddc == 0 || odn_table->max_vddc > 2000)
+		odn_table->max_vddc = dep_table[0]->entries[dep_table[0]->count - 1].vddc;
+	if (odn_table->min_vddc == 0 || odn_table->min_vddc > 2000)
+		odn_table->min_vddc = dep_table[0]->entries[0].vddc;
+
+	i = od_table[2]->count - 1;
+	od_table[2]->entries[i].clk = hwmgr->platform_descriptor.overdriveLimit.memoryClock > od_table[2]->entries[i].clk ?
+					hwmgr->platform_descriptor.overdriveLimit.memoryClock :
+					od_table[2]->entries[i].clk;
+	od_table[2]->entries[i].vddc = odn_table->max_vddc > od_table[2]->entries[i].vddc ?
+					odn_table->max_vddc :
+					od_table[2]->entries[i].vddc;
+
+	return 0;
+}
+
 static void vega10_init_dpm_defaults(struct pp_hwmgr *hwmgr)
 {
 	struct vega10_hwmgr *data = hwmgr->backend;
@@ -427,7 +461,6 @@ static void vega10_init_dpm_defaults(struct pp_hwmgr *hwmgr)
 		/* ACG firmware has major version 5 */
 	if ((hwmgr->smu_version & 0xff000000) == 0x5000000)
 		data->smu_features[GNLD_ACG].supported = true;
-
 	if (data->registry_data.didt_support)
 		data->smu_features[GNLD_DIDT].supported = true;
 
@@ -754,7 +787,6 @@ static int vega10_hwmgr_backend_init(struct pp_hwmgr *hwmgr)
 	uint32_t config_telemetry = 0;
 	struct pp_atomfwctrl_voltage_table vol_table;
 	struct amdgpu_device *adev = hwmgr->adev;
-	uint32_t reg;
 
 	data = kzalloc(sizeof(struct vega10_hwmgr), GFP_KERNEL);
 	if (data == NULL)
@@ -860,10 +892,7 @@ static int vega10_hwmgr_backend_init(struct pp_hwmgr *hwmgr)
 			advanceFanControlParameters.usFanPWMMinLimit *
 			hwmgr->thermal_controller.fanInfo.ulMaxRPM / 100;
 
-	reg = soc15_get_register_offset(DF_HWID, 0,
-			mmDF_CS_AON0_DramBaseAddress0_BASE_IDX,
-			mmDF_CS_AON0_DramBaseAddress0);
-	data->mem_channels = (cgs_read_register(hwmgr->device, reg) &
+	data->mem_channels = (RREG32_SOC15(DF, 0, mmDF_CS_AON0_DramBaseAddress0) &
 			DF_CS_AON0_DramBaseAddress0__IntLvNumChan_MASK) >>
 			DF_CS_AON0_DramBaseAddress0__IntLvNumChan__SHIFT;
 	PP_ASSERT_WITH_CODE(data->mem_channels < ARRAY_SIZE(channel_number),
@@ -1280,6 +1309,9 @@ static int vega10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 	vega10_setup_default_single_dpm_table(hwmgr,
 			dpm_table,
 			dep_gfx_table);
+	if (hwmgr->platform_descriptor.overdriveLimit.engineClock == 0)
+		hwmgr->platform_descriptor.overdriveLimit.engineClock =
+					dpm_table->dpm_levels[dpm_table->count-1].value;
 	vega10_init_dpm_state(&(dpm_table->dpm_state));
 
 	/* Initialize Mclk DPM table based on allow Mclk values */
@@ -1288,6 +1320,10 @@ static int vega10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 	vega10_setup_default_single_dpm_table(hwmgr,
 			dpm_table,
 			dep_mclk_table);
+	if (hwmgr->platform_descriptor.overdriveLimit.memoryClock == 0)
+		hwmgr->platform_descriptor.overdriveLimit.memoryClock =
+					dpm_table->dpm_levels[dpm_table->count-1].value;
+
 	vega10_init_dpm_state(&(dpm_table->dpm_state));
 
 	data->dpm_table.eclk_table.count = 0;
@@ -1370,48 +1406,6 @@ static int vega10_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 	memcpy(&(data->golden_dpm_table), &(data->dpm_table),
 			sizeof(struct vega10_dpm_table));
 
-	if (PP_CAP(PHM_PlatformCaps_ODNinACSupport) ||
-	    PP_CAP(PHM_PlatformCaps_ODNinDCSupport)) {
-		data->odn_dpm_table.odn_core_clock_dpm_levels.num_of_pl =
-						data->dpm_table.gfx_table.count;
-		for (i = 0; i < data->dpm_table.gfx_table.count; i++) {
-			data->odn_dpm_table.odn_core_clock_dpm_levels.entries[i].clock =
-					data->dpm_table.gfx_table.dpm_levels[i].value;
-			data->odn_dpm_table.odn_core_clock_dpm_levels.entries[i].enabled = true;
-		}
-
-		data->odn_dpm_table.vdd_dependency_on_sclk.count =
-				dep_gfx_table->count;
-		for (i = 0; i < dep_gfx_table->count; i++) {
-			data->odn_dpm_table.vdd_dependency_on_sclk.entries[i].clk =
-					dep_gfx_table->entries[i].clk;
-			data->odn_dpm_table.vdd_dependency_on_sclk.entries[i].vddInd =
-					dep_gfx_table->entries[i].vddInd;
-			data->odn_dpm_table.vdd_dependency_on_sclk.entries[i].cks_enable =
-					dep_gfx_table->entries[i].cks_enable;
-			data->odn_dpm_table.vdd_dependency_on_sclk.entries[i].cks_voffset =
-					dep_gfx_table->entries[i].cks_voffset;
-		}
-
-		data->odn_dpm_table.odn_memory_clock_dpm_levels.num_of_pl =
-						data->dpm_table.mem_table.count;
-		for (i = 0; i < data->dpm_table.mem_table.count; i++) {
-			data->odn_dpm_table.odn_memory_clock_dpm_levels.entries[i].clock =
-					data->dpm_table.mem_table.dpm_levels[i].value;
-			data->odn_dpm_table.odn_memory_clock_dpm_levels.entries[i].enabled = true;
-		}
-
-		data->odn_dpm_table.vdd_dependency_on_mclk.count = dep_mclk_table->count;
-		for (i = 0; i < dep_mclk_table->count; i++) {
-			data->odn_dpm_table.vdd_dependency_on_mclk.entries[i].clk =
-					dep_mclk_table->entries[i].clk;
-			data->odn_dpm_table.vdd_dependency_on_mclk.entries[i].vddInd =
-					dep_mclk_table->entries[i].vddInd;
-			data->odn_dpm_table.vdd_dependency_on_mclk.entries[i].vddci =
-					dep_mclk_table->entries[i].vddci;
-		}
-	}
-
 	return 0;
 }
 
@@ -1514,18 +1508,18 @@ static int vega10_populate_single_gfx_level(struct pp_hwmgr *hwmgr,
 {
 	struct phm_ppt_v2_information *table_info =
 			(struct phm_ppt_v2_information *)(hwmgr->pptable);
-	struct phm_ppt_v1_clock_voltage_dependency_table *dep_on_sclk =
-			table_info->vdd_dep_on_sclk;
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_on_sclk;
 	struct vega10_hwmgr *data = hwmgr->backend;
 	struct pp_atomfwctrl_clock_dividers_soc15 dividers;
 	uint32_t gfx_max_clock =
 			hwmgr->platform_descriptor.overdriveLimit.engineClock;
 	uint32_t i = 0;
 
-	if (data->apply_overdrive_next_settings_mask &
-			DPMTABLE_OD_UPDATE_VDDC)
+	if (hwmgr->od_enabled)
 		dep_on_sclk = (struct phm_ppt_v1_clock_voltage_dependency_table *)
-						&(data->odn_dpm_table.vdd_dependency_on_sclk);
+						&(data->odn_dpm_table.vdd_dep_on_sclk);
+	else
+		dep_on_sclk = table_info->vdd_dep_on_sclk;
 
 	PP_ASSERT_WITH_CODE(dep_on_sclk,
 			"Invalid SOC_VDD-GFX_CLK Dependency Table!",
@@ -1577,23 +1571,32 @@ static int vega10_populate_single_soc_level(struct pp_hwmgr *hwmgr,
 		uint32_t soc_clock, uint8_t *current_soc_did,
 		uint8_t *current_vol_index)
 {
+	struct vega10_hwmgr *data = hwmgr->backend;
 	struct phm_ppt_v2_information *table_info =
 			(struct phm_ppt_v2_information *)(hwmgr->pptable);
-	struct phm_ppt_v1_clock_voltage_dependency_table *dep_on_soc =
-			table_info->vdd_dep_on_socclk;
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_on_soc;
 	struct pp_atomfwctrl_clock_dividers_soc15 dividers;
 	uint32_t i;
 
-	PP_ASSERT_WITH_CODE(dep_on_soc,
-			"Invalid SOC_VDD-SOC_CLK Dependency Table!",
-			return -EINVAL);
-	for (i = 0; i < dep_on_soc->count; i++) {
-		if (dep_on_soc->entries[i].clk == soc_clock)
-			break;
+	if (hwmgr->od_enabled) {
+		dep_on_soc = (struct phm_ppt_v1_clock_voltage_dependency_table *)
+						&data->odn_dpm_table.vdd_dep_on_socclk;
+		for (i = 0; i < dep_on_soc->count; i++) {
+			if (dep_on_soc->entries[i].clk >= soc_clock)
+				break;
+		}
+	} else {
+		dep_on_soc = table_info->vdd_dep_on_socclk;
+		for (i = 0; i < dep_on_soc->count; i++) {
+			if (dep_on_soc->entries[i].clk == soc_clock)
+				break;
+		}
 	}
+
 	PP_ASSERT_WITH_CODE(dep_on_soc->count > i,
 			"Cannot find SOC_CLK in SOC_VDD-SOC_CLK Dependency Table",
 			return -EINVAL);
+
 	PP_ASSERT_WITH_CODE(!pp_atomfwctrl_get_gpu_pll_dividers_vega10(hwmgr,
 			COMPUTE_GPUCLK_INPUT_FLAG_DEFAULT_GPUCLK,
 			soc_clock, &dividers),
@@ -1602,22 +1605,6 @@ static int vega10_populate_single_soc_level(struct pp_hwmgr *hwmgr,
 
 	*current_soc_did = (uint8_t)dividers.ulDid;
 	*current_vol_index = (uint8_t)(dep_on_soc->entries[i].vddInd);
-
-	return 0;
-}
-
-uint16_t vega10_locate_vddc_given_clock(struct pp_hwmgr *hwmgr,
-		uint32_t clk,
-		struct phm_ppt_v1_clock_voltage_dependency_table *dep_table)
-{
-	uint16_t i;
-
-	for (i = 0; i < dep_table->count; i++) {
-		if (dep_table->entries[i].clk == clk)
-			return dep_table->entries[i].vddc;
-	}
-
-	pr_info("[LocateVddcGivenClock] Cannot locate SOC Vddc for this clock!");
 	return 0;
 }
 
@@ -1631,8 +1618,6 @@ static int vega10_populate_all_graphic_levels(struct pp_hwmgr *hwmgr)
 	struct vega10_hwmgr *data = hwmgr->backend;
 	struct phm_ppt_v2_information *table_info =
 			(struct phm_ppt_v2_information *)(hwmgr->pptable);
-	struct phm_ppt_v1_clock_voltage_dependency_table *dep_table =
-			table_info->vdd_dep_on_socclk;
 	PPTable_t *pp_table = &(data->smc_state_table.pp_table);
 	struct vega10_single_dpm_table *dpm_table = &(data->dpm_table.gfx_table);
 	int result = 0;
@@ -1663,11 +1648,6 @@ static int vega10_populate_all_graphic_levels(struct pp_hwmgr *hwmgr)
 
 	dpm_table = &(data->dpm_table.soc_table);
 	for (i = 0; i < dpm_table->count; i++) {
-		pp_table->SocVid[i] =
-				(uint8_t)convert_to_vid(
-				vega10_locate_vddc_given_clock(hwmgr,
-						dpm_table->dpm_levels[i].value,
-						dep_table));
 		result = vega10_populate_single_soc_level(hwmgr,
 				dpm_table->dpm_levels[i].value,
 				&(pp_table->SocclkDid[i]),
@@ -1678,7 +1658,6 @@ static int vega10_populate_all_graphic_levels(struct pp_hwmgr *hwmgr)
 
 	j = i - 1;
 	while (i < NUM_SOCCLK_DPM_LEVELS) {
-		pp_table->SocVid[i] = pp_table->SocVid[j];
 		result = vega10_populate_single_soc_level(hwmgr,
 				dpm_table->dpm_levels[j].value,
 				&(pp_table->SocclkDid[i]),
@@ -1691,6 +1670,32 @@ static int vega10_populate_all_graphic_levels(struct pp_hwmgr *hwmgr)
 	return result;
 }
 
+static void vega10_populate_vddc_soc_levels(struct pp_hwmgr *hwmgr)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+	PPTable_t *pp_table = &(data->smc_state_table.pp_table);
+	struct phm_ppt_v2_information *table_info = hwmgr->pptable;
+	struct phm_ppt_v1_voltage_lookup_table *vddc_lookup_table;
+
+	uint8_t soc_vid = 0;
+	uint32_t i, max_vddc_level;
+
+	if (hwmgr->od_enabled)
+		vddc_lookup_table = (struct phm_ppt_v1_voltage_lookup_table *)&data->odn_dpm_table.vddc_lookup_table;
+	else
+		vddc_lookup_table = table_info->vddc_lookup_table;
+
+	max_vddc_level = vddc_lookup_table->count;
+	for (i = 0; i < max_vddc_level; i++) {
+		soc_vid = (uint8_t)convert_to_vid(vddc_lookup_table->entries[i].us_vdd);
+		pp_table->SocVid[i] = soc_vid;
+	}
+	while (i < MAX_REGULAR_DPM_NUMBER) {
+		pp_table->SocVid[i] = soc_vid;
+		i++;
+	}
+}
+
 /**
  * @brief Populates single SMC GFXCLK structure using the provided clock.
  *
@@ -1705,25 +1710,25 @@ static int vega10_populate_single_memory_level(struct pp_hwmgr *hwmgr,
 	struct vega10_hwmgr *data = hwmgr->backend;
 	struct phm_ppt_v2_information *table_info =
 			(struct phm_ppt_v2_information *)(hwmgr->pptable);
-	struct phm_ppt_v1_clock_voltage_dependency_table *dep_on_mclk =
-			table_info->vdd_dep_on_mclk;
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_on_mclk;
 	struct pp_atomfwctrl_clock_dividers_soc15 dividers;
 	uint32_t mem_max_clock =
 			hwmgr->platform_descriptor.overdriveLimit.memoryClock;
 	uint32_t i = 0;
 
-	if (data->apply_overdrive_next_settings_mask &
-			DPMTABLE_OD_UPDATE_VDDC)
+	if (hwmgr->od_enabled)
 		dep_on_mclk = (struct phm_ppt_v1_clock_voltage_dependency_table *)
-					&data->odn_dpm_table.vdd_dependency_on_mclk;
+					&data->odn_dpm_table.vdd_dep_on_mclk;
+	else
+		dep_on_mclk = table_info->vdd_dep_on_mclk;
 
 	PP_ASSERT_WITH_CODE(dep_on_mclk,
 			"Invalid SOC_VDD-UCLK Dependency Table!",
 			return -EINVAL);
 
-	if (data->need_update_dpm_table & DPMTABLE_OD_UPDATE_MCLK)
+	if (data->need_update_dpm_table & DPMTABLE_OD_UPDATE_MCLK) {
 		mem_clock = mem_clock > mem_max_clock ? mem_max_clock : mem_clock;
-	else {
+	} else {
 		for (i = 0; i < dep_on_mclk->count; i++) {
 			if (dep_on_mclk->entries[i].clk == mem_clock)
 				break;
@@ -2067,6 +2072,9 @@ static int vega10_populate_avfs_parameters(struct pp_hwmgr *hwmgr)
 	if (data->smu_features[GNLD_AVFS].supported) {
 		result = pp_atomfwctrl_get_avfs_information(hwmgr, &avfs_params);
 		if (!result) {
+			data->odn_dpm_table.max_vddc = avfs_params.ulMaxVddc;
+			data->odn_dpm_table.min_vddc = avfs_params.ulMinVddc;
+
 			pp_table->MinVoltageVid = (uint8_t)
 					convert_to_vid((uint16_t)(avfs_params.ulMinVddc));
 			pp_table->MaxVoltageVid = (uint8_t)
@@ -2345,6 +2353,22 @@ static int vega10_avfs_enable(struct pp_hwmgr *hwmgr, bool enable)
 	return 0;
 }
 
+static int vega10_update_avfs(struct pp_hwmgr *hwmgr)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+
+	if (data->need_update_dpm_table & DPMTABLE_OD_UPDATE_VDDC) {
+		vega10_avfs_enable(hwmgr, false);
+	} else if (data->need_update_dpm_table) {
+		vega10_avfs_enable(hwmgr, false);
+		vega10_avfs_enable(hwmgr, true);
+	} else {
+		vega10_avfs_enable(hwmgr, true);
+	}
+
+	return 0;
+}
+
 static int vega10_populate_and_upload_avfs_fuse_override(struct pp_hwmgr *hwmgr)
 {
 	int result = 0;
@@ -2384,6 +2408,40 @@ static int vega10_populate_and_upload_avfs_fuse_override(struct pp_hwmgr *hwmgr)
 	return result;
 }
 
+static void vega10_check_dpm_table_updated(struct pp_hwmgr *hwmgr)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+	struct vega10_odn_dpm_table *odn_table = &(data->odn_dpm_table);
+	struct phm_ppt_v2_information *table_info = hwmgr->pptable;
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_table;
+	struct phm_ppt_v1_clock_voltage_dependency_table *odn_dep_table;
+	uint32_t i;
+
+	dep_table = table_info->vdd_dep_on_mclk;
+	odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dep_on_mclk);
+
+	for (i = 0; i < dep_table->count; i++) {
+		if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) {
+			data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_MCLK;
+			return;
+		}
+	}
+
+	dep_table = table_info->vdd_dep_on_sclk;
+	odn_dep_table = (struct phm_ppt_v1_clock_voltage_dependency_table *)&(odn_table->vdd_dep_on_sclk);
+	for (i = 0; i < dep_table->count; i++) {
+		if (dep_table->entries[i].vddc != odn_dep_table->entries[i].vddc) {
+			data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_VDDC | DPMTABLE_OD_UPDATE_SCLK;
+			return;
+		}
+	}
+
+	if (data->need_update_dpm_table & DPMTABLE_OD_UPDATE_VDDC) {
+		data->need_update_dpm_table &= ~DPMTABLE_OD_UPDATE_VDDC;
+		data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_SCLK | DPMTABLE_OD_UPDATE_MCLK;
+	}
+}
+
 /**
 * Initializes the SMC table and uploads it
 *
@@ -2400,12 +2458,23 @@ static int vega10_init_smc_table(struct pp_hwmgr *hwmgr)
 	PPTable_t *pp_table = &(data->smc_state_table.pp_table);
 	struct pp_atomfwctrl_voltage_table voltage_table;
 	struct pp_atomfwctrl_bios_boot_up_values boot_up_values;
+	struct vega10_odn_dpm_table *odn_table = &(data->odn_dpm_table);
 
 	result = vega10_setup_default_dpm_tables(hwmgr);
 	PP_ASSERT_WITH_CODE(!result,
 			"Failed to setup default DPM tables!",
 			return result);
 
+	/* initialize ODN table */
+	if (hwmgr->od_enabled) {
+		if (odn_table->max_vddc) {
+			data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_SCLK | DPMTABLE_OD_UPDATE_MCLK;
+			vega10_check_dpm_table_updated(hwmgr);
+		} else {
+			vega10_odn_initial_default_setting(hwmgr);
+		}
+	}
+
 	pp_atomfwctrl_get_voltage_table_v4(hwmgr, VOLTAGE_TYPE_VDDC,
 			VOLTAGE_OBJ_SVID2,  &voltage_table);
 	pp_table->MaxVidStep = voltage_table.max_vid_step;
@@ -2452,6 +2521,8 @@ static int vega10_init_smc_table(struct pp_hwmgr *hwmgr)
 			"Failed to initialize Memory Level!",
 			return result);
 
+	vega10_populate_vddc_soc_levels(hwmgr);
+
 	result = vega10_populate_all_display_clock_levels(hwmgr);
 	PP_ASSERT_WITH_CODE(!result,
 			"Failed to initialize Display Level!",
@@ -2481,6 +2552,12 @@ static int vega10_init_smc_table(struct pp_hwmgr *hwmgr)
 		data->vbios_boot_state.mvddc    = boot_up_values.usMvddc;
 		data->vbios_boot_state.gfx_clock = boot_up_values.ulGfxClk;
 		data->vbios_boot_state.mem_clock = boot_up_values.ulUClk;
+		pp_atomfwctrl_get_clk_information_by_clkid(hwmgr,
+				SMU9_SYSPLL0_SOCCLK_ID, &boot_up_values.ulSocClk);
+
+		pp_atomfwctrl_get_clk_information_by_clkid(hwmgr,
+				SMU9_SYSPLL0_DCEFCLK_ID, &boot_up_values.ulDCEFClk);
+
 		data->vbios_boot_state.soc_clock = boot_up_values.ulSocClk;
 		data->vbios_boot_state.dcef_clock = boot_up_values.ulDCEFClk;
 		if (0 != boot_up_values.usVddc) {
@@ -2819,17 +2896,12 @@ static int vega10_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
 
 	vega10_enable_disable_PCC_limit_feature(hwmgr, true);
 
-	if ((hwmgr->smu_version == 0x001c2c00) ||
-			(hwmgr->smu_version == 0x001c2d00))
-		smum_send_msg_to_smc_with_parameter(hwmgr,
-				PPSMC_MSG_UpdatePkgPwrPidAlpha, 1);
-
 	smum_send_msg_to_smc_with_parameter(hwmgr,
 		PPSMC_MSG_ConfigureTelemetry, data->config_telemetry);
 
 	tmp_result = vega10_construct_voltage_tables(hwmgr);
 	PP_ASSERT_WITH_CODE(!tmp_result,
-			"Failed to contruct voltage tables!",
+			"Failed to construct voltage tables!",
 			result = tmp_result);
 
 	tmp_result = vega10_init_smc_table(hwmgr);
@@ -3019,6 +3091,7 @@ static int vega10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 				struct pp_power_state  *request_ps,
 			const struct pp_power_state *current_ps)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	struct vega10_power_state *vega10_ps =
 				cast_phw_vega10_power_state(&request_ps->hardware);
 	uint32_t sclk;
@@ -3028,7 +3101,6 @@ static int vega10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 	bool disable_mclk_switching_for_frame_lock;
 	bool disable_mclk_switching_for_vr;
 	bool force_mclk_high;
-	struct cgs_display_info info = {0};
 	const struct phm_clock_and_voltage_limits *max_limits;
 	uint32_t i;
 	struct vega10_hwmgr *data = hwmgr->backend;
@@ -3045,12 +3117,12 @@ static int vega10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 	if (vega10_ps->performance_level_count != 2)
 		pr_info("VI should always have 2 performance levels");
 
-	max_limits = (PP_PowerSource_AC == hwmgr->power_source) ?
+	max_limits = adev->pm.ac_power ?
 			&(hwmgr->dyn_state.max_clock_voltage_on_ac) :
 			&(hwmgr->dyn_state.max_clock_voltage_on_dc);
 
 	/* Cap clock DPM tables at DC MAX if it is in DC. */
-	if (PP_PowerSource_DC == hwmgr->power_source) {
+	if (!adev->pm.ac_power) {
 		for (i = 0; i < vega10_ps->performance_level_count; i++) {
 			if (vega10_ps->performance_levels[i].mem_clock >
 				max_limits->mclk)
@@ -3063,11 +3135,9 @@ static int vega10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 		}
 	}
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-
 	/* result = PHM_CheckVBlankTime(hwmgr, &vblankTooShort);*/
-	minimum_clocks.engineClock = hwmgr->display_config.min_core_set_clock;
-	minimum_clocks.memoryClock = hwmgr->display_config.min_mem_set_clock;
+	minimum_clocks.engineClock = hwmgr->display_config->min_core_set_clock;
+	minimum_clocks.memoryClock = hwmgr->display_config->min_mem_set_clock;
 
 	if (PP_CAP(PHM_PlatformCaps_StablePState)) {
 		stable_pstate_sclk_dpm_percentage =
@@ -3107,10 +3177,10 @@ static int vega10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 		PP_CAP(PHM_PlatformCaps_DisableMclkSwitchForVR);
 	force_mclk_high = PP_CAP(PHM_PlatformCaps_ForceMclkHigh);
 
-	if (info.display_count == 0)
+	if (hwmgr->display_config->num_display == 0)
 		disable_mclk_switching = false;
 	else
-		disable_mclk_switching = (info.display_count > 1) ||
+		disable_mclk_switching = (hwmgr->display_config->num_display > 1) ||
 			disable_mclk_switching_for_frame_lock ||
 			disable_mclk_switching_for_vr ||
 			force_mclk_high;
@@ -3142,7 +3212,7 @@ static int vega10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 		/* Find the lowest MCLK frequency that is within
 		 * the tolerable latency defined in DAL
 		 */
-		latency = 0;
+		latency = hwmgr->display_config->dce_tolerable_mclk_in_active_latency;
 		for (i = 0; i < data->mclk_latency_table.count; i++) {
 			if ((data->mclk_latency_table.entries[i].latency <= latency) &&
 				(data->mclk_latency_table.entries[i].frequency >=
@@ -3171,87 +3241,11 @@ static int vega10_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 
 static int vega10_find_dpm_states_clocks_in_dpm_table(struct pp_hwmgr *hwmgr, const void *input)
 {
-	const struct phm_set_power_state_input *states =
-			(const struct phm_set_power_state_input *)input;
-	const struct vega10_power_state *vega10_ps =
-			cast_const_phw_vega10_power_state(states->pnew_state);
 	struct vega10_hwmgr *data = hwmgr->backend;
-	struct vega10_single_dpm_table *sclk_table =
-			&(data->dpm_table.gfx_table);
-	uint32_t sclk = vega10_ps->performance_levels
-			[vega10_ps->performance_level_count - 1].gfx_clock;
-	struct vega10_single_dpm_table *mclk_table =
-			&(data->dpm_table.mem_table);
-	uint32_t mclk = vega10_ps->performance_levels
-			[vega10_ps->performance_level_count - 1].mem_clock;
-	struct PP_Clocks min_clocks = {0};
-	uint32_t i;
-	struct cgs_display_info info = {0};
-
-	data->need_update_dpm_table = 0;
-
-	if (PP_CAP(PHM_PlatformCaps_ODNinACSupport) ||
-	    PP_CAP(PHM_PlatformCaps_ODNinDCSupport)) {
-		for (i = 0; i < sclk_table->count; i++) {
-			if (sclk == sclk_table->dpm_levels[i].value)
-				break;
-		}
-
-		if (!(data->apply_overdrive_next_settings_mask &
-				DPMTABLE_OD_UPDATE_SCLK) && i >= sclk_table->count) {
-			/* Check SCLK in DAL's minimum clocks
-			 * in case DeepSleep divider update is required.
-			 */
-			if (data->display_timing.min_clock_in_sr !=
-					min_clocks.engineClockInSR &&
-				(min_clocks.engineClockInSR >=
-						VEGA10_MINIMUM_ENGINE_CLOCK ||
-					data->display_timing.min_clock_in_sr >=
-						VEGA10_MINIMUM_ENGINE_CLOCK))
-				data->need_update_dpm_table |= DPMTABLE_UPDATE_SCLK;
-		}
-
-		cgs_get_active_displays_info(hwmgr->device, &info);
-
-		if (data->display_timing.num_existing_displays !=
-				info.display_count)
-			data->need_update_dpm_table |= DPMTABLE_UPDATE_MCLK;
-	} else {
-		for (i = 0; i < sclk_table->count; i++) {
-			if (sclk == sclk_table->dpm_levels[i].value)
-				break;
-		}
-
-		if (i >= sclk_table->count)
-			data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_SCLK;
-		else {
-			/* Check SCLK in DAL's minimum clocks
-			 * in case DeepSleep divider update is required.
-			 */
-			if (data->display_timing.min_clock_in_sr !=
-					min_clocks.engineClockInSR &&
-				(min_clocks.engineClockInSR >=
-						VEGA10_MINIMUM_ENGINE_CLOCK ||
-					data->display_timing.min_clock_in_sr >=
-						VEGA10_MINIMUM_ENGINE_CLOCK))
-				data->need_update_dpm_table |= DPMTABLE_UPDATE_SCLK;
-		}
-
-		for (i = 0; i < mclk_table->count; i++) {
-			if (mclk == mclk_table->dpm_levels[i].value)
-				break;
-		}
 
-		cgs_get_active_displays_info(hwmgr->device, &info);
+	if (data->display_timing.num_existing_displays != hwmgr->display_config->num_display)
+		data->need_update_dpm_table |= DPMTABLE_UPDATE_MCLK;
 
-		if (i >= mclk_table->count)
-			data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_MCLK;
-
-		if (data->display_timing.num_existing_displays !=
-				info.display_count ||
-				i >= mclk_table->count)
-			data->need_update_dpm_table |= DPMTABLE_UPDATE_MCLK;
-	}
 	return 0;
 }
 
@@ -3259,194 +3253,29 @@ static int vega10_populate_and_upload_sclk_mclk_dpm_levels(
 		struct pp_hwmgr *hwmgr, const void *input)
 {
 	int result = 0;
-	const struct phm_set_power_state_input *states =
-			(const struct phm_set_power_state_input *)input;
-	const struct vega10_power_state *vega10_ps =
-			cast_const_phw_vega10_power_state(states->pnew_state);
 	struct vega10_hwmgr *data = hwmgr->backend;
-	uint32_t sclk = vega10_ps->performance_levels
-			[vega10_ps->performance_level_count - 1].gfx_clock;
-	uint32_t mclk = vega10_ps->performance_levels
-			[vega10_ps->performance_level_count - 1].mem_clock;
-	struct vega10_dpm_table *dpm_table = &data->dpm_table;
-	struct vega10_dpm_table *golden_dpm_table =
-			&data->golden_dpm_table;
-	uint32_t dpm_count, clock_percent;
-	uint32_t i;
-
-	if (PP_CAP(PHM_PlatformCaps_ODNinACSupport) ||
-	    PP_CAP(PHM_PlatformCaps_ODNinDCSupport)) {
-
-		if (!data->need_update_dpm_table &&
-			!data->apply_optimized_settings &&
-			!data->apply_overdrive_next_settings_mask)
-			return 0;
-
-		if (data->apply_overdrive_next_settings_mask &
-				DPMTABLE_OD_UPDATE_SCLK) {
-			for (dpm_count = 0;
-					dpm_count < dpm_table->gfx_table.count;
-					dpm_count++) {
-				dpm_table->gfx_table.dpm_levels[dpm_count].enabled =
-					data->odn_dpm_table.odn_core_clock_dpm_levels.entries[dpm_count].enabled;
-				dpm_table->gfx_table.dpm_levels[dpm_count].value =
-					data->odn_dpm_table.odn_core_clock_dpm_levels.entries[dpm_count].clock;
-			}
-		}
-
-		if (data->apply_overdrive_next_settings_mask &
-				DPMTABLE_OD_UPDATE_MCLK) {
-			for (dpm_count = 0;
-					dpm_count < dpm_table->mem_table.count;
-					dpm_count++) {
-				dpm_table->mem_table.dpm_levels[dpm_count].enabled =
-					data->odn_dpm_table.odn_memory_clock_dpm_levels.entries[dpm_count].enabled;
-				dpm_table->mem_table.dpm_levels[dpm_count].value =
-					data->odn_dpm_table.odn_memory_clock_dpm_levels.entries[dpm_count].clock;
-			}
-		}
-
-		if ((data->need_update_dpm_table & DPMTABLE_UPDATE_SCLK) ||
-			data->apply_optimized_settings ||
-			(data->apply_overdrive_next_settings_mask &
-					DPMTABLE_OD_UPDATE_SCLK)) {
-			result = vega10_populate_all_graphic_levels(hwmgr);
-			PP_ASSERT_WITH_CODE(!result,
-					"Failed to populate SCLK during PopulateNewDPMClocksStates Function!",
-					return result);
-		}
-
-		if ((data->need_update_dpm_table & DPMTABLE_UPDATE_MCLK) ||
-			(data->apply_overdrive_next_settings_mask &
-					DPMTABLE_OD_UPDATE_MCLK)){
-			result = vega10_populate_all_memory_levels(hwmgr);
-			PP_ASSERT_WITH_CODE(!result,
-					"Failed to populate MCLK during PopulateNewDPMClocksStates Function!",
-					return result);
-		}
-	} else {
-		if (!data->need_update_dpm_table &&
-				!data->apply_optimized_settings)
-			return 0;
 
-		if (data->need_update_dpm_table & DPMTABLE_OD_UPDATE_SCLK &&
-				data->smu_features[GNLD_DPM_GFXCLK].supported) {
-				dpm_table->
-				gfx_table.dpm_levels[dpm_table->gfx_table.count - 1].
-				value = sclk;
-				if (hwmgr->od_enabled) {
-					/* Need to do calculation based on the golden DPM table
-					 * as the Heatmap GPU Clock axis is also based on
-					 * the default values
-					 */
-					PP_ASSERT_WITH_CODE(
-							golden_dpm_table->gfx_table.dpm_levels
-							[golden_dpm_table->gfx_table.count - 1].value,
-							"Divide by 0!",
-							return -1);
-
-					dpm_count = dpm_table->gfx_table.count < 2 ?
-							0 : dpm_table->gfx_table.count - 2;
-					for (i = dpm_count; i > 1; i--) {
-						if (sclk > golden_dpm_table->gfx_table.dpm_levels
-							[golden_dpm_table->gfx_table.count - 1].value) {
-							clock_percent =
-								((sclk - golden_dpm_table->gfx_table.dpm_levels
-								[golden_dpm_table->gfx_table.count - 1].value) *
-								100) /
-								golden_dpm_table->gfx_table.dpm_levels
-								[golden_dpm_table->gfx_table.count - 1].value;
-
-							dpm_table->gfx_table.dpm_levels[i].value =
-								golden_dpm_table->gfx_table.dpm_levels[i].value +
-								(golden_dpm_table->gfx_table.dpm_levels[i].value *
-								clock_percent) / 100;
-						} else if (golden_dpm_table->
-								gfx_table.dpm_levels[dpm_table->gfx_table.count-1].value >
-								sclk) {
-							clock_percent =
-								((golden_dpm_table->gfx_table.dpm_levels
-								[golden_dpm_table->gfx_table.count - 1].value -
-								sclk) *	100) /
-								golden_dpm_table->gfx_table.dpm_levels
-								[golden_dpm_table->gfx_table.count-1].value;
-
-							dpm_table->gfx_table.dpm_levels[i].value =
-								golden_dpm_table->gfx_table.dpm_levels[i].value -
-								(golden_dpm_table->gfx_table.dpm_levels[i].value *
-								clock_percent) / 100;
-						} else
-							dpm_table->gfx_table.dpm_levels[i].value =
-								golden_dpm_table->gfx_table.dpm_levels[i].value;
-					}
-				}
-			}
-
-		if (data->need_update_dpm_table & DPMTABLE_OD_UPDATE_MCLK &&
-				data->smu_features[GNLD_DPM_UCLK].supported) {
-			dpm_table->
-			mem_table.dpm_levels[dpm_table->mem_table.count - 1].
-			value = mclk;
+	if (!data->need_update_dpm_table)
+		return 0;
 
-			if (hwmgr->od_enabled) {
-				PP_ASSERT_WITH_CODE(
-					golden_dpm_table->mem_table.dpm_levels
-					[golden_dpm_table->mem_table.count - 1].value,
-					"Divide by 0!",
-					return -1);
+	if (data->need_update_dpm_table &
+			(DPMTABLE_OD_UPDATE_SCLK + DPMTABLE_UPDATE_SCLK + DPMTABLE_UPDATE_SOCCLK)) {
+		result = vega10_populate_all_graphic_levels(hwmgr);
+		PP_ASSERT_WITH_CODE((0 == result),
+				"Failed to populate SCLK during PopulateNewDPMClocksStates Function!",
+				return result);
+	}
 
-				dpm_count = dpm_table->mem_table.count < 2 ?
-						0 : dpm_table->mem_table.count - 2;
-				for (i = dpm_count; i > 1; i--) {
-					if (mclk > golden_dpm_table->mem_table.dpm_levels
-						[golden_dpm_table->mem_table.count-1].value) {
-						clock_percent = ((mclk -
-							golden_dpm_table->mem_table.dpm_levels
-							[golden_dpm_table->mem_table.count-1].value) *
-							100) /
-							golden_dpm_table->mem_table.dpm_levels
-							[golden_dpm_table->mem_table.count-1].value;
-
-						dpm_table->mem_table.dpm_levels[i].value =
-							golden_dpm_table->mem_table.dpm_levels[i].value +
-							(golden_dpm_table->mem_table.dpm_levels[i].value *
-							clock_percent) / 100;
-					} else if (golden_dpm_table->mem_table.dpm_levels
-							[dpm_table->mem_table.count-1].value > mclk) {
-						clock_percent = ((golden_dpm_table->mem_table.dpm_levels
-							[golden_dpm_table->mem_table.count-1].value - mclk) *
-							100) /
-							golden_dpm_table->mem_table.dpm_levels
-							[golden_dpm_table->mem_table.count-1].value;
-
-						dpm_table->mem_table.dpm_levels[i].value =
-							golden_dpm_table->mem_table.dpm_levels[i].value -
-							(golden_dpm_table->mem_table.dpm_levels[i].value *
-							clock_percent) / 100;
-					} else
-						dpm_table->mem_table.dpm_levels[i].value =
-							golden_dpm_table->mem_table.dpm_levels[i].value;
-				}
-			}
-		}
+	if (data->need_update_dpm_table &
+			(DPMTABLE_OD_UPDATE_MCLK + DPMTABLE_UPDATE_MCLK)) {
+		result = vega10_populate_all_memory_levels(hwmgr);
+		PP_ASSERT_WITH_CODE((0 == result),
+				"Failed to populate MCLK during PopulateNewDPMClocksStates Function!",
+				return result);
+	}
 
-		if ((data->need_update_dpm_table &
-			(DPMTABLE_OD_UPDATE_SCLK + DPMTABLE_UPDATE_SCLK)) ||
-			data->apply_optimized_settings) {
-			result = vega10_populate_all_graphic_levels(hwmgr);
-			PP_ASSERT_WITH_CODE(!result,
-					"Failed to populate SCLK during PopulateNewDPMClocksStates Function!",
-					return result);
-		}
+	vega10_populate_vddc_soc_levels(hwmgr);
 
-		if (data->need_update_dpm_table &
-				(DPMTABLE_OD_UPDATE_MCLK + DPMTABLE_UPDATE_MCLK)) {
-			result = vega10_populate_all_memory_levels(hwmgr);
-			PP_ASSERT_WITH_CODE(!result,
-					"Failed to populate MCLK during PopulateNewDPMClocksStates Function!",
-					return result);
-		}
-	}
 	return result;
 }
 
@@ -3742,8 +3571,9 @@ static int vega10_set_power_state_tasks(struct pp_hwmgr *hwmgr,
 	PP_ASSERT_WITH_CODE(!result,
 			"Failed to upload PPtable!", return result);
 
-	data->apply_optimized_settings = false;
-	data->apply_overdrive_next_settings_mask = 0;
+	vega10_update_avfs(hwmgr);
+
+	data->need_update_dpm_table &= DPMTABLE_OD_UPDATE_VDDC;
 
 	return 0;
 }
@@ -3793,16 +3623,18 @@ static uint32_t vega10_dpm_get_mclk(struct pp_hwmgr *hwmgr, bool low)
 }
 
 static int vega10_get_gpu_power(struct pp_hwmgr *hwmgr,
-		struct pp_gpu_power *query)
+		uint32_t *query)
 {
 	uint32_t value;
 
+	if (!query)
+		return -EINVAL;
+
 	smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetCurrPkgPwr);
 	value = smum_get_argument(hwmgr);
 
-	/* power value is an integer */
-	memset(query, 0, sizeof *query);
-	query->average_gpu_power = value << 8;
+	/* SMC returning actual watts, keep consistent with legacy asics, low 8 bit as 8 fractional bits */
+	*query = value << 8;
 
 	return 0;
 }
@@ -3810,22 +3642,18 @@ static int vega10_get_gpu_power(struct pp_hwmgr *hwmgr,
 static int vega10_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 			      void *value, int *size)
 {
-	uint32_t sclk_idx, mclk_idx, activity_percent = 0;
+	struct amdgpu_device *adev = hwmgr->adev;
+	uint32_t sclk_mhz, mclk_idx, activity_percent = 0;
 	struct vega10_hwmgr *data = hwmgr->backend;
 	struct vega10_dpm_table *dpm_table = &data->dpm_table;
 	int ret = 0;
-	uint32_t reg, val_vid;
+	uint32_t val_vid;
 
 	switch (idx) {
 	case AMDGPU_PP_SENSOR_GFX_SCLK:
-		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetCurrentGfxclkIndex);
-		sclk_idx = smum_get_argument(hwmgr);
-		if (sclk_idx <  dpm_table->gfx_table.count) {
-			*((uint32_t *)value) = dpm_table->gfx_table.dpm_levels[sclk_idx].value;
-			*size = 4;
-		} else {
-			ret = -EINVAL;
-		}
+		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetAverageGfxclkActualFrequency);
+		sclk_mhz = smum_get_argument(hwmgr);
+		*((uint32_t *)value) = sclk_mhz * 100;
 		break;
 	case AMDGPU_PP_SENSOR_GFX_MCLK:
 		smum_send_msg_to_smc(hwmgr, PPSMC_MSG_GetCurrentUclkIndex);
@@ -3856,18 +3684,10 @@ static int vega10_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 		*size = 4;
 		break;
 	case AMDGPU_PP_SENSOR_GPU_POWER:
-		if (*size < sizeof(struct pp_gpu_power))
-			ret = -EINVAL;
-		else {
-			*size = sizeof(struct pp_gpu_power);
-			ret = vega10_get_gpu_power(hwmgr, (struct pp_gpu_power *)value);
-		}
+		ret = vega10_get_gpu_power(hwmgr, (uint32_t *)value);
 		break;
 	case AMDGPU_PP_SENSOR_VDDGFX:
-		reg = soc15_get_register_offset(SMUIO_HWID, 0,
-			mmSMUSVI0_PLANE0_CURRENTVID_BASE_IDX,
-			mmSMUSVI0_PLANE0_CURRENTVID);
-		val_vid = (cgs_read_register(hwmgr->device, reg) &
+		val_vid = (RREG32_SOC15(SMUIO, 0, mmSMUSVI0_PLANE0_CURRENTVID) &
 			SMUSVI0_PLANE0_CURRENTVID__CURRENT_SVI0_PLANE0_VID_MASK) >>
 			SMUSVI0_PLANE0_CURRENTVID__CURRENT_SVI0_PLANE0_VID__SHIFT;
 		*((uint32_t *)value) = (uint32_t)convert_to_vddc((uint8_t)val_vid);
@@ -3956,26 +3776,18 @@ static int vega10_notify_smc_display_config_after_ps_adjustment(
 			(struct phm_ppt_v2_information *)hwmgr->pptable;
 	struct phm_ppt_v1_clock_voltage_dependency_table *mclk_table = table_info->vdd_dep_on_mclk;
 	uint32_t idx;
-	uint32_t num_active_disps = 0;
-	struct cgs_display_info info = {0};
 	struct PP_Clocks min_clocks = {0};
 	uint32_t i;
 	struct pp_display_clock_request clock_req;
 
-	info.mode_info = NULL;
-
-	cgs_get_active_displays_info(hwmgr->device, &info);
-
-	num_active_disps = info.display_count;
-
-	if (num_active_disps > 1)
+	if (hwmgr->display_config->num_display > 1)
 		vega10_notify_smc_display_change(hwmgr, false);
 	else
 		vega10_notify_smc_display_change(hwmgr, true);
 
-	min_clocks.dcefClock = hwmgr->display_config.min_dcef_set_clk;
-	min_clocks.dcefClockInSR = hwmgr->display_config.min_dcef_deep_sleep_set_clk;
-	min_clocks.memoryClock = hwmgr->display_config.min_mem_set_clock;
+	min_clocks.dcefClock = hwmgr->display_config->min_dcef_set_clk;
+	min_clocks.dcefClockInSR = hwmgr->display_config->min_dcef_deep_sleep_set_clk;
+	min_clocks.memoryClock = hwmgr->display_config->min_mem_set_clock;
 
 	for (i = 0; i < dpm_table->count; i++) {
 		if (dpm_table->dpm_levels[i].value == min_clocks.dcefClock)
@@ -3984,7 +3796,7 @@ static int vega10_notify_smc_display_config_after_ps_adjustment(
 
 	if (i < dpm_table->count) {
 		clock_req.clock_type = amd_pp_dcef_clock;
-		clock_req.clock_freq_in_khz = dpm_table->dpm_levels[i].value;
+		clock_req.clock_freq_in_khz = dpm_table->dpm_levels[i].value * 10;
 		if (!vega10_display_clock_voltage_request(hwmgr, &clock_req)) {
 			smum_send_msg_to_smc_with_parameter(
 					hwmgr, PPSMC_MSG_SetMinDeepSleepDcefclk,
@@ -4120,6 +3932,47 @@ static void vega10_set_fan_control_mode(struct pp_hwmgr *hwmgr, uint32_t mode)
 	}
 }
 
+static int vega10_force_clock_level(struct pp_hwmgr *hwmgr,
+		enum pp_clock_type type, uint32_t mask)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+
+	switch (type) {
+	case PP_SCLK:
+		data->smc_state_table.gfx_boot_level = mask ? (ffs(mask) - 1) : 0;
+		data->smc_state_table.gfx_max_level = mask ? (fls(mask) - 1) : 0;
+
+		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_bootup_level(hwmgr),
+			"Failed to upload boot level to lowest!",
+			return -EINVAL);
+
+		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_max_level(hwmgr),
+			"Failed to upload dpm max level to highest!",
+			return -EINVAL);
+		break;
+
+	case PP_MCLK:
+		data->smc_state_table.mem_boot_level = mask ? (ffs(mask) - 1) : 0;
+		data->smc_state_table.mem_max_level = mask ? (fls(mask) - 1) : 0;
+
+		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_bootup_level(hwmgr),
+			"Failed to upload boot level to lowest!",
+			return -EINVAL);
+
+		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_max_level(hwmgr),
+			"Failed to upload dpm max level to highest!",
+			return -EINVAL);
+
+		break;
+
+	case PP_PCIE:
+	default:
+		break;
+	}
+
+	return 0;
+}
+
 static int vega10_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 				enum amd_dpm_forced_level level)
 {
@@ -4200,28 +4053,17 @@ static void vega10_get_sclks(struct pp_hwmgr *hwmgr,
 			table_info->vdd_dep_on_sclk;
 	uint32_t i;
 
+	clocks->num_levels = 0;
 	for (i = 0; i < dep_table->count; i++) {
 		if (dep_table->entries[i].clk) {
 			clocks->data[clocks->num_levels].clocks_in_khz =
-					dep_table->entries[i].clk;
+					dep_table->entries[i].clk * 10;
 			clocks->num_levels++;
 		}
 	}
 
 }
 
-static uint32_t vega10_get_mem_latency(struct pp_hwmgr *hwmgr,
-		uint32_t clock)
-{
-	if (clock >= MEM_FREQ_LOW_LATENCY &&
-			clock < MEM_FREQ_HIGH_LATENCY)
-		return MEM_LATENCY_HIGH;
-	else if (clock >= MEM_FREQ_HIGH_LATENCY)
-		return MEM_LATENCY_LOW;
-	else
-		return MEM_LATENCY_ERR;
-}
-
 static void vega10_get_memclocks(struct pp_hwmgr *hwmgr,
 		struct pp_clock_levels_with_latency *clocks)
 {
@@ -4230,26 +4072,22 @@ static void vega10_get_memclocks(struct pp_hwmgr *hwmgr,
 	struct phm_ppt_v1_clock_voltage_dependency_table *dep_table =
 			table_info->vdd_dep_on_mclk;
 	struct vega10_hwmgr *data = hwmgr->backend;
+	uint32_t j = 0;
 	uint32_t i;
 
-	clocks->num_levels = 0;
-	data->mclk_latency_table.count = 0;
-
 	for (i = 0; i < dep_table->count; i++) {
 		if (dep_table->entries[i].clk) {
-			clocks->data[clocks->num_levels].clocks_in_khz =
-			data->mclk_latency_table.entries
-			[data->mclk_latency_table.count].frequency =
-					dep_table->entries[i].clk;
-			clocks->data[clocks->num_levels].latency_in_us =
-			data->mclk_latency_table.entries
-			[data->mclk_latency_table.count].latency =
-					vega10_get_mem_latency(hwmgr,
-						dep_table->entries[i].clk);
-			clocks->num_levels++;
-			data->mclk_latency_table.count++;
+
+			clocks->data[j].clocks_in_khz =
+						dep_table->entries[i].clk * 10;
+			data->mclk_latency_table.entries[j].frequency =
+							dep_table->entries[i].clk;
+			clocks->data[j].latency_in_us =
+				data->mclk_latency_table.entries[j].latency = 25;
+			j++;
 		}
 	}
+	clocks->num_levels = data->mclk_latency_table.count = j;
 }
 
 static void vega10_get_dcefclocks(struct pp_hwmgr *hwmgr,
@@ -4262,7 +4100,7 @@ static void vega10_get_dcefclocks(struct pp_hwmgr *hwmgr,
 	uint32_t i;
 
 	for (i = 0; i < dep_table->count; i++) {
-		clocks->data[i].clocks_in_khz = dep_table->entries[i].clk;
+		clocks->data[i].clocks_in_khz = dep_table->entries[i].clk * 10;
 		clocks->data[i].latency_in_us = 0;
 		clocks->num_levels++;
 	}
@@ -4278,7 +4116,7 @@ static void vega10_get_socclocks(struct pp_hwmgr *hwmgr,
 	uint32_t i;
 
 	for (i = 0; i < dep_table->count; i++) {
-		clocks->data[i].clocks_in_khz = dep_table->entries[i].clk;
+		clocks->data[i].clocks_in_khz = dep_table->entries[i].clk * 10;
 		clocks->data[i].latency_in_us = 0;
 		clocks->num_levels++;
 	}
@@ -4338,7 +4176,7 @@ static int vega10_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr,
 	}
 
 	for (i = 0; i < dep_table->count; i++) {
-		clocks->data[i].clocks_in_khz = dep_table->entries[i].clk;
+		clocks->data[i].clocks_in_khz = dep_table->entries[i].clk  * 10;
 		clocks->data[i].voltage_in_mv = (uint32_t)(table_info->vddc_lookup_table->
 				entries[dep_table->entries[i].vddInd].us_vdd);
 		clocks->num_levels++;
@@ -4351,102 +4189,21 @@ static int vega10_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr,
 }
 
 static int vega10_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr,
-		struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges)
+							void *clock_range)
 {
 	struct vega10_hwmgr *data = hwmgr->backend;
+	struct dm_pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges = clock_range;
 	Watermarks_t *table = &(data->smc_state_table.water_marks_table);
 	int result = 0;
-	uint32_t i;
 
 	if (!data->registry_data.disable_water_mark) {
-		for (i = 0; i < wm_with_clock_ranges->num_wm_sets_dmif; i++) {
-			table->WatermarkRow[WM_DCEFCLK][i].MinClock =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_dmif[i].wm_min_dcefclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_DCEFCLK][i].MaxClock =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_dmif[i].wm_max_dcefclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_DCEFCLK][i].MinUclk =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_dmif[i].wm_min_memclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_DCEFCLK][i].MaxUclk =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_dmif[i].wm_max_memclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_DCEFCLK][i].WmSetting = (uint8_t)
-					wm_with_clock_ranges->wm_sets_dmif[i].wm_set_id;
-		}
-
-		for (i = 0; i < wm_with_clock_ranges->num_wm_sets_mcif; i++) {
-			table->WatermarkRow[WM_SOCCLK][i].MinClock =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_mcif[i].wm_min_socclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_SOCCLK][i].MaxClock =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_mcif[i].wm_max_socclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_SOCCLK][i].MinUclk =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_mcif[i].wm_min_memclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_SOCCLK][i].MaxUclk =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_mcif[i].wm_max_memclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_SOCCLK][i].WmSetting = (uint8_t)
-					wm_with_clock_ranges->wm_sets_mcif[i].wm_set_id;
-		}
+		smu_set_watermarks_for_clocks_ranges(table, wm_with_clock_ranges);
 		data->water_marks_bitmap = WaterMarksExist;
 	}
 
 	return result;
 }
 
-static int vega10_force_clock_level(struct pp_hwmgr *hwmgr,
-		enum pp_clock_type type, uint32_t mask)
-{
-	struct vega10_hwmgr *data = hwmgr->backend;
-
-	switch (type) {
-	case PP_SCLK:
-		data->smc_state_table.gfx_boot_level = mask ? (ffs(mask) - 1) : 0;
-		data->smc_state_table.gfx_max_level = mask ? (fls(mask) - 1) : 0;
-
-		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_bootup_level(hwmgr),
-			"Failed to upload boot level to lowest!",
-			return -EINVAL);
-
-		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_max_level(hwmgr),
-			"Failed to upload dpm max level to highest!",
-			return -EINVAL);
-		break;
-
-	case PP_MCLK:
-		data->smc_state_table.mem_boot_level = mask ? (ffs(mask) - 1) : 0;
-		data->smc_state_table.mem_max_level = mask ? (fls(mask) - 1) : 0;
-
-		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_bootup_level(hwmgr),
-			"Failed to upload boot level to lowest!",
-			return -EINVAL);
-
-		PP_ASSERT_WITH_CODE(!vega10_upload_dpm_max_level(hwmgr),
-			"Failed to upload dpm max level to highest!",
-			return -EINVAL);
-
-		break;
-
-	case PP_PCIE:
-	default:
-		break;
-	}
-
-	return 0;
-}
-
 static int vega10_print_clock_levels(struct pp_hwmgr *hwmgr,
 		enum pp_clock_type type, char *buf)
 {
@@ -4454,6 +4211,8 @@ static int vega10_print_clock_levels(struct pp_hwmgr *hwmgr,
 	struct vega10_single_dpm_table *sclk_table = &(data->dpm_table.gfx_table);
 	struct vega10_single_dpm_table *mclk_table = &(data->dpm_table.mem_table);
 	struct vega10_pcie_table *pcie_table = &(data->dpm_table.pcie_table);
+	struct vega10_odn_clock_voltage_dependency_table *podn_vdd_dep = NULL;
+
 	int i, now, size = 0;
 
 	switch (type) {
@@ -4492,6 +4251,40 @@ static int vega10_print_clock_levels(struct pp_hwmgr *hwmgr,
 					(pcie_table->pcie_gen[i] == 2) ? "8.0GT/s, x16" : "",
 					(i == now) ? "*" : "");
 		break;
+	case OD_SCLK:
+		if (hwmgr->od_enabled) {
+			size = sprintf(buf, "%s:\n", "OD_SCLK");
+			podn_vdd_dep = &data->odn_dpm_table.vdd_dep_on_sclk;
+			for (i = 0; i < podn_vdd_dep->count; i++)
+				size += sprintf(buf + size, "%d: %10uMhz %10umV\n",
+					i, podn_vdd_dep->entries[i].clk / 100,
+						podn_vdd_dep->entries[i].vddc);
+		}
+		break;
+	case OD_MCLK:
+		if (hwmgr->od_enabled) {
+			size = sprintf(buf, "%s:\n", "OD_MCLK");
+			podn_vdd_dep = &data->odn_dpm_table.vdd_dep_on_mclk;
+			for (i = 0; i < podn_vdd_dep->count; i++)
+				size += sprintf(buf + size, "%d: %10uMhz %10umV\n",
+					i, podn_vdd_dep->entries[i].clk/100,
+						podn_vdd_dep->entries[i].vddc);
+		}
+		break;
+	case OD_RANGE:
+		if (hwmgr->od_enabled) {
+			size = sprintf(buf, "%s:\n", "OD_RANGE");
+			size += sprintf(buf + size, "SCLK: %7uMHz %10uMHz\n",
+				data->golden_dpm_table.gfx_table.dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.engineClock/100);
+			size += sprintf(buf + size, "MCLK: %7uMHz %10uMHz\n",
+				data->golden_dpm_table.mem_table.dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.memoryClock/100);
+			size += sprintf(buf + size, "VDDC: %7umV %11umV\n",
+				data->odn_dpm_table.min_vddc,
+				data->odn_dpm_table.max_vddc);
+		}
+		break;
 	default:
 		break;
 	}
@@ -4501,10 +4294,8 @@ static int vega10_print_clock_levels(struct pp_hwmgr *hwmgr,
 static int vega10_display_configuration_changed_task(struct pp_hwmgr *hwmgr)
 {
 	struct vega10_hwmgr *data = hwmgr->backend;
-	int result = 0;
-	uint32_t num_turned_on_displays = 1;
 	Watermarks_t *wm_table = &(data->smc_state_table.water_marks_table);
-	struct cgs_display_info info = {0};
+	int result = 0;
 
 	if ((data->water_marks_bitmap & WaterMarksExist) &&
 			!(data->water_marks_bitmap & WaterMarksLoaded)) {
@@ -4514,10 +4305,8 @@ static int vega10_display_configuration_changed_task(struct pp_hwmgr *hwmgr)
 	}
 
 	if (data->water_marks_bitmap & WaterMarksLoaded) {
-		cgs_get_active_displays_info(hwmgr->device, &info);
-		num_turned_on_displays = info.display_count;
 		smum_send_msg_to_smc_with_parameter(hwmgr,
-			PPSMC_MSG_NumOfDisplays, num_turned_on_displays);
+			PPSMC_MSG_NumOfDisplays, hwmgr->display_config->num_display);
 	}
 
 	return result;
@@ -4603,15 +4392,12 @@ vega10_check_smc_update_required_for_display_configuration(struct pp_hwmgr *hwmg
 {
 	struct vega10_hwmgr *data = hwmgr->backend;
 	bool is_update_required = false;
-	struct cgs_display_info info = {0, 0, NULL};
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-
-	if (data->display_timing.num_existing_displays != info.display_count)
+	if (data->display_timing.num_existing_displays != hwmgr->display_config->num_display)
 		is_update_required = true;
 
 	if (PP_CAP(PHM_PlatformCaps_SclkDeepSleep)) {
-		if (data->display_timing.min_clock_in_sr != hwmgr->display_config.min_core_set_clock_in_sr)
+		if (data->display_timing.min_clock_in_sr != hwmgr->display_config->min_core_set_clock_in_sr)
 			is_update_required = true;
 	}
 
@@ -4886,6 +4672,166 @@ static int vega10_set_power_profile_mode(struct pp_hwmgr *hwmgr, long *input, ui
 	return 0;
 }
 
+
+static bool vega10_check_clk_voltage_valid(struct pp_hwmgr *hwmgr,
+					enum PP_OD_DPM_TABLE_COMMAND type,
+					uint32_t clk,
+					uint32_t voltage)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+	struct vega10_odn_dpm_table *odn_table = &(data->odn_dpm_table);
+	struct vega10_single_dpm_table *golden_table;
+
+	if (voltage < odn_table->min_vddc || voltage > odn_table->max_vddc) {
+		pr_info("OD voltage is out of range [%d - %d] mV\n", odn_table->min_vddc, odn_table->max_vddc);
+		return false;
+	}
+
+	if (type == PP_OD_EDIT_SCLK_VDDC_TABLE) {
+		golden_table = &(data->golden_dpm_table.gfx_table);
+		if (golden_table->dpm_levels[0].value > clk ||
+			hwmgr->platform_descriptor.overdriveLimit.engineClock < clk) {
+			pr_info("OD engine clock is out of range [%d - %d] MHz\n",
+				golden_table->dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.engineClock/100);
+			return false;
+		}
+	} else if (type == PP_OD_EDIT_MCLK_VDDC_TABLE) {
+		golden_table = &(data->golden_dpm_table.mem_table);
+		if (golden_table->dpm_levels[0].value > clk ||
+			hwmgr->platform_descriptor.overdriveLimit.memoryClock < clk) {
+			pr_info("OD memory clock is out of range [%d - %d] MHz\n",
+				golden_table->dpm_levels[0].value/100,
+				hwmgr->platform_descriptor.overdriveLimit.memoryClock/100);
+			return false;
+		}
+	} else {
+		return false;
+	}
+
+	return true;
+}
+
+static void vega10_odn_update_soc_table(struct pp_hwmgr *hwmgr,
+						enum PP_OD_DPM_TABLE_COMMAND type)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+	struct phm_ppt_v2_information *table_info = hwmgr->pptable;
+	struct phm_ppt_v1_clock_voltage_dependency_table *dep_table = table_info->vdd_dep_on_socclk;
+	struct vega10_single_dpm_table *dpm_table = &data->golden_dpm_table.soc_table;
+
+	struct vega10_odn_clock_voltage_dependency_table *podn_vdd_dep_on_socclk =
+							&data->odn_dpm_table.vdd_dep_on_socclk;
+	struct vega10_odn_vddc_lookup_table *od_vddc_lookup_table = &data->odn_dpm_table.vddc_lookup_table;
+
+	struct vega10_odn_clock_voltage_dependency_table *podn_vdd_dep;
+	uint8_t i, j;
+
+	if (type == PP_OD_EDIT_SCLK_VDDC_TABLE) {
+		podn_vdd_dep = &data->odn_dpm_table.vdd_dep_on_sclk;
+		for (i = 0; i < podn_vdd_dep->count - 1; i++)
+			od_vddc_lookup_table->entries[i].us_vdd = podn_vdd_dep->entries[i].vddc;
+		if (od_vddc_lookup_table->entries[i].us_vdd < podn_vdd_dep->entries[i].vddc)
+			od_vddc_lookup_table->entries[i].us_vdd = podn_vdd_dep->entries[i].vddc;
+	} else if (type == PP_OD_EDIT_MCLK_VDDC_TABLE) {
+		podn_vdd_dep = &data->odn_dpm_table.vdd_dep_on_mclk;
+		for (i = 0; i < dpm_table->count; i++) {
+			for (j = 0; j < od_vddc_lookup_table->count; j++) {
+				if (od_vddc_lookup_table->entries[j].us_vdd >
+					podn_vdd_dep->entries[i].vddc)
+					break;
+			}
+			if (j == od_vddc_lookup_table->count) {
+				od_vddc_lookup_table->entries[j-1].us_vdd =
+					podn_vdd_dep->entries[i].vddc;
+				data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_VDDC;
+			}
+			podn_vdd_dep->entries[i].vddInd = j;
+		}
+		dpm_table = &data->dpm_table.soc_table;
+		for (i = 0; i < dep_table->count; i++) {
+			if (dep_table->entries[i].vddInd == podn_vdd_dep->entries[dep_table->count-1].vddInd &&
+					dep_table->entries[i].clk < podn_vdd_dep->entries[dep_table->count-1].clk) {
+				data->need_update_dpm_table |= DPMTABLE_UPDATE_SOCCLK;
+				podn_vdd_dep_on_socclk->entries[i].clk = podn_vdd_dep->entries[dep_table->count-1].clk;
+				dpm_table->dpm_levels[i].value = podn_vdd_dep_on_socclk->entries[i].clk;
+			}
+		}
+		if (podn_vdd_dep_on_socclk->entries[podn_vdd_dep_on_socclk->count - 1].clk <
+					podn_vdd_dep->entries[dep_table->count-1].clk) {
+			data->need_update_dpm_table |= DPMTABLE_UPDATE_SOCCLK;
+			podn_vdd_dep_on_socclk->entries[podn_vdd_dep_on_socclk->count - 1].clk = podn_vdd_dep->entries[dep_table->count-1].clk;
+			dpm_table->dpm_levels[podn_vdd_dep_on_socclk->count - 1].value = podn_vdd_dep->entries[dep_table->count-1].clk;
+		}
+		if (podn_vdd_dep_on_socclk->entries[podn_vdd_dep_on_socclk->count - 1].vddInd <
+					podn_vdd_dep->entries[dep_table->count-1].vddInd) {
+			data->need_update_dpm_table |= DPMTABLE_UPDATE_SOCCLK;
+			podn_vdd_dep_on_socclk->entries[podn_vdd_dep_on_socclk->count - 1].vddInd = podn_vdd_dep->entries[dep_table->count-1].vddInd;
+		}
+	}
+}
+
+static int vega10_odn_edit_dpm_table(struct pp_hwmgr *hwmgr,
+					enum PP_OD_DPM_TABLE_COMMAND type,
+					long *input, uint32_t size)
+{
+	struct vega10_hwmgr *data = hwmgr->backend;
+	struct vega10_odn_clock_voltage_dependency_table *podn_vdd_dep_table;
+	struct vega10_single_dpm_table *dpm_table;
+
+	uint32_t input_clk;
+	uint32_t input_vol;
+	uint32_t input_level;
+	uint32_t i;
+
+	PP_ASSERT_WITH_CODE(input, "NULL user input for clock and voltage",
+				return -EINVAL);
+
+	if (!hwmgr->od_enabled) {
+		pr_info("OverDrive feature not enabled\n");
+		return -EINVAL;
+	}
+
+	if (PP_OD_EDIT_SCLK_VDDC_TABLE == type) {
+		dpm_table = &data->dpm_table.gfx_table;
+		podn_vdd_dep_table = &data->odn_dpm_table.vdd_dep_on_sclk;
+		data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_SCLK;
+	} else if (PP_OD_EDIT_MCLK_VDDC_TABLE == type) {
+		dpm_table = &data->dpm_table.mem_table;
+		podn_vdd_dep_table = &data->odn_dpm_table.vdd_dep_on_mclk;
+		data->need_update_dpm_table |= DPMTABLE_OD_UPDATE_MCLK;
+	} else if (PP_OD_RESTORE_DEFAULT_TABLE == type) {
+		memcpy(&(data->dpm_table), &(data->golden_dpm_table), sizeof(struct vega10_dpm_table));
+		vega10_odn_initial_default_setting(hwmgr);
+		return 0;
+	} else if (PP_OD_COMMIT_DPM_TABLE == type) {
+		vega10_check_dpm_table_updated(hwmgr);
+		return 0;
+	} else {
+		return -EINVAL;
+	}
+
+	for (i = 0; i < size; i += 3) {
+		if (i + 3 > size || input[i] >= podn_vdd_dep_table->count) {
+			pr_info("invalid clock voltage input\n");
+			return 0;
+		}
+		input_level = input[i];
+		input_clk = input[i+1] * 100;
+		input_vol = input[i+2];
+
+		if (vega10_check_clk_voltage_valid(hwmgr, type, input_clk, input_vol)) {
+			dpm_table->dpm_levels[input_level].value = input_clk;
+			podn_vdd_dep_table->entries[input_level].clk = input_clk;
+			podn_vdd_dep_table->entries[input_level].vddc = input_vol;
+		} else {
+			return -EINVAL;
+		}
+	}
+	vega10_odn_update_soc_table(hwmgr, type);
+	return 0;
+}
+
 static const struct pp_hwmgr_func vega10_hwmgr_funcs = {
 	.backend_init = vega10_hwmgr_backend_init,
 	.backend_fini = vega10_hwmgr_backend_fini,
@@ -4944,6 +4890,7 @@ static const struct pp_hwmgr_func vega10_hwmgr_funcs = {
 	.get_power_profile_mode = vega10_get_power_profile_mode,
 	.set_power_profile_mode = vega10_set_power_profile_mode,
 	.set_power_limit = vega10_set_power_limit,
+	.odn_edit_dpm_table = vega10_odn_edit_dpm_table,
 };
 
 int vega10_enable_smc_features(struct pp_hwmgr *hwmgr,
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h
index 5339ea1f3dce..339820da9e6a 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.h
@@ -282,15 +282,21 @@ struct vega10_registry_data {
 
 struct vega10_odn_clock_voltage_dependency_table {
 	uint32_t count;
-	struct phm_ppt_v1_clock_voltage_dependency_record
-		entries[MAX_REGULAR_DPM_NUMBER];
+	struct phm_ppt_v1_clock_voltage_dependency_record entries[MAX_REGULAR_DPM_NUMBER];
+};
+
+struct vega10_odn_vddc_lookup_table {
+	uint32_t count;
+	struct phm_ppt_v1_voltage_lookup_record entries[MAX_REGULAR_DPM_NUMBER];
 };
 
 struct vega10_odn_dpm_table {
-	struct phm_odn_clock_levels		odn_core_clock_dpm_levels;
-	struct phm_odn_clock_levels		odn_memory_clock_dpm_levels;
-	struct vega10_odn_clock_voltage_dependency_table		vdd_dependency_on_sclk;
-	struct vega10_odn_clock_voltage_dependency_table		vdd_dependency_on_mclk;
+	struct vega10_odn_clock_voltage_dependency_table vdd_dep_on_sclk;
+	struct vega10_odn_clock_voltage_dependency_table vdd_dep_on_mclk;
+	struct vega10_odn_clock_voltage_dependency_table vdd_dep_on_socclk;
+	struct vega10_odn_vddc_lookup_table vddc_lookup_table;
+	uint32_t max_vddc;
+	uint32_t min_vddc;
 };
 
 struct vega10_odn_fan_table {
@@ -301,8 +307,8 @@ struct vega10_odn_fan_table {
 };
 
 struct vega10_hwmgr {
-	struct vega10_dpm_table			dpm_table;
-	struct vega10_dpm_table			golden_dpm_table;
+	struct vega10_dpm_table          dpm_table;
+	struct vega10_dpm_table          golden_dpm_table;
 	struct vega10_registry_data      registry_data;
 	struct vega10_vbios_boot_state   vbios_boot_state;
 	struct vega10_mclk_latency_table mclk_latency_table;
@@ -364,16 +370,11 @@ struct vega10_hwmgr {
 	/* ---- Power Gating States ---- */
 	bool                           uvd_power_gated;
 	bool                           vce_power_gated;
-	bool                           samu_power_gated;
 	bool                           need_long_memory_training;
 
 	/* Internal settings to apply the application power optimization parameters */
-	bool                           apply_optimized_settings;
 	uint32_t                       disable_dpm_mask;
 
-	/* ---- Overdrive next setting ---- */
-	uint32_t                       apply_overdrive_next_settings_mask;
-
 	/* ---- SMU9 ---- */
 	struct smu_features            smu_features[GNLD_FEATURES_MAX];
 	struct vega10_smc_state_table  smc_state_table;
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c
index ba63faefc61f..22364875a943 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_powertune.c
@@ -27,7 +27,7 @@
 #include "vega10_ppsmc.h"
 #include "vega10_inc.h"
 #include "pp_debug.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 
 static const struct vega10_didt_config_reg SEDiDtTuningCtrlConfig_Vega10[] =
 {
@@ -888,36 +888,36 @@ static void vega10_didt_set_mask(struct pp_hwmgr *hwmgr, const bool enable)
 	if (PP_CAP(PHM_PlatformCaps_DiDtEDCEnable)) {
 		if (PP_CAP(PHM_PlatformCaps_SQRamping)) {
 			data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_SQ_EDC_CTRL);
-			data = CGS_REG_SET_FIELD(data, DIDT_SQ_EDC_CTRL, EDC_EN, en);
-			data = CGS_REG_SET_FIELD(data, DIDT_SQ_EDC_CTRL, EDC_SW_RST, ~en);
+			data = REG_SET_FIELD(data, DIDT_SQ_EDC_CTRL, EDC_EN, en);
+			data = REG_SET_FIELD(data, DIDT_SQ_EDC_CTRL, EDC_SW_RST, ~en);
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_SQ_EDC_CTRL, data);
 		}
 
 		if (PP_CAP(PHM_PlatformCaps_DBRamping)) {
 			data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DB_EDC_CTRL);
-			data = CGS_REG_SET_FIELD(data, DIDT_DB_EDC_CTRL, EDC_EN, en);
-			data = CGS_REG_SET_FIELD(data, DIDT_DB_EDC_CTRL, EDC_SW_RST, ~en);
+			data = REG_SET_FIELD(data, DIDT_DB_EDC_CTRL, EDC_EN, en);
+			data = REG_SET_FIELD(data, DIDT_DB_EDC_CTRL, EDC_SW_RST, ~en);
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DB_EDC_CTRL, data);
 		}
 
 		if (PP_CAP(PHM_PlatformCaps_TDRamping)) {
 			data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TD_EDC_CTRL);
-			data = CGS_REG_SET_FIELD(data, DIDT_TD_EDC_CTRL, EDC_EN, en);
-			data = CGS_REG_SET_FIELD(data, DIDT_TD_EDC_CTRL, EDC_SW_RST, ~en);
+			data = REG_SET_FIELD(data, DIDT_TD_EDC_CTRL, EDC_EN, en);
+			data = REG_SET_FIELD(data, DIDT_TD_EDC_CTRL, EDC_SW_RST, ~en);
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TD_EDC_CTRL, data);
 		}
 
 		if (PP_CAP(PHM_PlatformCaps_TCPRamping)) {
 			data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TCP_EDC_CTRL);
-			data = CGS_REG_SET_FIELD(data, DIDT_TCP_EDC_CTRL, EDC_EN, en);
-			data = CGS_REG_SET_FIELD(data, DIDT_TCP_EDC_CTRL, EDC_SW_RST, ~en);
+			data = REG_SET_FIELD(data, DIDT_TCP_EDC_CTRL, EDC_EN, en);
+			data = REG_SET_FIELD(data, DIDT_TCP_EDC_CTRL, EDC_SW_RST, ~en);
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_TCP_EDC_CTRL, data);
 		}
 
 		if (PP_CAP(PHM_PlatformCaps_DBRRamping)) {
 			data = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DBR_EDC_CTRL);
-			data = CGS_REG_SET_FIELD(data, DIDT_DBR_EDC_CTRL, EDC_EN, en);
-			data = CGS_REG_SET_FIELD(data, DIDT_DBR_EDC_CTRL, EDC_SW_RST, ~en);
+			data = REG_SET_FIELD(data, DIDT_DBR_EDC_CTRL, EDC_EN, en);
+			data = REG_SET_FIELD(data, DIDT_DBR_EDC_CTRL, EDC_SW_RST, ~en);
 			cgs_write_ind_register(hwmgr->device, CGS_IND_REG__DIDT, ixDIDT_DBR_EDC_CTRL, data);
 		}
 	}
@@ -930,20 +930,18 @@ static void vega10_didt_set_mask(struct pp_hwmgr *hwmgr, const bool enable)
 
 static int vega10_enable_cac_driving_se_didt_config(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 	uint32_t num_se = 0, count, data;
-	struct amdgpu_device *adev = hwmgr->adev;
-	uint32_t reg;
 
 	num_se = adev->gfx.config.max_shader_engines;
 
-	cgs_enter_safe_mode(hwmgr->device, true);
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
-	cgs_lock_grbm_idx(hwmgr->device, true);
-	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
+	mutex_lock(&adev->grbm_idx_mutex);
 	for (count = 0; count < num_se; count++) {
 		data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT);
-		cgs_write_register(hwmgr->device, reg, data);
+		WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, data);
 
 		result =  vega10_program_didt_config_registers(hwmgr, SEDiDtStallCtrlConfig_vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, SEDiDtStallPatternConfig_vega10, VEGA10_CONFIGREG_DIDT);
@@ -958,43 +956,43 @@ static int vega10_enable_cac_driving_se_didt_config(struct pp_hwmgr *hwmgr)
 		if (0 != result)
 			break;
 	}
-	cgs_write_register(hwmgr->device, reg, 0xE0000000);
-	cgs_lock_grbm_idx(hwmgr->device, false);
+	WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, 0xE0000000);
+	mutex_unlock(&adev->grbm_idx_mutex);
 
 	vega10_didt_set_mask(hwmgr, true);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	return 0;
 }
 
 static int vega10_disable_cac_driving_se_didt_config(struct pp_hwmgr *hwmgr)
 {
-	cgs_enter_safe_mode(hwmgr->device, true);
+	struct amdgpu_device *adev = hwmgr->adev;
+
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	vega10_didt_set_mask(hwmgr, false);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	return 0;
 }
 
 static int vega10_enable_psm_gc_didt_config(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 	uint32_t num_se = 0, count, data;
-	struct amdgpu_device *adev = hwmgr->adev;
-	uint32_t reg;
 
 	num_se = adev->gfx.config.max_shader_engines;
 
-	cgs_enter_safe_mode(hwmgr->device, true);
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
-	cgs_lock_grbm_idx(hwmgr->device, true);
-	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
+	mutex_lock(&adev->grbm_idx_mutex);
 	for (count = 0; count < num_se; count++) {
 		data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT);
-		cgs_write_register(hwmgr->device, reg, data);
+		WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, data);
 
 		result = vega10_program_didt_config_registers(hwmgr, SEDiDtStallCtrlConfig_vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, SEDiDtStallPatternConfig_vega10, VEGA10_CONFIGREG_DIDT);
@@ -1003,12 +1001,12 @@ static int vega10_enable_psm_gc_didt_config(struct pp_hwmgr *hwmgr)
 		if (0 != result)
 			break;
 	}
-	cgs_write_register(hwmgr->device, reg, 0xE0000000);
-	cgs_lock_grbm_idx(hwmgr->device, false);
+	WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, 0xE0000000);
+	mutex_unlock(&adev->grbm_idx_mutex);
 
 	vega10_didt_set_mask(hwmgr, true);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	vega10_program_gc_didt_config_registers(hwmgr, GCDiDtDroopCtrlConfig_vega10);
 	if (PP_CAP(PHM_PlatformCaps_GCEDC))
@@ -1022,13 +1020,14 @@ static int vega10_enable_psm_gc_didt_config(struct pp_hwmgr *hwmgr)
 
 static int vega10_disable_psm_gc_didt_config(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t data;
 
-	cgs_enter_safe_mode(hwmgr->device, true);
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	vega10_didt_set_mask(hwmgr, false);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	if (PP_CAP(PHM_PlatformCaps_GCEDC)) {
 		data = 0x00000000;
@@ -1043,20 +1042,18 @@ static int vega10_disable_psm_gc_didt_config(struct pp_hwmgr *hwmgr)
 
 static int vega10_enable_se_edc_config(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 	uint32_t num_se = 0, count, data;
-	struct amdgpu_device *adev = hwmgr->adev;
-	uint32_t reg;
 
 	num_se = adev->gfx.config.max_shader_engines;
 
-	cgs_enter_safe_mode(hwmgr->device, true);
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
-	cgs_lock_grbm_idx(hwmgr->device, true);
-	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
+	mutex_lock(&adev->grbm_idx_mutex);
 	for (count = 0; count < num_se; count++) {
 		data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT);
-		cgs_write_register(hwmgr->device, reg, data);
+		WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, data);
 		result = vega10_program_didt_config_registers(hwmgr, SEDiDtWeightConfig_Vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, SEEDCStallPatternConfig_Vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, SEEDCStallDelayConfig_Vega10, VEGA10_CONFIGREG_DIDT);
@@ -1067,47 +1064,47 @@ static int vega10_enable_se_edc_config(struct pp_hwmgr *hwmgr)
 		if (0 != result)
 			break;
 	}
-	cgs_write_register(hwmgr->device, reg, 0xE0000000);
-	cgs_lock_grbm_idx(hwmgr->device, false);
+	WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, 0xE0000000);
+	mutex_unlock(&adev->grbm_idx_mutex);
 
 	vega10_didt_set_mask(hwmgr, true);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	return 0;
 }
 
 static int vega10_disable_se_edc_config(struct pp_hwmgr *hwmgr)
 {
-	cgs_enter_safe_mode(hwmgr->device, true);
+	struct amdgpu_device *adev = hwmgr->adev;
+
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	vega10_didt_set_mask(hwmgr, false);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	return 0;
 }
 
 static int vega10_enable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 {
-	int result;
+	struct amdgpu_device *adev = hwmgr->adev;
+	int result = 0;
 	uint32_t num_se = 0;
 	uint32_t count, data;
-	struct amdgpu_device *adev = hwmgr->adev;
-	uint32_t reg;
 
 	num_se = adev->gfx.config.max_shader_engines;
 
-	cgs_enter_safe_mode(hwmgr->device, true);
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	vega10_program_gc_didt_config_registers(hwmgr, AvfsPSMResetConfig_vega10);
 
-	cgs_lock_grbm_idx(hwmgr->device, true);
-	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
+	mutex_lock(&adev->grbm_idx_mutex);
 	for (count = 0; count < num_se; count++) {
 		data = GRBM_GFX_INDEX__INSTANCE_BROADCAST_WRITES_MASK | GRBM_GFX_INDEX__SH_BROADCAST_WRITES_MASK | ( count << GRBM_GFX_INDEX__SE_INDEX__SHIFT);
-		cgs_write_register(hwmgr->device, reg, data);
-		result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCStallPatternConfig_Vega10, VEGA10_CONFIGREG_DIDT);
+		WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, data);
+		result = vega10_program_didt_config_registers(hwmgr, PSMSEEDCStallPatternConfig_Vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCStallDelayConfig_Vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCCtrlResetConfig_Vega10, VEGA10_CONFIGREG_DIDT);
 		result |= vega10_program_didt_config_registers(hwmgr, PSMSEEDCCtrlConfig_Vega10, VEGA10_CONFIGREG_DIDT);
@@ -1115,12 +1112,12 @@ static int vega10_enable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 		if (0 != result)
 			break;
 	}
-	cgs_write_register(hwmgr->device, reg, 0xE0000000);
-	cgs_lock_grbm_idx(hwmgr->device, false);
+	WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, 0xE0000000);
+	mutex_unlock(&adev->grbm_idx_mutex);
 
 	vega10_didt_set_mask(hwmgr, true);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	vega10_program_gc_didt_config_registers(hwmgr, PSMGCEDCDroopCtrlConfig_vega10);
 
@@ -1137,13 +1134,14 @@ static int vega10_enable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 
 static int vega10_disable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t data;
 
-	cgs_enter_safe_mode(hwmgr->device, true);
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
 	vega10_didt_set_mask(hwmgr, false);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	if (PP_CAP(PHM_PlatformCaps_GCEDC)) {
 		data = 0x00000000;
@@ -1158,15 +1156,14 @@ static int vega10_disable_psm_gc_edc_config(struct pp_hwmgr *hwmgr)
 
 static int vega10_enable_se_edc_force_stall_config(struct pp_hwmgr *hwmgr)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 	int result;
 
-	cgs_enter_safe_mode(hwmgr->device, true);
+	adev->gfx.rlc.funcs->enter_safe_mode(adev);
 
-	cgs_lock_grbm_idx(hwmgr->device, true);
-	reg = soc15_get_register_offset(GC_HWID, 0, mmGRBM_GFX_INDEX_BASE_IDX, mmGRBM_GFX_INDEX);
-	cgs_write_register(hwmgr->device, reg, 0xE0000000);
-	cgs_lock_grbm_idx(hwmgr->device, false);
+	mutex_lock(&adev->grbm_idx_mutex);
+	WREG32_SOC15(GC, 0, mmGRBM_GFX_INDEX, 0xE0000000);
+	mutex_unlock(&adev->grbm_idx_mutex);
 
 	result = vega10_program_didt_config_registers(hwmgr, SEEDCForceStallPatternConfig_Vega10, VEGA10_CONFIGREG_DIDT);
 	result |= vega10_program_didt_config_registers(hwmgr, SEEDCCtrlForceStallConfig_Vega10, VEGA10_CONFIGREG_DIDT);
@@ -1175,7 +1172,7 @@ static int vega10_enable_se_edc_force_stall_config(struct pp_hwmgr *hwmgr)
 
 	vega10_didt_set_mask(hwmgr, false);
 
-	cgs_enter_safe_mode(hwmgr->device, false);
+	adev->gfx.rlc.funcs->exit_safe_mode(adev);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c
index c61d0744860d..16b1a9cf6cf0 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c
@@ -52,7 +52,7 @@ static const void *get_powerplay_table(struct pp_hwmgr *hwmgr)
 
 	if (!table_address) {
 		table_address = (ATOM_Vega10_POWERPLAYTABLE *)
-				cgs_atom_get_data_table(hwmgr->device, index,
+				smu_atom_get_data_table(hwmgr->adev, index,
 						&size, &frev, &crev);
 
 		hwmgr->soft_pp_table = table_address;	/*Cache the result in RAM.*/
@@ -267,12 +267,6 @@ static int init_over_drive_limits(
 	hwmgr->platform_descriptor.maxOverdriveVDDC = 0;
 	hwmgr->platform_descriptor.overdriveVDDCStep = 0;
 
-	if (hwmgr->platform_descriptor.overdriveLimit.engineClock == 0 ||
-		hwmgr->platform_descriptor.overdriveLimit.memoryClock == 0) {
-		hwmgr->od_enabled = false;
-		pr_debug("OverDrive feature not support by VBIOS\n");
-	}
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c
index 9f18226a56ea..aa044c1955fe 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_thermal.c
@@ -25,7 +25,7 @@
 #include "vega10_hwmgr.h"
 #include "vega10_ppsmc.h"
 #include "vega10_inc.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 #include "pp_debug.h"
 
 static int vega10_get_current_rpm(struct pp_hwmgr *hwmgr, uint32_t *current_rpm)
@@ -89,6 +89,7 @@ int vega10_fan_ctrl_get_fan_speed_percent(struct pp_hwmgr *hwmgr,
 
 int vega10_fan_ctrl_get_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t *speed)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	struct vega10_hwmgr *data = hwmgr->backend;
 	uint32_t tach_period;
 	uint32_t crystal_clock_freq;
@@ -100,10 +101,8 @@ int vega10_fan_ctrl_get_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t *speed)
 	if (data->smu_features[GNLD_FAN_CONTROL].supported) {
 		result = vega10_get_current_rpm(hwmgr, speed);
 	} else {
-		uint32_t reg = soc15_get_register_offset(THM_HWID, 0,
-				mmCG_TACH_STATUS_BASE_IDX, mmCG_TACH_STATUS);
 		tach_period =
-			CGS_REG_GET_FIELD(cgs_read_register(hwmgr->device, reg),
+			REG_GET_FIELD(RREG32_SOC15(THM, 0, mmCG_TACH_STATUS),
 					  CG_TACH_STATUS,
 					  TACH_PERIOD);
 
@@ -127,26 +126,23 @@ int vega10_fan_ctrl_get_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t *speed)
 */
 int vega10_fan_ctrl_set_static_mode(struct pp_hwmgr *hwmgr, uint32_t mode)
 {
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmCG_FDO_CTRL2_BASE_IDX, mmCG_FDO_CTRL2);
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	if (hwmgr->fan_ctrl_is_in_default_mode) {
 		hwmgr->fan_ctrl_default_mode =
-			CGS_REG_GET_FIELD(cgs_read_register(hwmgr->device, reg),
+			REG_GET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2),
 				CG_FDO_CTRL2, FDO_PWM_MODE);
 		hwmgr->tmin =
-			CGS_REG_GET_FIELD(cgs_read_register(hwmgr->device, reg),
+			REG_GET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2),
 				CG_FDO_CTRL2, TMIN);
 		hwmgr->fan_ctrl_is_in_default_mode = false;
 	}
 
-	cgs_write_register(hwmgr->device, reg,
-			CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+	WREG32_SOC15(THM, 0, mmCG_FDO_CTRL2,
+			REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2),
 				CG_FDO_CTRL2, TMIN, 0));
-	cgs_write_register(hwmgr->device, reg,
-			CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+	WREG32_SOC15(THM, 0, mmCG_FDO_CTRL2,
+			REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2),
 				CG_FDO_CTRL2, FDO_PWM_MODE, mode));
 
 	return 0;
@@ -159,18 +155,15 @@ int vega10_fan_ctrl_set_static_mode(struct pp_hwmgr *hwmgr, uint32_t mode)
 */
 int vega10_fan_ctrl_set_default_mode(struct pp_hwmgr *hwmgr)
 {
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmCG_FDO_CTRL2_BASE_IDX, mmCG_FDO_CTRL2);
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	if (!hwmgr->fan_ctrl_is_in_default_mode) {
-		cgs_write_register(hwmgr->device, reg,
-			CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+		WREG32_SOC15(THM, 0, mmCG_FDO_CTRL2,
+			REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2),
 				CG_FDO_CTRL2, FDO_PWM_MODE,
 				hwmgr->fan_ctrl_default_mode));
-		cgs_write_register(hwmgr->device, reg,
-			CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+		WREG32_SOC15(THM, 0, mmCG_FDO_CTRL2,
+			REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2),
 				CG_FDO_CTRL2, TMIN,
 				hwmgr->tmin << CG_FDO_CTRL2__TMIN__SHIFT));
 		hwmgr->fan_ctrl_is_in_default_mode = true;
@@ -257,10 +250,10 @@ int vega10_fan_ctrl_stop_smc_fan_control(struct pp_hwmgr *hwmgr)
 int vega10_fan_ctrl_set_fan_speed_percent(struct pp_hwmgr *hwmgr,
 		uint32_t speed)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t duty100;
 	uint32_t duty;
 	uint64_t tmp64;
-	uint32_t reg;
 
 	if (hwmgr->thermal_controller.fanInfo.bNoFan)
 		return 0;
@@ -271,10 +264,7 @@ int vega10_fan_ctrl_set_fan_speed_percent(struct pp_hwmgr *hwmgr,
 	if (PP_CAP(PHM_PlatformCaps_MicrocodeFanControl))
 		vega10_fan_ctrl_stop_smc_fan_control(hwmgr);
 
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmCG_FDO_CTRL1_BASE_IDX, mmCG_FDO_CTRL1);
-
-	duty100 = CGS_REG_GET_FIELD(cgs_read_register(hwmgr->device, reg),
+	duty100 = REG_GET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL1),
 				    CG_FDO_CTRL1, FMAX_DUTY100);
 
 	if (duty100 == 0)
@@ -284,10 +274,8 @@ int vega10_fan_ctrl_set_fan_speed_percent(struct pp_hwmgr *hwmgr,
 	do_div(tmp64, 100);
 	duty = (uint32_t)tmp64;
 
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmCG_FDO_CTRL0_BASE_IDX, mmCG_FDO_CTRL0);
-	cgs_write_register(hwmgr->device, reg,
-		CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+	WREG32_SOC15(THM, 0, mmCG_FDO_CTRL0,
+		REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL0),
 			CG_FDO_CTRL0, FDO_STATIC_DUTY, duty));
 
 	return vega10_fan_ctrl_set_static_mode(hwmgr, FDO_PWM_MODE_STATIC);
@@ -317,10 +305,10 @@ int vega10_fan_ctrl_reset_fan_speed_to_default(struct pp_hwmgr *hwmgr)
 */
 int vega10_fan_ctrl_set_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t speed)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t tach_period;
 	uint32_t crystal_clock_freq;
 	int result = 0;
-	uint32_t reg;
 
 	if (hwmgr->thermal_controller.fanInfo.bNoFan ||
 	    (speed < hwmgr->thermal_controller.fanInfo.ulMinRPM) ||
@@ -333,10 +321,8 @@ int vega10_fan_ctrl_set_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t speed)
 	if (!result) {
 		crystal_clock_freq = amdgpu_asic_get_xclk((struct amdgpu_device *)hwmgr->adev);
 		tach_period = 60 * crystal_clock_freq * 10000 / (8 * speed);
-		reg = soc15_get_register_offset(THM_HWID, 0,
-				mmCG_TACH_STATUS_BASE_IDX, mmCG_TACH_STATUS);
-		cgs_write_register(hwmgr->device, reg,
-				CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+		WREG32_SOC15(THM, 0, mmCG_TACH_STATUS,
+				REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_TACH_STATUS),
 					CG_TACH_STATUS, TACH_PERIOD,
 					tach_period));
 	}
@@ -350,13 +336,10 @@ int vega10_fan_ctrl_set_fan_speed_rpm(struct pp_hwmgr *hwmgr, uint32_t speed)
 */
 int vega10_thermal_get_temperature(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int temp;
-	uint32_t reg;
 
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmCG_MULT_THERMAL_STATUS_BASE_IDX,  mmCG_MULT_THERMAL_STATUS);
-
-	temp = cgs_read_register(hwmgr->device, reg);
+	temp = RREG32_SOC15(THM, 0, mmCG_MULT_THERMAL_STATUS);
 
 	temp = (temp & CG_MULT_THERMAL_STATUS__CTF_TEMP_MASK) >>
 			CG_MULT_THERMAL_STATUS__CTF_TEMP__SHIFT;
@@ -379,11 +362,12 @@ int vega10_thermal_get_temperature(struct pp_hwmgr *hwmgr)
 static int vega10_thermal_set_temperature_range(struct pp_hwmgr *hwmgr,
 		struct PP_TemperatureRange *range)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int low = VEGA10_THERMAL_MINIMUM_ALERT_TEMP *
 			PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 	int high = VEGA10_THERMAL_MAXIMUM_ALERT_TEMP *
 			PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
-	uint32_t val, reg;
+	uint32_t val;
 
 	if (low < range->min)
 		low = range->min;
@@ -393,20 +377,17 @@ static int vega10_thermal_set_temperature_range(struct pp_hwmgr *hwmgr,
 	if (low > high)
 		return -EINVAL;
 
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmTHM_THERMAL_INT_CTRL_BASE_IDX, mmTHM_THERMAL_INT_CTRL);
-
-	val = cgs_read_register(hwmgr->device, reg);
+	val = RREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_CTRL);
 
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, MAX_IH_CREDIT, 5);
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, THERM_IH_HW_ENA, 1);
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTH, (high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTL, (low / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, MAX_IH_CREDIT, 5);
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, THERM_IH_HW_ENA, 1);
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTH, (high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTL, (low / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
 	val &= (~THM_THERMAL_INT_CTRL__THERM_TRIGGER_MASK_MASK) &
 			(~THM_THERMAL_INT_CTRL__THERM_INTH_MASK_MASK) &
 			(~THM_THERMAL_INT_CTRL__THERM_INTL_MASK_MASK);
 
-	cgs_write_register(hwmgr->device, reg, val);
+	WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_CTRL, val);
 
 	return 0;
 }
@@ -418,21 +399,17 @@ static int vega10_thermal_set_temperature_range(struct pp_hwmgr *hwmgr,
 */
 static int vega10_thermal_initialize(struct pp_hwmgr *hwmgr)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	if (hwmgr->thermal_controller.fanInfo.ucTachometerPulsesPerRevolution) {
-		reg = soc15_get_register_offset(THM_HWID, 0,
-				mmCG_TACH_CTRL_BASE_IDX, mmCG_TACH_CTRL);
-		cgs_write_register(hwmgr->device, reg,
-			CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+		WREG32_SOC15(THM, 0, mmCG_TACH_CTRL,
+			REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_TACH_CTRL),
 				CG_TACH_CTRL, EDGE_PER_REV,
 				hwmgr->thermal_controller.fanInfo.ucTachometerPulsesPerRevolution - 1));
 	}
 
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmCG_FDO_CTRL2_BASE_IDX, mmCG_FDO_CTRL2);
-	cgs_write_register(hwmgr->device, reg,
-		CGS_REG_SET_FIELD(cgs_read_register(hwmgr->device, reg),
+	WREG32_SOC15(THM, 0, mmCG_FDO_CTRL2,
+		REG_SET_FIELD(RREG32_SOC15(THM, 0, mmCG_FDO_CTRL2),
 			CG_FDO_CTRL2, TACH_PWM_RESP_RATE, 0x28));
 
 	return 0;
@@ -445,9 +422,9 @@ static int vega10_thermal_initialize(struct pp_hwmgr *hwmgr)
 */
 static int vega10_thermal_enable_alert(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	struct vega10_hwmgr *data = hwmgr->backend;
 	uint32_t val = 0;
-	uint32_t reg;
 
 	if (data->smu_features[GNLD_FW_CTF].supported) {
 		if (data->smu_features[GNLD_FW_CTF].enabled)
@@ -465,8 +442,7 @@ static int vega10_thermal_enable_alert(struct pp_hwmgr *hwmgr)
 	val |= (1 << THM_THERMAL_INT_ENA__THERM_INTL_CLR__SHIFT);
 	val |= (1 << THM_THERMAL_INT_ENA__THERM_TRIGGER_CLR__SHIFT);
 
-	reg = soc15_get_register_offset(THM_HWID, 0, mmTHM_THERMAL_INT_ENA_BASE_IDX, mmTHM_THERMAL_INT_ENA);
-	cgs_write_register(hwmgr->device, reg, val);
+	WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_ENA, val);
 
 	return 0;
 }
@@ -477,8 +453,8 @@ static int vega10_thermal_enable_alert(struct pp_hwmgr *hwmgr)
 */
 int vega10_thermal_disable_alert(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	struct vega10_hwmgr *data = hwmgr->backend;
-	uint32_t reg;
 
 	if (data->smu_features[GNLD_FW_CTF].supported) {
 		if (!data->smu_features[GNLD_FW_CTF].enabled)
@@ -493,8 +469,7 @@ int vega10_thermal_disable_alert(struct pp_hwmgr *hwmgr)
 		data->smu_features[GNLD_FW_CTF].enabled = false;
 	}
 
-	reg = soc15_get_register_offset(THM_HWID, 0, mmTHM_THERMAL_INT_ENA_BASE_IDX, mmTHM_THERMAL_INT_ENA);
-	cgs_write_register(hwmgr->device, reg, 0);
+	WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_ENA, 0);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
index 200de46bd06b..4ed218dd8ba7 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.c
@@ -34,7 +34,6 @@
 #include "atomfirmware.h"
 #include "cgs_common.h"
 #include "vega12_inc.h"
-#include "pp_soc15.h"
 #include "pppcielanes.h"
 #include "vega12_hwmgr.h"
 #include "vega12_processpptables.h"
@@ -82,6 +81,7 @@ static void vega12_set_default_registry_data(struct pp_hwmgr *hwmgr)
 
 	data->registry_data.disallowed_features = 0x0;
 	data->registry_data.od_state_in_dc_support = 0;
+	data->registry_data.thermal_support = 1;
 	data->registry_data.skip_baco_hardware = 0;
 
 	data->registry_data.log_avfs_param = 0;
@@ -423,6 +423,11 @@ static int vega12_hwmgr_backend_init(struct pp_hwmgr *hwmgr)
 			hwmgr->thermal_controller.advanceFanControlParameters.usFanPWMMinLimit *
 			hwmgr->thermal_controller.fanInfo.ulMaxRPM / 100;
 
+	if (hwmgr->feature_mask & PP_GFXOFF_MASK)
+		data->gfxoff_controlled_by_driver = true;
+	else
+		data->gfxoff_controlled_by_driver = false;
+
 	return result;
 }
 
@@ -454,43 +459,36 @@ static int vega12_setup_asic_task(struct pp_hwmgr *hwmgr)
  */
 static void vega12_init_dpm_state(struct vega12_dpm_state *dpm_state)
 {
-	dpm_state->soft_min_level = 0xff;
-	dpm_state->soft_max_level = 0xff;
-	dpm_state->hard_min_level = 0xff;
-	dpm_state->hard_max_level = 0xff;
+	dpm_state->soft_min_level = 0x0;
+	dpm_state->soft_max_level = 0xffff;
+	dpm_state->hard_min_level = 0x0;
+	dpm_state->hard_max_level = 0xffff;
 }
 
-static int vega12_get_number_dpm_level(struct pp_hwmgr *hwmgr,
-		PPCLK_e clkID, uint32_t *num_dpm_level)
+static int vega12_get_number_of_dpm_level(struct pp_hwmgr *hwmgr,
+		PPCLK_e clk_id, uint32_t *num_of_levels)
 {
-	int result;
-	/*
-	 * SMU expects the Clock ID to be in the top 16 bits.
-	 * Lower 16 bits specify the level however 0xFF is a
-	 * special argument the returns the total number of levels
-	 */
-	PP_ASSERT_WITH_CODE(smum_send_msg_to_smc_with_parameter(hwmgr,
-		PPSMC_MSG_GetDpmFreqByIndex, (clkID << 16 | 0xFF)) == 0,
-		"[GetNumberDpmLevel] Failed to get DPM levels from SMU for CLKID!",
-		return -EINVAL);
-
-	result = vega12_read_arg_from_smc(hwmgr, num_dpm_level);
+	int ret = 0;
 
-	PP_ASSERT_WITH_CODE(*num_dpm_level < MAX_REGULAR_DPM_NUMBER,
-		"[GetNumberDPMLevel] Number of DPM levels is greater than limit",
-		return -EINVAL);
+	ret = smum_send_msg_to_smc_with_parameter(hwmgr,
+			PPSMC_MSG_GetDpmFreqByIndex,
+			(clk_id << 16 | 0xFF));
+	PP_ASSERT_WITH_CODE(!ret,
+			"[GetNumOfDpmLevel] failed to get dpm levels!",
+			return ret);
 
-	PP_ASSERT_WITH_CODE(*num_dpm_level != 0,
-		"[GetNumberDPMLevel] Number of CLK Levels is zero!",
-		return -EINVAL);
+	*num_of_levels = smum_get_argument(hwmgr);
+	PP_ASSERT_WITH_CODE(*num_of_levels > 0,
+			"[GetNumOfDpmLevel] number of clk levels is invalid!",
+			return -EINVAL);
 
-	return result;
+	return ret;
 }
 
 static int vega12_get_dpm_frequency_by_index(struct pp_hwmgr *hwmgr,
 		PPCLK_e clkID, uint32_t index, uint32_t *clock)
 {
-	int result;
+	int result = 0;
 
 	/*
 	 *SMU expects the Clock ID to be in the top 16 bits.
@@ -501,15 +499,36 @@ static int vega12_get_dpm_frequency_by_index(struct pp_hwmgr *hwmgr,
 		"[GetDpmFrequencyByIndex] Failed to get dpm frequency from SMU!",
 		return -EINVAL);
 
-	result = vega12_read_arg_from_smc(hwmgr, clock);
-
-	PP_ASSERT_WITH_CODE(*clock != 0,
-		"[GetDPMFrequencyByIndex] Failed to get dpm frequency by index.!",
-		return -EINVAL);
+	*clock = smum_get_argument(hwmgr);
 
 	return result;
 }
 
+static int vega12_setup_single_dpm_table(struct pp_hwmgr *hwmgr,
+		struct vega12_single_dpm_table *dpm_table, PPCLK_e clk_id)
+{
+	int ret = 0;
+	uint32_t i, num_of_levels, clk;
+
+	ret = vega12_get_number_of_dpm_level(hwmgr, clk_id, &num_of_levels);
+	PP_ASSERT_WITH_CODE(!ret,
+			"[SetupSingleDpmTable] failed to get clk levels!",
+			return ret);
+
+	dpm_table->count = num_of_levels;
+
+	for (i = 0; i < num_of_levels; i++) {
+		ret = vega12_get_dpm_frequency_by_index(hwmgr, clk_id, i, &clk);
+		PP_ASSERT_WITH_CODE(!ret,
+			"[SetupSingleDpmTable] failed to get clk of specific level!",
+			return ret);
+		dpm_table->dpm_levels[i].value = clk;
+		dpm_table->dpm_levels[i].enabled = true;
+	}
+
+	return ret;
+}
+
 /*
  * This function is to initialize all DPM state tables
  * for SMU based on the dependency table.
@@ -520,214 +539,136 @@ static int vega12_get_dpm_frequency_by_index(struct pp_hwmgr *hwmgr,
  */
 static int vega12_setup_default_dpm_tables(struct pp_hwmgr *hwmgr)
 {
-	uint32_t num_levels, i, clock;
 
 	struct vega12_hwmgr *data =
 			(struct vega12_hwmgr *)(hwmgr->backend);
-
 	struct vega12_single_dpm_table *dpm_table;
+	int ret = 0;
 
 	memset(&data->dpm_table, 0, sizeof(data->dpm_table));
 
-	/* Initialize Sclk DPM and SOC DPM table based on allow Sclk values */
+	/* socclk */
 	dpm_table = &(data->dpm_table.soc_table);
-
-	PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, PPCLK_SOCCLK,
-		&num_levels) == 0,
-		"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for SOCCLK!",
-		return -EINVAL);
-
-	dpm_table->count = num_levels;
-
-	for (i = 0; i < num_levels; i++) {
-		PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr,
-			PPCLK_SOCCLK, i, &clock) == 0,
-			"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for SOCCLK!",
-			return -EINVAL);
-
-		dpm_table->dpm_levels[i].value = clock;
+	if (data->smu_features[GNLD_DPM_SOCCLK].enabled) {
+		ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_SOCCLK);
+		PP_ASSERT_WITH_CODE(!ret,
+				"[SetupDefaultDpmTable] failed to get socclk dpm levels!",
+				return ret);
+	} else {
+		dpm_table->count = 1;
+		dpm_table->dpm_levels[0].value = data->vbios_boot_state.soc_clock / 100;
 	}
-
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
 
+	/* gfxclk */
 	dpm_table = &(data->dpm_table.gfx_table);
-
-	PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, PPCLK_GFXCLK,
-		&num_levels) == 0,
-		"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for GFXCLK!",
-		return -EINVAL);
-
-	dpm_table->count = num_levels;
-	for (i = 0; i < num_levels; i++) {
-		PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr,
-			PPCLK_GFXCLK, i, &clock) == 0,
-			"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for GFXCLK!",
-			return -EINVAL);
-
-		dpm_table->dpm_levels[i].value = clock;
+	if (data->smu_features[GNLD_DPM_GFXCLK].enabled) {
+		ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_GFXCLK);
+		PP_ASSERT_WITH_CODE(!ret,
+				"[SetupDefaultDpmTable] failed to get gfxclk dpm levels!",
+				return ret);
+	} else {
+		dpm_table->count = 1;
+		dpm_table->dpm_levels[0].value = data->vbios_boot_state.gfx_clock / 100;
 	}
-
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
-	/* Initialize Mclk DPM table based on allow Mclk values */
-	dpm_table = &(data->dpm_table.mem_table);
-
-	PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, PPCLK_UCLK,
-		&num_levels) == 0,
-		"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for UCLK!",
-		return -EINVAL);
-
-	dpm_table->count = num_levels;
-
-	for (i = 0; i < num_levels; i++) {
-		PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr,
-			PPCLK_UCLK, i, &clock) == 0,
-			"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for UCLK!",
-			return -EINVAL);
 
-		dpm_table->dpm_levels[i].value = clock;
+	/* memclk */
+	dpm_table = &(data->dpm_table.mem_table);
+	if (data->smu_features[GNLD_DPM_UCLK].enabled) {
+		ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_UCLK);
+		PP_ASSERT_WITH_CODE(!ret,
+				"[SetupDefaultDpmTable] failed to get memclk dpm levels!",
+				return ret);
+	} else {
+		dpm_table->count = 1;
+		dpm_table->dpm_levels[0].value = data->vbios_boot_state.mem_clock / 100;
 	}
-
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
 
+	/* eclk */
 	dpm_table = &(data->dpm_table.eclk_table);
-
-	PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, PPCLK_ECLK,
-		&num_levels) == 0,
-		"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for ECLK!",
-		return -EINVAL);
-
-	dpm_table->count = num_levels;
-
-	for (i = 0; i < num_levels; i++) {
-		PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr,
-		PPCLK_ECLK, i, &clock) == 0,
-		"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for ECLK!",
-		return -EINVAL);
-
-		dpm_table->dpm_levels[i].value = clock;
+	if (data->smu_features[GNLD_DPM_VCE].enabled) {
+		ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_ECLK);
+		PP_ASSERT_WITH_CODE(!ret,
+				"[SetupDefaultDpmTable] failed to get eclk dpm levels!",
+				return ret);
+	} else {
+		dpm_table->count = 1;
+		dpm_table->dpm_levels[0].value = data->vbios_boot_state.eclock / 100;
 	}
-
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
 
+	/* vclk */
 	dpm_table = &(data->dpm_table.vclk_table);
-
-	PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, PPCLK_VCLK,
-		&num_levels) == 0,
-		"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for VCLK!",
-		return -EINVAL);
-
-	dpm_table->count = num_levels;
-
-	for (i = 0; i < num_levels; i++) {
-		PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr,
-			PPCLK_VCLK, i, &clock) == 0,
-			"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for VCLK!",
-			return -EINVAL);
-
-		dpm_table->dpm_levels[i].value = clock;
+	if (data->smu_features[GNLD_DPM_UVD].enabled) {
+		ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_VCLK);
+		PP_ASSERT_WITH_CODE(!ret,
+				"[SetupDefaultDpmTable] failed to get vclk dpm levels!",
+				return ret);
+	} else {
+		dpm_table->count = 1;
+		dpm_table->dpm_levels[0].value = data->vbios_boot_state.vclock / 100;
 	}
-
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
 
+	/* dclk */
 	dpm_table = &(data->dpm_table.dclk_table);
-
-	PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr, PPCLK_DCLK,
-		&num_levels) == 0,
-		"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for DCLK!",
-		return -EINVAL);
-
-	dpm_table->count = num_levels;
-
-	for (i = 0; i < num_levels; i++) {
-		PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr,
-			PPCLK_DCLK, i, &clock) == 0,
-		"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for DCLK!",
-		return -EINVAL);
-
-		dpm_table->dpm_levels[i].value = clock;
+	if (data->smu_features[GNLD_DPM_UVD].enabled) {
+		ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_DCLK);
+		PP_ASSERT_WITH_CODE(!ret,
+				"[SetupDefaultDpmTable] failed to get dclk dpm levels!",
+				return ret);
+	} else {
+		dpm_table->count = 1;
+		dpm_table->dpm_levels[0].value = data->vbios_boot_state.dclock / 100;
 	}
-
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
 
-	/* Assume there is no headless Vega12 for now */
+	/* dcefclk */
 	dpm_table = &(data->dpm_table.dcef_table);
-
-	PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr,
-		PPCLK_DCEFCLK, &num_levels) == 0,
-		"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for DCEFCLK!",
-		return -EINVAL);
-
-	dpm_table->count = num_levels;
-
-	for (i = 0; i < num_levels; i++) {
-		PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr,
-			PPCLK_DCEFCLK, i, &clock) == 0,
-			"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for DCEFCLK!",
-			return -EINVAL);
-
-		dpm_table->dpm_levels[i].value = clock;
+	if (data->smu_features[GNLD_DPM_DCEFCLK].enabled) {
+		ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_DCEFCLK);
+		PP_ASSERT_WITH_CODE(!ret,
+				"[SetupDefaultDpmTable] failed to get dcefclk dpm levels!",
+				return ret);
+	} else {
+		dpm_table->count = 1;
+		dpm_table->dpm_levels[0].value = data->vbios_boot_state.dcef_clock / 100;
 	}
-
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
 
+	/* pixclk */
 	dpm_table = &(data->dpm_table.pixel_table);
-
-	PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr,
-		PPCLK_PIXCLK, &num_levels) == 0,
-		"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for PIXCLK!",
-		return -EINVAL);
-
-	dpm_table->count = num_levels;
-
-	for (i = 0; i < num_levels; i++) {
-		PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr,
-			PPCLK_PIXCLK, i, &clock) == 0,
-			"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for PIXCLK!",
-			return -EINVAL);
-
-		dpm_table->dpm_levels[i].value = clock;
-	}
-
+	if (data->smu_features[GNLD_DPM_DCEFCLK].enabled) {
+		ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_PIXCLK);
+		PP_ASSERT_WITH_CODE(!ret,
+				"[SetupDefaultDpmTable] failed to get pixclk dpm levels!",
+				return ret);
+	} else
+		dpm_table->count = 0;
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
 
+	/* dispclk */
 	dpm_table = &(data->dpm_table.display_table);
-
-	PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr,
-		PPCLK_DISPCLK, &num_levels) == 0,
-		"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for DISPCLK!",
-		return -EINVAL);
-
-	dpm_table->count = num_levels;
-
-	for (i = 0; i < num_levels; i++) {
-		PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr,
-			PPCLK_DISPCLK, i, &clock) == 0,
-			"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for DISPCLK!",
-			return -EINVAL);
-
-		dpm_table->dpm_levels[i].value = clock;
-	}
-
+	if (data->smu_features[GNLD_DPM_DCEFCLK].enabled) {
+		ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_DISPCLK);
+		PP_ASSERT_WITH_CODE(!ret,
+				"[SetupDefaultDpmTable] failed to get dispclk dpm levels!",
+				return ret);
+	} else
+		dpm_table->count = 0;
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
 
+	/* phyclk */
 	dpm_table = &(data->dpm_table.phy_table);
-
-	PP_ASSERT_WITH_CODE(vega12_get_number_dpm_level(hwmgr,
-		PPCLK_PHYCLK, &num_levels) == 0,
-		"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for PHYCLK!",
-		return -EINVAL);
-
-	dpm_table->count = num_levels;
-
-	for (i = 0; i < num_levels; i++) {
-		PP_ASSERT_WITH_CODE(vega12_get_dpm_frequency_by_index(hwmgr,
-			PPCLK_PHYCLK, i, &clock) == 0,
-			"[SetupDefaultDPMTables] Failed to get DPM levels from SMU for PHYCLK!",
-			return -EINVAL);
-
-		dpm_table->dpm_levels[i].value = clock;
-	}
-
+	if (data->smu_features[GNLD_DPM_DCEFCLK].enabled) {
+		ret = vega12_setup_single_dpm_table(hwmgr, dpm_table, PPCLK_PHYCLK);
+		PP_ASSERT_WITH_CODE(!ret,
+				"[SetupDefaultDpmTable] failed to get phyclk dpm levels!",
+				return ret);
+	} else
+		dpm_table->count = 0;
 	vega12_init_dpm_state(&(dpm_table->dpm_state));
 
 	/* save a copy of the default DPM table */
@@ -794,6 +735,9 @@ static int vega12_init_smc_table(struct pp_hwmgr *hwmgr)
 		data->vbios_boot_state.soc_clock = boot_up_values.ulSocClk;
 		data->vbios_boot_state.dcef_clock = boot_up_values.ulDCEFClk;
 		data->vbios_boot_state.uc_cooling_id = boot_up_values.ucCoolingID;
+		data->vbios_boot_state.eclock = boot_up_values.ulEClk;
+		data->vbios_boot_state.dclock = boot_up_values.ulDClk;
+		data->vbios_boot_state.vclock = boot_up_values.ulVClk;
 		smum_send_msg_to_smc_with_parameter(hwmgr,
 				PPSMC_MSG_SetMinDeepSleepDcefclk,
 			(uint32_t)(data->vbios_boot_state.dcef_clock / 100));
@@ -835,6 +779,21 @@ static int vega12_set_allowed_featuresmask(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
+static void vega12_init_powergate_state(struct pp_hwmgr *hwmgr)
+{
+	struct vega12_hwmgr *data =
+			(struct vega12_hwmgr *)(hwmgr->backend);
+
+	data->uvd_power_gated = true;
+	data->vce_power_gated = true;
+
+	if (data->smu_features[GNLD_DPM_UVD].enabled)
+		data->uvd_power_gated = false;
+
+	if (data->smu_features[GNLD_DPM_VCE].enabled)
+		data->vce_power_gated = false;
+}
+
 static int vega12_enable_all_smu_features(struct pp_hwmgr *hwmgr)
 {
 	struct vega12_hwmgr *data =
@@ -853,12 +812,11 @@ static int vega12_enable_all_smu_features(struct pp_hwmgr *hwmgr)
 			enabled = (features_enabled & data->smu_features[i].smu_feature_bitmap) ? true : false;
 			data->smu_features[i].enabled = enabled;
 			data->smu_features[i].supported = enabled;
-			PP_ASSERT(
-				!data->smu_features[i].allowed || enabled,
-				"[EnableAllSMUFeatures] Enabled feature is different from allowed, expected disabled!");
 		}
 	}
 
+	vega12_init_powergate_state(hwmgr);
+
 	return 0;
 }
 
@@ -914,6 +872,48 @@ static int vega12_power_control_set_level(struct pp_hwmgr *hwmgr)
 	return result;
 }
 
+static int vega12_get_all_clock_ranges_helper(struct pp_hwmgr *hwmgr,
+		PPCLK_e clkid, struct vega12_clock_range *clock)
+{
+	/* AC Max */
+	PP_ASSERT_WITH_CODE(
+		smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetMaxDpmFreq, (clkid << 16)) == 0,
+		"[GetClockRanges] Failed to get max ac clock from SMC!",
+		return -EINVAL);
+	clock->ACMax = smum_get_argument(hwmgr);
+
+	/* AC Min */
+	PP_ASSERT_WITH_CODE(
+		smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetMinDpmFreq, (clkid << 16)) == 0,
+		"[GetClockRanges] Failed to get min ac clock from SMC!",
+		return -EINVAL);
+	clock->ACMin = smum_get_argument(hwmgr);
+
+	/* DC Max */
+	PP_ASSERT_WITH_CODE(
+		smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetDcModeMaxDpmFreq, (clkid << 16)) == 0,
+		"[GetClockRanges] Failed to get max dc clock from SMC!",
+		return -EINVAL);
+	clock->DCMax = smum_get_argument(hwmgr);
+
+	return 0;
+}
+
+static int vega12_get_all_clock_ranges(struct pp_hwmgr *hwmgr)
+{
+	struct vega12_hwmgr *data =
+			(struct vega12_hwmgr *)(hwmgr->backend);
+	uint32_t i;
+
+	for (i = 0; i < PPCLK_COUNT; i++)
+		PP_ASSERT_WITH_CODE(!vega12_get_all_clock_ranges_helper(hwmgr,
+					i, &(data->clk_range[i])),
+				"Failed to get clk range from SMC!",
+				return -EINVAL);
+
+	return 0;
+}
+
 static int vega12_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
 {
 	int tmp_result, result = 0;
@@ -941,6 +941,11 @@ static int vega12_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
 			"Failed to power control set level!",
 			result = tmp_result);
 
+	result = vega12_get_all_clock_ranges(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to get all clock ranges!",
+			return result);
+
 	result = vega12_odn_initialize_default_settings(hwmgr);
 	PP_ASSERT_WITH_CODE(!result,
 			"Failed to power control set level!",
@@ -969,37 +974,173 @@ static uint32_t vega12_find_lowest_dpm_level(
 			break;
 	}
 
+	if (i >= table->count) {
+		i = 0;
+		table->dpm_levels[i].enabled = true;
+	}
+
 	return i;
 }
 
 static uint32_t vega12_find_highest_dpm_level(
 		struct vega12_single_dpm_table *table)
 {
-	uint32_t i = 0;
+	int32_t i = 0;
+	PP_ASSERT_WITH_CODE(table->count <= MAX_REGULAR_DPM_NUMBER,
+			"[FindHighestDPMLevel] DPM Table has too many entries!",
+			return MAX_REGULAR_DPM_NUMBER - 1);
 
-	if (table->count <= MAX_REGULAR_DPM_NUMBER) {
-		for (i = table->count; i > 0; i--) {
-			if (table->dpm_levels[i - 1].enabled)
-				return i - 1;
-		}
-	} else {
-		pr_info("DPM Table Has Too Many Entries!");
-		return MAX_REGULAR_DPM_NUMBER - 1;
+	for (i = table->count - 1; i >= 0; i--) {
+		if (table->dpm_levels[i].enabled)
+			break;
 	}
 
-	return i;
+	if (i < 0) {
+		i = 0;
+		table->dpm_levels[i].enabled = true;
+	}
+
+	return (uint32_t)i;
 }
 
 static int vega12_upload_dpm_min_level(struct pp_hwmgr *hwmgr)
 {
-	return 0;
+	struct vega12_hwmgr *data = hwmgr->backend;
+	uint32_t min_freq;
+	int ret = 0;
+
+	if (data->smu_features[GNLD_DPM_GFXCLK].enabled) {
+		min_freq = data->dpm_table.gfx_table.dpm_state.soft_min_level;
+		PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(
+					hwmgr, PPSMC_MSG_SetSoftMinByFreq,
+					(PPCLK_GFXCLK << 16) | (min_freq & 0xffff))),
+					"Failed to set soft min gfxclk !",
+					return ret);
+	}
+
+	if (data->smu_features[GNLD_DPM_UCLK].enabled) {
+		min_freq = data->dpm_table.mem_table.dpm_state.soft_min_level;
+		PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(
+					hwmgr, PPSMC_MSG_SetSoftMinByFreq,
+					(PPCLK_UCLK << 16) | (min_freq & 0xffff))),
+					"Failed to set soft min memclk !",
+					return ret);
+
+		min_freq = data->dpm_table.mem_table.dpm_state.hard_min_level;
+		PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(
+					hwmgr, PPSMC_MSG_SetHardMinByFreq,
+					(PPCLK_UCLK << 16) | (min_freq & 0xffff))),
+					"Failed to set hard min memclk !",
+					return ret);
+	}
+
+	if (data->smu_features[GNLD_DPM_UVD].enabled) {
+		min_freq = data->dpm_table.vclk_table.dpm_state.soft_min_level;
+
+		PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(
+					hwmgr, PPSMC_MSG_SetSoftMinByFreq,
+					(PPCLK_VCLK << 16) | (min_freq & 0xffff))),
+					"Failed to set soft min vclk!",
+					return ret);
+
+		min_freq = data->dpm_table.dclk_table.dpm_state.soft_min_level;
+
+		PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(
+					hwmgr, PPSMC_MSG_SetSoftMinByFreq,
+					(PPCLK_DCLK << 16) | (min_freq & 0xffff))),
+					"Failed to set soft min dclk!",
+					return ret);
+	}
+
+	if (data->smu_features[GNLD_DPM_VCE].enabled) {
+		min_freq = data->dpm_table.eclk_table.dpm_state.soft_min_level;
+
+		PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(
+					hwmgr, PPSMC_MSG_SetSoftMinByFreq,
+					(PPCLK_ECLK << 16) | (min_freq & 0xffff))),
+					"Failed to set soft min eclk!",
+					return ret);
+	}
+
+	if (data->smu_features[GNLD_DPM_SOCCLK].enabled) {
+		min_freq = data->dpm_table.soc_table.dpm_state.soft_min_level;
+
+		PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(
+					hwmgr, PPSMC_MSG_SetSoftMinByFreq,
+					(PPCLK_SOCCLK << 16) | (min_freq & 0xffff))),
+					"Failed to set soft min socclk!",
+					return ret);
+	}
+
+	return ret;
+
 }
 
 static int vega12_upload_dpm_max_level(struct pp_hwmgr *hwmgr)
 {
-	return 0;
-}
+	struct vega12_hwmgr *data = hwmgr->backend;
+	uint32_t max_freq;
+	int ret = 0;
 
+	if (data->smu_features[GNLD_DPM_GFXCLK].enabled) {
+		max_freq = data->dpm_table.gfx_table.dpm_state.soft_max_level;
+
+		PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(
+					hwmgr, PPSMC_MSG_SetSoftMaxByFreq,
+					(PPCLK_GFXCLK << 16) | (max_freq & 0xffff))),
+					"Failed to set soft max gfxclk!",
+					return ret);
+	}
+
+	if (data->smu_features[GNLD_DPM_UCLK].enabled) {
+		max_freq = data->dpm_table.mem_table.dpm_state.soft_max_level;
+
+		PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(
+					hwmgr, PPSMC_MSG_SetSoftMaxByFreq,
+					(PPCLK_UCLK << 16) | (max_freq & 0xffff))),
+					"Failed to set soft max memclk!",
+					return ret);
+	}
+
+	if (data->smu_features[GNLD_DPM_UVD].enabled) {
+		max_freq = data->dpm_table.vclk_table.dpm_state.soft_max_level;
+
+		PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(
+					hwmgr, PPSMC_MSG_SetSoftMaxByFreq,
+					(PPCLK_VCLK << 16) | (max_freq & 0xffff))),
+					"Failed to set soft max vclk!",
+					return ret);
+
+		max_freq = data->dpm_table.dclk_table.dpm_state.soft_max_level;
+		PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(
+					hwmgr, PPSMC_MSG_SetSoftMaxByFreq,
+					(PPCLK_DCLK << 16) | (max_freq & 0xffff))),
+					"Failed to set soft max dclk!",
+					return ret);
+	}
+
+	if (data->smu_features[GNLD_DPM_VCE].enabled) {
+		max_freq = data->dpm_table.eclk_table.dpm_state.soft_max_level;
+
+		PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(
+					hwmgr, PPSMC_MSG_SetSoftMaxByFreq,
+					(PPCLK_ECLK << 16) | (max_freq & 0xffff))),
+					"Failed to set soft max eclk!",
+					return ret);
+	}
+
+	if (data->smu_features[GNLD_DPM_SOCCLK].enabled) {
+		max_freq = data->dpm_table.soc_table.dpm_state.soft_max_level;
+
+		PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(
+					hwmgr, PPSMC_MSG_SetSoftMaxByFreq,
+					(PPCLK_SOCCLK << 16) | (max_freq & 0xffff))),
+					"Failed to set soft max socclk!",
+					return ret);
+	}
+
+	return ret;
+}
 
 int vega12_enable_disable_vce_dpm(struct pp_hwmgr *hwmgr, bool enable)
 {
@@ -1064,8 +1205,7 @@ static uint32_t vega12_dpm_get_mclk(struct pp_hwmgr *hwmgr, bool low)
 	return (mem_clk * 100);
 }
 
-static int vega12_get_gpu_power(struct pp_hwmgr *hwmgr,
-		struct pp_gpu_power *query)
+static int vega12_get_gpu_power(struct pp_hwmgr *hwmgr, uint32_t *query)
 {
 #if 0
 	uint32_t value;
@@ -1075,9 +1215,9 @@ static int vega12_get_gpu_power(struct pp_hwmgr *hwmgr,
 			"Failed to get current package power!",
 			return -EINVAL);
 
-	vega12_read_arg_from_smc(hwmgr, &value);
+	value = smum_get_argument(hwmgr);
 	/* power value is an integer */
-	query->average_gpu_power = value << 8;
+	*query = value << 8;
 #endif
 	return 0;
 }
@@ -1088,14 +1228,11 @@ static int vega12_get_current_gfx_clk_freq(struct pp_hwmgr *hwmgr, uint32_t *gfx
 
 	*gfx_freq = 0;
 
-	PP_ASSERT_WITH_CODE(
-			smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetDpmClockFreq, (PPCLK_GFXCLK << 16)) == 0,
+	PP_ASSERT_WITH_CODE(smum_send_msg_to_smc_with_parameter(hwmgr,
+			PPSMC_MSG_GetDpmClockFreq, (PPCLK_GFXCLK << 16)) == 0,
 			"[GetCurrentGfxClkFreq] Attempt to get Current GFXCLK Frequency Failed!",
-			return -1);
-	PP_ASSERT_WITH_CODE(
-			vega12_read_arg_from_smc(hwmgr, &gfx_clk) == 0,
-			"[GetCurrentGfxClkFreq] Attempt to read arg from SMC Failed",
-			return -1);
+			return -EINVAL);
+	gfx_clk = smum_get_argument(hwmgr);
 
 	*gfx_freq = gfx_clk * 100;
 
@@ -1111,11 +1248,8 @@ static int vega12_get_current_mclk_freq(struct pp_hwmgr *hwmgr, uint32_t *mclk_f
 	PP_ASSERT_WITH_CODE(
 			smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetDpmClockFreq, (PPCLK_UCLK << 16)) == 0,
 			"[GetCurrentMClkFreq] Attempt to get Current MCLK Frequency Failed!",
-			return -1);
-	PP_ASSERT_WITH_CODE(
-			vega12_read_arg_from_smc(hwmgr, &mem_clk) == 0,
-			"[GetCurrentMClkFreq] Attempt to read arg from SMC Failed",
-			return -1);
+			return -EINVAL);
+	mem_clk = smum_get_argument(hwmgr);
 
 	*mclk_freq = mem_clk * 100;
 
@@ -1132,16 +1266,12 @@ static int vega12_get_current_activity_percent(
 #if 0
 	ret = smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetAverageGfxActivity, 0);
 	if (!ret) {
-		ret = vega12_read_arg_from_smc(hwmgr, &current_activity);
-		if (!ret) {
-			if (current_activity > 100) {
-				PP_ASSERT(false,
-					"[GetCurrentActivityPercent] Activity Percentage Exceeds 100!");
-				current_activity = 100;
-			}
-		} else
+		current_activity = smum_get_argument(hwmgr);
+		if (current_activity > 100) {
 			PP_ASSERT(false,
-				"[GetCurrentActivityPercent] Attempt To Read Average Graphics Activity from SMU Failed!");
+				  "[GetCurrentActivityPercent] Activity Percentage Exceeds 100!");
+			current_activity = 100;
+		}
 	} else
 		PP_ASSERT(false,
 			"[GetCurrentActivityPercent] Attempt To Send Get Average Graphics Activity to SMU Failed!");
@@ -1186,12 +1316,8 @@ static int vega12_read_sensor(struct pp_hwmgr *hwmgr, int idx,
 		*size = 4;
 		break;
 	case AMDGPU_PP_SENSOR_GPU_POWER:
-		if (*size < sizeof(struct pp_gpu_power))
-			ret = -EINVAL;
-		else {
-			*size = sizeof(struct pp_gpu_power);
-			ret = vega12_get_gpu_power(hwmgr, (struct pp_gpu_power *)value);
-		}
+		ret = vega12_get_gpu_power(hwmgr, (uint32_t *)value);
+
 		break;
 	default:
 		ret = -EINVAL;
@@ -1226,7 +1352,6 @@ int vega12_display_clock_voltage_request(struct pp_hwmgr *hwmgr,
 	if (data->smu_features[GNLD_DPM_DCEFCLK].enabled) {
 		switch (clk_type) {
 		case amd_pp_dcef_clock:
-			clk_freq = clock_req->clock_freq_in_khz / 100;
 			clk_select = PPCLK_DCEFCLK;
 			break;
 		case amd_pp_disp_clock:
@@ -1260,27 +1385,22 @@ static int vega12_notify_smc_display_config_after_ps_adjustment(
 {
 	struct vega12_hwmgr *data =
 			(struct vega12_hwmgr *)(hwmgr->backend);
-	uint32_t num_active_disps = 0;
-	struct cgs_display_info info = {0};
 	struct PP_Clocks min_clocks = {0};
 	struct pp_display_clock_request clock_req;
-	uint32_t clk_request;
 
-	info.mode_info = NULL;
-	cgs_get_active_displays_info(hwmgr->device, &info);
-	num_active_disps = info.display_count;
-	if (num_active_disps > 1)
+	if ((hwmgr->display_config->num_display > 1) &&
+		!hwmgr->display_config->multi_monitor_in_sync)
 		vega12_notify_smc_display_change(hwmgr, false);
 	else
 		vega12_notify_smc_display_change(hwmgr, true);
 
-	min_clocks.dcefClock = hwmgr->display_config.min_dcef_set_clk;
-	min_clocks.dcefClockInSR = hwmgr->display_config.min_dcef_deep_sleep_set_clk;
-	min_clocks.memoryClock = hwmgr->display_config.min_mem_set_clock;
+	min_clocks.dcefClock = hwmgr->display_config->min_dcef_set_clk;
+	min_clocks.dcefClockInSR = hwmgr->display_config->min_dcef_deep_sleep_set_clk;
+	min_clocks.memoryClock = hwmgr->display_config->min_mem_set_clock;
 
 	if (data->smu_features[GNLD_DPM_DCEFCLK].supported) {
 		clock_req.clock_type = amd_pp_dcef_clock;
-		clock_req.clock_freq_in_khz = min_clocks.dcefClock;
+		clock_req.clock_freq_in_khz = min_clocks.dcefClock/10;
 		if (!vega12_display_clock_voltage_request(hwmgr, &clock_req)) {
 			if (data->smu_features[GNLD_DS_DCEFCLK].supported)
 				PP_ASSERT_WITH_CODE(
@@ -1294,15 +1414,6 @@ static int vega12_notify_smc_display_config_after_ps_adjustment(
 		}
 	}
 
-	if (data->smu_features[GNLD_DPM_UCLK].enabled) {
-		clk_request = (PPCLK_UCLK << 16) | (min_clocks.memoryClock) / 100;
-		PP_ASSERT_WITH_CODE(
-			smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_SetHardMinByFreq, clk_request) == 0,
-			"[PhwVega12_NotifySMCDisplayConfigAfterPowerStateAdjustment] Attempt to set UCLK HardMin Failed!",
-			return -1);
-		data->dpm_table.mem_table.dpm_state.hard_min_level = min_clocks.memoryClock;
-	}
-
 	return 0;
 }
 
@@ -1311,12 +1422,19 @@ static int vega12_force_dpm_highest(struct pp_hwmgr *hwmgr)
 	struct vega12_hwmgr *data =
 			(struct vega12_hwmgr *)(hwmgr->backend);
 
-	data->smc_state_table.gfx_boot_level =
-	data->smc_state_table.gfx_max_level =
-			vega12_find_highest_dpm_level(&(data->dpm_table.gfx_table));
-	data->smc_state_table.mem_boot_level =
-	data->smc_state_table.mem_max_level =
-			vega12_find_highest_dpm_level(&(data->dpm_table.mem_table));
+	uint32_t soft_level;
+
+	soft_level = vega12_find_highest_dpm_level(&(data->dpm_table.gfx_table));
+
+	data->dpm_table.gfx_table.dpm_state.soft_min_level =
+		data->dpm_table.gfx_table.dpm_state.soft_max_level =
+		data->dpm_table.gfx_table.dpm_levels[soft_level].value;
+
+	soft_level = vega12_find_highest_dpm_level(&(data->dpm_table.mem_table));
+
+	data->dpm_table.mem_table.dpm_state.soft_min_level =
+		data->dpm_table.mem_table.dpm_state.soft_max_level =
+		data->dpm_table.mem_table.dpm_levels[soft_level].value;
 
 	PP_ASSERT_WITH_CODE(!vega12_upload_dpm_min_level(hwmgr),
 			"Failed to upload boot level to highest!",
@@ -1333,13 +1451,19 @@ static int vega12_force_dpm_lowest(struct pp_hwmgr *hwmgr)
 {
 	struct vega12_hwmgr *data =
 			(struct vega12_hwmgr *)(hwmgr->backend);
+	uint32_t soft_level;
 
-	data->smc_state_table.gfx_boot_level =
-	data->smc_state_table.gfx_max_level =
-			vega12_find_lowest_dpm_level(&(data->dpm_table.gfx_table));
-	data->smc_state_table.mem_boot_level =
-	data->smc_state_table.mem_max_level =
-			vega12_find_lowest_dpm_level(&(data->dpm_table.mem_table));
+	soft_level = vega12_find_lowest_dpm_level(&(data->dpm_table.gfx_table));
+
+	data->dpm_table.gfx_table.dpm_state.soft_min_level =
+		data->dpm_table.gfx_table.dpm_state.soft_max_level =
+		data->dpm_table.gfx_table.dpm_levels[soft_level].value;
+
+	soft_level = vega12_find_lowest_dpm_level(&(data->dpm_table.mem_table));
+
+	data->dpm_table.mem_table.dpm_state.soft_min_level =
+		data->dpm_table.mem_table.dpm_state.soft_max_level =
+		data->dpm_table.mem_table.dpm_levels[soft_level].value;
 
 	PP_ASSERT_WITH_CODE(!vega12_upload_dpm_min_level(hwmgr),
 			"Failed to upload boot level to highest!",
@@ -1355,17 +1479,6 @@ static int vega12_force_dpm_lowest(struct pp_hwmgr *hwmgr)
 
 static int vega12_unforce_dpm_levels(struct pp_hwmgr *hwmgr)
 {
-	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
-
-	data->smc_state_table.gfx_boot_level =
-			vega12_find_lowest_dpm_level(&(data->dpm_table.gfx_table));
-	data->smc_state_table.gfx_max_level =
-			vega12_find_highest_dpm_level(&(data->dpm_table.gfx_table));
-	data->smc_state_table.mem_boot_level =
-			vega12_find_lowest_dpm_level(&(data->dpm_table.mem_table));
-	data->smc_state_table.mem_max_level =
-			vega12_find_highest_dpm_level(&(data->dpm_table.mem_table));
-
 	PP_ASSERT_WITH_CODE(!vega12_upload_dpm_min_level(hwmgr),
 			"Failed to upload DPM Bootup Levels!",
 			return -1);
@@ -1373,22 +1486,28 @@ static int vega12_unforce_dpm_levels(struct pp_hwmgr *hwmgr)
 	PP_ASSERT_WITH_CODE(!vega12_upload_dpm_max_level(hwmgr),
 			"Failed to upload DPM Max Levels!",
 			return -1);
+
 	return 0;
 }
 
-#if 0
 static int vega12_get_profiling_clk_mask(struct pp_hwmgr *hwmgr, enum amd_dpm_forced_level level,
 				uint32_t *sclk_mask, uint32_t *mclk_mask, uint32_t *soc_mask)
 {
-	struct phm_ppt_v2_information *table_info =
-			(struct phm_ppt_v2_information *)(hwmgr->pptable);
+	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
+	struct vega12_single_dpm_table *gfx_dpm_table = &(data->dpm_table.gfx_table);
+	struct vega12_single_dpm_table *mem_dpm_table = &(data->dpm_table.mem_table);
+	struct vega12_single_dpm_table *soc_dpm_table = &(data->dpm_table.soc_table);
+
+	*sclk_mask = 0;
+	*mclk_mask = 0;
+	*soc_mask  = 0;
 
-	if (table_info->vdd_dep_on_sclk->count > VEGA12_UMD_PSTATE_GFXCLK_LEVEL &&
-		table_info->vdd_dep_on_socclk->count > VEGA12_UMD_PSTATE_SOCCLK_LEVEL &&
-		table_info->vdd_dep_on_mclk->count > VEGA12_UMD_PSTATE_MCLK_LEVEL) {
+	if (gfx_dpm_table->count > VEGA12_UMD_PSTATE_GFXCLK_LEVEL &&
+	    mem_dpm_table->count > VEGA12_UMD_PSTATE_MCLK_LEVEL &&
+	    soc_dpm_table->count > VEGA12_UMD_PSTATE_SOCCLK_LEVEL) {
 		*sclk_mask = VEGA12_UMD_PSTATE_GFXCLK_LEVEL;
-		*soc_mask = VEGA12_UMD_PSTATE_SOCCLK_LEVEL;
 		*mclk_mask = VEGA12_UMD_PSTATE_MCLK_LEVEL;
+		*soc_mask  = VEGA12_UMD_PSTATE_SOCCLK_LEVEL;
 	}
 
 	if (level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK) {
@@ -1396,13 +1515,13 @@ static int vega12_get_profiling_clk_mask(struct pp_hwmgr *hwmgr, enum amd_dpm_fo
 	} else if (level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK) {
 		*mclk_mask = 0;
 	} else if (level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) {
-		*sclk_mask = table_info->vdd_dep_on_sclk->count - 1;
-		*soc_mask = table_info->vdd_dep_on_socclk->count - 1;
-		*mclk_mask = table_info->vdd_dep_on_mclk->count - 1;
+		*sclk_mask = gfx_dpm_table->count - 1;
+		*mclk_mask = mem_dpm_table->count - 1;
+		*soc_mask  = soc_dpm_table->count - 1;
 	}
+
 	return 0;
 }
-#endif
 
 static void vega12_set_fan_control_mode(struct pp_hwmgr *hwmgr, uint32_t mode)
 {
@@ -1426,11 +1545,9 @@ static int vega12_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 				enum amd_dpm_forced_level level)
 {
 	int ret = 0;
-#if 0
 	uint32_t sclk_mask = 0;
 	uint32_t mclk_mask = 0;
 	uint32_t soc_mask = 0;
-#endif
 
 	switch (level) {
 	case AMD_DPM_FORCED_LEVEL_HIGH:
@@ -1446,27 +1563,18 @@ static int vega12_dpm_force_dpm_level(struct pp_hwmgr *hwmgr,
 	case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK:
 	case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK:
 	case AMD_DPM_FORCED_LEVEL_PROFILE_PEAK:
-#if 0
 		ret = vega12_get_profiling_clk_mask(hwmgr, level, &sclk_mask, &mclk_mask, &soc_mask);
 		if (ret)
 			return ret;
-		vega12_force_clock_level(hwmgr, PP_SCLK, 1<<sclk_mask);
-		vega12_force_clock_level(hwmgr, PP_MCLK, 1<<mclk_mask);
-#endif
+		vega12_force_clock_level(hwmgr, PP_SCLK, 1 << sclk_mask);
+		vega12_force_clock_level(hwmgr, PP_MCLK, 1 << mclk_mask);
 		break;
 	case AMD_DPM_FORCED_LEVEL_MANUAL:
 	case AMD_DPM_FORCED_LEVEL_PROFILE_EXIT:
 	default:
 		break;
 	}
-#if 0
-	if (!ret) {
-		if (level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK && hwmgr->dpm_level != AMD_DPM_FORCED_LEVEL_PROFILE_PEAK)
-			vega12_set_fan_control_mode(hwmgr, AMD_FAN_CTRL_NONE);
-		else if (level != AMD_DPM_FORCED_LEVEL_PROFILE_PEAK && hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK)
-			vega12_set_fan_control_mode(hwmgr, AMD_FAN_CTRL_AUTO);
-	}
-#endif
+
 	return ret;
 }
 
@@ -1500,24 +1608,14 @@ static int vega12_get_clock_ranges(struct pp_hwmgr *hwmgr,
 		PPCLK_e clock_select,
 		bool max)
 {
-	int result;
-	*clock = 0;
+	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
 
-	if (max) {
-		 PP_ASSERT_WITH_CODE(
-			smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetMaxDpmFreq, (clock_select << 16)) == 0,
-			"[GetClockRanges] Failed to get max clock from SMC!",
-			return -1);
-		result = vega12_read_arg_from_smc(hwmgr, clock);
-	} else {
-		PP_ASSERT_WITH_CODE(
-			smum_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_GetMinDpmFreq, (clock_select << 16)) == 0,
-			"[GetClockRanges] Failed to get min clock from SMC!",
-			return -1);
-		result = vega12_read_arg_from_smc(hwmgr, clock);
-	}
+	if (max)
+		*clock = data->clk_range[clock_select].ACMax;
+	else
+		*clock = data->clk_range[clock_select].ACMin;
 
-	return result;
+	return 0;
 }
 
 static int vega12_get_sclks(struct pp_hwmgr *hwmgr,
@@ -1532,12 +1630,12 @@ static int vega12_get_sclks(struct pp_hwmgr *hwmgr,
 		return -1;
 
 	dpm_table = &(data->dpm_table.gfx_table);
-	ucount = (dpm_table->count > VG12_PSUEDO_NUM_GFXCLK_DPM_LEVELS) ?
-		VG12_PSUEDO_NUM_GFXCLK_DPM_LEVELS : dpm_table->count;
+	ucount = (dpm_table->count > MAX_NUM_CLOCKS) ?
+		MAX_NUM_CLOCKS : dpm_table->count;
 
 	for (i = 0; i < ucount; i++) {
 		clocks->data[i].clocks_in_khz =
-			dpm_table->dpm_levels[i].value * 100;
+			dpm_table->dpm_levels[i].value * 1000;
 
 		clocks->data[i].latency_in_us = 0;
 	}
@@ -1564,13 +1662,12 @@ static int vega12_get_memclocks(struct pp_hwmgr *hwmgr,
 		return -1;
 
 	dpm_table = &(data->dpm_table.mem_table);
-	ucount = (dpm_table->count > VG12_PSUEDO_NUM_UCLK_DPM_LEVELS) ?
-		VG12_PSUEDO_NUM_UCLK_DPM_LEVELS : dpm_table->count;
+	ucount = (dpm_table->count > MAX_NUM_CLOCKS) ?
+		MAX_NUM_CLOCKS : dpm_table->count;
 
 	for (i = 0; i < ucount; i++) {
-		clocks->data[i].clocks_in_khz =
-			dpm_table->dpm_levels[i].value * 100;
-
+		clocks->data[i].clocks_in_khz = dpm_table->dpm_levels[i].value * 1000;
+		data->mclk_latency_table.entries[i].frequency = dpm_table->dpm_levels[i].value * 100;
 		clocks->data[i].latency_in_us =
 			data->mclk_latency_table.entries[i].latency =
 			vega12_get_mem_latency(hwmgr, dpm_table->dpm_levels[i].value);
@@ -1594,12 +1691,12 @@ static int vega12_get_dcefclocks(struct pp_hwmgr *hwmgr,
 
 
 	dpm_table = &(data->dpm_table.dcef_table);
-	ucount = (dpm_table->count > VG12_PSUEDO_NUM_DCEFCLK_DPM_LEVELS) ?
-		VG12_PSUEDO_NUM_DCEFCLK_DPM_LEVELS : dpm_table->count;
+	ucount = (dpm_table->count > MAX_NUM_CLOCKS) ?
+		MAX_NUM_CLOCKS : dpm_table->count;
 
 	for (i = 0; i < ucount; i++) {
 		clocks->data[i].clocks_in_khz =
-			dpm_table->dpm_levels[i].value * 100;
+			dpm_table->dpm_levels[i].value * 1000;
 
 		clocks->data[i].latency_in_us = 0;
 	}
@@ -1622,12 +1719,12 @@ static int vega12_get_socclocks(struct pp_hwmgr *hwmgr,
 
 
 	dpm_table = &(data->dpm_table.soc_table);
-	ucount = (dpm_table->count > VG12_PSUEDO_NUM_SOCCLK_DPM_LEVELS) ?
-		VG12_PSUEDO_NUM_SOCCLK_DPM_LEVELS : dpm_table->count;
+	ucount = (dpm_table->count > MAX_NUM_CLOCKS) ?
+		MAX_NUM_CLOCKS : dpm_table->count;
 
 	for (i = 0; i < ucount; i++) {
 		clocks->data[i].clocks_in_khz =
-			dpm_table->dpm_levels[i].value * 100;
+			dpm_table->dpm_levels[i].value * 1000;
 
 		clocks->data[i].latency_in_us = 0;
 	}
@@ -1674,99 +1771,69 @@ static int vega12_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr,
 }
 
 static int vega12_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr,
-		struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges)
+							void *clock_ranges)
 {
 	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
 	Watermarks_t *table = &(data->smc_state_table.water_marks_table);
-	int result = 0;
-	uint32_t i;
+	struct dm_pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges = clock_ranges;
 
 	if (!data->registry_data.disable_water_mark &&
 			data->smu_features[GNLD_DPM_DCEFCLK].supported &&
 			data->smu_features[GNLD_DPM_SOCCLK].supported) {
-		for (i = 0; i < wm_with_clock_ranges->num_wm_sets_dmif; i++) {
-			table->WatermarkRow[WM_DCEFCLK][i].MinClock =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_dmif[i].wm_min_dcefclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_DCEFCLK][i].MaxClock =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_dmif[i].wm_max_dcefclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_DCEFCLK][i].MinUclk =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_dmif[i].wm_min_memclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_DCEFCLK][i].MaxUclk =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_dmif[i].wm_max_memclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_DCEFCLK][i].WmSetting = (uint8_t)
-					wm_with_clock_ranges->wm_sets_dmif[i].wm_set_id;
-		}
-
-		for (i = 0; i < wm_with_clock_ranges->num_wm_sets_mcif; i++) {
-			table->WatermarkRow[WM_SOCCLK][i].MinClock =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_mcif[i].wm_min_socclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_SOCCLK][i].MaxClock =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_mcif[i].wm_max_socclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_SOCCLK][i].MinUclk =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_mcif[i].wm_min_memclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_SOCCLK][i].MaxUclk =
-				cpu_to_le16((uint16_t)
-				(wm_with_clock_ranges->wm_sets_mcif[i].wm_max_memclk_in_khz) /
-				100);
-			table->WatermarkRow[WM_SOCCLK][i].WmSetting = (uint8_t)
-					wm_with_clock_ranges->wm_sets_mcif[i].wm_set_id;
-		}
+		smu_set_watermarks_for_clocks_ranges(table, wm_with_clock_ranges);
 		data->water_marks_bitmap |= WaterMarksExist;
 		data->water_marks_bitmap &= ~WaterMarksLoaded;
 	}
 
-	return result;
+	return 0;
 }
 
 static int vega12_force_clock_level(struct pp_hwmgr *hwmgr,
 		enum pp_clock_type type, uint32_t mask)
 {
 	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
-
-	if (hwmgr->request_dpm_level & (AMD_DPM_FORCED_LEVEL_AUTO |
-				AMD_DPM_FORCED_LEVEL_LOW |
-				AMD_DPM_FORCED_LEVEL_HIGH))
-		return -EINVAL;
+	uint32_t soft_min_level, soft_max_level;
+	int ret = 0;
 
 	switch (type) {
 	case PP_SCLK:
-		data->smc_state_table.gfx_boot_level = mask ? (ffs(mask) - 1) : 0;
-		data->smc_state_table.gfx_max_level = mask ? (fls(mask) - 1) : 0;
+		soft_min_level = mask ? (ffs(mask) - 1) : 0;
+		soft_max_level = mask ? (fls(mask) - 1) : 0;
+
+		data->dpm_table.gfx_table.dpm_state.soft_min_level =
+			data->dpm_table.gfx_table.dpm_levels[soft_min_level].value;
+		data->dpm_table.gfx_table.dpm_state.soft_max_level =
+			data->dpm_table.gfx_table.dpm_levels[soft_max_level].value;
 
-		PP_ASSERT_WITH_CODE(!vega12_upload_dpm_min_level(hwmgr),
+		ret = vega12_upload_dpm_min_level(hwmgr);
+		PP_ASSERT_WITH_CODE(!ret,
 			"Failed to upload boot level to lowest!",
-			return -EINVAL);
+			return ret);
 
-		PP_ASSERT_WITH_CODE(!vega12_upload_dpm_max_level(hwmgr),
+		ret = vega12_upload_dpm_max_level(hwmgr);
+		PP_ASSERT_WITH_CODE(!ret,
 			"Failed to upload dpm max level to highest!",
-			return -EINVAL);
+			return ret);
 		break;
 
 	case PP_MCLK:
-		data->smc_state_table.mem_boot_level = mask ? (ffs(mask) - 1) : 0;
-		data->smc_state_table.mem_max_level = mask ? (fls(mask) - 1) : 0;
+		soft_min_level = mask ? (ffs(mask) - 1) : 0;
+		soft_max_level = mask ? (fls(mask) - 1) : 0;
+
+		data->dpm_table.mem_table.dpm_state.soft_min_level =
+			data->dpm_table.mem_table.dpm_levels[soft_min_level].value;
+		data->dpm_table.mem_table.dpm_state.soft_max_level =
+			data->dpm_table.mem_table.dpm_levels[soft_max_level].value;
 
-		PP_ASSERT_WITH_CODE(!vega12_upload_dpm_min_level(hwmgr),
+		ret = vega12_upload_dpm_min_level(hwmgr);
+		PP_ASSERT_WITH_CODE(!ret,
 			"Failed to upload boot level to lowest!",
-			return -EINVAL);
+			return ret);
 
-		PP_ASSERT_WITH_CODE(!vega12_upload_dpm_max_level(hwmgr),
+		ret = vega12_upload_dpm_max_level(hwmgr);
+		PP_ASSERT_WITH_CODE(!ret,
 			"Failed to upload dpm max level to highest!",
-			return -EINVAL);
+			return ret);
 
 		break;
 
@@ -1799,8 +1866,8 @@ static int vega12_print_clock_levels(struct pp_hwmgr *hwmgr,
 				return -1);
 		for (i = 0; i < clocks.num_levels; i++)
 			size += sprintf(buf + size, "%d: %uMhz %s\n",
-				i, clocks.data[i].clocks_in_khz / 100,
-				(clocks.data[i].clocks_in_khz == now) ? "*" : "");
+				i, clocks.data[i].clocks_in_khz / 1000,
+				(clocks.data[i].clocks_in_khz / 1000 == now / 100) ? "*" : "");
 		break;
 
 	case PP_MCLK:
@@ -1815,8 +1882,8 @@ static int vega12_print_clock_levels(struct pp_hwmgr *hwmgr,
 				return -1);
 		for (i = 0; i < clocks.num_levels; i++)
 			size += sprintf(buf + size, "%d: %uMhz %s\n",
-				i, clocks.data[i].clocks_in_khz / 100,
-				(clocks.data[i].clocks_in_khz == now) ? "*" : "");
+				i, clocks.data[i].clocks_in_khz / 1000,
+				(clocks.data[i].clocks_in_khz / 1000 == now / 100) ? "*" : "");
 		break;
 
 	case PP_PCIE:
@@ -1828,13 +1895,210 @@ static int vega12_print_clock_levels(struct pp_hwmgr *hwmgr,
 	return size;
 }
 
+static int vega12_apply_clocks_adjust_rules(struct pp_hwmgr *hwmgr)
+{
+	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
+	struct vega12_single_dpm_table *dpm_table;
+	bool vblank_too_short = false;
+	bool disable_mclk_switching;
+	uint32_t i, latency;
+
+	disable_mclk_switching = ((1 < hwmgr->display_config->num_display) &&
+			          !hwmgr->display_config->multi_monitor_in_sync) ||
+			          vblank_too_short;
+	latency = hwmgr->display_config->dce_tolerable_mclk_in_active_latency;
+
+	/* gfxclk */
+	dpm_table = &(data->dpm_table.gfx_table);
+	dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value;
+	dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+	dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value;
+	dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+
+	if (PP_CAP(PHM_PlatformCaps_UMDPState)) {
+		if (VEGA12_UMD_PSTATE_GFXCLK_LEVEL < dpm_table->count) {
+			dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_GFXCLK_LEVEL].value;
+			dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_GFXCLK_LEVEL].value;
+		}
+
+		if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK) {
+			dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value;
+			dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[0].value;
+		}
+
+		if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) {
+			dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+			dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+		}
+	}
+
+	/* memclk */
+	dpm_table = &(data->dpm_table.mem_table);
+	dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value;
+	dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+	dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value;
+	dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+
+	if (PP_CAP(PHM_PlatformCaps_UMDPState)) {
+		if (VEGA12_UMD_PSTATE_MCLK_LEVEL < dpm_table->count) {
+			dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_MCLK_LEVEL].value;
+			dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_MCLK_LEVEL].value;
+		}
+
+		if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK) {
+			dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value;
+			dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[0].value;
+		}
+
+		if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) {
+			dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+			dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+		}
+	}
+
+	/* honour DAL's UCLK Hardmin */
+	if (dpm_table->dpm_state.hard_min_level < (hwmgr->display_config->min_mem_set_clock / 100))
+		dpm_table->dpm_state.hard_min_level = hwmgr->display_config->min_mem_set_clock / 100;
+
+	/* Hardmin is dependent on displayconfig */
+	if (disable_mclk_switching) {
+		dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+		for (i = 0; i < data->mclk_latency_table.count - 1; i++) {
+			if (data->mclk_latency_table.entries[i].latency <= latency) {
+				if (dpm_table->dpm_levels[i].value >= (hwmgr->display_config->min_mem_set_clock / 100)) {
+					dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[i].value;
+					break;
+				}
+			}
+		}
+	}
+
+	if (hwmgr->display_config->nb_pstate_switch_disable)
+		dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+
+	/* vclk */
+	dpm_table = &(data->dpm_table.vclk_table);
+	dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value;
+	dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+	dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value;
+	dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+
+	if (PP_CAP(PHM_PlatformCaps_UMDPState)) {
+		if (VEGA12_UMD_PSTATE_UVDCLK_LEVEL < dpm_table->count) {
+			dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_UVDCLK_LEVEL].value;
+			dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_UVDCLK_LEVEL].value;
+		}
+
+		if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) {
+			dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+			dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+		}
+	}
+
+	/* dclk */
+	dpm_table = &(data->dpm_table.dclk_table);
+	dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value;
+	dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+	dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value;
+	dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+
+	if (PP_CAP(PHM_PlatformCaps_UMDPState)) {
+		if (VEGA12_UMD_PSTATE_UVDCLK_LEVEL < dpm_table->count) {
+			dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_UVDCLK_LEVEL].value;
+			dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_UVDCLK_LEVEL].value;
+		}
+
+		if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) {
+			dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+			dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+		}
+	}
+
+	/* socclk */
+	dpm_table = &(data->dpm_table.soc_table);
+	dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value;
+	dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+	dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value;
+	dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+
+	if (PP_CAP(PHM_PlatformCaps_UMDPState)) {
+		if (VEGA12_UMD_PSTATE_SOCCLK_LEVEL < dpm_table->count) {
+			dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_SOCCLK_LEVEL].value;
+			dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_SOCCLK_LEVEL].value;
+		}
+
+		if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) {
+			dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+			dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+		}
+	}
+
+	/* eclk */
+	dpm_table = &(data->dpm_table.eclk_table);
+	dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[0].value;
+	dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+	dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[0].value;
+	dpm_table->dpm_state.hard_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+
+	if (PP_CAP(PHM_PlatformCaps_UMDPState)) {
+		if (VEGA12_UMD_PSTATE_VCEMCLK_LEVEL < dpm_table->count) {
+			dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_VCEMCLK_LEVEL].value;
+			dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[VEGA12_UMD_PSTATE_VCEMCLK_LEVEL].value;
+		}
+
+		if (hwmgr->dpm_level == AMD_DPM_FORCED_LEVEL_PROFILE_PEAK) {
+			dpm_table->dpm_state.soft_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+			dpm_table->dpm_state.soft_max_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+		}
+	}
+
+	return 0;
+}
+
+static int vega12_set_uclk_to_highest_dpm_level(struct pp_hwmgr *hwmgr,
+		struct vega12_single_dpm_table *dpm_table)
+{
+	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
+	int ret = 0;
+
+	if (data->smu_features[GNLD_DPM_UCLK].enabled) {
+		PP_ASSERT_WITH_CODE(dpm_table->count > 0,
+				"[SetUclkToHightestDpmLevel] Dpm table has no entry!",
+				return -EINVAL);
+		PP_ASSERT_WITH_CODE(dpm_table->count <= NUM_UCLK_DPM_LEVELS,
+				"[SetUclkToHightestDpmLevel] Dpm table has too many entries!",
+				return -EINVAL);
+
+		dpm_table->dpm_state.hard_min_level = dpm_table->dpm_levels[dpm_table->count - 1].value;
+		PP_ASSERT_WITH_CODE(!(ret = smum_send_msg_to_smc_with_parameter(hwmgr,
+				PPSMC_MSG_SetHardMinByFreq,
+				(PPCLK_UCLK << 16 ) | dpm_table->dpm_state.hard_min_level)),
+				"[SetUclkToHightestDpmLevel] Set hard min uclk failed!",
+				return ret);
+	}
+
+	return ret;
+}
+
+static int vega12_pre_display_configuration_changed_task(struct pp_hwmgr *hwmgr)
+{
+	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
+	int ret = 0;
+
+	smum_send_msg_to_smc_with_parameter(hwmgr,
+			PPSMC_MSG_NumOfDisplays, 0);
+
+	ret = vega12_set_uclk_to_highest_dpm_level(hwmgr,
+			&data->dpm_table.mem_table);
+
+	return ret;
+}
+
 static int vega12_display_configuration_changed_task(struct pp_hwmgr *hwmgr)
 {
 	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
 	int result = 0;
-	uint32_t num_turned_on_displays = 1;
 	Watermarks_t *wm_table = &(data->smc_state_table.water_marks_table);
-	struct cgs_display_info info = {0};
 
 	if ((data->water_marks_bitmap & WaterMarksExist) &&
 			!(data->water_marks_bitmap & WaterMarksLoaded)) {
@@ -1846,12 +2110,9 @@ static int vega12_display_configuration_changed_task(struct pp_hwmgr *hwmgr)
 
 	if ((data->water_marks_bitmap & WaterMarksExist) &&
 		data->smu_features[GNLD_DPM_DCEFCLK].supported &&
-		data->smu_features[GNLD_DPM_SOCCLK].supported) {
-		cgs_get_active_displays_info(hwmgr->device, &info);
-		num_turned_on_displays = info.display_count;
+		data->smu_features[GNLD_DPM_SOCCLK].supported)
 		smum_send_msg_to_smc_with_parameter(hwmgr,
-			PPSMC_MSG_NumOfDisplays, num_turned_on_displays);
-	}
+			PPSMC_MSG_NumOfDisplays, hwmgr->display_config->num_display);
 
 	return result;
 }
@@ -1877,6 +2138,9 @@ static void vega12_power_gate_vce(struct pp_hwmgr *hwmgr, bool bgate)
 {
 	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
 
+	if (data->vce_power_gated == bgate)
+		return;
+
 	data->vce_power_gated = bgate;
 	vega12_enable_disable_vce_dpm(hwmgr, !bgate);
 }
@@ -1885,6 +2149,9 @@ static void vega12_power_gate_uvd(struct pp_hwmgr *hwmgr, bool bgate)
 {
 	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
 
+	if (data->uvd_power_gated == bgate)
+		return;
+
 	data->uvd_power_gated = bgate;
 	vega12_enable_disable_uvd_dpm(hwmgr, !bgate);
 }
@@ -1894,15 +2161,12 @@ vega12_check_smc_update_required_for_display_configuration(struct pp_hwmgr *hwmg
 {
 	struct vega12_hwmgr *data = (struct vega12_hwmgr *)(hwmgr->backend);
 	bool is_update_required = false;
-	struct cgs_display_info info = {0, 0, NULL};
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-
-	if (data->display_timing.num_existing_displays != info.display_count)
+	if (data->display_timing.num_existing_displays != hwmgr->display_config->num_display)
 		is_update_required = true;
 
 	if (data->registry_data.gfx_clk_deep_sleep_support) {
-		if (data->display_timing.min_clock_in_sr != hwmgr->display_config.min_core_set_clock_in_sr)
+		if (data->display_timing.min_clock_in_sr != hwmgr->display_config->min_core_set_clock_in_sr)
 			is_update_required = true;
 	}
 
@@ -2055,6 +2319,38 @@ static int vega12_get_thermal_temperature_range(struct pp_hwmgr *hwmgr,
 	return 0;
 }
 
+static int vega12_enable_gfx_off(struct pp_hwmgr *hwmgr)
+{
+	struct vega12_hwmgr *data =
+			(struct vega12_hwmgr *)(hwmgr->backend);
+	int ret = 0;
+
+	if (data->gfxoff_controlled_by_driver)
+		ret = smum_send_msg_to_smc(hwmgr, PPSMC_MSG_AllowGfxOff);
+
+	return ret;
+}
+
+static int vega12_disable_gfx_off(struct pp_hwmgr *hwmgr)
+{
+	struct vega12_hwmgr *data =
+			(struct vega12_hwmgr *)(hwmgr->backend);
+	int ret = 0;
+
+	if (data->gfxoff_controlled_by_driver)
+		ret = smum_send_msg_to_smc(hwmgr, PPSMC_MSG_DisallowGfxOff);
+
+	return ret;
+}
+
+static int vega12_gfx_off_control(struct pp_hwmgr *hwmgr, bool enable)
+{
+	if (enable)
+		return vega12_enable_gfx_off(hwmgr);
+	else
+		return vega12_disable_gfx_off(hwmgr);
+}
+
 static const struct pp_hwmgr_func vega12_hwmgr_funcs = {
 	.backend_init = vega12_hwmgr_backend_init,
 	.backend_fini = vega12_hwmgr_backend_fini,
@@ -2082,6 +2378,10 @@ static const struct pp_hwmgr_func vega12_hwmgr_funcs = {
 	.display_clock_voltage_request = vega12_display_clock_voltage_request,
 	.force_clock_level = vega12_force_clock_level,
 	.print_clock_levels = vega12_print_clock_levels,
+	.apply_clocks_adjust_rules =
+		vega12_apply_clocks_adjust_rules,
+	.pre_display_config_changed =
+		vega12_pre_display_configuration_changed_task,
 	.display_config_changed = vega12_display_configuration_changed_task,
 	.powergate_uvd = vega12_power_gate_uvd,
 	.powergate_vce = vega12_power_gate_vce,
@@ -2100,6 +2400,7 @@ static const struct pp_hwmgr_func vega12_hwmgr_funcs = {
 	.get_thermal_temperature_range = vega12_get_thermal_temperature_range,
 	.register_irq_handlers = smu9_register_irq_handlers,
 	.start_thermal_controller = vega12_start_thermal_controller,
+	.powergate_gfx = vega12_gfx_off_control,
 };
 
 int vega12_hwmgr_init(struct pp_hwmgr *hwmgr)
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h
index bc98b1df3b65..b3e424d28994 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_hwmgr.h
@@ -33,7 +33,7 @@
 #define WaterMarksExist  1
 #define WaterMarksLoaded 2
 
-#define VG12_PSUEDO_NUM_GFXCLK_DPM_LEVELS   8
+#define VG12_PSUEDO_NUM_GFXCLK_DPM_LEVELS   16
 #define VG12_PSUEDO_NUM_SOCCLK_DPM_LEVELS   8
 #define VG12_PSUEDO_NUM_DCEFCLK_DPM_LEVELS  8
 #define VG12_PSUEDO_NUM_UCLK_DPM_LEVELS     4
@@ -167,6 +167,9 @@ struct vega12_vbios_boot_state {
 	uint32_t    mem_clock;
 	uint32_t    soc_clock;
 	uint32_t    dcef_clock;
+	uint32_t    eclock;
+	uint32_t    dclock;
+	uint32_t    vclock;
 };
 
 #define DPMTABLE_OD_UPDATE_SCLK     0x00000001
@@ -301,6 +304,12 @@ struct vega12_odn_fan_table {
 	bool		force_fan_pwm;
 };
 
+struct vega12_clock_range {
+	uint32_t	ACMax;
+	uint32_t	ACMin;
+	uint32_t	DCMax;
+};
+
 struct vega12_hwmgr {
 	struct vega12_dpm_table          dpm_table;
 	struct vega12_dpm_table          golden_dpm_table;
@@ -382,6 +391,11 @@ struct vega12_hwmgr {
 	uint32_t                       smu_version;
 	struct smu_features            smu_features[GNLD_FEATURES_MAX];
 	struct vega12_smc_state_table  smc_state_table;
+
+	struct vega12_clock_range      clk_range[PPCLK_COUNT];
+
+	/* ---- Gfxoff ---- */
+	bool                           gfxoff_controlled_by_driver;
 };
 
 #define VEGA12_DPM2_NEAR_TDP_DEC                      10
@@ -432,6 +446,8 @@ struct vega12_hwmgr {
 #define VEGA12_UMD_PSTATE_GFXCLK_LEVEL         0x3
 #define VEGA12_UMD_PSTATE_SOCCLK_LEVEL         0x3
 #define VEGA12_UMD_PSTATE_MCLK_LEVEL           0x2
+#define VEGA12_UMD_PSTATE_UVDCLK_LEVEL         0x3
+#define VEGA12_UMD_PSTATE_VCEMCLK_LEVEL        0x3
 
 int vega12_enable_disable_vce_dpm(struct pp_hwmgr *hwmgr, bool enable);
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c
index b34113f45904..f4f366b26fd1 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_processpptables.c
@@ -51,7 +51,7 @@ static const void *get_powerplay_table(struct pp_hwmgr *hwmgr)
 
 	if (!table_address) {
 		table_address = (ATOM_Vega12_POWERPLAYTABLE *)
-				cgs_atom_get_data_table(hwmgr->device, index,
+				smu_atom_get_data_table(hwmgr->adev, index,
 						&size, &frev, &crev);
 
 		hwmgr->soft_pp_table = table_address;	/*Cache the result in RAM.*/
@@ -224,6 +224,7 @@ static int append_vbios_pptable(struct pp_hwmgr *hwmgr, PPTable_t *ppsmc_pptable
 	ppsmc_pptable->AcgGfxclkSpreadPercent = smc_dpm_table.acggfxclkspreadpercent;
 	ppsmc_pptable->AcgGfxclkSpreadFreq = smc_dpm_table.acggfxclkspreadfreq;
 
+	ppsmc_pptable->Vr2_I2C_address = smc_dpm_table.Vr2_I2C_address;
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_thermal.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_thermal.c
index df0fa815cd6e..904eb2c9155b 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_thermal.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega12_thermal.c
@@ -26,7 +26,7 @@
 #include "vega12_smumgr.h"
 #include "vega12_ppsmc.h"
 #include "vega12_inc.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 #include "pp_debug.h"
 
 static int vega12_get_current_rpm(struct pp_hwmgr *hwmgr, uint32_t *current_rpm)
@@ -34,11 +34,9 @@ static int vega12_get_current_rpm(struct pp_hwmgr *hwmgr, uint32_t *current_rpm)
 	PP_ASSERT_WITH_CODE(!smum_send_msg_to_smc(hwmgr,
 				PPSMC_MSG_GetCurrentRpm),
 			"Attempt to get current RPM from SMC Failed!",
-			return -1);
-	PP_ASSERT_WITH_CODE(!vega12_read_arg_from_smc(hwmgr,
-			current_rpm),
-			"Attempt to read current RPM from SMC Failed!",
-			return -1);
+			return -EINVAL);
+	*current_rpm = smum_get_argument(hwmgr);
+
 	return 0;
 }
 
@@ -147,13 +145,10 @@ int vega12_fan_ctrl_reset_fan_speed_to_default(struct pp_hwmgr *hwmgr)
 */
 int vega12_thermal_get_temperature(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int temp = 0;
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmCG_MULT_THERMAL_STATUS_BASE_IDX,  mmCG_MULT_THERMAL_STATUS);
 
-	temp = cgs_read_register(hwmgr->device, reg);
+	temp = RREG32_SOC15(THM, 0, mmCG_MULT_THERMAL_STATUS);
 
 	temp = (temp & CG_MULT_THERMAL_STATUS__CTF_TEMP_MASK) >>
 			CG_MULT_THERMAL_STATUS__CTF_TEMP__SHIFT;
@@ -175,11 +170,12 @@ int vega12_thermal_get_temperature(struct pp_hwmgr *hwmgr)
 static int vega12_thermal_set_temperature_range(struct pp_hwmgr *hwmgr,
 		struct PP_TemperatureRange *range)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	int low = VEGA12_THERMAL_MINIMUM_ALERT_TEMP *
 			PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
 	int high = VEGA12_THERMAL_MAXIMUM_ALERT_TEMP *
 			PP_TEMPERATURE_UNITS_PER_CENTIGRADES;
-	uint32_t val, reg;
+	uint32_t val;
 
 	if (low < range->min)
 		low = range->min;
@@ -189,18 +185,15 @@ static int vega12_thermal_set_temperature_range(struct pp_hwmgr *hwmgr,
 	if (low > high)
 		return -EINVAL;
 
-	reg = soc15_get_register_offset(THM_HWID, 0,
-			mmTHM_THERMAL_INT_CTRL_BASE_IDX, mmTHM_THERMAL_INT_CTRL);
-
-	val = cgs_read_register(hwmgr->device, reg);
+	val = RREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_CTRL);
 
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, MAX_IH_CREDIT, 5);
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, THERM_IH_HW_ENA, 1);
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTH, (high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
-	val = CGS_REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTL, (low / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, MAX_IH_CREDIT, 5);
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, THERM_IH_HW_ENA, 1);
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTH, (high / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
+	val = REG_SET_FIELD(val, THM_THERMAL_INT_CTRL, DIG_THERM_INTL, (low / PP_TEMPERATURE_UNITS_PER_CENTIGRADES));
 	val = val & (~THM_THERMAL_INT_CTRL__THERM_TRIGGER_MASK_MASK);
 
-	cgs_write_register(hwmgr->device, reg, val);
+	WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_CTRL, val);
 
 	return 0;
 }
@@ -212,15 +205,14 @@ static int vega12_thermal_set_temperature_range(struct pp_hwmgr *hwmgr,
 */
 static int vega12_thermal_enable_alert(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t val = 0;
-	uint32_t reg;
 
 	val |= (1 << THM_THERMAL_INT_ENA__THERM_INTH_CLR__SHIFT);
 	val |= (1 << THM_THERMAL_INT_ENA__THERM_INTL_CLR__SHIFT);
 	val |= (1 << THM_THERMAL_INT_ENA__THERM_TRIGGER_CLR__SHIFT);
 
-	reg = soc15_get_register_offset(THM_HWID, 0, mmTHM_THERMAL_INT_ENA_BASE_IDX, mmTHM_THERMAL_INT_ENA);
-	cgs_write_register(hwmgr->device, reg, val);
+	WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_ENA, val);
 
 	return 0;
 }
@@ -231,10 +223,9 @@ static int vega12_thermal_enable_alert(struct pp_hwmgr *hwmgr)
 */
 int vega12_thermal_disable_alert(struct pp_hwmgr *hwmgr)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
-	reg = soc15_get_register_offset(THM_HWID, 0, mmTHM_THERMAL_INT_ENA_BASE_IDX, mmTHM_THERMAL_INT_ENA);
-	cgs_write_register(hwmgr->device, reg, 0);
+	WREG32_SOC15(THM, 0, mmTHM_THERMAL_INT_ENA, 0);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
index 8b78bbecd1bc..429c9c4322da 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hardwaremanager.h
@@ -377,11 +377,7 @@ struct phm_clocks {
 #define DPMTABLE_UPDATE_SCLK        0x00000004
 #define DPMTABLE_UPDATE_MCLK        0x00000008
 #define DPMTABLE_OD_UPDATE_VDDC     0x00000010
-
-/* To determine if sclk and mclk are in overdrive state */
-#define SCLK_OVERDRIVE_ENABLED           0x00000001
-#define MCLK_OVERDRIVE_ENABLED           0x00000002
-#define VDDC_OVERDRIVE_ENABLED           0x00000010
+#define DPMTABLE_UPDATE_SOCCLK      0x00000020
 
 struct phm_odn_performance_level {
 	uint32_t clock;
@@ -414,7 +410,10 @@ extern int phm_apply_state_adjust_rules(struct pp_hwmgr *hwmgr,
 				   struct pp_power_state *adjusted_ps,
 			     const struct pp_power_state *current_ps);
 
+extern int phm_apply_clock_adjust_rules(struct pp_hwmgr *hwmgr);
+
 extern int phm_force_dpm_levels(struct pp_hwmgr *hwmgr, enum amd_dpm_forced_level level);
+extern int phm_pre_display_configuration_changed(struct pp_hwmgr *hwmgr);
 extern int phm_display_configuration_changed(struct pp_hwmgr *hwmgr);
 extern int phm_notify_smc_display_config_after_ps_adjustment(struct pp_hwmgr *hwmgr);
 extern int phm_register_irq_handlers(struct pp_hwmgr *hwmgr);
@@ -456,7 +455,7 @@ extern int phm_get_clock_by_type_with_voltage(struct pp_hwmgr *hwmgr,
 		enum amd_pp_clock_type type,
 		struct pp_clock_levels_with_voltage *clocks);
 extern int phm_set_watermarks_for_clocks_ranges(struct pp_hwmgr *hwmgr,
-		struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges);
+						void *clock_ranges);
 extern int phm_display_clock_voltage_request(struct pp_hwmgr *hwmgr,
 		struct pp_display_clock_request *clock);
 
diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
index 17f811d181c8..d3d96260f440 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h
@@ -26,7 +26,6 @@
 #include <linux/seq_file.h>
 #include "amd_powerplay.h"
 #include "hardwaremanager.h"
-#include "pp_power_source.h"
 #include "hwmgr_ppt.h"
 #include "ppatomctrl.h"
 #include "hwmgr_ppt.h"
@@ -38,6 +37,8 @@ struct phm_fan_speed_info;
 struct pp_atomctrl_voltage_table;
 
 #define VOLTAGE_SCALE 4
+#define VOLTAGE_VID_OFFSET_SCALE1   625
+#define VOLTAGE_VID_OFFSET_SCALE2   100
 
 enum DISPLAY_GAP {
 	DISPLAY_GAP_VBLANK_OR_WM = 0,   /* Wait for vblank or MCHG watermark. */
@@ -64,24 +65,6 @@ struct vi_dpm_table {
 #define PCIE_PERF_REQ_GEN2         3
 #define PCIE_PERF_REQ_GEN3         4
 
-enum PP_FEATURE_MASK {
-	PP_SCLK_DPM_MASK = 0x1,
-	PP_MCLK_DPM_MASK = 0x2,
-	PP_PCIE_DPM_MASK = 0x4,
-	PP_SCLK_DEEP_SLEEP_MASK = 0x8,
-	PP_POWER_CONTAINMENT_MASK = 0x10,
-	PP_UVD_HANDSHAKE_MASK = 0x20,
-	PP_SMC_VOLTAGE_CONTROL_MASK = 0x40,
-	PP_VBI_TIME_SUPPORT_MASK = 0x80,
-	PP_ULV_MASK = 0x100,
-	PP_ENABLE_GFX_CG_THRU_SMU = 0x200,
-	PP_CLOCK_STRETCH_MASK = 0x400,
-	PP_OD_FUZZY_FAN_CONTROL_MASK = 0x800,
-	PP_SOCCLK_DPM_MASK = 0x1000,
-	PP_DCEFCLK_DPM_MASK = 0x2000,
-	PP_OVERDRIVE_MASK = 0x4000,
-};
-
 enum PHM_BackEnd_Magic {
 	PHM_Dummy_Magic       = 0xAA5555AA,
 	PHM_RV770_Magic       = 0xDCBAABCD,
@@ -211,7 +194,7 @@ struct pp_smumgr_func {
 	int (*request_smu_load_fw)(struct pp_hwmgr  *hwmgr);
 	int (*request_smu_load_specific_fw)(struct pp_hwmgr  *hwmgr,
 					    uint32_t firmware);
-	int (*get_argument)(struct pp_hwmgr  *hwmgr);
+	uint32_t (*get_argument)(struct pp_hwmgr  *hwmgr);
 	int (*send_msg_to_smc)(struct pp_hwmgr  *hwmgr, uint16_t msg);
 	int (*send_msg_to_smc_with_parameter)(struct pp_hwmgr  *hwmgr,
 					  uint16_t msg, uint32_t parameter);
@@ -245,6 +228,8 @@ struct pp_hwmgr_func {
 				struct pp_power_state  *prequest_ps,
 			const struct pp_power_state *pcurrent_ps);
 
+	int (*apply_clocks_adjust_rules)(struct pp_hwmgr *hwmgr);
+
 	int (*force_dpm_level)(struct pp_hwmgr *hw_mgr,
 					enum amd_dpm_forced_level level);
 
@@ -268,6 +253,7 @@ struct pp_hwmgr_func {
 						const void *state);
 	int (*enable_clock_power_gating)(struct pp_hwmgr *hwmgr);
 	int (*notify_smc_display_config_after_ps_adjustment)(struct pp_hwmgr *hwmgr);
+	int (*pre_display_config_changed)(struct pp_hwmgr *hwmgr);
 	int (*display_config_changed)(struct pp_hwmgr *hwmgr);
 	int (*disable_clock_power_gating)(struct pp_hwmgr *hwmgr);
 	int (*update_clock_gatings)(struct pp_hwmgr *hwmgr,
@@ -307,15 +293,15 @@ struct pp_hwmgr_func {
 	int (*get_clock_by_type_with_voltage)(struct pp_hwmgr *hwmgr,
 			enum amd_pp_clock_type type,
 			struct pp_clock_levels_with_voltage *clocks);
-	int (*set_watermarks_for_clocks_ranges)(struct pp_hwmgr *hwmgr,
-			struct pp_wm_sets_with_clock_ranges_soc15 *wm_with_clock_ranges);
+	int (*set_watermarks_for_clocks_ranges)(struct pp_hwmgr *hwmgr, void *clock_ranges);
 	int (*display_clock_voltage_request)(struct pp_hwmgr *hwmgr,
 			struct pp_display_clock_request *clock);
 	int (*get_max_high_clocks)(struct pp_hwmgr *hwmgr, struct amd_pp_simple_clock_info *clocks);
+	int (*gfx_off_control)(struct pp_hwmgr *hwmgr, bool enable);
 	int (*power_off_asic)(struct pp_hwmgr *hwmgr);
 	int (*force_clock_level)(struct pp_hwmgr *hwmgr, enum pp_clock_type type, uint32_t mask);
 	int (*print_clock_levels)(struct pp_hwmgr *hwmgr, enum pp_clock_type type, char *buf);
-	int (*enable_per_cu_power_gating)(struct pp_hwmgr *hwmgr, bool enable);
+	int (*powergate_gfx)(struct pp_hwmgr *hwmgr, bool enable);
 	int (*get_sclk_od)(struct pp_hwmgr *hwmgr);
 	int (*set_sclk_od)(struct pp_hwmgr *hwmgr, uint32_t value);
 	int (*get_mclk_od)(struct pp_hwmgr *hwmgr);
@@ -340,7 +326,8 @@ struct pp_hwmgr_func {
 					enum PP_OD_DPM_TABLE_COMMAND type,
 					long *input, uint32_t size);
 	int (*set_power_limit)(struct pp_hwmgr *hwmgr, uint32_t n);
-	int (*set_mmhub_powergating_by_smu)(struct pp_hwmgr *hwmgr);
+	int (*powergate_mmhub)(struct pp_hwmgr *hwmgr);
+	int (*smus_notify_pwe)(struct pp_hwmgr *hwmgr);
 };
 
 struct pp_table_func {
@@ -718,6 +705,7 @@ struct pp_hwmgr {
 	uint32_t chip_family;
 	uint32_t chip_id;
 	uint32_t smu_version;
+	bool not_vf;
 	bool pm_en;
 	struct mutex smu_lock;
 
@@ -751,7 +739,6 @@ struct pp_hwmgr {
 	const struct pp_table_func *pptable_func;
 
 	struct pp_power_state    *ps;
-	enum pp_power_source  power_source;
 	uint32_t num_ps;
 	struct pp_thermal_controller_info thermal_controller;
 	bool fan_ctrl_is_in_default_mode;
@@ -764,7 +751,7 @@ struct pp_hwmgr {
 	struct pp_power_state    *request_ps;
 	struct pp_power_state    *boot_ps;
 	struct pp_power_state    *uvd_ps;
-	struct amd_pp_display_configuration display_config;
+	const struct amd_pp_display_configuration *display_config;
 	uint32_t feature_mask;
 	bool avfs_supported;
 	/* UMD Pstate */
@@ -782,10 +769,13 @@ struct pp_hwmgr {
 };
 
 int hwmgr_early_init(struct pp_hwmgr *hwmgr);
+int hwmgr_sw_init(struct pp_hwmgr *hwmgr);
+int hwmgr_sw_fini(struct pp_hwmgr *hwmgr);
 int hwmgr_hw_init(struct pp_hwmgr *hwmgr);
 int hwmgr_hw_fini(struct pp_hwmgr *hwmgr);
-int hwmgr_hw_suspend(struct pp_hwmgr *hwmgr);
-int hwmgr_hw_resume(struct pp_hwmgr *hwmgr);
+int hwmgr_suspend(struct pp_hwmgr *hwmgr);
+int hwmgr_resume(struct pp_hwmgr *hwmgr);
+
 int hwmgr_handle_task(struct pp_hwmgr *hwmgr,
 				enum amd_pp_task task_id,
 				enum amd_pm_state_type *user_state);
diff --git a/drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h
index 426bff2aad2b..a2991fa2e6f8 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/rv_ppsmc.h
@@ -75,13 +75,15 @@
 #define PPSMC_MSG_GetMinGfxclkFrequency         0x2C
 #define PPSMC_MSG_GetMaxGfxclkFrequency         0x2D
 #define PPSMC_MSG_SoftReset                     0x2E
+#define PPSMC_MSG_SetGfxCGPG			0x2F
 #define PPSMC_MSG_SetSoftMaxGfxClk              0x30
 #define PPSMC_MSG_SetHardMinGfxClk              0x31
 #define PPSMC_MSG_SetSoftMaxSocclkByFreq        0x32
 #define PPSMC_MSG_SetSoftMaxFclkByFreq          0x33
 #define PPSMC_MSG_SetSoftMaxVcn                 0x34
 #define PPSMC_MSG_PowerGateMmHub                0x35
-#define PPSMC_Message_Count                     0x36
+#define PPSMC_MSG_SetRccPfcPmeRestoreRegister   0x36
+#define PPSMC_Message_Count                     0x37
 
 
 typedef uint16_t PPSMC_Result;
diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu75.h b/drivers/gpu/drm/amd/powerplay/inc/smu75.h
new file mode 100644
index 000000000000..771523001533
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/inc/smu75.h
@@ -0,0 +1,760 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef SMU75_H
+#define SMU75_H
+
+#pragma pack(push, 1)
+
+typedef struct {
+	uint32_t high;
+	uint32_t low;
+} data_64_t;
+
+typedef struct {
+	data_64_t high;
+	data_64_t low;
+} data_128_t;
+
+#define SMU__DGPU_ONLY
+
+#define SMU__NUM_SCLK_DPM_STATE  8
+#define SMU__NUM_MCLK_DPM_LEVELS 4
+#define SMU__NUM_LCLK_DPM_LEVELS 8
+#define SMU__NUM_PCIE_DPM_LEVELS 8
+
+#define SMU7_CONTEXT_ID_SMC        1
+#define SMU7_CONTEXT_ID_VBIOS      2
+
+#define SMU75_MAX_LEVELS_VDDC            16
+#define SMU75_MAX_LEVELS_VDDGFX          16
+#define SMU75_MAX_LEVELS_VDDCI           8
+#define SMU75_MAX_LEVELS_MVDD            4
+
+#define SMU_MAX_SMIO_LEVELS              4
+
+#define SMU75_MAX_LEVELS_GRAPHICS        SMU__NUM_SCLK_DPM_STATE
+#define SMU75_MAX_LEVELS_MEMORY          SMU__NUM_MCLK_DPM_LEVELS
+#define SMU75_MAX_LEVELS_GIO             SMU__NUM_LCLK_DPM_LEVELS
+#define SMU75_MAX_LEVELS_LINK            SMU__NUM_PCIE_DPM_LEVELS
+#define SMU75_MAX_LEVELS_UVD             8
+#define SMU75_MAX_LEVELS_VCE             8
+#define SMU75_MAX_LEVELS_ACP             8
+#define SMU75_MAX_LEVELS_SAMU            8
+#define SMU75_MAX_ENTRIES_SMIO           32
+
+#define DPM_NO_LIMIT 0
+#define DPM_NO_UP 1
+#define DPM_GO_DOWN 2
+#define DPM_GO_UP 3
+
+#define SMU7_FIRST_DPM_GRAPHICS_LEVEL    0
+#define SMU7_FIRST_DPM_MEMORY_LEVEL      0
+
+#define GPIO_CLAMP_MODE_VRHOT      1
+#define GPIO_CLAMP_MODE_THERM      2
+#define GPIO_CLAMP_MODE_DC         4
+
+#define SCRATCH_B_TARG_PCIE_INDEX_SHIFT 0
+#define SCRATCH_B_TARG_PCIE_INDEX_MASK  (0x7<<SCRATCH_B_TARG_PCIE_INDEX_SHIFT)
+#define SCRATCH_B_CURR_PCIE_INDEX_SHIFT 3
+#define SCRATCH_B_CURR_PCIE_INDEX_MASK  (0x7<<SCRATCH_B_CURR_PCIE_INDEX_SHIFT)
+#define SCRATCH_B_TARG_UVD_INDEX_SHIFT  6
+#define SCRATCH_B_TARG_UVD_INDEX_MASK   (0x7<<SCRATCH_B_TARG_UVD_INDEX_SHIFT)
+#define SCRATCH_B_CURR_UVD_INDEX_SHIFT  9
+#define SCRATCH_B_CURR_UVD_INDEX_MASK   (0x7<<SCRATCH_B_CURR_UVD_INDEX_SHIFT)
+#define SCRATCH_B_TARG_VCE_INDEX_SHIFT  12
+#define SCRATCH_B_TARG_VCE_INDEX_MASK   (0x7<<SCRATCH_B_TARG_VCE_INDEX_SHIFT)
+#define SCRATCH_B_CURR_VCE_INDEX_SHIFT  15
+#define SCRATCH_B_CURR_VCE_INDEX_MASK   (0x7<<SCRATCH_B_CURR_VCE_INDEX_SHIFT)
+#define SCRATCH_B_TARG_ACP_INDEX_SHIFT  18
+#define SCRATCH_B_TARG_ACP_INDEX_MASK   (0x7<<SCRATCH_B_TARG_ACP_INDEX_SHIFT)
+#define SCRATCH_B_CURR_ACP_INDEX_SHIFT  21
+#define SCRATCH_B_CURR_ACP_INDEX_MASK   (0x7<<SCRATCH_B_CURR_ACP_INDEX_SHIFT)
+#define SCRATCH_B_TARG_SAMU_INDEX_SHIFT 24
+#define SCRATCH_B_TARG_SAMU_INDEX_MASK  (0x7<<SCRATCH_B_TARG_SAMU_INDEX_SHIFT)
+#define SCRATCH_B_CURR_SAMU_INDEX_SHIFT 27
+#define SCRATCH_B_CURR_SAMU_INDEX_MASK  (0x7<<SCRATCH_B_CURR_SAMU_INDEX_SHIFT)
+
+/* Virtualization Defines */
+#define CG_XDMA_MASK  0x1
+#define CG_XDMA_SHIFT 0
+#define CG_UVD_MASK   0x2
+#define CG_UVD_SHIFT  1
+#define CG_VCE_MASK   0x4
+#define CG_VCE_SHIFT  2
+#define CG_SAMU_MASK  0x8
+#define CG_SAMU_SHIFT 3
+#define CG_GFX_MASK   0x10
+#define CG_GFX_SHIFT  4
+#define CG_SDMA_MASK  0x20
+#define CG_SDMA_SHIFT 5
+#define CG_HDP_MASK   0x40
+#define CG_HDP_SHIFT  6
+#define CG_MC_MASK    0x80
+#define CG_MC_SHIFT   7
+#define CG_DRM_MASK   0x100
+#define CG_DRM_SHIFT  8
+#define CG_ROM_MASK   0x200
+#define CG_ROM_SHIFT  9
+#define CG_BIF_MASK   0x400
+#define CG_BIF_SHIFT  10
+
+#if defined SMU__DGPU_ONLY
+#define SMU75_DTE_ITERATIONS 5
+#define SMU75_DTE_SOURCES 3
+#define SMU75_DTE_SINKS 1
+#define SMU75_NUM_CPU_TES 0
+#define SMU75_NUM_GPU_TES 1
+#define SMU75_NUM_NON_TES 2
+#define SMU75_DTE_FAN_SCALAR_MIN 0x100
+#define SMU75_DTE_FAN_SCALAR_MAX 0x166
+#define SMU75_DTE_FAN_TEMP_MAX 93
+#define SMU75_DTE_FAN_TEMP_MIN 83
+#endif
+#define SMU75_THERMAL_INPUT_LOOP_COUNT 2
+#define SMU75_THERMAL_CLAMP_MODE_COUNT 2
+
+#define EXP_M1_1  93
+#define EXP_M2_1  195759
+#define EXP_B_1   111176531
+
+#define EXP_M1_2  67
+#define EXP_M2_2  153720
+#define EXP_B_2   94415767
+
+#define EXP_M1_3  48
+#define EXP_M2_3  119796
+#define EXP_B_3   79195279
+
+#define EXP_M1_4  550
+#define EXP_M2_4  1484190
+#define EXP_B_4   1051432828
+
+#define EXP_M1_5  394
+#define EXP_M2_5  1143049
+#define EXP_B_5   864288432
+
+struct SMU7_HystController_Data {
+	uint16_t waterfall_up;
+	uint16_t waterfall_down;
+	uint16_t waterfall_limit;
+	uint16_t release_cnt;
+	uint16_t release_limit;
+	uint16_t spare;
+};
+
+typedef struct SMU7_HystController_Data SMU7_HystController_Data;
+
+struct SMU75_PIDController {
+	uint32_t Ki;
+	int32_t LFWindupUpperLim;
+	int32_t LFWindupLowerLim;
+	uint32_t StatePrecision;
+	uint32_t LfPrecision;
+	uint32_t LfOffset;
+	uint32_t MaxState;
+	uint32_t MaxLfFraction;
+	uint32_t StateShift;
+};
+
+typedef struct SMU75_PIDController SMU75_PIDController;
+
+struct SMU7_LocalDpmScoreboard {
+	uint32_t PercentageBusy;
+
+	int32_t  PIDError;
+	int32_t  PIDIntegral;
+	int32_t  PIDOutput;
+
+	uint32_t SigmaDeltaAccum;
+	uint32_t SigmaDeltaOutput;
+	uint32_t SigmaDeltaLevel;
+
+	uint32_t UtilizationSetpoint;
+
+	uint8_t  TdpClampMode;
+	uint8_t  TdcClampMode;
+	uint8_t  ThermClampMode;
+	uint8_t  VoltageBusy;
+
+	int8_t   CurrLevel;
+	int8_t   TargLevel;
+	uint8_t  LevelChangeInProgress;
+	uint8_t  UpHyst;
+
+	uint8_t  DownHyst;
+	uint8_t  VoltageDownHyst;
+	uint8_t  DpmEnable;
+	uint8_t  DpmRunning;
+
+	uint8_t  DpmForce;
+	uint8_t  DpmForceLevel;
+	uint8_t  DisplayWatermark;
+	uint8_t  McArbIndex;
+
+	uint32_t MinimumPerfSclk;
+
+	uint8_t  AcpiReq;
+	uint8_t  AcpiAck;
+	uint8_t  GfxClkSlow;
+	uint8_t  GpioClampMode;
+
+	uint8_t  EnableModeSwitchRLCNotification;
+	uint8_t  EnabledLevelsChange;
+	uint8_t  DteClampMode;
+	uint8_t  FpsClampMode;
+
+	uint16_t LevelResidencyCounters [SMU75_MAX_LEVELS_GRAPHICS];
+	uint16_t LevelSwitchCounters [SMU75_MAX_LEVELS_GRAPHICS];
+
+	void     (*TargetStateCalculator)(uint8_t);
+	void     (*SavedTargetStateCalculator)(uint8_t);
+
+	uint16_t AutoDpmInterval;
+	uint16_t AutoDpmRange;
+
+	uint8_t  FpsEnabled;
+	uint8_t  MaxPerfLevel;
+	uint8_t  AllowLowClkInterruptToHost;
+	uint8_t  FpsRunning;
+
+	uint32_t MaxAllowedFrequency;
+
+	uint32_t FilteredSclkFrequency;
+	uint32_t LastSclkFrequency;
+	uint32_t FilteredSclkFrequencyCnt;
+
+	uint8_t MinPerfLevel;
+#ifdef SMU__FIRMWARE_SCKS_PRESENT__1
+	uint8_t ScksClampMode;
+	uint8_t padding[2];
+#else
+	uint8_t padding[3];
+#endif
+
+	uint16_t FpsAlpha;
+	uint16_t DeltaTime;
+	uint32_t CurrentFps;
+	uint32_t FilteredFps;
+	uint32_t FrameCount;
+	uint32_t FrameCountLast;
+	uint16_t FpsTargetScalar;
+	uint16_t FpsWaterfallLimitScalar;
+	uint16_t FpsAlphaScalar;
+	uint16_t spare8;
+	SMU7_HystController_Data HystControllerData;
+};
+
+typedef struct SMU7_LocalDpmScoreboard SMU7_LocalDpmScoreboard;
+
+#define SMU7_MAX_VOLTAGE_CLIENTS 12
+
+typedef uint8_t (*VoltageChangeHandler_t)(uint16_t, uint8_t);
+
+#define VDDC_MASK    0x00007FFF
+#define VDDC_SHIFT   0
+#define VDDCI_MASK   0x3FFF8000
+#define VDDCI_SHIFT  15
+#define PHASES_MASK  0xC0000000
+#define PHASES_SHIFT 30
+
+typedef uint32_t SMU_VoltageLevel;
+
+struct SMU7_VoltageScoreboard {
+	SMU_VoltageLevel TargetVoltage;
+	uint16_t MaxVid;
+	uint8_t  HighestVidOffset;
+	uint8_t  CurrentVidOffset;
+
+	uint16_t CurrentVddc;
+	uint16_t CurrentVddci;
+
+	uint8_t  ControllerBusy;
+	uint8_t  CurrentVid;
+	uint8_t  CurrentVddciVid;
+	uint8_t  padding;
+
+	SMU_VoltageLevel RequestedVoltage[SMU7_MAX_VOLTAGE_CLIENTS];
+	SMU_VoltageLevel TargetVoltageState;
+	uint8_t  EnabledRequest[SMU7_MAX_VOLTAGE_CLIENTS];
+
+	uint8_t  padding2;
+	uint8_t  padding3;
+	uint8_t  ControllerEnable;
+	uint8_t  ControllerRunning;
+	uint16_t CurrentStdVoltageHiSidd;
+	uint16_t CurrentStdVoltageLoSidd;
+	uint8_t  OverrideVoltage;
+	uint8_t  padding4;
+	uint8_t  padding5;
+	uint8_t  CurrentPhases;
+
+	VoltageChangeHandler_t ChangeVddc;
+	VoltageChangeHandler_t ChangeVddci;
+	VoltageChangeHandler_t ChangePhase;
+	VoltageChangeHandler_t ChangeMvdd;
+
+	VoltageChangeHandler_t functionLinks[6];
+
+	uint16_t * VddcFollower1;
+	int16_t  Driver_OD_RequestedVidOffset1;
+	int16_t  Driver_OD_RequestedVidOffset2;
+};
+
+typedef struct SMU7_VoltageScoreboard SMU7_VoltageScoreboard;
+
+#define SMU7_MAX_PCIE_LINK_SPEEDS 3
+
+struct SMU7_PCIeLinkSpeedScoreboard {
+	uint8_t     DpmEnable;
+	uint8_t     DpmRunning;
+	uint8_t     DpmForce;
+	uint8_t     DpmForceLevel;
+
+	uint8_t     CurrentLinkSpeed;
+	uint8_t     EnabledLevelsChange;
+	uint16_t    AutoDpmInterval;
+
+	uint16_t    AutoDpmRange;
+	uint16_t    AutoDpmCount;
+
+	uint8_t     DpmMode;
+	uint8_t     AcpiReq;
+	uint8_t     AcpiAck;
+	uint8_t     CurrentLinkLevel;
+};
+
+typedef struct SMU7_PCIeLinkSpeedScoreboard SMU7_PCIeLinkSpeedScoreboard;
+
+#define SMU7_LKGE_LUT_NUM_OF_TEMP_ENTRIES 16
+#define SMU7_LKGE_LUT_NUM_OF_VOLT_ENTRIES 16
+
+#define SMU7_SCALE_I  7
+#define SMU7_SCALE_R 12
+
+struct SMU7_PowerScoreboard {
+	uint32_t GpuPower;
+
+	uint32_t VddcPower;
+	uint32_t VddcVoltage;
+	uint32_t VddcCurrent;
+
+	uint32_t VddciPower;
+	uint32_t VddciVoltage;
+	uint32_t VddciCurrent;
+
+	uint32_t RocPower;
+
+	uint16_t Telemetry_1_slope;
+	uint16_t Telemetry_2_slope;
+	int32_t  Telemetry_1_offset;
+	int32_t  Telemetry_2_offset;
+
+	uint8_t MCLK_patch_flag;
+	uint8_t reserved[3];
+};
+
+typedef struct SMU7_PowerScoreboard SMU7_PowerScoreboard;
+
+#define SMU7_SCLK_DPM_CONFIG_MASK                        0x01
+#define SMU7_VOLTAGE_CONTROLLER_CONFIG_MASK              0x02
+#define SMU7_THERMAL_CONTROLLER_CONFIG_MASK              0x04
+#define SMU7_MCLK_DPM_CONFIG_MASK                        0x08
+#define SMU7_UVD_DPM_CONFIG_MASK                         0x10
+#define SMU7_VCE_DPM_CONFIG_MASK                         0x20
+#define SMU7_ACP_DPM_CONFIG_MASK                         0x40
+#define SMU7_SAMU_DPM_CONFIG_MASK                        0x80
+#define SMU7_PCIEGEN_DPM_CONFIG_MASK                    0x100
+
+#define SMU7_ACP_MCLK_HANDSHAKE_DISABLE                  0x00000001
+#define SMU7_ACP_SCLK_HANDSHAKE_DISABLE                  0x00000002
+#define SMU7_UVD_MCLK_HANDSHAKE_DISABLE                  0x00000100
+#define SMU7_UVD_SCLK_HANDSHAKE_DISABLE                  0x00000200
+#define SMU7_VCE_MCLK_HANDSHAKE_DISABLE                  0x00010000
+#define SMU7_VCE_SCLK_HANDSHAKE_DISABLE                  0x00020000
+
+struct SMU75_SoftRegisters {
+	uint32_t        RefClockFrequency;
+	uint32_t        PmTimerPeriod;
+	uint32_t        FeatureEnables;
+#if defined (SMU__DGPU_ONLY)
+	uint32_t        PreVBlankGap;
+	uint32_t        VBlankTimeout;
+	uint32_t        TrainTimeGap;
+	uint32_t        MvddSwitchTime;
+	uint32_t        LongestAcpiTrainTime;
+	uint32_t        AcpiDelay;
+	uint32_t        G5TrainTime;
+	uint32_t        DelayMpllPwron;
+	uint32_t        VoltageChangeTimeout;
+#endif
+	uint32_t        HandshakeDisables;
+
+	uint8_t         DisplayPhy1Config;
+	uint8_t         DisplayPhy2Config;
+	uint8_t         DisplayPhy3Config;
+	uint8_t         DisplayPhy4Config;
+
+	uint8_t         DisplayPhy5Config;
+	uint8_t         DisplayPhy6Config;
+	uint8_t         DisplayPhy7Config;
+	uint8_t         DisplayPhy8Config;
+
+	uint32_t        AverageGraphicsActivity;
+	uint32_t        AverageMemoryActivity;
+	uint32_t        AverageGioActivity;
+
+	uint8_t         SClkDpmEnabledLevels;
+	uint8_t         MClkDpmEnabledLevels;
+	uint8_t         LClkDpmEnabledLevels;
+	uint8_t         PCIeDpmEnabledLevels;
+
+	uint8_t         UVDDpmEnabledLevels;
+	uint8_t         SAMUDpmEnabledLevels;
+	uint8_t         ACPDpmEnabledLevels;
+	uint8_t         VCEDpmEnabledLevels;
+
+	uint32_t        DRAM_LOG_ADDR_H;
+	uint32_t        DRAM_LOG_ADDR_L;
+	uint32_t        DRAM_LOG_PHY_ADDR_H;
+	uint32_t        DRAM_LOG_PHY_ADDR_L;
+	uint32_t        DRAM_LOG_BUFF_SIZE;
+	uint32_t        UlvEnterCount;
+	uint32_t        UlvTime;
+	uint32_t        UcodeLoadStatus;
+	uint32_t        AllowMvddSwitch;
+	uint8_t         Activity_Weight;
+	uint8_t         Reserved8[3];
+};
+
+typedef struct SMU75_SoftRegisters SMU75_SoftRegisters;
+
+struct SMU75_Firmware_Header {
+	uint32_t Digest[5];
+	uint32_t Version;
+	uint32_t HeaderSize;
+	uint32_t Flags;
+	uint32_t EntryPoint;
+	uint32_t CodeSize;
+	uint32_t ImageSize;
+
+	uint32_t Rtos;
+	uint32_t SoftRegisters;
+	uint32_t DpmTable;
+	uint32_t FanTable;
+	uint32_t CacConfigTable;
+	uint32_t CacStatusTable;
+	uint32_t mcRegisterTable;
+	uint32_t mcArbDramTimingTable;
+	uint32_t PmFuseTable;
+	uint32_t Globals;
+	uint32_t ClockStretcherTable;
+	uint32_t VftTable;
+	uint32_t Reserved1;
+	uint32_t AvfsCksOff_AvfsGbvTable;
+	uint32_t AvfsCksOff_BtcGbvTable;
+	uint32_t MM_AvfsTable;
+	uint32_t PowerSharingTable;
+	uint32_t AvfsTable;
+	uint32_t AvfsCksOffGbvTable;
+	uint32_t AvfsMeanNSigma;
+	uint32_t AvfsSclkOffsetTable;
+	uint32_t Reserved[12];
+	uint32_t Signature;
+};
+
+typedef struct SMU75_Firmware_Header SMU75_Firmware_Header;
+
+#define SMU7_FIRMWARE_HEADER_LOCATION 0x20000
+
+enum  DisplayConfig {
+	PowerDown = 1,
+	DP54x4,
+	DP54x2,
+	DP54x1,
+	DP27x4,
+	DP27x2,
+	DP27x1,
+	HDMI297,
+	HDMI162,
+	LVDS,
+	DP324x4,
+	DP324x2,
+	DP324x1
+};
+
+#define MC_BLOCK_COUNT 1
+#define CPL_BLOCK_COUNT 5
+#define SE_BLOCK_COUNT 15
+#define GC_BLOCK_COUNT 24
+
+struct SMU7_Local_Cac {
+	uint8_t BlockId;
+	uint8_t SignalId;
+	uint8_t Threshold;
+	uint8_t Padding;
+};
+
+typedef struct SMU7_Local_Cac SMU7_Local_Cac;
+
+struct SMU7_Local_Cac_Table {
+	SMU7_Local_Cac CplLocalCac[CPL_BLOCK_COUNT];
+	SMU7_Local_Cac McLocalCac[MC_BLOCK_COUNT];
+	SMU7_Local_Cac SeLocalCac[SE_BLOCK_COUNT];
+	SMU7_Local_Cac GcLocalCac[GC_BLOCK_COUNT];
+};
+
+typedef struct SMU7_Local_Cac_Table SMU7_Local_Cac_Table;
+
+#pragma pack(pop)
+
+#define CG_SYS_BITMASK_FIRST_BIT      0
+#define CG_SYS_BITMASK_LAST_BIT       10
+#define CG_SYS_BIF_MGLS_SHIFT         0
+#define CG_SYS_ROM_SHIFT              1
+#define CG_SYS_MC_MGCG_SHIFT          2
+#define CG_SYS_MC_MGLS_SHIFT          3
+#define CG_SYS_SDMA_MGCG_SHIFT        4
+#define CG_SYS_SDMA_MGLS_SHIFT        5
+#define CG_SYS_DRM_MGCG_SHIFT         6
+#define CG_SYS_HDP_MGCG_SHIFT         7
+#define CG_SYS_HDP_MGLS_SHIFT         8
+#define CG_SYS_DRM_MGLS_SHIFT         9
+#define CG_SYS_BIF_MGCG_SHIFT         10
+
+#define CG_SYS_BIF_MGLS_MASK          0x1
+#define CG_SYS_ROM_MASK               0x2
+#define CG_SYS_MC_MGCG_MASK           0x4
+#define CG_SYS_MC_MGLS_MASK           0x8
+#define CG_SYS_SDMA_MGCG_MASK         0x10
+#define CG_SYS_SDMA_MGLS_MASK         0x20
+#define CG_SYS_DRM_MGCG_MASK          0x40
+#define CG_SYS_HDP_MGCG_MASK          0x80
+#define CG_SYS_HDP_MGLS_MASK          0x100
+#define CG_SYS_DRM_MGLS_MASK          0x200
+#define CG_SYS_BIF_MGCG_MASK          0x400
+
+#define CG_GFX_BITMASK_FIRST_BIT      16
+#define CG_GFX_BITMASK_LAST_BIT       24
+
+#define CG_GFX_CGCG_SHIFT             16
+#define CG_GFX_CGLS_SHIFT             17
+#define CG_CPF_MGCG_SHIFT             18
+#define CG_RLC_MGCG_SHIFT             19
+#define CG_GFX_OTHERS_MGCG_SHIFT      20
+#define CG_GFX_3DCG_SHIFT             21
+#define CG_GFX_3DLS_SHIFT             22
+#define CG_GFX_RLC_LS_SHIFT           23
+#define CG_GFX_CP_LS_SHIFT            24
+
+#define CG_GFX_CGCG_MASK              0x00010000
+#define CG_GFX_CGLS_MASK              0x00020000
+#define CG_CPF_MGCG_MASK              0x00040000
+#define CG_RLC_MGCG_MASK              0x00080000
+#define CG_GFX_OTHERS_MGCG_MASK       0x00100000
+#define CG_GFX_3DCG_MASK              0x00200000
+#define CG_GFX_3DLS_MASK              0x00400000
+#define CG_GFX_RLC_LS_MASK            0x00800000
+#define CG_GFX_CP_LS_MASK             0x01000000
+
+
+#define VRCONF_VDDC_MASK         0x000000FF
+#define VRCONF_VDDC_SHIFT        0
+#define VRCONF_VDDGFX_MASK       0x0000FF00
+#define VRCONF_VDDGFX_SHIFT      8
+#define VRCONF_VDDCI_MASK        0x00FF0000
+#define VRCONF_VDDCI_SHIFT       16
+#define VRCONF_MVDD_MASK         0xFF000000
+#define VRCONF_MVDD_SHIFT        24
+
+#define VR_MERGED_WITH_VDDC      0
+#define VR_SVI2_PLANE_1          1
+#define VR_SVI2_PLANE_2          2
+#define VR_SMIO_PATTERN_1        3
+#define VR_SMIO_PATTERN_2        4
+#define VR_STATIC_VOLTAGE        5
+
+#define CLOCK_STRETCHER_MAX_ENTRIES 0x4
+#define CKS_LOOKUPTable_MAX_ENTRIES 0x4
+
+#define CLOCK_STRETCHER_SETTING_DDT_MASK             0x01
+#define CLOCK_STRETCHER_SETTING_DDT_SHIFT            0x0
+#define CLOCK_STRETCHER_SETTING_STRETCH_AMOUNT_MASK  0x1E
+#define CLOCK_STRETCHER_SETTING_STRETCH_AMOUNT_SHIFT 0x1
+#define CLOCK_STRETCHER_SETTING_ENABLE_MASK          0x80
+#define CLOCK_STRETCHER_SETTING_ENABLE_SHIFT         0x7
+
+struct SMU_ClockStretcherDataTableEntry {
+	uint8_t minVID;
+	uint8_t maxVID;
+
+	uint16_t setting;
+};
+typedef struct SMU_ClockStretcherDataTableEntry SMU_ClockStretcherDataTableEntry;
+
+struct SMU_ClockStretcherDataTable {
+	SMU_ClockStretcherDataTableEntry ClockStretcherDataTableEntry[CLOCK_STRETCHER_MAX_ENTRIES];
+};
+typedef struct SMU_ClockStretcherDataTable SMU_ClockStretcherDataTable;
+
+struct SMU_CKS_LOOKUPTableEntry {
+	uint16_t minFreq;
+	uint16_t maxFreq;
+
+	uint8_t setting;
+	uint8_t padding[3];
+};
+typedef struct SMU_CKS_LOOKUPTableEntry SMU_CKS_LOOKUPTableEntry;
+
+struct SMU_CKS_LOOKUPTable {
+	SMU_CKS_LOOKUPTableEntry CKS_LOOKUPTableEntry[CKS_LOOKUPTable_MAX_ENTRIES];
+};
+typedef struct SMU_CKS_LOOKUPTable SMU_CKS_LOOKUPTable;
+
+struct AgmAvfsData_t {
+	uint16_t avgPsmCount[28];
+	uint16_t minPsmCount[28];
+};
+typedef struct AgmAvfsData_t AgmAvfsData_t;
+
+enum VFT_COLUMNS {
+	SCLK0,
+	SCLK1,
+	SCLK2,
+	SCLK3,
+	SCLK4,
+	SCLK5,
+	SCLK6,
+	SCLK7,
+
+	NUM_VFT_COLUMNS
+};
+enum {
+  SCS_FUSE_T0,
+  SCS_FUSE_T1,
+  NUM_SCS_FUSE_TEMPERATURE
+};
+enum {
+  SCKS_ON,
+  SCKS_OFF,
+  NUM_SCKS_STATE_TYPES
+};
+
+#define VFT_TABLE_DEFINED
+
+#define TEMP_RANGE_MAXSTEPS 12
+struct VFT_CELL_t {
+	uint16_t Voltage;
+};
+
+typedef struct VFT_CELL_t VFT_CELL_t;
+#ifdef SMU__FIRMWARE_SCKS_PRESENT__1
+struct SCS_CELL_t {
+	uint16_t PsmCnt[NUM_SCKS_STATE_TYPES];
+};
+typedef struct SCS_CELL_t SCS_CELL_t;
+#endif
+
+struct VFT_TABLE_t {
+	VFT_CELL_t    Cell[TEMP_RANGE_MAXSTEPS][NUM_VFT_COLUMNS];
+	uint16_t      AvfsGbv [NUM_VFT_COLUMNS];
+	uint16_t      BtcGbv  [NUM_VFT_COLUMNS];
+	int16_t       Temperature [TEMP_RANGE_MAXSTEPS];
+
+#ifdef SMU__FIRMWARE_SCKS_PRESENT__1
+	SCS_CELL_t    ScksCell[TEMP_RANGE_MAXSTEPS][NUM_VFT_COLUMNS];
+#endif
+
+	uint8_t       NumTemperatureSteps;
+	uint8_t       padding[3];
+};
+typedef struct VFT_TABLE_t VFT_TABLE_t;
+
+#define BTCGB_VDROOP_TABLE_MAX_ENTRIES 2
+#define AVFSGB_VDROOP_TABLE_MAX_ENTRIES 2
+
+struct GB_VDROOP_TABLE_t {
+	int32_t a0;
+	int32_t a1;
+	int32_t a2;
+	uint32_t spare;
+};
+typedef struct GB_VDROOP_TABLE_t GB_VDROOP_TABLE_t;
+
+struct SMU_QuadraticCoeffs {
+	int32_t m1;
+	int32_t b;
+
+	int16_t m2;
+	uint8_t m1_shift;
+	uint8_t m2_shift;
+};
+typedef struct SMU_QuadraticCoeffs SMU_QuadraticCoeffs;
+
+struct AVFS_Margin_t {
+	VFT_CELL_t Cell[NUM_VFT_COLUMNS];
+};
+typedef struct AVFS_Margin_t AVFS_Margin_t;
+
+struct AVFS_CksOff_Gbv_t {
+	VFT_CELL_t Cell[NUM_VFT_COLUMNS];
+};
+typedef struct AVFS_CksOff_Gbv_t AVFS_CksOff_Gbv_t;
+
+struct AVFS_CksOff_AvfsGbv_t {
+	VFT_CELL_t Cell[NUM_VFT_COLUMNS];
+};
+typedef struct AVFS_CksOff_AvfsGbv_t AVFS_CksOff_AvfsGbv_t;
+
+struct AVFS_CksOff_BtcGbv_t {
+	VFT_CELL_t Cell[NUM_VFT_COLUMNS];
+};
+typedef struct AVFS_CksOff_BtcGbv_t AVFS_CksOff_BtcGbv_t;
+
+struct AVFS_meanNsigma_t {
+	uint32_t Aconstant[3];
+	uint16_t DC_tol_sigma;
+	uint16_t Platform_mean;
+	uint16_t Platform_sigma;
+	uint16_t PSM_Age_CompFactor;
+	uint8_t  Static_Voltage_Offset[NUM_VFT_COLUMNS];
+};
+typedef struct AVFS_meanNsigma_t AVFS_meanNsigma_t;
+
+struct AVFS_Sclk_Offset_t {
+	uint16_t Sclk_Offset[8];
+};
+typedef struct AVFS_Sclk_Offset_t AVFS_Sclk_Offset_t;
+
+struct Power_Sharing_t {
+	uint32_t EnergyCounter;
+	uint32_t EngeryThreshold;
+	uint64_t AM_SCLK_CNT;
+	uint64_t AM_0_BUSY_CNT;
+};
+typedef struct Power_Sharing_t  Power_Sharing_t;
+
+
+#endif
+
+
diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu75_discrete.h b/drivers/gpu/drm/amd/powerplay/inc/smu75_discrete.h
new file mode 100644
index 000000000000..b64e58a22ddf
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/inc/smu75_discrete.h
@@ -0,0 +1,886 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef SMU75_DISCRETE_H
+#define SMU75_DISCRETE_H
+
+#include "smu75.h"
+
+#pragma pack(push, 1)
+
+#define NUM_SCLK_RANGE 8
+
+#define VCO_3_6 1
+#define VCO_2_4 3
+
+#define POSTDIV_DIV_BY_1  0
+#define POSTDIV_DIV_BY_2  1
+#define POSTDIV_DIV_BY_4  2
+#define POSTDIV_DIV_BY_8  3
+#define POSTDIV_DIV_BY_16 4
+
+struct sclkFcwRange_t {
+	uint8_t  vco_setting; /* 1: 3-6GHz, 3: 2-4GHz */
+	uint8_t  postdiv;     /* divide by 2^n */
+	uint16_t fcw_pcc;
+	uint16_t fcw_trans_upper;
+	uint16_t fcw_trans_lower;
+};
+typedef struct sclkFcwRange_t sclkFcwRange_t;
+
+struct SMIO_Pattern {
+	uint16_t Voltage;
+	uint8_t  Smio;
+	uint8_t  padding;
+};
+
+typedef struct SMIO_Pattern SMIO_Pattern;
+
+struct SMIO_Table {
+	SMIO_Pattern Pattern[SMU_MAX_SMIO_LEVELS];
+};
+
+typedef struct SMIO_Table SMIO_Table;
+
+struct SMU_SclkSetting {
+	uint32_t    SclkFrequency;
+	uint16_t    Fcw_int;
+	uint16_t    Fcw_frac;
+	uint16_t    Pcc_fcw_int;
+	uint8_t     PllRange;
+	uint8_t     SSc_En;
+	uint16_t    Sclk_slew_rate;
+	uint16_t    Pcc_up_slew_rate;
+	uint16_t    Pcc_down_slew_rate;
+	uint16_t    Fcw1_int;
+	uint16_t    Fcw1_frac;
+	uint16_t    Sclk_ss_slew_rate;
+};
+typedef struct SMU_SclkSetting SMU_SclkSetting;
+
+struct SMU75_Discrete_GraphicsLevel {
+	SMU_VoltageLevel MinVoltage;
+
+	uint8_t     pcieDpmLevel;
+	uint8_t     DeepSleepDivId;
+	uint16_t    ActivityLevel;
+
+	uint32_t    CgSpllFuncCntl3;
+	uint32_t    CgSpllFuncCntl4;
+	uint32_t    CcPwrDynRm;
+	uint32_t    CcPwrDynRm1;
+
+	uint8_t     SclkDid;
+	uint8_t     padding;
+	uint8_t     EnabledForActivity;
+	uint8_t     EnabledForThrottle;
+	uint8_t     UpHyst;
+	uint8_t     DownHyst;
+	uint8_t     VoltageDownHyst;
+	uint8_t     PowerThrottle;
+
+	SMU_SclkSetting SclkSetting;
+
+	uint8_t  ScksStretchThreshVid[NUM_SCKS_STATE_TYPES];
+	uint16_t Padding;
+};
+
+typedef struct SMU75_Discrete_GraphicsLevel SMU75_Discrete_GraphicsLevel;
+
+struct SMU75_Discrete_ACPILevel {
+	uint32_t    Flags;
+	SMU_VoltageLevel MinVoltage;
+	uint32_t    SclkFrequency;
+	uint8_t     SclkDid;
+	uint8_t     DisplayWatermark;
+	uint8_t     DeepSleepDivId;
+	uint8_t     padding;
+	uint32_t    CcPwrDynRm;
+	uint32_t    CcPwrDynRm1;
+
+	SMU_SclkSetting SclkSetting;
+};
+
+typedef struct SMU75_Discrete_ACPILevel SMU75_Discrete_ACPILevel;
+
+struct SMU75_Discrete_Ulv {
+	uint32_t    CcPwrDynRm;
+	uint32_t    CcPwrDynRm1;
+	uint16_t    VddcOffset;
+	uint8_t     VddcOffsetVid;
+	uint8_t     VddcPhase;
+	uint16_t    BifSclkDfs;
+	uint16_t    Reserved;
+};
+
+typedef struct SMU75_Discrete_Ulv SMU75_Discrete_Ulv;
+
+struct SMU75_Discrete_MemoryLevel {
+	SMU_VoltageLevel MinVoltage;
+	uint32_t    MinMvdd;
+
+	uint32_t    MclkFrequency;
+
+	uint8_t     StutterEnable;
+	uint8_t     EnabledForThrottle;
+	uint8_t     EnabledForActivity;
+	uint8_t     padding_0;
+
+	uint8_t     UpHyst;
+	uint8_t     DownHyst;
+	uint8_t     VoltageDownHyst;
+	uint8_t     padding_1;
+
+	uint16_t    ActivityLevel;
+	uint8_t     DisplayWatermark;
+	uint8_t     padding_2;
+
+	uint16_t    Fcw_int;
+	uint16_t    Fcw_frac;
+	uint8_t     Postdiv;
+	uint8_t     padding_3[3];
+};
+
+typedef struct SMU75_Discrete_MemoryLevel SMU75_Discrete_MemoryLevel;
+
+struct SMU75_Discrete_LinkLevel {
+	uint8_t     PcieGenSpeed;
+	uint8_t     PcieLaneCount;
+	uint8_t     EnabledForActivity;
+	uint8_t     SPC;
+	uint32_t    DownThreshold;
+	uint32_t    UpThreshold;
+	uint16_t    BifSclkDfs;
+	uint16_t    Reserved;
+};
+
+typedef struct SMU75_Discrete_LinkLevel SMU75_Discrete_LinkLevel;
+
+
+/* MC ARB DRAM Timing registers. */
+struct SMU75_Discrete_MCArbDramTimingTableEntry {
+	uint32_t McArbDramTiming;
+	uint32_t McArbDramTiming2;
+	uint32_t McArbBurstTime;
+	uint32_t McArbRfshRate;
+	uint32_t McArbMisc3;
+};
+
+typedef struct SMU75_Discrete_MCArbDramTimingTableEntry SMU75_Discrete_MCArbDramTimingTableEntry;
+
+struct SMU75_Discrete_MCArbDramTimingTable {
+	SMU75_Discrete_MCArbDramTimingTableEntry entries[SMU__NUM_SCLK_DPM_STATE][SMU__NUM_MCLK_DPM_LEVELS];
+};
+
+typedef struct SMU75_Discrete_MCArbDramTimingTable SMU75_Discrete_MCArbDramTimingTable;
+
+/* UVD VCLK/DCLK state (level) definition. */
+struct SMU75_Discrete_UvdLevel {
+	uint32_t VclkFrequency;
+	uint32_t DclkFrequency;
+	SMU_VoltageLevel MinVoltage;
+	uint8_t  VclkDivider;
+	uint8_t  DclkDivider;
+	uint8_t  padding[2];
+};
+
+typedef struct SMU75_Discrete_UvdLevel SMU75_Discrete_UvdLevel;
+
+/* Clocks for other external blocks (VCE, ACP, SAMU). */
+struct SMU75_Discrete_ExtClkLevel {
+	uint32_t Frequency;
+	SMU_VoltageLevel MinVoltage;
+	uint8_t  Divider;
+	uint8_t  padding[3];
+};
+
+typedef struct SMU75_Discrete_ExtClkLevel SMU75_Discrete_ExtClkLevel;
+
+struct SMU75_Discrete_StateInfo {
+	uint32_t SclkFrequency;
+	uint32_t MclkFrequency;
+	uint32_t VclkFrequency;
+	uint32_t DclkFrequency;
+	uint32_t SamclkFrequency;
+	uint32_t AclkFrequency;
+	uint32_t EclkFrequency;
+	uint16_t MvddVoltage;
+	uint16_t padding16;
+	uint8_t  DisplayWatermark;
+	uint8_t  McArbIndex;
+	uint8_t  McRegIndex;
+	uint8_t  SeqIndex;
+	uint8_t  SclkDid;
+	int8_t   SclkIndex;
+	int8_t   MclkIndex;
+	uint8_t  PCIeGen;
+};
+
+typedef struct SMU75_Discrete_StateInfo SMU75_Discrete_StateInfo;
+
+struct SMU75_Discrete_DpmTable {
+	SMU75_PIDController                  GraphicsPIDController;
+	SMU75_PIDController                  MemoryPIDController;
+	SMU75_PIDController                  LinkPIDController;
+
+	uint32_t                            SystemFlags;
+
+	uint32_t                            VRConfig;
+	uint32_t                            SmioMask1;
+	uint32_t                            SmioMask2;
+	SMIO_Table                          SmioTable1;
+	SMIO_Table                          SmioTable2;
+
+	uint32_t                            MvddLevelCount;
+
+	uint8_t                             BapmVddcVidHiSidd        [SMU75_MAX_LEVELS_VDDC];
+	uint8_t                             BapmVddcVidLoSidd        [SMU75_MAX_LEVELS_VDDC];
+	uint8_t                             BapmVddcVidHiSidd2       [SMU75_MAX_LEVELS_VDDC];
+
+	uint8_t                             GraphicsDpmLevelCount;
+	uint8_t                             MemoryDpmLevelCount;
+	uint8_t                             LinkLevelCount;
+	uint8_t                             MasterDeepSleepControl;
+
+	uint8_t                             UvdLevelCount;
+	uint8_t                             VceLevelCount;
+	uint8_t                             AcpLevelCount;
+	uint8_t                             SamuLevelCount;
+
+	uint8_t                             ThermOutGpio;
+	uint8_t                             ThermOutPolarity;
+	uint8_t                             ThermOutMode;
+	uint8_t                             BootPhases;
+
+	uint8_t                             VRHotLevel;
+	uint8_t                             LdoRefSel;
+
+	uint8_t                             Reserved1[2];
+
+	uint16_t                            FanStartTemperature;
+	uint16_t                            FanStopTemperature;
+
+	uint16_t                            MaxVoltage;
+	uint16_t                            Reserved2;
+	uint32_t                            Reserved;
+
+	SMU75_Discrete_GraphicsLevel        GraphicsLevel           [SMU75_MAX_LEVELS_GRAPHICS];
+	SMU75_Discrete_MemoryLevel          MemoryACPILevel;
+	SMU75_Discrete_MemoryLevel          MemoryLevel             [SMU75_MAX_LEVELS_MEMORY];
+	SMU75_Discrete_LinkLevel            LinkLevel               [SMU75_MAX_LEVELS_LINK];
+	SMU75_Discrete_ACPILevel            ACPILevel;
+	SMU75_Discrete_UvdLevel             UvdLevel                [SMU75_MAX_LEVELS_UVD];
+	SMU75_Discrete_ExtClkLevel          VceLevel                [SMU75_MAX_LEVELS_VCE];
+	SMU75_Discrete_ExtClkLevel          AcpLevel                [SMU75_MAX_LEVELS_ACP];
+	SMU75_Discrete_ExtClkLevel          SamuLevel               [SMU75_MAX_LEVELS_SAMU];
+	SMU75_Discrete_Ulv                  Ulv;
+
+	uint8_t                             DisplayWatermark        [SMU75_MAX_LEVELS_MEMORY][SMU75_MAX_LEVELS_GRAPHICS];
+
+	uint32_t                            SclkStepSize;
+	uint32_t                            Smio                    [SMU75_MAX_ENTRIES_SMIO];
+
+	uint8_t                             UvdBootLevel;
+	uint8_t                             VceBootLevel;
+	uint8_t                             AcpBootLevel;
+	uint8_t                             SamuBootLevel;
+
+	uint8_t                             GraphicsBootLevel;
+	uint8_t                             GraphicsVoltageChangeEnable;
+	uint8_t                             GraphicsThermThrottleEnable;
+	uint8_t                             GraphicsInterval;
+
+	uint8_t                             VoltageInterval;
+	uint8_t                             ThermalInterval;
+	uint16_t                            TemperatureLimitHigh;
+
+	uint16_t                            TemperatureLimitLow;
+	uint8_t                             MemoryBootLevel;
+	uint8_t                             MemoryVoltageChangeEnable;
+
+	uint16_t                            BootMVdd;
+	uint8_t                             MemoryInterval;
+	uint8_t                             MemoryThermThrottleEnable;
+
+	uint16_t                            VoltageResponseTime;
+	uint16_t                            PhaseResponseTime;
+
+	uint8_t                             PCIeBootLinkLevel;
+	uint8_t                             PCIeGenInterval;
+	uint8_t                             DTEInterval;
+	uint8_t                             DTEMode;
+
+	uint8_t                             SVI2Enable;
+	uint8_t                             VRHotGpio;
+	uint8_t                             AcDcGpio;
+	uint8_t                             ThermGpio;
+
+	uint16_t                            PPM_PkgPwrLimit;
+	uint16_t                            PPM_TemperatureLimit;
+
+	uint16_t                            DefaultTdp;
+	uint16_t                            TargetTdp;
+
+	uint16_t                            FpsHighThreshold;
+	uint16_t                            FpsLowThreshold;
+
+	uint16_t                            BAPMTI_R  [SMU75_DTE_ITERATIONS][SMU75_DTE_SOURCES][SMU75_DTE_SINKS];
+	uint16_t                            BAPMTI_RC [SMU75_DTE_ITERATIONS][SMU75_DTE_SOURCES][SMU75_DTE_SINKS];
+
+	uint16_t                            TemperatureLimitEdge;
+	uint16_t                            TemperatureLimitHotspot;
+
+	uint16_t                            BootVddc;
+	uint16_t                            BootVddci;
+
+	uint16_t                            FanGainEdge;
+	uint16_t                            FanGainHotspot;
+
+	uint32_t                            LowSclkInterruptThreshold;
+	uint32_t                            VddGfxReChkWait;
+
+	uint8_t                             ClockStretcherAmount;
+	uint8_t                             Sclk_CKS_masterEn0_7;
+	uint8_t                             Sclk_CKS_masterEn8_15;
+	uint8_t                             DPMFreezeAndForced;
+
+	uint8_t                             Sclk_voltageOffset[8];
+
+	SMU_ClockStretcherDataTable         ClockStretcherDataTable;
+	SMU_CKS_LOOKUPTable                 CKS_LOOKUPTable;
+
+	uint32_t                            CurrSclkPllRange;
+	sclkFcwRange_t                      SclkFcwRangeTable[NUM_SCLK_RANGE];
+
+	GB_VDROOP_TABLE_t                   BTCGB_VDROOP_TABLE[BTCGB_VDROOP_TABLE_MAX_ENTRIES];
+	SMU_QuadraticCoeffs                 AVFSGB_FUSE_TABLE[AVFSGB_VDROOP_TABLE_MAX_ENTRIES];
+};
+
+typedef struct SMU75_Discrete_DpmTable SMU75_Discrete_DpmTable;
+
+struct SMU75_Discrete_FanTable {
+	uint16_t FdoMode;
+	int16_t  TempMin;
+	int16_t  TempMed;
+	int16_t  TempMax;
+	int16_t  Slope1;
+	int16_t  Slope2;
+	int16_t  FdoMin;
+	int16_t  HystUp;
+	int16_t  HystDown;
+	int16_t  HystSlope;
+	int16_t  TempRespLim;
+	int16_t  TempCurr;
+	int16_t  SlopeCurr;
+	int16_t  PwmCurr;
+	uint32_t RefreshPeriod;
+	int16_t  FdoMax;
+	uint8_t  TempSrc;
+	int8_t   Padding;
+};
+
+typedef struct SMU75_Discrete_FanTable SMU75_Discrete_FanTable;
+
+#define SMU7_DISCRETE_GPIO_SCLK_DEBUG             4
+#define SMU7_DISCRETE_GPIO_SCLK_DEBUG_BIT         (0x1 << SMU7_DISCRETE_GPIO_SCLK_DEBUG)
+
+
+
+struct SMU7_MclkDpmScoreboard {
+	uint32_t PercentageBusy;
+
+	int32_t  PIDError;
+	int32_t  PIDIntegral;
+	int32_t  PIDOutput;
+
+	uint32_t SigmaDeltaAccum;
+	uint32_t SigmaDeltaOutput;
+	uint32_t SigmaDeltaLevel;
+
+	uint32_t UtilizationSetpoint;
+
+	uint8_t  TdpClampMode;
+	uint8_t  TdcClampMode;
+	uint8_t  ThermClampMode;
+	uint8_t  VoltageBusy;
+
+	int8_t   CurrLevel;
+	int8_t   TargLevel;
+	uint8_t  LevelChangeInProgress;
+	uint8_t  UpHyst;
+
+	uint8_t  DownHyst;
+	uint8_t  VoltageDownHyst;
+	uint8_t  DpmEnable;
+	uint8_t  DpmRunning;
+
+	uint8_t  DpmForce;
+	uint8_t  DpmForceLevel;
+	uint8_t  padding2;
+	uint8_t  McArbIndex;
+
+	uint32_t MinimumPerfMclk;
+
+	uint8_t  AcpiReq;
+	uint8_t  AcpiAck;
+	uint8_t  MclkSwitchInProgress;
+	uint8_t  MclkSwitchCritical;
+
+	uint8_t  IgnoreVBlank;
+	uint8_t  TargetMclkIndex;
+	uint8_t  TargetMvddIndex;
+	uint8_t  MclkSwitchResult;
+
+	uint16_t VbiFailureCount;
+	uint8_t  VbiWaitCounter;
+	uint8_t  EnabledLevelsChange;
+
+	uint16_t LevelResidencyCounters [SMU75_MAX_LEVELS_MEMORY];
+	uint16_t LevelSwitchCounters [SMU75_MAX_LEVELS_MEMORY];
+
+	void     (*TargetStateCalculator)(uint8_t);
+	void     (*SavedTargetStateCalculator)(uint8_t);
+
+	uint16_t AutoDpmInterval;
+	uint16_t AutoDpmRange;
+
+	uint16_t VbiTimeoutCount;
+	uint16_t MclkSwitchingTime;
+
+	uint8_t  fastSwitch;
+	uint8_t  Save_PIC_VDDGFX_EXIT;
+	uint8_t  Save_PIC_VDDGFX_ENTER;
+	uint8_t  VbiTimeout;
+
+	uint32_t HbmTempRegBackup;
+};
+
+typedef struct SMU7_MclkDpmScoreboard SMU7_MclkDpmScoreboard;
+
+struct SMU7_UlvScoreboard {
+	uint8_t     EnterUlv;
+	uint8_t     ExitUlv;
+	uint8_t     UlvActive;
+	uint8_t     WaitingForUlv;
+	uint8_t     UlvEnable;
+	uint8_t     UlvRunning;
+	uint8_t     UlvMasterEnable;
+	uint8_t     padding;
+	uint32_t    UlvAbortedCount;
+	uint32_t    UlvTimeStamp;
+};
+
+typedef struct SMU7_UlvScoreboard SMU7_UlvScoreboard;
+
+struct VddgfxSavedRegisters {
+	uint32_t GPU_DBG[3];
+	uint32_t MEC_BaseAddress_Hi;
+	uint32_t MEC_BaseAddress_Lo;
+	uint32_t THM_TMON0_CTRL2__RDIR_PRESENT;
+	uint32_t THM_TMON1_CTRL2__RDIR_PRESENT;
+	uint32_t CP_INT_CNTL;
+};
+
+typedef struct VddgfxSavedRegisters VddgfxSavedRegisters;
+
+struct SMU7_VddGfxScoreboard {
+	uint8_t     VddGfxEnable;
+	uint8_t     VddGfxActive;
+	uint8_t     VPUResetOccured;
+	uint8_t     padding;
+
+	uint32_t    VddGfxEnteredCount;
+	uint32_t    VddGfxAbortedCount;
+
+	uint32_t    VddGfxVid;
+
+	VddgfxSavedRegisters SavedRegisters;
+};
+
+typedef struct SMU7_VddGfxScoreboard SMU7_VddGfxScoreboard;
+
+struct SMU7_TdcLimitScoreboard {
+	uint8_t  Enable;
+	uint8_t  Running;
+	uint16_t Alpha;
+	uint32_t FilteredIddc;
+	uint32_t IddcLimit;
+	uint32_t IddcHyst;
+	SMU7_HystController_Data HystControllerData;
+};
+
+typedef struct SMU7_TdcLimitScoreboard SMU7_TdcLimitScoreboard;
+
+struct SMU7_PkgPwrLimitScoreboard {
+	uint8_t  Enable;
+	uint8_t  Running;
+	uint16_t Alpha;
+	uint32_t FilteredPkgPwr;
+	uint32_t Limit;
+	uint32_t Hyst;
+	uint32_t LimitFromDriver;
+	uint8_t PowerSharingEnabled;
+	uint8_t PowerSharingCounter;
+	uint8_t PowerSharingINTEnabled;
+	uint8_t GFXActivityCounterEnabled;
+	uint32_t EnergyCount;
+	uint32_t PSACTCount;
+	uint8_t RollOverRequired;
+	uint8_t RollOverCount;
+	uint8_t padding[2];
+	SMU7_HystController_Data HystControllerData;
+};
+
+typedef struct SMU7_PkgPwrLimitScoreboard SMU7_PkgPwrLimitScoreboard;
+
+struct SMU7_BapmScoreboard {
+	uint32_t source_powers[SMU75_DTE_SOURCES];
+	uint32_t source_powers_last[SMU75_DTE_SOURCES];
+	int32_t entity_temperatures[SMU75_NUM_GPU_TES];
+	int32_t initial_entity_temperatures[SMU75_NUM_GPU_TES];
+	int32_t Limit;
+	int32_t Hyst;
+	int32_t therm_influence_coeff_table[SMU75_DTE_ITERATIONS * SMU75_DTE_SOURCES * SMU75_DTE_SINKS * 2];
+	int32_t therm_node_table[SMU75_DTE_ITERATIONS * SMU75_DTE_SOURCES * SMU75_DTE_SINKS];
+	uint16_t ConfigTDPPowerScalar;
+	uint16_t FanSpeedPowerScalar;
+	uint16_t OverDrivePowerScalar;
+	uint16_t OverDriveLimitScalar;
+	uint16_t FinalPowerScalar;
+	uint8_t VariantID;
+	uint8_t spare997;
+
+	SMU7_HystController_Data HystControllerData;
+
+	int32_t temperature_gradient_slope;
+	int32_t temperature_gradient;
+	uint32_t measured_temperature;
+};
+
+
+typedef struct SMU7_BapmScoreboard SMU7_BapmScoreboard;
+
+struct SMU7_AcpiScoreboard {
+	uint32_t SavedInterruptMask[2];
+	uint8_t LastACPIRequest;
+	uint8_t CgBifResp;
+	uint8_t RequestType;
+	uint8_t Padding;
+	SMU75_Discrete_ACPILevel D0Level;
+};
+
+typedef struct SMU7_AcpiScoreboard SMU7_AcpiScoreboard;
+
+struct SMU75_Discrete_PmFuses {
+	uint8_t BapmVddCVidHiSidd[8];
+
+	uint8_t BapmVddCVidLoSidd[8];
+
+	uint8_t VddCVid[8];
+
+	uint8_t SviLoadLineEn;
+	uint8_t SviLoadLineVddC;
+	uint8_t SviLoadLineTrimVddC;
+	uint8_t SviLoadLineOffsetVddC;
+
+	uint16_t TDC_VDDC_PkgLimit;
+	uint8_t TDC_VDDC_ThrottleReleaseLimitPerc;
+	uint8_t TDC_MAWt;
+
+	uint8_t TdcWaterfallCtl;
+	uint8_t LPMLTemperatureMin;
+	uint8_t LPMLTemperatureMax;
+	uint8_t Reserved;
+
+	uint8_t LPMLTemperatureScaler[16];
+
+	int16_t FuzzyFan_ErrorSetDelta;
+	int16_t FuzzyFan_ErrorRateSetDelta;
+	int16_t FuzzyFan_PwmSetDelta;
+	uint16_t Reserved6;
+
+	uint8_t GnbLPML[16];
+
+	uint8_t GnbLPMLMaxVid;
+	uint8_t GnbLPMLMinVid;
+	uint8_t Reserved1[2];
+
+	uint16_t BapmVddCBaseLeakageHiSidd;
+	uint16_t BapmVddCBaseLeakageLoSidd;
+
+	uint16_t  VFT_Temp[3];
+	uint8_t   Version;
+	uint8_t   padding;
+
+	SMU_QuadraticCoeffs VFT_ATE[3];
+
+	SMU_QuadraticCoeffs AVFS_GB;
+	SMU_QuadraticCoeffs ATE_ACBTC_GB;
+
+	SMU_QuadraticCoeffs P2V;
+
+	uint32_t PsmCharzFreq;
+
+	uint16_t InversionVoltage;
+	uint16_t PsmCharzTemp;
+
+	uint32_t EnabledAvfsModules;
+
+	SMU_QuadraticCoeffs BtcGbv_CksOff;
+};
+
+typedef struct SMU75_Discrete_PmFuses SMU75_Discrete_PmFuses;
+
+struct SMU7_Discrete_Log_Header_Table {
+	uint32_t    version;
+	uint32_t    asic_id;
+	uint16_t    flags;
+	uint16_t    entry_size;
+	uint32_t    total_size;
+	uint32_t    num_of_entries;
+	uint8_t     type;
+	uint8_t     mode;
+	uint8_t     filler_0[2];
+	uint32_t    filler_1[2];
+};
+
+typedef struct SMU7_Discrete_Log_Header_Table SMU7_Discrete_Log_Header_Table;
+
+struct SMU7_Discrete_Log_Cntl {
+	uint8_t             Enabled;
+	uint8_t             Type;
+	uint8_t             padding[2];
+	uint32_t            BufferSize;
+	uint32_t            SamplesLogged;
+	uint32_t            SampleSize;
+	uint32_t            AddrL;
+	uint32_t            AddrH;
+};
+
+typedef struct SMU7_Discrete_Log_Cntl SMU7_Discrete_Log_Cntl;
+
+#if defined SMU__DGPU_ONLY
+#define CAC_ACC_NW_NUM_OF_SIGNALS 87
+#endif
+
+
+struct SMU7_Discrete_Cac_Collection_Table {
+	uint32_t temperature;
+	uint32_t cac_acc_nw[CAC_ACC_NW_NUM_OF_SIGNALS];
+};
+
+typedef struct SMU7_Discrete_Cac_Collection_Table SMU7_Discrete_Cac_Collection_Table;
+
+struct SMU7_Discrete_Cac_Verification_Table {
+	uint32_t VddcTotalPower;
+	uint32_t VddcLeakagePower;
+	uint32_t VddcConstantPower;
+	uint32_t VddcGfxDynamicPower;
+	uint32_t VddcUvdDynamicPower;
+	uint32_t VddcVceDynamicPower;
+	uint32_t VddcAcpDynamicPower;
+	uint32_t VddcPcieDynamicPower;
+	uint32_t VddcDceDynamicPower;
+	uint32_t VddcCurrent;
+	uint32_t VddcVoltage;
+	uint32_t VddciTotalPower;
+	uint32_t VddciLeakagePower;
+	uint32_t VddciConstantPower;
+	uint32_t VddciDynamicPower;
+	uint32_t Vddr1TotalPower;
+	uint32_t Vddr1LeakagePower;
+	uint32_t Vddr1ConstantPower;
+	uint32_t Vddr1DynamicPower;
+	uint32_t spare[4];
+	uint32_t temperature;
+};
+
+typedef struct SMU7_Discrete_Cac_Verification_Table SMU7_Discrete_Cac_Verification_Table;
+
+struct SMU7_Discrete_Pm_Status_Table {
+	int32_t T_meas_max[SMU75_THERMAL_INPUT_LOOP_COUNT];
+	int32_t T_meas_acc[SMU75_THERMAL_INPUT_LOOP_COUNT];
+
+	uint32_t I_calc_max;
+	uint32_t I_calc_acc;
+	uint32_t P_meas_acc;
+	uint32_t V_meas_load_acc;
+	uint32_t I_meas_acc;
+	uint32_t P_meas_acc_vddci;
+	uint32_t V_meas_load_acc_vddci;
+	uint32_t I_meas_acc_vddci;
+
+	uint16_t Sclk_dpm_residency[8];
+	uint16_t Uvd_dpm_residency[8];
+	uint16_t Vce_dpm_residency[8];
+	uint16_t Mclk_dpm_residency[4];
+
+	uint32_t P_roc_acc;
+	uint32_t PkgPwr_max;
+	uint32_t PkgPwr_acc;
+	uint32_t MclkSwitchingTime_max;
+	uint32_t MclkSwitchingTime_acc;
+	uint32_t FanPwm_acc;
+	uint32_t FanRpm_acc;
+	uint32_t Gfx_busy_acc;
+	uint32_t Mc_busy_acc;
+	uint32_t Fps_acc;
+
+	uint32_t AccCnt;
+};
+
+typedef struct SMU7_Discrete_Pm_Status_Table SMU7_Discrete_Pm_Status_Table;
+
+struct SMU7_Discrete_AutoWattMan_Status_Table {
+	int32_t T_meas_acc[SMU75_THERMAL_INPUT_LOOP_COUNT];
+	uint16_t Sclk_dpm_residency[8];
+	uint16_t Mclk_dpm_residency[4];
+	uint32_t TgpPwr_acc;
+	uint32_t Gfx_busy_acc;
+	uint32_t Mc_busy_acc;
+	uint32_t AccCnt;
+};
+
+typedef struct SMU7_Discrete_AutoWattMan_Status_Table SMU7_Discrete_AutoWattMan_Status_Table;
+
+#define SMU7_MAX_GFX_CU_COUNT 24
+#define SMU7_MIN_GFX_CU_COUNT  8
+#define SMU7_GFX_CU_PG_ENABLE_DC_MAX_CU_SHIFT 0
+#define SMU7_GFX_CU_PG_ENABLE_DC_MAX_CU_MASK  (0xFFFF << SMU7_GFX_CU_PG_ENABLE_DC_MAX_CU_SHIFT)
+#define SMU7_GFX_CU_PG_ENABLE_AC_MAX_CU_SHIFT 16
+#define SMU7_GFX_CU_PG_ENABLE_AC_MAX_CU_MASK  (0xFFFF << SMU7_GFX_CU_PG_ENABLE_AC_MAX_CU_SHIFT)
+
+struct SMU7_GfxCuPgScoreboard {
+	uint8_t Enabled;
+	uint8_t WaterfallUp;
+	uint8_t WaterfallDown;
+	uint8_t WaterfallLimit;
+	uint8_t CurrMaxCu;
+	uint8_t TargMaxCu;
+	uint8_t ClampMode;
+	uint8_t Active;
+	uint8_t MaxSupportedCu;
+	uint8_t MinSupportedCu;
+	uint8_t PendingGfxCuHostInterrupt;
+	uint8_t LastFilteredMaxCuInteger;
+	uint16_t FilteredMaxCu;
+	uint16_t FilteredMaxCuAlpha;
+	uint16_t FilterResetCount;
+	uint16_t FilterResetCountLimit;
+	uint8_t ForceCu;
+	uint8_t ForceCuCount;
+	uint8_t AcModeMaxCu;
+	uint8_t DcModeMaxCu;
+};
+
+typedef struct SMU7_GfxCuPgScoreboard SMU7_GfxCuPgScoreboard;
+
+#define SMU7_SCLK_CAC 0x561
+#define SMU7_MCLK_CAC 0xF9
+#define SMU7_VCLK_CAC 0x2DE
+#define SMU7_DCLK_CAC 0x2DE
+#define SMU7_ECLK_CAC 0x25E
+#define SMU7_ACLK_CAC 0x25E
+#define SMU7_SAMCLK_CAC 0x25E
+#define SMU7_DISPCLK_CAC 0x100
+#define SMU7_CAC_CONSTANT 0x2EE3430
+#define SMU7_CAC_CONSTANT_SHIFT 18
+
+#define SMU7_VDDCI_MCLK_CONST        1765
+#define SMU7_VDDCI_MCLK_CONST_SHIFT  16
+#define SMU7_VDDCI_VDDCI_CONST       50958
+#define SMU7_VDDCI_VDDCI_CONST_SHIFT 14
+#define SMU7_VDDCI_CONST             11781
+#define SMU7_VDDCI_STROBE_PWR        1331
+
+#define SMU7_VDDR1_CONST            693
+#define SMU7_VDDR1_CAC_WEIGHT       20
+#define SMU7_VDDR1_CAC_WEIGHT_SHIFT 19
+#define SMU7_VDDR1_STROBE_PWR       512
+
+#define SMU7_AREA_COEFF_UVD 0xA78
+#define SMU7_AREA_COEFF_VCE 0x190A
+#define SMU7_AREA_COEFF_ACP 0x22D1
+#define SMU7_AREA_COEFF_SAMU 0x534
+
+#define SMU7_THERM_OUT_MODE_DISABLE       0x0
+#define SMU7_THERM_OUT_MODE_THERM_ONLY    0x1
+#define SMU7_THERM_OUT_MODE_THERM_VRHOT   0x2
+
+#define SQ_Enable_MASK 0x1
+#define SQ_IR_MASK 0x2
+#define SQ_PCC_MASK 0x4
+#define SQ_EDC_MASK 0x8
+
+#define TCP_Enable_MASK 0x100
+#define TCP_IR_MASK 0x200
+#define TCP_PCC_MASK 0x400
+#define TCP_EDC_MASK 0x800
+
+#define TD_Enable_MASK 0x10000
+#define TD_IR_MASK 0x20000
+#define TD_PCC_MASK 0x40000
+#define TD_EDC_MASK 0x80000
+
+#define DB_Enable_MASK 0x1000000
+#define DB_IR_MASK 0x2000000
+#define DB_PCC_MASK 0x4000000
+#define DB_EDC_MASK 0x8000000
+
+#define SQ_Enable_SHIFT 0
+#define SQ_IR_SHIFT 1
+#define SQ_PCC_SHIFT 2
+#define SQ_EDC_SHIFT 3
+
+#define TCP_Enable_SHIFT 8
+#define TCP_IR_SHIFT 9
+#define TCP_PCC_SHIFT 10
+#define TCP_EDC_SHIFT 11
+
+#define TD_Enable_SHIFT 16
+#define TD_IR_SHIFT 17
+#define TD_PCC_SHIFT 18
+#define TD_EDC_SHIFT 19
+
+#define DB_Enable_SHIFT 24
+#define DB_IR_SHIFT 25
+#define DB_PCC_SHIFT 26
+#define DB_EDC_SHIFT 27
+
+#define PMFUSES_AVFSSIZE 104
+
+#define BTCGB0_Vdroop_Enable_MASK  0x1
+#define BTCGB1_Vdroop_Enable_MASK  0x2
+#define AVFSGB0_Vdroop_Enable_MASK 0x4
+#define AVFSGB1_Vdroop_Enable_MASK 0x8
+
+#define BTCGB0_Vdroop_Enable_SHIFT  0
+#define BTCGB1_Vdroop_Enable_SHIFT  1
+#define AVFSGB0_Vdroop_Enable_SHIFT 2
+#define AVFSGB1_Vdroop_Enable_SHIFT 3
+
+#pragma pack(pop)
+
+
+#endif
+
diff --git a/drivers/gpu/drm/amd/powerplay/inc/smumgr.h b/drivers/gpu/drm/amd/powerplay/inc/smumgr.h
index 6c22ed9249bf..82550a8a3a3f 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/smumgr.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/smumgr.h
@@ -29,7 +29,6 @@
 enum SMU_TABLE {
 	SMU_UVD_TABLE = 0,
 	SMU_VCE_TABLE,
-	SMU_SAMU_TABLE,
 	SMU_BIF_TABLE,
 };
 
@@ -47,7 +46,6 @@ enum SMU_MEMBER {
 	UcodeLoadStatus,
 	UvdBootLevel,
 	VceBootLevel,
-	SamuBootLevel,
 	LowSclkInterruptThreshold,
 	DRAM_LOG_ADDR_H,
 	DRAM_LOG_ADDR_L,
@@ -82,7 +80,7 @@ enum SMU10_TABLE_ID {
 	SMU10_CLOCKTABLE,
 };
 
-extern int smum_get_argument(struct pp_hwmgr *hwmgr);
+extern uint32_t smum_get_argument(struct pp_hwmgr *hwmgr);
 
 extern int smum_download_powerplay_table(struct pp_hwmgr *hwmgr, void **table);
 
diff --git a/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h
index c3ed737ab951..715b5a168831 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h
@@ -131,6 +131,7 @@ typedef uint16_t PPSMC_Result;
 #define PPSMC_MSG_RunAcgInOpenLoop               0x5E
 #define PPSMC_MSG_InitializeAcg                  0x5F
 #define PPSMC_MSG_GetCurrPkgPwr                  0x61
+#define PPSMC_MSG_GetAverageGfxclkActualFrequency 0x63
 #define PPSMC_MSG_SetPccThrottleLevel            0x67
 #define PPSMC_MSG_UpdatePkgPwrPidAlpha           0x68
 #define PPSMC_Message_Count                      0x69
diff --git a/drivers/gpu/drm/amd/powerplay/inc/vega12/smu9_driver_if.h b/drivers/gpu/drm/amd/powerplay/inc/vega12/smu9_driver_if.h
index 2f8a3b983cce..b6ffd08784e7 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/vega12/smu9_driver_if.h
+++ b/drivers/gpu/drm/amd/powerplay/inc/vega12/smu9_driver_if.h
@@ -412,10 +412,10 @@ typedef struct {
   QuadraticInt_t    ReservedEquation2;
   QuadraticInt_t    ReservedEquation3;
 
-	uint16_t     MinVoltageUlvGfx;
-	uint16_t     MinVoltageUlvSoc;
+  uint16_t     MinVoltageUlvGfx;
+  uint16_t     MinVoltageUlvSoc;
 
-	uint32_t     Reserved[14];
+  uint32_t     Reserved[14];
 
 
 
@@ -483,9 +483,9 @@ typedef struct {
   uint8_t      padding8_4;
 
 
-	uint8_t      PllGfxclkSpreadEnabled;
-	uint8_t      PllGfxclkSpreadPercent;
-	uint16_t     PllGfxclkSpreadFreq;
+  uint8_t      PllGfxclkSpreadEnabled;
+  uint8_t      PllGfxclkSpreadPercent;
+  uint16_t     PllGfxclkSpreadFreq;
 
   uint8_t      UclkSpreadEnabled;
   uint8_t      UclkSpreadPercent;
@@ -495,11 +495,14 @@ typedef struct {
   uint8_t      SocclkSpreadPercent;
   uint16_t     SocclkSpreadFreq;
 
-	uint8_t      AcgGfxclkSpreadEnabled;
-	uint8_t      AcgGfxclkSpreadPercent;
-	uint16_t     AcgGfxclkSpreadFreq;
+  uint8_t      AcgGfxclkSpreadEnabled;
+  uint8_t      AcgGfxclkSpreadPercent;
+  uint16_t     AcgGfxclkSpreadFreq;
 
-	uint32_t     BoardReserved[10];
+  uint8_t      Vr2_I2C_address;
+  uint8_t      padding_vr2[3];
+
+  uint32_t     BoardReserved[9];
 
 
   uint32_t     MmHubPadding[7];
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/Makefile b/drivers/gpu/drm/amd/powerplay/smumgr/Makefile
index 958755075421..8d557accaef2 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/Makefile
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/Makefile
@@ -26,7 +26,7 @@
 SMU_MGR = smumgr.o smu8_smumgr.o tonga_smumgr.o fiji_smumgr.o \
 	  polaris10_smumgr.o iceland_smumgr.o \
 	  smu7_smumgr.o vega10_smumgr.o smu10_smumgr.o ci_smumgr.o \
-	  vega12_smumgr.o
+	  vega12_smumgr.o vegam_smumgr.o smu9_smumgr.o
 
 AMD_PP_SMUMGR = $(addprefix $(AMD_PP_PATH)/smumgr/,$(SMU_MGR))
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
index 08d000140eca..fbe3ef4ee45c 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c
@@ -61,9 +61,6 @@
 
 #define SMC_RAM_END 0x40000
 
-#define VOLTAGE_SCALE               4
-#define VOLTAGE_VID_OFFSET_SCALE1    625
-#define VOLTAGE_VID_OFFSET_SCALE2    100
 #define CISLAND_MINIMUM_ENGINE_CLOCK 800
 #define CISLAND_MAX_DEEPSLEEP_DIVIDER_ID 5
 
@@ -211,9 +208,7 @@ static int ci_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 {
 	int ret;
 
-	if (!ci_is_smc_ram_running(hwmgr))
-		return -EINVAL;
-
+	cgs_write_register(hwmgr->device, mmSMC_RESP_0, 0);
 	cgs_write_register(hwmgr->device, mmSMC_MESSAGE_0, msg);
 
 	PHM_WAIT_FIELD_UNEQUAL(hwmgr, SMC_RESP_0, SMC_RESP, 0);
@@ -1182,7 +1177,6 @@ static int ci_populate_single_memory_level(
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 	int result = 0;
 	bool dll_state_on;
-	struct cgs_display_info info = {0};
 	uint32_t mclk_edc_wr_enable_threshold = 40000;
 	uint32_t mclk_edc_enable_threshold = 40000;
 	uint32_t mclk_strobe_mode_threshold = 40000;
@@ -1236,8 +1230,7 @@ static int ci_populate_single_memory_level(
 	/* default set to low watermark. Highest level will be set to high later.*/
 	memory_level->DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW;
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-	data->display_timing.num_existing_displays = info.display_count;
+	data->display_timing.num_existing_displays = hwmgr->display_config->num_display;
 
 	/* stutter mode not support on ci */
 
@@ -1621,37 +1614,6 @@ static int ci_populate_smc_acp_level(struct pp_hwmgr *hwmgr,
 	return result;
 }
 
-static int ci_populate_smc_samu_level(struct pp_hwmgr *hwmgr,
-					SMU7_Discrete_DpmTable *table)
-{
-	int result = -EINVAL;
-	uint8_t count;
-	struct pp_atomctrl_clock_dividers_vi dividers;
-	struct phm_samu_clock_voltage_dependency_table *samu_table =
-				hwmgr->dyn_state.samu_clock_voltage_dependency_table;
-
-	table->SamuBootLevel = 0;
-	table->SamuLevelCount = (uint8_t)(samu_table->count);
-
-	for (count = 0; count < table->SamuLevelCount; count++) {
-		table->SamuLevel[count].Frequency = samu_table->entries[count].samclk;
-		table->SamuLevel[count].MinVoltage = samu_table->entries[count].v * VOLTAGE_SCALE;
-		table->SamuLevel[count].MinPhases = 1;
-
-		/* retrieve divider value for VBIOS */
-		result = atomctrl_get_dfs_pll_dividers_vi(hwmgr,
-				table->SamuLevel[count].Frequency, &dividers);
-		PP_ASSERT_WITH_CODE((0 == result),
-				"can not find divide id for samu clock", return result);
-
-		table->SamuLevel[count].Divider = (uint8_t)dividers.pll_post_divider;
-
-		CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].Frequency);
-		CONVERT_FROM_HOST_TO_SMC_US(table->SamuLevel[count].MinVoltage);
-	}
-	return result;
-}
-
 static int ci_populate_memory_timing_parameters(
 		struct pp_hwmgr *hwmgr,
 		uint32_t engine_clock,
@@ -2033,10 +1995,6 @@ static int ci_init_smc_table(struct pp_hwmgr *hwmgr)
 	PP_ASSERT_WITH_CODE(0 == result,
 		"Failed to initialize ACP Level!", return result);
 
-	result = ci_populate_smc_samu_level(hwmgr, table);
-	PP_ASSERT_WITH_CODE(0 == result,
-		"Failed to initialize SAMU Level!", return result);
-
 	/* Since only the initial state is completely set up at this point (the other states are just copies of the boot state) we only */
 	/* need to populate the  ARB settings for the initial state. */
 	result = ci_program_memory_timing_parameters(hwmgr);
@@ -2784,7 +2742,6 @@ static int ci_smu_fini(struct pp_hwmgr *hwmgr)
 {
 	kfree(hwmgr->smu_backend);
 	hwmgr->smu_backend = NULL;
-	cgs_rel_firmware(hwmgr->device, CGS_UCODE_ID_SMU);
 	return 0;
 }
 
@@ -2889,6 +2846,89 @@ static int ci_update_dpm_settings(struct pp_hwmgr *hwmgr,
 	return 0;
 }
 
+static int ci_update_uvd_smc_table(struct pp_hwmgr *hwmgr)
+{
+	struct amdgpu_device *adev = hwmgr->adev;
+	struct smu7_hwmgr *data = hwmgr->backend;
+	struct ci_smumgr *smu_data = hwmgr->smu_backend;
+	struct phm_uvd_clock_voltage_dependency_table *uvd_table =
+			hwmgr->dyn_state.uvd_clock_voltage_dependency_table;
+	uint32_t profile_mode_mask = AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD |
+					AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK |
+					AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK |
+					AMD_DPM_FORCED_LEVEL_PROFILE_PEAK;
+	uint32_t max_vddc = adev->pm.ac_power ? hwmgr->dyn_state.max_clock_voltage_on_ac.vddc :
+						hwmgr->dyn_state.max_clock_voltage_on_dc.vddc;
+	int32_t i;
+
+	if (PP_CAP(PHM_PlatformCaps_UVDDPM) || uvd_table->count <= 0)
+		smu_data->smc_state_table.UvdBootLevel = 0;
+	else
+		smu_data->smc_state_table.UvdBootLevel = uvd_table->count - 1;
+
+	PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, DPM_TABLE_475,
+				UvdBootLevel, smu_data->smc_state_table.UvdBootLevel);
+
+	data->dpm_level_enable_mask.uvd_dpm_enable_mask = 0;
+
+	for (i = uvd_table->count - 1; i >= 0; i--) {
+		if (uvd_table->entries[i].v <= max_vddc)
+			data->dpm_level_enable_mask.uvd_dpm_enable_mask |= 1 << i;
+		if (hwmgr->dpm_level & profile_mode_mask || !PP_CAP(PHM_PlatformCaps_UVDDPM))
+			break;
+	}
+	ci_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_UVDDPM_SetEnabledMask,
+				data->dpm_level_enable_mask.uvd_dpm_enable_mask);
+
+	return 0;
+}
+
+static int ci_update_vce_smc_table(struct pp_hwmgr *hwmgr)
+{
+	struct amdgpu_device *adev = hwmgr->adev;
+	struct smu7_hwmgr *data = hwmgr->backend;
+	struct phm_vce_clock_voltage_dependency_table *vce_table =
+			hwmgr->dyn_state.vce_clock_voltage_dependency_table;
+	uint32_t profile_mode_mask = AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD |
+					AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK |
+					AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK |
+					AMD_DPM_FORCED_LEVEL_PROFILE_PEAK;
+	uint32_t max_vddc = adev->pm.ac_power ? hwmgr->dyn_state.max_clock_voltage_on_ac.vddc :
+						hwmgr->dyn_state.max_clock_voltage_on_dc.vddc;
+	int32_t i;
+
+	PHM_WRITE_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, DPM_TABLE_475,
+				VceBootLevel, 0); /* temp hard code to level 0, vce can set min evclk*/
+
+	data->dpm_level_enable_mask.vce_dpm_enable_mask = 0;
+
+	for (i = vce_table->count - 1; i >= 0; i--) {
+		if (vce_table->entries[i].v <= max_vddc)
+			data->dpm_level_enable_mask.vce_dpm_enable_mask |= 1 << i;
+		if (hwmgr->dpm_level & profile_mode_mask || !PP_CAP(PHM_PlatformCaps_VCEDPM))
+			break;
+	}
+	ci_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_VCEDPM_SetEnabledMask,
+				data->dpm_level_enable_mask.vce_dpm_enable_mask);
+
+	return 0;
+}
+
+static int ci_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type)
+{
+	switch (type) {
+	case SMU_UVD_TABLE:
+		ci_update_uvd_smc_table(hwmgr);
+		break;
+	case SMU_VCE_TABLE:
+		ci_update_vce_smc_table(hwmgr);
+		break;
+	default:
+		break;
+	}
+	return 0;
+}
+
 const struct pp_smumgr_func ci_smu_funcs = {
 	.smu_init = ci_smu_init,
 	.smu_fini = ci_smu_fini,
@@ -2911,4 +2951,5 @@ const struct pp_smumgr_func ci_smu_funcs = {
 	.initialize_mc_reg_table = ci_initialize_mc_reg_table,
 	.is_dpm_running = ci_is_dpm_running,
 	.update_dpm_settings = ci_update_dpm_settings,
+	.update_smc_table = ci_update_smc_table,
 };
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
index faef78321446..18048f8e2f13 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/fiji_smumgr.c
@@ -53,10 +53,7 @@
 
 #define FIJI_SMC_SIZE 0x20000
 
-#define VOLTAGE_SCALE 4
 #define POWERTUNE_DEFAULT_SET_MAX    1
-#define VOLTAGE_VID_OFFSET_SCALE1   625
-#define VOLTAGE_VID_OFFSET_SCALE2   100
 #define VDDC_VDDCI_DELTA            300
 #define MC_CG_ARB_FREQ_F1           0x0b
 
@@ -288,8 +285,7 @@ static int fiji_start_smu(struct pp_hwmgr *hwmgr)
 	struct fiji_smumgr *priv = (struct fiji_smumgr *)(hwmgr->smu_backend);
 
 	/* Only start SMC if SMC RAM is not running */
-	if (!(smu7_is_smc_ram_running(hwmgr)
-		|| cgs_is_virtualization_enabled(hwmgr->device))) {
+	if (!smu7_is_smc_ram_running(hwmgr) && hwmgr->not_vf) {
 		/* Check if SMU is running in protected mode */
 		if (0 == PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device,
 				CGS_IND_REG__SMC,
@@ -307,13 +303,13 @@ static int fiji_start_smu(struct pp_hwmgr *hwmgr)
 	}
 
 	/* To initialize all clock gating before RLC loaded and running.*/
-	cgs_set_clockgating_state(hwmgr->device,
+	amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 			AMD_IP_BLOCK_TYPE_GFX, AMD_CG_STATE_GATE);
-	cgs_set_clockgating_state(hwmgr->device,
+	amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 			AMD_IP_BLOCK_TYPE_GMC, AMD_CG_STATE_GATE);
-	cgs_set_clockgating_state(hwmgr->device,
+	amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 			AMD_IP_BLOCK_TYPE_SDMA, AMD_CG_STATE_GATE);
-	cgs_set_clockgating_state(hwmgr->device,
+	amdgpu_device_ip_set_clockgating_state(hwmgr->adev,
 			AMD_IP_BLOCK_TYPE_COMMON, AMD_CG_STATE_GATE);
 
 	/* Setup SoftRegsStart here for register lookup in case
@@ -335,10 +331,10 @@ static bool fiji_is_hw_avfs_present(struct pp_hwmgr *hwmgr)
 	uint32_t efuse = 0;
 	uint32_t mask = (1 << ((AVFS_EN_MSB - AVFS_EN_LSB) + 1)) - 1;
 
-	if (cgs_is_virtualization_enabled(hwmgr->device))
-		return 0;
+	if (!hwmgr->not_vf)
+		return false;
 
-	if (!atomctrl_read_efuse(hwmgr->device, AVFS_EN_LSB, AVFS_EN_MSB,
+	if (!atomctrl_read_efuse(hwmgr, AVFS_EN_LSB, AVFS_EN_MSB,
 			mask, &efuse)) {
 		if (efuse)
 			return true;
@@ -989,11 +985,11 @@ static int fiji_populate_single_graphic_level(struct pp_hwmgr *hwmgr,
 
 	threshold = clock * data->fast_watermark_threshold / 100;
 
-	data->display_timing.min_clock_in_sr = hwmgr->display_config.min_core_set_clock_in_sr;
+	data->display_timing.min_clock_in_sr = hwmgr->display_config->min_core_set_clock_in_sr;
 
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SclkDeepSleep))
 		level->DeepSleepDivId = smu7_get_sleep_divider_id_from_clock(clock,
-								hwmgr->display_config.min_core_set_clock_in_sr);
+								hwmgr->display_config->min_core_set_clock_in_sr);
 
 
 	/* Default to slow, highest DPM level will be
@@ -1507,44 +1503,6 @@ static int fiji_populate_smc_acp_level(struct pp_hwmgr *hwmgr,
 	return result;
 }
 
-static int fiji_populate_smc_samu_level(struct pp_hwmgr *hwmgr,
-		SMU73_Discrete_DpmTable *table)
-{
-	int result = -EINVAL;
-	uint8_t count;
-	struct pp_atomctrl_clock_dividers_vi dividers;
-	struct phm_ppt_v1_information *table_info =
-			(struct phm_ppt_v1_information *)(hwmgr->pptable);
-	struct phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table =
-			table_info->mm_dep_table;
-
-	table->SamuBootLevel = 0;
-	table->SamuLevelCount = (uint8_t)(mm_table->count);
-
-	for (count = 0; count < table->SamuLevelCount; count++) {
-		/* not sure whether we need evclk or not */
-		table->SamuLevel[count].MinVoltage = 0;
-		table->SamuLevel[count].Frequency = mm_table->entries[count].samclock;
-		table->SamuLevel[count].MinVoltage |= (mm_table->entries[count].vddc *
-				VOLTAGE_SCALE) << VDDC_SHIFT;
-		table->SamuLevel[count].MinVoltage |= ((mm_table->entries[count].vddc -
-				VDDC_VDDCI_DELTA) * VOLTAGE_SCALE) << VDDCI_SHIFT;
-		table->SamuLevel[count].MinVoltage |= 1 << PHASES_SHIFT;
-
-		/* retrieve divider value for VBIOS */
-		result = atomctrl_get_dfs_pll_dividers_vi(hwmgr,
-				table->SamuLevel[count].Frequency, &dividers);
-		PP_ASSERT_WITH_CODE((0 == result),
-				"can not find divide id for samu clock", return result);
-
-		table->SamuLevel[count].Divider = (uint8_t)dividers.pll_post_divider;
-
-		CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].Frequency);
-		CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].MinVoltage);
-	}
-	return result;
-}
-
 static int fiji_populate_memory_timing_parameters(struct pp_hwmgr *hwmgr,
 		int32_t eng_clock, int32_t mem_clock,
 		struct SMU73_Discrete_MCArbDramTimingTableEntry *arb_regs)
@@ -2032,10 +1990,6 @@ static int fiji_init_smc_table(struct pp_hwmgr *hwmgr)
 	PP_ASSERT_WITH_CODE(0 == result,
 			"Failed to initialize ACP Level!", return result);
 
-	result = fiji_populate_smc_samu_level(hwmgr, table);
-	PP_ASSERT_WITH_CODE(0 == result,
-			"Failed to initialize SAMU Level!", return result);
-
 	/* Since only the initial state is completely set up at this point
 	 * (the other states are just copies of the boot state) we only
 	 * need to populate the  ARB settings for the initial state.
@@ -2382,8 +2336,6 @@ static uint32_t fiji_get_offsetof(uint32_t type, uint32_t member)
 			return offsetof(SMU73_Discrete_DpmTable, UvdBootLevel);
 		case VceBootLevel:
 			return offsetof(SMU73_Discrete_DpmTable, VceBootLevel);
-		case SamuBootLevel:
-			return offsetof(SMU73_Discrete_DpmTable, SamuBootLevel);
 		case LowSclkInterruptThreshold:
 			return offsetof(SMU73_Discrete_DpmTable, LowSclkInterruptThreshold);
 		}
@@ -2482,33 +2434,6 @@ static int fiji_update_vce_smc_table(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
-static int fiji_update_samu_smc_table(struct pp_hwmgr *hwmgr)
-{
-	struct fiji_smumgr *smu_data = (struct fiji_smumgr *)(hwmgr->smu_backend);
-	uint32_t mm_boot_level_offset, mm_boot_level_value;
-
-
-	smu_data->smc_state_table.SamuBootLevel = 0;
-	mm_boot_level_offset = smu_data->smu7_data.dpm_table_start +
-				offsetof(SMU73_Discrete_DpmTable, SamuBootLevel);
-
-	mm_boot_level_offset /= 4;
-	mm_boot_level_offset *= 4;
-	mm_boot_level_value = cgs_read_ind_register(hwmgr->device,
-			CGS_IND_REG__SMC, mm_boot_level_offset);
-	mm_boot_level_value &= 0xFFFFFF00;
-	mm_boot_level_value |= smu_data->smc_state_table.SamuBootLevel << 0;
-	cgs_write_ind_register(hwmgr->device,
-			CGS_IND_REG__SMC, mm_boot_level_offset, mm_boot_level_value);
-
-	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
-			PHM_PlatformCaps_StablePState))
-		smum_send_msg_to_smc_with_parameter(hwmgr,
-				PPSMC_MSG_SAMUDPM_SetEnabledMask,
-				(uint32_t)(1 << smu_data->smc_state_table.SamuBootLevel));
-	return 0;
-}
-
 static int fiji_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type)
 {
 	switch (type) {
@@ -2518,9 +2443,6 @@ static int fiji_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type)
 	case SMU_VCE_TABLE:
 		fiji_update_vce_smc_table(hwmgr);
 		break;
-	case SMU_SAMU_TABLE:
-		fiji_update_samu_smc_table(hwmgr);
-		break;
 	default:
 		break;
 	}
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c
index d4bb934e7334..9299b93aa09a 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c
@@ -60,10 +60,7 @@
 
 #define ICELAND_SMC_SIZE               0x20000
 
-#define VOLTAGE_SCALE 4
 #define POWERTUNE_DEFAULT_SET_MAX    1
-#define VOLTAGE_VID_OFFSET_SCALE1   625
-#define VOLTAGE_VID_OFFSET_SCALE2   100
 #define MC_CG_ARB_FREQ_F1           0x0b
 #define VDDC_VDDCI_DELTA            200
 
@@ -932,7 +929,7 @@ static int iceland_populate_single_graphic_level(struct pp_hwmgr *hwmgr,
 	graphic_level->PowerThrottle = 0;
 
 	data->display_timing.min_clock_in_sr =
-			hwmgr->display_config.min_core_set_clock_in_sr;
+			hwmgr->display_config->min_core_set_clock_in_sr;
 
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
 			PHM_PlatformCaps_SclkDeepSleep))
@@ -1236,7 +1233,6 @@ static int iceland_populate_single_memory_level(
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 	int result = 0;
 	bool dll_state_on;
-	struct cgs_display_info info = {0};
 	uint32_t mclk_edc_wr_enable_threshold = 40000;
 	uint32_t mclk_edc_enable_threshold = 40000;
 	uint32_t mclk_strobe_mode_threshold = 40000;
@@ -1283,8 +1279,7 @@ static int iceland_populate_single_memory_level(
 	/* default set to low watermark. Highest level will be set to high later.*/
 	memory_level->DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW;
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-	data->display_timing.num_existing_displays = info.display_count;
+	data->display_timing.num_existing_displays = hwmgr->display_config->num_display;
 
 	/* stutter mode not support on iceland */
 
@@ -1583,12 +1578,6 @@ static int iceland_populate_smc_acp_level(struct pp_hwmgr *hwmgr,
 	return 0;
 }
 
-static int iceland_populate_smc_samu_level(struct pp_hwmgr *hwmgr,
-	SMU71_Discrete_DpmTable *table)
-{
-	return 0;
-}
-
 static int iceland_populate_memory_timing_parameters(
 		struct pp_hwmgr *hwmgr,
 		uint32_t engine_clock,
@@ -1997,10 +1986,6 @@ static int iceland_init_smc_table(struct pp_hwmgr *hwmgr)
 	PP_ASSERT_WITH_CODE(0 == result,
 		"Failed to initialize ACP Level!", return result;);
 
-	result = iceland_populate_smc_samu_level(hwmgr, table);
-	PP_ASSERT_WITH_CODE(0 == result,
-		"Failed to initialize SAMU Level!", return result;);
-
 	/* Since only the initial state is completely set up at this point (the other states are just copies of the boot state) we only */
 	/* need to populate the  ARB settings for the initial state. */
 	result = iceland_program_memory_timing_parameters(hwmgr);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
index 997a777dd35b..a4ce199af475 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c
@@ -52,8 +52,6 @@
 #include "dce/dce_10_0_sh_mask.h"
 
 #define POLARIS10_SMC_SIZE 0x20000
-#define VOLTAGE_VID_OFFSET_SCALE1   625
-#define VOLTAGE_VID_OFFSET_SCALE2   100
 #define POWERTUNE_DEFAULT_SET_MAX    1
 #define VDDC_VDDCI_DELTA            200
 #define MC_CG_ARB_FREQ_F1           0x0b
@@ -295,25 +293,16 @@ static int polaris10_start_smu(struct pp_hwmgr *hwmgr)
 	struct polaris10_smumgr *smu_data = (struct polaris10_smumgr *)(hwmgr->smu_backend);
 
 	/* Only start SMC if SMC RAM is not running */
-	if (!(smu7_is_smc_ram_running(hwmgr)
-		|| cgs_is_virtualization_enabled(hwmgr->device))) {
+	if (!smu7_is_smc_ram_running(hwmgr) && hwmgr->not_vf) {
 		smu_data->protected_mode = (uint8_t) (PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, SMU_FIRMWARE, SMU_MODE));
 		smu_data->smu7_data.security_hard_key = (uint8_t) (PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC, SMU_FIRMWARE, SMU_SEL));
 
 		/* Check if SMU is running in protected mode */
-		if (smu_data->protected_mode == 0) {
+		if (smu_data->protected_mode == 0)
 			result = polaris10_start_smu_in_non_protection_mode(hwmgr);
-		} else {
+		else
 			result = polaris10_start_smu_in_protection_mode(hwmgr);
 
-			/* If failed, try with different security Key. */
-			if (result != 0) {
-				smu_data->smu7_data.security_hard_key ^= 1;
-				cgs_rel_firmware(hwmgr->device, CGS_UCODE_ID_SMU);
-				result = polaris10_start_smu_in_protection_mode(hwmgr);
-			}
-		}
-
 		if (result != 0)
 			PP_ASSERT_WITH_CODE(0, "Failed to load SMU ucode.", return result);
 
@@ -951,11 +940,11 @@ static int polaris10_populate_single_graphic_level(struct pp_hwmgr *hwmgr,
 	level->DownHyst = data->current_profile_setting.sclk_down_hyst;
 	level->VoltageDownHyst = 0;
 	level->PowerThrottle = 0;
-	data->display_timing.min_clock_in_sr = hwmgr->display_config.min_core_set_clock_in_sr;
+	data->display_timing.min_clock_in_sr = hwmgr->display_config->min_core_set_clock_in_sr;
 
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SclkDeepSleep))
 		level->DeepSleepDivId = smu7_get_sleep_divider_id_from_clock(clock,
-								hwmgr->display_config.min_core_set_clock_in_sr);
+								hwmgr->display_config->min_core_set_clock_in_sr);
 
 	/* Default to slow, highest DPM level will be
 	 * set to PPSMC_DISPLAY_WATERMARK_LOW later.
@@ -1085,11 +1074,9 @@ static int polaris10_populate_single_memory_level(struct pp_hwmgr *hwmgr,
 	struct phm_ppt_v1_information *table_info =
 			(struct phm_ppt_v1_information *)(hwmgr->pptable);
 	int result = 0;
-	struct cgs_display_info info = {0, 0, NULL};
 	uint32_t mclk_stutter_mode_threshold = 40000;
 	phm_ppt_v1_clock_voltage_dependency_table *vdd_dep_table = NULL;
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
 
 	if (hwmgr->od_enabled)
 		vdd_dep_table = (phm_ppt_v1_clock_voltage_dependency_table *)&data->odn_dpm_table.vdd_dependency_on_mclk;
@@ -1115,7 +1102,7 @@ static int polaris10_populate_single_memory_level(struct pp_hwmgr *hwmgr,
 	mem_level->StutterEnable = false;
 	mem_level->DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW;
 
-	data->display_timing.num_existing_displays = info.display_count;
+	data->display_timing.num_existing_displays = hwmgr->display_config->num_display;
 
 	if (mclk_stutter_mode_threshold &&
 		(clock <= mclk_stutter_mode_threshold) &&
@@ -1350,55 +1337,6 @@ static int polaris10_populate_smc_vce_level(struct pp_hwmgr *hwmgr,
 	return result;
 }
 
-
-static int polaris10_populate_smc_samu_level(struct pp_hwmgr *hwmgr,
-		SMU74_Discrete_DpmTable *table)
-{
-	int result = -EINVAL;
-	uint8_t count;
-	struct pp_atomctrl_clock_dividers_vi dividers;
-	struct phm_ppt_v1_information *table_info =
-			(struct phm_ppt_v1_information *)(hwmgr->pptable);
-	struct phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table =
-			table_info->mm_dep_table;
-	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
-	uint32_t vddci;
-
-	table->SamuBootLevel = 0;
-	table->SamuLevelCount = (uint8_t)(mm_table->count);
-
-	for (count = 0; count < table->SamuLevelCount; count++) {
-		/* not sure whether we need evclk or not */
-		table->SamuLevel[count].MinVoltage = 0;
-		table->SamuLevel[count].Frequency = mm_table->entries[count].samclock;
-		table->SamuLevel[count].MinVoltage |= (mm_table->entries[count].vddc *
-				VOLTAGE_SCALE) << VDDC_SHIFT;
-
-		if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control)
-			vddci = (uint32_t)phm_find_closest_vddci(&(data->vddci_voltage_table),
-						mm_table->entries[count].vddc - VDDC_VDDCI_DELTA);
-		else if (SMU7_VOLTAGE_CONTROL_BY_SVID2 == data->vddci_control)
-			vddci = mm_table->entries[count].vddc - VDDC_VDDCI_DELTA;
-		else
-			vddci = (data->vbios_boot_state.vddci_bootup_value * VOLTAGE_SCALE) << VDDCI_SHIFT;
-
-		table->SamuLevel[count].MinVoltage |= (vddci * VOLTAGE_SCALE) << VDDCI_SHIFT;
-		table->SamuLevel[count].MinVoltage |= 1 << PHASES_SHIFT;
-
-		/* retrieve divider value for VBIOS */
-		result = atomctrl_get_dfs_pll_dividers_vi(hwmgr,
-				table->SamuLevel[count].Frequency, &dividers);
-		PP_ASSERT_WITH_CODE((0 == result),
-				"can not find divide id for samu clock", return result);
-
-		table->SamuLevel[count].Divider = (uint8_t)dividers.pll_post_divider;
-
-		CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].Frequency);
-		CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].MinVoltage);
-	}
-	return result;
-}
-
 static int polaris10_populate_memory_timing_parameters(struct pp_hwmgr *hwmgr,
 		int32_t eng_clock, int32_t mem_clock,
 		SMU74_Discrete_MCArbDramTimingTableEntry *arb_regs)
@@ -1878,10 +1816,6 @@ static int polaris10_init_smc_table(struct pp_hwmgr *hwmgr)
 	PP_ASSERT_WITH_CODE(0 == result,
 			"Failed to initialize VCE Level!", return result);
 
-	result = polaris10_populate_smc_samu_level(hwmgr, table);
-	PP_ASSERT_WITH_CODE(0 == result,
-			"Failed to initialize SAMU Level!", return result);
-
 	/* Since only the initial state is completely set up at this point
 	 * (the other states are just copies of the boot state) we only
 	 * need to populate the  ARB settings for the initial state.
@@ -2235,34 +2169,6 @@ static int polaris10_update_vce_smc_table(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
-static int polaris10_update_samu_smc_table(struct pp_hwmgr *hwmgr)
-{
-	struct polaris10_smumgr *smu_data = (struct polaris10_smumgr *)(hwmgr->smu_backend);
-	uint32_t mm_boot_level_offset, mm_boot_level_value;
-
-
-	smu_data->smc_state_table.SamuBootLevel = 0;
-	mm_boot_level_offset = smu_data->smu7_data.dpm_table_start +
-				offsetof(SMU74_Discrete_DpmTable, SamuBootLevel);
-
-	mm_boot_level_offset /= 4;
-	mm_boot_level_offset *= 4;
-	mm_boot_level_value = cgs_read_ind_register(hwmgr->device,
-			CGS_IND_REG__SMC, mm_boot_level_offset);
-	mm_boot_level_value &= 0xFFFFFF00;
-	mm_boot_level_value |= smu_data->smc_state_table.SamuBootLevel << 0;
-	cgs_write_ind_register(hwmgr->device,
-			CGS_IND_REG__SMC, mm_boot_level_offset, mm_boot_level_value);
-
-	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
-			PHM_PlatformCaps_StablePState))
-		smum_send_msg_to_smc_with_parameter(hwmgr,
-				PPSMC_MSG_SAMUDPM_SetEnabledMask,
-				(uint32_t)(1 << smu_data->smc_state_table.SamuBootLevel));
-	return 0;
-}
-
-
 static int polaris10_update_bif_smc_table(struct pp_hwmgr *hwmgr)
 {
 	struct polaris10_smumgr *smu_data = (struct polaris10_smumgr *)(hwmgr->smu_backend);
@@ -2289,9 +2195,6 @@ static int polaris10_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type)
 	case SMU_VCE_TABLE:
 		polaris10_update_vce_smc_table(hwmgr);
 		break;
-	case SMU_SAMU_TABLE:
-		polaris10_update_samu_smc_table(hwmgr);
-		break;
 	case SMU_BIF_TABLE:
 		polaris10_update_bif_smc_table(hwmgr);
 	default:
@@ -2370,8 +2273,6 @@ static uint32_t polaris10_get_offsetof(uint32_t type, uint32_t member)
 			return offsetof(SMU74_Discrete_DpmTable, UvdBootLevel);
 		case VceBootLevel:
 			return offsetof(SMU74_Discrete_DpmTable, VceBootLevel);
-		case SamuBootLevel:
-			return offsetof(SMU74_Discrete_DpmTable, SamuBootLevel);
 		case LowSclkInterruptThreshold:
 			return offsetof(SMU74_Discrete_DpmTable, LowSclkInterruptThreshold);
 		}
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c
index bc53f2beda30..bb07d43f3874 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu10_smumgr.c
@@ -23,7 +23,7 @@
 
 #include "smumgr.h"
 #include "smu10_inc.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 #include "smu10_smumgr.h"
 #include "ppatomctrl.h"
 #include "rv_ppsmc.h"
@@ -33,8 +33,6 @@
 #include "pp_debug.h"
 
 
-#define VOLTAGE_SCALE 4
-
 #define BUFFER_SIZE                 80000
 #define MAX_STRING_SIZE             15
 #define BUFFER_SIZETWO              131072
@@ -49,48 +47,41 @@
 
 static uint32_t smu10_wait_for_response(struct pp_hwmgr *hwmgr)
 {
+	struct amdgpu_device *adev = hwmgr->adev;
 	uint32_t reg;
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
+	reg = SOC15_REG_OFFSET(MP1, 0, mmMP1_SMN_C2PMSG_90);
 
 	phm_wait_for_register_unequal(hwmgr, reg,
 			0, MP1_C2PMSG_90__CONTENT_MASK);
 
-	return cgs_read_register(hwmgr->device, reg);
+	return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90);
 }
 
 static int smu10_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr,
 		uint16_t msg)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_66_BASE_IDX, mmMP1_SMN_C2PMSG_66);
-	cgs_write_register(hwmgr->device, reg, msg);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_66, msg);
 
 	return 0;
 }
 
-static int smu10_read_arg_from_smc(struct pp_hwmgr *hwmgr)
+static uint32_t smu10_read_arg_from_smc(struct pp_hwmgr *hwmgr)
 {
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_82_BASE_IDX, mmMP1_SMN_C2PMSG_82);
+	struct amdgpu_device *adev = hwmgr->adev;
 
-	return cgs_read_register(hwmgr->device, reg);
+	return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82);
 }
 
 static int smu10_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	smu10_wait_for_response(hwmgr);
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
-	cgs_write_register(hwmgr->device, reg, 0);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0);
 
 	smu10_send_msg_to_smc_without_waiting(hwmgr, msg);
 
@@ -104,17 +95,13 @@ static int smu10_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 static int smu10_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr,
 		uint16_t msg, uint32_t parameter)
 {
-	uint32_t reg;
+	struct amdgpu_device *adev = hwmgr->adev;
 
 	smu10_wait_for_response(hwmgr);
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
-	cgs_write_register(hwmgr->device, reg, 0);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0);
 
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_82_BASE_IDX, mmMP1_SMN_C2PMSG_82);
-	cgs_write_register(hwmgr->device, reg, parameter);
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82, parameter);
 
 	smu10_send_msg_to_smc_without_waiting(hwmgr, msg);
 
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
index 0399c10d2be0..a029e47c2319 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.c
@@ -167,24 +167,25 @@ int smu7_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
 {
 	int ret;
 
-	if (!smu7_is_smc_ram_running(hwmgr))
-		return -EINVAL;
-
-
 	PHM_WAIT_FIELD_UNEQUAL(hwmgr, SMC_RESP_0, SMC_RESP, 0);
 
 	ret = PHM_READ_FIELD(hwmgr->device, SMC_RESP_0, SMC_RESP);
 
-	if (ret != 1)
-		pr_info("\n failed to send pre message %x ret is %d \n",  msg, ret);
+	if (ret == 0xFE)
+		pr_debug("last message was not supported\n");
+	else if (ret != 1)
+		pr_info("\n last message was failed ret is %d\n", ret);
 
+	cgs_write_register(hwmgr->device, mmSMC_RESP_0, 0);
 	cgs_write_register(hwmgr->device, mmSMC_MESSAGE_0, msg);
 
 	PHM_WAIT_FIELD_UNEQUAL(hwmgr, SMC_RESP_0, SMC_RESP, 0);
 
 	ret = PHM_READ_FIELD(hwmgr->device, SMC_RESP_0, SMC_RESP);
 
-	if (ret != 1)
+	if (ret == 0xFE)
+		pr_debug("message %x was not supported\n", msg);
+	else if (ret != 1)
 		pr_info("\n failed to send message %x ret is %d \n",  msg, ret);
 
 	return 0;
@@ -199,10 +200,6 @@ int smu7_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr, uint16_t msg)
 
 int smu7_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr, uint16_t msg, uint32_t parameter)
 {
-	if (!smu7_is_smc_ram_running(hwmgr)) {
-		return -EINVAL;
-	}
-
 	PHM_WAIT_FIELD_UNEQUAL(hwmgr, SMC_RESP_0, SMC_RESP, 0);
 
 	cgs_write_register(hwmgr->device, mmSMC_MSG_ARG_0, parameter);
@@ -231,16 +228,6 @@ int smu7_send_msg_to_smc_offset(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
-int smu7_wait_for_smc_inactive(struct pp_hwmgr *hwmgr)
-{
-	if (!smu7_is_smc_ram_running(hwmgr))
-		return -EINVAL;
-
-	PHM_WAIT_VFPF_INDIRECT_FIELD(hwmgr, SMC_IND, SMC_SYSCON_CLOCK_CNTL_0, cken, 0);
-	return 0;
-}
-
-
 enum cgs_ucode_id smu7_convert_fw_type_to_cgs(uint32_t fw_type)
 {
 	enum cgs_ucode_id result = CGS_UCODE_ID_MAXIMUM;
@@ -296,11 +283,9 @@ int smu7_read_smc_sram_dword(struct pp_hwmgr *hwmgr, uint32_t smc_addr, uint32_t
 
 	result = smu7_set_smc_sram_address(hwmgr, smc_addr, limit);
 
-	if (result)
-		return result;
+	*value = result ? 0 : cgs_read_register(hwmgr->device, mmSMC_IND_DATA_11);
 
-	*value = cgs_read_register(hwmgr->device, mmSMC_IND_DATA_11);
-	return 0;
+	return result;
 }
 
 int smu7_write_smc_sram_dword(struct pp_hwmgr *hwmgr, uint32_t smc_addr, uint32_t value, uint32_t limit)
@@ -375,7 +360,7 @@ static int smu7_populate_single_firmware_entry(struct pp_hwmgr *hwmgr,
 		entry->meta_data_addr_low = 0;
 
 		/* digest need be excluded out */
-		if (cgs_is_virtualization_enabled(hwmgr->device))
+		if (!hwmgr->not_vf)
 			info.image_size -= 20;
 		entry->data_size_byte = info.image_size;
 		entry->num_register_entries = 0;
@@ -394,8 +379,7 @@ int smu7_request_smu_load_fw(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_smumgr *smu_data = (struct smu7_smumgr *)(hwmgr->smu_backend);
 	uint32_t fw_to_load;
-	int result = 0;
-	struct SMU_DRAMData_TOC *toc;
+	int r = 0;
 
 	if (!hwmgr->reload_fw) {
 		pr_info("skip reloading...\n");
@@ -409,7 +393,7 @@ int smu7_request_smu_load_fw(struct pp_hwmgr *hwmgr)
 					0x0);
 
 	if (hwmgr->chip_id > CHIP_TOPAZ) { /* add support for Topaz */
-		if (!cgs_is_virtualization_enabled(hwmgr->device)) {
+		if (hwmgr->not_vf) {
 			smu7_send_msg_to_smc_with_parameter(hwmgr,
 						PPSMC_MSG_SMU_DRAM_ADDR_HI,
 						upper_32_bits(smu_data->smu_buffer.mc_addr));
@@ -436,49 +420,62 @@ int smu7_request_smu_load_fw(struct pp_hwmgr *hwmgr)
 			   + UCODE_ID_CP_MEC_JT2_MASK;
 	}
 
-	toc = (struct SMU_DRAMData_TOC *)smu_data->header;
-	toc->num_entries = 0;
-	toc->structure_version = 1;
+	if (!smu_data->toc) {
+		struct SMU_DRAMData_TOC *toc;
 
-	PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
+		smu_data->toc = kzalloc(sizeof(struct SMU_DRAMData_TOC), GFP_KERNEL);
+		if (!smu_data->toc)
+			return -ENOMEM;
+		toc = smu_data->toc;
+		toc->num_entries = 0;
+		toc->structure_version = 1;
+
+		PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
 				UCODE_ID_RLC_G, &toc->entry[toc->num_entries++]),
-				"Failed to Get Firmware Entry.", return -EINVAL);
-	PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
+				"Failed to Get Firmware Entry.", r = -EINVAL; goto failed);
+		PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
 				UCODE_ID_CP_CE, &toc->entry[toc->num_entries++]),
-				"Failed to Get Firmware Entry.", return -EINVAL);
-	PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
+				"Failed to Get Firmware Entry.", r = -EINVAL; goto failed);
+		PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
 				UCODE_ID_CP_PFP, &toc->entry[toc->num_entries++]),
-				"Failed to Get Firmware Entry.", return -EINVAL);
-	PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
+				"Failed to Get Firmware Entry.", r = -EINVAL; goto failed);
+		PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
 				UCODE_ID_CP_ME, &toc->entry[toc->num_entries++]),
-				"Failed to Get Firmware Entry.", return -EINVAL);
-	PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
+				"Failed to Get Firmware Entry.", r = -EINVAL; goto failed);
+		PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
 				UCODE_ID_CP_MEC, &toc->entry[toc->num_entries++]),
-				"Failed to Get Firmware Entry.", return -EINVAL);
-	PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
+				"Failed to Get Firmware Entry.", r = -EINVAL; goto failed);
+		PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
 				UCODE_ID_CP_MEC_JT1, &toc->entry[toc->num_entries++]),
-				"Failed to Get Firmware Entry.", return -EINVAL);
-	PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
+				"Failed to Get Firmware Entry.", r = -EINVAL; goto failed);
+		PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
 				UCODE_ID_CP_MEC_JT2, &toc->entry[toc->num_entries++]),
-				"Failed to Get Firmware Entry.", return -EINVAL);
-	PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
+				"Failed to Get Firmware Entry.", r = -EINVAL; goto failed);
+		PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
 				UCODE_ID_SDMA0, &toc->entry[toc->num_entries++]),
-				"Failed to Get Firmware Entry.", return -EINVAL);
-	PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
-				UCODE_ID_SDMA1, &toc->entry[toc->num_entries++]),
-				"Failed to Get Firmware Entry.", return -EINVAL);
-	if (cgs_is_virtualization_enabled(hwmgr->device))
+				"Failed to Get Firmware Entry.", r = -EINVAL; goto failed);
 		PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
+				UCODE_ID_SDMA1, &toc->entry[toc->num_entries++]),
+				"Failed to Get Firmware Entry.", r = -EINVAL; goto failed);
+		if (!hwmgr->not_vf)
+			PP_ASSERT_WITH_CODE(0 == smu7_populate_single_firmware_entry(hwmgr,
 				UCODE_ID_MEC_STORAGE, &toc->entry[toc->num_entries++]),
-				"Failed to Get Firmware Entry.", return -EINVAL);
-
+				"Failed to Get Firmware Entry.", r = -EINVAL; goto failed);
+	}
+	memcpy_toio(smu_data->header_buffer.kaddr, smu_data->toc,
+		    sizeof(struct SMU_DRAMData_TOC));
 	smu7_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_DRV_DRAM_ADDR_HI, upper_32_bits(smu_data->header_buffer.mc_addr));
 	smu7_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_DRV_DRAM_ADDR_LO, lower_32_bits(smu_data->header_buffer.mc_addr));
 
 	if (smu7_send_msg_to_smc_with_parameter(hwmgr, PPSMC_MSG_LoadUcodes, fw_to_load))
 		pr_err("Fail to Request SMU Load uCode");
 
-	return result;
+	return r;
+
+failed:
+	kfree(smu_data->toc);
+	smu_data->toc = NULL;
+	return r;
 }
 
 /* Check if the FW has been loaded, SMU will not return if loading has not finished. */
@@ -585,7 +582,6 @@ int smu7_setup_pwr_virus(struct pp_hwmgr *hwmgr)
 int smu7_init(struct pp_hwmgr *hwmgr)
 {
 	struct smu7_smumgr *smu_data;
-	uint64_t mc_addr = 0;
 	int r;
 	/* Allocate memory for backend private data */
 	smu_data = (struct smu7_smumgr *)(hwmgr->smu_backend);
@@ -599,16 +595,13 @@ int smu7_init(struct pp_hwmgr *hwmgr)
 		PAGE_SIZE,
 		AMDGPU_GEM_DOMAIN_VRAM,
 		&smu_data->header_buffer.handle,
-		&mc_addr,
+		&smu_data->header_buffer.mc_addr,
 		&smu_data->header_buffer.kaddr);
 
 	if (r)
 		return -EINVAL;
 
-	smu_data->header = smu_data->header_buffer.kaddr;
-	smu_data->header_buffer.mc_addr = mc_addr;
-
-	if (cgs_is_virtualization_enabled(hwmgr->device))
+	if (!hwmgr->not_vf)
 		return 0;
 
 	smu_data->smu_buffer.data_size = 200*4096;
@@ -617,7 +610,7 @@ int smu7_init(struct pp_hwmgr *hwmgr)
 		PAGE_SIZE,
 		AMDGPU_GEM_DOMAIN_VRAM,
 		&smu_data->smu_buffer.handle,
-		&mc_addr,
+		&smu_data->smu_buffer.mc_addr,
 		&smu_data->smu_buffer.kaddr);
 
 	if (r) {
@@ -626,7 +619,6 @@ int smu7_init(struct pp_hwmgr *hwmgr)
 					&smu_data->header_buffer.kaddr);
 		return -EINVAL;
 	}
-	smu_data->smu_buffer.mc_addr = mc_addr;
 
 	if (smum_is_hw_avfs_present(hwmgr))
 		hwmgr->avfs_supported = true;
@@ -643,13 +635,15 @@ int smu7_smu_fini(struct pp_hwmgr *hwmgr)
 					&smu_data->header_buffer.mc_addr,
 					&smu_data->header_buffer.kaddr);
 
-	if (!cgs_is_virtualization_enabled(hwmgr->device))
+	if (hwmgr->not_vf)
 		amdgpu_bo_free_kernel(&smu_data->smu_buffer.handle,
 					&smu_data->smu_buffer.mc_addr,
 					&smu_data->smu_buffer.kaddr);
 
+
+	kfree(smu_data->toc);
+	smu_data->toc = NULL;
 	kfree(hwmgr->smu_backend);
 	hwmgr->smu_backend = NULL;
-	cgs_rel_firmware(hwmgr->device, CGS_UCODE_ID_SMU);
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h
index 126d300259ba..01f0538fba6b 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu7_smumgr.h
@@ -37,10 +37,9 @@ struct smu7_buffer_entry {
 };
 
 struct smu7_smumgr {
-	uint8_t *header;
-	uint8_t *mec_image;
 	struct smu7_buffer_entry smu_buffer;
 	struct smu7_buffer_entry header_buffer;
+	struct SMU_DRAMData_TOC *toc;
 
 	uint32_t                             soft_regs_start;
 	uint32_t                             dpm_table_start;
@@ -67,7 +66,6 @@ int smu7_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr, uint16_t msg,
 int smu7_send_msg_to_smc_with_parameter_without_waiting(struct pp_hwmgr *hwmgr,
 						uint16_t msg, uint32_t parameter);
 int smu7_send_msg_to_smc_offset(struct pp_hwmgr *hwmgr);
-int smu7_wait_for_smc_inactive(struct pp_hwmgr *hwmgr);
 
 enum cgs_ucode_id smu7_convert_fw_type_to_cgs(uint32_t fw_type);
 int smu7_read_smc_sram_dword(struct pp_hwmgr *hwmgr, uint32_t smc_addr,
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu8_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu8_smumgr.c
index c861d3023474..f7e3bc22bb93 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smu8_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu8_smumgr.c
@@ -52,10 +52,10 @@ static const enum smu8_scratch_entry firmware_list[] = {
 	SMU8_SCRATCH_ENTRY_UCODE_ID_RLC_G,
 };
 
-static int smu8_get_argument(struct pp_hwmgr *hwmgr)
+static uint32_t smu8_get_argument(struct pp_hwmgr *hwmgr)
 {
 	if (hwmgr == NULL || hwmgr->device == NULL)
-		return -EINVAL;
+		return 0;
 
 	return cgs_read_register(hwmgr->device,
 					mmSMU_MP1_SRBM2P_ARG_0);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu9_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smu9_smumgr.c
new file mode 100644
index 000000000000..079fc8e8f709
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu9_smumgr.c
@@ -0,0 +1,150 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#include "smumgr.h"
+#include "vega10_inc.h"
+#include "soc15_common.h"
+#include "pp_debug.h"
+
+
+/* MP Apertures */
+#define MP0_Public                  0x03800000
+#define MP0_SRAM                    0x03900000
+#define MP1_Public                  0x03b00000
+#define MP1_SRAM                    0x03c00004
+
+#define smnMP1_FIRMWARE_FLAGS                                                                           0x3010028
+
+bool smu9_is_smc_ram_running(struct pp_hwmgr *hwmgr)
+{
+	struct amdgpu_device *adev = hwmgr->adev;
+	uint32_t mp1_fw_flags;
+
+	WREG32_SOC15(NBIF, 0, mmPCIE_INDEX2,
+			(MP1_Public | (smnMP1_FIRMWARE_FLAGS & 0xffffffff)));
+
+	mp1_fw_flags = RREG32_SOC15(NBIF, 0, mmPCIE_DATA2);
+
+	if (mp1_fw_flags & MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED_MASK)
+		return true;
+
+	return false;
+}
+
+/*
+ * Check if SMC has responded to previous message.
+ *
+ * @param    smumgr  the address of the powerplay hardware manager.
+ * @return   TRUE    SMC has responded, FALSE otherwise.
+ */
+static uint32_t smu9_wait_for_response(struct pp_hwmgr *hwmgr)
+{
+	struct amdgpu_device *adev = hwmgr->adev;
+	uint32_t reg;
+	uint32_t ret;
+
+	reg = SOC15_REG_OFFSET(MP1, 0, mmMP1_SMN_C2PMSG_90);
+
+	ret = phm_wait_for_register_unequal(hwmgr, reg,
+			0, MP1_C2PMSG_90__CONTENT_MASK);
+
+	if (ret)
+		pr_err("No response from smu\n");
+
+	return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90);
+}
+
+/*
+ * Send a message to the SMC, and do not wait for its response.
+ * @param    smumgr  the address of the powerplay hardware manager.
+ * @param    msg the message to send.
+ * @return   Always return 0.
+ */
+static int smu9_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr,
+						uint16_t msg)
+{
+	struct amdgpu_device *adev = hwmgr->adev;
+
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_66, msg);
+
+	return 0;
+}
+
+/*
+ * Send a message to the SMC, and wait for its response.
+ * @param    hwmgr  the address of the powerplay hardware manager.
+ * @param    msg the message to send.
+ * @return   Always return 0.
+ */
+int smu9_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
+{
+	struct amdgpu_device *adev = hwmgr->adev;
+	uint32_t ret;
+
+	smu9_wait_for_response(hwmgr);
+
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0);
+
+	smu9_send_msg_to_smc_without_waiting(hwmgr, msg);
+
+	ret = smu9_wait_for_response(hwmgr);
+	if (ret != 1)
+		pr_err("Failed to send message: 0x%x, ret value: 0x%x\n", msg, ret);
+
+	return 0;
+}
+
+/*
+ * Send a message to the SMC with parameter
+ * @param    hwmgr:  the address of the powerplay hardware manager.
+ * @param    msg: the message to send.
+ * @param    parameter: the parameter to send
+ * @return   Always return 0.
+ */
+int smu9_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr,
+					uint16_t msg, uint32_t parameter)
+{
+	struct amdgpu_device *adev = hwmgr->adev;
+	uint32_t ret;
+
+	smu9_wait_for_response(hwmgr);
+
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0);
+
+	WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82, parameter);
+
+	smu9_send_msg_to_smc_without_waiting(hwmgr, msg);
+
+	ret = smu9_wait_for_response(hwmgr);
+	if (ret != 1)
+		pr_err("Failed message: 0x%x, input parameter: 0x%x, error code: 0x%x\n", msg, parameter, ret);
+
+	return 0;
+}
+
+uint32_t smu9_get_argument(struct pp_hwmgr *hwmgr)
+{
+	struct amdgpu_device *adev = hwmgr->adev;
+
+	return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82);
+}
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smu9_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/smu9_smumgr.h
new file mode 100644
index 000000000000..1462279ca128
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smu9_smumgr.h
@@ -0,0 +1,32 @@
+/*
+ * Copyright 2018 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#ifndef _SMU9_SMUMANAGER_H_
+#define _SMU9_SMUMANAGER_H_
+
+bool smu9_is_smc_ram_running(struct pp_hwmgr *hwmgr);
+int smu9_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg);
+int smu9_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr,
+					uint16_t msg, uint32_t parameter);
+uint32_t smu9_get_argument(struct pp_hwmgr *hwmgr);
+
+#endif
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
index c28b60aae5f8..99d5e4f98f49 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/smumgr.c
@@ -41,9 +41,11 @@ MODULE_FIRMWARE("amdgpu/polaris11_smc.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_smc_sk.bin");
 MODULE_FIRMWARE("amdgpu/polaris11_k_smc.bin");
 MODULE_FIRMWARE("amdgpu/polaris12_smc.bin");
+MODULE_FIRMWARE("amdgpu/vegam_smc.bin");
 MODULE_FIRMWARE("amdgpu/vega10_smc.bin");
 MODULE_FIRMWARE("amdgpu/vega10_acg_smc.bin");
 MODULE_FIRMWARE("amdgpu/vega12_smc.bin");
+MODULE_FIRMWARE("amdgpu/vega20_smc.bin");
 
 int smum_thermal_avfs_enable(struct pp_hwmgr *hwmgr)
 {
@@ -94,7 +96,7 @@ int smum_process_firmware_header(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
-int smum_get_argument(struct pp_hwmgr *hwmgr)
+uint32_t smum_get_argument(struct pp_hwmgr *hwmgr)
 {
 	if (NULL != hwmgr->smumgr_funcs->get_argument)
 		return hwmgr->smumgr_funcs->get_argument(hwmgr);
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c
index b51d7468c3e7..7dabc6c456e1 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c
@@ -55,11 +55,7 @@
 #include "dce/dce_10_0_d.h"
 #include "dce/dce_10_0_sh_mask.h"
 
-
-#define VOLTAGE_SCALE 4
 #define POWERTUNE_DEFAULT_SET_MAX    1
-#define VOLTAGE_VID_OFFSET_SCALE1   625
-#define VOLTAGE_VID_OFFSET_SCALE2   100
 #define MC_CG_ARB_FREQ_F1           0x0b
 #define VDDC_VDDCI_DELTA            200
 
@@ -199,8 +195,7 @@ static int tonga_start_smu(struct pp_hwmgr *hwmgr)
 	int result;
 
 	/* Only start SMC if SMC RAM is not running */
-	if (!(smu7_is_smc_ram_running(hwmgr) ||
-		cgs_is_virtualization_enabled(hwmgr->device))) {
+	if (!smu7_is_smc_ram_running(hwmgr) && hwmgr->not_vf) {
 		/*Check if SMU is running in protected mode*/
 		if (0 == PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
 					SMU_FIRMWARE, SMU_MODE)) {
@@ -651,7 +646,7 @@ static int tonga_populate_single_graphic_level(struct pp_hwmgr *hwmgr,
 	graphic_level->PowerThrottle = 0;
 
 	data->display_timing.min_clock_in_sr =
-			hwmgr->display_config.min_core_set_clock_in_sr;
+			hwmgr->display_config->min_core_set_clock_in_sr;
 
 	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
 			PHM_PlatformCaps_SclkDeepSleep))
@@ -957,18 +952,17 @@ static int tonga_populate_single_memory_level(
 		SMU72_Discrete_MemoryLevel *memory_level
 		)
 {
-	uint32_t mvdd = 0;
 	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
 	struct phm_ppt_v1_information *pptable_info =
 			  (struct phm_ppt_v1_information *)(hwmgr->pptable);
-	int result = 0;
-	bool dll_state_on;
-	struct cgs_display_info info = {0};
 	uint32_t mclk_edc_wr_enable_threshold = 40000;
 	uint32_t mclk_stutter_mode_threshold = 30000;
 	uint32_t mclk_edc_enable_threshold = 40000;
 	uint32_t mclk_strobe_mode_threshold = 40000;
 	phm_ppt_v1_clock_voltage_dependency_table *vdd_dep_table = NULL;
+	int result = 0;
+	bool dll_state_on;
+	uint32_t mvdd = 0;
 
 	if (hwmgr->od_enabled)
 		vdd_dep_table = (phm_ppt_v1_clock_voltage_dependency_table *)&data->odn_dpm_table.vdd_dependency_on_mclk;
@@ -1009,8 +1003,7 @@ static int tonga_populate_single_memory_level(
 	/* default set to low watermark. Highest level will be set to high later.*/
 	memory_level->DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW;
 
-	cgs_get_active_displays_info(hwmgr->device, &info);
-	data->display_timing.num_existing_displays = info.display_count;
+	data->display_timing.num_existing_displays = hwmgr->display_config->num_display;
 
 	if ((mclk_stutter_mode_threshold != 0) &&
 	    (memory_clock <= mclk_stutter_mode_threshold) &&
@@ -1450,51 +1443,6 @@ static int tonga_populate_smc_acp_level(struct pp_hwmgr *hwmgr,
 	return result;
 }
 
-static int tonga_populate_smc_samu_level(struct pp_hwmgr *hwmgr,
-		SMU72_Discrete_DpmTable *table)
-{
-	int result = 0;
-	uint8_t count;
-	pp_atomctrl_clock_dividers_vi dividers;
-	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
-	struct phm_ppt_v1_information *pptable_info =
-			     (struct phm_ppt_v1_information *)(hwmgr->pptable);
-	phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table =
-						    pptable_info->mm_dep_table;
-
-	table->SamuBootLevel = 0;
-	table->SamuLevelCount = (uint8_t) (mm_table->count);
-
-	for (count = 0; count < table->SamuLevelCount; count++) {
-		/* not sure whether we need evclk or not */
-		table->SamuLevel[count].Frequency =
-			pptable_info->mm_dep_table->entries[count].samclock;
-		table->SamuLevel[count].MinVoltage.Vddc =
-			phm_get_voltage_index(pptable_info->vddc_lookup_table,
-				mm_table->entries[count].vddc);
-		table->SamuLevel[count].MinVoltage.VddGfx =
-			(data->vdd_gfx_control == SMU7_VOLTAGE_CONTROL_BY_SVID2) ?
-			phm_get_voltage_index(pptable_info->vddgfx_lookup_table,
-				mm_table->entries[count].vddgfx) : 0;
-		table->SamuLevel[count].MinVoltage.Vddci =
-			phm_get_voltage_id(&data->vddci_voltage_table,
-				mm_table->entries[count].vddc - VDDC_VDDCI_DELTA);
-		table->SamuLevel[count].MinVoltage.Phases = 1;
-
-		/* retrieve divider value for VBIOS */
-		result = atomctrl_get_dfs_pll_dividers_vi(hwmgr,
-					table->SamuLevel[count].Frequency, &dividers);
-		PP_ASSERT_WITH_CODE((!result),
-			"can not find divide id for samu clock", return result);
-
-		table->SamuLevel[count].Divider = (uint8_t)dividers.pll_post_divider;
-
-		CONVERT_FROM_HOST_TO_SMC_UL(table->SamuLevel[count].Frequency);
-	}
-
-	return result;
-}
-
 static int tonga_populate_memory_timing_parameters(
 		struct pp_hwmgr *hwmgr,
 		uint32_t engine_clock,
@@ -2330,10 +2278,6 @@ static int tonga_init_smc_table(struct pp_hwmgr *hwmgr)
 	PP_ASSERT_WITH_CODE(!result,
 		"Failed to initialize ACP Level !", return result);
 
-	result = tonga_populate_smc_samu_level(hwmgr, table);
-	PP_ASSERT_WITH_CODE(!result,
-		"Failed to initialize SAMU Level !", return result);
-
 	/* Since only the initial state is completely set up at this
 	* point (the other states are just copies of the boot state) we only
 	* need to populate the  ARB settings for the initial state.
@@ -2680,8 +2624,6 @@ static uint32_t tonga_get_offsetof(uint32_t type, uint32_t member)
 			return offsetof(SMU72_Discrete_DpmTable, UvdBootLevel);
 		case VceBootLevel:
 			return offsetof(SMU72_Discrete_DpmTable, VceBootLevel);
-		case SamuBootLevel:
-			return offsetof(SMU72_Discrete_DpmTable, SamuBootLevel);
 		case LowSclkInterruptThreshold:
 			return offsetof(SMU72_Discrete_DpmTable, LowSclkInterruptThreshold);
 		}
@@ -2780,32 +2722,6 @@ static int tonga_update_vce_smc_table(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
-static int tonga_update_samu_smc_table(struct pp_hwmgr *hwmgr)
-{
-	struct tonga_smumgr *smu_data = (struct tonga_smumgr *)(hwmgr->smu_backend);
-	uint32_t mm_boot_level_offset, mm_boot_level_value;
-
-	smu_data->smc_state_table.SamuBootLevel = 0;
-	mm_boot_level_offset = smu_data->smu7_data.dpm_table_start +
-				offsetof(SMU72_Discrete_DpmTable, SamuBootLevel);
-
-	mm_boot_level_offset /= 4;
-	mm_boot_level_offset *= 4;
-	mm_boot_level_value = cgs_read_ind_register(hwmgr->device,
-			CGS_IND_REG__SMC, mm_boot_level_offset);
-	mm_boot_level_value &= 0xFFFFFF00;
-	mm_boot_level_value |= smu_data->smc_state_table.SamuBootLevel << 0;
-	cgs_write_ind_register(hwmgr->device,
-			CGS_IND_REG__SMC, mm_boot_level_offset, mm_boot_level_value);
-
-	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
-			PHM_PlatformCaps_StablePState))
-		smum_send_msg_to_smc_with_parameter(hwmgr,
-				PPSMC_MSG_SAMUDPM_SetEnabledMask,
-				(uint32_t)(1 << smu_data->smc_state_table.SamuBootLevel));
-	return 0;
-}
-
 static int tonga_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type)
 {
 	switch (type) {
@@ -2815,9 +2731,6 @@ static int tonga_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type)
 	case SMU_VCE_TABLE:
 		tonga_update_vce_smc_table(hwmgr);
 		break;
-	case SMU_SAMU_TABLE:
-		tonga_update_samu_smc_table(hwmgr);
-		break;
 	default:
 		break;
 	}
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c
index 4aafb043bcb0..5d19115f410c 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega10_smumgr.c
@@ -23,165 +23,16 @@
 
 #include "smumgr.h"
 #include "vega10_inc.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
 #include "vega10_smumgr.h"
 #include "vega10_hwmgr.h"
 #include "vega10_ppsmc.h"
 #include "smu9_driver_if.h"
+#include "smu9_smumgr.h"
 #include "ppatomctrl.h"
 #include "pp_debug.h"
 
 
-#define AVFS_EN_MSB		1568
-#define AVFS_EN_LSB		1568
-
-#define VOLTAGE_SCALE	4
-
-/* Microcode file is stored in this buffer */
-#define BUFFER_SIZE                 80000
-#define MAX_STRING_SIZE             15
-#define BUFFER_SIZETWO              131072 /* 128 *1024 */
-
-/* MP Apertures */
-#define MP0_Public                  0x03800000
-#define MP0_SRAM                    0x03900000
-#define MP1_Public                  0x03b00000
-#define MP1_SRAM                    0x03c00004
-
-#define smnMP1_FIRMWARE_FLAGS                                                                           0x3010028
-#define smnMP0_FW_INTF                                                                                  0x3010104
-#define smnMP1_PUB_CTRL                                                                                 0x3010b14
-
-static bool vega10_is_smc_ram_running(struct pp_hwmgr *hwmgr)
-{
-	uint32_t mp1_fw_flags, reg;
-
-	reg = soc15_get_register_offset(NBIF_HWID, 0,
-			mmPCIE_INDEX2_BASE_IDX, mmPCIE_INDEX2);
-
-	cgs_write_register(hwmgr->device, reg,
-			(MP1_Public | (smnMP1_FIRMWARE_FLAGS & 0xffffffff)));
-
-	reg = soc15_get_register_offset(NBIF_HWID, 0,
-			mmPCIE_DATA2_BASE_IDX, mmPCIE_DATA2);
-
-	mp1_fw_flags = cgs_read_register(hwmgr->device, reg);
-
-	if (mp1_fw_flags & MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED_MASK)
-		return true;
-
-	return false;
-}
-
-/*
- * Check if SMC has responded to previous message.
- *
- * @param    smumgr  the address of the powerplay hardware manager.
- * @return   TRUE    SMC has responded, FALSE otherwise.
- */
-static uint32_t vega10_wait_for_response(struct pp_hwmgr *hwmgr)
-{
-	uint32_t reg;
-	uint32_t ret;
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
-
-	ret = phm_wait_for_register_unequal(hwmgr, reg,
-			0, MP1_C2PMSG_90__CONTENT_MASK);
-
-	if (ret)
-		pr_err("No response from smu\n");
-
-	return cgs_read_register(hwmgr->device, reg);
-}
-
-/*
- * Send a message to the SMC, and do not wait for its response.
- * @param    smumgr  the address of the powerplay hardware manager.
- * @param    msg the message to send.
- * @return   Always return 0.
- */
-static int vega10_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr,
-		uint16_t msg)
-{
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_66_BASE_IDX, mmMP1_SMN_C2PMSG_66);
-	cgs_write_register(hwmgr->device, reg, msg);
-
-	return 0;
-}
-
-/*
- * Send a message to the SMC, and wait for its response.
- * @param    hwmgr  the address of the powerplay hardware manager.
- * @param    msg the message to send.
- * @return   Always return 0.
- */
-static int vega10_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
-{
-	uint32_t reg;
-	uint32_t ret;
-
-	vega10_wait_for_response(hwmgr);
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
-	cgs_write_register(hwmgr->device, reg, 0);
-
-	vega10_send_msg_to_smc_without_waiting(hwmgr, msg);
-
-	ret = vega10_wait_for_response(hwmgr);
-	if (ret != 1)
-		pr_err("Failed to send message: 0x%x, ret value: 0x%x\n", msg, ret);
-
-	return 0;
-}
-
-/*
- * Send a message to the SMC with parameter
- * @param    hwmgr:  the address of the powerplay hardware manager.
- * @param    msg: the message to send.
- * @param    parameter: the parameter to send
- * @return   Always return 0.
- */
-static int vega10_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr,
-		uint16_t msg, uint32_t parameter)
-{
-	uint32_t reg;
-	uint32_t ret;
-
-	vega10_wait_for_response(hwmgr);
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
-	cgs_write_register(hwmgr->device, reg, 0);
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_82_BASE_IDX, mmMP1_SMN_C2PMSG_82);
-	cgs_write_register(hwmgr->device, reg, parameter);
-
-	vega10_send_msg_to_smc_without_waiting(hwmgr, msg);
-
-	ret = vega10_wait_for_response(hwmgr);
-	if (ret != 1)
-		pr_err("Failed message: 0x%x, input parameter: 0x%x, error code: 0x%x\n", msg, parameter, ret);
-
-	return 0;
-}
-
-static int vega10_get_argument(struct pp_hwmgr *hwmgr)
-{
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_82_BASE_IDX, mmMP1_SMN_C2PMSG_82);
-
-	return cgs_read_register(hwmgr->device, reg);
-}
-
 static int vega10_copy_table_from_smc(struct pp_hwmgr *hwmgr,
 		uint8_t *table, int16_t table_id)
 {
@@ -193,13 +44,13 @@ static int vega10_copy_table_from_smc(struct pp_hwmgr *hwmgr,
 			"Invalid SMU Table version!", return -EINVAL);
 	PP_ASSERT_WITH_CODE(priv->smu_tables.entry[table_id].size != 0,
 			"Invalid SMU Table Length!", return -EINVAL);
-	vega10_send_msg_to_smc_with_parameter(hwmgr,
+	smu9_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_SetDriverDramAddrHigh,
 			upper_32_bits(priv->smu_tables.entry[table_id].mc_addr));
-	vega10_send_msg_to_smc_with_parameter(hwmgr,
+	smu9_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_SetDriverDramAddrLow,
 			lower_32_bits(priv->smu_tables.entry[table_id].mc_addr));
-	vega10_send_msg_to_smc_with_parameter(hwmgr,
+	smu9_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_TransferTableSmu2Dram,
 			priv->smu_tables.entry[table_id].table_id);
 
@@ -224,13 +75,13 @@ static int vega10_copy_table_to_smc(struct pp_hwmgr *hwmgr,
 	memcpy(priv->smu_tables.entry[table_id].table, table,
 			priv->smu_tables.entry[table_id].size);
 
-	vega10_send_msg_to_smc_with_parameter(hwmgr,
+	smu9_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_SetDriverDramAddrHigh,
 			upper_32_bits(priv->smu_tables.entry[table_id].mc_addr));
-	vega10_send_msg_to_smc_with_parameter(hwmgr,
+	smu9_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_SetDriverDramAddrLow,
 			lower_32_bits(priv->smu_tables.entry[table_id].mc_addr));
-	vega10_send_msg_to_smc_with_parameter(hwmgr,
+	smu9_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_TransferTableDram2Smu,
 			priv->smu_tables.entry[table_id].table_id);
 
@@ -243,8 +94,8 @@ static int vega10_get_smc_features(struct pp_hwmgr *hwmgr,
 	if (features_enabled == NULL)
 		return -EINVAL;
 
-	vega10_send_msg_to_smc(hwmgr, PPSMC_MSG_GetEnabledSmuFeatures);
-	*features_enabled = vega10_get_argument(hwmgr);
+	smu9_send_msg_to_smc(hwmgr, PPSMC_MSG_GetEnabledSmuFeatures);
+	*features_enabled = smu9_get_argument(hwmgr);
 
 	return 0;
 }
@@ -266,10 +117,10 @@ static int vega10_set_tools_address(struct pp_hwmgr *hwmgr)
 	struct vega10_smumgr *priv = hwmgr->smu_backend;
 
 	if (priv->smu_tables.entry[TOOLSTABLE].mc_addr) {
-		vega10_send_msg_to_smc_with_parameter(hwmgr,
+		smu9_send_msg_to_smc_with_parameter(hwmgr,
 				PPSMC_MSG_SetToolsDramAddrHigh,
 				upper_32_bits(priv->smu_tables.entry[TOOLSTABLE].mc_addr));
-		vega10_send_msg_to_smc_with_parameter(hwmgr,
+		smu9_send_msg_to_smc_with_parameter(hwmgr,
 				PPSMC_MSG_SetToolsDramAddrLow,
 				lower_32_bits(priv->smu_tables.entry[TOOLSTABLE].mc_addr));
 	}
@@ -283,11 +134,11 @@ static int vega10_verify_smc_interface(struct pp_hwmgr *hwmgr)
 	uint32_t dev_id;
 	uint32_t rev_id;
 
-	PP_ASSERT_WITH_CODE(!vega10_send_msg_to_smc(hwmgr,
+	PP_ASSERT_WITH_CODE(!smu9_send_msg_to_smc(hwmgr,
 			PPSMC_MSG_GetDriverIfVersion),
 			"Attempt to get SMC IF Version Number Failed!",
 			return -EINVAL);
-	smc_driver_if_version = vega10_get_argument(hwmgr);
+	smc_driver_if_version = smu9_get_argument(hwmgr);
 
 	dev_id = adev->pdev->device;
 	rev_id = adev->pdev->revision;
@@ -459,7 +310,7 @@ static int vega10_smu_fini(struct pp_hwmgr *hwmgr)
 
 static int vega10_start_smu(struct pp_hwmgr *hwmgr)
 {
-	if (!vega10_is_smc_ram_running(hwmgr))
+	if (!smu9_is_smc_ram_running(hwmgr))
 		return -EINVAL;
 
 	PP_ASSERT_WITH_CODE(!vega10_verify_smc_interface(hwmgr),
@@ -471,7 +322,8 @@ static int vega10_start_smu(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
-static int vega10_smc_table_manager(struct pp_hwmgr *hwmgr, uint8_t *table, uint16_t table_id, bool rw)
+static int vega10_smc_table_manager(struct pp_hwmgr *hwmgr, uint8_t *table,
+				    uint16_t table_id, bool rw)
 {
 	int ret;
 
@@ -488,11 +340,11 @@ const struct pp_smumgr_func vega10_smu_funcs = {
 	.smu_fini = &vega10_smu_fini,
 	.start_smu = &vega10_start_smu,
 	.request_smu_load_specific_fw = NULL,
-	.send_msg_to_smc = &vega10_send_msg_to_smc,
-	.send_msg_to_smc_with_parameter = &vega10_send_msg_to_smc_with_parameter,
+	.send_msg_to_smc = &smu9_send_msg_to_smc,
+	.send_msg_to_smc_with_parameter = &smu9_send_msg_to_smc_with_parameter,
 	.download_pptable_settings = NULL,
 	.upload_pptable_settings = NULL,
 	.is_dpm_running = vega10_is_dpm_running,
-	.get_argument = vega10_get_argument,
+	.get_argument = smu9_get_argument,
 	.smc_table_manager = vega10_smc_table_manager,
 };
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c
index 651a3f28734b..7f0e2109f40d 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.c
@@ -23,176 +23,15 @@
 
 #include "smumgr.h"
 #include "vega12_inc.h"
-#include "pp_soc15.h"
+#include "soc15_common.h"
+#include "smu9_smumgr.h"
 #include "vega12_smumgr.h"
 #include "vega12_ppsmc.h"
 #include "vega12/smu9_driver_if.h"
-
 #include "ppatomctrl.h"
 #include "pp_debug.h"
 
 
-/* MP Apertures */
-#define MP0_Public                  0x03800000
-#define MP0_SRAM                    0x03900000
-#define MP1_Public                  0x03b00000
-#define MP1_SRAM                    0x03c00004
-
-#define smnMP1_FIRMWARE_FLAGS                                                                           0x3010028
-#define smnMP0_FW_INTF                                                                                  0x3010104
-#define smnMP1_PUB_CTRL                                                                                 0x3010b14
-
-static bool vega12_is_smc_ram_running(struct pp_hwmgr *hwmgr)
-{
-	uint32_t mp1_fw_flags, reg;
-
-	reg = soc15_get_register_offset(NBIF_HWID, 0,
-			mmPCIE_INDEX2_BASE_IDX, mmPCIE_INDEX2);
-
-	cgs_write_register(hwmgr->device, reg,
-			(MP1_Public | (smnMP1_FIRMWARE_FLAGS & 0xffffffff)));
-
-	reg = soc15_get_register_offset(NBIF_HWID, 0,
-			mmPCIE_DATA2_BASE_IDX, mmPCIE_DATA2);
-
-	mp1_fw_flags = cgs_read_register(hwmgr->device, reg);
-
-	if ((mp1_fw_flags & MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED_MASK) >>
-				MP1_FIRMWARE_FLAGS__INTERRUPTS_ENABLED__SHIFT)
-		return true;
-
-	return false;
-}
-
-/*
- * Check if SMC has responded to previous message.
- *
- * @param    smumgr  the address of the powerplay hardware manager.
- * @return   TRUE    SMC has responded, FALSE otherwise.
- */
-static uint32_t vega12_wait_for_response(struct pp_hwmgr *hwmgr)
-{
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
-
-	phm_wait_for_register_unequal(hwmgr, reg,
-			0, MP1_C2PMSG_90__CONTENT_MASK);
-
-	return cgs_read_register(hwmgr->device, reg);
-}
-
-/*
- * Send a message to the SMC, and do not wait for its response.
- * @param    smumgr  the address of the powerplay hardware manager.
- * @param    msg the message to send.
- * @return   Always return 0.
- */
-int vega12_send_msg_to_smc_without_waiting(struct pp_hwmgr *hwmgr,
-		uint16_t msg)
-{
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_66_BASE_IDX, mmMP1_SMN_C2PMSG_66);
-	cgs_write_register(hwmgr->device, reg, msg);
-
-	return 0;
-}
-
-/*
- * Send a message to the SMC, and wait for its response.
- * @param    hwmgr  the address of the powerplay hardware manager.
- * @param    msg the message to send.
- * @return   Always return 0.
- */
-int vega12_send_msg_to_smc(struct pp_hwmgr *hwmgr, uint16_t msg)
-{
-	uint32_t reg;
-
-	vega12_wait_for_response(hwmgr);
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
-	cgs_write_register(hwmgr->device, reg, 0);
-
-	vega12_send_msg_to_smc_without_waiting(hwmgr, msg);
-
-	if (vega12_wait_for_response(hwmgr) != 1)
-		pr_err("Failed to send message: 0x%x\n", msg);
-
-	return 0;
-}
-
-/*
- * Send a message to the SMC with parameter
- * @param    hwmgr:  the address of the powerplay hardware manager.
- * @param    msg: the message to send.
- * @param    parameter: the parameter to send
- * @return   Always return 0.
- */
-int vega12_send_msg_to_smc_with_parameter(struct pp_hwmgr *hwmgr,
-		uint16_t msg, uint32_t parameter)
-{
-	uint32_t reg;
-
-	vega12_wait_for_response(hwmgr);
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_90_BASE_IDX, mmMP1_SMN_C2PMSG_90);
-	cgs_write_register(hwmgr->device, reg, 0);
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_82_BASE_IDX, mmMP1_SMN_C2PMSG_82);
-	cgs_write_register(hwmgr->device, reg, parameter);
-
-	vega12_send_msg_to_smc_without_waiting(hwmgr, msg);
-
-	if (vega12_wait_for_response(hwmgr) != 1)
-		pr_err("Failed to send message: 0x%x\n", msg);
-
-	return 0;
-}
-
-
-/*
- * Send a message to the SMC with parameter, do not wait for response
- * @param    hwmgr:  the address of the powerplay hardware manager.
- * @param    msg: the message to send.
- * @param    parameter: the parameter to send
- * @return   The response that came from the SMC.
- */
-int vega12_send_msg_to_smc_with_parameter_without_waiting(
-		struct pp_hwmgr *hwmgr, uint16_t msg, uint32_t parameter)
-{
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_66_BASE_IDX, mmMP1_SMN_C2PMSG_66);
-	cgs_write_register(hwmgr->device, reg, parameter);
-
-	return vega12_send_msg_to_smc_without_waiting(hwmgr, msg);
-}
-
-/*
- * Retrieve an argument from SMC.
- * @param    hwmgr  the address of the powerplay hardware manager.
- * @param    arg     pointer to store the argument from SMC.
- * @return   Always return 0.
- */
-int vega12_read_arg_from_smc(struct pp_hwmgr *hwmgr, uint32_t *arg)
-{
-	uint32_t reg;
-
-	reg = soc15_get_register_offset(MP1_HWID, 0,
-			mmMP1_SMN_C2PMSG_82_BASE_IDX, mmMP1_SMN_C2PMSG_82);
-
-	*arg = cgs_read_register(hwmgr->device, reg);
-
-	return 0;
-}
-
 /*
  * Copy table from SMC into driver FB
  * @param   hwmgr    the address of the HW manager
@@ -210,16 +49,16 @@ int vega12_copy_table_from_smc(struct pp_hwmgr *hwmgr,
 			"Invalid SMU Table version!", return -EINVAL);
 	PP_ASSERT_WITH_CODE(priv->smu_tables.entry[table_id].size != 0,
 			"Invalid SMU Table Length!", return -EINVAL);
-	PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr,
+	PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_SetDriverDramAddrHigh,
 			upper_32_bits(priv->smu_tables.entry[table_id].mc_addr)) == 0,
 			"[CopyTableFromSMC] Attempt to Set Dram Addr High Failed!", return -EINVAL);
-	PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr,
+	PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_SetDriverDramAddrLow,
 			lower_32_bits(priv->smu_tables.entry[table_id].mc_addr)) == 0,
 			"[CopyTableFromSMC] Attempt to Set Dram Addr Low Failed!",
 			return -EINVAL);
-	PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr,
+	PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_TransferTableSmu2Dram,
 			table_id) == 0,
 			"[CopyTableFromSMC] Attempt to Transfer Table From SMU Failed!",
@@ -252,17 +91,17 @@ int vega12_copy_table_to_smc(struct pp_hwmgr *hwmgr,
 	memcpy(priv->smu_tables.entry[table_id].table, table,
 			priv->smu_tables.entry[table_id].size);
 
-	PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr,
+	PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_SetDriverDramAddrHigh,
 			upper_32_bits(priv->smu_tables.entry[table_id].mc_addr)) == 0,
 			"[CopyTableToSMC] Attempt to Set Dram Addr High Failed!",
 			return -EINVAL;);
-	PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr,
+	PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_SetDriverDramAddrLow,
 			lower_32_bits(priv->smu_tables.entry[table_id].mc_addr)) == 0,
 			"[CopyTableToSMC] Attempt to Set Dram Addr Low Failed!",
 			return -EINVAL);
-	PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr,
+	PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr,
 			PPSMC_MSG_TransferTableDram2Smu,
 			table_id) == 0,
 			"[CopyTableToSMC] Attempt to Transfer Table To SMU Failed!",
@@ -280,20 +119,20 @@ int vega12_enable_smc_features(struct pp_hwmgr *hwmgr,
 	smu_features_high = (uint32_t)((feature_mask & SMU_FEATURES_HIGH_MASK) >> SMU_FEATURES_HIGH_SHIFT);
 
 	if (enable) {
-		PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr,
+		PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr,
 				PPSMC_MSG_EnableSmuFeaturesLow, smu_features_low) == 0,
 				"[EnableDisableSMCFeatures] Attemp to enable SMU features Low failed!",
 				return -EINVAL);
-		PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr,
+		PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr,
 				PPSMC_MSG_EnableSmuFeaturesHigh, smu_features_high) == 0,
 				"[EnableDisableSMCFeatures] Attemp to enable SMU features High failed!",
 				return -EINVAL);
 	} else {
-		PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr,
+		PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr,
 				PPSMC_MSG_DisableSmuFeaturesLow, smu_features_low) == 0,
 				"[EnableDisableSMCFeatures] Attemp to disable SMU features Low failed!",
 				return -EINVAL);
-		PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc_with_parameter(hwmgr,
+		PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc_with_parameter(hwmgr,
 				PPSMC_MSG_DisableSmuFeaturesHigh, smu_features_high) == 0,
 				"[EnableDisableSMCFeatures] Attemp to disable SMU features High failed!",
 				return -EINVAL);
@@ -310,22 +149,17 @@ int vega12_get_enabled_smc_features(struct pp_hwmgr *hwmgr,
 	if (features_enabled == NULL)
 		return -EINVAL;
 
-	PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc(hwmgr,
+	PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc(hwmgr,
 			PPSMC_MSG_GetEnabledSmuFeaturesLow) == 0,
 			"[GetEnabledSMCFeatures] Attemp to get SMU features Low failed!",
 			return -EINVAL);
-	PP_ASSERT_WITH_CODE(vega12_read_arg_from_smc(hwmgr,
-			&smc_features_low) == 0,
-			"[GetEnabledSMCFeatures] Attemp to read SMU features Low argument failed!",
-			return -EINVAL);
-	PP_ASSERT_WITH_CODE(vega12_send_msg_to_smc(hwmgr,
+	smc_features_low = smu9_get_argument(hwmgr);
+
+	PP_ASSERT_WITH_CODE(smu9_send_msg_to_smc(hwmgr,
 			PPSMC_MSG_GetEnabledSmuFeaturesHigh) == 0,
 			"[GetEnabledSMCFeatures] Attemp to get SMU features High failed!",
 			return -EINVAL);
-	PP_ASSERT_WITH_CODE(vega12_read_arg_from_smc(hwmgr,
-			&smc_features_high) == 0,
-			"[GetEnabledSMCFeatures] Attemp to read SMU features High argument failed!",
-			return -EINVAL);
+	smc_features_high = smu9_get_argument(hwmgr);
 
 	*features_enabled = ((((uint64_t)smc_features_low << SMU_FEATURES_LOW_SHIFT) & SMU_FEATURES_LOW_MASK) |
 			(((uint64_t)smc_features_high << SMU_FEATURES_HIGH_SHIFT) & SMU_FEATURES_HIGH_MASK));
@@ -351,39 +185,16 @@ static int vega12_set_tools_address(struct pp_hwmgr *hwmgr)
 			(struct vega12_smumgr *)(hwmgr->smu_backend);
 
 	if (priv->smu_tables.entry[TABLE_PMSTATUSLOG].mc_addr) {
-		if (!vega12_send_msg_to_smc_with_parameter(hwmgr,
+		if (!smu9_send_msg_to_smc_with_parameter(hwmgr,
 				PPSMC_MSG_SetToolsDramAddrHigh,
 				upper_32_bits(priv->smu_tables.entry[TABLE_PMSTATUSLOG].mc_addr)))
-			vega12_send_msg_to_smc_with_parameter(hwmgr,
+			smu9_send_msg_to_smc_with_parameter(hwmgr,
 					PPSMC_MSG_SetToolsDramAddrLow,
 					lower_32_bits(priv->smu_tables.entry[TABLE_PMSTATUSLOG].mc_addr));
 	}
 	return 0;
 }
 
-#if 0 /* tentatively remove */
-static int vega12_verify_smc_interface(struct pp_hwmgr *hwmgr)
-{
-	uint32_t smc_driver_if_version;
-
-	PP_ASSERT_WITH_CODE(!vega12_send_msg_to_smc(hwmgr,
-			PPSMC_MSG_GetDriverIfVersion),
-			"Attempt to get SMC IF Version Number Failed!",
-			return -EINVAL);
-	vega12_read_arg_from_smc(hwmgr, &smc_driver_if_version);
-
-	if (smc_driver_if_version != SMU9_DRIVER_IF_VERSION) {
-		pr_err("Your firmware(0x%x) doesn't match \
-			SMU9_DRIVER_IF_VERSION(0x%x). \
-			Please update your firmware!\n",
-			smc_driver_if_version, SMU9_DRIVER_IF_VERSION);
-		return -EINVAL;
-	}
-
-	return 0;
-}
-#endif
-
 static int vega12_smu_init(struct pp_hwmgr *hwmgr)
 {
 	struct vega12_smumgr *priv;
@@ -531,16 +342,10 @@ static int vega12_smu_fini(struct pp_hwmgr *hwmgr)
 
 static int vega12_start_smu(struct pp_hwmgr *hwmgr)
 {
-	PP_ASSERT_WITH_CODE(vega12_is_smc_ram_running(hwmgr),
+	PP_ASSERT_WITH_CODE(smu9_is_smc_ram_running(hwmgr),
 			"SMC is not running!",
 			return -EINVAL);
 
-#if 0 /* tentatively remove */
-	PP_ASSERT_WITH_CODE(!vega12_verify_smc_interface(hwmgr),
-			"Failed to verify SMC interface!",
-			return -EINVAL);
-#endif
-
 	vega12_set_tools_address(hwmgr);
 
 	return 0;
@@ -551,9 +356,10 @@ const struct pp_smumgr_func vega12_smu_funcs = {
 	.smu_fini = &vega12_smu_fini,
 	.start_smu = &vega12_start_smu,
 	.request_smu_load_specific_fw = NULL,
-	.send_msg_to_smc = &vega12_send_msg_to_smc,
-	.send_msg_to_smc_with_parameter = &vega12_send_msg_to_smc_with_parameter,
+	.send_msg_to_smc = &smu9_send_msg_to_smc,
+	.send_msg_to_smc_with_parameter = &smu9_send_msg_to_smc_with_parameter,
 	.download_pptable_settings = NULL,
 	.upload_pptable_settings = NULL,
 	.is_dpm_running = vega12_is_dpm_running,
+	.get_argument = smu9_get_argument,
 };
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.h
index 2810d387b611..b285cbc04019 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.h
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vega12_smumgr.h
@@ -48,7 +48,6 @@ struct vega12_smumgr {
 #define SMU_FEATURES_HIGH_MASK       0xFFFFFFFF00000000
 #define SMU_FEATURES_HIGH_SHIFT      32
 
-int vega12_read_arg_from_smc(struct pp_hwmgr *hwmgr, uint32_t *arg);
 int vega12_copy_table_from_smc(struct pp_hwmgr *hwmgr,
 		uint8_t *table, int16_t table_id);
 int vega12_copy_table_to_smc(struct pp_hwmgr *hwmgr,
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c
new file mode 100644
index 000000000000..57420d7caa4e
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.c
@@ -0,0 +1,2298 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+#include "pp_debug.h"
+#include "smumgr.h"
+#include "smu_ucode_xfer_vi.h"
+#include "vegam_smumgr.h"
+#include "smu/smu_7_1_3_d.h"
+#include "smu/smu_7_1_3_sh_mask.h"
+#include "gmc/gmc_8_1_d.h"
+#include "gmc/gmc_8_1_sh_mask.h"
+#include "oss/oss_3_0_d.h"
+#include "gca/gfx_8_0_d.h"
+#include "bif/bif_5_0_d.h"
+#include "bif/bif_5_0_sh_mask.h"
+#include "ppatomctrl.h"
+#include "cgs_common.h"
+#include "smu7_ppsmc.h"
+
+#include "smu7_dyn_defaults.h"
+
+#include "smu7_hwmgr.h"
+#include "hardwaremanager.h"
+#include "ppatomctrl.h"
+#include "atombios.h"
+#include "pppcielanes.h"
+
+#include "dce/dce_11_2_d.h"
+#include "dce/dce_11_2_sh_mask.h"
+
+#define PPVEGAM_TARGETACTIVITY_DFLT                     50
+
+#define VOLTAGE_VID_OFFSET_SCALE1   625
+#define VOLTAGE_VID_OFFSET_SCALE2   100
+#define POWERTUNE_DEFAULT_SET_MAX    1
+#define VDDC_VDDCI_DELTA            200
+#define MC_CG_ARB_FREQ_F1           0x0b
+
+#define STRAP_ASIC_RO_LSB    2168
+#define STRAP_ASIC_RO_MSB    2175
+
+#define PPSMC_MSG_ApplyAvfsCksOffVoltage      ((uint16_t) 0x415)
+#define PPSMC_MSG_EnableModeSwitchRLCNotification  ((uint16_t) 0x305)
+
+static const struct vegam_pt_defaults
+vegam_power_tune_data_set_array[POWERTUNE_DEFAULT_SET_MAX] = {
+	/* sviLoadLIneEn, SviLoadLineVddC, TDC_VDDC_ThrottleReleaseLimitPerc, TDC_MAWt,
+	 * TdcWaterfallCtl, DTEAmbientTempBase, DisplayCac, BAPM_TEMP_GRADIENT */
+	{ 1, 0xF, 0xFD, 0x19, 5, 45, 0, 0xB0000,
+	{ 0x79, 0x253, 0x25D, 0xAE, 0x72, 0x80, 0x83, 0x86, 0x6F, 0xC8, 0xC9, 0xC9, 0x2F, 0x4D, 0x61},
+	{ 0x17C, 0x172, 0x180, 0x1BC, 0x1B3, 0x1BD, 0x206, 0x200, 0x203, 0x25D, 0x25A, 0x255, 0x2C3, 0x2C5, 0x2B4 } },
+};
+
+static const sclkFcwRange_t Range_Table[NUM_SCLK_RANGE] = {
+			{VCO_2_4, POSTDIV_DIV_BY_16,  75, 160, 112},
+			{VCO_3_6, POSTDIV_DIV_BY_16, 112, 224, 160},
+			{VCO_2_4, POSTDIV_DIV_BY_8,   75, 160, 112},
+			{VCO_3_6, POSTDIV_DIV_BY_8,  112, 224, 160},
+			{VCO_2_4, POSTDIV_DIV_BY_4,   75, 160, 112},
+			{VCO_3_6, POSTDIV_DIV_BY_4,  112, 216, 160},
+			{VCO_2_4, POSTDIV_DIV_BY_2,   75, 160, 108},
+			{VCO_3_6, POSTDIV_DIV_BY_2,  112, 216, 160} };
+
+static int vegam_smu_init(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data;
+
+	smu_data = kzalloc(sizeof(struct vegam_smumgr), GFP_KERNEL);
+	if (smu_data == NULL)
+		return -ENOMEM;
+
+	hwmgr->smu_backend = smu_data;
+
+	if (smu7_init(hwmgr)) {
+		kfree(smu_data);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+static int vegam_start_smu_in_protection_mode(struct pp_hwmgr *hwmgr)
+{
+	int result = 0;
+
+	/* Wait for smc boot up */
+	/* PHM_WAIT_VFPF_INDIRECT_FIELD_UNEQUAL(smumgr, SMC_IND, RCU_UC_EVENTS, boot_seq_done, 0) */
+
+	/* Assert reset */
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_RESET_CNTL, rst_reg, 1);
+
+	result = smu7_upload_smu_firmware_image(hwmgr);
+	if (result != 0)
+		return result;
+
+	/* Clear status */
+	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixSMU_STATUS, 0);
+
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_CLOCK_CNTL_0, ck_disable, 0);
+
+	/* De-assert reset */
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_RESET_CNTL, rst_reg, 0);
+
+
+	PHM_WAIT_VFPF_INDIRECT_FIELD(hwmgr, SMC_IND, RCU_UC_EVENTS, INTERRUPTS_ENABLED, 1);
+
+
+	/* Call Test SMU message with 0x20000 offset to trigger SMU start */
+	smu7_send_msg_to_smc_offset(hwmgr);
+
+	/* Wait done bit to be set */
+	/* Check pass/failed indicator */
+
+	PHM_WAIT_VFPF_INDIRECT_FIELD_UNEQUAL(hwmgr, SMC_IND, SMU_STATUS, SMU_DONE, 0);
+
+	if (1 != PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+						SMU_STATUS, SMU_PASS))
+		PP_ASSERT_WITH_CODE(false, "SMU Firmware start failed!", return -1);
+
+	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixFIRMWARE_FLAGS, 0);
+
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_RESET_CNTL, rst_reg, 1);
+
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_RESET_CNTL, rst_reg, 0);
+
+	/* Wait for firmware to initialize */
+	PHM_WAIT_VFPF_INDIRECT_FIELD(hwmgr, SMC_IND, FIRMWARE_FLAGS, INTERRUPTS_ENABLED, 1);
+
+	return result;
+}
+
+static int vegam_start_smu_in_non_protection_mode(struct pp_hwmgr *hwmgr)
+{
+	int result = 0;
+
+	/* wait for smc boot up */
+	PHM_WAIT_VFPF_INDIRECT_FIELD_UNEQUAL(hwmgr, SMC_IND, RCU_UC_EVENTS, boot_seq_done, 0);
+
+	/* Clear firmware interrupt enable flag */
+	/* PHM_WRITE_VFPF_INDIRECT_FIELD(pSmuMgr, SMC_IND, SMC_SYSCON_MISC_CNTL, pre_fetcher_en, 1); */
+	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+				ixFIRMWARE_FLAGS, 0);
+
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_RESET_CNTL,
+					rst_reg, 1);
+
+	result = smu7_upload_smu_firmware_image(hwmgr);
+	if (result != 0)
+		return result;
+
+	/* Set smc instruct start point at 0x0 */
+	smu7_program_jump_on_start(hwmgr);
+
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_CLOCK_CNTL_0, ck_disable, 0);
+
+	PHM_WRITE_VFPF_INDIRECT_FIELD(hwmgr->device, CGS_IND_REG__SMC,
+					SMC_SYSCON_RESET_CNTL, rst_reg, 0);
+
+	/* Wait for firmware to initialize */
+
+	PHM_WAIT_VFPF_INDIRECT_FIELD(hwmgr, SMC_IND,
+					FIRMWARE_FLAGS, INTERRUPTS_ENABLED, 1);
+
+	return result;
+}
+
+static int vegam_start_smu(struct pp_hwmgr *hwmgr)
+{
+	int result = 0;
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+
+	/* Only start SMC if SMC RAM is not running */
+	if (!smu7_is_smc_ram_running(hwmgr) && hwmgr->not_vf) {
+		smu_data->protected_mode = (uint8_t)(PHM_READ_VFPF_INDIRECT_FIELD(hwmgr->device,
+				CGS_IND_REG__SMC, SMU_FIRMWARE, SMU_MODE));
+		smu_data->smu7_data.security_hard_key = (uint8_t)(PHM_READ_VFPF_INDIRECT_FIELD(
+				hwmgr->device, CGS_IND_REG__SMC, SMU_FIRMWARE, SMU_SEL));
+
+		/* Check if SMU is running in protected mode */
+		if (smu_data->protected_mode == 0)
+			result = vegam_start_smu_in_non_protection_mode(hwmgr);
+		else
+			result = vegam_start_smu_in_protection_mode(hwmgr);
+
+		if (result != 0)
+			PP_ASSERT_WITH_CODE(0, "Failed to load SMU ucode.", return result);
+	}
+
+	/* Setup SoftRegsStart here for register lookup in case DummyBackEnd is used and ProcessFirmwareHeader is not executed */
+	smu7_read_smc_sram_dword(hwmgr,
+			SMU7_FIRMWARE_HEADER_LOCATION + offsetof(SMU75_Firmware_Header, SoftRegisters),
+			&(smu_data->smu7_data.soft_regs_start),
+			0x40000);
+
+	result = smu7_request_smu_load_fw(hwmgr);
+
+	return result;
+}
+
+static int vegam_process_firmware_header(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	uint32_t tmp;
+	int result;
+	bool error = false;
+
+	result = smu7_read_smc_sram_dword(hwmgr,
+			SMU7_FIRMWARE_HEADER_LOCATION +
+			offsetof(SMU75_Firmware_Header, DpmTable),
+			&tmp, SMC_RAM_END);
+
+	if (0 == result)
+		smu_data->smu7_data.dpm_table_start = tmp;
+
+	error |= (0 != result);
+
+	result = smu7_read_smc_sram_dword(hwmgr,
+			SMU7_FIRMWARE_HEADER_LOCATION +
+			offsetof(SMU75_Firmware_Header, SoftRegisters),
+			&tmp, SMC_RAM_END);
+
+	if (!result) {
+		data->soft_regs_start = tmp;
+		smu_data->smu7_data.soft_regs_start = tmp;
+	}
+
+	error |= (0 != result);
+
+	result = smu7_read_smc_sram_dword(hwmgr,
+			SMU7_FIRMWARE_HEADER_LOCATION +
+			offsetof(SMU75_Firmware_Header, mcRegisterTable),
+			&tmp, SMC_RAM_END);
+
+	if (!result)
+		smu_data->smu7_data.mc_reg_table_start = tmp;
+
+	result = smu7_read_smc_sram_dword(hwmgr,
+			SMU7_FIRMWARE_HEADER_LOCATION +
+			offsetof(SMU75_Firmware_Header, FanTable),
+			&tmp, SMC_RAM_END);
+
+	if (!result)
+		smu_data->smu7_data.fan_table_start = tmp;
+
+	error |= (0 != result);
+
+	result = smu7_read_smc_sram_dword(hwmgr,
+			SMU7_FIRMWARE_HEADER_LOCATION +
+			offsetof(SMU75_Firmware_Header, mcArbDramTimingTable),
+			&tmp, SMC_RAM_END);
+
+	if (!result)
+		smu_data->smu7_data.arb_table_start = tmp;
+
+	error |= (0 != result);
+
+	result = smu7_read_smc_sram_dword(hwmgr,
+			SMU7_FIRMWARE_HEADER_LOCATION +
+			offsetof(SMU75_Firmware_Header, Version),
+			&tmp, SMC_RAM_END);
+
+	if (!result)
+		hwmgr->microcode_version_info.SMC = tmp;
+
+	error |= (0 != result);
+
+	return error ? -1 : 0;
+}
+
+static bool vegam_is_dpm_running(struct pp_hwmgr *hwmgr)
+{
+	return (1 == PHM_READ_INDIRECT_FIELD(hwmgr->device,
+			CGS_IND_REG__SMC, FEATURE_STATUS, VOLTAGE_CONTROLLER_ON))
+			? true : false;
+}
+
+static uint32_t vegam_get_mac_definition(uint32_t value)
+{
+	switch (value) {
+	case SMU_MAX_LEVELS_GRAPHICS:
+		return SMU75_MAX_LEVELS_GRAPHICS;
+	case SMU_MAX_LEVELS_MEMORY:
+		return SMU75_MAX_LEVELS_MEMORY;
+	case SMU_MAX_LEVELS_LINK:
+		return SMU75_MAX_LEVELS_LINK;
+	case SMU_MAX_ENTRIES_SMIO:
+		return SMU75_MAX_ENTRIES_SMIO;
+	case SMU_MAX_LEVELS_VDDC:
+		return SMU75_MAX_LEVELS_VDDC;
+	case SMU_MAX_LEVELS_VDDGFX:
+		return SMU75_MAX_LEVELS_VDDGFX;
+	case SMU_MAX_LEVELS_VDDCI:
+		return SMU75_MAX_LEVELS_VDDCI;
+	case SMU_MAX_LEVELS_MVDD:
+		return SMU75_MAX_LEVELS_MVDD;
+	case SMU_UVD_MCLK_HANDSHAKE_DISABLE:
+		return SMU7_UVD_MCLK_HANDSHAKE_DISABLE |
+				SMU7_VCE_MCLK_HANDSHAKE_DISABLE;
+	}
+
+	pr_warn("can't get the mac of %x\n", value);
+	return 0;
+}
+
+static int vegam_update_uvd_smc_table(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	uint32_t mm_boot_level_offset, mm_boot_level_value;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+
+	smu_data->smc_state_table.UvdBootLevel = 0;
+	if (table_info->mm_dep_table->count > 0)
+		smu_data->smc_state_table.UvdBootLevel =
+				(uint8_t) (table_info->mm_dep_table->count - 1);
+	mm_boot_level_offset = smu_data->smu7_data.dpm_table_start + offsetof(SMU75_Discrete_DpmTable,
+						UvdBootLevel);
+	mm_boot_level_offset /= 4;
+	mm_boot_level_offset *= 4;
+	mm_boot_level_value = cgs_read_ind_register(hwmgr->device,
+			CGS_IND_REG__SMC, mm_boot_level_offset);
+	mm_boot_level_value &= 0x00FFFFFF;
+	mm_boot_level_value |= smu_data->smc_state_table.UvdBootLevel << 24;
+	cgs_write_ind_register(hwmgr->device,
+			CGS_IND_REG__SMC, mm_boot_level_offset, mm_boot_level_value);
+
+	if (!phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_UVDDPM) ||
+		phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_StablePState))
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+				PPSMC_MSG_UVDDPM_SetEnabledMask,
+				(uint32_t)(1 << smu_data->smc_state_table.UvdBootLevel));
+	return 0;
+}
+
+static int vegam_update_vce_smc_table(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	uint32_t mm_boot_level_offset, mm_boot_level_value;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+					PHM_PlatformCaps_StablePState))
+		smu_data->smc_state_table.VceBootLevel =
+			(uint8_t) (table_info->mm_dep_table->count - 1);
+	else
+		smu_data->smc_state_table.VceBootLevel = 0;
+
+	mm_boot_level_offset = smu_data->smu7_data.dpm_table_start +
+					offsetof(SMU75_Discrete_DpmTable, VceBootLevel);
+	mm_boot_level_offset /= 4;
+	mm_boot_level_offset *= 4;
+	mm_boot_level_value = cgs_read_ind_register(hwmgr->device,
+			CGS_IND_REG__SMC, mm_boot_level_offset);
+	mm_boot_level_value &= 0xFF00FFFF;
+	mm_boot_level_value |= smu_data->smc_state_table.VceBootLevel << 16;
+	cgs_write_ind_register(hwmgr->device,
+			CGS_IND_REG__SMC, mm_boot_level_offset, mm_boot_level_value);
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_StablePState))
+		smum_send_msg_to_smc_with_parameter(hwmgr,
+				PPSMC_MSG_VCEDPM_SetEnabledMask,
+				(uint32_t)1 << smu_data->smc_state_table.VceBootLevel);
+	return 0;
+}
+
+static int vegam_update_bif_smc_table(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_ppt_v1_pcie_table *pcie_table = table_info->pcie_table;
+	int max_entry, i;
+
+	max_entry = (SMU75_MAX_LEVELS_LINK < pcie_table->count) ?
+						SMU75_MAX_LEVELS_LINK :
+						pcie_table->count;
+	/* Setup BIF_SCLK levels */
+	for (i = 0; i < max_entry; i++)
+		smu_data->bif_sclk_table[i] = pcie_table->entries[i].pcie_sclk;
+	return 0;
+}
+
+static int vegam_update_smc_table(struct pp_hwmgr *hwmgr, uint32_t type)
+{
+	switch (type) {
+	case SMU_UVD_TABLE:
+		vegam_update_uvd_smc_table(hwmgr);
+		break;
+	case SMU_VCE_TABLE:
+		vegam_update_vce_smc_table(hwmgr);
+		break;
+	case SMU_BIF_TABLE:
+		vegam_update_bif_smc_table(hwmgr);
+		break;
+	default:
+		break;
+	}
+	return 0;
+}
+
+static void vegam_initialize_power_tune_defaults(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct  phm_ppt_v1_information *table_info =
+			(struct  phm_ppt_v1_information *)(hwmgr->pptable);
+
+	if (table_info &&
+			table_info->cac_dtp_table->usPowerTuneDataSetID <= POWERTUNE_DEFAULT_SET_MAX &&
+			table_info->cac_dtp_table->usPowerTuneDataSetID)
+		smu_data->power_tune_defaults =
+				&vegam_power_tune_data_set_array
+				[table_info->cac_dtp_table->usPowerTuneDataSetID - 1];
+	else
+		smu_data->power_tune_defaults = &vegam_power_tune_data_set_array[0];
+
+}
+
+static int vegam_populate_smc_mvdd_table(struct pp_hwmgr *hwmgr,
+			SMU75_Discrete_DpmTable *table)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	uint32_t count, level;
+
+	if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->mvdd_control) {
+		count = data->mvdd_voltage_table.count;
+		if (count > SMU_MAX_SMIO_LEVELS)
+			count = SMU_MAX_SMIO_LEVELS;
+		for (level = 0; level < count; level++) {
+			table->SmioTable2.Pattern[level].Voltage = PP_HOST_TO_SMC_US(
+					data->mvdd_voltage_table.entries[count].value * VOLTAGE_SCALE);
+			/* Index into DpmTable.Smio. Drive bits from Smio entry to get this voltage level.*/
+			table->SmioTable2.Pattern[level].Smio =
+				(uint8_t) level;
+			table->Smio[level] |=
+				data->mvdd_voltage_table.entries[level].smio_low;
+		}
+		table->SmioMask2 = data->mvdd_voltage_table.mask_low;
+
+		table->MvddLevelCount = (uint32_t) PP_HOST_TO_SMC_UL(count);
+	}
+
+	return 0;
+}
+
+static int vegam_populate_smc_vddci_table(struct pp_hwmgr *hwmgr,
+					struct SMU75_Discrete_DpmTable *table)
+{
+	uint32_t count, level;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+
+	count = data->vddci_voltage_table.count;
+
+	if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control) {
+		if (count > SMU_MAX_SMIO_LEVELS)
+			count = SMU_MAX_SMIO_LEVELS;
+		for (level = 0; level < count; ++level) {
+			table->SmioTable1.Pattern[level].Voltage = PP_HOST_TO_SMC_US(
+					data->vddci_voltage_table.entries[level].value * VOLTAGE_SCALE);
+			table->SmioTable1.Pattern[level].Smio = (uint8_t) level;
+
+			table->Smio[level] |= data->vddci_voltage_table.entries[level].smio_low;
+		}
+	}
+
+	table->SmioMask1 = data->vddci_voltage_table.mask_low;
+
+	return 0;
+}
+
+static int vegam_populate_cac_table(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_DpmTable *table)
+{
+	uint32_t count;
+	uint8_t index;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_ppt_v1_voltage_lookup_table *lookup_table =
+			table_info->vddc_lookup_table;
+	/* tables is already swapped, so in order to use the value from it,
+	 * we need to swap it back.
+	 * We are populating vddc CAC data to BapmVddc table
+	 * in split and merged mode
+	 */
+	for (count = 0; count < lookup_table->count; count++) {
+		index = phm_get_voltage_index(lookup_table,
+				data->vddc_voltage_table.entries[count].value);
+		table->BapmVddcVidLoSidd[count] =
+				convert_to_vid(lookup_table->entries[index].us_cac_low);
+		table->BapmVddcVidHiSidd[count] =
+				convert_to_vid(lookup_table->entries[index].us_cac_mid);
+		table->BapmVddcVidHiSidd2[count] =
+				convert_to_vid(lookup_table->entries[index].us_cac_high);
+	}
+
+	return 0;
+}
+
+static int vegam_populate_smc_voltage_tables(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_DpmTable *table)
+{
+	vegam_populate_smc_vddci_table(hwmgr, table);
+	vegam_populate_smc_mvdd_table(hwmgr, table);
+	vegam_populate_cac_table(hwmgr, table);
+
+	return 0;
+}
+
+static int vegam_populate_ulv_level(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_Ulv *state)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+
+	state->CcPwrDynRm = 0;
+	state->CcPwrDynRm1 = 0;
+
+	state->VddcOffset = (uint16_t) table_info->us_ulv_voltage_offset;
+	state->VddcOffsetVid = (uint8_t)(table_info->us_ulv_voltage_offset *
+			VOLTAGE_VID_OFFSET_SCALE2 / VOLTAGE_VID_OFFSET_SCALE1);
+
+	state->VddcPhase = data->vddc_phase_shed_control ^ 0x3;
+
+	CONVERT_FROM_HOST_TO_SMC_UL(state->CcPwrDynRm);
+	CONVERT_FROM_HOST_TO_SMC_UL(state->CcPwrDynRm1);
+	CONVERT_FROM_HOST_TO_SMC_US(state->VddcOffset);
+
+	return 0;
+}
+
+static int vegam_populate_ulv_state(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_DpmTable *table)
+{
+	return vegam_populate_ulv_level(hwmgr, &table->Ulv);
+}
+
+static int vegam_populate_smc_link_level(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_DpmTable *table)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data =
+			(struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct smu7_dpm_table *dpm_table = &data->dpm_table;
+	int i;
+
+	/* Index (dpm_table->pcie_speed_table.count)
+	 * is reserved for PCIE boot level. */
+	for (i = 0; i <= dpm_table->pcie_speed_table.count; i++) {
+		table->LinkLevel[i].PcieGenSpeed  =
+				(uint8_t)dpm_table->pcie_speed_table.dpm_levels[i].value;
+		table->LinkLevel[i].PcieLaneCount = (uint8_t)encode_pcie_lane_width(
+				dpm_table->pcie_speed_table.dpm_levels[i].param1);
+		table->LinkLevel[i].EnabledForActivity = 1;
+		table->LinkLevel[i].SPC = (uint8_t)(data->pcie_spc_cap & 0xff);
+		table->LinkLevel[i].DownThreshold = PP_HOST_TO_SMC_UL(5);
+		table->LinkLevel[i].UpThreshold = PP_HOST_TO_SMC_UL(30);
+	}
+
+	smu_data->smc_state_table.LinkLevelCount =
+			(uint8_t)dpm_table->pcie_speed_table.count;
+
+/* To Do move to hwmgr */
+	data->dpm_level_enable_mask.pcie_dpm_enable_mask =
+			phm_get_dpm_level_enable_mask_value(&dpm_table->pcie_speed_table);
+
+	return 0;
+}
+
+static int vegam_get_dependency_volt_by_clk(struct pp_hwmgr *hwmgr,
+		struct phm_ppt_v1_clock_voltage_dependency_table *dep_table,
+		uint32_t clock, SMU_VoltageLevel *voltage, uint32_t *mvdd)
+{
+	uint32_t i;
+	uint16_t vddci;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+
+	*voltage = *mvdd = 0;
+
+	/* clock - voltage dependency table is empty table */
+	if (dep_table->count == 0)
+		return -EINVAL;
+
+	for (i = 0; i < dep_table->count; i++) {
+		/* find first sclk bigger than request */
+		if (dep_table->entries[i].clk >= clock) {
+			*voltage |= (dep_table->entries[i].vddc *
+					VOLTAGE_SCALE) << VDDC_SHIFT;
+			if (SMU7_VOLTAGE_CONTROL_NONE == data->vddci_control)
+				*voltage |= (data->vbios_boot_state.vddci_bootup_value *
+						VOLTAGE_SCALE) << VDDCI_SHIFT;
+			else if (dep_table->entries[i].vddci)
+				*voltage |= (dep_table->entries[i].vddci *
+						VOLTAGE_SCALE) << VDDCI_SHIFT;
+			else {
+				vddci = phm_find_closest_vddci(&(data->vddci_voltage_table),
+						(dep_table->entries[i].vddc -
+								(uint16_t)VDDC_VDDCI_DELTA));
+				*voltage |= (vddci * VOLTAGE_SCALE) << VDDCI_SHIFT;
+			}
+
+			if (SMU7_VOLTAGE_CONTROL_NONE == data->mvdd_control)
+				*mvdd = data->vbios_boot_state.mvdd_bootup_value *
+					VOLTAGE_SCALE;
+			else if (dep_table->entries[i].mvdd)
+				*mvdd = (uint32_t) dep_table->entries[i].mvdd *
+					VOLTAGE_SCALE;
+
+			*voltage |= 1 << PHASES_SHIFT;
+			return 0;
+		}
+	}
+
+	/* sclk is bigger than max sclk in the dependence table */
+	*voltage |= (dep_table->entries[i - 1].vddc * VOLTAGE_SCALE) << VDDC_SHIFT;
+	vddci = phm_find_closest_vddci(&(data->vddci_voltage_table),
+			(dep_table->entries[i - 1].vddc -
+					(uint16_t)VDDC_VDDCI_DELTA));
+
+	if (SMU7_VOLTAGE_CONTROL_NONE == data->vddci_control)
+		*voltage |= (data->vbios_boot_state.vddci_bootup_value *
+				VOLTAGE_SCALE) << VDDCI_SHIFT;
+	else if (dep_table->entries[i - 1].vddci)
+		*voltage |= (dep_table->entries[i - 1].vddci *
+				VOLTAGE_SCALE) << VDDC_SHIFT;
+	else
+		*voltage |= (vddci * VOLTAGE_SCALE) << VDDCI_SHIFT;
+
+	if (SMU7_VOLTAGE_CONTROL_NONE == data->mvdd_control)
+		*mvdd = data->vbios_boot_state.mvdd_bootup_value * VOLTAGE_SCALE;
+	else if (dep_table->entries[i].mvdd)
+		*mvdd = (uint32_t) dep_table->entries[i - 1].mvdd * VOLTAGE_SCALE;
+
+	return 0;
+}
+
+static void vegam_get_sclk_range_table(struct pp_hwmgr *hwmgr,
+				   SMU75_Discrete_DpmTable  *table)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	uint32_t i, ref_clk;
+
+	struct pp_atom_ctrl_sclk_range_table range_table_from_vbios = { { {0} } };
+
+	ref_clk = amdgpu_asic_get_xclk((struct amdgpu_device *)hwmgr->adev);
+
+	if (0 == atomctrl_get_smc_sclk_range_table(hwmgr, &range_table_from_vbios)) {
+		for (i = 0; i < NUM_SCLK_RANGE; i++) {
+			table->SclkFcwRangeTable[i].vco_setting =
+					range_table_from_vbios.entry[i].ucVco_setting;
+			table->SclkFcwRangeTable[i].postdiv =
+					range_table_from_vbios.entry[i].ucPostdiv;
+			table->SclkFcwRangeTable[i].fcw_pcc =
+					range_table_from_vbios.entry[i].usFcw_pcc;
+
+			table->SclkFcwRangeTable[i].fcw_trans_upper =
+					range_table_from_vbios.entry[i].usFcw_trans_upper;
+			table->SclkFcwRangeTable[i].fcw_trans_lower =
+					range_table_from_vbios.entry[i].usRcw_trans_lower;
+
+			CONVERT_FROM_HOST_TO_SMC_US(table->SclkFcwRangeTable[i].fcw_pcc);
+			CONVERT_FROM_HOST_TO_SMC_US(table->SclkFcwRangeTable[i].fcw_trans_upper);
+			CONVERT_FROM_HOST_TO_SMC_US(table->SclkFcwRangeTable[i].fcw_trans_lower);
+		}
+		return;
+	}
+
+	for (i = 0; i < NUM_SCLK_RANGE; i++) {
+		smu_data->range_table[i].trans_lower_frequency =
+				(ref_clk * Range_Table[i].fcw_trans_lower) >> Range_Table[i].postdiv;
+		smu_data->range_table[i].trans_upper_frequency =
+				(ref_clk * Range_Table[i].fcw_trans_upper) >> Range_Table[i].postdiv;
+
+		table->SclkFcwRangeTable[i].vco_setting = Range_Table[i].vco_setting;
+		table->SclkFcwRangeTable[i].postdiv = Range_Table[i].postdiv;
+		table->SclkFcwRangeTable[i].fcw_pcc = Range_Table[i].fcw_pcc;
+
+		table->SclkFcwRangeTable[i].fcw_trans_upper = Range_Table[i].fcw_trans_upper;
+		table->SclkFcwRangeTable[i].fcw_trans_lower = Range_Table[i].fcw_trans_lower;
+
+		CONVERT_FROM_HOST_TO_SMC_US(table->SclkFcwRangeTable[i].fcw_pcc);
+		CONVERT_FROM_HOST_TO_SMC_US(table->SclkFcwRangeTable[i].fcw_trans_upper);
+		CONVERT_FROM_HOST_TO_SMC_US(table->SclkFcwRangeTable[i].fcw_trans_lower);
+	}
+}
+
+static int vegam_calculate_sclk_params(struct pp_hwmgr *hwmgr,
+		uint32_t clock, SMU_SclkSetting *sclk_setting)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	const SMU75_Discrete_DpmTable *table = &(smu_data->smc_state_table);
+	struct pp_atomctrl_clock_dividers_ai dividers;
+	uint32_t ref_clock;
+	uint32_t pcc_target_percent, pcc_target_freq, ss_target_percent, ss_target_freq;
+	uint8_t i;
+	int result;
+	uint64_t temp;
+
+	sclk_setting->SclkFrequency = clock;
+	/* get the engine clock dividers for this clock value */
+	result = atomctrl_get_engine_pll_dividers_ai(hwmgr, clock,  &dividers);
+	if (result == 0) {
+		sclk_setting->Fcw_int = dividers.usSclk_fcw_int;
+		sclk_setting->Fcw_frac = dividers.usSclk_fcw_frac;
+		sclk_setting->Pcc_fcw_int = dividers.usPcc_fcw_int;
+		sclk_setting->PllRange = dividers.ucSclkPllRange;
+		sclk_setting->Sclk_slew_rate = 0x400;
+		sclk_setting->Pcc_up_slew_rate = dividers.usPcc_fcw_slew_frac;
+		sclk_setting->Pcc_down_slew_rate = 0xffff;
+		sclk_setting->SSc_En = dividers.ucSscEnable;
+		sclk_setting->Fcw1_int = dividers.usSsc_fcw1_int;
+		sclk_setting->Fcw1_frac = dividers.usSsc_fcw1_frac;
+		sclk_setting->Sclk_ss_slew_rate = dividers.usSsc_fcw_slew_frac;
+		return result;
+	}
+
+	ref_clock = amdgpu_asic_get_xclk((struct amdgpu_device *)hwmgr->adev);
+
+	for (i = 0; i < NUM_SCLK_RANGE; i++) {
+		if (clock > smu_data->range_table[i].trans_lower_frequency
+		&& clock <= smu_data->range_table[i].trans_upper_frequency) {
+			sclk_setting->PllRange = i;
+			break;
+		}
+	}
+
+	sclk_setting->Fcw_int = (uint16_t)
+			((clock << table->SclkFcwRangeTable[sclk_setting->PllRange].postdiv) /
+					ref_clock);
+	temp = clock << table->SclkFcwRangeTable[sclk_setting->PllRange].postdiv;
+	temp <<= 0x10;
+	do_div(temp, ref_clock);
+	sclk_setting->Fcw_frac = temp & 0xffff;
+
+	pcc_target_percent = 10; /*  Hardcode 10% for now. */
+	pcc_target_freq = clock - (clock * pcc_target_percent / 100);
+	sclk_setting->Pcc_fcw_int = (uint16_t)
+			((pcc_target_freq << table->SclkFcwRangeTable[sclk_setting->PllRange].postdiv) /
+					ref_clock);
+
+	ss_target_percent = 2; /*  Hardcode 2% for now. */
+	sclk_setting->SSc_En = 0;
+	if (ss_target_percent) {
+		sclk_setting->SSc_En = 1;
+		ss_target_freq = clock - (clock * ss_target_percent / 100);
+		sclk_setting->Fcw1_int = (uint16_t)
+				((ss_target_freq << table->SclkFcwRangeTable[sclk_setting->PllRange].postdiv) /
+						ref_clock);
+		temp = ss_target_freq << table->SclkFcwRangeTable[sclk_setting->PllRange].postdiv;
+		temp <<= 0x10;
+		do_div(temp, ref_clock);
+		sclk_setting->Fcw1_frac = temp & 0xffff;
+	}
+
+	return 0;
+}
+
+static uint8_t vegam_get_sleep_divider_id_from_clock(uint32_t clock,
+		uint32_t clock_insr)
+{
+	uint8_t i;
+	uint32_t temp;
+	uint32_t min = max(clock_insr, (uint32_t)SMU7_MINIMUM_ENGINE_CLOCK);
+
+	PP_ASSERT_WITH_CODE((clock >= min),
+			"Engine clock can't satisfy stutter requirement!",
+			return 0);
+	for (i = 31;  ; i--) {
+		temp = clock / (i + 1);
+
+		if (temp >= min || i == 0)
+			break;
+	}
+	return i;
+}
+
+static int vegam_populate_single_graphic_level(struct pp_hwmgr *hwmgr,
+		uint32_t clock, struct SMU75_Discrete_GraphicsLevel *level)
+{
+	int result;
+	/* PP_Clocks minClocks; */
+	uint32_t mvdd;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	SMU_SclkSetting curr_sclk_setting = { 0 };
+
+	result = vegam_calculate_sclk_params(hwmgr, clock, &curr_sclk_setting);
+
+	/* populate graphics levels */
+	result = vegam_get_dependency_volt_by_clk(hwmgr,
+			table_info->vdd_dep_on_sclk, clock,
+			&level->MinVoltage, &mvdd);
+
+	PP_ASSERT_WITH_CODE((0 == result),
+			"can not find VDDC voltage value for "
+			"VDDC engine clock dependency table",
+			return result);
+	level->ActivityLevel = (uint16_t)(SclkDPMTuning_VEGAM >> DPMTuning_Activity_Shift);
+
+	level->CcPwrDynRm = 0;
+	level->CcPwrDynRm1 = 0;
+	level->EnabledForActivity = 0;
+	level->EnabledForThrottle = 1;
+	level->VoltageDownHyst = 0;
+	level->PowerThrottle = 0;
+	data->display_timing.min_clock_in_sr = hwmgr->display_config->min_core_set_clock_in_sr;
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_SclkDeepSleep))
+		level->DeepSleepDivId = vegam_get_sleep_divider_id_from_clock(clock,
+								hwmgr->display_config->min_core_set_clock_in_sr);
+
+	level->SclkSetting = curr_sclk_setting;
+
+	CONVERT_FROM_HOST_TO_SMC_UL(level->MinVoltage);
+	CONVERT_FROM_HOST_TO_SMC_UL(level->CcPwrDynRm);
+	CONVERT_FROM_HOST_TO_SMC_UL(level->CcPwrDynRm1);
+	CONVERT_FROM_HOST_TO_SMC_US(level->ActivityLevel);
+	CONVERT_FROM_HOST_TO_SMC_UL(level->SclkSetting.SclkFrequency);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Fcw_int);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Fcw_frac);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Pcc_fcw_int);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Sclk_slew_rate);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Pcc_up_slew_rate);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Pcc_down_slew_rate);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Fcw1_int);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Fcw1_frac);
+	CONVERT_FROM_HOST_TO_SMC_US(level->SclkSetting.Sclk_ss_slew_rate);
+	return 0;
+}
+
+static int vegam_populate_all_graphic_levels(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *hw_data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct smu7_dpm_table *dpm_table = &hw_data->dpm_table;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_ppt_v1_pcie_table *pcie_table = table_info->pcie_table;
+	uint8_t pcie_entry_cnt = (uint8_t) hw_data->dpm_table.pcie_speed_table.count;
+	int result = 0;
+	uint32_t array = smu_data->smu7_data.dpm_table_start +
+			offsetof(SMU75_Discrete_DpmTable, GraphicsLevel);
+	uint32_t array_size = sizeof(struct SMU75_Discrete_GraphicsLevel) *
+			SMU75_MAX_LEVELS_GRAPHICS;
+	struct SMU75_Discrete_GraphicsLevel *levels =
+			smu_data->smc_state_table.GraphicsLevel;
+	uint32_t i, max_entry;
+	uint8_t hightest_pcie_level_enabled = 0,
+		lowest_pcie_level_enabled = 0,
+		mid_pcie_level_enabled = 0,
+		count = 0;
+
+	vegam_get_sclk_range_table(hwmgr, &(smu_data->smc_state_table));
+
+	for (i = 0; i < dpm_table->sclk_table.count; i++) {
+
+		result = vegam_populate_single_graphic_level(hwmgr,
+				dpm_table->sclk_table.dpm_levels[i].value,
+				&(smu_data->smc_state_table.GraphicsLevel[i]));
+		if (result)
+			return result;
+
+		levels[i].UpHyst = (uint8_t)
+				(SclkDPMTuning_VEGAM >> DPMTuning_Uphyst_Shift);
+		levels[i].DownHyst = (uint8_t)
+				(SclkDPMTuning_VEGAM >> DPMTuning_Downhyst_Shift);
+		/* Making sure only DPM level 0-1 have Deep Sleep Div ID populated. */
+		if (i > 1)
+			levels[i].DeepSleepDivId = 0;
+	}
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+					PHM_PlatformCaps_SPLLShutdownSupport))
+		smu_data->smc_state_table.GraphicsLevel[0].SclkSetting.SSc_En = 0;
+
+	smu_data->smc_state_table.GraphicsDpmLevelCount =
+			(uint8_t)dpm_table->sclk_table.count;
+	hw_data->dpm_level_enable_mask.sclk_dpm_enable_mask =
+			phm_get_dpm_level_enable_mask_value(&dpm_table->sclk_table);
+
+	for (i = 0; i < dpm_table->sclk_table.count; i++)
+		levels[i].EnabledForActivity =
+				(hw_data->dpm_level_enable_mask.sclk_dpm_enable_mask >> i) & 0x1;
+
+	if (pcie_table != NULL) {
+		PP_ASSERT_WITH_CODE((1 <= pcie_entry_cnt),
+				"There must be 1 or more PCIE levels defined in PPTable.",
+				return -EINVAL);
+		max_entry = pcie_entry_cnt - 1;
+		for (i = 0; i < dpm_table->sclk_table.count; i++)
+			levels[i].pcieDpmLevel =
+					(uint8_t) ((i < max_entry) ? i : max_entry);
+	} else {
+		while (hw_data->dpm_level_enable_mask.pcie_dpm_enable_mask &&
+				((hw_data->dpm_level_enable_mask.pcie_dpm_enable_mask &
+						(1 << (hightest_pcie_level_enabled + 1))) != 0))
+			hightest_pcie_level_enabled++;
+
+		while (hw_data->dpm_level_enable_mask.pcie_dpm_enable_mask &&
+				((hw_data->dpm_level_enable_mask.pcie_dpm_enable_mask &
+						(1 << lowest_pcie_level_enabled)) == 0))
+			lowest_pcie_level_enabled++;
+
+		while ((count < hightest_pcie_level_enabled) &&
+				((hw_data->dpm_level_enable_mask.pcie_dpm_enable_mask &
+						(1 << (lowest_pcie_level_enabled + 1 + count))) == 0))
+			count++;
+
+		mid_pcie_level_enabled = (lowest_pcie_level_enabled + 1 + count) <
+				hightest_pcie_level_enabled ?
+						(lowest_pcie_level_enabled + 1 + count) :
+						hightest_pcie_level_enabled;
+
+		/* set pcieDpmLevel to hightest_pcie_level_enabled */
+		for (i = 2; i < dpm_table->sclk_table.count; i++)
+			levels[i].pcieDpmLevel = hightest_pcie_level_enabled;
+
+		/* set pcieDpmLevel to lowest_pcie_level_enabled */
+		levels[0].pcieDpmLevel = lowest_pcie_level_enabled;
+
+		/* set pcieDpmLevel to mid_pcie_level_enabled */
+		levels[1].pcieDpmLevel = mid_pcie_level_enabled;
+	}
+	/* level count will send to smc once at init smc table and never change */
+	result = smu7_copy_bytes_to_smc(hwmgr, array, (uint8_t *)levels,
+			(uint32_t)array_size, SMC_RAM_END);
+
+	return result;
+}
+
+static int vegam_calculate_mclk_params(struct pp_hwmgr *hwmgr,
+		uint32_t clock, struct SMU75_Discrete_MemoryLevel *mem_level)
+{
+	struct pp_atomctrl_memory_clock_param_ai mpll_param;
+
+	PP_ASSERT_WITH_CODE(!atomctrl_get_memory_pll_dividers_ai(hwmgr,
+			clock, &mpll_param),
+			"Failed to retrieve memory pll parameter.",
+			return -EINVAL);
+
+	mem_level->MclkFrequency = (uint32_t)mpll_param.ulClock;
+	mem_level->Fcw_int = (uint16_t)mpll_param.ulMclk_fcw_int;
+	mem_level->Fcw_frac = (uint16_t)mpll_param.ulMclk_fcw_frac;
+	mem_level->Postdiv = (uint8_t)mpll_param.ulPostDiv;
+
+	return 0;
+}
+
+static int vegam_populate_single_memory_level(struct pp_hwmgr *hwmgr,
+		uint32_t clock, struct SMU75_Discrete_MemoryLevel *mem_level)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	int result = 0;
+	uint32_t mclk_stutter_mode_threshold = 60000;
+
+
+	if (table_info->vdd_dep_on_mclk) {
+		result = vegam_get_dependency_volt_by_clk(hwmgr,
+				table_info->vdd_dep_on_mclk, clock,
+				&mem_level->MinVoltage, &mem_level->MinMvdd);
+		PP_ASSERT_WITH_CODE(!result,
+				"can not find MinVddc voltage value from memory "
+				"VDDC voltage dependency table", return result);
+	}
+
+	result = vegam_calculate_mclk_params(hwmgr, clock, mem_level);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to calculate mclk params.",
+			return -EINVAL);
+
+	mem_level->EnabledForThrottle = 1;
+	mem_level->EnabledForActivity = 0;
+	mem_level->VoltageDownHyst = 0;
+	mem_level->ActivityLevel = (uint16_t)
+			(MemoryDPMTuning_VEGAM >> DPMTuning_Activity_Shift);
+	mem_level->StutterEnable = false;
+	mem_level->DisplayWatermark = PPSMC_DISPLAY_WATERMARK_LOW;
+
+	data->display_timing.num_existing_displays = hwmgr->display_config->num_display;
+
+	if (mclk_stutter_mode_threshold &&
+		(clock <= mclk_stutter_mode_threshold) &&
+		(PHM_READ_FIELD(hwmgr->device, DPG_PIPE_STUTTER_CONTROL,
+				STUTTER_ENABLE) & 0x1))
+		mem_level->StutterEnable = true;
+
+	if (!result) {
+		CONVERT_FROM_HOST_TO_SMC_UL(mem_level->MinMvdd);
+		CONVERT_FROM_HOST_TO_SMC_UL(mem_level->MclkFrequency);
+		CONVERT_FROM_HOST_TO_SMC_US(mem_level->Fcw_int);
+		CONVERT_FROM_HOST_TO_SMC_US(mem_level->Fcw_frac);
+		CONVERT_FROM_HOST_TO_SMC_US(mem_level->ActivityLevel);
+		CONVERT_FROM_HOST_TO_SMC_UL(mem_level->MinVoltage);
+	}
+
+	return result;
+}
+
+static int vegam_populate_all_memory_levels(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *hw_data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct smu7_dpm_table *dpm_table = &hw_data->dpm_table;
+	int result;
+	/* populate MCLK dpm table to SMU7 */
+	uint32_t array = smu_data->smu7_data.dpm_table_start +
+			offsetof(SMU75_Discrete_DpmTable, MemoryLevel);
+	uint32_t array_size = sizeof(SMU75_Discrete_MemoryLevel) *
+			SMU75_MAX_LEVELS_MEMORY;
+	struct SMU75_Discrete_MemoryLevel *levels =
+			smu_data->smc_state_table.MemoryLevel;
+	uint32_t i;
+
+	for (i = 0; i < dpm_table->mclk_table.count; i++) {
+		PP_ASSERT_WITH_CODE((0 != dpm_table->mclk_table.dpm_levels[i].value),
+				"can not populate memory level as memory clock is zero",
+				return -EINVAL);
+		result = vegam_populate_single_memory_level(hwmgr,
+				dpm_table->mclk_table.dpm_levels[i].value,
+				&levels[i]);
+
+		if (result)
+			return result;
+
+		levels[i].UpHyst = (uint8_t)
+				(MemoryDPMTuning_VEGAM >> DPMTuning_Uphyst_Shift);
+		levels[i].DownHyst = (uint8_t)
+				(MemoryDPMTuning_VEGAM >> DPMTuning_Downhyst_Shift);
+	}
+
+	smu_data->smc_state_table.MemoryDpmLevelCount =
+			(uint8_t)dpm_table->mclk_table.count;
+	hw_data->dpm_level_enable_mask.mclk_dpm_enable_mask =
+			phm_get_dpm_level_enable_mask_value(&dpm_table->mclk_table);
+
+	for (i = 0; i < dpm_table->mclk_table.count; i++)
+		levels[i].EnabledForActivity =
+				(hw_data->dpm_level_enable_mask.mclk_dpm_enable_mask >> i) & 0x1;
+
+	levels[dpm_table->mclk_table.count - 1].DisplayWatermark =
+			PPSMC_DISPLAY_WATERMARK_HIGH;
+
+	/* level count will send to smc once at init smc table and never change */
+	result = smu7_copy_bytes_to_smc(hwmgr, array, (uint8_t *)levels,
+			(uint32_t)array_size, SMC_RAM_END);
+
+	return result;
+}
+
+static int vegam_populate_mvdd_value(struct pp_hwmgr *hwmgr,
+		uint32_t mclk, SMIO_Pattern *smio_pat)
+{
+	const struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	uint32_t i = 0;
+
+	if (SMU7_VOLTAGE_CONTROL_NONE != data->mvdd_control) {
+		/* find mvdd value which clock is more than request */
+		for (i = 0; i < table_info->vdd_dep_on_mclk->count; i++) {
+			if (mclk <= table_info->vdd_dep_on_mclk->entries[i].clk) {
+				smio_pat->Voltage = data->mvdd_voltage_table.entries[i].value;
+				break;
+			}
+		}
+		PP_ASSERT_WITH_CODE(i < table_info->vdd_dep_on_mclk->count,
+				"MVDD Voltage is outside the supported range.",
+				return -EINVAL);
+	} else
+		return -EINVAL;
+
+	return 0;
+}
+
+static int vegam_populate_smc_acpi_level(struct pp_hwmgr *hwmgr,
+		SMU75_Discrete_DpmTable *table)
+{
+	int result = 0;
+	uint32_t sclk_frequency;
+	const struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	SMIO_Pattern vol_level;
+	uint32_t mvdd;
+	uint16_t us_mvdd;
+
+	table->ACPILevel.Flags &= ~PPSMC_SWSTATE_FLAG_DC;
+
+	/* Get MinVoltage and Frequency from DPM0,
+	 * already converted to SMC_UL */
+	sclk_frequency = data->vbios_boot_state.sclk_bootup_value;
+	result = vegam_get_dependency_volt_by_clk(hwmgr,
+			table_info->vdd_dep_on_sclk,
+			sclk_frequency,
+			&table->ACPILevel.MinVoltage, &mvdd);
+	PP_ASSERT_WITH_CODE(!result,
+			"Cannot find ACPI VDDC voltage value "
+			"in Clock Dependency Table",
+			);
+
+	result = vegam_calculate_sclk_params(hwmgr, sclk_frequency,
+			&(table->ACPILevel.SclkSetting));
+	PP_ASSERT_WITH_CODE(!result,
+			"Error retrieving Engine Clock dividers from VBIOS.",
+			return result);
+
+	table->ACPILevel.DeepSleepDivId = 0;
+	table->ACPILevel.CcPwrDynRm = 0;
+	table->ACPILevel.CcPwrDynRm1 = 0;
+
+	CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.Flags);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.MinVoltage);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.CcPwrDynRm);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.CcPwrDynRm1);
+
+	CONVERT_FROM_HOST_TO_SMC_UL(table->ACPILevel.SclkSetting.SclkFrequency);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Fcw_int);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Fcw_frac);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Pcc_fcw_int);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Sclk_slew_rate);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Pcc_up_slew_rate);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Pcc_down_slew_rate);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Fcw1_int);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Fcw1_frac);
+	CONVERT_FROM_HOST_TO_SMC_US(table->ACPILevel.SclkSetting.Sclk_ss_slew_rate);
+
+
+	/* Get MinVoltage and Frequency from DPM0, already converted to SMC_UL */
+	table->MemoryACPILevel.MclkFrequency = data->vbios_boot_state.mclk_bootup_value;
+	result = vegam_get_dependency_volt_by_clk(hwmgr,
+			table_info->vdd_dep_on_mclk,
+			table->MemoryACPILevel.MclkFrequency,
+			&table->MemoryACPILevel.MinVoltage, &mvdd);
+	PP_ASSERT_WITH_CODE((0 == result),
+			"Cannot find ACPI VDDCI voltage value "
+			"in Clock Dependency Table",
+			);
+
+	us_mvdd = 0;
+	if ((SMU7_VOLTAGE_CONTROL_NONE == data->mvdd_control) ||
+			(data->mclk_dpm_key_disabled))
+		us_mvdd = data->vbios_boot_state.mvdd_bootup_value;
+	else {
+		if (!vegam_populate_mvdd_value(hwmgr,
+				data->dpm_table.mclk_table.dpm_levels[0].value,
+				&vol_level))
+			us_mvdd = vol_level.Voltage;
+	}
+
+	if (!vegam_populate_mvdd_value(hwmgr, 0, &vol_level))
+		table->MemoryACPILevel.MinMvdd = PP_HOST_TO_SMC_UL(vol_level.Voltage);
+	else
+		table->MemoryACPILevel.MinMvdd = 0;
+
+	table->MemoryACPILevel.StutterEnable = false;
+
+	table->MemoryACPILevel.EnabledForThrottle = 0;
+	table->MemoryACPILevel.EnabledForActivity = 0;
+	table->MemoryACPILevel.UpHyst = 0;
+	table->MemoryACPILevel.DownHyst = 100;
+	table->MemoryACPILevel.VoltageDownHyst = 0;
+	table->MemoryACPILevel.ActivityLevel =
+		PP_HOST_TO_SMC_US(data->current_profile_setting.mclk_activity);
+
+	CONVERT_FROM_HOST_TO_SMC_UL(table->MemoryACPILevel.MclkFrequency);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->MemoryACPILevel.MinVoltage);
+
+	return result;
+}
+
+static int vegam_populate_smc_vce_level(struct pp_hwmgr *hwmgr,
+		SMU75_Discrete_DpmTable *table)
+{
+	int result = -EINVAL;
+	uint8_t count;
+	struct pp_atomctrl_clock_dividers_vi dividers;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table =
+			table_info->mm_dep_table;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	uint32_t vddci;
+
+	table->VceLevelCount = (uint8_t)(mm_table->count);
+	table->VceBootLevel = 0;
+
+	for (count = 0; count < table->VceLevelCount; count++) {
+		table->VceLevel[count].Frequency = mm_table->entries[count].eclk;
+		table->VceLevel[count].MinVoltage = 0;
+		table->VceLevel[count].MinVoltage |=
+				(mm_table->entries[count].vddc * VOLTAGE_SCALE) << VDDC_SHIFT;
+
+		if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control)
+			vddci = (uint32_t)phm_find_closest_vddci(&(data->vddci_voltage_table),
+						mm_table->entries[count].vddc - VDDC_VDDCI_DELTA);
+		else if (SMU7_VOLTAGE_CONTROL_BY_SVID2 == data->vddci_control)
+			vddci = mm_table->entries[count].vddc - VDDC_VDDCI_DELTA;
+		else
+			vddci = (data->vbios_boot_state.vddci_bootup_value * VOLTAGE_SCALE) << VDDCI_SHIFT;
+
+
+		table->VceLevel[count].MinVoltage |=
+				(vddci * VOLTAGE_SCALE) << VDDCI_SHIFT;
+		table->VceLevel[count].MinVoltage |= 1 << PHASES_SHIFT;
+
+		/*retrieve divider value for VBIOS */
+		result = atomctrl_get_dfs_pll_dividers_vi(hwmgr,
+				table->VceLevel[count].Frequency, &dividers);
+		PP_ASSERT_WITH_CODE((0 == result),
+				"can not find divide id for VCE engine clock",
+				return result);
+
+		table->VceLevel[count].Divider = (uint8_t)dividers.pll_post_divider;
+
+		CONVERT_FROM_HOST_TO_SMC_UL(table->VceLevel[count].Frequency);
+		CONVERT_FROM_HOST_TO_SMC_UL(table->VceLevel[count].MinVoltage);
+	}
+	return result;
+}
+
+static int vegam_populate_memory_timing_parameters(struct pp_hwmgr *hwmgr,
+		int32_t eng_clock, int32_t mem_clock,
+		SMU75_Discrete_MCArbDramTimingTableEntry *arb_regs)
+{
+	uint32_t dram_timing;
+	uint32_t dram_timing2;
+	uint32_t burst_time;
+	uint32_t rfsh_rate;
+	uint32_t misc3;
+
+	int result;
+
+	result = atomctrl_set_engine_dram_timings_rv770(hwmgr,
+			eng_clock, mem_clock);
+	PP_ASSERT_WITH_CODE(result == 0,
+			"Error calling VBIOS to set DRAM_TIMING.",
+			return result);
+
+	dram_timing = cgs_read_register(hwmgr->device, mmMC_ARB_DRAM_TIMING);
+	dram_timing2 = cgs_read_register(hwmgr->device, mmMC_ARB_DRAM_TIMING2);
+	burst_time = cgs_read_register(hwmgr->device, mmMC_ARB_BURST_TIME);
+	rfsh_rate = cgs_read_register(hwmgr->device, mmMC_ARB_RFSH_RATE);
+	misc3 = cgs_read_register(hwmgr->device, mmMC_ARB_MISC3);
+
+	arb_regs->McArbDramTiming  = PP_HOST_TO_SMC_UL(dram_timing);
+	arb_regs->McArbDramTiming2 = PP_HOST_TO_SMC_UL(dram_timing2);
+	arb_regs->McArbBurstTime   = PP_HOST_TO_SMC_UL(burst_time);
+	arb_regs->McArbRfshRate = PP_HOST_TO_SMC_UL(rfsh_rate);
+	arb_regs->McArbMisc3 = PP_HOST_TO_SMC_UL(misc3);
+
+	return 0;
+}
+
+static int vegam_program_memory_timing_parameters(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *hw_data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct SMU75_Discrete_MCArbDramTimingTable arb_regs;
+	uint32_t i, j;
+	int result = 0;
+
+	memset(&arb_regs, 0, sizeof(SMU75_Discrete_MCArbDramTimingTable));
+
+	for (i = 0; i < hw_data->dpm_table.sclk_table.count; i++) {
+		for (j = 0; j < hw_data->dpm_table.mclk_table.count; j++) {
+			result = vegam_populate_memory_timing_parameters(hwmgr,
+					hw_data->dpm_table.sclk_table.dpm_levels[i].value,
+					hw_data->dpm_table.mclk_table.dpm_levels[j].value,
+					&arb_regs.entries[i][j]);
+			if (result)
+				return result;
+		}
+	}
+
+	result = smu7_copy_bytes_to_smc(
+			hwmgr,
+			smu_data->smu7_data.arb_table_start,
+			(uint8_t *)&arb_regs,
+			sizeof(SMU75_Discrete_MCArbDramTimingTable),
+			SMC_RAM_END);
+	return result;
+}
+
+static int vegam_populate_smc_uvd_level(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_DpmTable *table)
+{
+	int result = -EINVAL;
+	uint8_t count;
+	struct pp_atomctrl_clock_dividers_vi dividers;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_ppt_v1_mm_clock_voltage_dependency_table *mm_table =
+			table_info->mm_dep_table;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	uint32_t vddci;
+
+	table->UvdLevelCount = (uint8_t)(mm_table->count);
+	table->UvdBootLevel = 0;
+
+	for (count = 0; count < table->UvdLevelCount; count++) {
+		table->UvdLevel[count].MinVoltage = 0;
+		table->UvdLevel[count].VclkFrequency = mm_table->entries[count].vclk;
+		table->UvdLevel[count].DclkFrequency = mm_table->entries[count].dclk;
+		table->UvdLevel[count].MinVoltage |=
+				(mm_table->entries[count].vddc * VOLTAGE_SCALE) << VDDC_SHIFT;
+
+		if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control)
+			vddci = (uint32_t)phm_find_closest_vddci(&(data->vddci_voltage_table),
+						mm_table->entries[count].vddc - VDDC_VDDCI_DELTA);
+		else if (SMU7_VOLTAGE_CONTROL_BY_SVID2 == data->vddci_control)
+			vddci = mm_table->entries[count].vddc - VDDC_VDDCI_DELTA;
+		else
+			vddci = (data->vbios_boot_state.vddci_bootup_value * VOLTAGE_SCALE) << VDDCI_SHIFT;
+
+		table->UvdLevel[count].MinVoltage |= (vddci * VOLTAGE_SCALE) << VDDCI_SHIFT;
+		table->UvdLevel[count].MinVoltage |= 1 << PHASES_SHIFT;
+
+		/* retrieve divider value for VBIOS */
+		result = atomctrl_get_dfs_pll_dividers_vi(hwmgr,
+				table->UvdLevel[count].VclkFrequency, &dividers);
+		PP_ASSERT_WITH_CODE((0 == result),
+				"can not find divide id for Vclk clock", return result);
+
+		table->UvdLevel[count].VclkDivider = (uint8_t)dividers.pll_post_divider;
+
+		result = atomctrl_get_dfs_pll_dividers_vi(hwmgr,
+				table->UvdLevel[count].DclkFrequency, &dividers);
+		PP_ASSERT_WITH_CODE((0 == result),
+				"can not find divide id for Dclk clock", return result);
+
+		table->UvdLevel[count].DclkDivider = (uint8_t)dividers.pll_post_divider;
+
+		CONVERT_FROM_HOST_TO_SMC_UL(table->UvdLevel[count].VclkFrequency);
+		CONVERT_FROM_HOST_TO_SMC_UL(table->UvdLevel[count].DclkFrequency);
+		CONVERT_FROM_HOST_TO_SMC_UL(table->UvdLevel[count].MinVoltage);
+	}
+
+	return result;
+}
+
+static int vegam_populate_smc_boot_level(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_DpmTable *table)
+{
+	int result = 0;
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+
+	table->GraphicsBootLevel = 0;
+	table->MemoryBootLevel = 0;
+
+	/* find boot level from dpm table */
+	result = phm_find_boot_level(&(data->dpm_table.sclk_table),
+			data->vbios_boot_state.sclk_bootup_value,
+			(uint32_t *)&(table->GraphicsBootLevel));
+
+	result = phm_find_boot_level(&(data->dpm_table.mclk_table),
+			data->vbios_boot_state.mclk_bootup_value,
+			(uint32_t *)&(table->MemoryBootLevel));
+
+	table->BootVddc  = data->vbios_boot_state.vddc_bootup_value *
+			VOLTAGE_SCALE;
+	table->BootVddci = data->vbios_boot_state.vddci_bootup_value *
+			VOLTAGE_SCALE;
+	table->BootMVdd  = data->vbios_boot_state.mvdd_bootup_value *
+			VOLTAGE_SCALE;
+
+	CONVERT_FROM_HOST_TO_SMC_US(table->BootVddc);
+	CONVERT_FROM_HOST_TO_SMC_US(table->BootVddci);
+	CONVERT_FROM_HOST_TO_SMC_US(table->BootMVdd);
+
+	return 0;
+}
+
+static int vegam_populate_smc_initial_state(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *hw_data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	uint8_t count, level;
+
+	count = (uint8_t)(table_info->vdd_dep_on_sclk->count);
+
+	for (level = 0; level < count; level++) {
+		if (table_info->vdd_dep_on_sclk->entries[level].clk >=
+				hw_data->vbios_boot_state.sclk_bootup_value) {
+			smu_data->smc_state_table.GraphicsBootLevel = level;
+			break;
+		}
+	}
+
+	count = (uint8_t)(table_info->vdd_dep_on_mclk->count);
+	for (level = 0; level < count; level++) {
+		if (table_info->vdd_dep_on_mclk->entries[level].clk >=
+				hw_data->vbios_boot_state.mclk_bootup_value) {
+			smu_data->smc_state_table.MemoryBootLevel = level;
+			break;
+		}
+	}
+
+	return 0;
+}
+
+static uint16_t scale_fan_gain_settings(uint16_t raw_setting)
+{
+	uint32_t tmp;
+	tmp = raw_setting * 4096 / 100;
+	return (uint16_t)tmp;
+}
+
+static int vegam_populate_bapm_parameters_in_dpm_table(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+
+	const struct vegam_pt_defaults *defaults = smu_data->power_tune_defaults;
+	SMU75_Discrete_DpmTable  *table = &(smu_data->smc_state_table);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_cac_tdp_table *cac_dtp_table = table_info->cac_dtp_table;
+	struct pp_advance_fan_control_parameters *fan_table =
+			&hwmgr->thermal_controller.advanceFanControlParameters;
+	int i, j, k;
+	const uint16_t *pdef1;
+	const uint16_t *pdef2;
+
+	table->DefaultTdp = PP_HOST_TO_SMC_US((uint16_t)(cac_dtp_table->usTDP * 128));
+	table->TargetTdp  = PP_HOST_TO_SMC_US((uint16_t)(cac_dtp_table->usTDP * 128));
+
+	PP_ASSERT_WITH_CODE(cac_dtp_table->usTargetOperatingTemp <= 255,
+				"Target Operating Temp is out of Range!",
+				);
+
+	table->TemperatureLimitEdge = PP_HOST_TO_SMC_US(
+			cac_dtp_table->usTargetOperatingTemp * 256);
+	table->TemperatureLimitHotspot = PP_HOST_TO_SMC_US(
+			cac_dtp_table->usTemperatureLimitHotspot * 256);
+	table->FanGainEdge = PP_HOST_TO_SMC_US(
+			scale_fan_gain_settings(fan_table->usFanGainEdge));
+	table->FanGainHotspot = PP_HOST_TO_SMC_US(
+			scale_fan_gain_settings(fan_table->usFanGainHotspot));
+
+	pdef1 = defaults->BAPMTI_R;
+	pdef2 = defaults->BAPMTI_RC;
+
+	for (i = 0; i < SMU75_DTE_ITERATIONS; i++) {
+		for (j = 0; j < SMU75_DTE_SOURCES; j++) {
+			for (k = 0; k < SMU75_DTE_SINKS; k++) {
+				table->BAPMTI_R[i][j][k] = PP_HOST_TO_SMC_US(*pdef1);
+				table->BAPMTI_RC[i][j][k] = PP_HOST_TO_SMC_US(*pdef2);
+				pdef1++;
+				pdef2++;
+			}
+		}
+	}
+
+	return 0;
+}
+
+static int vegam_populate_clock_stretcher_data_table(struct pp_hwmgr *hwmgr)
+{
+	uint32_t ro, efuse, volt_without_cks, volt_with_cks, value, max, min;
+	struct vegam_smumgr *smu_data =
+			(struct vegam_smumgr *)(hwmgr->smu_backend);
+
+	uint8_t i, stretch_amount, stretch_amount2, volt_offset = 0;
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct phm_ppt_v1_clock_voltage_dependency_table *sclk_table =
+			table_info->vdd_dep_on_sclk;
+	uint32_t mask = (1 << ((STRAP_ASIC_RO_MSB - STRAP_ASIC_RO_LSB) + 1)) - 1;
+
+	stretch_amount = (uint8_t)table_info->cac_dtp_table->usClockStretchAmount;
+
+	atomctrl_read_efuse(hwmgr, STRAP_ASIC_RO_LSB, STRAP_ASIC_RO_MSB,
+			mask, &efuse);
+
+	min = 1200;
+	max = 2500;
+
+	ro = efuse * (max - min) / 255 + min;
+
+	/* Populate Sclk_CKS_masterEn0_7 and Sclk_voltageOffset */
+	for (i = 0; i < sclk_table->count; i++) {
+		smu_data->smc_state_table.Sclk_CKS_masterEn0_7 |=
+				sclk_table->entries[i].cks_enable << i;
+		volt_without_cks = (uint32_t)((2753594000U + (sclk_table->entries[i].clk/100) *
+				136418 - (ro - 70) * 1000000) /
+				(2424180 - (sclk_table->entries[i].clk/100) * 1132925/1000));
+		volt_with_cks = (uint32_t)((2797202000U + sclk_table->entries[i].clk/100 *
+				3232 - (ro - 65) * 1000000) /
+				(2522480 - sclk_table->entries[i].clk/100 * 115764/100));
+
+		if (volt_without_cks >= volt_with_cks)
+			volt_offset = (uint8_t)(((volt_without_cks - volt_with_cks +
+					sclk_table->entries[i].cks_voffset) * 100 + 624) / 625);
+
+		smu_data->smc_state_table.Sclk_voltageOffset[i] = volt_offset;
+	}
+
+	smu_data->smc_state_table.LdoRefSel =
+			(table_info->cac_dtp_table->ucCKS_LDO_REFSEL != 0) ?
+			table_info->cac_dtp_table->ucCKS_LDO_REFSEL : 5;
+	/* Populate CKS Lookup Table */
+	if (stretch_amount == 1 || stretch_amount == 2 || stretch_amount == 5)
+		stretch_amount2 = 0;
+	else if (stretch_amount == 3 || stretch_amount == 4)
+		stretch_amount2 = 1;
+	else {
+		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+				PHM_PlatformCaps_ClockStretcher);
+		PP_ASSERT_WITH_CODE(false,
+				"Stretch Amount in PPTable not supported\n",
+				return -EINVAL);
+	}
+
+	value = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixPWR_CKS_CNTL);
+	value &= 0xFFFFFFFE;
+	cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC, ixPWR_CKS_CNTL, value);
+
+	return 0;
+}
+
+static bool vegam_is_hw_avfs_present(struct pp_hwmgr *hwmgr)
+{
+	uint32_t efuse;
+
+	efuse = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+			ixSMU_EFUSE_0 + (49 * 4));
+	efuse &= 0x00000001;
+
+	if (efuse)
+		return true;
+
+	return false;
+}
+
+static int vegam_populate_avfs_parameters(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+
+	SMU75_Discrete_DpmTable  *table = &(smu_data->smc_state_table);
+	int result = 0;
+	struct pp_atom_ctrl__avfs_parameters avfs_params = {0};
+	AVFS_meanNsigma_t AVFS_meanNsigma = { {0} };
+	AVFS_Sclk_Offset_t AVFS_SclkOffset = { {0} };
+	uint32_t tmp, i;
+
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)hwmgr->pptable;
+	struct phm_ppt_v1_clock_voltage_dependency_table *sclk_table =
+			table_info->vdd_dep_on_sclk;
+
+	if (!hwmgr->avfs_supported)
+		return 0;
+
+	result = atomctrl_get_avfs_information(hwmgr, &avfs_params);
+
+	if (0 == result) {
+		table->BTCGB_VDROOP_TABLE[0].a0 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSON_a0);
+		table->BTCGB_VDROOP_TABLE[0].a1 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSON_a1);
+		table->BTCGB_VDROOP_TABLE[0].a2 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSON_a2);
+		table->BTCGB_VDROOP_TABLE[1].a0 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSOFF_a0);
+		table->BTCGB_VDROOP_TABLE[1].a1 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSOFF_a1);
+		table->BTCGB_VDROOP_TABLE[1].a2 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulGB_VDROOP_TABLE_CKSOFF_a2);
+		table->AVFSGB_FUSE_TABLE[0].m1 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulAVFSGB_FUSE_TABLE_CKSON_m1);
+		table->AVFSGB_FUSE_TABLE[0].m2 =
+				PP_HOST_TO_SMC_US(avfs_params.usAVFSGB_FUSE_TABLE_CKSON_m2);
+		table->AVFSGB_FUSE_TABLE[0].b =
+				PP_HOST_TO_SMC_UL(avfs_params.ulAVFSGB_FUSE_TABLE_CKSON_b);
+		table->AVFSGB_FUSE_TABLE[0].m1_shift = 24;
+		table->AVFSGB_FUSE_TABLE[0].m2_shift = 12;
+		table->AVFSGB_FUSE_TABLE[1].m1 =
+				PP_HOST_TO_SMC_UL(avfs_params.ulAVFSGB_FUSE_TABLE_CKSOFF_m1);
+		table->AVFSGB_FUSE_TABLE[1].m2 =
+				PP_HOST_TO_SMC_US(avfs_params.usAVFSGB_FUSE_TABLE_CKSOFF_m2);
+		table->AVFSGB_FUSE_TABLE[1].b =
+				PP_HOST_TO_SMC_UL(avfs_params.ulAVFSGB_FUSE_TABLE_CKSOFF_b);
+		table->AVFSGB_FUSE_TABLE[1].m1_shift = 24;
+		table->AVFSGB_FUSE_TABLE[1].m2_shift = 12;
+		table->MaxVoltage = PP_HOST_TO_SMC_US(avfs_params.usMaxVoltage_0_25mv);
+		AVFS_meanNsigma.Aconstant[0] =
+				PP_HOST_TO_SMC_UL(avfs_params.ulAVFS_meanNsigma_Acontant0);
+		AVFS_meanNsigma.Aconstant[1] =
+				PP_HOST_TO_SMC_UL(avfs_params.ulAVFS_meanNsigma_Acontant1);
+		AVFS_meanNsigma.Aconstant[2] =
+				PP_HOST_TO_SMC_UL(avfs_params.ulAVFS_meanNsigma_Acontant2);
+		AVFS_meanNsigma.DC_tol_sigma =
+				PP_HOST_TO_SMC_US(avfs_params.usAVFS_meanNsigma_DC_tol_sigma);
+		AVFS_meanNsigma.Platform_mean =
+				PP_HOST_TO_SMC_US(avfs_params.usAVFS_meanNsigma_Platform_mean);
+		AVFS_meanNsigma.PSM_Age_CompFactor =
+				PP_HOST_TO_SMC_US(avfs_params.usPSM_Age_ComFactor);
+		AVFS_meanNsigma.Platform_sigma =
+				PP_HOST_TO_SMC_US(avfs_params.usAVFS_meanNsigma_Platform_sigma);
+
+		for (i = 0; i < sclk_table->count; i++) {
+			AVFS_meanNsigma.Static_Voltage_Offset[i] =
+					(uint8_t)(sclk_table->entries[i].cks_voffset * 100 / 625);
+			AVFS_SclkOffset.Sclk_Offset[i] =
+					PP_HOST_TO_SMC_US((uint16_t)
+							(sclk_table->entries[i].sclk_offset) / 100);
+		}
+
+		result = smu7_read_smc_sram_dword(hwmgr,
+				SMU7_FIRMWARE_HEADER_LOCATION +
+				offsetof(SMU75_Firmware_Header, AvfsMeanNSigma),
+				&tmp, SMC_RAM_END);
+		smu7_copy_bytes_to_smc(hwmgr,
+					tmp,
+					(uint8_t *)&AVFS_meanNsigma,
+					sizeof(AVFS_meanNsigma_t),
+					SMC_RAM_END);
+
+		result = smu7_read_smc_sram_dword(hwmgr,
+				SMU7_FIRMWARE_HEADER_LOCATION +
+				offsetof(SMU75_Firmware_Header, AvfsSclkOffsetTable),
+				&tmp, SMC_RAM_END);
+		smu7_copy_bytes_to_smc(hwmgr,
+					tmp,
+					(uint8_t *)&AVFS_SclkOffset,
+					sizeof(AVFS_Sclk_Offset_t),
+					SMC_RAM_END);
+
+		data->avfs_vdroop_override_setting =
+				(avfs_params.ucEnableGB_VDROOP_TABLE_CKSON << BTCGB0_Vdroop_Enable_SHIFT) |
+				(avfs_params.ucEnableGB_VDROOP_TABLE_CKSOFF << BTCGB1_Vdroop_Enable_SHIFT) |
+				(avfs_params.ucEnableGB_FUSE_TABLE_CKSON << AVFSGB0_Vdroop_Enable_SHIFT) |
+				(avfs_params.ucEnableGB_FUSE_TABLE_CKSOFF << AVFSGB1_Vdroop_Enable_SHIFT);
+		data->apply_avfs_cks_off_voltage =
+				(avfs_params.ucEnableApplyAVFS_CKS_OFF_Voltage == 1) ? true : false;
+	}
+	return result;
+}
+
+static int vegam_populate_vr_config(struct pp_hwmgr *hwmgr,
+		struct SMU75_Discrete_DpmTable *table)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data =
+			(struct vegam_smumgr *)(hwmgr->smu_backend);
+	uint16_t config;
+
+	config = VR_MERGED_WITH_VDDC;
+	table->VRConfig |= (config << VRCONF_VDDGFX_SHIFT);
+
+	/* Set Vddc Voltage Controller */
+	if (SMU7_VOLTAGE_CONTROL_BY_SVID2 == data->voltage_control) {
+		config = VR_SVI2_PLANE_1;
+		table->VRConfig |= config;
+	} else {
+		PP_ASSERT_WITH_CODE(false,
+				"VDDC should be on SVI2 control in merged mode!",
+				);
+	}
+	/* Set Vddci Voltage Controller */
+	if (SMU7_VOLTAGE_CONTROL_BY_SVID2 == data->vddci_control) {
+		config = VR_SVI2_PLANE_2;  /* only in merged mode */
+		table->VRConfig |= (config << VRCONF_VDDCI_SHIFT);
+	} else if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->vddci_control) {
+		config = VR_SMIO_PATTERN_1;
+		table->VRConfig |= (config << VRCONF_VDDCI_SHIFT);
+	} else {
+		config = VR_STATIC_VOLTAGE;
+		table->VRConfig |= (config << VRCONF_VDDCI_SHIFT);
+	}
+	/* Set Mvdd Voltage Controller */
+	if (SMU7_VOLTAGE_CONTROL_BY_SVID2 == data->mvdd_control) {
+		if (config != VR_SVI2_PLANE_2) {
+			config = VR_SVI2_PLANE_2;
+			table->VRConfig |= (config << VRCONF_MVDD_SHIFT);
+			cgs_write_ind_register(hwmgr->device,
+					CGS_IND_REG__SMC,
+					smu_data->smu7_data.soft_regs_start +
+					offsetof(SMU75_SoftRegisters, AllowMvddSwitch),
+					0x1);
+		} else {
+			PP_ASSERT_WITH_CODE(false,
+					"SVI2 Plane 2 is already taken, set MVDD as Static",);
+			config = VR_STATIC_VOLTAGE;
+			table->VRConfig = (config << VRCONF_MVDD_SHIFT);
+		}
+	} else if (SMU7_VOLTAGE_CONTROL_BY_GPIO == data->mvdd_control) {
+		config = VR_SMIO_PATTERN_2;
+		table->VRConfig = (config << VRCONF_MVDD_SHIFT);
+		cgs_write_ind_register(hwmgr->device,
+				CGS_IND_REG__SMC,
+				smu_data->smu7_data.soft_regs_start +
+				offsetof(SMU75_SoftRegisters, AllowMvddSwitch),
+				0x1);
+	} else {
+		config = VR_STATIC_VOLTAGE;
+		table->VRConfig |= (config << VRCONF_MVDD_SHIFT);
+	}
+
+	return 0;
+}
+
+static int vegam_populate_svi_load_line(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	const struct vegam_pt_defaults *defaults = smu_data->power_tune_defaults;
+
+	smu_data->power_tune_table.SviLoadLineEn = defaults->SviLoadLineEn;
+	smu_data->power_tune_table.SviLoadLineVddC = defaults->SviLoadLineVddC;
+	smu_data->power_tune_table.SviLoadLineTrimVddC = 3;
+	smu_data->power_tune_table.SviLoadLineOffsetVddC = 0;
+
+	return 0;
+}
+
+static int vegam_populate_tdc_limit(struct pp_hwmgr *hwmgr)
+{
+	uint16_t tdc_limit;
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	const struct vegam_pt_defaults *defaults = smu_data->power_tune_defaults;
+
+	tdc_limit = (uint16_t)(table_info->cac_dtp_table->usTDC * 128);
+	smu_data->power_tune_table.TDC_VDDC_PkgLimit =
+			CONVERT_FROM_HOST_TO_SMC_US(tdc_limit);
+	smu_data->power_tune_table.TDC_VDDC_ThrottleReleaseLimitPerc =
+			defaults->TDC_VDDC_ThrottleReleaseLimitPerc;
+	smu_data->power_tune_table.TDC_MAWt = defaults->TDC_MAWt;
+
+	return 0;
+}
+
+static int vegam_populate_dw8(struct pp_hwmgr *hwmgr, uint32_t fuse_table_offset)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	const struct vegam_pt_defaults *defaults = smu_data->power_tune_defaults;
+	uint32_t temp;
+
+	if (smu7_read_smc_sram_dword(hwmgr,
+			fuse_table_offset +
+			offsetof(SMU75_Discrete_PmFuses, TdcWaterfallCtl),
+			(uint32_t *)&temp, SMC_RAM_END))
+		PP_ASSERT_WITH_CODE(false,
+				"Attempt to read PmFuses.DW6 (SviLoadLineEn) from SMC Failed!",
+				return -EINVAL);
+	else {
+		smu_data->power_tune_table.TdcWaterfallCtl = defaults->TdcWaterfallCtl;
+		smu_data->power_tune_table.LPMLTemperatureMin =
+				(uint8_t)((temp >> 16) & 0xff);
+		smu_data->power_tune_table.LPMLTemperatureMax =
+				(uint8_t)((temp >> 8) & 0xff);
+		smu_data->power_tune_table.Reserved = (uint8_t)(temp & 0xff);
+	}
+	return 0;
+}
+
+static int vegam_populate_temperature_scaler(struct pp_hwmgr *hwmgr)
+{
+	int i;
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+
+	/* Currently not used. Set all to zero. */
+	for (i = 0; i < 16; i++)
+		smu_data->power_tune_table.LPMLTemperatureScaler[i] = 0;
+
+	return 0;
+}
+
+static int vegam_populate_fuzzy_fan(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+
+/* TO DO move to hwmgr */
+	if ((hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity & (1 << 15))
+		|| 0 == hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity)
+		hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity =
+			hwmgr->thermal_controller.advanceFanControlParameters.usDefaultFanOutputSensitivity;
+
+	smu_data->power_tune_table.FuzzyFan_PwmSetDelta = PP_HOST_TO_SMC_US(
+				hwmgr->thermal_controller.advanceFanControlParameters.usFanOutputSensitivity);
+	return 0;
+}
+
+static int vegam_populate_gnb_lpml(struct pp_hwmgr *hwmgr)
+{
+	int i;
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+
+	/* Currently not used. Set all to zero. */
+	for (i = 0; i < 16; i++)
+		smu_data->power_tune_table.GnbLPML[i] = 0;
+
+	return 0;
+}
+
+static int vegam_populate_bapm_vddc_base_leakage_sidd(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	uint16_t hi_sidd = smu_data->power_tune_table.BapmVddCBaseLeakageHiSidd;
+	uint16_t lo_sidd = smu_data->power_tune_table.BapmVddCBaseLeakageLoSidd;
+	struct phm_cac_tdp_table *cac_table = table_info->cac_dtp_table;
+
+	hi_sidd = (uint16_t)(cac_table->usHighCACLeakage / 100 * 256);
+	lo_sidd = (uint16_t)(cac_table->usLowCACLeakage / 100 * 256);
+
+	smu_data->power_tune_table.BapmVddCBaseLeakageHiSidd =
+			CONVERT_FROM_HOST_TO_SMC_US(hi_sidd);
+	smu_data->power_tune_table.BapmVddCBaseLeakageLoSidd =
+			CONVERT_FROM_HOST_TO_SMC_US(lo_sidd);
+
+	return 0;
+}
+
+static int vegam_populate_pm_fuses(struct pp_hwmgr *hwmgr)
+{
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+	uint32_t pm_fuse_table_offset;
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_PowerContainment)) {
+		if (smu7_read_smc_sram_dword(hwmgr,
+				SMU7_FIRMWARE_HEADER_LOCATION +
+				offsetof(SMU75_Firmware_Header, PmFuseTable),
+				&pm_fuse_table_offset, SMC_RAM_END))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to get pm_fuse_table_offset Failed!",
+					return -EINVAL);
+
+		if (vegam_populate_svi_load_line(hwmgr))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to populate SviLoadLine Failed!",
+					return -EINVAL);
+
+		if (vegam_populate_tdc_limit(hwmgr))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to populate TDCLimit Failed!", return -EINVAL);
+
+		if (vegam_populate_dw8(hwmgr, pm_fuse_table_offset))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to populate TdcWaterfallCtl, "
+					"LPMLTemperature Min and Max Failed!",
+					return -EINVAL);
+
+		if (0 != vegam_populate_temperature_scaler(hwmgr))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to populate LPMLTemperatureScaler Failed!",
+					return -EINVAL);
+
+		if (vegam_populate_fuzzy_fan(hwmgr))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to populate Fuzzy Fan Control parameters Failed!",
+					return -EINVAL);
+
+		if (vegam_populate_gnb_lpml(hwmgr))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to populate GnbLPML Failed!",
+					return -EINVAL);
+
+		if (vegam_populate_bapm_vddc_base_leakage_sidd(hwmgr))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to populate BapmVddCBaseLeakage Hi and Lo "
+					"Sidd Failed!", return -EINVAL);
+
+		if (smu7_copy_bytes_to_smc(hwmgr, pm_fuse_table_offset,
+				(uint8_t *)&smu_data->power_tune_table,
+				(sizeof(struct SMU75_Discrete_PmFuses) - PMFUSES_AVFSSIZE),
+				SMC_RAM_END))
+			PP_ASSERT_WITH_CODE(false,
+					"Attempt to download PmFuseTable Failed!",
+					return -EINVAL);
+	}
+	return 0;
+}
+
+static int vegam_enable_reconfig_cus(struct pp_hwmgr *hwmgr)
+{
+	struct amdgpu_device *adev = hwmgr->adev;
+
+	smum_send_msg_to_smc_with_parameter(hwmgr,
+					    PPSMC_MSG_EnableModeSwitchRLCNotification,
+					    adev->gfx.cu_info.number);
+
+	return 0;
+}
+
+static int vegam_init_smc_table(struct pp_hwmgr *hwmgr)
+{
+	int result;
+	struct smu7_hwmgr *hw_data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data = (struct vegam_smumgr *)(hwmgr->smu_backend);
+
+	struct phm_ppt_v1_information *table_info =
+			(struct phm_ppt_v1_information *)(hwmgr->pptable);
+	struct SMU75_Discrete_DpmTable *table = &(smu_data->smc_state_table);
+	uint8_t i;
+	struct pp_atomctrl_gpio_pin_assignment gpio_pin;
+	struct phm_ppt_v1_gpio_table *gpio_table =
+			(struct phm_ppt_v1_gpio_table *)table_info->gpio_table;
+	pp_atomctrl_clock_dividers_vi dividers;
+
+	phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_AutomaticDCTransition);
+
+	vegam_initialize_power_tune_defaults(hwmgr);
+
+	if (SMU7_VOLTAGE_CONTROL_NONE != hw_data->voltage_control)
+		vegam_populate_smc_voltage_tables(hwmgr, table);
+
+	table->SystemFlags = 0;
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_AutomaticDCTransition))
+		table->SystemFlags |= PPSMC_SYSTEMFLAG_GPIO_DC;
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_StepVddc))
+		table->SystemFlags |= PPSMC_SYSTEMFLAG_STEPVDDC;
+
+	if (hw_data->is_memory_gddr5)
+		table->SystemFlags |= PPSMC_SYSTEMFLAG_GDDR5;
+
+	if (hw_data->ulv_supported && table_info->us_ulv_voltage_offset) {
+		result = vegam_populate_ulv_state(hwmgr, table);
+		PP_ASSERT_WITH_CODE(!result,
+				"Failed to initialize ULV state!", return result);
+		cgs_write_ind_register(hwmgr->device, CGS_IND_REG__SMC,
+				ixCG_ULV_PARAMETER, SMU7_CGULVPARAMETER_DFLT);
+	}
+
+	result = vegam_populate_smc_link_level(hwmgr, table);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize Link Level!", return result);
+
+	result = vegam_populate_all_graphic_levels(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize Graphics Level!", return result);
+
+	result = vegam_populate_all_memory_levels(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize Memory Level!", return result);
+
+	result = vegam_populate_smc_acpi_level(hwmgr, table);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize ACPI Level!", return result);
+
+	result = vegam_populate_smc_vce_level(hwmgr, table);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize VCE Level!", return result);
+
+	/* Since only the initial state is completely set up at this point
+	 * (the other states are just copies of the boot state) we only
+	 * need to populate the  ARB settings for the initial state.
+	 */
+	result = vegam_program_memory_timing_parameters(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to Write ARB settings for the initial state.", return result);
+
+	result = vegam_populate_smc_uvd_level(hwmgr, table);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize UVD Level!", return result);
+
+	result = vegam_populate_smc_boot_level(hwmgr, table);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize Boot Level!", return result);
+
+	result = vegam_populate_smc_initial_state(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to initialize Boot State!", return result);
+
+	result = vegam_populate_bapm_parameters_in_dpm_table(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to populate BAPM Parameters!", return result);
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_ClockStretcher)) {
+		result = vegam_populate_clock_stretcher_data_table(hwmgr);
+		PP_ASSERT_WITH_CODE(!result,
+				"Failed to populate Clock Stretcher Data Table!",
+				return result);
+	}
+
+	result = vegam_populate_avfs_parameters(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to populate AVFS Parameters!", return result;);
+
+	table->CurrSclkPllRange = 0xff;
+	table->GraphicsVoltageChangeEnable  = 1;
+	table->GraphicsThermThrottleEnable  = 1;
+	table->GraphicsInterval = 1;
+	table->VoltageInterval  = 1;
+	table->ThermalInterval  = 1;
+	table->TemperatureLimitHigh =
+			table_info->cac_dtp_table->usTargetOperatingTemp *
+			SMU7_Q88_FORMAT_CONVERSION_UNIT;
+	table->TemperatureLimitLow  =
+			(table_info->cac_dtp_table->usTargetOperatingTemp - 1) *
+			SMU7_Q88_FORMAT_CONVERSION_UNIT;
+	table->MemoryVoltageChangeEnable = 1;
+	table->MemoryInterval = 1;
+	table->VoltageResponseTime = 0;
+	table->PhaseResponseTime = 0;
+	table->MemoryThermThrottleEnable = 1;
+
+	PP_ASSERT_WITH_CODE(hw_data->dpm_table.pcie_speed_table.count >= 1,
+			"There must be 1 or more PCIE levels defined in PPTable.",
+			return -EINVAL);
+	table->PCIeBootLinkLevel =
+			hw_data->dpm_table.pcie_speed_table.count;
+	table->PCIeGenInterval = 1;
+	table->VRConfig = 0;
+
+	result = vegam_populate_vr_config(hwmgr, table);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to populate VRConfig setting!", return result);
+
+	table->ThermGpio = 17;
+	table->SclkStepSize = 0x4000;
+
+	if (atomctrl_get_pp_assign_pin(hwmgr,
+			VDDC_VRHOT_GPIO_PINID, &gpio_pin)) {
+		table->VRHotGpio = gpio_pin.uc_gpio_pin_bit_shift;
+		if (gpio_table)
+			table->VRHotLevel =
+					table_info->gpio_table->vrhot_triggered_sclk_dpm_index;
+	} else {
+		table->VRHotGpio = SMU7_UNUSED_GPIO_PIN;
+		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+				PHM_PlatformCaps_RegulatorHot);
+	}
+
+	if (atomctrl_get_pp_assign_pin(hwmgr,
+			PP_AC_DC_SWITCH_GPIO_PINID,	&gpio_pin)) {
+		table->AcDcGpio = gpio_pin.uc_gpio_pin_bit_shift;
+		if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+				PHM_PlatformCaps_AutomaticDCTransition) &&
+				!smum_send_msg_to_smc(hwmgr, PPSMC_MSG_UseNewGPIOScheme))
+			phm_cap_set(hwmgr->platform_descriptor.platformCaps,
+					PHM_PlatformCaps_SMCtoPPLIBAcdcGpioScheme);
+	} else {
+		table->AcDcGpio = SMU7_UNUSED_GPIO_PIN;
+		phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+				PHM_PlatformCaps_AutomaticDCTransition);
+	}
+
+	/* Thermal Output GPIO */
+	if (atomctrl_get_pp_assign_pin(hwmgr,
+			THERMAL_INT_OUTPUT_GPIO_PINID, &gpio_pin)) {
+		table->ThermOutGpio = gpio_pin.uc_gpio_pin_bit_shift;
+
+		/* For porlarity read GPIOPAD_A with assigned Gpio pin
+		 * since VBIOS will program this register to set 'inactive state',
+		 * driver can then determine 'active state' from this and
+		 * program SMU with correct polarity
+		 */
+		table->ThermOutPolarity =
+				(0 == (cgs_read_register(hwmgr->device, mmGPIOPAD_A) &
+				(1 << gpio_pin.uc_gpio_pin_bit_shift))) ? 1:0;
+		table->ThermOutMode = SMU7_THERM_OUT_MODE_THERM_ONLY;
+
+		/* if required, combine VRHot/PCC with thermal out GPIO */
+		if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+				PHM_PlatformCaps_RegulatorHot) &&
+			phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+				PHM_PlatformCaps_CombinePCCWithThermalSignal))
+			table->ThermOutMode = SMU7_THERM_OUT_MODE_THERM_VRHOT;
+	} else {
+		table->ThermOutGpio = 17;
+		table->ThermOutPolarity = 1;
+		table->ThermOutMode = SMU7_THERM_OUT_MODE_DISABLE;
+	}
+
+	/* Populate BIF_SCLK levels into SMC DPM table */
+	for (i = 0; i <= hw_data->dpm_table.pcie_speed_table.count; i++) {
+		result = atomctrl_get_dfs_pll_dividers_vi(hwmgr,
+				smu_data->bif_sclk_table[i], &dividers);
+		PP_ASSERT_WITH_CODE(!result,
+				"Can not find DFS divide id for Sclk",
+				return result);
+
+		if (i == 0)
+			table->Ulv.BifSclkDfs =
+					PP_HOST_TO_SMC_US((uint16_t)(dividers.pll_post_divider));
+		else
+			table->LinkLevel[i - 1].BifSclkDfs =
+					PP_HOST_TO_SMC_US((uint16_t)(dividers.pll_post_divider));
+	}
+
+	for (i = 0; i < SMU75_MAX_ENTRIES_SMIO; i++)
+		table->Smio[i] = PP_HOST_TO_SMC_UL(table->Smio[i]);
+
+	CONVERT_FROM_HOST_TO_SMC_UL(table->SystemFlags);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->VRConfig);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->SmioMask1);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->SmioMask2);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->SclkStepSize);
+	CONVERT_FROM_HOST_TO_SMC_UL(table->CurrSclkPllRange);
+	CONVERT_FROM_HOST_TO_SMC_US(table->TemperatureLimitHigh);
+	CONVERT_FROM_HOST_TO_SMC_US(table->TemperatureLimitLow);
+	CONVERT_FROM_HOST_TO_SMC_US(table->VoltageResponseTime);
+	CONVERT_FROM_HOST_TO_SMC_US(table->PhaseResponseTime);
+
+	/* Upload all dpm data to SMC memory.(dpm level, dpm level count etc) */
+	result = smu7_copy_bytes_to_smc(hwmgr,
+			smu_data->smu7_data.dpm_table_start +
+			offsetof(SMU75_Discrete_DpmTable, SystemFlags),
+			(uint8_t *)&(table->SystemFlags),
+			sizeof(SMU75_Discrete_DpmTable) - 3 * sizeof(SMU75_PIDController),
+			SMC_RAM_END);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to upload dpm data to SMC memory!", return result);
+
+	result = vegam_populate_pm_fuses(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to  populate PM fuses to SMC memory!", return result);
+
+	result = vegam_enable_reconfig_cus(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"Failed to enable reconfigurable CUs!", return result);
+
+	return 0;
+}
+
+static uint32_t vegam_get_offsetof(uint32_t type, uint32_t member)
+{
+	switch (type) {
+	case SMU_SoftRegisters:
+		switch (member) {
+		case HandshakeDisables:
+			return offsetof(SMU75_SoftRegisters, HandshakeDisables);
+		case VoltageChangeTimeout:
+			return offsetof(SMU75_SoftRegisters, VoltageChangeTimeout);
+		case AverageGraphicsActivity:
+			return offsetof(SMU75_SoftRegisters, AverageGraphicsActivity);
+		case PreVBlankGap:
+			return offsetof(SMU75_SoftRegisters, PreVBlankGap);
+		case VBlankTimeout:
+			return offsetof(SMU75_SoftRegisters, VBlankTimeout);
+		case UcodeLoadStatus:
+			return offsetof(SMU75_SoftRegisters, UcodeLoadStatus);
+		case DRAM_LOG_ADDR_H:
+			return offsetof(SMU75_SoftRegisters, DRAM_LOG_ADDR_H);
+		case DRAM_LOG_ADDR_L:
+			return offsetof(SMU75_SoftRegisters, DRAM_LOG_ADDR_L);
+		case DRAM_LOG_PHY_ADDR_H:
+			return offsetof(SMU75_SoftRegisters, DRAM_LOG_PHY_ADDR_H);
+		case DRAM_LOG_PHY_ADDR_L:
+			return offsetof(SMU75_SoftRegisters, DRAM_LOG_PHY_ADDR_L);
+		case DRAM_LOG_BUFF_SIZE:
+			return offsetof(SMU75_SoftRegisters, DRAM_LOG_BUFF_SIZE);
+		}
+	case SMU_Discrete_DpmTable:
+		switch (member) {
+		case UvdBootLevel:
+			return offsetof(SMU75_Discrete_DpmTable, UvdBootLevel);
+		case VceBootLevel:
+			return offsetof(SMU75_Discrete_DpmTable, VceBootLevel);
+		case LowSclkInterruptThreshold:
+			return offsetof(SMU75_Discrete_DpmTable, LowSclkInterruptThreshold);
+		}
+	}
+	pr_warn("can't get the offset of type %x member %x\n", type, member);
+	return 0;
+}
+
+static int vegam_program_mem_timing_parameters(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+
+	if (data->need_update_smu7_dpm_table &
+		(DPMTABLE_OD_UPDATE_SCLK +
+		DPMTABLE_UPDATE_SCLK +
+		DPMTABLE_UPDATE_MCLK))
+		return vegam_program_memory_timing_parameters(hwmgr);
+
+	return 0;
+}
+
+static int vegam_update_sclk_threshold(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	struct vegam_smumgr *smu_data =
+			(struct vegam_smumgr *)(hwmgr->smu_backend);
+	int result = 0;
+	uint32_t low_sclk_interrupt_threshold = 0;
+
+	if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_SclkThrottleLowNotification)
+	    && (data->low_sclk_interrupt_threshold != 0)) {
+		low_sclk_interrupt_threshold =
+				data->low_sclk_interrupt_threshold;
+
+		CONVERT_FROM_HOST_TO_SMC_UL(low_sclk_interrupt_threshold);
+
+		result = smu7_copy_bytes_to_smc(
+				hwmgr,
+				smu_data->smu7_data.dpm_table_start +
+				offsetof(SMU75_Discrete_DpmTable,
+					LowSclkInterruptThreshold),
+				(uint8_t *)&low_sclk_interrupt_threshold,
+				sizeof(uint32_t),
+				SMC_RAM_END);
+	}
+	PP_ASSERT_WITH_CODE((result == 0),
+			"Failed to update SCLK threshold!", return result);
+
+	result = vegam_program_mem_timing_parameters(hwmgr);
+	PP_ASSERT_WITH_CODE((result == 0),
+			"Failed to program memory timing parameters!",
+			);
+
+	return result;
+}
+
+int vegam_thermal_avfs_enable(struct pp_hwmgr *hwmgr)
+{
+	struct smu7_hwmgr *data = (struct smu7_hwmgr *)(hwmgr->backend);
+	int ret;
+
+	if (!hwmgr->avfs_supported)
+		return 0;
+
+	ret = smum_send_msg_to_smc(hwmgr, PPSMC_MSG_EnableAvfs);
+	if (!ret) {
+		if (data->apply_avfs_cks_off_voltage)
+			ret = smum_send_msg_to_smc(hwmgr, PPSMC_MSG_ApplyAvfsCksOffVoltage);
+	}
+
+	return ret;
+}
+
+static int vegam_thermal_setup_fan_table(struct pp_hwmgr *hwmgr)
+{
+	PP_ASSERT_WITH_CODE(hwmgr->thermal_controller.fanInfo.bNoFan,
+			"VBIOS fan info is not correct!",
+			);
+	phm_cap_unset(hwmgr->platform_descriptor.platformCaps,
+			PHM_PlatformCaps_MicrocodeFanControl);
+	return 0;
+}
+
+const struct pp_smumgr_func vegam_smu_funcs = {
+	.smu_init = vegam_smu_init,
+	.smu_fini = smu7_smu_fini,
+	.start_smu = vegam_start_smu,
+	.check_fw_load_finish = smu7_check_fw_load_finish,
+	.request_smu_load_fw = smu7_reload_firmware,
+	.request_smu_load_specific_fw = NULL,
+	.send_msg_to_smc = smu7_send_msg_to_smc,
+	.send_msg_to_smc_with_parameter = smu7_send_msg_to_smc_with_parameter,
+	.process_firmware_header = vegam_process_firmware_header,
+	.is_dpm_running = vegam_is_dpm_running,
+	.get_mac_definition = vegam_get_mac_definition,
+	.update_smc_table = vegam_update_smc_table,
+	.init_smc_table = vegam_init_smc_table,
+	.get_offsetof = vegam_get_offsetof,
+	.populate_all_graphic_levels = vegam_populate_all_graphic_levels,
+	.populate_all_memory_levels = vegam_populate_all_memory_levels,
+	.update_sclk_threshold = vegam_update_sclk_threshold,
+	.is_hw_avfs_present = vegam_is_hw_avfs_present,
+	.thermal_avfs_enable = vegam_thermal_avfs_enable,
+	.thermal_setup_fan_table = vegam_thermal_setup_fan_table,
+};
diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.h b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.h
new file mode 100644
index 000000000000..2b6558238500
--- /dev/null
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/vegam_smumgr.h
@@ -0,0 +1,75 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#ifndef _VEGAM_SMUMANAGER_H
+#define _VEGAM_SMUMANAGER_H
+
+
+#include <pp_endian.h>
+#include "smu75_discrete.h"
+#include "smu7_smumgr.h"
+
+#define SMC_RAM_END 0x40000
+
+#define DPMTuning_Uphyst_Shift    0
+#define DPMTuning_Downhyst_Shift  8
+#define DPMTuning_Activity_Shift  16
+
+#define GraphicsDPMTuning_VEGAM    0x001e6400
+#define MemoryDPMTuning_VEGAM      0x000f3c0a
+#define SclkDPMTuning_VEGAM        0x002d000a
+#define MclkDPMTuning_VEGAM        0x001f100a
+
+
+struct vegam_pt_defaults {
+	uint8_t   SviLoadLineEn;
+	uint8_t   SviLoadLineVddC;
+	uint8_t   TDC_VDDC_ThrottleReleaseLimitPerc;
+	uint8_t   TDC_MAWt;
+	uint8_t   TdcWaterfallCtl;
+	uint8_t   DTEAmbientTempBase;
+
+	uint32_t  DisplayCac;
+	uint32_t  BAPM_TEMP_GRADIENT;
+	uint16_t  BAPMTI_R[SMU75_DTE_ITERATIONS * SMU75_DTE_SOURCES * SMU75_DTE_SINKS];
+	uint16_t  BAPMTI_RC[SMU75_DTE_ITERATIONS * SMU75_DTE_SOURCES * SMU75_DTE_SINKS];
+};
+
+struct vegam_range_table {
+	uint32_t trans_lower_frequency; /* in 10khz */
+	uint32_t trans_upper_frequency;
+};
+
+struct vegam_smumgr {
+	struct smu7_smumgr smu7_data;
+	uint8_t protected_mode;
+	SMU75_Discrete_DpmTable              smc_state_table;
+	struct SMU75_Discrete_Ulv            ulv_setting;
+	struct SMU75_Discrete_PmFuses  power_tune_table;
+	struct vegam_range_table                range_table[NUM_SCLK_RANGE];
+	const struct vegam_pt_defaults       *power_tune_defaults;
+	uint32_t               bif_sclk_table[SMU75_MAX_LEVELS_LINK];
+};
+
+
+#endif
diff --git a/drivers/gpu/drm/arc/arcpgu_crtc.c b/drivers/gpu/drm/arc/arcpgu_crtc.c
index 16903dc7fe0d..965cda48dc13 100644
--- a/drivers/gpu/drm/arc/arcpgu_crtc.c
+++ b/drivers/gpu/drm/arc/arcpgu_crtc.c
@@ -136,9 +136,6 @@ static void arc_pgu_crtc_atomic_disable(struct drm_crtc *crtc,
 {
 	struct arcpgu_drm_private *arcpgu = crtc_to_arcpgu_priv(crtc);
 
-	if (!crtc->primary->fb)
-		return;
-
 	clk_disable_unprepare(arcpgu->clk);
 	arc_pgu_write(arcpgu, ARCPGU_REG_CTRL,
 			      arc_pgu_read(arcpgu, ARCPGU_REG_CTRL) &
@@ -189,7 +186,7 @@ static const struct drm_plane_helper_funcs arc_pgu_plane_helper_funcs = {
 
 static void arc_pgu_plane_destroy(struct drm_plane *plane)
 {
-	drm_plane_helper_disable(plane);
+	drm_plane_helper_disable(plane, NULL);
 	drm_plane_cleanup(plane);
 }
 
diff --git a/drivers/gpu/drm/arc/arcpgu_sim.c b/drivers/gpu/drm/arc/arcpgu_sim.c
index b8f6f9a5dfbe..68629e614990 100644
--- a/drivers/gpu/drm/arc/arcpgu_sim.c
+++ b/drivers/gpu/drm/arc/arcpgu_sim.c
@@ -99,7 +99,7 @@ int arcpgu_drm_sim_init(struct drm_device *drm, struct device_node *np)
 		goto error_encoder_cleanup;
 	}
 
-	ret = drm_mode_connector_attach_encoder(connector, encoder);
+	ret = drm_connector_attach_encoder(connector, encoder);
 	if (ret < 0) {
 		dev_err(drm->dev, "could not attach connector to encoder\n");
 		drm_connector_unregister(connector);
diff --git a/drivers/gpu/drm/arm/Makefile b/drivers/gpu/drm/arm/Makefile
index bb8b158ff90d..3bf31d1a4722 100644
--- a/drivers/gpu/drm/arm/Makefile
+++ b/drivers/gpu/drm/arm/Makefile
@@ -1,4 +1,5 @@
 hdlcd-y := hdlcd_drv.o hdlcd_crtc.o
 obj-$(CONFIG_DRM_HDLCD)	+= hdlcd.o
 mali-dp-y := malidp_drv.o malidp_hw.o malidp_planes.o malidp_crtc.o
+mali-dp-y += malidp_mw.o
 obj-$(CONFIG_DRM_MALI_DISPLAY)	+= mali-dp.o
diff --git a/drivers/gpu/drm/arm/hdlcd_crtc.c b/drivers/gpu/drm/arm/hdlcd_crtc.c
index cf5cbd63ecdf..f3f08cd6e9ef 100644
--- a/drivers/gpu/drm/arm/hdlcd_crtc.c
+++ b/drivers/gpu/drm/arm/hdlcd_crtc.c
@@ -282,7 +282,7 @@ static const struct drm_plane_helper_funcs hdlcd_plane_helper_funcs = {
 
 static void hdlcd_plane_destroy(struct drm_plane *plane)
 {
-	drm_plane_helper_disable(plane);
+	drm_plane_helper_disable(plane, NULL);
 	drm_plane_cleanup(plane);
 }
 
diff --git a/drivers/gpu/drm/arm/malidp_crtc.c b/drivers/gpu/drm/arm/malidp_crtc.c
index fcc62bc60f6a..ef44202fb43f 100644
--- a/drivers/gpu/drm/arm/malidp_crtc.c
+++ b/drivers/gpu/drm/arm/malidp_crtc.c
@@ -411,6 +411,16 @@ static int malidp_crtc_atomic_check(struct drm_crtc *crtc,
 		}
 	}
 
+	/* If only the writeback routing has changed, we don't need a modeset */
+	if (state->connectors_changed) {
+		u32 old_mask = crtc->state->connector_mask;
+		u32 new_mask = state->connector_mask;
+
+		if ((old_mask ^ new_mask) ==
+		    (1 << drm_connector_index(&malidp->mw_connector.base)))
+			state->connectors_changed = false;
+	}
+
 	ret = malidp_crtc_atomic_check_gamma(crtc, state);
 	ret = ret ? ret : malidp_crtc_atomic_check_ctm(crtc, state);
 	ret = ret ? ret : malidp_crtc_atomic_check_scaling(crtc, state);
diff --git a/drivers/gpu/drm/arm/malidp_drv.c b/drivers/gpu/drm/arm/malidp_drv.c
index 8d20faa198cf..5b7260557391 100644
--- a/drivers/gpu/drm/arm/malidp_drv.c
+++ b/drivers/gpu/drm/arm/malidp_drv.c
@@ -17,6 +17,7 @@
 #include <linux/of_graph.h>
 #include <linux/of_reserved_mem.h>
 #include <linux/pm_runtime.h>
+#include <linux/debugfs.h>
 
 #include <drm/drmP.h>
 #include <drm/drm_atomic.h>
@@ -31,6 +32,7 @@
 #include <drm/drm_of.h>
 
 #include "malidp_drv.h"
+#include "malidp_mw.h"
 #include "malidp_regs.h"
 #include "malidp_hw.h"
 
@@ -170,14 +172,15 @@ static int malidp_set_and_wait_config_valid(struct drm_device *drm)
 	struct malidp_hw_device *hwdev = malidp->dev;
 	int ret;
 
-	atomic_set(&malidp->config_valid, 0);
-	hwdev->hw->set_config_valid(hwdev);
+	hwdev->hw->set_config_valid(hwdev, 1);
 	/* don't wait for config_valid flag if we are in config mode */
-	if (hwdev->hw->in_config_mode(hwdev))
+	if (hwdev->hw->in_config_mode(hwdev)) {
+		atomic_set(&malidp->config_valid, MALIDP_CONFIG_VALID_DONE);
 		return 0;
+	}
 
 	ret = wait_event_interruptible_timeout(malidp->wq,
-			atomic_read(&malidp->config_valid) == 1,
+			atomic_read(&malidp->config_valid) == MALIDP_CONFIG_VALID_DONE,
 			msecs_to_jiffies(MALIDP_CONF_VALID_TIMEOUT));
 
 	return (ret > 0) ? 0 : -ETIMEDOUT;
@@ -216,12 +219,20 @@ static void malidp_atomic_commit_hw_done(struct drm_atomic_state *state)
 static void malidp_atomic_commit_tail(struct drm_atomic_state *state)
 {
 	struct drm_device *drm = state->dev;
+	struct malidp_drm *malidp = drm->dev_private;
 	struct drm_crtc *crtc;
 	struct drm_crtc_state *old_crtc_state;
 	int i;
 
 	pm_runtime_get_sync(drm->dev);
 
+	/*
+	 * set config_valid to a special value to let IRQ handlers
+	 * know that we are updating registers
+	 */
+	atomic_set(&malidp->config_valid, MALIDP_CONFIG_START);
+	malidp->dev->hw->set_config_valid(malidp->dev, 0);
+
 	drm_atomic_helper_commit_modeset_disables(drm, state);
 
 	for_each_old_crtc_in_state(state, crtc, old_crtc_state, i) {
@@ -230,7 +241,9 @@ static void malidp_atomic_commit_tail(struct drm_atomic_state *state)
 		malidp_atomic_commit_se_config(crtc, old_crtc_state);
 	}
 
-	drm_atomic_helper_commit_planes(drm, state, 0);
+	drm_atomic_helper_commit_planes(drm, state, DRM_PLANE_COMMIT_ACTIVE_ONLY);
+
+	malidp_mw_atomic_commit(drm, state);
 
 	drm_atomic_helper_commit_modeset_enables(drm, state);
 
@@ -268,17 +281,22 @@ static int malidp_init(struct drm_device *drm)
 	drm->mode_config.helper_private = &malidp_mode_config_helpers;
 
 	ret = malidp_crtc_init(drm);
-	if (ret) {
-		drm_mode_config_cleanup(drm);
-		return ret;
-	}
+	if (ret)
+		goto crtc_fail;
+
+	ret = malidp_mw_connector_init(drm);
+	if (ret)
+		goto crtc_fail;
 
 	return 0;
+
+crtc_fail:
+	drm_mode_config_cleanup(drm);
+	return ret;
 }
 
 static void malidp_fini(struct drm_device *drm)
 {
-	drm_atomic_helper_shutdown(drm);
 	drm_mode_config_cleanup(drm);
 }
 
@@ -286,6 +304,8 @@ static int malidp_irq_init(struct platform_device *pdev)
 {
 	int irq_de, irq_se, ret = 0;
 	struct drm_device *drm = dev_get_drvdata(&pdev->dev);
+	struct malidp_drm *malidp = drm->dev_private;
+	struct malidp_hw_device *hwdev = malidp->dev;
 
 	/* fetch the interrupts from DT */
 	irq_de = platform_get_irq_byname(pdev, "DE");
@@ -305,7 +325,7 @@ static int malidp_irq_init(struct platform_device *pdev)
 
 	ret = malidp_se_irq_init(drm, irq_se);
 	if (ret) {
-		malidp_de_irq_fini(drm);
+		malidp_de_irq_fini(hwdev);
 		return ret;
 	}
 
@@ -327,6 +347,106 @@ static int malidp_dumb_create(struct drm_file *file_priv,
 	return drm_gem_cma_dumb_create_internal(file_priv, drm, args);
 }
 
+#ifdef CONFIG_DEBUG_FS
+
+static void malidp_error_stats_init(struct malidp_error_stats *error_stats)
+{
+	error_stats->num_errors = 0;
+	error_stats->last_error_status = 0;
+	error_stats->last_error_vblank = -1;
+}
+
+void malidp_error(struct malidp_drm *malidp,
+		  struct malidp_error_stats *error_stats, u32 status,
+		  u64 vblank)
+{
+	unsigned long irqflags;
+
+	spin_lock_irqsave(&malidp->errors_lock, irqflags);
+	error_stats->last_error_status = status;
+	error_stats->last_error_vblank = vblank;
+	error_stats->num_errors++;
+	spin_unlock_irqrestore(&malidp->errors_lock, irqflags);
+}
+
+void malidp_error_stats_dump(const char *prefix,
+			     struct malidp_error_stats error_stats,
+			     struct seq_file *m)
+{
+	seq_printf(m, "[%s] num_errors : %d\n", prefix,
+		   error_stats.num_errors);
+	seq_printf(m, "[%s] last_error_status  : 0x%08x\n", prefix,
+		   error_stats.last_error_status);
+	seq_printf(m, "[%s] last_error_vblank : %lld\n", prefix,
+		   error_stats.last_error_vblank);
+}
+
+static int malidp_show_stats(struct seq_file *m, void *arg)
+{
+	struct drm_device *drm = m->private;
+	struct malidp_drm *malidp = drm->dev_private;
+	unsigned long irqflags;
+	struct malidp_error_stats de_errors, se_errors;
+
+	spin_lock_irqsave(&malidp->errors_lock, irqflags);
+	de_errors = malidp->de_errors;
+	se_errors = malidp->se_errors;
+	spin_unlock_irqrestore(&malidp->errors_lock, irqflags);
+	malidp_error_stats_dump("DE", de_errors, m);
+	malidp_error_stats_dump("SE", se_errors, m);
+	return 0;
+}
+
+static int malidp_debugfs_open(struct inode *inode, struct file *file)
+{
+	return single_open(file, malidp_show_stats, inode->i_private);
+}
+
+static ssize_t malidp_debugfs_write(struct file *file, const char __user *ubuf,
+				    size_t len, loff_t *offp)
+{
+	struct seq_file *m = file->private_data;
+	struct drm_device *drm = m->private;
+	struct malidp_drm *malidp = drm->dev_private;
+	unsigned long irqflags;
+
+	spin_lock_irqsave(&malidp->errors_lock, irqflags);
+	malidp_error_stats_init(&malidp->de_errors);
+	malidp_error_stats_init(&malidp->se_errors);
+	spin_unlock_irqrestore(&malidp->errors_lock, irqflags);
+	return len;
+}
+
+static const struct file_operations malidp_debugfs_fops = {
+	.owner = THIS_MODULE,
+	.open = malidp_debugfs_open,
+	.read = seq_read,
+	.write = malidp_debugfs_write,
+	.llseek = seq_lseek,
+	.release = single_release,
+};
+
+static int malidp_debugfs_init(struct drm_minor *minor)
+{
+	struct malidp_drm *malidp = minor->dev->dev_private;
+	struct dentry *dentry = NULL;
+
+	malidp_error_stats_init(&malidp->de_errors);
+	malidp_error_stats_init(&malidp->se_errors);
+	spin_lock_init(&malidp->errors_lock);
+	dentry = debugfs_create_file("debug",
+				     S_IRUGO | S_IWUSR,
+				     minor->debugfs_root, minor->dev,
+				     &malidp_debugfs_fops);
+	if (!dentry) {
+		DRM_ERROR("Cannot create debug file\n");
+		return -ENOMEM;
+	}
+	return 0;
+}
+
+#endif //CONFIG_DEBUG_FS
+
 static struct drm_driver malidp_driver = {
 	.driver_features = DRIVER_GEM | DRIVER_MODESET | DRIVER_ATOMIC |
 			   DRIVER_PRIME,
@@ -343,6 +463,9 @@ static struct drm_driver malidp_driver = {
 	.gem_prime_vmap = drm_gem_cma_prime_vmap,
 	.gem_prime_vunmap = drm_gem_cma_prime_vunmap,
 	.gem_prime_mmap = drm_gem_cma_prime_mmap,
+#ifdef CONFIG_DEBUG_FS
+	.debugfs_init = malidp_debugfs_init,
+#endif
 	.fops = &fops,
 	.name = "mali-dp",
 	.desc = "ARM Mali Display Processor driver",
@@ -459,6 +582,8 @@ static int malidp_runtime_pm_suspend(struct device *dev)
 	/* we can only suspend if the hardware is in config mode */
 	WARN_ON(!hwdev->hw->in_config_mode(hwdev));
 
+	malidp_se_irq_fini(hwdev);
+	malidp_de_irq_fini(hwdev);
 	hwdev->pm_suspended = true;
 	clk_disable_unprepare(hwdev->mclk);
 	clk_disable_unprepare(hwdev->aclk);
@@ -477,6 +602,8 @@ static int malidp_runtime_pm_resume(struct device *dev)
 	clk_prepare_enable(hwdev->aclk);
 	clk_prepare_enable(hwdev->mclk);
 	hwdev->pm_suspended = false;
+	malidp_de_irq_hw_init(hwdev);
+	malidp_se_irq_hw_init(hwdev);
 
 	return 0;
 }
@@ -588,8 +715,9 @@ static int malidp_bind(struct device *dev)
 	for (i = 0; i < MAX_OUTPUT_CHANNELS; i++)
 		out_depth = (out_depth << 8) | (output_width[i] & 0xf);
 	malidp_hw_write(hwdev, out_depth, hwdev->hw->map.out_depth_base);
+	hwdev->output_color_depth = out_depth;
 
-	atomic_set(&malidp->config_valid, 0);
+	atomic_set(&malidp->config_valid, MALIDP_CONFIG_VALID_INIT);
 	init_waitqueue_head(&malidp->wq);
 
 	ret = malidp_init(drm);
@@ -642,10 +770,11 @@ register_fail:
 fbdev_fail:
 	pm_runtime_get_sync(dev);
 vblank_fail:
-	malidp_se_irq_fini(drm);
-	malidp_de_irq_fini(drm);
+	malidp_se_irq_fini(hwdev);
+	malidp_de_irq_fini(hwdev);
 	drm->irq_enabled = false;
 irq_init_fail:
+	drm_atomic_helper_shutdown(drm);
 	component_unbind_all(dev, drm);
 bind_fail:
 	of_node_put(malidp->crtc.port);
@@ -672,15 +801,17 @@ static void malidp_unbind(struct device *dev)
 {
 	struct drm_device *drm = dev_get_drvdata(dev);
 	struct malidp_drm *malidp = drm->dev_private;
+	struct malidp_hw_device *hwdev = malidp->dev;
 
 	drm_dev_unregister(drm);
 	drm_fb_cma_fbdev_fini(drm);
 	drm_kms_helper_poll_fini(drm);
 	pm_runtime_get_sync(dev);
 	drm_crtc_vblank_off(&malidp->crtc);
-	malidp_se_irq_fini(drm);
-	malidp_de_irq_fini(drm);
+	malidp_se_irq_fini(hwdev);
+	malidp_de_irq_fini(hwdev);
 	drm->irq_enabled = false;
+	drm_atomic_helper_shutdown(drm);
 	component_unbind_all(dev, drm);
 	of_node_put(malidp->crtc.port);
 	malidp->crtc.port = NULL;
@@ -751,8 +882,25 @@ static int __maybe_unused malidp_pm_resume(struct device *dev)
 	return 0;
 }
 
+static int __maybe_unused malidp_pm_suspend_late(struct device *dev)
+{
+	if (!pm_runtime_status_suspended(dev)) {
+		malidp_runtime_pm_suspend(dev);
+		pm_runtime_set_suspended(dev);
+	}
+	return 0;
+}
+
+static int __maybe_unused malidp_pm_resume_early(struct device *dev)
+{
+	malidp_runtime_pm_resume(dev);
+	pm_runtime_set_active(dev);
+	return 0;
+}
+
 static const struct dev_pm_ops malidp_pm_ops = {
 	SET_SYSTEM_SLEEP_PM_OPS(malidp_pm_suspend, malidp_pm_resume) \
+	SET_LATE_SYSTEM_SLEEP_PM_OPS(malidp_pm_suspend_late, malidp_pm_resume_early) \
 	SET_RUNTIME_PM_OPS(malidp_runtime_pm_suspend, malidp_runtime_pm_resume, NULL)
 };
 
diff --git a/drivers/gpu/drm/arm/malidp_drv.h b/drivers/gpu/drm/arm/malidp_drv.h
index c70989b93387..e3eb0cb1f385 100644
--- a/drivers/gpu/drm/arm/malidp_drv.h
+++ b/drivers/gpu/drm/arm/malidp_drv.h
@@ -13,18 +13,38 @@
 #ifndef __MALIDP_DRV_H__
 #define __MALIDP_DRV_H__
 
+#include <drm/drm_writeback.h>
+#include <drm/drm_encoder.h>
 #include <linux/mutex.h>
 #include <linux/wait.h>
+#include <linux/spinlock.h>
 #include <drm/drmP.h>
 #include "malidp_hw.h"
 
+#define MALIDP_CONFIG_VALID_INIT	0
+#define MALIDP_CONFIG_VALID_DONE	1
+#define MALIDP_CONFIG_START		0xd0
+
+struct malidp_error_stats {
+	s32 num_errors;
+	u32 last_error_status;
+	s64 last_error_vblank;
+};
+
 struct malidp_drm {
 	struct malidp_hw_device *dev;
 	struct drm_crtc crtc;
+	struct drm_writeback_connector mw_connector;
 	wait_queue_head_t wq;
 	struct drm_pending_vblank_event *event;
 	atomic_t config_valid;
 	u32 core_id;
+#ifdef CONFIG_DEBUG_FS
+	struct malidp_error_stats de_errors;
+	struct malidp_error_stats se_errors;
+	/* Protects errors stats */
+	spinlock_t errors_lock;
+#endif
 };
 
 #define crtc_to_malidp_device(x) container_of(x, struct malidp_drm, crtc)
@@ -62,6 +82,12 @@ struct malidp_crtc_state {
 int malidp_de_planes_init(struct drm_device *drm);
 int malidp_crtc_init(struct drm_device *drm);
 
+#ifdef CONFIG_DEBUG_FS
+void malidp_error(struct malidp_drm *malidp,
+		  struct malidp_error_stats *error_stats, u32 status,
+		  u64 vblank);
+#endif
+
 /* often used combination of rotational bits */
 #define MALIDP_ROTATED_MASK	(DRM_MODE_ROTATE_90 | DRM_MODE_ROTATE_270)
 
diff --git a/drivers/gpu/drm/arm/malidp_hw.c b/drivers/gpu/drm/arm/malidp_hw.c
index d789b46dc817..c94a4422e0e9 100644
--- a/drivers/gpu/drm/arm/malidp_hw.c
+++ b/drivers/gpu/drm/arm/malidp_hw.c
@@ -21,15 +21,24 @@
 
 #include "malidp_drv.h"
 #include "malidp_hw.h"
+#include "malidp_mw.h"
+
+enum {
+	MW_NOT_ENABLED = 0,	/* SE writeback not enabled */
+	MW_ONESHOT,		/* SE in one-shot mode for writeback */
+	MW_START,		/* SE started writeback */
+	MW_RESTART,		/* SE will start another writeback after this one */
+	MW_STOP,		/* SE needs to stop after this writeback */
+};
 
 static const struct malidp_format_id malidp500_de_formats[] = {
 	/*    fourcc,   layers supporting the format,     internal id  */
-	{ DRM_FORMAT_ARGB2101010, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  0 },
-	{ DRM_FORMAT_ABGR2101010, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  1 },
+	{ DRM_FORMAT_ARGB2101010, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2 | SE_MEMWRITE,  0 },
+	{ DRM_FORMAT_ABGR2101010, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2 | SE_MEMWRITE,  1 },
 	{ DRM_FORMAT_ARGB8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  2 },
 	{ DRM_FORMAT_ABGR8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  3 },
-	{ DRM_FORMAT_XRGB8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  4 },
-	{ DRM_FORMAT_XBGR8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  5 },
+	{ DRM_FORMAT_XRGB8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2 | SE_MEMWRITE,  4 },
+	{ DRM_FORMAT_XBGR8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2 | SE_MEMWRITE,  5 },
 	{ DRM_FORMAT_RGB888, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  6 },
 	{ DRM_FORMAT_BGR888, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  7 },
 	{ DRM_FORMAT_RGBA5551, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2,  8 },
@@ -38,7 +47,7 @@ static const struct malidp_format_id malidp500_de_formats[] = {
 	{ DRM_FORMAT_BGR565, DE_VIDEO1 | DE_GRAPHICS1 | DE_GRAPHICS2, 11 },
 	{ DRM_FORMAT_UYVY, DE_VIDEO1, 12 },
 	{ DRM_FORMAT_YUYV, DE_VIDEO1, 13 },
-	{ DRM_FORMAT_NV12, DE_VIDEO1, 14 },
+	{ DRM_FORMAT_NV12, DE_VIDEO1 | SE_MEMWRITE, 14 },
 	{ DRM_FORMAT_YUV420, DE_VIDEO1, 15 },
 };
 
@@ -47,27 +56,27 @@ static const struct malidp_format_id malidp500_de_formats[] = {
 
 #define MALIDP_COMMON_FORMATS \
 	/*    fourcc,   layers supporting the format,      internal id   */ \
-	{ DRM_FORMAT_ARGB2101010, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2, MALIDP_ID(0, 0) }, \
-	{ DRM_FORMAT_ABGR2101010, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2, MALIDP_ID(0, 1) }, \
-	{ DRM_FORMAT_RGBA1010102, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2, MALIDP_ID(0, 2) }, \
-	{ DRM_FORMAT_BGRA1010102, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2, MALIDP_ID(0, 3) }, \
+	{ DRM_FORMAT_ARGB2101010, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | SE_MEMWRITE, MALIDP_ID(0, 0) }, \
+	{ DRM_FORMAT_ABGR2101010, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | SE_MEMWRITE, MALIDP_ID(0, 1) }, \
+	{ DRM_FORMAT_RGBA1010102, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | SE_MEMWRITE, MALIDP_ID(0, 2) }, \
+	{ DRM_FORMAT_BGRA1010102, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | SE_MEMWRITE, MALIDP_ID(0, 3) }, \
 	{ DRM_FORMAT_ARGB8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART, MALIDP_ID(1, 0) }, \
 	{ DRM_FORMAT_ABGR8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART, MALIDP_ID(1, 1) }, \
 	{ DRM_FORMAT_RGBA8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART, MALIDP_ID(1, 2) }, \
 	{ DRM_FORMAT_BGRA8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART, MALIDP_ID(1, 3) }, \
-	{ DRM_FORMAT_XRGB8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART, MALIDP_ID(2, 0) }, \
-	{ DRM_FORMAT_XBGR8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART, MALIDP_ID(2, 1) }, \
-	{ DRM_FORMAT_RGBX8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART, MALIDP_ID(2, 2) }, \
-	{ DRM_FORMAT_BGRX8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART, MALIDP_ID(2, 3) }, \
-	{ DRM_FORMAT_RGB888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2, MALIDP_ID(3, 0) }, \
-	{ DRM_FORMAT_BGR888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2, MALIDP_ID(3, 1) }, \
+	{ DRM_FORMAT_XRGB8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART | SE_MEMWRITE, MALIDP_ID(2, 0) }, \
+	{ DRM_FORMAT_XBGR8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART | SE_MEMWRITE, MALIDP_ID(2, 1) }, \
+	{ DRM_FORMAT_RGBX8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART | SE_MEMWRITE, MALIDP_ID(2, 2) }, \
+	{ DRM_FORMAT_BGRX8888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | DE_SMART | SE_MEMWRITE, MALIDP_ID(2, 3) }, \
+	{ DRM_FORMAT_RGB888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | SE_MEMWRITE, MALIDP_ID(3, 0) }, \
+	{ DRM_FORMAT_BGR888, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2 | SE_MEMWRITE, MALIDP_ID(3, 1) }, \
 	{ DRM_FORMAT_RGBA5551, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2, MALIDP_ID(4, 0) }, \
 	{ DRM_FORMAT_ABGR1555, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2, MALIDP_ID(4, 1) }, \
 	{ DRM_FORMAT_RGB565, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2, MALIDP_ID(4, 2) }, \
 	{ DRM_FORMAT_BGR565, DE_VIDEO1 | DE_GRAPHICS1 | DE_VIDEO2, MALIDP_ID(4, 3) }, \
 	{ DRM_FORMAT_YUYV, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(5, 2) },	\
 	{ DRM_FORMAT_UYVY, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(5, 3) },	\
-	{ DRM_FORMAT_NV12, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(5, 6) },	\
+	{ DRM_FORMAT_NV12, DE_VIDEO1 | DE_VIDEO2 | SE_MEMWRITE, MALIDP_ID(5, 6) },	\
 	{ DRM_FORMAT_YUV420, DE_VIDEO1 | DE_VIDEO2, MALIDP_ID(5, 7) }
 
 static const struct malidp_format_id malidp550_de_formats[] = {
@@ -223,15 +232,20 @@ static bool malidp500_in_config_mode(struct malidp_hw_device *hwdev)
 	return false;
 }
 
-static void malidp500_set_config_valid(struct malidp_hw_device *hwdev)
+static void malidp500_set_config_valid(struct malidp_hw_device *hwdev, u8 value)
 {
-	malidp_hw_setbits(hwdev, MALIDP_CFG_VALID, MALIDP500_CONFIG_VALID);
+	if (value)
+		malidp_hw_setbits(hwdev, MALIDP_CFG_VALID, MALIDP500_CONFIG_VALID);
+	else
+		malidp_hw_clearbits(hwdev, MALIDP_CFG_VALID, MALIDP500_CONFIG_VALID);
 }
 
 static void malidp500_modeset(struct malidp_hw_device *hwdev, struct videomode *mode)
 {
 	u32 val = 0;
 
+	malidp_hw_write(hwdev, hwdev->output_color_depth,
+		hwdev->hw->map.out_depth_base);
 	malidp_hw_clearbits(hwdev, MALIDP500_DC_CLEAR_MASK, MALIDP500_DC_CONTROL);
 	if (mode->flags & DISPLAY_FLAGS_HSYNC_HIGH)
 		val |= MALIDP500_HSYNCPOL;
@@ -368,6 +382,55 @@ static long malidp500_se_calc_mclk(struct malidp_hw_device *hwdev,
 	return ret;
 }
 
+static int malidp500_enable_memwrite(struct malidp_hw_device *hwdev,
+				     dma_addr_t *addrs, s32 *pitches,
+				     int num_planes, u16 w, u16 h, u32 fmt_id)
+{
+	u32 base = MALIDP500_SE_MEMWRITE_BASE;
+	u32 de_base = malidp_get_block_base(hwdev, MALIDP_DE_BLOCK);
+
+	/* enable the scaling engine block */
+	malidp_hw_setbits(hwdev, MALIDP_SCALE_ENGINE_EN, de_base + MALIDP_DE_DISPLAY_FUNC);
+
+	/* restart the writeback if already enabled */
+	if (hwdev->mw_state != MW_NOT_ENABLED)
+		hwdev->mw_state = MW_RESTART;
+	else
+		hwdev->mw_state = MW_START;
+
+	malidp_hw_write(hwdev, fmt_id, base + MALIDP_MW_FORMAT);
+	switch (num_planes) {
+	case 2:
+		malidp_hw_write(hwdev, lower_32_bits(addrs[1]), base + MALIDP_MW_P2_PTR_LOW);
+		malidp_hw_write(hwdev, upper_32_bits(addrs[1]), base + MALIDP_MW_P2_PTR_HIGH);
+		malidp_hw_write(hwdev, pitches[1], base + MALIDP_MW_P2_STRIDE);
+		/* fall through */
+	case 1:
+		malidp_hw_write(hwdev, lower_32_bits(addrs[0]), base + MALIDP_MW_P1_PTR_LOW);
+		malidp_hw_write(hwdev, upper_32_bits(addrs[0]), base + MALIDP_MW_P1_PTR_HIGH);
+		malidp_hw_write(hwdev, pitches[0], base + MALIDP_MW_P1_STRIDE);
+		break;
+	default:
+		WARN(1, "Invalid number of planes");
+	}
+
+	malidp_hw_write(hwdev, MALIDP_DE_H_ACTIVE(w) | MALIDP_DE_V_ACTIVE(h),
+			MALIDP500_SE_MEMWRITE_OUT_SIZE);
+	malidp_hw_setbits(hwdev, MALIDP_SE_MEMWRITE_EN, MALIDP500_SE_CONTROL);
+
+	return 0;
+}
+
+static void malidp500_disable_memwrite(struct malidp_hw_device *hwdev)
+{
+	u32 base = malidp_get_block_base(hwdev, MALIDP_DE_BLOCK);
+
+	if (hwdev->mw_state == MW_START || hwdev->mw_state == MW_RESTART)
+		hwdev->mw_state = MW_STOP;
+	malidp_hw_clearbits(hwdev, MALIDP_SE_MEMWRITE_EN, MALIDP500_SE_CONTROL);
+	malidp_hw_clearbits(hwdev, MALIDP_SCALE_ENGINE_EN, base + MALIDP_DE_DISPLAY_FUNC);
+}
+
 static int malidp550_query_hw(struct malidp_hw_device *hwdev)
 {
 	u32 conf = malidp_hw_read(hwdev, MALIDP550_CONFIG_ID);
@@ -447,15 +510,20 @@ static bool malidp550_in_config_mode(struct malidp_hw_device *hwdev)
 	return false;
 }
 
-static void malidp550_set_config_valid(struct malidp_hw_device *hwdev)
+static void malidp550_set_config_valid(struct malidp_hw_device *hwdev, u8 value)
 {
-	malidp_hw_setbits(hwdev, MALIDP_CFG_VALID, MALIDP550_CONFIG_VALID);
+	if (value)
+		malidp_hw_setbits(hwdev, MALIDP_CFG_VALID, MALIDP550_CONFIG_VALID);
+	else
+		malidp_hw_clearbits(hwdev, MALIDP_CFG_VALID, MALIDP550_CONFIG_VALID);
 }
 
 static void malidp550_modeset(struct malidp_hw_device *hwdev, struct videomode *mode)
 {
 	u32 val = MALIDP_DE_DEFAULT_PREFETCH_START;
 
+	malidp_hw_write(hwdev, hwdev->output_color_depth,
+		hwdev->hw->map.out_depth_base);
 	malidp_hw_write(hwdev, val, MALIDP550_DE_CONTROL);
 	/*
 	 * Mali-DP550 and Mali-DP650 encode the background color like this:
@@ -588,6 +656,51 @@ static long malidp550_se_calc_mclk(struct malidp_hw_device *hwdev,
 	return ret;
 }
 
+static int malidp550_enable_memwrite(struct malidp_hw_device *hwdev,
+				     dma_addr_t *addrs, s32 *pitches,
+				     int num_planes, u16 w, u16 h, u32 fmt_id)
+{
+	u32 base = MALIDP550_SE_MEMWRITE_BASE;
+	u32 de_base = malidp_get_block_base(hwdev, MALIDP_DE_BLOCK);
+
+	/* enable the scaling engine block */
+	malidp_hw_setbits(hwdev, MALIDP_SCALE_ENGINE_EN, de_base + MALIDP_DE_DISPLAY_FUNC);
+
+	hwdev->mw_state = MW_ONESHOT;
+
+	malidp_hw_write(hwdev, fmt_id, base + MALIDP_MW_FORMAT);
+	switch (num_planes) {
+	case 2:
+		malidp_hw_write(hwdev, lower_32_bits(addrs[1]), base + MALIDP_MW_P2_PTR_LOW);
+		malidp_hw_write(hwdev, upper_32_bits(addrs[1]), base + MALIDP_MW_P2_PTR_HIGH);
+		malidp_hw_write(hwdev, pitches[1], base + MALIDP_MW_P2_STRIDE);
+		/* fall through */
+	case 1:
+		malidp_hw_write(hwdev, lower_32_bits(addrs[0]), base + MALIDP_MW_P1_PTR_LOW);
+		malidp_hw_write(hwdev, upper_32_bits(addrs[0]), base + MALIDP_MW_P1_PTR_HIGH);
+		malidp_hw_write(hwdev, pitches[0], base + MALIDP_MW_P1_STRIDE);
+		break;
+	default:
+		WARN(1, "Invalid number of planes");
+	}
+
+	malidp_hw_write(hwdev, MALIDP_DE_H_ACTIVE(w) | MALIDP_DE_V_ACTIVE(h),
+			MALIDP550_SE_MEMWRITE_OUT_SIZE);
+	malidp_hw_setbits(hwdev, MALIDP550_SE_MEMWRITE_ONESHOT | MALIDP_SE_MEMWRITE_EN,
+			  MALIDP550_SE_CONTROL);
+
+	return 0;
+}
+
+static void malidp550_disable_memwrite(struct malidp_hw_device *hwdev)
+{
+	u32 base = malidp_get_block_base(hwdev, MALIDP_DE_BLOCK);
+
+	malidp_hw_clearbits(hwdev, MALIDP550_SE_MEMWRITE_ONESHOT | MALIDP_SE_MEMWRITE_EN,
+			    MALIDP550_SE_CONTROL);
+	malidp_hw_clearbits(hwdev, MALIDP_SCALE_ENGINE_EN, base + MALIDP_DE_DISPLAY_FUNC);
+}
+
 static int malidp650_query_hw(struct malidp_hw_device *hwdev)
 {
 	u32 conf = malidp_hw_read(hwdev, MALIDP550_CONFIG_ID);
@@ -632,10 +745,18 @@ const struct malidp_hw malidp_device[MALIDP_MAX_DEVICES] = {
 					    MALIDP500_DE_IRQ_VSYNC |
 					    MALIDP500_DE_IRQ_GLOBAL,
 				.vsync_irq = MALIDP500_DE_IRQ_VSYNC,
+				.err_mask = MALIDP_DE_IRQ_UNDERRUN |
+					    MALIDP500_DE_IRQ_AXI_ERR |
+					    MALIDP500_DE_IRQ_SATURATION,
 			},
 			.se_irq_map = {
-				.irq_mask = MALIDP500_SE_IRQ_CONF_MODE,
-				.vsync_irq = 0,
+				.irq_mask = MALIDP500_SE_IRQ_CONF_MODE |
+					    MALIDP500_SE_IRQ_CONF_VALID |
+					    MALIDP500_SE_IRQ_GLOBAL,
+				.vsync_irq = MALIDP500_SE_IRQ_CONF_VALID,
+				.err_mask = MALIDP500_SE_IRQ_INIT_BUSY |
+					    MALIDP500_SE_IRQ_AXI_ERROR |
+					    MALIDP500_SE_IRQ_OVERRUN,
 			},
 			.dc_irq_map = {
 				.irq_mask = MALIDP500_DE_IRQ_CONF_VALID,
@@ -654,6 +775,8 @@ const struct malidp_hw malidp_device[MALIDP_MAX_DEVICES] = {
 		.rotmem_required = malidp500_rotmem_required,
 		.se_set_scaling_coeffs = malidp500_se_set_scaling_coeffs,
 		.se_calc_mclk = malidp500_se_calc_mclk,
+		.enable_memwrite = malidp500_enable_memwrite,
+		.disable_memwrite = malidp500_disable_memwrite,
 		.features = MALIDP_DEVICE_LV_HAS_3_STRIDES,
 	},
 	[MALIDP_550] = {
@@ -669,13 +792,20 @@ const struct malidp_hw malidp_device[MALIDP_MAX_DEVICES] = {
 				.irq_mask = MALIDP_DE_IRQ_UNDERRUN |
 					    MALIDP550_DE_IRQ_VSYNC,
 				.vsync_irq = MALIDP550_DE_IRQ_VSYNC,
+				.err_mask = MALIDP_DE_IRQ_UNDERRUN |
+					    MALIDP550_DE_IRQ_SATURATION |
+					    MALIDP550_DE_IRQ_AXI_ERR,
 			},
 			.se_irq_map = {
-				.irq_mask = MALIDP550_SE_IRQ_EOW |
-					    MALIDP550_SE_IRQ_AXI_ERR,
+				.irq_mask = MALIDP550_SE_IRQ_EOW,
+				.vsync_irq = MALIDP550_SE_IRQ_EOW,
+				.err_mask  = MALIDP550_SE_IRQ_AXI_ERR |
+					     MALIDP550_SE_IRQ_OVR |
+					     MALIDP550_SE_IRQ_IBSY,
 			},
 			.dc_irq_map = {
-				.irq_mask = MALIDP550_DC_IRQ_CONF_VALID,
+				.irq_mask = MALIDP550_DC_IRQ_CONF_VALID |
+					    MALIDP550_DC_IRQ_SE,
 				.vsync_irq = MALIDP550_DC_IRQ_CONF_VALID,
 			},
 			.pixel_formats = malidp550_de_formats,
@@ -691,6 +821,8 @@ const struct malidp_hw malidp_device[MALIDP_MAX_DEVICES] = {
 		.rotmem_required = malidp550_rotmem_required,
 		.se_set_scaling_coeffs = malidp550_se_set_scaling_coeffs,
 		.se_calc_mclk = malidp550_se_calc_mclk,
+		.enable_memwrite = malidp550_enable_memwrite,
+		.disable_memwrite = malidp550_disable_memwrite,
 		.features = 0,
 	},
 	[MALIDP_650] = {
@@ -707,13 +839,25 @@ const struct malidp_hw malidp_device[MALIDP_MAX_DEVICES] = {
 					    MALIDP650_DE_IRQ_DRIFT |
 					    MALIDP550_DE_IRQ_VSYNC,
 				.vsync_irq = MALIDP550_DE_IRQ_VSYNC,
+				.err_mask = MALIDP_DE_IRQ_UNDERRUN |
+					    MALIDP650_DE_IRQ_DRIFT |
+					    MALIDP550_DE_IRQ_SATURATION |
+					    MALIDP550_DE_IRQ_AXI_ERR |
+					    MALIDP650_DE_IRQ_ACEV1 |
+					    MALIDP650_DE_IRQ_ACEV2 |
+					    MALIDP650_DE_IRQ_ACEG |
+					    MALIDP650_DE_IRQ_AXIEP,
 			},
 			.se_irq_map = {
-				.irq_mask = MALIDP550_SE_IRQ_EOW |
-					    MALIDP550_SE_IRQ_AXI_ERR,
+				.irq_mask = MALIDP550_SE_IRQ_EOW,
+				.vsync_irq = MALIDP550_SE_IRQ_EOW,
+				.err_mask = MALIDP550_SE_IRQ_AXI_ERR |
+					    MALIDP550_SE_IRQ_OVR |
+					    MALIDP550_SE_IRQ_IBSY,
 			},
 			.dc_irq_map = {
-				.irq_mask = MALIDP550_DC_IRQ_CONF_VALID,
+				.irq_mask = MALIDP550_DC_IRQ_CONF_VALID |
+					    MALIDP550_DC_IRQ_SE,
 				.vsync_irq = MALIDP550_DC_IRQ_CONF_VALID,
 			},
 			.pixel_formats = malidp550_de_formats,
@@ -729,6 +873,8 @@ const struct malidp_hw malidp_device[MALIDP_MAX_DEVICES] = {
 		.rotmem_required = malidp550_rotmem_required,
 		.se_set_scaling_coeffs = malidp550_se_set_scaling_coeffs,
 		.se_calc_mclk = malidp550_se_calc_mclk,
+		.enable_memwrite = malidp550_enable_memwrite,
+		.disable_memwrite = malidp550_disable_memwrite,
 		.features = 0,
 	},
 };
@@ -790,7 +936,7 @@ static irqreturn_t malidp_de_irq(int irq, void *arg)
 			malidp->event = NULL;
 			spin_unlock(&drm->event_lock);
 		}
-		atomic_set(&malidp->config_valid, 1);
+		atomic_set(&malidp->config_valid, MALIDP_CONFIG_VALID_DONE);
 		ret = IRQ_WAKE_THREAD;
 	}
 
@@ -799,10 +945,17 @@ static irqreturn_t malidp_de_irq(int irq, void *arg)
 		return ret;
 
 	mask = malidp_hw_read(hwdev, MALIDP_REG_MASKIRQ);
-	status &= mask;
+	/* keep the status of the enabled interrupts, plus the error bits */
+	status &= (mask | de->err_mask);
 	if ((status & de->vsync_irq) && malidp->crtc.enabled)
 		drm_crtc_handle_vblank(&malidp->crtc);
 
+#ifdef CONFIG_DEBUG_FS
+	if (status & de->err_mask) {
+		malidp_error(malidp, &malidp->de_errors, status,
+			     drm_crtc_vblank_count(&malidp->crtc));
+	}
+#endif
 	malidp_hw_clear_irq(hwdev, MALIDP_DE_BLOCK, status);
 
 	return (ret == IRQ_NONE) ? IRQ_HANDLED : ret;
@@ -818,6 +971,23 @@ static irqreturn_t malidp_de_irq_thread_handler(int irq, void *arg)
 	return IRQ_HANDLED;
 }
 
+void malidp_de_irq_hw_init(struct malidp_hw_device *hwdev)
+{
+	/* ensure interrupts are disabled */
+	malidp_hw_disable_irq(hwdev, MALIDP_DE_BLOCK, 0xffffffff);
+	malidp_hw_clear_irq(hwdev, MALIDP_DE_BLOCK, 0xffffffff);
+	malidp_hw_disable_irq(hwdev, MALIDP_DC_BLOCK, 0xffffffff);
+	malidp_hw_clear_irq(hwdev, MALIDP_DC_BLOCK, 0xffffffff);
+
+	/* first enable the DC block IRQs */
+	malidp_hw_enable_irq(hwdev, MALIDP_DC_BLOCK,
+			     hwdev->hw->map.dc_irq_map.irq_mask);
+
+	/* now enable the DE block IRQs */
+	malidp_hw_enable_irq(hwdev, MALIDP_DE_BLOCK,
+			     hwdev->hw->map.de_irq_map.irq_mask);
+}
+
 int malidp_de_irq_init(struct drm_device *drm, int irq)
 {
 	struct malidp_drm *malidp = drm->dev_private;
@@ -838,22 +1008,13 @@ int malidp_de_irq_init(struct drm_device *drm, int irq)
 		return ret;
 	}
 
-	/* first enable the DC block IRQs */
-	malidp_hw_enable_irq(hwdev, MALIDP_DC_BLOCK,
-			     hwdev->hw->map.dc_irq_map.irq_mask);
-
-	/* now enable the DE block IRQs */
-	malidp_hw_enable_irq(hwdev, MALIDP_DE_BLOCK,
-			     hwdev->hw->map.de_irq_map.irq_mask);
+	malidp_de_irq_hw_init(hwdev);
 
 	return 0;
 }
 
-void malidp_de_irq_fini(struct drm_device *drm)
+void malidp_de_irq_fini(struct malidp_hw_device *hwdev)
 {
-	struct malidp_drm *malidp = drm->dev_private;
-	struct malidp_hw_device *hwdev = malidp->dev;
-
 	malidp_hw_disable_irq(hwdev, MALIDP_DE_BLOCK,
 			      hwdev->hw->map.de_irq_map.irq_mask);
 	malidp_hw_disable_irq(hwdev, MALIDP_DC_BLOCK,
@@ -878,19 +1039,61 @@ static irqreturn_t malidp_se_irq(int irq, void *arg)
 		return IRQ_NONE;
 
 	status = malidp_hw_read(hwdev, hw->map.se_base + MALIDP_REG_STATUS);
-	if (!(status & se->irq_mask))
+	if (!(status & (se->irq_mask | se->err_mask)))
 		return IRQ_NONE;
 
+#ifdef CONFIG_DEBUG_FS
+	if (status & se->err_mask)
+		malidp_error(malidp, &malidp->se_errors, status,
+			     drm_crtc_vblank_count(&malidp->crtc));
+#endif
 	mask = malidp_hw_read(hwdev, hw->map.se_base + MALIDP_REG_MASKIRQ);
-	status = malidp_hw_read(hwdev, hw->map.se_base + MALIDP_REG_STATUS);
 	status &= mask;
-	/* ToDo: status decoding and firing up of VSYNC and page flip events */
+
+	if (status & se->vsync_irq) {
+		switch (hwdev->mw_state) {
+		case MW_ONESHOT:
+			drm_writeback_signal_completion(&malidp->mw_connector, 0);
+			break;
+		case MW_STOP:
+			drm_writeback_signal_completion(&malidp->mw_connector, 0);
+			/* disable writeback after stop */
+			hwdev->mw_state = MW_NOT_ENABLED;
+			break;
+		case MW_RESTART:
+			drm_writeback_signal_completion(&malidp->mw_connector, 0);
+			/* fall through to a new start */
+		case MW_START:
+			/* writeback started, need to emulate one-shot mode */
+			hw->disable_memwrite(hwdev);
+			/*
+			 * only set config_valid HW bit if there is no other update
+			 * in progress or if we raced ahead of the DE IRQ handler
+			 * and config_valid flag will not be update until later
+			 */
+			status = malidp_hw_read(hwdev, hw->map.dc_base + MALIDP_REG_STATUS);
+			if ((atomic_read(&malidp->config_valid) != MALIDP_CONFIG_START) ||
+			    (status & hw->map.dc_irq_map.vsync_irq))
+				hw->set_config_valid(hwdev, 1);
+			break;
+		}
+	}
 
 	malidp_hw_clear_irq(hwdev, MALIDP_SE_BLOCK, status);
 
 	return IRQ_HANDLED;
 }
 
+void malidp_se_irq_hw_init(struct malidp_hw_device *hwdev)
+{
+	/* ensure interrupts are disabled */
+	malidp_hw_disable_irq(hwdev, MALIDP_SE_BLOCK, 0xffffffff);
+	malidp_hw_clear_irq(hwdev, MALIDP_SE_BLOCK, 0xffffffff);
+
+	malidp_hw_enable_irq(hwdev, MALIDP_SE_BLOCK,
+			     hwdev->hw->map.se_irq_map.irq_mask);
+}
+
 static irqreturn_t malidp_se_irq_thread_handler(int irq, void *arg)
 {
 	return IRQ_HANDLED;
@@ -914,17 +1117,14 @@ int malidp_se_irq_init(struct drm_device *drm, int irq)
 		return ret;
 	}
 
-	malidp_hw_enable_irq(hwdev, MALIDP_SE_BLOCK,
-			     hwdev->hw->map.se_irq_map.irq_mask);
+	hwdev->mw_state = MW_NOT_ENABLED;
+	malidp_se_irq_hw_init(hwdev);
 
 	return 0;
 }
 
-void malidp_se_irq_fini(struct drm_device *drm)
+void malidp_se_irq_fini(struct malidp_hw_device *hwdev)
 {
-	struct malidp_drm *malidp = drm->dev_private;
-	struct malidp_hw_device *hwdev = malidp->dev;
-
 	malidp_hw_disable_irq(hwdev, MALIDP_SE_BLOCK,
 			      hwdev->hw->map.se_irq_map.irq_mask);
 }
diff --git a/drivers/gpu/drm/arm/malidp_hw.h b/drivers/gpu/drm/arm/malidp_hw.h
index b5dd6c73ec9f..ad2e96915d44 100644
--- a/drivers/gpu/drm/arm/malidp_hw.h
+++ b/drivers/gpu/drm/arm/malidp_hw.h
@@ -33,6 +33,7 @@ enum {
 	DE_GRAPHICS2 = BIT(2), /* used only in DP500 */
 	DE_VIDEO2 = BIT(3),
 	DE_SMART = BIT(4),
+	SE_MEMWRITE = BIT(5),
 };
 
 struct malidp_format_id {
@@ -52,6 +53,7 @@ struct malidp_format_id {
 struct malidp_irq_map {
 	u32 irq_mask;		/* mask of IRQs that can be enabled in the block */
 	u32 vsync_irq;		/* IRQ bit used for signaling during VSYNC */
+	u32 err_mask;		/* mask of bits that represent errors */
 };
 
 struct malidp_layer {
@@ -151,12 +153,13 @@ struct malidp_hw {
 	bool (*in_config_mode)(struct malidp_hw_device *hwdev);
 
 	/*
-	 * Set configuration valid flag for hardware parameters that can
-	 * be changed outside the configuration mode. Hardware will use
-	 * the new settings when config valid is set after the end of the
-	 * current buffer scanout
+	 * Set/clear configuration valid flag for hardware parameters that can
+	 * be changed outside the configuration mode to the given value.
+	 * Hardware will use the new settings when config valid is set,
+	 * after the end of the current buffer scanout, and will ignore
+	 * any new values for those parameters if config valid flag is cleared
 	 */
-	void (*set_config_valid)(struct malidp_hw_device *hwdev);
+	void (*set_config_valid)(struct malidp_hw_device *hwdev, u8 value);
 
 	/*
 	 * Set a new mode in hardware. Requires the hardware to be in
@@ -177,6 +180,23 @@ struct malidp_hw {
 	long (*se_calc_mclk)(struct malidp_hw_device *hwdev,
 			     struct malidp_se_config *se_config,
 			     struct videomode *vm);
+	/*
+	 * Enable writing to memory the content of the next frame
+	 * @param hwdev - malidp_hw_device structure containing the HW description
+	 * @param addrs - array of addresses for each plane
+	 * @param pitches - array of pitches for each plane
+	 * @param num_planes - number of planes to be written
+	 * @param w - width of the output frame
+	 * @param h - height of the output frame
+	 * @param fmt_id - internal format ID of output buffer
+	 */
+	int (*enable_memwrite)(struct malidp_hw_device *hwdev, dma_addr_t *addrs,
+			       s32 *pitches, int num_planes, u16 w, u16 h, u32 fmt_id);
+
+	/*
+	 * Disable the writing to memory of the next frame's content.
+	 */
+	void (*disable_memwrite)(struct malidp_hw_device *hwdev);
 
 	u8 features;
 };
@@ -210,10 +230,14 @@ struct malidp_hw_device {
 
 	u8 min_line_size;
 	u16 max_line_size;
+	u32 output_color_depth;
 
 	/* track the device PM state */
 	bool pm_suspended;
 
+	/* track the SE memory writeback state */
+	u8 mw_state;
+
 	/* size of memory used for rotating layers, up to two banks available */
 	u32 rotation_memory[2];
 };
@@ -279,9 +303,11 @@ static inline void malidp_hw_enable_irq(struct malidp_hw_device *hwdev,
 }
 
 int malidp_de_irq_init(struct drm_device *drm, int irq);
-void malidp_de_irq_fini(struct drm_device *drm);
+void malidp_se_irq_hw_init(struct malidp_hw_device *hwdev);
+void malidp_de_irq_hw_init(struct malidp_hw_device *hwdev);
+void malidp_de_irq_fini(struct malidp_hw_device *hwdev);
 int malidp_se_irq_init(struct drm_device *drm, int irq);
-void malidp_se_irq_fini(struct drm_device *drm);
+void malidp_se_irq_fini(struct malidp_hw_device *hwdev);
 
 u8 malidp_hw_get_format_id(const struct malidp_hw_regmap *map,
 			   u8 layer_id, u32 format);
diff --git a/drivers/gpu/drm/arm/malidp_mw.c b/drivers/gpu/drm/arm/malidp_mw.c
new file mode 100644
index 000000000000..cfd718e7e97c
--- /dev/null
+++ b/drivers/gpu/drm/arm/malidp_mw.c
@@ -0,0 +1,250 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2016 ARM Limited. All rights reserved.
+ * Author: Brian Starkey <brian.starkey@arm.com>
+ *
+ * ARM Mali DP Writeback connector implementation
+ */
+#include <drm/drm_atomic.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_fb_cma_helper.h>
+#include <drm/drm_gem_cma_helper.h>
+#include <drm/drmP.h>
+#include <drm/drm_writeback.h>
+
+#include "malidp_drv.h"
+#include "malidp_hw.h"
+#include "malidp_mw.h"
+
+#define to_mw_state(_state) (struct malidp_mw_connector_state *)(_state)
+
+struct malidp_mw_connector_state {
+	struct drm_connector_state base;
+	dma_addr_t addrs[2];
+	s32 pitches[2];
+	u8 format;
+	u8 n_planes;
+};
+
+static int malidp_mw_connector_get_modes(struct drm_connector *connector)
+{
+	struct drm_device *dev = connector->dev;
+
+	return drm_add_modes_noedid(connector, dev->mode_config.max_width,
+				    dev->mode_config.max_height);
+}
+
+static enum drm_mode_status
+malidp_mw_connector_mode_valid(struct drm_connector *connector,
+			       struct drm_display_mode *mode)
+{
+	struct drm_device *dev = connector->dev;
+	struct drm_mode_config *mode_config = &dev->mode_config;
+	int w = mode->hdisplay, h = mode->vdisplay;
+
+	if ((w < mode_config->min_width) || (w > mode_config->max_width))
+		return MODE_BAD_HVALUE;
+
+	if ((h < mode_config->min_height) || (h > mode_config->max_height))
+		return MODE_BAD_VVALUE;
+
+	return MODE_OK;
+}
+
+const struct drm_connector_helper_funcs malidp_mw_connector_helper_funcs = {
+	.get_modes = malidp_mw_connector_get_modes,
+	.mode_valid = malidp_mw_connector_mode_valid,
+};
+
+static void malidp_mw_connector_reset(struct drm_connector *connector)
+{
+	struct malidp_mw_connector_state *mw_state =
+		kzalloc(sizeof(*mw_state), GFP_KERNEL);
+
+	if (connector->state)
+		__drm_atomic_helper_connector_destroy_state(connector->state);
+
+	kfree(connector->state);
+	__drm_atomic_helper_connector_reset(connector, &mw_state->base);
+}
+
+static enum drm_connector_status
+malidp_mw_connector_detect(struct drm_connector *connector, bool force)
+{
+	return connector_status_disconnected;
+}
+
+static void malidp_mw_connector_destroy(struct drm_connector *connector)
+{
+	drm_connector_cleanup(connector);
+}
+
+static struct drm_connector_state *
+malidp_mw_connector_duplicate_state(struct drm_connector *connector)
+{
+	struct malidp_mw_connector_state *mw_state;
+
+	if (WARN_ON(!connector->state))
+		return NULL;
+
+	mw_state = kzalloc(sizeof(*mw_state), GFP_KERNEL);
+	if (!mw_state)
+		return NULL;
+
+	/* No need to preserve any of our driver-local data */
+	__drm_atomic_helper_connector_duplicate_state(connector, &mw_state->base);
+
+	return &mw_state->base;
+}
+
+static const struct drm_connector_funcs malidp_mw_connector_funcs = {
+	.reset = malidp_mw_connector_reset,
+	.detect = malidp_mw_connector_detect,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = malidp_mw_connector_destroy,
+	.atomic_duplicate_state = malidp_mw_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+static int
+malidp_mw_encoder_atomic_check(struct drm_encoder *encoder,
+			       struct drm_crtc_state *crtc_state,
+			       struct drm_connector_state *conn_state)
+{
+	struct malidp_mw_connector_state *mw_state = to_mw_state(conn_state);
+	struct malidp_drm *malidp = encoder->dev->dev_private;
+	struct drm_framebuffer *fb;
+	int i, n_planes;
+
+	if (!conn_state->writeback_job || !conn_state->writeback_job->fb)
+		return 0;
+
+	fb = conn_state->writeback_job->fb;
+	if ((fb->width != crtc_state->mode.hdisplay) ||
+	    (fb->height != crtc_state->mode.vdisplay)) {
+		DRM_DEBUG_KMS("Invalid framebuffer size %ux%u\n",
+				fb->width, fb->height);
+		return -EINVAL;
+	}
+
+	mw_state->format =
+		malidp_hw_get_format_id(&malidp->dev->hw->map, SE_MEMWRITE,
+					fb->format->format);
+	if (mw_state->format == MALIDP_INVALID_FORMAT_ID) {
+		struct drm_format_name_buf format_name;
+
+		DRM_DEBUG_KMS("Invalid pixel format %s\n",
+			      drm_get_format_name(fb->format->format,
+						  &format_name));
+		return -EINVAL;
+	}
+
+	n_planes = drm_format_num_planes(fb->format->format);
+	for (i = 0; i < n_planes; i++) {
+		struct drm_gem_cma_object *obj = drm_fb_cma_get_gem_obj(fb, i);
+		/* memory write buffers are never rotated */
+		u8 alignment = malidp_hw_get_pitch_align(malidp->dev, 0);
+
+		if (fb->pitches[i] & (alignment - 1)) {
+			DRM_DEBUG_KMS("Invalid pitch %u for plane %d\n",
+				      fb->pitches[i], i);
+			return -EINVAL;
+		}
+		mw_state->pitches[i] = fb->pitches[i];
+		mw_state->addrs[i] = obj->paddr + fb->offsets[i];
+	}
+	mw_state->n_planes = n_planes;
+
+	return 0;
+}
+
+static const struct drm_encoder_helper_funcs malidp_mw_encoder_helper_funcs = {
+	.atomic_check = malidp_mw_encoder_atomic_check,
+};
+
+static u32 *get_writeback_formats(struct malidp_drm *malidp, int *n_formats)
+{
+	const struct malidp_hw_regmap *map = &malidp->dev->hw->map;
+	u32 *formats;
+	int n, i;
+
+	formats = kcalloc(map->n_pixel_formats, sizeof(*formats),
+			  GFP_KERNEL);
+	if (!formats)
+		return NULL;
+
+	for (n = 0, i = 0;  i < map->n_pixel_formats; i++) {
+		if (map->pixel_formats[i].layer & SE_MEMWRITE)
+			formats[n++] = map->pixel_formats[i].format;
+	}
+
+	*n_formats = n;
+
+	return formats;
+}
+
+int malidp_mw_connector_init(struct drm_device *drm)
+{
+	struct malidp_drm *malidp = drm->dev_private;
+	u32 *formats;
+	int ret, n_formats;
+
+	if (!malidp->dev->hw->enable_memwrite)
+		return 0;
+
+	malidp->mw_connector.encoder.possible_crtcs = 1 << drm_crtc_index(&malidp->crtc);
+	drm_connector_helper_add(&malidp->mw_connector.base,
+				 &malidp_mw_connector_helper_funcs);
+
+	formats = get_writeback_formats(malidp, &n_formats);
+	if (!formats)
+		return -ENOMEM;
+
+	ret = drm_writeback_connector_init(drm, &malidp->mw_connector,
+					   &malidp_mw_connector_funcs,
+					   &malidp_mw_encoder_helper_funcs,
+					   formats, n_formats);
+	kfree(formats);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+void malidp_mw_atomic_commit(struct drm_device *drm,
+			     struct drm_atomic_state *old_state)
+{
+	struct malidp_drm *malidp = drm->dev_private;
+	struct drm_writeback_connector *mw_conn = &malidp->mw_connector;
+	struct drm_connector_state *conn_state = mw_conn->base.state;
+	struct malidp_hw_device *hwdev = malidp->dev;
+	struct malidp_mw_connector_state *mw_state;
+
+	if (!conn_state)
+		return;
+
+	mw_state = to_mw_state(conn_state);
+
+	if (conn_state->writeback_job && conn_state->writeback_job->fb) {
+		struct drm_framebuffer *fb = conn_state->writeback_job->fb;
+
+		DRM_DEV_DEBUG_DRIVER(drm->dev,
+				     "Enable memwrite %ux%u:%d %pad fmt: %u\n",
+				     fb->width, fb->height,
+				     mw_state->pitches[0],
+				     &mw_state->addrs[0],
+				     mw_state->format);
+
+		drm_writeback_queue_job(mw_conn, conn_state->writeback_job);
+		conn_state->writeback_job = NULL;
+
+		hwdev->hw->enable_memwrite(hwdev, mw_state->addrs,
+					   mw_state->pitches, mw_state->n_planes,
+					   fb->width, fb->height, mw_state->format);
+	} else {
+		DRM_DEV_DEBUG_DRIVER(drm->dev, "Disable memwrite\n");
+		hwdev->hw->disable_memwrite(hwdev);
+	}
+}
diff --git a/drivers/gpu/drm/arm/malidp_mw.h b/drivers/gpu/drm/arm/malidp_mw.h
new file mode 100644
index 000000000000..19a007676a1d
--- /dev/null
+++ b/drivers/gpu/drm/arm/malidp_mw.h
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * (C) COPYRIGHT 2016 ARM Limited. All rights reserved.
+ * Author: Brian Starkey <brian.starkey@arm.com>
+ *
+ */
+
+#ifndef __MALIDP_MW_H__
+#define __MALIDP_MW_H__
+
+int malidp_mw_connector_init(struct drm_device *drm);
+void malidp_mw_atomic_commit(struct drm_device *drm,
+			     struct drm_atomic_state *old_state);
+#endif
diff --git a/drivers/gpu/drm/arm/malidp_planes.c b/drivers/gpu/drm/arm/malidp_planes.c
index 7a44897c50fe..29409a65d864 100644
--- a/drivers/gpu/drm/arm/malidp_planes.c
+++ b/drivers/gpu/drm/arm/malidp_planes.c
@@ -23,6 +23,7 @@
 
 /* Layer specific register offsets */
 #define MALIDP_LAYER_FORMAT		0x000
+#define   LAYER_FORMAT_MASK		0x3f
 #define MALIDP_LAYER_CONTROL		0x004
 #define   LAYER_ENABLE			(1 << 0)
 #define   LAYER_FLOWCFG_MASK		7
@@ -235,8 +236,8 @@ static int malidp_de_plane_check(struct drm_plane *plane,
 	if (state->rotation & MALIDP_ROTATED_MASK) {
 		int val;
 
-		val = mp->hwdev->hw->rotmem_required(mp->hwdev, state->crtc_h,
-						     state->crtc_w,
+		val = mp->hwdev->hw->rotmem_required(mp->hwdev, state->crtc_w,
+						     state->crtc_h,
 						     fb->format->format);
 		if (val < 0)
 			return val;
@@ -337,7 +338,9 @@ static void malidp_de_plane_update(struct drm_plane *plane,
 	dest_w = plane->state->crtc_w;
 	dest_h = plane->state->crtc_h;
 
-	malidp_hw_write(mp->hwdev, ms->format, mp->layer->base);
+	val = malidp_hw_read(mp->hwdev, mp->layer->base);
+	val = (val & ~LAYER_FORMAT_MASK) | ms->format;
+	malidp_hw_write(mp->hwdev, val, mp->layer->base);
 
 	for (i = 0; i < ms->n_planes; i++) {
 		/* calculate the offset for the layer's plane registers */
diff --git a/drivers/gpu/drm/arm/malidp_regs.h b/drivers/gpu/drm/arm/malidp_regs.h
index 149024fb4432..3579d36b2a71 100644
--- a/drivers/gpu/drm/arm/malidp_regs.h
+++ b/drivers/gpu/drm/arm/malidp_regs.h
@@ -53,6 +53,8 @@
 #define MALIDP550_DE_IRQ_AXI_ERR		(1 << 16)
 #define MALIDP550_SE_IRQ_EOW			(1 << 0)
 #define MALIDP550_SE_IRQ_AXI_ERR		(1 << 16)
+#define MALIDP550_SE_IRQ_OVR			(1 << 17)
+#define MALIDP550_SE_IRQ_IBSY			(1 << 18)
 #define MALIDP550_DC_IRQ_CONF_VALID		(1 << 0)
 #define MALIDP550_DC_IRQ_CONF_MODE		(1 << 4)
 #define MALIDP550_DC_IRQ_CONF_ACTIVE		(1 << 16)
@@ -60,12 +62,18 @@
 #define MALIDP550_DC_IRQ_SE			(1 << 24)
 
 #define MALIDP650_DE_IRQ_DRIFT			(1 << 4)
+#define MALIDP650_DE_IRQ_ACEV1			(1 << 17)
+#define MALIDP650_DE_IRQ_ACEV2			(1 << 18)
+#define MALIDP650_DE_IRQ_ACEG			(1 << 19)
+#define MALIDP650_DE_IRQ_AXIEP			(1 << 28)
 
 /* bit masks that are common between products */
 #define   MALIDP_CFG_VALID		(1 << 0)
 #define   MALIDP_DISP_FUNC_GAMMA	(1 << 0)
 #define   MALIDP_DISP_FUNC_CADJ		(1 << 4)
 #define   MALIDP_DISP_FUNC_ILACED	(1 << 8)
+#define   MALIDP_SCALE_ENGINE_EN	(1 << 16)
+#define   MALIDP_SE_MEMWRITE_EN		(2 << 5)
 
 /* register offsets for IRQ management */
 #define MALIDP_REG_STATUS		0x00000
@@ -153,6 +161,16 @@
 		(((x) & MALIDP_SE_ENH_LIMIT_MASK) << 16)
 #define   MALIDP_SE_ENH_COEFF0			0x04
 
+
+/* register offsets relative to MALIDP5x0_SE_MEMWRITE_BASE */
+#define MALIDP_MW_FORMAT		0x00000
+#define MALIDP_MW_P1_STRIDE		0x00004
+#define MALIDP_MW_P2_STRIDE		0x00008
+#define MALIDP_MW_P1_PTR_LOW		0x0000c
+#define MALIDP_MW_P1_PTR_HIGH		0x00010
+#define MALIDP_MW_P2_PTR_LOW		0x0002c
+#define MALIDP_MW_P2_PTR_HIGH		0x00030
+
 /* register offsets and bits specific to DP500 */
 #define MALIDP500_ADDR_SPACE_SIZE	0x01000
 #define MALIDP500_DC_BASE		0x00000
@@ -186,7 +204,8 @@
 #define MALIDP500_DE_LG2_PTR_BASE	0x0031c
 #define MALIDP500_SE_BASE		0x00c00
 #define MALIDP500_SE_CONTROL		0x00c0c
-#define MALIDP500_SE_PTR_BASE		0x00e0c
+#define MALIDP500_SE_MEMWRITE_OUT_SIZE	0x00c2c
+#define MALIDP500_SE_MEMWRITE_BASE	0x00e00
 #define MALIDP500_DC_IRQ_BASE		0x00f00
 #define MALIDP500_CONFIG_VALID		0x00f00
 #define MALIDP500_CONFIG_ID		0x00fd4
@@ -217,6 +236,9 @@
 #define MALIDP550_DE_PERF_BASE		0x00500
 #define MALIDP550_SE_BASE		0x08000
 #define MALIDP550_SE_CONTROL		0x08010
+#define   MALIDP550_SE_MEMWRITE_ONESHOT	(1 << 7)
+#define MALIDP550_SE_MEMWRITE_OUT_SIZE	0x08030
+#define MALIDP550_SE_MEMWRITE_BASE	0x08100
 #define MALIDP550_DC_BASE		0x0c000
 #define MALIDP550_DC_CONTROL		0x0c010
 #define   MALIDP550_DC_CONFIG_REQ	(1 << 16)
diff --git a/drivers/gpu/drm/armada/armada_fb.c b/drivers/gpu/drm/armada/armada_fb.c
index ac92bce07ecd..edd15126bde9 100644
--- a/drivers/gpu/drm/armada/armada_fb.c
+++ b/drivers/gpu/drm/armada/armada_fb.c
@@ -7,30 +7,15 @@
  */
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_fb_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 #include "armada_drm.h"
 #include "armada_fb.h"
 #include "armada_gem.h"
 #include "armada_hw.h"
 
-static void armada_fb_destroy(struct drm_framebuffer *fb)
-{
-	struct armada_framebuffer *dfb = drm_fb_to_armada_fb(fb);
-
-	drm_framebuffer_cleanup(&dfb->fb);
-	drm_gem_object_put_unlocked(&dfb->obj->obj);
-	kfree(dfb);
-}
-
-static int armada_fb_create_handle(struct drm_framebuffer *fb,
-	struct drm_file *dfile, unsigned int *handle)
-{
-	struct armada_framebuffer *dfb = drm_fb_to_armada_fb(fb);
-	return drm_gem_handle_create(dfile, &dfb->obj->obj, handle);
-}
-
 static const struct drm_framebuffer_funcs armada_fb_funcs = {
-	.destroy	= armada_fb_destroy,
-	.create_handle	= armada_fb_create_handle,
+	.destroy	= drm_gem_fb_destroy,
+	.create_handle	= drm_gem_fb_create_handle,
 };
 
 struct armada_framebuffer *armada_framebuffer_create(struct drm_device *dev,
@@ -78,7 +63,7 @@ struct armada_framebuffer *armada_framebuffer_create(struct drm_device *dev,
 
 	dfb->fmt = format;
 	dfb->mod = config;
-	dfb->obj = obj;
+	dfb->fb.obj[0] = &obj->obj;
 
 	drm_helper_mode_fill_fb_struct(dev, &dfb->fb, mode);
 
diff --git a/drivers/gpu/drm/armada/armada_fb.h b/drivers/gpu/drm/armada/armada_fb.h
index 48073c4f54d8..5c130ff5da77 100644
--- a/drivers/gpu/drm/armada/armada_fb.h
+++ b/drivers/gpu/drm/armada/armada_fb.h
@@ -10,13 +10,12 @@
 
 struct armada_framebuffer {
 	struct drm_framebuffer	fb;
-	struct armada_gem_object *obj;
 	uint8_t			fmt;
 	uint8_t			mod;
 };
 #define drm_fb_to_armada_fb(dfb) \
 	container_of(dfb, struct armada_framebuffer, fb)
-#define drm_fb_obj(fb) drm_fb_to_armada_fb(fb)->obj
+#define drm_fb_obj(fb) drm_to_armada_gem((fb)->obj[0])
 
 struct armada_framebuffer *armada_framebuffer_create(struct drm_device *,
 	const struct drm_mode_fb_cmd2 *, struct armada_gem_object *);
diff --git a/drivers/gpu/drm/armada/armada_gem.c b/drivers/gpu/drm/armada/armada_gem.c
index a97f509743a5..3fb37c75c065 100644
--- a/drivers/gpu/drm/armada/armada_gem.c
+++ b/drivers/gpu/drm/armada/armada_gem.c
@@ -490,8 +490,6 @@ static const struct dma_buf_ops armada_gem_prime_dmabuf_ops = {
 	.map_dma_buf	= armada_gem_prime_map_dma_buf,
 	.unmap_dma_buf	= armada_gem_prime_unmap_dma_buf,
 	.release	= drm_gem_dmabuf_release,
-	.map_atomic	= armada_gem_dmabuf_no_kmap,
-	.unmap_atomic	= armada_gem_dmabuf_no_kunmap,
 	.map		= armada_gem_dmabuf_no_kmap,
 	.unmap		= armada_gem_dmabuf_no_kunmap,
 	.mmap		= armada_gem_dmabuf_mmap,
diff --git a/drivers/gpu/drm/ast/ast_mode.c b/drivers/gpu/drm/ast/ast_mode.c
index 036dff8a1f33..5e77d456d9bb 100644
--- a/drivers/gpu/drm/ast/ast_mode.c
+++ b/drivers/gpu/drm/ast/ast_mode.c
@@ -790,12 +790,12 @@ static int ast_get_modes(struct drm_connector *connector)
 	if (!flags)
 		edid = drm_get_edid(connector, &ast_connector->i2c->adapter);
 	if (edid) {
-		drm_mode_connector_update_edid_property(&ast_connector->base, edid);
+		drm_connector_update_edid_property(&ast_connector->base, edid);
 		ret = drm_add_edid_modes(connector, edid);
 		kfree(edid);
 		return ret;
 	} else
-		drm_mode_connector_update_edid_property(&ast_connector->base, NULL);
+		drm_connector_update_edid_property(&ast_connector->base, NULL);
 	return 0;
 }
 
@@ -900,7 +900,7 @@ static int ast_connector_init(struct drm_device *dev)
 	connector->polled = DRM_CONNECTOR_POLL_CONNECT;
 
 	encoder = list_first_entry(&dev->mode_config.encoder_list, struct drm_encoder, head);
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	ast_connector->i2c = ast_i2c_create(dev);
 	if (!ast_connector->i2c)
diff --git a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_dc.c b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_dc.c
index c1ea5c36b006..843cac222e60 100644
--- a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_dc.c
+++ b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_dc.c
@@ -681,6 +681,7 @@ static void atmel_hlcdc_dc_unload(struct drm_device *dev)
 	drm_fb_cma_fbdev_fini(dev);
 	flush_workqueue(dc->wq);
 	drm_kms_helper_poll_fini(dev);
+	drm_atomic_helper_shutdown(dev);
 	drm_mode_config_cleanup(dev);
 
 	pm_runtime_get_sync(dev->dev);
diff --git a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c
index 73c875db45f4..04440064b9b7 100644
--- a/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c
+++ b/drivers/gpu/drm/atmel-hlcdc/atmel_hlcdc_plane.c
@@ -412,9 +412,10 @@ static void atmel_hlcdc_plane_update_format(struct atmel_hlcdc_plane *plane,
 				    ATMEL_HLCDC_LAYER_FORMAT_CFG, cfg);
 }
 
-static void atmel_hlcdc_plane_update_clut(struct atmel_hlcdc_plane *plane)
+static void atmel_hlcdc_plane_update_clut(struct atmel_hlcdc_plane *plane,
+					  struct atmel_hlcdc_plane_state *state)
 {
-	struct drm_crtc *crtc = plane->base.crtc;
+	struct drm_crtc *crtc = state->base.crtc;
 	struct drm_color_lut *lut;
 	int idx;
 
@@ -779,7 +780,7 @@ static void atmel_hlcdc_plane_atomic_update(struct drm_plane *p,
 	atmel_hlcdc_plane_update_pos_and_size(plane, state);
 	atmel_hlcdc_plane_update_general_settings(plane, state);
 	atmel_hlcdc_plane_update_format(plane, state);
-	atmel_hlcdc_plane_update_clut(plane);
+	atmel_hlcdc_plane_update_clut(plane, state);
 	atmel_hlcdc_plane_update_buffers(plane, state);
 	atmel_hlcdc_plane_update_disc_area(plane, state);
 
@@ -816,16 +817,6 @@ static void atmel_hlcdc_plane_atomic_disable(struct drm_plane *p,
 	atmel_hlcdc_layer_read_reg(&plane->layer, ATMEL_HLCDC_LAYER_ISR);
 }
 
-static void atmel_hlcdc_plane_destroy(struct drm_plane *p)
-{
-	struct atmel_hlcdc_plane *plane = drm_plane_to_atmel_hlcdc_plane(p);
-
-	if (plane->base.fb)
-		drm_framebuffer_put(plane->base.fb);
-
-	drm_plane_cleanup(p);
-}
-
 static int atmel_hlcdc_plane_init_properties(struct atmel_hlcdc_plane *plane)
 {
 	const struct atmel_hlcdc_layer_desc *desc = plane->layer.desc;
@@ -839,7 +830,7 @@ static int atmel_hlcdc_plane_init_properties(struct atmel_hlcdc_plane *plane)
 			return ret;
 	}
 
-	if (desc->layout.xstride && desc->layout.pstride) {
+	if (desc->layout.xstride[0] && desc->layout.pstride[0]) {
 		int ret;
 
 		ret = drm_plane_create_rotation_property(&plane->base,
@@ -1002,7 +993,7 @@ static void atmel_hlcdc_plane_atomic_destroy_state(struct drm_plane *p,
 static const struct drm_plane_funcs layer_plane_funcs = {
 	.update_plane = drm_atomic_helper_update_plane,
 	.disable_plane = drm_atomic_helper_disable_plane,
-	.destroy = atmel_hlcdc_plane_destroy,
+	.destroy = drm_plane_cleanup,
 	.reset = atmel_hlcdc_plane_reset,
 	.atomic_duplicate_state = atmel_hlcdc_plane_atomic_duplicate_state,
 	.atomic_destroy_state = atmel_hlcdc_plane_atomic_destroy_state,
diff --git a/drivers/gpu/drm/bochs/bochs_kms.c b/drivers/gpu/drm/bochs/bochs_kms.c
index 233980a78591..ca5a9afdd5cf 100644
--- a/drivers/gpu/drm/bochs/bochs_kms.c
+++ b/drivers/gpu/drm/bochs/bochs_kms.c
@@ -259,7 +259,7 @@ int bochs_kms_init(struct bochs_device *bochs)
 	bochs_crtc_init(bochs->dev);
 	bochs_encoder_init(bochs->dev);
 	bochs_connector_init(bochs->dev);
-	drm_mode_connector_attach_encoder(&bochs->connector,
+	drm_connector_attach_encoder(&bochs->connector,
 					  &bochs->encoder);
 
 	return 0;
diff --git a/drivers/gpu/drm/bridge/Kconfig b/drivers/gpu/drm/bridge/Kconfig
index 1d75d3a1f951..bf6cad6c9178 100644
--- a/drivers/gpu/drm/bridge/Kconfig
+++ b/drivers/gpu/drm/bridge/Kconfig
@@ -82,8 +82,11 @@ config DRM_PARADE_PS8622
 
 config DRM_SIL_SII8620
 	tristate "Silicon Image SII8620 HDMI/MHL bridge"
-	depends on OF && RC_CORE
+	depends on OF
 	select DRM_KMS_HELPER
+	imply EXTCON
+	select INPUT
+	select RC_CORE
 	help
 	  Silicon Image SII8620 HDMI/MHL bridge chip driver.
 
diff --git a/drivers/gpu/drm/bridge/adv7511/Kconfig b/drivers/gpu/drm/bridge/adv7511/Kconfig
index 592b9d2ec034..944e440c4fde 100644
--- a/drivers/gpu/drm/bridge/adv7511/Kconfig
+++ b/drivers/gpu/drm/bridge/adv7511/Kconfig
@@ -1,5 +1,5 @@
 config DRM_I2C_ADV7511
-	tristate "AV7511 encoder"
+	tristate "ADV7511 encoder"
 	depends on OF
 	select DRM_KMS_HELPER
 	select REGMAP_I2C
diff --git a/drivers/gpu/drm/bridge/adv7511/adv7511_drv.c b/drivers/gpu/drm/bridge/adv7511/adv7511_drv.c
index 2614cea538e2..6437b878724a 100644
--- a/drivers/gpu/drm/bridge/adv7511/adv7511_drv.c
+++ b/drivers/gpu/drm/bridge/adv7511/adv7511_drv.c
@@ -601,7 +601,7 @@ static int adv7511_get_modes(struct adv7511 *adv7511,
 		__adv7511_power_off(adv7511);
 
 
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 	count = drm_add_edid_modes(connector, edid);
 
 	adv7511_set_config_csc(adv7511, connector, adv7511->rgb,
@@ -860,7 +860,7 @@ static int adv7511_bridge_attach(struct drm_bridge *bridge)
 	}
 	drm_connector_helper_add(&adv->connector,
 				 &adv7511_connector_helper_funcs);
-	drm_mode_connector_attach_encoder(&adv->connector, bridge->encoder);
+	drm_connector_attach_encoder(&adv->connector, bridge->encoder);
 
 	if (adv->type == ADV7533)
 		ret = adv7533_attach_dsi(adv);
@@ -1127,7 +1127,7 @@ static int adv7511_probe(struct i2c_client *i2c, const struct i2c_device_id *id)
 	}
 
 	if (adv7511->gpio_pd) {
-		mdelay(5);
+		usleep_range(5000, 6000);
 		gpiod_set_value_cansleep(adv7511->gpio_pd, 0);
 	}
 
diff --git a/drivers/gpu/drm/bridge/analogix-anx78xx.c b/drivers/gpu/drm/bridge/analogix-anx78xx.c
index b49043866be6..f8433c93f463 100644
--- a/drivers/gpu/drm/bridge/analogix-anx78xx.c
+++ b/drivers/gpu/drm/bridge/analogix-anx78xx.c
@@ -969,8 +969,8 @@ static int anx78xx_get_modes(struct drm_connector *connector)
 		goto unlock;
 	}
 
-	err = drm_mode_connector_update_edid_property(connector,
-						      anx78xx->edid);
+	err = drm_connector_update_edid_property(connector,
+						 anx78xx->edid);
 	if (err) {
 		DRM_ERROR("Failed to update EDID property: %d\n", err);
 		goto unlock;
@@ -1048,8 +1048,8 @@ static int anx78xx_bridge_attach(struct drm_bridge *bridge)
 
 	anx78xx->connector.polled = DRM_CONNECTOR_POLL_HPD;
 
-	err = drm_mode_connector_attach_encoder(&anx78xx->connector,
-						bridge->encoder);
+	err = drm_connector_attach_encoder(&anx78xx->connector,
+					   bridge->encoder);
 	if (err) {
 		DRM_ERROR("Failed to link up connector to encoder: %d\n", err);
 		return err;
diff --git a/drivers/gpu/drm/bridge/analogix/analogix_dp_core.c b/drivers/gpu/drm/bridge/analogix/analogix_dp_core.c
index 2bcbfadb6ac5..d68986cea132 100644
--- a/drivers/gpu/drm/bridge/analogix/analogix_dp_core.c
+++ b/drivers/gpu/drm/bridge/analogix/analogix_dp_core.c
@@ -1119,8 +1119,8 @@ static int analogix_dp_get_modes(struct drm_connector *connector)
 		edid = drm_get_edid(connector, &dp->aux.ddc);
 		pm_runtime_put(dp->dev);
 		if (edid) {
-			drm_mode_connector_update_edid_property(&dp->connector,
-								edid);
+			drm_connector_update_edid_property(&dp->connector,
+							   edid);
 			num_modes += drm_add_edid_modes(&dp->connector, edid);
 			kfree(edid);
 		}
@@ -1210,7 +1210,7 @@ static int analogix_dp_bridge_attach(struct drm_bridge *bridge)
 
 		drm_connector_helper_add(connector,
 					 &analogix_dp_connector_helper_funcs);
-		drm_mode_connector_attach_encoder(connector, encoder);
+		drm_connector_attach_encoder(connector, encoder);
 	}
 
 	/*
diff --git a/drivers/gpu/drm/bridge/cdns-dsi.c b/drivers/gpu/drm/bridge/cdns-dsi.c
index c255fc3e1be5..ce9496d13986 100644
--- a/drivers/gpu/drm/bridge/cdns-dsi.c
+++ b/drivers/gpu/drm/bridge/cdns-dsi.c
@@ -1152,7 +1152,7 @@ static int cdns_dsi_attach(struct mipi_dsi_host *host,
 		np = of_node_get(dev->dev.of_node);
 
 	panel = of_drm_find_panel(np);
-	if (panel) {
+	if (!IS_ERR(panel)) {
 		bridge = drm_panel_bridge_add(panel, DRM_MODE_CONNECTOR_DSI);
 	} else {
 		bridge = of_drm_find_bridge(dev->dev.of_node);
@@ -1337,7 +1337,7 @@ static const struct mipi_dsi_host_ops cdns_dsi_ops = {
 	.transfer = cdns_dsi_transfer,
 };
 
-static int cdns_dsi_resume(struct device *dev)
+static int __maybe_unused cdns_dsi_resume(struct device *dev)
 {
 	struct cdns_dsi *dsi = dev_get_drvdata(dev);
 
@@ -1350,7 +1350,7 @@ static int cdns_dsi_resume(struct device *dev)
 	return 0;
 }
 
-static int cdns_dsi_suspend(struct device *dev)
+static int __maybe_unused cdns_dsi_suspend(struct device *dev)
 {
 	struct cdns_dsi *dsi = dev_get_drvdata(dev);
 
diff --git a/drivers/gpu/drm/bridge/dumb-vga-dac.c b/drivers/gpu/drm/bridge/dumb-vga-dac.c
index 498d5948d1a8..9b706789a341 100644
--- a/drivers/gpu/drm/bridge/dumb-vga-dac.c
+++ b/drivers/gpu/drm/bridge/dumb-vga-dac.c
@@ -55,8 +55,10 @@ static int dumb_vga_get_modes(struct drm_connector *connector)
 		goto fallback;
 	}
 
-	drm_mode_connector_update_edid_property(connector, edid);
-	return drm_add_edid_modes(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
+	ret = drm_add_edid_modes(connector, edid);
+	kfree(edid);
+	return ret;
 
 fallback:
 	/*
@@ -120,7 +122,7 @@ static int dumb_vga_attach(struct drm_bridge *bridge)
 		return ret;
 	}
 
-	drm_mode_connector_attach_encoder(&vga->connector,
+	drm_connector_attach_encoder(&vga->connector,
 					  bridge->encoder);
 
 	return 0;
diff --git a/drivers/gpu/drm/bridge/lvds-encoder.c b/drivers/gpu/drm/bridge/lvds-encoder.c
index 75b0d3f6e4de..f56c92f7af7c 100644
--- a/drivers/gpu/drm/bridge/lvds-encoder.c
+++ b/drivers/gpu/drm/bridge/lvds-encoder.c
@@ -68,9 +68,9 @@ static int lvds_encoder_probe(struct platform_device *pdev)
 
 	panel = of_drm_find_panel(panel_node);
 	of_node_put(panel_node);
-	if (!panel) {
+	if (IS_ERR(panel)) {
 		dev_dbg(&pdev->dev, "panel not found, deferring probe\n");
-		return -EPROBE_DEFER;
+		return PTR_ERR(panel);
 	}
 
 	lvds_encoder->panel_bridge =
diff --git a/drivers/gpu/drm/bridge/megachips-stdpxxxx-ge-b850v3-fw.c b/drivers/gpu/drm/bridge/megachips-stdpxxxx-ge-b850v3-fw.c
index 7ccadba7c98c..2136c97aeb8e 100644
--- a/drivers/gpu/drm/bridge/megachips-stdpxxxx-ge-b850v3-fw.c
+++ b/drivers/gpu/drm/bridge/megachips-stdpxxxx-ge-b850v3-fw.c
@@ -152,7 +152,7 @@ static int ge_b850v3_lvds_get_modes(struct drm_connector *connector)
 	ge_b850v3_lvds_ptr->edid = (struct edid *)stdp2690_get_edid(client);
 
 	if (ge_b850v3_lvds_ptr->edid) {
-		drm_mode_connector_update_edid_property(connector,
+		drm_connector_update_edid_property(connector,
 						      ge_b850v3_lvds_ptr->edid);
 		num_modes = drm_add_edid_modes(connector,
 					       ge_b850v3_lvds_ptr->edid);
@@ -241,7 +241,7 @@ static int ge_b850v3_lvds_attach(struct drm_bridge *bridge)
 		return ret;
 	}
 
-	ret = drm_mode_connector_attach_encoder(connector, bridge->encoder);
+	ret = drm_connector_attach_encoder(connector, bridge->encoder);
 	if (ret)
 		return ret;
 
diff --git a/drivers/gpu/drm/bridge/nxp-ptn3460.c b/drivers/gpu/drm/bridge/nxp-ptn3460.c
index d64a3283822a..a3e817abace1 100644
--- a/drivers/gpu/drm/bridge/nxp-ptn3460.c
+++ b/drivers/gpu/drm/bridge/nxp-ptn3460.c
@@ -222,7 +222,7 @@ static int ptn3460_get_modes(struct drm_connector *connector)
 	}
 
 	ptn_bridge->edid = (struct edid *)edid;
-	drm_mode_connector_update_edid_property(connector, ptn_bridge->edid);
+	drm_connector_update_edid_property(connector, ptn_bridge->edid);
 
 	num_modes = drm_add_edid_modes(connector, ptn_bridge->edid);
 
@@ -265,7 +265,7 @@ static int ptn3460_bridge_attach(struct drm_bridge *bridge)
 	drm_connector_helper_add(&ptn_bridge->connector,
 					&ptn3460_connector_helper_funcs);
 	drm_connector_register(&ptn_bridge->connector);
-	drm_mode_connector_attach_encoder(&ptn_bridge->connector,
+	drm_connector_attach_encoder(&ptn_bridge->connector,
 							bridge->encoder);
 
 	if (ptn_bridge->panel)
diff --git a/drivers/gpu/drm/bridge/panel.c b/drivers/gpu/drm/bridge/panel.c
index 6d99d4a3beb3..7cbaba213ef6 100644
--- a/drivers/gpu/drm/bridge/panel.c
+++ b/drivers/gpu/drm/bridge/panel.c
@@ -79,7 +79,7 @@ static int panel_bridge_attach(struct drm_bridge *bridge)
 		return ret;
 	}
 
-	drm_mode_connector_attach_encoder(&panel_bridge->connector,
+	drm_connector_attach_encoder(&panel_bridge->connector,
 					  bridge->encoder);
 
 	ret = drm_panel_attach(panel_bridge->panel, &panel_bridge->connector);
diff --git a/drivers/gpu/drm/bridge/parade-ps8622.c b/drivers/gpu/drm/bridge/parade-ps8622.c
index 81198f5e9afa..7334d1b62b71 100644
--- a/drivers/gpu/drm/bridge/parade-ps8622.c
+++ b/drivers/gpu/drm/bridge/parade-ps8622.c
@@ -503,7 +503,7 @@ static int ps8622_attach(struct drm_bridge *bridge)
 	drm_connector_helper_add(&ps8622->connector,
 					&ps8622_connector_helper_funcs);
 	drm_connector_register(&ps8622->connector);
-	drm_mode_connector_attach_encoder(&ps8622->connector,
+	drm_connector_attach_encoder(&ps8622->connector,
 							bridge->encoder);
 
 	if (ps8622->panel)
diff --git a/drivers/gpu/drm/bridge/sii902x.c b/drivers/gpu/drm/bridge/sii902x.c
index 60373d7eb220..e59a13542333 100644
--- a/drivers/gpu/drm/bridge/sii902x.c
+++ b/drivers/gpu/drm/bridge/sii902x.c
@@ -170,7 +170,7 @@ static int sii902x_get_modes(struct drm_connector *connector)
 		return ret;
 
 	edid = drm_get_edid(connector, sii902x->i2c->adapter);
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 	if (edid) {
 		num = drm_add_edid_modes(connector, edid);
 		kfree(edid);
@@ -324,7 +324,7 @@ static int sii902x_bridge_attach(struct drm_bridge *bridge)
 	else
 		sii902x->connector.polled = DRM_CONNECTOR_POLL_CONNECT;
 
-	drm_mode_connector_attach_encoder(&sii902x->connector, bridge->encoder);
+	drm_connector_attach_encoder(&sii902x->connector, bridge->encoder);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/bridge/sil-sii8620.c b/drivers/gpu/drm/bridge/sil-sii8620.c
index 7ab36042a822..250effa0e6b8 100644
--- a/drivers/gpu/drm/bridge/sil-sii8620.c
+++ b/drivers/gpu/drm/bridge/sil-sii8620.c
@@ -36,8 +36,11 @@
 
 #define SII8620_BURST_BUF_LEN 288
 #define VAL_RX_HDMI_CTRL2_DEFVAL VAL_RX_HDMI_CTRL2_IDLE_CNT(3)
-#define MHL1_MAX_LCLK 225000
-#define MHL3_MAX_LCLK 600000
+
+#define MHL1_MAX_PCLK 75000
+#define MHL1_MAX_PCLK_PP_MODE 150000
+#define MHL3_MAX_PCLK 200000
+#define MHL3_MAX_PCLK_PP_MODE 300000
 
 enum sii8620_mode {
 	CM_DISCONNECTED,
@@ -80,6 +83,9 @@ struct sii8620 {
 	u8 devcap[MHL_DCAP_SIZE];
 	u8 xdevcap[MHL_XDC_SIZE];
 	u8 avif[HDMI_INFOFRAME_SIZE(AVI)];
+	bool feature_complete;
+	bool devcap_read;
+	bool sink_detected;
 	struct edid *edid;
 	unsigned int gen2_write_burst:1;
 	enum sii8620_mt_state mt_state;
@@ -476,7 +482,7 @@ static void sii8620_update_array(u8 *dst, u8 *src, int count)
 	}
 }
 
-static void sii8620_sink_detected(struct sii8620 *ctx, int ret)
+static void sii8620_identify_sink(struct sii8620 *ctx)
 {
 	static const char * const sink_str[] = {
 		[SINK_NONE] = "NONE",
@@ -487,7 +493,7 @@ static void sii8620_sink_detected(struct sii8620 *ctx, int ret)
 	char sink_name[20];
 	struct device *dev = ctx->dev;
 
-	if (ret < 0)
+	if (!ctx->sink_detected || !ctx->devcap_read)
 		return;
 
 	sii8620_fetch_edid(ctx);
@@ -496,6 +502,7 @@ static void sii8620_sink_detected(struct sii8620 *ctx, int ret)
 		sii8620_mhl_disconnected(ctx);
 		return;
 	}
+	sii8620_set_upstream_edid(ctx);
 
 	if (drm_detect_hdmi_monitor(ctx->edid))
 		ctx->sink_type = SINK_HDMI;
@@ -508,53 +515,6 @@ static void sii8620_sink_detected(struct sii8620 *ctx, int ret)
 		 sink_str[ctx->sink_type], sink_name);
 }
 
-static void sii8620_hsic_init(struct sii8620 *ctx)
-{
-	if (!sii8620_is_mhl3(ctx))
-		return;
-
-	sii8620_write(ctx, REG_FCGC,
-		BIT_FCGC_HSIC_HOSTMODE | BIT_FCGC_HSIC_ENABLE);
-	sii8620_setbits(ctx, REG_HRXCTRL3,
-		BIT_HRXCTRL3_HRX_STAY_RESET | BIT_HRXCTRL3_STATUS_EN, ~0);
-	sii8620_setbits(ctx, REG_TTXNUMB, MSK_TTXNUMB_TTX_NUMBPS, 4);
-	sii8620_setbits(ctx, REG_TRXCTRL, BIT_TRXCTRL_TRX_FROM_SE_COC, ~0);
-	sii8620_setbits(ctx, REG_HTXCTRL, BIT_HTXCTRL_HTX_DRVCONN1, 0);
-	sii8620_setbits(ctx, REG_KEEPER, MSK_KEEPER_MODE, VAL_KEEPER_MODE_HOST);
-	sii8620_write_seq_static(ctx,
-		REG_TDMLLCTL, 0,
-		REG_UTSRST, BIT_UTSRST_HRX_SRST | BIT_UTSRST_HTX_SRST |
-			BIT_UTSRST_KEEPER_SRST | BIT_UTSRST_FC_SRST,
-		REG_UTSRST, BIT_UTSRST_HRX_SRST | BIT_UTSRST_HTX_SRST,
-		REG_HRXINTL, 0xff,
-		REG_HRXINTH, 0xff,
-		REG_TTXINTL, 0xff,
-		REG_TTXINTH, 0xff,
-		REG_TRXINTL, 0xff,
-		REG_TRXINTH, 0xff,
-		REG_HTXINTL, 0xff,
-		REG_HTXINTH, 0xff,
-		REG_FCINTR0, 0xff,
-		REG_FCINTR1, 0xff,
-		REG_FCINTR2, 0xff,
-		REG_FCINTR3, 0xff,
-		REG_FCINTR4, 0xff,
-		REG_FCINTR5, 0xff,
-		REG_FCINTR6, 0xff,
-		REG_FCINTR7, 0xff
-	);
-}
-
-static void sii8620_edid_read(struct sii8620 *ctx, int ret)
-{
-	if (ret < 0)
-		return;
-
-	sii8620_set_upstream_edid(ctx);
-	sii8620_hsic_init(ctx);
-	sii8620_enable_hpd(ctx);
-}
-
 static void sii8620_mr_devcap(struct sii8620 *ctx)
 {
 	u8 dcap[MHL_DCAP_SIZE];
@@ -570,6 +530,8 @@ static void sii8620_mr_devcap(struct sii8620 *ctx)
 		 dcap[MHL_DCAP_ADOPTER_ID_H], dcap[MHL_DCAP_ADOPTER_ID_L],
 		 dcap[MHL_DCAP_DEVICE_ID_H], dcap[MHL_DCAP_DEVICE_ID_L]);
 	sii8620_update_array(ctx->devcap, dcap, MHL_DCAP_SIZE);
+	ctx->devcap_read = true;
+	sii8620_identify_sink(ctx);
 }
 
 static void sii8620_mr_xdevcap(struct sii8620 *ctx)
@@ -807,6 +769,7 @@ static void sii8620_burst_rx_all(struct sii8620 *ctx)
 static void sii8620_fetch_edid(struct sii8620 *ctx)
 {
 	u8 lm_ddc, ddc_cmd, int3, cbus;
+	unsigned long timeout;
 	int fetched, i;
 	int edid_len = EDID_LENGTH;
 	u8 *edid;
@@ -856,23 +819,31 @@ static void sii8620_fetch_edid(struct sii8620 *ctx)
 			REG_DDC_CMD, ddc_cmd | VAL_DDC_CMD_ENH_DDC_READ_NO_ACK
 		);
 
-		do {
-			int3 = sii8620_readb(ctx, REG_INTR3);
+		int3 = 0;
+		timeout = jiffies + msecs_to_jiffies(200);
+		for (;;) {
 			cbus = sii8620_readb(ctx, REG_CBUS_STATUS);
-
-			if (int3 & BIT_DDC_CMD_DONE)
-				break;
-
-			if (!(cbus & BIT_CBUS_STATUS_CBUS_CONNECTED)) {
+			if (~cbus & BIT_CBUS_STATUS_CBUS_CONNECTED) {
+				kfree(edid);
+				edid = NULL;
+				goto end;
+			}
+			if (int3 & BIT_DDC_CMD_DONE) {
+				if (sii8620_readb(ctx, REG_DDC_DOUT_CNT)
+				    >= FETCH_SIZE)
+					break;
+			} else {
+				int3 = sii8620_readb(ctx, REG_INTR3);
+			}
+			if (time_is_before_jiffies(timeout)) {
+				ctx->error = -ETIMEDOUT;
+				dev_err(ctx->dev, "timeout during EDID read\n");
 				kfree(edid);
 				edid = NULL;
 				goto end;
 			}
-		} while (1);
-
-		sii8620_readb(ctx, REG_DDC_STATUS);
-		while (sii8620_readb(ctx, REG_DDC_DOUT_CNT) < FETCH_SIZE)
 			usleep_range(10, 20);
+		}
 
 		sii8620_read_buf(ctx, REG_DDC_DATA, edid + fetched, FETCH_SIZE);
 		if (fetched + FETCH_SIZE == EDID_LENGTH) {
@@ -971,8 +942,17 @@ static int sii8620_hw_on(struct sii8620 *ctx)
 	ret = regulator_bulk_enable(ARRAY_SIZE(ctx->supplies), ctx->supplies);
 	if (ret)
 		return ret;
+
 	usleep_range(10000, 20000);
-	return clk_prepare_enable(ctx->clk_xtal);
+	ret = clk_prepare_enable(ctx->clk_xtal);
+	if (ret)
+		return ret;
+
+	msleep(100);
+	gpiod_set_value(ctx->gpio_reset, 0);
+	msleep(100);
+
+	return 0;
 }
 
 static int sii8620_hw_off(struct sii8620 *ctx)
@@ -982,17 +962,6 @@ static int sii8620_hw_off(struct sii8620 *ctx)
 	return regulator_bulk_disable(ARRAY_SIZE(ctx->supplies), ctx->supplies);
 }
 
-static void sii8620_hw_reset(struct sii8620 *ctx)
-{
-	usleep_range(10000, 20000);
-	gpiod_set_value(ctx->gpio_reset, 0);
-	usleep_range(5000, 20000);
-	gpiod_set_value(ctx->gpio_reset, 1);
-	usleep_range(10000, 20000);
-	gpiod_set_value(ctx->gpio_reset, 0);
-	msleep(300);
-}
-
 static void sii8620_cbus_reset(struct sii8620 *ctx)
 {
 	sii8620_write(ctx, REG_PWD_SRST, BIT_PWD_SRST_CBUS_RST
@@ -1048,20 +1017,11 @@ static void sii8620_stop_video(struct sii8620 *ctx)
 
 static void sii8620_set_format(struct sii8620 *ctx)
 {
-	u8 out_fmt;
-
 	if (sii8620_is_mhl3(ctx)) {
 		sii8620_setbits(ctx, REG_M3_P0CTRL,
 				BIT_M3_P0CTRL_MHL3_P0_PIXEL_MODE_PACKED,
 				ctx->use_packed_pixel ? ~0 : 0);
 	} else {
-		if (ctx->use_packed_pixel)
-			sii8620_write_seq_static(ctx,
-				REG_VID_MODE, BIT_VID_MODE_M1080P,
-				REG_MHL_TOP_CTL, BIT_MHL_TOP_CTL_MHL_PP_SEL | 1,
-				REG_MHLTX_CTL6, 0x60
-			);
-		else
 			sii8620_write_seq_static(ctx,
 				REG_VID_MODE, 0,
 				REG_MHL_TOP_CTL, 1,
@@ -1069,15 +1029,9 @@ static void sii8620_set_format(struct sii8620 *ctx)
 			);
 	}
 
-	if (ctx->use_packed_pixel)
-		out_fmt = VAL_TPI_FORMAT(YCBCR422, FULL) |
-			BIT_TPI_OUTPUT_CSCMODE709;
-	else
-		out_fmt = VAL_TPI_FORMAT(RGB, FULL);
-
 	sii8620_write_seq(ctx,
 		REG_TPI_INPUT, VAL_TPI_FORMAT(RGB, FULL),
-		REG_TPI_OUTPUT, out_fmt,
+		REG_TPI_OUTPUT, VAL_TPI_FORMAT(RGB, FULL),
 	);
 }
 
@@ -1216,7 +1170,7 @@ static void sii8620_start_video(struct sii8620 *ctx)
 		int clk = ctx->pixel_clock * (ctx->use_packed_pixel ? 2 : 3);
 		int i;
 
-		for (i = 0; i < ARRAY_SIZE(clk_spec); ++i)
+		for (i = 0; i < ARRAY_SIZE(clk_spec) - 1; ++i)
 			if (clk < clk_spec[i].max_clk)
 				break;
 
@@ -1534,6 +1488,16 @@ static void sii8620_set_mode(struct sii8620 *ctx, enum sii8620_mode mode)
 	);
 }
 
+static void sii8620_hpd_unplugged(struct sii8620 *ctx)
+{
+	sii8620_disable_hpd(ctx);
+	ctx->sink_type = SINK_NONE;
+	ctx->sink_detected = false;
+	ctx->feature_complete = false;
+	kfree(ctx->edid);
+	ctx->edid = NULL;
+}
+
 static void sii8620_disconnect(struct sii8620 *ctx)
 {
 	sii8620_disable_gen2_write_burst(ctx);
@@ -1561,7 +1525,7 @@ static void sii8620_disconnect(struct sii8620 *ctx)
 		REG_MHL_DP_CTL6, 0x2A,
 		REG_MHL_DP_CTL7, 0x03
 	);
-	sii8620_disable_hpd(ctx);
+	sii8620_hpd_unplugged(ctx);
 	sii8620_write_seq_static(ctx,
 		REG_M3_CTRL, VAL_M3_CTRL_MHL3_VALUE,
 		REG_MHL_COC_CTL1, 0x07,
@@ -1609,10 +1573,8 @@ static void sii8620_disconnect(struct sii8620 *ctx)
 	memset(ctx->xstat, 0, sizeof(ctx->xstat));
 	memset(ctx->devcap, 0, sizeof(ctx->devcap));
 	memset(ctx->xdevcap, 0, sizeof(ctx->xdevcap));
+	ctx->devcap_read = false;
 	ctx->cbus_status = 0;
-	ctx->sink_type = SINK_NONE;
-	kfree(ctx->edid);
-	ctx->edid = NULL;
 	sii8620_mt_cleanup(ctx);
 }
 
@@ -1703,9 +1665,6 @@ static void sii8620_status_changed_path(struct sii8620 *ctx)
 		sii8620_mt_write_stat(ctx, MHL_DST_REG(LINK_MODE),
 				      MHL_DST_LM_CLK_MODE_NORMAL
 				      | MHL_DST_LM_PATH_ENABLED);
-		if (!sii8620_is_mhl3(ctx))
-			sii8620_mt_read_devcap(ctx, false);
-		sii8620_mt_set_cont(ctx, sii8620_sink_detected);
 	} else {
 		sii8620_mt_write_stat(ctx, MHL_DST_REG(LINK_MODE),
 				      MHL_DST_LM_CLK_MODE_NORMAL);
@@ -1722,9 +1681,14 @@ static void sii8620_msc_mr_write_stat(struct sii8620 *ctx)
 	sii8620_update_array(ctx->stat, st, MHL_DST_SIZE);
 	sii8620_update_array(ctx->xstat, xst, MHL_XDS_SIZE);
 
-	if (ctx->stat[MHL_DST_CONNECTED_RDY] & MHL_DST_CONN_DCAP_RDY)
+	if (ctx->stat[MHL_DST_CONNECTED_RDY] & st[MHL_DST_CONNECTED_RDY] &
+	    MHL_DST_CONN_DCAP_RDY) {
 		sii8620_status_dcap_ready(ctx);
 
+		if (!sii8620_is_mhl3(ctx))
+			sii8620_mt_read_devcap(ctx, false);
+	}
+
 	if (st[MHL_DST_LINK_MODE] & MHL_DST_LM_PATH_ENABLED)
 		sii8620_status_changed_path(ctx);
 }
@@ -1808,8 +1772,11 @@ static void sii8620_msc_mr_set_int(struct sii8620 *ctx)
 	}
 	if (ints[MHL_INT_RCHANGE] & MHL_INT_RC_FEAT_REQ)
 		sii8620_send_features(ctx);
-	if (ints[MHL_INT_RCHANGE] & MHL_INT_RC_FEAT_COMPLETE)
-		sii8620_edid_read(ctx, 0);
+	if (ints[MHL_INT_RCHANGE] & MHL_INT_RC_FEAT_COMPLETE) {
+		ctx->feature_complete = true;
+		if (ctx->edid)
+			sii8620_enable_hpd(ctx);
+	}
 }
 
 static struct sii8620_mt_msg *sii8620_msc_msg_first(struct sii8620 *ctx)
@@ -1884,6 +1851,15 @@ static void sii8620_irq_msc(struct sii8620 *ctx)
 	if (stat & BIT_CBUS_MSC_MR_WRITE_STAT)
 		sii8620_msc_mr_write_stat(ctx);
 
+	if (stat & BIT_CBUS_HPD_CHG) {
+		if (ctx->cbus_status & BIT_CBUS_STATUS_CBUS_HPD) {
+			ctx->sink_detected = true;
+			sii8620_identify_sink(ctx);
+		} else {
+			sii8620_hpd_unplugged(ctx);
+		}
+	}
+
 	if (stat & BIT_CBUS_MSC_MR_SET_INT)
 		sii8620_msc_mr_set_int(ctx);
 
@@ -1931,14 +1907,6 @@ static void sii8620_irq_edid(struct sii8620 *ctx)
 		ctx->mt_state = MT_STATE_DONE;
 }
 
-static void sii8620_scdt_high(struct sii8620 *ctx)
-{
-	sii8620_write_seq_static(ctx,
-		REG_INTR8_MASK, BIT_CEA_NEW_AVI | BIT_CEA_NEW_VSI,
-		REG_TPI_SC, BIT_TPI_SC_TPI_OUTPUT_MODE_0_HDMI,
-	);
-}
-
 static void sii8620_irq_scdt(struct sii8620 *ctx)
 {
 	u8 stat = sii8620_readb(ctx, REG_INTR5);
@@ -1946,53 +1914,13 @@ static void sii8620_irq_scdt(struct sii8620 *ctx)
 	if (stat & BIT_INTR_SCDT_CHANGE) {
 		u8 cstat = sii8620_readb(ctx, REG_TMDS_CSTAT_P3);
 
-		if (cstat & BIT_TMDS_CSTAT_P3_SCDT) {
-			if (ctx->sink_type == SINK_HDMI)
-				/* enable infoframe interrupt */
-				sii8620_scdt_high(ctx);
-			else
-				sii8620_start_video(ctx);
-		}
+		if (cstat & BIT_TMDS_CSTAT_P3_SCDT)
+			sii8620_start_video(ctx);
 	}
 
 	sii8620_write(ctx, REG_INTR5, stat);
 }
 
-static void sii8620_new_vsi(struct sii8620 *ctx)
-{
-	u8 vsif[11];
-
-	sii8620_write(ctx, REG_RX_HDMI_CTRL2,
-		      VAL_RX_HDMI_CTRL2_DEFVAL |
-		      BIT_RX_HDMI_CTRL2_VSI_MON_SEL_VSI);
-	sii8620_read_buf(ctx, REG_RX_HDMI_MON_PKT_HEADER1, vsif,
-			 ARRAY_SIZE(vsif));
-}
-
-static void sii8620_new_avi(struct sii8620 *ctx)
-{
-	sii8620_write(ctx, REG_RX_HDMI_CTRL2, VAL_RX_HDMI_CTRL2_DEFVAL);
-	sii8620_read_buf(ctx, REG_RX_HDMI_MON_PKT_HEADER1, ctx->avif,
-			 ARRAY_SIZE(ctx->avif));
-}
-
-static void sii8620_irq_infr(struct sii8620 *ctx)
-{
-	u8 stat = sii8620_readb(ctx, REG_INTR8)
-		& (BIT_CEA_NEW_VSI | BIT_CEA_NEW_AVI);
-
-	sii8620_write(ctx, REG_INTR8, stat);
-
-	if (stat & BIT_CEA_NEW_VSI)
-		sii8620_new_vsi(ctx);
-
-	if (stat & BIT_CEA_NEW_AVI)
-		sii8620_new_avi(ctx);
-
-	if (stat & (BIT_CEA_NEW_VSI | BIT_CEA_NEW_AVI))
-		sii8620_start_video(ctx);
-}
-
 static void sii8620_got_xdevcap(struct sii8620 *ctx, int ret)
 {
 	if (ret < 0)
@@ -2043,11 +1971,11 @@ static void sii8620_irq_ddc(struct sii8620 *ctx)
 
 	if (stat & BIT_DDC_CMD_DONE) {
 		sii8620_write(ctx, REG_INTR3_MASK, 0);
-		if (sii8620_is_mhl3(ctx))
+		if (sii8620_is_mhl3(ctx) && !ctx->feature_complete)
 			sii8620_mt_set_int(ctx, MHL_INT_REG(RCHANGE),
 					   MHL_INT_RC_FEAT_REQ);
 		else
-			sii8620_edid_read(ctx, 0);
+			sii8620_enable_hpd(ctx);
 	}
 	sii8620_write(ctx, REG_INTR3, stat);
 }
@@ -2074,7 +2002,6 @@ static irqreturn_t sii8620_irq_thread(int irq, void *data)
 		{ BIT_FAST_INTR_STAT_EDID, sii8620_irq_edid },
 		{ BIT_FAST_INTR_STAT_DDC, sii8620_irq_ddc },
 		{ BIT_FAST_INTR_STAT_SCDT, sii8620_irq_scdt },
-		{ BIT_FAST_INTR_STAT_INFR, sii8620_irq_infr },
 	};
 	struct sii8620 *ctx = data;
 	u8 stats[LEN_FAST_INTR_STAT];
@@ -2112,7 +2039,6 @@ static void sii8620_cable_in(struct sii8620 *ctx)
 		dev_err(dev, "Error powering on, %d.\n", ret);
 		return;
 	}
-	sii8620_hw_reset(ctx);
 
 	sii8620_read_buf(ctx, REG_VND_IDL, ver, ARRAY_SIZE(ver));
 	ret = sii8620_clear_error(ctx);
@@ -2268,17 +2194,43 @@ static void sii8620_detach(struct drm_bridge *bridge)
 	rc_unregister_device(ctx->rc_dev);
 }
 
+static int sii8620_is_packing_required(struct sii8620 *ctx,
+				       const struct drm_display_mode *mode)
+{
+	int max_pclk, max_pclk_pp_mode;
+
+	if (sii8620_is_mhl3(ctx)) {
+		max_pclk = MHL3_MAX_PCLK;
+		max_pclk_pp_mode = MHL3_MAX_PCLK_PP_MODE;
+	} else {
+		max_pclk = MHL1_MAX_PCLK;
+		max_pclk_pp_mode = MHL1_MAX_PCLK_PP_MODE;
+	}
+
+	if (mode->clock < max_pclk)
+		return 0;
+	else if (mode->clock < max_pclk_pp_mode)
+		return 1;
+	else
+		return -1;
+}
+
 static enum drm_mode_status sii8620_mode_valid(struct drm_bridge *bridge,
 					 const struct drm_display_mode *mode)
 {
 	struct sii8620 *ctx = bridge_to_sii8620(bridge);
+	int pack_required = sii8620_is_packing_required(ctx, mode);
 	bool can_pack = ctx->devcap[MHL_DCAP_VID_LINK_MODE] &
 			MHL_DCAP_VID_LINK_PPIXEL;
-	unsigned int max_pclk = sii8620_is_mhl3(ctx) ? MHL3_MAX_LCLK :
-						       MHL1_MAX_LCLK;
-	max_pclk /= can_pack ? 2 : 3;
 
-	return (mode->clock > max_pclk) ? MODE_CLOCK_HIGH : MODE_OK;
+	switch (pack_required) {
+	case 0:
+		return MODE_OK;
+	case 1:
+		return (can_pack) ? MODE_OK : MODE_CLOCK_HIGH;
+	default:
+		return MODE_CLOCK_HIGH;
+	}
 }
 
 static bool sii8620_mode_fixup(struct drm_bridge *bridge,
@@ -2286,43 +2238,16 @@ static bool sii8620_mode_fixup(struct drm_bridge *bridge,
 			       struct drm_display_mode *adjusted_mode)
 {
 	struct sii8620 *ctx = bridge_to_sii8620(bridge);
-	int max_lclk;
-	bool ret = true;
 
 	mutex_lock(&ctx->lock);
 
-	max_lclk = sii8620_is_mhl3(ctx) ? MHL3_MAX_LCLK : MHL1_MAX_LCLK;
-	if (max_lclk > 3 * adjusted_mode->clock) {
-		ctx->use_packed_pixel = 0;
-		goto end;
-	}
-	if ((ctx->devcap[MHL_DCAP_VID_LINK_MODE] & MHL_DCAP_VID_LINK_PPIXEL) &&
-	    max_lclk > 2 * adjusted_mode->clock) {
-		ctx->use_packed_pixel = 1;
-		goto end;
-	}
-	ret = false;
-end:
-	if (ret) {
-		u8 vic = drm_match_cea_mode(adjusted_mode);
-
-		if (!vic) {
-			union hdmi_infoframe frm;
-			u8 mhl_vic[] = { 0, 95, 94, 93, 98 };
-
-			/* FIXME: We need the connector here */
-			drm_hdmi_vendor_infoframe_from_display_mode(
-				&frm.vendor.hdmi, NULL, adjusted_mode);
-			vic = frm.vendor.hdmi.vic;
-			if (vic >= ARRAY_SIZE(mhl_vic))
-				vic = 0;
-			vic = mhl_vic[vic];
-		}
-		ctx->video_code = vic;
-		ctx->pixel_clock = adjusted_mode->clock;
-	}
+	ctx->use_packed_pixel = sii8620_is_packing_required(ctx, adjusted_mode);
+	ctx->video_code = drm_match_cea_mode(adjusted_mode);
+	ctx->pixel_clock = adjusted_mode->clock;
+
 	mutex_unlock(&ctx->lock);
-	return ret;
+
+	return true;
 }
 
 static const struct drm_bridge_funcs sii8620_bridge_funcs = {
diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
index ec8d0006ef7c..5971976284bf 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
+++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
@@ -1922,7 +1922,7 @@ static int dw_hdmi_connector_get_modes(struct drm_connector *connector)
 
 		hdmi->sink_is_hdmi = drm_detect_hdmi_monitor(edid);
 		hdmi->sink_has_audio = drm_detect_monitor_audio(edid);
-		drm_mode_connector_update_edid_property(connector, edid);
+		drm_connector_update_edid_property(connector, edid);
 		cec_notifier_set_phys_addr_from_edid(hdmi->cec_notifier, edid);
 		ret = drm_add_edid_modes(connector, edid);
 		kfree(edid);
@@ -1974,7 +1974,7 @@ static int dw_hdmi_bridge_attach(struct drm_bridge *bridge)
 	drm_connector_init(bridge->dev, connector, &dw_hdmi_connector_funcs,
 			   DRM_MODE_CONNECTOR_HDMIA);
 
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	return 0;
 }
@@ -2077,7 +2077,7 @@ static irqreturn_t dw_hdmi_hardirq(int irq, void *dev_id)
 	return ret;
 }
 
-void __dw_hdmi_setup_rx_sense(struct dw_hdmi *hdmi, bool hpd, bool rx_sense)
+void dw_hdmi_setup_rx_sense(struct dw_hdmi *hdmi, bool hpd, bool rx_sense)
 {
 	mutex_lock(&hdmi->mutex);
 
@@ -2103,13 +2103,6 @@ void __dw_hdmi_setup_rx_sense(struct dw_hdmi *hdmi, bool hpd, bool rx_sense)
 	}
 	mutex_unlock(&hdmi->mutex);
 }
-
-void dw_hdmi_setup_rx_sense(struct device *dev, bool hpd, bool rx_sense)
-{
-	struct dw_hdmi *hdmi = dev_get_drvdata(dev);
-
-	__dw_hdmi_setup_rx_sense(hdmi, hpd, rx_sense);
-}
 EXPORT_SYMBOL_GPL(dw_hdmi_setup_rx_sense);
 
 static irqreturn_t dw_hdmi_irq(int irq, void *dev_id)
@@ -2145,9 +2138,9 @@ static irqreturn_t dw_hdmi_irq(int irq, void *dev_id)
 	 */
 	if (intr_stat &
 	    (HDMI_IH_PHY_STAT0_RX_SENSE | HDMI_IH_PHY_STAT0_HPD)) {
-		__dw_hdmi_setup_rx_sense(hdmi,
-					 phy_stat & HDMI_PHY_HPD,
-					 phy_stat & HDMI_PHY_RX_SENSE);
+		dw_hdmi_setup_rx_sense(hdmi,
+				       phy_stat & HDMI_PHY_HPD,
+				       phy_stat & HDMI_PHY_RX_SENSE);
 
 		if ((phy_stat & (HDMI_PHY_RX_SENSE | HDMI_PHY_HPD)) == 0)
 			cec_notifier_set_phys_addr(hdmi->cec_notifier,
diff --git a/drivers/gpu/drm/bridge/tc358767.c b/drivers/gpu/drm/bridge/tc358767.c
index 0fd9cf27542c..8e28e738cb52 100644
--- a/drivers/gpu/drm/bridge/tc358767.c
+++ b/drivers/gpu/drm/bridge/tc358767.c
@@ -1140,7 +1140,7 @@ static int tc_connector_get_modes(struct drm_connector *connector)
 	if (!edid)
 		return 0;
 
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 	count = drm_add_edid_modes(connector, edid);
 
 	return count;
@@ -1195,7 +1195,7 @@ static int tc_bridge_attach(struct drm_bridge *bridge)
 
 	drm_display_info_set_bus_formats(&tc->connector.display_info,
 					 &bus_format, 1);
-	drm_mode_connector_attach_encoder(&tc->connector, tc->bridge.encoder);
+	drm_connector_attach_encoder(&tc->connector, tc->bridge.encoder);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/bridge/ti-tfp410.c b/drivers/gpu/drm/bridge/ti-tfp410.c
index acb857030951..c3e32138c6bb 100644
--- a/drivers/gpu/drm/bridge/ti-tfp410.c
+++ b/drivers/gpu/drm/bridge/ti-tfp410.c
@@ -62,7 +62,7 @@ static int tfp410_get_modes(struct drm_connector *connector)
 		goto fallback;
 	}
 
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 
 	return drm_add_edid_modes(connector, edid);
 fallback:
@@ -132,7 +132,7 @@ static int tfp410_attach(struct drm_bridge *bridge)
 		return ret;
 	}
 
-	drm_mode_connector_attach_encoder(&dvi->connector,
+	drm_connector_attach_encoder(&dvi->connector,
 					  bridge->encoder);
 
 	return 0;
diff --git a/drivers/gpu/drm/cirrus/cirrus_drv.h b/drivers/gpu/drm/cirrus/cirrus_drv.h
index be2d7e488062..ce9db7aab225 100644
--- a/drivers/gpu/drm/cirrus/cirrus_drv.h
+++ b/drivers/gpu/drm/cirrus/cirrus_drv.h
@@ -92,7 +92,6 @@
 
 #define to_cirrus_crtc(x) container_of(x, struct cirrus_crtc, base)
 #define to_cirrus_encoder(x) container_of(x, struct cirrus_encoder, base)
-#define to_cirrus_framebuffer(x) container_of(x, struct cirrus_framebuffer, base)
 
 struct cirrus_crtc {
 	struct drm_crtc			base;
@@ -117,11 +116,6 @@ struct cirrus_connector {
 	struct drm_connector		base;
 };
 
-struct cirrus_framebuffer {
-	struct drm_framebuffer		base;
-	struct drm_gem_object *obj;
-};
-
 struct cirrus_mc {
 	resource_size_t			vram_size;
 	resource_size_t			vram_base;
@@ -152,7 +146,7 @@ struct cirrus_device {
 
 struct cirrus_fbdev {
 	struct drm_fb_helper helper;
-	struct cirrus_framebuffer gfb;
+	struct drm_framebuffer gfb;
 	void *sysram;
 	int size;
 	int x1, y1, x2, y2; /* dirty rect */
@@ -198,7 +192,7 @@ int cirrus_dumb_create(struct drm_file *file,
 		       struct drm_mode_create_dumb *args);
 
 int cirrus_framebuffer_init(struct drm_device *dev,
-			   struct cirrus_framebuffer *gfb,
+			    struct drm_framebuffer *gfb,
 			    const struct drm_mode_fb_cmd2 *mode_cmd,
 			    struct drm_gem_object *obj);
 
diff --git a/drivers/gpu/drm/cirrus/cirrus_fbdev.c b/drivers/gpu/drm/cirrus/cirrus_fbdev.c
index 32fbfba2c623..b643ac92801c 100644
--- a/drivers/gpu/drm/cirrus/cirrus_fbdev.c
+++ b/drivers/gpu/drm/cirrus/cirrus_fbdev.c
@@ -22,14 +22,14 @@ static void cirrus_dirty_update(struct cirrus_fbdev *afbdev,
 	struct drm_gem_object *obj;
 	struct cirrus_bo *bo;
 	int src_offset, dst_offset;
-	int bpp = afbdev->gfb.base.format->cpp[0];
+	int bpp = afbdev->gfb.format->cpp[0];
 	int ret = -EBUSY;
 	bool unmap = false;
 	bool store_for_later = false;
 	int x2, y2;
 	unsigned long flags;
 
-	obj = afbdev->gfb.obj;
+	obj = afbdev->gfb.obj[0];
 	bo = gem_to_cirrus_bo(obj);
 
 	/*
@@ -82,7 +82,7 @@ static void cirrus_dirty_update(struct cirrus_fbdev *afbdev,
 	}
 	for (i = y; i < y + height; i++) {
 		/* assume equal stride for now */
-		src_offset = dst_offset = i * afbdev->gfb.base.pitches[0] + (x * bpp);
+		src_offset = dst_offset = i * afbdev->gfb.pitches[0] + (x * bpp);
 		memcpy_toio(bo->kmap.virtual + src_offset, afbdev->sysram + src_offset, width * bpp);
 
 	}
@@ -204,7 +204,7 @@ static int cirrusfb_create(struct drm_fb_helper *helper,
 	gfbdev->sysram = sysram;
 	gfbdev->size = size;
 
-	fb = &gfbdev->gfb.base;
+	fb = &gfbdev->gfb;
 	if (!fb) {
 		DRM_INFO("fb is NULL\n");
 		return -EINVAL;
@@ -246,19 +246,19 @@ static int cirrusfb_create(struct drm_fb_helper *helper,
 static int cirrus_fbdev_destroy(struct drm_device *dev,
 				struct cirrus_fbdev *gfbdev)
 {
-	struct cirrus_framebuffer *gfb = &gfbdev->gfb;
+	struct drm_framebuffer *gfb = &gfbdev->gfb;
 
 	drm_fb_helper_unregister_fbi(&gfbdev->helper);
 
-	if (gfb->obj) {
-		drm_gem_object_put_unlocked(gfb->obj);
-		gfb->obj = NULL;
+	if (gfb->obj[0]) {
+		drm_gem_object_put_unlocked(gfb->obj[0]);
+		gfb->obj[0] = NULL;
 	}
 
 	vfree(gfbdev->sysram);
 	drm_fb_helper_fini(&gfbdev->helper);
-	drm_framebuffer_unregister_private(&gfb->base);
-	drm_framebuffer_cleanup(&gfb->base);
+	drm_framebuffer_unregister_private(gfb);
+	drm_framebuffer_cleanup(gfb);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/cirrus/cirrus_main.c b/drivers/gpu/drm/cirrus/cirrus_main.c
index 26df1e8cd490..60d54e10a34d 100644
--- a/drivers/gpu/drm/cirrus/cirrus_main.c
+++ b/drivers/gpu/drm/cirrus/cirrus_main.c
@@ -10,42 +10,25 @@
  */
 #include <drm/drmP.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 
 #include "cirrus_drv.h"
 
-static int cirrus_create_handle(struct drm_framebuffer *fb,
-				struct drm_file* file_priv,
-				unsigned int* handle)
-{
-	struct cirrus_framebuffer *cirrus_fb = to_cirrus_framebuffer(fb);
-
-	return drm_gem_handle_create(file_priv, cirrus_fb->obj, handle);
-}
-
-static void cirrus_user_framebuffer_destroy(struct drm_framebuffer *fb)
-{
-	struct cirrus_framebuffer *cirrus_fb = to_cirrus_framebuffer(fb);
-
-	drm_gem_object_put_unlocked(cirrus_fb->obj);
-	drm_framebuffer_cleanup(fb);
-	kfree(fb);
-}
-
 static const struct drm_framebuffer_funcs cirrus_fb_funcs = {
-	.create_handle = cirrus_create_handle,
-	.destroy = cirrus_user_framebuffer_destroy,
+	.create_handle = drm_gem_fb_create_handle,
+	.destroy = drm_gem_fb_destroy,
 };
 
 int cirrus_framebuffer_init(struct drm_device *dev,
-			    struct cirrus_framebuffer *gfb,
+			    struct drm_framebuffer *gfb,
 			    const struct drm_mode_fb_cmd2 *mode_cmd,
 			    struct drm_gem_object *obj)
 {
 	int ret;
 
-	drm_helper_mode_fill_fb_struct(dev, &gfb->base, mode_cmd);
-	gfb->obj = obj;
-	ret = drm_framebuffer_init(dev, &gfb->base, &cirrus_fb_funcs);
+	drm_helper_mode_fill_fb_struct(dev, gfb, mode_cmd);
+	gfb->obj[0] = obj;
+	ret = drm_framebuffer_init(dev, gfb, &cirrus_fb_funcs);
 	if (ret) {
 		DRM_ERROR("drm_framebuffer_init failed: %d\n", ret);
 		return ret;
@@ -60,7 +43,7 @@ cirrus_user_framebuffer_create(struct drm_device *dev,
 {
 	struct cirrus_device *cdev = dev->dev_private;
 	struct drm_gem_object *obj;
-	struct cirrus_framebuffer *cirrus_fb;
+	struct drm_framebuffer *fb;
 	u32 bpp;
 	int ret;
 
@@ -74,19 +57,19 @@ cirrus_user_framebuffer_create(struct drm_device *dev,
 	if (obj == NULL)
 		return ERR_PTR(-ENOENT);
 
-	cirrus_fb = kzalloc(sizeof(*cirrus_fb), GFP_KERNEL);
-	if (!cirrus_fb) {
+	fb = kzalloc(sizeof(*fb), GFP_KERNEL);
+	if (!fb) {
 		drm_gem_object_put_unlocked(obj);
 		return ERR_PTR(-ENOMEM);
 	}
 
-	ret = cirrus_framebuffer_init(dev, cirrus_fb, mode_cmd, obj);
+	ret = cirrus_framebuffer_init(dev, fb, mode_cmd, obj);
 	if (ret) {
 		drm_gem_object_put_unlocked(obj);
-		kfree(cirrus_fb);
+		kfree(fb);
 		return ERR_PTR(ret);
 	}
-	return &cirrus_fb->base;
+	return fb;
 }
 
 static const struct drm_mode_config_funcs cirrus_mode_funcs = {
diff --git a/drivers/gpu/drm/cirrus/cirrus_mode.c b/drivers/gpu/drm/cirrus/cirrus_mode.c
index c91b9b054e3f..336bfda40125 100644
--- a/drivers/gpu/drm/cirrus/cirrus_mode.c
+++ b/drivers/gpu/drm/cirrus/cirrus_mode.c
@@ -101,17 +101,13 @@ static int cirrus_crtc_do_set_base(struct drm_crtc *crtc,
 				int x, int y, int atomic)
 {
 	struct cirrus_device *cdev = crtc->dev->dev_private;
-	struct drm_gem_object *obj;
-	struct cirrus_framebuffer *cirrus_fb;
 	struct cirrus_bo *bo;
 	int ret;
 	u64 gpu_addr;
 
 	/* push the previous fb to system ram */
 	if (!atomic && fb) {
-		cirrus_fb = to_cirrus_framebuffer(fb);
-		obj = cirrus_fb->obj;
-		bo = gem_to_cirrus_bo(obj);
+		bo = gem_to_cirrus_bo(fb->obj[0]);
 		ret = cirrus_bo_reserve(bo, false);
 		if (ret)
 			return ret;
@@ -119,9 +115,7 @@ static int cirrus_crtc_do_set_base(struct drm_crtc *crtc,
 		cirrus_bo_unreserve(bo);
 	}
 
-	cirrus_fb = to_cirrus_framebuffer(crtc->primary->fb);
-	obj = cirrus_fb->obj;
-	bo = gem_to_cirrus_bo(obj);
+	bo = gem_to_cirrus_bo(crtc->primary->fb->obj[0]);
 
 	ret = cirrus_bo_reserve(bo, false);
 	if (ret)
@@ -133,7 +127,7 @@ static int cirrus_crtc_do_set_base(struct drm_crtc *crtc,
 		return ret;
 	}
 
-	if (&cdev->mode_info.gfbdev->gfb == cirrus_fb) {
+	if (&cdev->mode_info.gfbdev->gfb == crtc->primary->fb) {
 		/* if pushing console in kmap it */
 		ret = ttm_bo_kmap(&bo->bo, 0, bo->bo.num_pages, &bo->kmap);
 		if (ret)
@@ -536,7 +530,7 @@ int cirrus_modeset_init(struct cirrus_device *cdev)
 		return -1;
 	}
 
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	ret = cirrus_fbdev_init(cdev);
 	if (ret) {
diff --git a/drivers/gpu/drm/drm_atomic.c b/drivers/gpu/drm/drm_atomic.c
index 3c181d217f8b..3eb061e11e2e 100644
--- a/drivers/gpu/drm/drm_atomic.c
+++ b/drivers/gpu/drm/drm_atomic.c
@@ -30,6 +30,7 @@
 #include <drm/drm_atomic.h>
 #include <drm/drm_mode.h>
 #include <drm/drm_print.h>
+#include <drm/drm_writeback.h>
 #include <linux/sync_file.h>
 
 #include "drm_crtc_internal.h"
@@ -155,6 +156,8 @@ void drm_atomic_state_default_clear(struct drm_atomic_state *state)
 						       state->connectors[i].state);
 		state->connectors[i].ptr = NULL;
 		state->connectors[i].state = NULL;
+		state->connectors[i].old_state = NULL;
+		state->connectors[i].new_state = NULL;
 		drm_connector_put(connector);
 	}
 
@@ -169,6 +172,8 @@ void drm_atomic_state_default_clear(struct drm_atomic_state *state)
 
 		state->crtcs[i].ptr = NULL;
 		state->crtcs[i].state = NULL;
+		state->crtcs[i].old_state = NULL;
+		state->crtcs[i].new_state = NULL;
 	}
 
 	for (i = 0; i < config->num_total_plane; i++) {
@@ -181,6 +186,8 @@ void drm_atomic_state_default_clear(struct drm_atomic_state *state)
 						   state->planes[i].state);
 		state->planes[i].ptr = NULL;
 		state->planes[i].state = NULL;
+		state->planes[i].old_state = NULL;
+		state->planes[i].new_state = NULL;
 	}
 
 	for (i = 0; i < state->num_private_objs; i++) {
@@ -190,6 +197,8 @@ void drm_atomic_state_default_clear(struct drm_atomic_state *state)
 						 state->private_objs[i].state);
 		state->private_objs[i].ptr = NULL;
 		state->private_objs[i].state = NULL;
+		state->private_objs[i].old_state = NULL;
+		state->private_objs[i].new_state = NULL;
 	}
 	state->num_private_objs = 0;
 
@@ -317,6 +326,35 @@ static s32 __user *get_out_fence_for_crtc(struct drm_atomic_state *state,
 	return fence_ptr;
 }
 
+static int set_out_fence_for_connector(struct drm_atomic_state *state,
+					struct drm_connector *connector,
+					s32 __user *fence_ptr)
+{
+	unsigned int index = drm_connector_index(connector);
+
+	if (!fence_ptr)
+		return 0;
+
+	if (put_user(-1, fence_ptr))
+		return -EFAULT;
+
+	state->connectors[index].out_fence_ptr = fence_ptr;
+
+	return 0;
+}
+
+static s32 __user *get_out_fence_for_connector(struct drm_atomic_state *state,
+					       struct drm_connector *connector)
+{
+	unsigned int index = drm_connector_index(connector);
+	s32 __user *fence_ptr;
+
+	fence_ptr = state->connectors[index].out_fence_ptr;
+	state->connectors[index].out_fence_ptr = NULL;
+
+	return fence_ptr;
+}
+
 /**
  * drm_atomic_set_mode_for_crtc - set mode for CRTC
  * @state: the CRTC whose incoming state to update
@@ -331,6 +369,7 @@ static s32 __user *get_out_fence_for_crtc(struct drm_atomic_state *state,
 int drm_atomic_set_mode_for_crtc(struct drm_crtc_state *state,
 				 const struct drm_display_mode *mode)
 {
+	struct drm_crtc *crtc = state->crtc;
 	struct drm_mode_modeinfo umode;
 
 	/* Early return for no change. */
@@ -351,13 +390,13 @@ int drm_atomic_set_mode_for_crtc(struct drm_crtc_state *state,
 
 		drm_mode_copy(&state->mode, mode);
 		state->enable = true;
-		DRM_DEBUG_ATOMIC("Set [MODE:%s] for CRTC state %p\n",
-				 mode->name, state);
+		DRM_DEBUG_ATOMIC("Set [MODE:%s] for [CRTC:%d:%s] state %p\n",
+				 mode->name, crtc->base.id, crtc->name, state);
 	} else {
 		memset(&state->mode, 0, sizeof(state->mode));
 		state->enable = false;
-		DRM_DEBUG_ATOMIC("Set [NOMODE] for CRTC state %p\n",
-				 state);
+		DRM_DEBUG_ATOMIC("Set [NOMODE] for [CRTC:%d:%s] state %p\n",
+				 crtc->base.id, crtc->name, state);
 	}
 
 	return 0;
@@ -380,6 +419,8 @@ EXPORT_SYMBOL(drm_atomic_set_mode_for_crtc);
 int drm_atomic_set_mode_prop_for_crtc(struct drm_crtc_state *state,
                                       struct drm_property_blob *blob)
 {
+	struct drm_crtc *crtc = state->crtc;
+
 	if (blob == state->mode_blob)
 		return 0;
 
@@ -389,19 +430,34 @@ int drm_atomic_set_mode_prop_for_crtc(struct drm_crtc_state *state,
 	memset(&state->mode, 0, sizeof(state->mode));
 
 	if (blob) {
-		if (blob->length != sizeof(struct drm_mode_modeinfo) ||
-		    drm_mode_convert_umode(state->crtc->dev, &state->mode,
-					   blob->data))
+		int ret;
+
+		if (blob->length != sizeof(struct drm_mode_modeinfo)) {
+			DRM_DEBUG_ATOMIC("[CRTC:%d:%s] bad mode blob length: %zu\n",
+					 crtc->base.id, crtc->name,
+					 blob->length);
 			return -EINVAL;
+		}
+
+		ret = drm_mode_convert_umode(crtc->dev,
+					     &state->mode, blob->data);
+		if (ret) {
+			DRM_DEBUG_ATOMIC("[CRTC:%d:%s] invalid mode (ret=%d, status=%s):\n",
+					 crtc->base.id, crtc->name,
+					 ret, drm_get_mode_status_name(state->mode.status));
+			drm_mode_debug_printmodeline(&state->mode);
+			return -EINVAL;
+		}
 
 		state->mode_blob = drm_property_blob_get(blob);
 		state->enable = true;
-		DRM_DEBUG_ATOMIC("Set [MODE:%s] for CRTC state %p\n",
-				 state->mode.name, state);
+		DRM_DEBUG_ATOMIC("Set [MODE:%s] for [CRTC:%d:%s] state %p\n",
+				 state->mode.name, crtc->base.id, crtc->name,
+				 state);
 	} else {
 		state->enable = false;
-		DRM_DEBUG_ATOMIC("Set [NOMODE] for CRTC state %p\n",
-				 state);
+		DRM_DEBUG_ATOMIC("Set [NOMODE] for [CRTC:%d:%s] state %p\n",
+				 crtc->base.id, crtc->name, state);
 	}
 
 	return 0;
@@ -531,10 +587,14 @@ int drm_atomic_crtc_set_property(struct drm_crtc *crtc,
 			return -EFAULT;
 
 		set_out_fence_for_crtc(state->state, crtc, fence_ptr);
-	} else if (crtc->funcs->atomic_set_property)
+	} else if (crtc->funcs->atomic_set_property) {
 		return crtc->funcs->atomic_set_property(crtc, state, property, val);
-	else
+	} else {
+		DRM_DEBUG_ATOMIC("[CRTC:%d:%s] unknown property [PROP:%d:%s]]\n",
+				 crtc->base.id, crtc->name,
+				 property->base.id, property->name);
 		return -EINVAL;
+	}
 
 	return 0;
 }
@@ -669,6 +729,51 @@ static void drm_atomic_crtc_print_state(struct drm_printer *p,
 }
 
 /**
+ * drm_atomic_connector_check - check connector state
+ * @connector: connector to check
+ * @state: connector state to check
+ *
+ * Provides core sanity checks for connector state.
+ *
+ * RETURNS:
+ * Zero on success, error code on failure
+ */
+static int drm_atomic_connector_check(struct drm_connector *connector,
+		struct drm_connector_state *state)
+{
+	struct drm_crtc_state *crtc_state;
+	struct drm_writeback_job *writeback_job = state->writeback_job;
+
+	if ((connector->connector_type != DRM_MODE_CONNECTOR_WRITEBACK) || !writeback_job)
+		return 0;
+
+	if (writeback_job->fb && !state->crtc) {
+		DRM_DEBUG_ATOMIC("[CONNECTOR:%d:%s] framebuffer without CRTC\n",
+				 connector->base.id, connector->name);
+		return -EINVAL;
+	}
+
+	if (state->crtc)
+		crtc_state = drm_atomic_get_existing_crtc_state(state->state,
+								state->crtc);
+
+	if (writeback_job->fb && !crtc_state->active) {
+		DRM_DEBUG_ATOMIC("[CONNECTOR:%d:%s] has framebuffer, but [CRTC:%d] is off\n",
+				 connector->base.id, connector->name,
+				 state->crtc->base.id);
+		return -EINVAL;
+	}
+
+	if (writeback_job->out_fence && !writeback_job->fb) {
+		DRM_DEBUG_ATOMIC("[CONNECTOR:%d:%s] requesting out-fence without framebuffer\n",
+				 connector->base.id, connector->name);
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+/**
  * drm_atomic_get_plane_state - get plane state
  * @state: global atomic state object
  * @plane: plane to get state object for
@@ -692,6 +797,11 @@ drm_atomic_get_plane_state(struct drm_atomic_state *state,
 
 	WARN_ON(!state->acquire_ctx);
 
+	/* the legacy pointers should never be set */
+	WARN_ON(plane->fb);
+	WARN_ON(plane->old_fb);
+	WARN_ON(plane->crtc);
+
 	plane_state = drm_atomic_get_existing_plane_state(state, plane);
 	if (plane_state)
 		return plane_state;
@@ -786,8 +896,11 @@ static int drm_atomic_plane_set_property(struct drm_plane *plane,
 	} else if (property == plane->alpha_property) {
 		state->alpha = val;
 	} else if (property == plane->rotation_property) {
-		if (!is_power_of_2(val & DRM_MODE_ROTATE_MASK))
+		if (!is_power_of_2(val & DRM_MODE_ROTATE_MASK)) {
+			DRM_DEBUG_ATOMIC("[PLANE:%d:%s] bad rotation bitmask: 0x%llx\n",
+					 plane->base.id, plane->name, val);
 			return -EINVAL;
+		}
 		state->rotation = val;
 	} else if (property == plane->zpos_property) {
 		state->zpos = val;
@@ -799,6 +912,9 @@ static int drm_atomic_plane_set_property(struct drm_plane *plane,
 		return plane->funcs->atomic_set_property(plane, state,
 				property, val);
 	} else {
+		DRM_DEBUG_ATOMIC("[PLANE:%d:%s] unknown property [PROP:%d:%s]]\n",
+				 plane->base.id, plane->name,
+				 property->base.id, property->name);
 		return -EINVAL;
 	}
 
@@ -906,10 +1022,12 @@ static int drm_atomic_plane_check(struct drm_plane *plane,
 
 	/* either *both* CRTC and FB must be set, or neither */
 	if (state->crtc && !state->fb) {
-		DRM_DEBUG_ATOMIC("CRTC set but no FB\n");
+		DRM_DEBUG_ATOMIC("[PLANE:%d:%s] CRTC set but no FB\n",
+				 plane->base.id, plane->name);
 		return -EINVAL;
 	} else if (state->fb && !state->crtc) {
-		DRM_DEBUG_ATOMIC("FB set but no CRTC\n");
+		DRM_DEBUG_ATOMIC("[PLANE:%d:%s] FB set but no CRTC\n",
+				 plane->base.id, plane->name);
 		return -EINVAL;
 	}
 
@@ -919,7 +1037,9 @@ static int drm_atomic_plane_check(struct drm_plane *plane,
 
 	/* Check whether this plane is usable on this CRTC */
 	if (!(plane->possible_crtcs & drm_crtc_mask(state->crtc))) {
-		DRM_DEBUG_ATOMIC("Invalid crtc for plane\n");
+		DRM_DEBUG_ATOMIC("Invalid [CRTC:%d:%s] for [PLANE:%d:%s]\n",
+				 state->crtc->base.id, state->crtc->name,
+				 plane->base.id, plane->name);
 		return -EINVAL;
 	}
 
@@ -928,7 +1048,8 @@ static int drm_atomic_plane_check(struct drm_plane *plane,
 					   state->fb->modifier);
 	if (ret) {
 		struct drm_format_name_buf format_name;
-		DRM_DEBUG_ATOMIC("Invalid pixel format %s, modifier 0x%llx\n",
+		DRM_DEBUG_ATOMIC("[PLANE:%d:%s] invalid pixel format %s, modifier 0x%llx\n",
+				 plane->base.id, plane->name,
 				 drm_get_format_name(state->fb->format->format,
 						     &format_name),
 				 state->fb->modifier);
@@ -940,7 +1061,8 @@ static int drm_atomic_plane_check(struct drm_plane *plane,
 	    state->crtc_x > INT_MAX - (int32_t) state->crtc_w ||
 	    state->crtc_h > INT_MAX ||
 	    state->crtc_y > INT_MAX - (int32_t) state->crtc_h) {
-		DRM_DEBUG_ATOMIC("Invalid CRTC coordinates %ux%u+%d+%d\n",
+		DRM_DEBUG_ATOMIC("[PLANE:%d:%s] invalid CRTC coordinates %ux%u+%d+%d\n",
+				 plane->base.id, plane->name,
 				 state->crtc_w, state->crtc_h,
 				 state->crtc_x, state->crtc_y);
 		return -ERANGE;
@@ -954,8 +1076,9 @@ static int drm_atomic_plane_check(struct drm_plane *plane,
 	    state->src_x > fb_width - state->src_w ||
 	    state->src_h > fb_height ||
 	    state->src_y > fb_height - state->src_h) {
-		DRM_DEBUG_ATOMIC("Invalid source coordinates "
+		DRM_DEBUG_ATOMIC("[PLANE:%d:%s] invalid source coordinates "
 				 "%u.%06ux%u.%06u+%u.%06u+%u.%06u (fb %ux%u)\n",
+				 plane->base.id, plane->name,
 				 state->src_w >> 16, ((state->src_w & 0xffff) * 15625) >> 10,
 				 state->src_h >> 16, ((state->src_h & 0xffff) * 15625) >> 10,
 				 state->src_x >> 16, ((state->src_x & 0xffff) * 15625) >> 10,
@@ -988,6 +1111,7 @@ static void drm_atomic_plane_print_state(struct drm_printer *p,
 	drm_printf(p, "\tcrtc-pos=" DRM_RECT_FMT "\n", DRM_RECT_ARG(&dest));
 	drm_printf(p, "\tsrc-pos=" DRM_RECT_FP_FMT "\n", DRM_RECT_FP_ARG(&src));
 	drm_printf(p, "\trotation=%x\n", state->rotation);
+	drm_printf(p, "\tnormalized-zpos=%x\n", state->normalized_zpos);
 	drm_printf(p, "\tcolor-encoding=%s\n",
 		   drm_get_color_encoding_name(state->color_encoding));
 	drm_printf(p, "\tcolor-range=%s\n",
@@ -1112,6 +1236,7 @@ drm_atomic_get_private_obj_state(struct drm_atomic_state *state,
 	state->private_objs[index].old_state = obj->state;
 	state->private_objs[index].new_state = obj_state;
 	state->private_objs[index].ptr = obj;
+	obj_state->state = state;
 
 	state->num_private_objs = num_objs;
 
@@ -1270,6 +1395,8 @@ static int drm_atomic_connector_set_property(struct drm_connector *connector,
 			state->link_status = val;
 	} else if (property == config->aspect_ratio_property) {
 		state->picture_aspect_ratio = val;
+	} else if (property == config->content_type_property) {
+		state->content_type = val;
 	} else if (property == connector->scaling_mode_property) {
 		state->scaling_mode = val;
 	} else if (property == connector->content_protection_property) {
@@ -1278,10 +1405,24 @@ static int drm_atomic_connector_set_property(struct drm_connector *connector,
 			return -EINVAL;
 		}
 		state->content_protection = val;
+	} else if (property == config->writeback_fb_id_property) {
+		struct drm_framebuffer *fb = drm_framebuffer_lookup(dev, NULL, val);
+		int ret = drm_atomic_set_writeback_fb_for_connector(state, fb);
+		if (fb)
+			drm_framebuffer_put(fb);
+		return ret;
+	} else if (property == config->writeback_out_fence_ptr_property) {
+		s32 __user *fence_ptr = u64_to_user_ptr(val);
+
+		return set_out_fence_for_connector(state->state, connector,
+						   fence_ptr);
 	} else if (connector->funcs->atomic_set_property) {
 		return connector->funcs->atomic_set_property(connector,
 				state, property, val);
 	} else {
+		DRM_DEBUG_ATOMIC("[CONNECTOR:%d:%s] unknown property [PROP:%d:%s]]\n",
+				 connector->base.id, connector->name,
+				 property->base.id, property->name);
 		return -EINVAL;
 	}
 
@@ -1296,6 +1437,10 @@ static void drm_atomic_connector_print_state(struct drm_printer *p,
 	drm_printf(p, "connector[%u]: %s\n", connector->base.id, connector->name);
 	drm_printf(p, "\tcrtc=%s\n", state->crtc ? state->crtc->name : "(null)");
 
+	if (connector->connector_type == DRM_MODE_CONNECTOR_WRITEBACK)
+		if (state->writeback_job && state->writeback_job->fb)
+			drm_printf(p, "\tfb=%d\n", state->writeback_job->fb->base.id);
+
 	if (connector->funcs->atomic_print_state)
 		connector->funcs->atomic_print_state(p, state);
 }
@@ -1355,10 +1500,17 @@ drm_atomic_connector_get_property(struct drm_connector *connector,
 		*val = state->link_status;
 	} else if (property == config->aspect_ratio_property) {
 		*val = state->picture_aspect_ratio;
+	} else if (property == config->content_type_property) {
+		*val = state->content_type;
 	} else if (property == connector->scaling_mode_property) {
 		*val = state->scaling_mode;
 	} else if (property == connector->content_protection_property) {
 		*val = state->content_protection;
+	} else if (property == config->writeback_fb_id_property) {
+		/* Writeback framebuffer is one-shot, write and forget */
+		*val = 0;
+	} else if (property == config->writeback_out_fence_ptr_property) {
+		*val = 0;
 	} else if (connector->funcs->atomic_get_property) {
 		return connector->funcs->atomic_get_property(connector,
 				state, property, val);
@@ -1425,14 +1577,16 @@ drm_atomic_set_crtc_for_plane(struct drm_plane_state *plane_state,
 {
 	struct drm_plane *plane = plane_state->plane;
 	struct drm_crtc_state *crtc_state;
-
+	/* Nothing to do for same crtc*/
+	if (plane_state->crtc == crtc)
+		return 0;
 	if (plane_state->crtc) {
 		crtc_state = drm_atomic_get_crtc_state(plane_state->state,
 						       plane_state->crtc);
 		if (WARN_ON(IS_ERR(crtc_state)))
 			return PTR_ERR(crtc_state);
 
-		crtc_state->plane_mask &= ~(1 << drm_plane_index(plane));
+		crtc_state->plane_mask &= ~drm_plane_mask(plane);
 	}
 
 	plane_state->crtc = crtc;
@@ -1442,15 +1596,16 @@ drm_atomic_set_crtc_for_plane(struct drm_plane_state *plane_state,
 						       crtc);
 		if (IS_ERR(crtc_state))
 			return PTR_ERR(crtc_state);
-		crtc_state->plane_mask |= (1 << drm_plane_index(plane));
+		crtc_state->plane_mask |= drm_plane_mask(plane);
 	}
 
 	if (crtc)
-		DRM_DEBUG_ATOMIC("Link plane state %p to [CRTC:%d:%s]\n",
-				 plane_state, crtc->base.id, crtc->name);
+		DRM_DEBUG_ATOMIC("Link [PLANE:%d:%s] state %p to [CRTC:%d:%s]\n",
+				 plane->base.id, plane->name, plane_state,
+				 crtc->base.id, crtc->name);
 	else
-		DRM_DEBUG_ATOMIC("Link plane state %p to [NOCRTC]\n",
-				 plane_state);
+		DRM_DEBUG_ATOMIC("Link [PLANE:%d:%s] state %p to [NOCRTC]\n",
+				 plane->base.id, plane->name, plane_state);
 
 	return 0;
 }
@@ -1470,12 +1625,15 @@ void
 drm_atomic_set_fb_for_plane(struct drm_plane_state *plane_state,
 			    struct drm_framebuffer *fb)
 {
+	struct drm_plane *plane = plane_state->plane;
+
 	if (fb)
-		DRM_DEBUG_ATOMIC("Set [FB:%d] for plane state %p\n",
-				 fb->base.id, plane_state);
-	else
-		DRM_DEBUG_ATOMIC("Set [NOFB] for plane state %p\n",
+		DRM_DEBUG_ATOMIC("Set [FB:%d] for [PLANE:%d:%s] state %p\n",
+				 fb->base.id, plane->base.id, plane->name,
 				 plane_state);
+	else
+		DRM_DEBUG_ATOMIC("Set [NOFB] for [PLANE:%d:%s] state %p\n",
+				 plane->base.id, plane->name, plane_state);
 
 	drm_framebuffer_assign(&plane_state->fb, fb);
 }
@@ -1536,6 +1694,7 @@ int
 drm_atomic_set_crtc_for_connector(struct drm_connector_state *conn_state,
 				  struct drm_crtc *crtc)
 {
+	struct drm_connector *connector = conn_state->connector;
 	struct drm_crtc_state *crtc_state;
 
 	if (conn_state->crtc == crtc)
@@ -1546,7 +1705,7 @@ drm_atomic_set_crtc_for_connector(struct drm_connector_state *conn_state,
 							   conn_state->crtc);
 
 		crtc_state->connector_mask &=
-			~(1 << drm_connector_index(conn_state->connector));
+			~drm_connector_mask(conn_state->connector);
 
 		drm_connector_put(conn_state->connector);
 		conn_state->crtc = NULL;
@@ -1558,15 +1717,17 @@ drm_atomic_set_crtc_for_connector(struct drm_connector_state *conn_state,
 			return PTR_ERR(crtc_state);
 
 		crtc_state->connector_mask |=
-			1 << drm_connector_index(conn_state->connector);
+			drm_connector_mask(conn_state->connector);
 
 		drm_connector_get(conn_state->connector);
 		conn_state->crtc = crtc;
 
-		DRM_DEBUG_ATOMIC("Link connector state %p to [CRTC:%d:%s]\n",
+		DRM_DEBUG_ATOMIC("Link [CONNECTOR:%d:%s] state %p to [CRTC:%d:%s]\n",
+				 connector->base.id, connector->name,
 				 conn_state, crtc->base.id, crtc->name);
 	} else {
-		DRM_DEBUG_ATOMIC("Link connector state %p to [NOCRTC]\n",
+		DRM_DEBUG_ATOMIC("Link [CONNECTOR:%d:%s] state %p to [NOCRTC]\n",
+				 connector->base.id, connector->name,
 				 conn_state);
 	}
 
@@ -1574,6 +1735,70 @@ drm_atomic_set_crtc_for_connector(struct drm_connector_state *conn_state,
 }
 EXPORT_SYMBOL(drm_atomic_set_crtc_for_connector);
 
+/*
+ * drm_atomic_get_writeback_job - return or allocate a writeback job
+ * @conn_state: Connector state to get the job for
+ *
+ * Writeback jobs have a different lifetime to the atomic state they are
+ * associated with. This convenience function takes care of allocating a job
+ * if there isn't yet one associated with the connector state, otherwise
+ * it just returns the existing job.
+ *
+ * Returns: The writeback job for the given connector state
+ */
+static struct drm_writeback_job *
+drm_atomic_get_writeback_job(struct drm_connector_state *conn_state)
+{
+	WARN_ON(conn_state->connector->connector_type != DRM_MODE_CONNECTOR_WRITEBACK);
+
+	if (!conn_state->writeback_job)
+		conn_state->writeback_job =
+			kzalloc(sizeof(*conn_state->writeback_job), GFP_KERNEL);
+
+	return conn_state->writeback_job;
+}
+
+/**
+ * drm_atomic_set_writeback_fb_for_connector - set writeback framebuffer
+ * @conn_state: atomic state object for the connector
+ * @fb: fb to use for the connector
+ *
+ * This is used to set the framebuffer for a writeback connector, which outputs
+ * to a buffer instead of an actual physical connector.
+ * Changing the assigned framebuffer requires us to grab a reference to the new
+ * fb and drop the reference to the old fb, if there is one. This function
+ * takes care of all these details besides updating the pointer in the
+ * state object itself.
+ *
+ * Note: The only way conn_state can already have an fb set is if the commit
+ * sets the property more than once.
+ *
+ * See also: drm_writeback_connector_init()
+ *
+ * Returns: 0 on success
+ */
+int drm_atomic_set_writeback_fb_for_connector(
+		struct drm_connector_state *conn_state,
+		struct drm_framebuffer *fb)
+{
+	struct drm_writeback_job *job =
+		drm_atomic_get_writeback_job(conn_state);
+	if (!job)
+		return -ENOMEM;
+
+	drm_framebuffer_assign(&job->fb, fb);
+
+	if (fb)
+		DRM_DEBUG_ATOMIC("Set [FB:%d] for connector state %p\n",
+				 fb->base.id, conn_state);
+	else
+		DRM_DEBUG_ATOMIC("Set [NOFB] for connector state %p\n",
+				 conn_state);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_atomic_set_writeback_fb_for_connector);
+
 /**
  * drm_atomic_add_affected_connectors - add connectors for crtc
  * @state: atomic state
@@ -1619,7 +1844,7 @@ drm_atomic_add_affected_connectors(struct drm_atomic_state *state,
 	 */
 	drm_connector_list_iter_begin(state->dev, &conn_iter);
 	drm_for_each_connector_iter(connector, &conn_iter) {
-		if (!(crtc_state->connector_mask & (1 << drm_connector_index(connector))))
+		if (!(crtc_state->connector_mask & drm_connector_mask(connector)))
 			continue;
 
 		conn_state = drm_atomic_get_connector_state(state, connector);
@@ -1662,6 +1887,9 @@ drm_atomic_add_affected_planes(struct drm_atomic_state *state,
 
 	WARN_ON(!drm_atomic_get_new_crtc_state(state, crtc));
 
+	DRM_DEBUG_ATOMIC("Adding all current planes for [CRTC:%d:%s] to %p\n",
+			 crtc->base.id, crtc->name, state);
+
 	drm_for_each_plane_mask(plane, state->dev, crtc->state->plane_mask) {
 		struct drm_plane_state *plane_state =
 			drm_atomic_get_plane_state(state, plane);
@@ -1692,6 +1920,8 @@ int drm_atomic_check_only(struct drm_atomic_state *state)
 	struct drm_plane_state *plane_state;
 	struct drm_crtc *crtc;
 	struct drm_crtc_state *crtc_state;
+	struct drm_connector *conn;
+	struct drm_connector_state *conn_state;
 	int i, ret = 0;
 
 	DRM_DEBUG_ATOMIC("checking %p\n", state);
@@ -1714,6 +1944,15 @@ int drm_atomic_check_only(struct drm_atomic_state *state)
 		}
 	}
 
+	for_each_new_connector_in_state(state, conn, conn_state, i) {
+		ret = drm_atomic_connector_check(conn, conn_state);
+		if (ret) {
+			DRM_DEBUG_ATOMIC("[CONNECTOR:%d:%s] atomic core check failed\n",
+					 conn->base.id, conn->name);
+			return ret;
+		}
+	}
+
 	if (config->funcs->atomic_check) {
 		ret = config->funcs->atomic_check(state->dev, state);
 
@@ -2038,45 +2277,6 @@ int drm_atomic_set_property(struct drm_atomic_state *state,
 }
 
 /**
- * drm_atomic_clean_old_fb -- Unset old_fb pointers and set plane->fb pointers.
- *
- * @dev: drm device to check.
- * @plane_mask: plane mask for planes that were updated.
- * @ret: return value, can be -EDEADLK for a retry.
- *
- * Before doing an update &drm_plane.old_fb is set to &drm_plane.fb, but before
- * dropping the locks old_fb needs to be set to NULL and plane->fb updated. This
- * is a common operation for each atomic update, so this call is split off as a
- * helper.
- */
-void drm_atomic_clean_old_fb(struct drm_device *dev,
-			     unsigned plane_mask,
-			     int ret)
-{
-	struct drm_plane *plane;
-
-	/* if succeeded, fixup legacy plane crtc/fb ptrs before dropping
-	 * locks (ie. while it is still safe to deref plane->state).  We
-	 * need to do this here because the driver entry points cannot
-	 * distinguish between legacy and atomic ioctls.
-	 */
-	drm_for_each_plane_mask(plane, dev, plane_mask) {
-		if (ret == 0) {
-			struct drm_framebuffer *new_fb = plane->state->fb;
-			if (new_fb)
-				drm_framebuffer_get(new_fb);
-			plane->fb = new_fb;
-			plane->crtc = plane->state->crtc;
-
-			if (plane->old_fb)
-				drm_framebuffer_put(plane->old_fb);
-		}
-		plane->old_fb = NULL;
-	}
-}
-EXPORT_SYMBOL(drm_atomic_clean_old_fb);
-
-/**
  * DOC: explicit fencing properties
  *
  * Explicit fencing allows userspace to control the buffer synchronization
@@ -2151,7 +2351,7 @@ static int setup_out_fence(struct drm_out_fence_state *fence_state,
 	return 0;
 }
 
-static int prepare_crtc_signaling(struct drm_device *dev,
+static int prepare_signaling(struct drm_device *dev,
 				  struct drm_atomic_state *state,
 				  struct drm_mode_atomic *arg,
 				  struct drm_file *file_priv,
@@ -2160,6 +2360,8 @@ static int prepare_crtc_signaling(struct drm_device *dev,
 {
 	struct drm_crtc *crtc;
 	struct drm_crtc_state *crtc_state;
+	struct drm_connector *conn;
+	struct drm_connector_state *conn_state;
 	int i, c = 0, ret;
 
 	if (arg->flags & DRM_MODE_ATOMIC_TEST_ONLY)
@@ -2225,6 +2427,45 @@ static int prepare_crtc_signaling(struct drm_device *dev,
 		c++;
 	}
 
+	for_each_new_connector_in_state(state, conn, conn_state, i) {
+		struct drm_writeback_connector *wb_conn;
+		struct drm_writeback_job *job;
+		struct drm_out_fence_state *f;
+		struct dma_fence *fence;
+		s32 __user *fence_ptr;
+
+		fence_ptr = get_out_fence_for_connector(state, conn);
+		if (!fence_ptr)
+			continue;
+
+		job = drm_atomic_get_writeback_job(conn_state);
+		if (!job)
+			return -ENOMEM;
+
+		f = krealloc(*fence_state, sizeof(**fence_state) *
+			     (*num_fences + 1), GFP_KERNEL);
+		if (!f)
+			return -ENOMEM;
+
+		memset(&f[*num_fences], 0, sizeof(*f));
+
+		f[*num_fences].out_fence_ptr = fence_ptr;
+		*fence_state = f;
+
+		wb_conn = drm_connector_to_writeback(conn);
+		fence = drm_writeback_get_out_fence(wb_conn);
+		if (!fence)
+			return -ENOMEM;
+
+		ret = setup_out_fence(&f[(*num_fences)++], fence);
+		if (ret) {
+			dma_fence_put(fence);
+			return ret;
+		}
+
+		job->out_fence = fence;
+	}
+
 	/*
 	 * Having this flag means user mode pends on event which will never
 	 * reach due to lack of at least one CRTC for signaling
@@ -2235,11 +2476,11 @@ static int prepare_crtc_signaling(struct drm_device *dev,
 	return 0;
 }
 
-static void complete_crtc_signaling(struct drm_device *dev,
-				    struct drm_atomic_state *state,
-				    struct drm_out_fence_state *fence_state,
-				    unsigned int num_fences,
-				    bool install_fds)
+static void complete_signaling(struct drm_device *dev,
+			       struct drm_atomic_state *state,
+			       struct drm_out_fence_state *fence_state,
+			       unsigned int num_fences,
+			       bool install_fds)
 {
 	struct drm_crtc *crtc;
 	struct drm_crtc_state *crtc_state;
@@ -2296,9 +2537,7 @@ int drm_mode_atomic_ioctl(struct drm_device *dev,
 	unsigned int copied_objs, copied_props;
 	struct drm_atomic_state *state;
 	struct drm_modeset_acquire_ctx ctx;
-	struct drm_plane *plane;
 	struct drm_out_fence_state *fence_state;
-	unsigned plane_mask;
 	int ret = 0;
 	unsigned int i, j, num_fences;
 
@@ -2338,7 +2577,6 @@ int drm_mode_atomic_ioctl(struct drm_device *dev,
 	state->allow_modeset = !!(arg->flags & DRM_MODE_ATOMIC_ALLOW_MODESET);
 
 retry:
-	plane_mask = 0;
 	copied_objs = 0;
 	copied_props = 0;
 	fence_state = NULL;
@@ -2409,17 +2647,11 @@ retry:
 			copied_props++;
 		}
 
-		if (obj->type == DRM_MODE_OBJECT_PLANE && count_props &&
-		    !(arg->flags & DRM_MODE_ATOMIC_TEST_ONLY)) {
-			plane = obj_to_plane(obj);
-			plane_mask |= (1 << drm_plane_index(plane));
-			plane->old_fb = plane->fb;
-		}
 		drm_mode_object_put(obj);
 	}
 
-	ret = prepare_crtc_signaling(dev, state, arg, file_priv, &fence_state,
-				     &num_fences);
+	ret = prepare_signaling(dev, state, arg, file_priv, &fence_state,
+				&num_fences);
 	if (ret)
 		goto out;
 
@@ -2435,9 +2667,7 @@ retry:
 	}
 
 out:
-	drm_atomic_clean_old_fb(dev, plane_mask, ret);
-
-	complete_crtc_signaling(dev, state, fence_state, num_fences, !ret);
+	complete_signaling(dev, state, fence_state, num_fences, !ret);
 
 	if (ret == -EDEADLK) {
 		drm_atomic_state_clear(state);
diff --git a/drivers/gpu/drm/drm_atomic_helper.c b/drivers/gpu/drm/drm_atomic_helper.c
index 9cb2209f6fc8..866a2cc72ef6 100644
--- a/drivers/gpu/drm/drm_atomic_helper.c
+++ b/drivers/gpu/drm/drm_atomic_helper.c
@@ -30,6 +30,7 @@
 #include <drm/drm_plane_helper.h>
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_writeback.h>
 #include <linux/dma-fence.h>
 
 #include "drm_crtc_helper_internal.h"
@@ -120,7 +121,7 @@ static int handle_conflicting_encoders(struct drm_atomic_state *state,
 			new_encoder = drm_atomic_helper_best_encoder(connector);
 
 		if (new_encoder) {
-			if (encoder_mask & (1 << drm_encoder_index(new_encoder))) {
+			if (encoder_mask & drm_encoder_mask(new_encoder)) {
 				DRM_DEBUG_ATOMIC("[ENCODER:%d:%s] on [CONNECTOR:%d:%s] already assigned\n",
 					new_encoder->base.id, new_encoder->name,
 					connector->base.id, connector->name);
@@ -128,7 +129,7 @@ static int handle_conflicting_encoders(struct drm_atomic_state *state,
 				return -EINVAL;
 			}
 
-			encoder_mask |= 1 << drm_encoder_index(new_encoder);
+			encoder_mask |= drm_encoder_mask(new_encoder);
 		}
 	}
 
@@ -154,7 +155,7 @@ static int handle_conflicting_encoders(struct drm_atomic_state *state,
 			continue;
 
 		encoder = connector->state->best_encoder;
-		if (!encoder || !(encoder_mask & (1 << drm_encoder_index(encoder))))
+		if (!encoder || !(encoder_mask & drm_encoder_mask(encoder)))
 			continue;
 
 		if (!disable_conflicting_encoders) {
@@ -222,7 +223,7 @@ set_best_encoder(struct drm_atomic_state *state,
 			crtc_state = drm_atomic_get_new_crtc_state(state, crtc);
 
 			crtc_state->encoder_mask &=
-				~(1 << drm_encoder_index(conn_state->best_encoder));
+				~drm_encoder_mask(conn_state->best_encoder);
 		}
 	}
 
@@ -233,7 +234,7 @@ set_best_encoder(struct drm_atomic_state *state,
 			crtc_state = drm_atomic_get_new_crtc_state(state, crtc);
 
 			crtc_state->encoder_mask |=
-				1 << drm_encoder_index(encoder);
+				drm_encoder_mask(encoder);
 		}
 	}
 
@@ -644,7 +645,7 @@ drm_atomic_helper_check_modeset(struct drm_device *dev,
 		if (ret)
 			return ret;
 
-		connectors_mask += BIT(i);
+		connectors_mask |= BIT(i);
 	}
 
 	/*
@@ -766,7 +767,7 @@ int drm_atomic_helper_check_plane_state(struct drm_plane_state *plane_state,
 	if (crtc_state->enable)
 		drm_mode_get_hv_timing(&crtc_state->mode, &clip.x2, &clip.y2);
 
-	plane_state->visible = drm_rect_clip_scaled(src, dst, &clip, hscale, vscale);
+	plane_state->visible = drm_rect_clip_scaled(src, dst, &clip);
 
 	drm_rect_rotate_inv(src, fb->width << 16, fb->height << 16, rotation);
 
@@ -1172,6 +1173,27 @@ void drm_atomic_helper_commit_modeset_disables(struct drm_device *dev,
 }
 EXPORT_SYMBOL(drm_atomic_helper_commit_modeset_disables);
 
+static void drm_atomic_helper_commit_writebacks(struct drm_device *dev,
+						struct drm_atomic_state *old_state)
+{
+	struct drm_connector *connector;
+	struct drm_connector_state *new_conn_state;
+	int i;
+
+	for_each_new_connector_in_state(old_state, connector, new_conn_state, i) {
+		const struct drm_connector_helper_funcs *funcs;
+
+		funcs = connector->helper_private;
+		if (!funcs->atomic_commit)
+			continue;
+
+		if (new_conn_state->writeback_job && new_conn_state->writeback_job->fb) {
+			WARN_ON(connector->connector_type != DRM_MODE_CONNECTOR_WRITEBACK);
+			funcs->atomic_commit(connector, new_conn_state);
+		}
+	}
+}
+
 /**
  * drm_atomic_helper_commit_modeset_enables - modeset commit to enable outputs
  * @dev: DRM device
@@ -1251,6 +1273,8 @@ void drm_atomic_helper_commit_modeset_enables(struct drm_device *dev,
 
 		drm_bridge_enable(encoder->bridge);
 	}
+
+	drm_atomic_helper_commit_writebacks(dev, old_state);
 }
 EXPORT_SYMBOL(drm_atomic_helper_commit_modeset_enables);
 
@@ -1426,6 +1450,8 @@ void drm_atomic_helper_commit_tail(struct drm_atomic_state *old_state)
 
 	drm_atomic_helper_commit_modeset_enables(dev, old_state);
 
+	drm_atomic_helper_fake_vblank(old_state);
+
 	drm_atomic_helper_commit_hw_done(old_state);
 
 	drm_atomic_helper_wait_for_vblanks(dev, old_state);
@@ -1455,6 +1481,8 @@ void drm_atomic_helper_commit_tail_rpm(struct drm_atomic_state *old_state)
 	drm_atomic_helper_commit_planes(dev, old_state,
 					DRM_PLANE_COMMIT_ACTIVE_ONLY);
 
+	drm_atomic_helper_fake_vblank(old_state);
+
 	drm_atomic_helper_commit_hw_done(old_state);
 
 	drm_atomic_helper_wait_for_vblanks(dev, old_state);
@@ -2030,6 +2058,45 @@ void drm_atomic_helper_wait_for_dependencies(struct drm_atomic_state *old_state)
 EXPORT_SYMBOL(drm_atomic_helper_wait_for_dependencies);
 
 /**
+ * drm_atomic_helper_fake_vblank - fake VBLANK events if needed
+ * @old_state: atomic state object with old state structures
+ *
+ * This function walks all CRTCs and fake VBLANK events on those with
+ * &drm_crtc_state.no_vblank set to true and &drm_crtc_state.event != NULL.
+ * The primary use of this function is writeback connectors working in oneshot
+ * mode and faking VBLANK events. In this case they only fake the VBLANK event
+ * when a job is queued, and any change to the pipeline that does not touch the
+ * connector is leading to timeouts when calling
+ * drm_atomic_helper_wait_for_vblanks() or
+ * drm_atomic_helper_wait_for_flip_done().
+ *
+ * This is part of the atomic helper support for nonblocking commits, see
+ * drm_atomic_helper_setup_commit() for an overview.
+ */
+void drm_atomic_helper_fake_vblank(struct drm_atomic_state *old_state)
+{
+	struct drm_crtc_state *new_crtc_state;
+	struct drm_crtc *crtc;
+	int i;
+
+	for_each_new_crtc_in_state(old_state, crtc, new_crtc_state, i) {
+		unsigned long flags;
+
+		if (!new_crtc_state->no_vblank)
+			continue;
+
+		spin_lock_irqsave(&old_state->dev->event_lock, flags);
+		if (new_crtc_state->event) {
+			drm_crtc_send_vblank_event(crtc,
+						   new_crtc_state->event);
+			new_crtc_state->event = NULL;
+		}
+		spin_unlock_irqrestore(&old_state->dev->event_lock, flags);
+	}
+}
+EXPORT_SYMBOL(drm_atomic_helper_fake_vblank);
+
+/**
  * drm_atomic_helper_commit_hw_done - setup possible nonblocking commit
  * @old_state: atomic state object with old state structures
  *
@@ -2320,11 +2387,13 @@ drm_atomic_helper_commit_planes_on_crtc(struct drm_crtc_state *old_crtc_state)
 	const struct drm_crtc_helper_funcs *crtc_funcs;
 	struct drm_crtc *crtc = old_crtc_state->crtc;
 	struct drm_atomic_state *old_state = old_crtc_state->state;
+	struct drm_crtc_state *new_crtc_state =
+		drm_atomic_get_new_crtc_state(old_state, crtc);
 	struct drm_plane *plane;
 	unsigned plane_mask;
 
 	plane_mask = old_crtc_state->plane_mask;
-	plane_mask |= crtc->state->plane_mask;
+	plane_mask |= new_crtc_state->plane_mask;
 
 	crtc_funcs = crtc->helper_private;
 	if (crtc_funcs && crtc_funcs->atomic_begin)
@@ -2333,6 +2402,8 @@ drm_atomic_helper_commit_planes_on_crtc(struct drm_crtc_state *old_crtc_state)
 	drm_for_each_plane_mask(plane, crtc->dev, plane_mask) {
 		struct drm_plane_state *old_plane_state =
 			drm_atomic_get_old_plane_state(old_state, plane);
+		struct drm_plane_state *new_plane_state =
+			drm_atomic_get_new_plane_state(old_state, plane);
 		const struct drm_plane_helper_funcs *plane_funcs;
 
 		plane_funcs = plane->helper_private;
@@ -2340,13 +2411,14 @@ drm_atomic_helper_commit_planes_on_crtc(struct drm_crtc_state *old_crtc_state)
 		if (!old_plane_state || !plane_funcs)
 			continue;
 
-		WARN_ON(plane->state->crtc && plane->state->crtc != crtc);
+		WARN_ON(new_plane_state->crtc &&
+			new_plane_state->crtc != crtc);
 
-		if (drm_atomic_plane_disabling(old_plane_state, plane->state) &&
+		if (drm_atomic_plane_disabling(old_plane_state, new_plane_state) &&
 		    plane_funcs->atomic_disable)
 			plane_funcs->atomic_disable(plane, old_plane_state);
-		else if (plane->state->crtc ||
-			 drm_atomic_plane_disabling(old_plane_state, plane->state))
+		else if (new_plane_state->crtc ||
+			 drm_atomic_plane_disabling(old_plane_state, new_plane_state))
 			plane_funcs->atomic_update(plane, old_plane_state);
 	}
 
@@ -2795,7 +2867,7 @@ static int update_output_state(struct drm_atomic_state *state,
  * resets the "link-status" property to GOOD, to force any link
  * re-training. The SETCRTC ioctl does not define whether an update does
  * need a full modeset or just a plane update, hence we're allowed to do
- * that. See also drm_mode_connector_set_link_status_property().
+ * that. See also drm_connector_set_link_status_property().
  *
  * Returns:
  * Returns 0 on success, negative errno numbers on failure.
@@ -2914,7 +2986,6 @@ static int __drm_atomic_helper_disable_all(struct drm_device *dev,
 	struct drm_plane *plane;
 	struct drm_crtc_state *crtc_state;
 	struct drm_crtc *crtc;
-	unsigned plane_mask = 0;
 	int ret, i;
 
 	state = drm_atomic_state_alloc(dev);
@@ -2957,17 +3028,10 @@ static int __drm_atomic_helper_disable_all(struct drm_device *dev,
 			goto free;
 
 		drm_atomic_set_fb_for_plane(plane_state, NULL);
-
-		if (clean_old_fbs) {
-			plane->old_fb = plane->fb;
-			plane_mask |= BIT(drm_plane_index(plane));
-		}
 	}
 
 	ret = drm_atomic_commit(state);
 free:
-	if (plane_mask)
-		drm_atomic_clean_old_fb(dev, plane_mask, ret);
 	drm_atomic_state_put(state);
 	return ret;
 }
@@ -3129,13 +3193,8 @@ int drm_atomic_helper_commit_duplicated_state(struct drm_atomic_state *state,
 
 	state->acquire_ctx = ctx;
 
-	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
-		WARN_ON(plane->crtc != new_plane_state->crtc);
-		WARN_ON(plane->fb != new_plane_state->fb);
-		WARN_ON(plane->old_fb);
-
+	for_each_new_plane_in_state(state, plane, new_plane_state, i)
 		state->planes[i].old_state = plane->state;
-	}
 
 	for_each_new_crtc_in_state(state, crtc, new_crtc_state, i)
 		state->crtcs[i].old_state = crtc->state;
@@ -3660,6 +3719,9 @@ __drm_atomic_helper_connector_duplicate_state(struct drm_connector *connector,
 	if (state->crtc)
 		drm_connector_get(connector);
 	state->commit = NULL;
+
+	/* Don't copy over a writeback job, they are used only once */
+	state->writeback_job = NULL;
 }
 EXPORT_SYMBOL(__drm_atomic_helper_connector_duplicate_state);
 
diff --git a/drivers/gpu/drm/drm_client.c b/drivers/gpu/drm/drm_client.c
new file mode 100644
index 000000000000..baff50a4c234
--- /dev/null
+++ b/drivers/gpu/drm/drm_client.c
@@ -0,0 +1,406 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright 2018 Noralf Trønnes
+ */
+
+#include <linux/list.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/seq_file.h>
+#include <linux/slab.h>
+
+#include <drm/drm_client.h>
+#include <drm/drm_debugfs.h>
+#include <drm/drm_device.h>
+#include <drm/drm_drv.h>
+#include <drm/drm_file.h>
+#include <drm/drm_fourcc.h>
+#include <drm/drm_gem.h>
+#include <drm/drm_mode.h>
+#include <drm/drm_print.h>
+#include <drm/drmP.h>
+
+#include "drm_crtc_internal.h"
+#include "drm_internal.h"
+
+/**
+ * DOC: overview
+ *
+ * This library provides support for clients running in the kernel like fbdev and bootsplash.
+ * Currently it's only partially implemented, just enough to support fbdev.
+ *
+ * GEM drivers which provide a GEM based dumb buffer with a virtual address are supported.
+ */
+
+static int drm_client_open(struct drm_client_dev *client)
+{
+	struct drm_device *dev = client->dev;
+	struct drm_file *file;
+
+	file = drm_file_alloc(dev->primary);
+	if (IS_ERR(file))
+		return PTR_ERR(file);
+
+	mutex_lock(&dev->filelist_mutex);
+	list_add(&file->lhead, &dev->filelist_internal);
+	mutex_unlock(&dev->filelist_mutex);
+
+	client->file = file;
+
+	return 0;
+}
+
+static void drm_client_close(struct drm_client_dev *client)
+{
+	struct drm_device *dev = client->dev;
+
+	mutex_lock(&dev->filelist_mutex);
+	list_del(&client->file->lhead);
+	mutex_unlock(&dev->filelist_mutex);
+
+	drm_file_free(client->file);
+}
+EXPORT_SYMBOL(drm_client_close);
+
+/**
+ * drm_client_new - Create a DRM client
+ * @dev: DRM device
+ * @client: DRM client
+ * @name: Client name
+ * @funcs: DRM client functions (optional)
+ *
+ * The caller needs to hold a reference on @dev before calling this function.
+ * The client is freed when the &drm_device is unregistered. See drm_client_release().
+ *
+ * Returns:
+ * Zero on success or negative error code on failure.
+ */
+int drm_client_new(struct drm_device *dev, struct drm_client_dev *client,
+		   const char *name, const struct drm_client_funcs *funcs)
+{
+	int ret;
+
+	if (!drm_core_check_feature(dev, DRIVER_MODESET) ||
+	    !dev->driver->dumb_create || !dev->driver->gem_prime_vmap)
+		return -ENOTSUPP;
+
+	if (funcs && !try_module_get(funcs->owner))
+		return -ENODEV;
+
+	client->dev = dev;
+	client->name = name;
+	client->funcs = funcs;
+
+	ret = drm_client_open(client);
+	if (ret)
+		goto err_put_module;
+
+	mutex_lock(&dev->clientlist_mutex);
+	list_add(&client->list, &dev->clientlist);
+	mutex_unlock(&dev->clientlist_mutex);
+
+	drm_dev_get(dev);
+
+	return 0;
+
+err_put_module:
+	if (funcs)
+		module_put(funcs->owner);
+
+	return ret;
+}
+EXPORT_SYMBOL(drm_client_new);
+
+/**
+ * drm_client_release - Release DRM client resources
+ * @client: DRM client
+ *
+ * Releases resources by closing the &drm_file that was opened by drm_client_new().
+ * It is called automatically if the &drm_client_funcs.unregister callback is _not_ set.
+ *
+ * This function should only be called from the unregister callback. An exception
+ * is fbdev which cannot free the buffer if userspace has open file descriptors.
+ *
+ * Note:
+ * Clients cannot initiate a release by themselves. This is done to keep the code simple.
+ * The driver has to be unloaded before the client can be unloaded.
+ */
+void drm_client_release(struct drm_client_dev *client)
+{
+	struct drm_device *dev = client->dev;
+
+	DRM_DEV_DEBUG_KMS(dev->dev, "%s\n", client->name);
+
+	drm_client_close(client);
+	drm_dev_put(dev);
+	if (client->funcs)
+		module_put(client->funcs->owner);
+}
+EXPORT_SYMBOL(drm_client_release);
+
+void drm_client_dev_unregister(struct drm_device *dev)
+{
+	struct drm_client_dev *client, *tmp;
+
+	if (!drm_core_check_feature(dev, DRIVER_MODESET))
+		return;
+
+	mutex_lock(&dev->clientlist_mutex);
+	list_for_each_entry_safe(client, tmp, &dev->clientlist, list) {
+		list_del(&client->list);
+		if (client->funcs && client->funcs->unregister) {
+			client->funcs->unregister(client);
+		} else {
+			drm_client_release(client);
+			kfree(client);
+		}
+	}
+	mutex_unlock(&dev->clientlist_mutex);
+}
+
+/**
+ * drm_client_dev_hotplug - Send hotplug event to clients
+ * @dev: DRM device
+ *
+ * This function calls the &drm_client_funcs.hotplug callback on the attached clients.
+ *
+ * drm_kms_helper_hotplug_event() calls this function, so drivers that use it
+ * don't need to call this function themselves.
+ */
+void drm_client_dev_hotplug(struct drm_device *dev)
+{
+	struct drm_client_dev *client;
+	int ret;
+
+	if (!drm_core_check_feature(dev, DRIVER_MODESET))
+		return;
+
+	mutex_lock(&dev->clientlist_mutex);
+	list_for_each_entry(client, &dev->clientlist, list) {
+		if (!client->funcs || !client->funcs->hotplug)
+			continue;
+
+		ret = client->funcs->hotplug(client);
+		DRM_DEV_DEBUG_KMS(dev->dev, "%s: ret=%d\n", client->name, ret);
+	}
+	mutex_unlock(&dev->clientlist_mutex);
+}
+EXPORT_SYMBOL(drm_client_dev_hotplug);
+
+void drm_client_dev_restore(struct drm_device *dev)
+{
+	struct drm_client_dev *client;
+	int ret;
+
+	if (!drm_core_check_feature(dev, DRIVER_MODESET))
+		return;
+
+	mutex_lock(&dev->clientlist_mutex);
+	list_for_each_entry(client, &dev->clientlist, list) {
+		if (!client->funcs || !client->funcs->restore)
+			continue;
+
+		ret = client->funcs->restore(client);
+		DRM_DEV_DEBUG_KMS(dev->dev, "%s: ret=%d\n", client->name, ret);
+		if (!ret) /* The first one to return zero gets the privilege to restore */
+			break;
+	}
+	mutex_unlock(&dev->clientlist_mutex);
+}
+
+static void drm_client_buffer_delete(struct drm_client_buffer *buffer)
+{
+	struct drm_device *dev = buffer->client->dev;
+
+	if (buffer->vaddr && dev->driver->gem_prime_vunmap)
+		dev->driver->gem_prime_vunmap(buffer->gem, buffer->vaddr);
+
+	if (buffer->gem)
+		drm_gem_object_put_unlocked(buffer->gem);
+
+	if (buffer->handle)
+		drm_mode_destroy_dumb(dev, buffer->handle, buffer->client->file);
+
+	kfree(buffer);
+}
+
+static struct drm_client_buffer *
+drm_client_buffer_create(struct drm_client_dev *client, u32 width, u32 height, u32 format)
+{
+	struct drm_mode_create_dumb dumb_args = { };
+	struct drm_device *dev = client->dev;
+	struct drm_client_buffer *buffer;
+	struct drm_gem_object *obj;
+	void *vaddr;
+	int ret;
+
+	buffer = kzalloc(sizeof(*buffer), GFP_KERNEL);
+	if (!buffer)
+		return ERR_PTR(-ENOMEM);
+
+	buffer->client = client;
+
+	dumb_args.width = width;
+	dumb_args.height = height;
+	dumb_args.bpp = drm_format_plane_cpp(format, 0) * 8;
+	ret = drm_mode_create_dumb(dev, &dumb_args, client->file);
+	if (ret)
+		goto err_delete;
+
+	buffer->handle = dumb_args.handle;
+	buffer->pitch = dumb_args.pitch;
+
+	obj = drm_gem_object_lookup(client->file, dumb_args.handle);
+	if (!obj)  {
+		ret = -ENOENT;
+		goto err_delete;
+	}
+
+	buffer->gem = obj;
+
+	/*
+	 * FIXME: The dependency on GEM here isn't required, we could
+	 * convert the driver handle to a dma-buf instead and use the
+	 * backend-agnostic dma-buf vmap support instead. This would
+	 * require that the handle2fd prime ioctl is reworked to pull the
+	 * fd_install step out of the driver backend hooks, to make that
+	 * final step optional for internal users.
+	 */
+	vaddr = dev->driver->gem_prime_vmap(obj);
+	if (!vaddr) {
+		ret = -ENOMEM;
+		goto err_delete;
+	}
+
+	buffer->vaddr = vaddr;
+
+	return buffer;
+
+err_delete:
+	drm_client_buffer_delete(buffer);
+
+	return ERR_PTR(ret);
+}
+
+static void drm_client_buffer_rmfb(struct drm_client_buffer *buffer)
+{
+	int ret;
+
+	if (!buffer->fb)
+		return;
+
+	ret = drm_mode_rmfb(buffer->client->dev, buffer->fb->base.id, buffer->client->file);
+	if (ret)
+		DRM_DEV_ERROR(buffer->client->dev->dev,
+			      "Error removing FB:%u (%d)\n", buffer->fb->base.id, ret);
+
+	buffer->fb = NULL;
+}
+
+static int drm_client_buffer_addfb(struct drm_client_buffer *buffer,
+				   u32 width, u32 height, u32 format)
+{
+	struct drm_client_dev *client = buffer->client;
+	struct drm_mode_fb_cmd fb_req = { };
+	const struct drm_format_info *info;
+	int ret;
+
+	info = drm_format_info(format);
+	fb_req.bpp = info->cpp[0] * 8;
+	fb_req.depth = info->depth;
+	fb_req.width = width;
+	fb_req.height = height;
+	fb_req.handle = buffer->handle;
+	fb_req.pitch = buffer->pitch;
+
+	ret = drm_mode_addfb(client->dev, &fb_req, client->file);
+	if (ret)
+		return ret;
+
+	buffer->fb = drm_framebuffer_lookup(client->dev, buffer->client->file, fb_req.fb_id);
+	if (WARN_ON(!buffer->fb))
+		return -ENOENT;
+
+	/* drop the reference we picked up in framebuffer lookup */
+	drm_framebuffer_put(buffer->fb);
+
+	strscpy(buffer->fb->comm, client->name, TASK_COMM_LEN);
+
+	return 0;
+}
+
+/**
+ * drm_client_framebuffer_create - Create a client framebuffer
+ * @client: DRM client
+ * @width: Framebuffer width
+ * @height: Framebuffer height
+ * @format: Buffer format
+ *
+ * This function creates a &drm_client_buffer which consists of a
+ * &drm_framebuffer backed by a dumb buffer.
+ * Call drm_client_framebuffer_delete() to free the buffer.
+ *
+ * Returns:
+ * Pointer to a client buffer or an error pointer on failure.
+ */
+struct drm_client_buffer *
+drm_client_framebuffer_create(struct drm_client_dev *client, u32 width, u32 height, u32 format)
+{
+	struct drm_client_buffer *buffer;
+	int ret;
+
+	buffer = drm_client_buffer_create(client, width, height, format);
+	if (IS_ERR(buffer))
+		return buffer;
+
+	ret = drm_client_buffer_addfb(buffer, width, height, format);
+	if (ret) {
+		drm_client_buffer_delete(buffer);
+		return ERR_PTR(ret);
+	}
+
+	return buffer;
+}
+EXPORT_SYMBOL(drm_client_framebuffer_create);
+
+/**
+ * drm_client_framebuffer_delete - Delete a client framebuffer
+ * @buffer: DRM client buffer (can be NULL)
+ */
+void drm_client_framebuffer_delete(struct drm_client_buffer *buffer)
+{
+	if (!buffer)
+		return;
+
+	drm_client_buffer_rmfb(buffer);
+	drm_client_buffer_delete(buffer);
+}
+EXPORT_SYMBOL(drm_client_framebuffer_delete);
+
+#ifdef CONFIG_DEBUG_FS
+static int drm_client_debugfs_internal_clients(struct seq_file *m, void *data)
+{
+	struct drm_info_node *node = m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct drm_printer p = drm_seq_file_printer(m);
+	struct drm_client_dev *client;
+
+	mutex_lock(&dev->clientlist_mutex);
+	list_for_each_entry(client, &dev->clientlist, list)
+		drm_printf(&p, "%s\n", client->name);
+	mutex_unlock(&dev->clientlist_mutex);
+
+	return 0;
+}
+
+static const struct drm_info_list drm_client_debugfs_list[] = {
+	{ "internal_clients", drm_client_debugfs_internal_clients, 0 },
+};
+
+int drm_client_debugfs_init(struct drm_minor *minor)
+{
+	return drm_debugfs_create_files(drm_client_debugfs_list,
+					ARRAY_SIZE(drm_client_debugfs_list),
+					minor->debugfs_root, minor);
+}
+#endif
diff --git a/drivers/gpu/drm/drm_connector.c b/drivers/gpu/drm/drm_connector.c
index b3cde897cd80..6011d769d50b 100644
--- a/drivers/gpu/drm/drm_connector.c
+++ b/drivers/gpu/drm/drm_connector.c
@@ -48,7 +48,7 @@
  *
  * Connectors must be attached to an encoder to be used. For devices that map
  * connectors to encoders 1:1, the connector should be attached at
- * initialization time with a call to drm_mode_connector_attach_encoder(). The
+ * initialization time with a call to drm_connector_attach_encoder(). The
  * driver must also set the &drm_connector.encoder field to point to the
  * attached encoder.
  *
@@ -87,6 +87,7 @@ static struct drm_conn_prop_enum_list drm_connector_enum_list[] = {
 	{ DRM_MODE_CONNECTOR_VIRTUAL, "Virtual" },
 	{ DRM_MODE_CONNECTOR_DSI, "DSI" },
 	{ DRM_MODE_CONNECTOR_DPI, "DPI" },
+	{ DRM_MODE_CONNECTOR_WRITEBACK, "Writeback" },
 };
 
 void drm_connector_ida_init(void)
@@ -195,6 +196,10 @@ int drm_connector_init(struct drm_device *dev,
 	struct ida *connector_ida =
 		&drm_connector_enum_list[connector_type].ida;
 
+	WARN_ON(drm_drv_uses_atomic_modeset(dev) &&
+		(!funcs->atomic_destroy_state ||
+		 !funcs->atomic_duplicate_state));
+
 	ret = __drm_mode_object_add(dev, &connector->base,
 				    DRM_MODE_OBJECT_CONNECTOR,
 				    false, drm_connector_free);
@@ -249,7 +254,8 @@ int drm_connector_init(struct drm_device *dev,
 	config->num_connector++;
 	spin_unlock_irq(&config->connector_list_lock);
 
-	if (connector_type != DRM_MODE_CONNECTOR_VIRTUAL)
+	if (connector_type != DRM_MODE_CONNECTOR_VIRTUAL &&
+	    connector_type != DRM_MODE_CONNECTOR_WRITEBACK)
 		drm_object_attach_property(&connector->base,
 					      config->edid_property,
 					      0);
@@ -285,7 +291,7 @@ out_put:
 EXPORT_SYMBOL(drm_connector_init);
 
 /**
- * drm_mode_connector_attach_encoder - attach a connector to an encoder
+ * drm_connector_attach_encoder - attach a connector to an encoder
  * @connector: connector to attach
  * @encoder: encoder to attach @connector to
  *
@@ -296,8 +302,8 @@ EXPORT_SYMBOL(drm_connector_init);
  * Returns:
  * Zero on success, negative errno on failure.
  */
-int drm_mode_connector_attach_encoder(struct drm_connector *connector,
-				      struct drm_encoder *encoder)
+int drm_connector_attach_encoder(struct drm_connector *connector,
+				 struct drm_encoder *encoder)
 {
 	int i;
 
@@ -315,7 +321,7 @@ int drm_mode_connector_attach_encoder(struct drm_connector *connector,
 	if (WARN_ON(connector->encoder))
 		return -EINVAL;
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
+	for (i = 0; i < ARRAY_SIZE(connector->encoder_ids); i++) {
 		if (connector->encoder_ids[i] == 0) {
 			connector->encoder_ids[i] = encoder->base.id;
 			return 0;
@@ -323,7 +329,30 @@ int drm_mode_connector_attach_encoder(struct drm_connector *connector,
 	}
 	return -ENOMEM;
 }
-EXPORT_SYMBOL(drm_mode_connector_attach_encoder);
+EXPORT_SYMBOL(drm_connector_attach_encoder);
+
+/**
+ * drm_connector_has_possible_encoder - check if the connector and encoder are assosicated with each other
+ * @connector: the connector
+ * @encoder: the encoder
+ *
+ * Returns:
+ * True if @encoder is one of the possible encoders for @connector.
+ */
+bool drm_connector_has_possible_encoder(struct drm_connector *connector,
+					struct drm_encoder *encoder)
+{
+	struct drm_encoder *enc;
+	int i;
+
+	drm_connector_for_each_possible_encoder(connector, enc, i) {
+		if (enc == encoder)
+			return true;
+	}
+
+	return false;
+}
+EXPORT_SYMBOL(drm_connector_has_possible_encoder);
 
 static void drm_mode_remove(struct drm_connector *connector,
 			    struct drm_display_mode *mode)
@@ -577,7 +606,7 @@ __drm_connector_put_safe(struct drm_connector *conn)
 
 /**
  * drm_connector_list_iter_next - return next connector
- * @iter: connectr_list iterator
+ * @iter: connector_list iterator
  *
  * Returns the next connector for @iter, or NULL when the list walk has
  * completed.
@@ -720,6 +749,14 @@ static const struct drm_prop_enum_list drm_aspect_ratio_enum_list[] = {
 	{ DRM_MODE_PICTURE_ASPECT_16_9, "16:9" },
 };
 
+static const struct drm_prop_enum_list drm_content_type_enum_list[] = {
+	{ DRM_MODE_CONTENT_TYPE_NO_DATA, "No Data" },
+	{ DRM_MODE_CONTENT_TYPE_GRAPHICS, "Graphics" },
+	{ DRM_MODE_CONTENT_TYPE_PHOTO, "Photo" },
+	{ DRM_MODE_CONTENT_TYPE_CINEMA, "Cinema" },
+	{ DRM_MODE_CONTENT_TYPE_GAME, "Game" },
+};
+
 static const struct drm_prop_enum_list drm_panel_orientation_enum_list[] = {
 	{ DRM_MODE_PANEL_ORIENTATION_NORMAL,	"Normal"	},
 	{ DRM_MODE_PANEL_ORIENTATION_BOTTOM_UP,	"Upside Down"	},
@@ -777,7 +814,7 @@ DRM_ENUM_NAME_FN(drm_get_content_protection_name, drm_cp_enum_list)
  * 	Blob property which contains the current EDID read from the sink. This
  * 	is useful to parse sink identification information like vendor, model
  * 	and serial. Drivers should update this property by calling
- * 	drm_mode_connector_update_edid_property(), usually after having parsed
+ * 	drm_connector_update_edid_property(), usually after having parsed
  * 	the EDID using drm_add_edid_modes(). Userspace cannot change this
  * 	property.
  * DPMS:
@@ -815,7 +852,7 @@ DRM_ENUM_NAME_FN(drm_get_content_protection_name, drm_cp_enum_list)
  * PATH:
  * 	Connector path property to identify how this sink is physically
  * 	connected. Used by DP MST. This should be set by calling
- * 	drm_mode_connector_set_path_property(), in the case of DP MST with the
+ * 	drm_connector_set_path_property(), in the case of DP MST with the
  * 	path property the MST manager created. Userspace cannot change this
  * 	property.
  * TILE:
@@ -826,14 +863,14 @@ DRM_ENUM_NAME_FN(drm_get_content_protection_name, drm_cp_enum_list)
  * 	are not gen-locked. Note that for tiled panels which are genlocked, like
  * 	dual-link LVDS or dual-link DSI, the driver should try to not expose the
  * 	tiling and virtualize both &drm_crtc and &drm_plane if needed. Drivers
- * 	should update this value using drm_mode_connector_set_tile_property().
+ * 	should update this value using drm_connector_set_tile_property().
  * 	Userspace cannot change this property.
  * link-status:
  *      Connector link-status property to indicate the status of link. The
  *      default value of link-status is "GOOD". If something fails during or
  *      after modeset, the kernel driver may set this to "BAD" and issue a
  *      hotplug uevent. Drivers should update this value using
- *      drm_mode_connector_set_link_status_property().
+ *      drm_connector_set_link_status_property().
  * non_desktop:
  * 	Indicates the output should be ignored for purposes of displaying a
  * 	standard desktop environment or console. This is most likely because
@@ -997,6 +1034,82 @@ int drm_mode_create_dvi_i_properties(struct drm_device *dev)
 EXPORT_SYMBOL(drm_mode_create_dvi_i_properties);
 
 /**
+ * DOC: HDMI connector properties
+ *
+ * content type (HDMI specific):
+ *	Indicates content type setting to be used in HDMI infoframes to indicate
+ *	content type for the external device, so that it adjusts it's display
+ *	settings accordingly.
+ *
+ *	The value of this property can be one of the following:
+ *
+ *	No Data:
+ *		Content type is unknown
+ *	Graphics:
+ *		Content type is graphics
+ *	Photo:
+ *		Content type is photo
+ *	Cinema:
+ *		Content type is cinema
+ *	Game:
+ *		Content type is game
+ *
+ *	Drivers can set up this property by calling
+ *	drm_connector_attach_content_type_property(). Decoding to
+ *	infoframe values is done through drm_hdmi_avi_infoframe_content_type().
+ */
+
+/**
+ * drm_connector_attach_content_type_property - attach content-type property
+ * @connector: connector to attach content type property on.
+ *
+ * Called by a driver the first time a HDMI connector is made.
+ */
+int drm_connector_attach_content_type_property(struct drm_connector *connector)
+{
+	if (!drm_mode_create_content_type_property(connector->dev))
+		drm_object_attach_property(&connector->base,
+					   connector->dev->mode_config.content_type_property,
+					   DRM_MODE_CONTENT_TYPE_NO_DATA);
+	return 0;
+}
+EXPORT_SYMBOL(drm_connector_attach_content_type_property);
+
+
+/**
+ * drm_hdmi_avi_infoframe_content_type() - fill the HDMI AVI infoframe
+ *                                         content type information, based
+ *                                         on correspondent DRM property.
+ * @frame: HDMI AVI infoframe
+ * @conn_state: DRM display connector state
+ *
+ */
+void drm_hdmi_avi_infoframe_content_type(struct hdmi_avi_infoframe *frame,
+					 const struct drm_connector_state *conn_state)
+{
+	switch (conn_state->content_type) {
+	case DRM_MODE_CONTENT_TYPE_GRAPHICS:
+		frame->content_type = HDMI_CONTENT_TYPE_GRAPHICS;
+		break;
+	case DRM_MODE_CONTENT_TYPE_CINEMA:
+		frame->content_type = HDMI_CONTENT_TYPE_CINEMA;
+		break;
+	case DRM_MODE_CONTENT_TYPE_GAME:
+		frame->content_type = HDMI_CONTENT_TYPE_GAME;
+		break;
+	case DRM_MODE_CONTENT_TYPE_PHOTO:
+		frame->content_type = HDMI_CONTENT_TYPE_PHOTO;
+		break;
+	default:
+		/* Graphics is the default(0) */
+		frame->content_type = HDMI_CONTENT_TYPE_GRAPHICS;
+	}
+
+	frame->itc = conn_state->content_type != DRM_MODE_CONTENT_TYPE_NO_DATA;
+}
+EXPORT_SYMBOL(drm_hdmi_avi_infoframe_content_type);
+
+/**
  * drm_create_tv_properties - create TV specific connector properties
  * @dev: DRM device
  * @num_modes: number of different TV formats (modes) supported
@@ -1069,7 +1182,7 @@ int drm_mode_create_tv_properties(struct drm_device *dev,
 		goto nomem;
 
 	for (i = 0; i < num_modes; i++)
-		drm_property_add_enum(dev->mode_config.tv_mode_property, i,
+		drm_property_add_enum(dev->mode_config.tv_mode_property,
 				      i, modes[i]);
 
 	dev->mode_config.tv_brightness_property =
@@ -1156,7 +1269,7 @@ int drm_connector_attach_scaling_mode_property(struct drm_connector *connector,
 {
 	struct drm_device *dev = connector->dev;
 	struct drm_property *scaling_mode_property;
-	int i, j = 0;
+	int i;
 	const unsigned valid_scaling_mode_mask =
 		(1U << ARRAY_SIZE(drm_scaling_mode_enum_list)) - 1;
 
@@ -1177,7 +1290,7 @@ int drm_connector_attach_scaling_mode_property(struct drm_connector *connector,
 		if (!(BIT(i) & scaling_mode_mask))
 			continue;
 
-		ret = drm_property_add_enum(scaling_mode_property, j++,
+		ret = drm_property_add_enum(scaling_mode_property,
 					    drm_scaling_mode_enum_list[i].type,
 					    drm_scaling_mode_enum_list[i].name);
 
@@ -1261,6 +1374,33 @@ int drm_mode_create_aspect_ratio_property(struct drm_device *dev)
 EXPORT_SYMBOL(drm_mode_create_aspect_ratio_property);
 
 /**
+ * drm_mode_create_content_type_property - create content type property
+ * @dev: DRM device
+ *
+ * Called by a driver the first time it's needed, must be attached to desired
+ * connectors.
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int drm_mode_create_content_type_property(struct drm_device *dev)
+{
+	if (dev->mode_config.content_type_property)
+		return 0;
+
+	dev->mode_config.content_type_property =
+		drm_property_create_enum(dev, 0, "content type",
+					 drm_content_type_enum_list,
+					 ARRAY_SIZE(drm_content_type_enum_list));
+
+	if (dev->mode_config.content_type_property == NULL)
+		return -ENOMEM;
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_mode_create_content_type_property);
+
+/**
  * drm_mode_create_suggested_offset_properties - create suggests offset properties
  * @dev: DRM device
  *
@@ -1285,7 +1425,7 @@ int drm_mode_create_suggested_offset_properties(struct drm_device *dev)
 EXPORT_SYMBOL(drm_mode_create_suggested_offset_properties);
 
 /**
- * drm_mode_connector_set_path_property - set tile property on connector
+ * drm_connector_set_path_property - set tile property on connector
  * @connector: connector to set property on.
  * @path: path to use for property; must not be NULL.
  *
@@ -1297,8 +1437,8 @@ EXPORT_SYMBOL(drm_mode_create_suggested_offset_properties);
  * Returns:
  * Zero on success, negative errno on failure.
  */
-int drm_mode_connector_set_path_property(struct drm_connector *connector,
-					 const char *path)
+int drm_connector_set_path_property(struct drm_connector *connector,
+				    const char *path)
 {
 	struct drm_device *dev = connector->dev;
 	int ret;
@@ -1311,10 +1451,10 @@ int drm_mode_connector_set_path_property(struct drm_connector *connector,
 	                                       dev->mode_config.path_property);
 	return ret;
 }
-EXPORT_SYMBOL(drm_mode_connector_set_path_property);
+EXPORT_SYMBOL(drm_connector_set_path_property);
 
 /**
- * drm_mode_connector_set_tile_property - set tile property on connector
+ * drm_connector_set_tile_property - set tile property on connector
  * @connector: connector to set property on.
  *
  * This looks up the tile information for a connector, and creates a
@@ -1324,7 +1464,7 @@ EXPORT_SYMBOL(drm_mode_connector_set_path_property);
  * Returns:
  * Zero on success, errno on failure.
  */
-int drm_mode_connector_set_tile_property(struct drm_connector *connector)
+int drm_connector_set_tile_property(struct drm_connector *connector)
 {
 	struct drm_device *dev = connector->dev;
 	char tile[256];
@@ -1354,10 +1494,10 @@ int drm_mode_connector_set_tile_property(struct drm_connector *connector)
 	                                       dev->mode_config.tile_property);
 	return ret;
 }
-EXPORT_SYMBOL(drm_mode_connector_set_tile_property);
+EXPORT_SYMBOL(drm_connector_set_tile_property);
 
 /**
- * drm_mode_connector_update_edid_property - update the edid property of a connector
+ * drm_connector_update_edid_property - update the edid property of a connector
  * @connector: drm connector
  * @edid: new value of the edid property
  *
@@ -1367,8 +1507,8 @@ EXPORT_SYMBOL(drm_mode_connector_set_tile_property);
  * Returns:
  * Zero on success, negative errno on failure.
  */
-int drm_mode_connector_update_edid_property(struct drm_connector *connector,
-					    const struct edid *edid)
+int drm_connector_update_edid_property(struct drm_connector *connector,
+				       const struct edid *edid)
 {
 	struct drm_device *dev = connector->dev;
 	size_t size = 0;
@@ -1406,10 +1546,10 @@ int drm_mode_connector_update_edid_property(struct drm_connector *connector,
 	                                       dev->mode_config.edid_property);
 	return ret;
 }
-EXPORT_SYMBOL(drm_mode_connector_update_edid_property);
+EXPORT_SYMBOL(drm_connector_update_edid_property);
 
 /**
- * drm_mode_connector_set_link_status_property - Set link status property of a connector
+ * drm_connector_set_link_status_property - Set link status property of a connector
  * @connector: drm connector
  * @link_status: new value of link status property (0: Good, 1: Bad)
  *
@@ -1427,8 +1567,8 @@ EXPORT_SYMBOL(drm_mode_connector_update_edid_property);
  * it is not limited to DP or link training. For example, if we implement
  * asynchronous setcrtc, this property can be used to report any failures in that.
  */
-void drm_mode_connector_set_link_status_property(struct drm_connector *connector,
-						 uint64_t link_status)
+void drm_connector_set_link_status_property(struct drm_connector *connector,
+					    uint64_t link_status)
 {
 	struct drm_device *dev = connector->dev;
 
@@ -1436,7 +1576,7 @@ void drm_mode_connector_set_link_status_property(struct drm_connector *connector
 	connector->state->link_status = link_status;
 	drm_modeset_unlock(&dev->mode_config.connection_mutex);
 }
-EXPORT_SYMBOL(drm_mode_connector_set_link_status_property);
+EXPORT_SYMBOL(drm_connector_set_link_status_property);
 
 /**
  * drm_connector_init_panel_orientation_property -
@@ -1489,7 +1629,7 @@ int drm_connector_init_panel_orientation_property(
 }
 EXPORT_SYMBOL(drm_connector_init_panel_orientation_property);
 
-int drm_mode_connector_set_obj_prop(struct drm_mode_object *obj,
+int drm_connector_set_obj_prop(struct drm_mode_object *obj,
 				    struct drm_property *property,
 				    uint64_t value)
 {
@@ -1507,8 +1647,8 @@ int drm_mode_connector_set_obj_prop(struct drm_mode_object *obj,
 	return ret;
 }
 
-int drm_mode_connector_property_set_ioctl(struct drm_device *dev,
-				       void *data, struct drm_file *file_priv)
+int drm_connector_property_set_ioctl(struct drm_device *dev,
+				     void *data, struct drm_file *file_priv)
 {
 	struct drm_mode_connector_set_property *conn_set_prop = data;
 	struct drm_mode_obj_set_property obj_set_prop = {
@@ -1531,8 +1671,10 @@ static struct drm_encoder *drm_connector_get_encoder(struct drm_connector *conne
 	return connector->encoder;
 }
 
-static bool drm_mode_expose_to_userspace(const struct drm_display_mode *mode,
-					 const struct drm_file *file_priv)
+static bool
+drm_mode_expose_to_userspace(const struct drm_display_mode *mode,
+			     const struct list_head *export_list,
+			     const struct drm_file *file_priv)
 {
 	/*
 	 * If user-space hasn't configured the driver to expose the stereo 3D
@@ -1540,6 +1682,23 @@ static bool drm_mode_expose_to_userspace(const struct drm_display_mode *mode,
 	 */
 	if (!file_priv->stereo_allowed && drm_mode_is_stereo(mode))
 		return false;
+	/*
+	 * If user-space hasn't configured the driver to expose the modes
+	 * with aspect-ratio, don't expose them. However if such a mode
+	 * is unique, let it be exposed, but reset the aspect-ratio flags
+	 * while preparing the list of user-modes.
+	 */
+	if (!file_priv->aspect_ratio_allowed) {
+		struct drm_display_mode *mode_itr;
+
+		list_for_each_entry(mode_itr, export_list, export_head)
+			if (drm_mode_match(mode_itr, mode,
+					   DRM_MODE_MATCH_TIMINGS |
+					   DRM_MODE_MATCH_CLOCK |
+					   DRM_MODE_MATCH_FLAGS |
+					   DRM_MODE_MATCH_3D_FLAGS))
+				return false;
+	}
 
 	return true;
 }
@@ -1559,6 +1718,7 @@ int drm_mode_getconnector(struct drm_device *dev, void *data,
 	struct drm_mode_modeinfo u_mode;
 	struct drm_mode_modeinfo __user *mode_ptr;
 	uint32_t __user *encoder_ptr;
+	LIST_HEAD(export_list);
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
@@ -1569,22 +1729,19 @@ int drm_mode_getconnector(struct drm_device *dev, void *data,
 	if (!connector)
 		return -ENOENT;
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++)
-		if (connector->encoder_ids[i] != 0)
-			encoders_count++;
+	drm_connector_for_each_possible_encoder(connector, encoder, i)
+		encoders_count++;
 
 	if ((out_resp->count_encoders >= encoders_count) && encoders_count) {
 		copied = 0;
 		encoder_ptr = (uint32_t __user *)(unsigned long)(out_resp->encoders_ptr);
-		for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-			if (connector->encoder_ids[i] != 0) {
-				if (put_user(connector->encoder_ids[i],
-					     encoder_ptr + copied)) {
-					ret = -EFAULT;
-					goto out;
-				}
-				copied++;
+
+		drm_connector_for_each_possible_encoder(connector, encoder, i) {
+			if (put_user(encoder->base.id, encoder_ptr + copied)) {
+				ret = -EFAULT;
+				goto out;
 			}
+			copied++;
 		}
 	}
 	out_resp->count_encoders = encoders_count;
@@ -1607,21 +1764,31 @@ int drm_mode_getconnector(struct drm_device *dev, void *data,
 
 	/* delayed so we get modes regardless of pre-fill_modes state */
 	list_for_each_entry(mode, &connector->modes, head)
-		if (drm_mode_expose_to_userspace(mode, file_priv))
+		if (drm_mode_expose_to_userspace(mode, &export_list,
+						 file_priv)) {
+			list_add_tail(&mode->export_head, &export_list);
 			mode_count++;
+		}
 
 	/*
 	 * This ioctl is called twice, once to determine how much space is
 	 * needed, and the 2nd time to fill it.
+	 * The modes that need to be exposed to the user are maintained in the
+	 * 'export_list'. When the ioctl is called first time to determine the,
+	 * space, the export_list gets filled, to find the no.of modes. In the
+	 * 2nd time, the user modes are filled, one by one from the export_list.
 	 */
 	if ((out_resp->count_modes >= mode_count) && mode_count) {
 		copied = 0;
 		mode_ptr = (struct drm_mode_modeinfo __user *)(unsigned long)out_resp->modes_ptr;
-		list_for_each_entry(mode, &connector->modes, head) {
-			if (!drm_mode_expose_to_userspace(mode, file_priv))
-				continue;
-
+		list_for_each_entry(mode, &export_list, export_head) {
 			drm_mode_convert_to_umode(&u_mode, mode);
+			/*
+			 * Reset aspect ratio flags of user-mode, if modes with
+			 * aspect-ratio are not supported.
+			 */
+			if (!file_priv->aspect_ratio_allowed)
+				u_mode.flags &= ~DRM_MODE_FLAG_PIC_AR_MASK;
 			if (copy_to_user(mode_ptr + copied,
 					 &u_mode, sizeof(u_mode))) {
 				ret = -EFAULT;
diff --git a/drivers/gpu/drm/drm_crtc.c b/drivers/gpu/drm/drm_crtc.c
index a231dd5dce16..bae43938c8f6 100644
--- a/drivers/gpu/drm/drm_crtc.c
+++ b/drivers/gpu/drm/drm_crtc.c
@@ -225,16 +225,9 @@ static const char *drm_crtc_fence_get_timeline_name(struct dma_fence *fence)
 	return crtc->timeline_name;
 }
 
-static bool drm_crtc_fence_enable_signaling(struct dma_fence *fence)
-{
-	return true;
-}
-
 static const struct dma_fence_ops drm_crtc_fence_ops = {
 	.get_driver_name = drm_crtc_fence_get_driver_name,
 	.get_timeline_name = drm_crtc_fence_get_timeline_name,
-	.enable_signaling = drm_crtc_fence_enable_signaling,
-	.wait = dma_fence_default_wait,
 };
 
 struct dma_fence *drm_crtc_create_fence(struct drm_crtc *crtc)
@@ -286,6 +279,10 @@ int drm_crtc_init_with_planes(struct drm_device *dev, struct drm_crtc *crtc,
 	if (WARN_ON(config->num_crtc >= 32))
 		return -EINVAL;
 
+	WARN_ON(drm_drv_uses_atomic_modeset(dev) &&
+		(!funcs->atomic_destroy_state ||
+		 !funcs->atomic_duplicate_state));
+
 	crtc->dev = dev;
 	crtc->funcs = funcs;
 
@@ -325,9 +322,9 @@ int drm_crtc_init_with_planes(struct drm_device *dev, struct drm_crtc *crtc,
 	crtc->primary = primary;
 	crtc->cursor = cursor;
 	if (primary && !primary->possible_crtcs)
-		primary->possible_crtcs = 1 << drm_crtc_index(crtc);
+		primary->possible_crtcs = drm_crtc_mask(crtc);
 	if (cursor && !cursor->possible_crtcs)
-		cursor->possible_crtcs = 1 << drm_crtc_index(crtc);
+		cursor->possible_crtcs = drm_crtc_mask(crtc);
 
 	ret = drm_crtc_crc_init(crtc);
 	if (ret) {
@@ -449,6 +446,8 @@ int drm_mode_getcrtc(struct drm_device *dev,
 			crtc_resp->mode_valid = 0;
 		}
 	}
+	if (!file_priv->aspect_ratio_allowed)
+		crtc_resp->mode.flags &= ~DRM_MODE_FLAG_PIC_AR_MASK;
 	drm_modeset_unlock(&crtc->mutex);
 
 	return 0;
@@ -462,32 +461,42 @@ static int __drm_mode_set_config_internal(struct drm_mode_set *set,
 	struct drm_crtc *tmp;
 	int ret;
 
+	WARN_ON(drm_drv_uses_atomic_modeset(crtc->dev));
+
 	/*
 	 * NOTE: ->set_config can also disable other crtcs (if we steal all
 	 * connectors from it), hence we need to refcount the fbs across all
 	 * crtcs. Atomic modeset will have saner semantics ...
 	 */
-	drm_for_each_crtc(tmp, crtc->dev)
-		tmp->primary->old_fb = tmp->primary->fb;
+	drm_for_each_crtc(tmp, crtc->dev) {
+		struct drm_plane *plane = tmp->primary;
+
+		plane->old_fb = plane->fb;
+	}
 
 	fb = set->fb;
 
 	ret = crtc->funcs->set_config(set, ctx);
 	if (ret == 0) {
-		crtc->primary->crtc = fb ? crtc : NULL;
-		crtc->primary->fb = fb;
+		struct drm_plane *plane = crtc->primary;
+
+		plane->crtc = fb ? crtc : NULL;
+		plane->fb = fb;
 	}
 
 	drm_for_each_crtc(tmp, crtc->dev) {
-		if (tmp->primary->fb)
-			drm_framebuffer_get(tmp->primary->fb);
-		if (tmp->primary->old_fb)
-			drm_framebuffer_put(tmp->primary->old_fb);
-		tmp->primary->old_fb = NULL;
+		struct drm_plane *plane = tmp->primary;
+
+		if (plane->fb)
+			drm_framebuffer_get(plane->fb);
+		if (plane->old_fb)
+			drm_framebuffer_put(plane->old_fb);
+		plane->old_fb = NULL;
 	}
 
 	return ret;
 }
+
 /**
  * drm_mode_set_config_internal - helper to call &drm_mode_config_funcs.set_config
  * @set: modeset config to set
@@ -628,10 +637,19 @@ retry:
 			ret = -ENOMEM;
 			goto out;
 		}
+		if (!file_priv->aspect_ratio_allowed &&
+		    (crtc_req->mode.flags & DRM_MODE_FLAG_PIC_AR_MASK) != DRM_MODE_FLAG_PIC_AR_NONE) {
+			DRM_DEBUG_KMS("Unexpected aspect-ratio flag bits\n");
+			ret = -EINVAL;
+			goto out;
+		}
+
 
 		ret = drm_mode_convert_umode(dev, mode, &crtc_req->mode);
 		if (ret) {
-			DRM_DEBUG_KMS("Invalid mode\n");
+			DRM_DEBUG_KMS("Invalid mode (ret=%d, status=%s)\n",
+				      ret, drm_get_mode_status_name(mode->status));
+			drm_mode_debug_printmodeline(mode);
 			goto out;
 		}
 
@@ -723,7 +741,11 @@ retry:
 	set.connectors = connector_set;
 	set.num_connectors = crtc_req->count_connectors;
 	set.fb = fb;
-	ret = __drm_mode_set_config_internal(&set, &ctx);
+
+	if (drm_drv_uses_atomic_modeset(dev))
+		ret = crtc->funcs->set_config(&set, &ctx);
+	else
+		ret = __drm_mode_set_config_internal(&set, &ctx);
 
 out:
 	if (fb)
diff --git a/drivers/gpu/drm/drm_crtc_internal.h b/drivers/gpu/drm/drm_crtc_internal.h
index 5d307b23a4e6..b61322763394 100644
--- a/drivers/gpu/drm/drm_crtc_internal.h
+++ b/drivers/gpu/drm/drm_crtc_internal.h
@@ -56,12 +56,21 @@ int drm_mode_setcrtc(struct drm_device *dev,
 int drm_modeset_register_all(struct drm_device *dev);
 void drm_modeset_unregister_all(struct drm_device *dev);
 
+/* drm_modes.c */
+const char *drm_get_mode_status_name(enum drm_mode_status status);
+
 /* IOCTLs */
 int drm_mode_getresources(struct drm_device *dev,
 			  void *data, struct drm_file *file_priv);
 
 
 /* drm_dumb_buffers.c */
+int drm_mode_create_dumb(struct drm_device *dev,
+			 struct drm_mode_create_dumb *args,
+			 struct drm_file *file_priv);
+int drm_mode_destroy_dumb(struct drm_device *dev, u32 handle,
+			  struct drm_file *file_priv);
+
 /* IOCTLs */
 int drm_mode_create_dumb_ioctl(struct drm_device *dev,
 			       void *data, struct drm_file *file_priv);
@@ -139,7 +148,7 @@ void drm_connector_ida_init(void);
 void drm_connector_ida_destroy(void);
 void drm_connector_unregister_all(struct drm_device *dev);
 int drm_connector_register_all(struct drm_device *dev);
-int drm_mode_connector_set_obj_prop(struct drm_mode_object *obj,
+int drm_connector_set_obj_prop(struct drm_mode_object *obj,
 				    struct drm_property *property,
 				    uint64_t value);
 int drm_connector_create_standard_properties(struct drm_device *dev);
@@ -147,8 +156,8 @@ const char *drm_get_connector_force_name(enum drm_connector_force force);
 void drm_connector_free_work_fn(struct work_struct *work);
 
 /* IOCTL */
-int drm_mode_connector_property_set_ioctl(struct drm_device *dev,
-					  void *data, struct drm_file *file_priv);
+int drm_connector_property_set_ioctl(struct drm_device *dev,
+				     void *data, struct drm_file *file_priv);
 int drm_mode_getconnector(struct drm_device *dev,
 			  void *data, struct drm_file *file_priv);
 
@@ -163,14 +172,19 @@ int drm_framebuffer_check_src_coords(uint32_t src_x, uint32_t src_y,
 				     const struct drm_framebuffer *fb);
 void drm_fb_release(struct drm_file *file_priv);
 
+int drm_mode_addfb(struct drm_device *dev, struct drm_mode_fb_cmd *or,
+		   struct drm_file *file_priv);
+int drm_mode_rmfb(struct drm_device *dev, u32 fb_id,
+		  struct drm_file *file_priv);
+
 
 /* IOCTL */
-int drm_mode_addfb(struct drm_device *dev,
-		   void *data, struct drm_file *file_priv);
+int drm_mode_addfb_ioctl(struct drm_device *dev,
+			 void *data, struct drm_file *file_priv);
 int drm_mode_addfb2(struct drm_device *dev,
 		    void *data, struct drm_file *file_priv);
-int drm_mode_rmfb(struct drm_device *dev,
-		  void *data, struct drm_file *file_priv);
+int drm_mode_rmfb_ioctl(struct drm_device *dev,
+			void *data, struct drm_file *file_priv);
 int drm_mode_getfb(struct drm_device *dev,
 		   void *data, struct drm_file *file_priv);
 int drm_mode_dirtyfb_ioctl(struct drm_device *dev,
diff --git a/drivers/gpu/drm/drm_debugfs.c b/drivers/gpu/drm/drm_debugfs.c
index b2482818fee8..6f28fe58f169 100644
--- a/drivers/gpu/drm/drm_debugfs.c
+++ b/drivers/gpu/drm/drm_debugfs.c
@@ -28,6 +28,7 @@
 #include <linux/slab.h>
 #include <linux/export.h>
 
+#include <drm/drm_client.h>
 #include <drm/drm_debugfs.h>
 #include <drm/drm_edid.h>
 #include <drm/drm_atomic.h>
@@ -164,6 +165,12 @@ int drm_debugfs_init(struct drm_minor *minor, int minor_id,
 			DRM_ERROR("Failed to create framebuffer debugfs file\n");
 			return ret;
 		}
+
+		ret = drm_client_debugfs_init(minor);
+		if (ret) {
+			DRM_ERROR("Failed to create client debugfs file\n");
+			return ret;
+		}
 	}
 
 	if (dev->driver->debugfs_init) {
@@ -307,13 +314,13 @@ static ssize_t edid_write(struct file *file, const char __user *ubuf,
 
 	if (len == 5 && !strncmp(buf, "reset", 5)) {
 		connector->override_edid = false;
-		ret = drm_mode_connector_update_edid_property(connector, NULL);
+		ret = drm_connector_update_edid_property(connector, NULL);
 	} else if (len < EDID_LENGTH ||
 		   EDID_LENGTH * (1 + edid->extensions) > len)
 		ret = -EINVAL;
 	else {
 		connector->override_edid = false;
-		ret = drm_mode_connector_update_edid_property(connector, edid);
+		ret = drm_connector_update_edid_property(connector, edid);
 		if (!ret)
 			connector->override_edid = true;
 	}
diff --git a/drivers/gpu/drm/drm_debugfs_crc.c b/drivers/gpu/drm/drm_debugfs_crc.c
index 9f8312137cad..99961192bf03 100644
--- a/drivers/gpu/drm/drm_debugfs_crc.c
+++ b/drivers/gpu/drm/drm_debugfs_crc.c
@@ -139,6 +139,7 @@ static int crtc_crc_data_count(struct drm_crtc_crc *crc)
 static void crtc_crc_cleanup(struct drm_crtc_crc *crc)
 {
 	kfree(crc->entries);
+	crc->overflow = false;
 	crc->entries = NULL;
 	crc->head = 0;
 	crc->tail = 0;
@@ -391,8 +392,14 @@ int drm_crtc_add_crc_entry(struct drm_crtc *crtc, bool has_frame,
 	tail = crc->tail;
 
 	if (CIRC_SPACE(head, tail, DRM_CRC_ENTRIES_NR) < 1) {
+		bool was_overflow = crc->overflow;
+
+		crc->overflow = true;
 		spin_unlock(&crc->lock);
-		DRM_ERROR("Overflow of CRC buffer, userspace reads too slow.\n");
+
+		if (!was_overflow)
+			DRM_ERROR("Overflow of CRC buffer, userspace reads too slow.\n");
+
 		return -ENOBUFS;
 	}
 
diff --git a/drivers/gpu/drm/drm_dp_cec.c b/drivers/gpu/drm/drm_dp_cec.c
new file mode 100644
index 000000000000..ddb1c5adebb9
--- /dev/null
+++ b/drivers/gpu/drm/drm_dp_cec.c
@@ -0,0 +1,428 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * DisplayPort CEC-Tunneling-over-AUX support
+ *
+ * Copyright 2018 Cisco Systems, Inc. and/or its affiliates. All rights reserved.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <drm/drm_dp_helper.h>
+#include <media/cec.h>
+
+/*
+ * Unfortunately it turns out that we have a chicken-and-egg situation
+ * here. Quite a few active (mini-)DP-to-HDMI or USB-C-to-HDMI adapters
+ * have a converter chip that supports CEC-Tunneling-over-AUX (usually the
+ * Parade PS176), but they do not wire up the CEC pin, thus making CEC
+ * useless.
+ *
+ * Sadly there is no way for this driver to know this. What happens is
+ * that a /dev/cecX device is created that is isolated and unable to see
+ * any of the other CEC devices. Quite literally the CEC wire is cut
+ * (or in this case, never connected in the first place).
+ *
+ * The reason so few adapters support this is that this tunneling protocol
+ * was never supported by any OS. So there was no easy way of testing it,
+ * and no incentive to correctly wire up the CEC pin.
+ *
+ * Hopefully by creating this driver it will be easier for vendors to
+ * finally fix their adapters and test the CEC functionality.
+ *
+ * I keep a list of known working adapters here:
+ *
+ * https://hverkuil.home.xs4all.nl/cec-status.txt
+ *
+ * Please mail me (hverkuil@xs4all.nl) if you find an adapter that works
+ * and is not yet listed there.
+ *
+ * Note that the current implementation does not support CEC over an MST hub.
+ * As far as I can see there is no mechanism defined in the DisplayPort
+ * standard to transport CEC interrupts over an MST device. It might be
+ * possible to do this through polling, but I have not been able to get that
+ * to work.
+ */
+
+/**
+ * DOC: dp cec helpers
+ *
+ * These functions take care of supporting the CEC-Tunneling-over-AUX
+ * feature of DisplayPort-to-HDMI adapters.
+ */
+
+/*
+ * When the EDID is unset because the HPD went low, then the CEC DPCD registers
+ * typically can no longer be read (true for a DP-to-HDMI adapter since it is
+ * powered by the HPD). However, some displays toggle the HPD off and on for a
+ * short period for one reason or another, and that would cause the CEC adapter
+ * to be removed and added again, even though nothing else changed.
+ *
+ * This module parameter sets a delay in seconds before the CEC adapter is
+ * actually unregistered. Only if the HPD does not return within that time will
+ * the CEC adapter be unregistered.
+ *
+ * If it is set to a value >= NEVER_UNREG_DELAY, then the CEC adapter will never
+ * be unregistered for as long as the connector remains registered.
+ *
+ * If it is set to 0, then the CEC adapter will be unregistered immediately as
+ * soon as the HPD disappears.
+ *
+ * The default is one second to prevent short HPD glitches from unregistering
+ * the CEC adapter.
+ *
+ * Note that for integrated HDMI branch devices that support CEC the DPCD
+ * registers remain available even if the HPD goes low since it is not powered
+ * by the HPD. In that case the CEC adapter will never be unregistered during
+ * the life time of the connector. At least, this is the theory since I do not
+ * have hardware with an integrated HDMI branch device that supports CEC.
+ */
+#define NEVER_UNREG_DELAY 1000
+static unsigned int drm_dp_cec_unregister_delay = 1;
+module_param(drm_dp_cec_unregister_delay, uint, 0600);
+MODULE_PARM_DESC(drm_dp_cec_unregister_delay,
+		 "CEC unregister delay in seconds, 0: no delay, >= 1000: never unregister");
+
+static int drm_dp_cec_adap_enable(struct cec_adapter *adap, bool enable)
+{
+	struct drm_dp_aux *aux = cec_get_drvdata(adap);
+	u32 val = enable ? DP_CEC_TUNNELING_ENABLE : 0;
+	ssize_t err = 0;
+
+	err = drm_dp_dpcd_writeb(aux, DP_CEC_TUNNELING_CONTROL, val);
+	return (enable && err < 0) ? err : 0;
+}
+
+static int drm_dp_cec_adap_log_addr(struct cec_adapter *adap, u8 addr)
+{
+	struct drm_dp_aux *aux = cec_get_drvdata(adap);
+	/* Bit 15 (logical address 15) should always be set */
+	u16 la_mask = 1 << CEC_LOG_ADDR_BROADCAST;
+	u8 mask[2];
+	ssize_t err;
+
+	if (addr != CEC_LOG_ADDR_INVALID)
+		la_mask |= adap->log_addrs.log_addr_mask | (1 << addr);
+	mask[0] = la_mask & 0xff;
+	mask[1] = la_mask >> 8;
+	err = drm_dp_dpcd_write(aux, DP_CEC_LOGICAL_ADDRESS_MASK, mask, 2);
+	return (addr != CEC_LOG_ADDR_INVALID && err < 0) ? err : 0;
+}
+
+static int drm_dp_cec_adap_transmit(struct cec_adapter *adap, u8 attempts,
+				    u32 signal_free_time, struct cec_msg *msg)
+{
+	struct drm_dp_aux *aux = cec_get_drvdata(adap);
+	unsigned int retries = min(5, attempts - 1);
+	ssize_t err;
+
+	err = drm_dp_dpcd_write(aux, DP_CEC_TX_MESSAGE_BUFFER,
+				msg->msg, msg->len);
+	if (err < 0)
+		return err;
+
+	err = drm_dp_dpcd_writeb(aux, DP_CEC_TX_MESSAGE_INFO,
+				 (msg->len - 1) | (retries << 4) |
+				 DP_CEC_TX_MESSAGE_SEND);
+	return err < 0 ? err : 0;
+}
+
+static int drm_dp_cec_adap_monitor_all_enable(struct cec_adapter *adap,
+					      bool enable)
+{
+	struct drm_dp_aux *aux = cec_get_drvdata(adap);
+	ssize_t err;
+	u8 val;
+
+	if (!(adap->capabilities & CEC_CAP_MONITOR_ALL))
+		return 0;
+
+	err = drm_dp_dpcd_readb(aux, DP_CEC_TUNNELING_CONTROL, &val);
+	if (err >= 0) {
+		if (enable)
+			val |= DP_CEC_SNOOPING_ENABLE;
+		else
+			val &= ~DP_CEC_SNOOPING_ENABLE;
+		err = drm_dp_dpcd_writeb(aux, DP_CEC_TUNNELING_CONTROL, val);
+	}
+	return (enable && err < 0) ? err : 0;
+}
+
+static void drm_dp_cec_adap_status(struct cec_adapter *adap,
+				   struct seq_file *file)
+{
+	struct drm_dp_aux *aux = cec_get_drvdata(adap);
+	struct drm_dp_desc desc;
+	struct drm_dp_dpcd_ident *id = &desc.ident;
+
+	if (drm_dp_read_desc(aux, &desc, true))
+		return;
+	seq_printf(file, "OUI: %*pdH\n",
+		   (int)sizeof(id->oui), id->oui);
+	seq_printf(file, "ID: %*pE\n",
+		   (int)strnlen(id->device_id, sizeof(id->device_id)),
+		   id->device_id);
+	seq_printf(file, "HW Rev: %d.%d\n", id->hw_rev >> 4, id->hw_rev & 0xf);
+	/*
+	 * Show this both in decimal and hex: at least one vendor
+	 * always reports this in hex.
+	 */
+	seq_printf(file, "FW/SW Rev: %d.%d (0x%02x.0x%02x)\n",
+		   id->sw_major_rev, id->sw_minor_rev,
+		   id->sw_major_rev, id->sw_minor_rev);
+}
+
+static const struct cec_adap_ops drm_dp_cec_adap_ops = {
+	.adap_enable = drm_dp_cec_adap_enable,
+	.adap_log_addr = drm_dp_cec_adap_log_addr,
+	.adap_transmit = drm_dp_cec_adap_transmit,
+	.adap_monitor_all_enable = drm_dp_cec_adap_monitor_all_enable,
+	.adap_status = drm_dp_cec_adap_status,
+};
+
+static int drm_dp_cec_received(struct drm_dp_aux *aux)
+{
+	struct cec_adapter *adap = aux->cec.adap;
+	struct cec_msg msg;
+	u8 rx_msg_info;
+	ssize_t err;
+
+	err = drm_dp_dpcd_readb(aux, DP_CEC_RX_MESSAGE_INFO, &rx_msg_info);
+	if (err < 0)
+		return err;
+
+	if (!(rx_msg_info & DP_CEC_RX_MESSAGE_ENDED))
+		return 0;
+
+	msg.len = (rx_msg_info & DP_CEC_RX_MESSAGE_LEN_MASK) + 1;
+	err = drm_dp_dpcd_read(aux, DP_CEC_RX_MESSAGE_BUFFER, msg.msg, msg.len);
+	if (err < 0)
+		return err;
+
+	cec_received_msg(adap, &msg);
+	return 0;
+}
+
+static void drm_dp_cec_handle_irq(struct drm_dp_aux *aux)
+{
+	struct cec_adapter *adap = aux->cec.adap;
+	u8 flags;
+
+	if (drm_dp_dpcd_readb(aux, DP_CEC_TUNNELING_IRQ_FLAGS, &flags) < 0)
+		return;
+
+	if (flags & DP_CEC_RX_MESSAGE_INFO_VALID)
+		drm_dp_cec_received(aux);
+
+	if (flags & DP_CEC_TX_MESSAGE_SENT)
+		cec_transmit_attempt_done(adap, CEC_TX_STATUS_OK);
+	else if (flags & DP_CEC_TX_LINE_ERROR)
+		cec_transmit_attempt_done(adap, CEC_TX_STATUS_ERROR |
+						CEC_TX_STATUS_MAX_RETRIES);
+	else if (flags &
+		 (DP_CEC_TX_ADDRESS_NACK_ERROR | DP_CEC_TX_DATA_NACK_ERROR))
+		cec_transmit_attempt_done(adap, CEC_TX_STATUS_NACK |
+						CEC_TX_STATUS_MAX_RETRIES);
+	drm_dp_dpcd_writeb(aux, DP_CEC_TUNNELING_IRQ_FLAGS, flags);
+}
+
+/**
+ * drm_dp_cec_irq() - handle CEC interrupt, if any
+ * @aux: DisplayPort AUX channel
+ *
+ * Should be called when handling an IRQ_HPD request. If CEC-tunneling-over-AUX
+ * is present, then it will check for a CEC_IRQ and handle it accordingly.
+ */
+void drm_dp_cec_irq(struct drm_dp_aux *aux)
+{
+	u8 cec_irq;
+	int ret;
+
+	mutex_lock(&aux->cec.lock);
+	if (!aux->cec.adap)
+		goto unlock;
+
+	ret = drm_dp_dpcd_readb(aux, DP_DEVICE_SERVICE_IRQ_VECTOR_ESI1,
+				&cec_irq);
+	if (ret < 0 || !(cec_irq & DP_CEC_IRQ))
+		goto unlock;
+
+	drm_dp_cec_handle_irq(aux);
+	drm_dp_dpcd_writeb(aux, DP_DEVICE_SERVICE_IRQ_VECTOR_ESI1, DP_CEC_IRQ);
+unlock:
+	mutex_unlock(&aux->cec.lock);
+}
+EXPORT_SYMBOL(drm_dp_cec_irq);
+
+static bool drm_dp_cec_cap(struct drm_dp_aux *aux, u8 *cec_cap)
+{
+	u8 cap = 0;
+
+	if (drm_dp_dpcd_readb(aux, DP_CEC_TUNNELING_CAPABILITY, &cap) != 1 ||
+	    !(cap & DP_CEC_TUNNELING_CAPABLE))
+		return false;
+	if (cec_cap)
+		*cec_cap = cap;
+	return true;
+}
+
+/*
+ * Called if the HPD was low for more than drm_dp_cec_unregister_delay
+ * seconds. This unregisters the CEC adapter.
+ */
+static void drm_dp_cec_unregister_work(struct work_struct *work)
+{
+	struct drm_dp_aux *aux = container_of(work, struct drm_dp_aux,
+					      cec.unregister_work.work);
+
+	mutex_lock(&aux->cec.lock);
+	cec_unregister_adapter(aux->cec.adap);
+	aux->cec.adap = NULL;
+	mutex_unlock(&aux->cec.lock);
+}
+
+/*
+ * A new EDID is set. If there is no CEC adapter, then create one. If
+ * there was a CEC adapter, then check if the CEC adapter properties
+ * were unchanged and just update the CEC physical address. Otherwise
+ * unregister the old CEC adapter and create a new one.
+ */
+void drm_dp_cec_set_edid(struct drm_dp_aux *aux, const struct edid *edid)
+{
+	u32 cec_caps = CEC_CAP_DEFAULTS | CEC_CAP_NEEDS_HPD;
+	unsigned int num_las = 1;
+	u8 cap;
+
+#ifndef CONFIG_MEDIA_CEC_RC
+	/*
+	 * CEC_CAP_RC is part of CEC_CAP_DEFAULTS, but it is stripped by
+	 * cec_allocate_adapter() if CONFIG_MEDIA_CEC_RC is undefined.
+	 *
+	 * Do this here as well to ensure the tests against cec_caps are
+	 * correct.
+	 */
+	cec_caps &= ~CEC_CAP_RC;
+#endif
+	cancel_delayed_work_sync(&aux->cec.unregister_work);
+
+	mutex_lock(&aux->cec.lock);
+	if (!drm_dp_cec_cap(aux, &cap)) {
+		/* CEC is not supported, unregister any existing adapter */
+		cec_unregister_adapter(aux->cec.adap);
+		aux->cec.adap = NULL;
+		goto unlock;
+	}
+
+	if (cap & DP_CEC_SNOOPING_CAPABLE)
+		cec_caps |= CEC_CAP_MONITOR_ALL;
+	if (cap & DP_CEC_MULTIPLE_LA_CAPABLE)
+		num_las = CEC_MAX_LOG_ADDRS;
+
+	if (aux->cec.adap) {
+		if (aux->cec.adap->capabilities == cec_caps &&
+		    aux->cec.adap->available_log_addrs == num_las) {
+			/* Unchanged, so just set the phys addr */
+			cec_s_phys_addr_from_edid(aux->cec.adap, edid);
+			goto unlock;
+		}
+		/*
+		 * The capabilities changed, so unregister the old
+		 * adapter first.
+		 */
+		cec_unregister_adapter(aux->cec.adap);
+	}
+
+	/* Create a new adapter */
+	aux->cec.adap = cec_allocate_adapter(&drm_dp_cec_adap_ops,
+					     aux, aux->cec.name, cec_caps,
+					     num_las);
+	if (IS_ERR(aux->cec.adap)) {
+		aux->cec.adap = NULL;
+		goto unlock;
+	}
+	if (cec_register_adapter(aux->cec.adap, aux->cec.parent)) {
+		cec_delete_adapter(aux->cec.adap);
+		aux->cec.adap = NULL;
+	} else {
+		/*
+		 * Update the phys addr for the new CEC adapter. When called
+		 * from drm_dp_cec_register_connector() edid == NULL, so in
+		 * that case the phys addr is just invalidated.
+		 */
+		cec_s_phys_addr_from_edid(aux->cec.adap, edid);
+	}
+unlock:
+	mutex_unlock(&aux->cec.lock);
+}
+EXPORT_SYMBOL(drm_dp_cec_set_edid);
+
+/*
+ * The EDID disappeared (likely because of the HPD going down).
+ */
+void drm_dp_cec_unset_edid(struct drm_dp_aux *aux)
+{
+	cancel_delayed_work_sync(&aux->cec.unregister_work);
+
+	mutex_lock(&aux->cec.lock);
+	if (!aux->cec.adap)
+		goto unlock;
+
+	cec_phys_addr_invalidate(aux->cec.adap);
+	/*
+	 * We're done if we want to keep the CEC device
+	 * (drm_dp_cec_unregister_delay is >= NEVER_UNREG_DELAY) or if the
+	 * DPCD still indicates the CEC capability (expected for an integrated
+	 * HDMI branch device).
+	 */
+	if (drm_dp_cec_unregister_delay < NEVER_UNREG_DELAY &&
+	    !drm_dp_cec_cap(aux, NULL)) {
+		/*
+		 * Unregister the CEC adapter after drm_dp_cec_unregister_delay
+		 * seconds. This to debounce short HPD off-and-on cycles from
+		 * displays.
+		 */
+		schedule_delayed_work(&aux->cec.unregister_work,
+				      drm_dp_cec_unregister_delay * HZ);
+	}
+unlock:
+	mutex_unlock(&aux->cec.lock);
+}
+EXPORT_SYMBOL(drm_dp_cec_unset_edid);
+
+/**
+ * drm_dp_cec_register_connector() - register a new connector
+ * @aux: DisplayPort AUX channel
+ * @name: name of the CEC device
+ * @parent: parent device
+ *
+ * A new connector was registered with associated CEC adapter name and
+ * CEC adapter parent device. After registering the name and parent
+ * drm_dp_cec_set_edid() is called to check if the connector supports
+ * CEC and to register a CEC adapter if that is the case.
+ */
+void drm_dp_cec_register_connector(struct drm_dp_aux *aux, const char *name,
+				   struct device *parent)
+{
+	WARN_ON(aux->cec.adap);
+	aux->cec.name = name;
+	aux->cec.parent = parent;
+	INIT_DELAYED_WORK(&aux->cec.unregister_work,
+			  drm_dp_cec_unregister_work);
+
+	drm_dp_cec_set_edid(aux, NULL);
+}
+EXPORT_SYMBOL(drm_dp_cec_register_connector);
+
+/**
+ * drm_dp_cec_unregister_connector() - unregister the CEC adapter, if any
+ * @aux: DisplayPort AUX channel
+ */
+void drm_dp_cec_unregister_connector(struct drm_dp_aux *aux)
+{
+	if (!aux->cec.adap)
+		return;
+	cancel_delayed_work_sync(&aux->cec.unregister_work);
+	cec_unregister_adapter(aux->cec.adap);
+	aux->cec.adap = NULL;
+}
+EXPORT_SYMBOL(drm_dp_cec_unregister_connector);
diff --git a/drivers/gpu/drm/drm_dp_helper.c b/drivers/gpu/drm/drm_dp_helper.c
index ffe14ec3e7f2..0cccbcb2d03e 100644
--- a/drivers/gpu/drm/drm_dp_helper.c
+++ b/drivers/gpu/drm/drm_dp_helper.c
@@ -119,18 +119,32 @@ u8 drm_dp_get_adjust_request_pre_emphasis(const u8 link_status[DP_LINK_STATUS_SI
 EXPORT_SYMBOL(drm_dp_get_adjust_request_pre_emphasis);
 
 void drm_dp_link_train_clock_recovery_delay(const u8 dpcd[DP_RECEIVER_CAP_SIZE]) {
-	if (dpcd[DP_TRAINING_AUX_RD_INTERVAL] == 0)
+	int rd_interval = dpcd[DP_TRAINING_AUX_RD_INTERVAL] &
+			  DP_TRAINING_AUX_RD_MASK;
+
+	if (rd_interval > 4)
+		DRM_DEBUG_KMS("AUX interval %d, out of range (max 4)\n",
+			      rd_interval);
+
+	if (rd_interval == 0 || dpcd[DP_DPCD_REV] >= DP_DPCD_REV_14)
 		udelay(100);
 	else
-		mdelay(dpcd[DP_TRAINING_AUX_RD_INTERVAL] * 4);
+		mdelay(rd_interval * 4);
 }
 EXPORT_SYMBOL(drm_dp_link_train_clock_recovery_delay);
 
 void drm_dp_link_train_channel_eq_delay(const u8 dpcd[DP_RECEIVER_CAP_SIZE]) {
-	if (dpcd[DP_TRAINING_AUX_RD_INTERVAL] == 0)
+	int rd_interval = dpcd[DP_TRAINING_AUX_RD_INTERVAL] &
+			  DP_TRAINING_AUX_RD_MASK;
+
+	if (rd_interval > 4)
+		DRM_DEBUG_KMS("AUX interval %d, out of range (max 4)\n",
+			      rd_interval);
+
+	if (rd_interval == 0)
 		udelay(400);
 	else
-		mdelay(dpcd[DP_TRAINING_AUX_RD_INTERVAL] * 4);
+		mdelay(rd_interval * 4);
 }
 EXPORT_SYMBOL(drm_dp_link_train_channel_eq_delay);
 
@@ -171,6 +185,20 @@ EXPORT_SYMBOL(drm_dp_bw_code_to_link_rate);
 
 #define AUX_RETRY_INTERVAL 500 /* us */
 
+static inline void
+drm_dp_dump_access(const struct drm_dp_aux *aux,
+		   u8 request, uint offset, void *buffer, int ret)
+{
+	const char *arrow = request == DP_AUX_NATIVE_READ ? "->" : "<-";
+
+	if (ret > 0)
+		drm_dbg(DRM_UT_DP, "%s: 0x%05x AUX %s (ret=%3d) %*ph\n",
+			aux->name, offset, arrow, ret, min(ret, 20), buffer);
+	else
+		drm_dbg(DRM_UT_DP, "%s: 0x%05x AUX %s (ret=%3d)\n",
+			aux->name, offset, arrow, ret);
+}
+
 /**
  * DOC: dp helpers
  *
@@ -274,10 +302,14 @@ ssize_t drm_dp_dpcd_read(struct drm_dp_aux *aux, unsigned int offset,
 	ret = drm_dp_dpcd_access(aux, DP_AUX_NATIVE_READ, DP_DPCD_REV, buffer,
 				 1);
 	if (ret != 1)
-		return ret;
+		goto out;
 
-	return drm_dp_dpcd_access(aux, DP_AUX_NATIVE_READ, offset, buffer,
-				  size);
+	ret = drm_dp_dpcd_access(aux, DP_AUX_NATIVE_READ, offset, buffer,
+				 size);
+
+out:
+	drm_dp_dump_access(aux, DP_AUX_NATIVE_READ, offset, buffer, ret);
+	return ret;
 }
 EXPORT_SYMBOL(drm_dp_dpcd_read);
 
@@ -298,8 +330,12 @@ EXPORT_SYMBOL(drm_dp_dpcd_read);
 ssize_t drm_dp_dpcd_write(struct drm_dp_aux *aux, unsigned int offset,
 			  void *buffer, size_t size)
 {
-	return drm_dp_dpcd_access(aux, DP_AUX_NATIVE_WRITE, offset, buffer,
-				  size);
+	int ret;
+
+	ret = drm_dp_dpcd_access(aux, DP_AUX_NATIVE_WRITE, offset, buffer,
+				 size);
+	drm_dp_dump_access(aux, DP_AUX_NATIVE_WRITE, offset, buffer, ret);
+	return ret;
 }
 EXPORT_SYMBOL(drm_dp_dpcd_write);
 
@@ -1073,6 +1109,7 @@ static void drm_dp_aux_crc_work(struct work_struct *work)
 void drm_dp_aux_init(struct drm_dp_aux *aux)
 {
 	mutex_init(&aux->hw_mutex);
+	mutex_init(&aux->cec.lock);
 	INIT_WORK(&aux->crc_work, drm_dp_aux_crc_work);
 
 	aux->ddc.algo = &drm_dp_i2c_algo;
@@ -1145,6 +1182,7 @@ int drm_dp_psr_setup_time(const u8 psr_cap[EDP_PSR_RECEIVER_CAP_SIZE])
 	static const u16 psr_setup_time_us[] = {
 		PSR_SETUP_TIME(330),
 		PSR_SETUP_TIME(275),
+		PSR_SETUP_TIME(220),
 		PSR_SETUP_TIME(165),
 		PSR_SETUP_TIME(110),
 		PSR_SETUP_TIME(55),
diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c b/drivers/gpu/drm/drm_dp_mst_topology.c
index 658830620ca3..7780567aa669 100644
--- a/drivers/gpu/drm/drm_dp_mst_topology.c
+++ b/drivers/gpu/drm/drm_dp_mst_topology.c
@@ -1215,7 +1215,7 @@ static void drm_dp_add_port(struct drm_dp_mst_branch *mstb,
 		     port->pdt == DP_PEER_DEVICE_SST_SINK) &&
 		    port->port_num >= DP_MST_LOGICAL_PORT_0) {
 			port->cached_edid = drm_get_edid(port->connector, &port->aux.ddc);
-			drm_mode_connector_set_tile_property(port->connector);
+			drm_connector_set_tile_property(port->connector);
 		}
 		(*mstb->mgr->cbs->register_connector)(port->connector);
 	}
@@ -2559,7 +2559,7 @@ struct edid *drm_dp_mst_get_edid(struct drm_connector *connector, struct drm_dp_
 		edid = drm_edid_duplicate(port->cached_edid);
 	else {
 		edid = drm_get_edid(connector, &port->aux.ddc);
-		drm_mode_connector_set_tile_property(connector);
+		drm_connector_set_tile_property(connector);
 	}
 	port->has_audio = drm_detect_monitor_audio(edid);
 	drm_dp_put_port(port);
diff --git a/drivers/gpu/drm/drm_drv.c b/drivers/gpu/drm/drm_drv.c
index 32a83b41ab61..ea4941da9b27 100644
--- a/drivers/gpu/drm/drm_drv.c
+++ b/drivers/gpu/drm/drm_drv.c
@@ -34,6 +34,7 @@
 #include <linux/slab.h>
 #include <linux/srcu.h>
 
+#include <drm/drm_client.h>
 #include <drm/drm_drv.h>
 #include <drm/drmP.h>
 
@@ -53,13 +54,14 @@ MODULE_AUTHOR("Gareth Hughes, Leif Delgass, José Fonseca, Jon Smirl");
 MODULE_DESCRIPTION("DRM shared core routines");
 MODULE_LICENSE("GPL and additional rights");
 MODULE_PARM_DESC(debug, "Enable debug output, where each bit enables a debug category.\n"
-"\t\tBit 0 (0x01) will enable CORE messages (drm core code)\n"
-"\t\tBit 1 (0x02) will enable DRIVER messages (drm controller code)\n"
-"\t\tBit 2 (0x04) will enable KMS messages (modesetting code)\n"
-"\t\tBit 3 (0x08) will enable PRIME messages (prime code)\n"
-"\t\tBit 4 (0x10) will enable ATOMIC messages (atomic code)\n"
-"\t\tBit 5 (0x20) will enable VBL messages (vblank code)\n"
-"\t\tBit 7 (0x80) will enable LEASE messages (leasing code)");
+"\t\tBit 0 (0x01)  will enable CORE messages (drm core code)\n"
+"\t\tBit 1 (0x02)  will enable DRIVER messages (drm controller code)\n"
+"\t\tBit 2 (0x04)  will enable KMS messages (modesetting code)\n"
+"\t\tBit 3 (0x08)  will enable PRIME messages (prime code)\n"
+"\t\tBit 4 (0x10)  will enable ATOMIC messages (atomic code)\n"
+"\t\tBit 5 (0x20)  will enable VBL messages (vblank code)\n"
+"\t\tBit 7 (0x80)  will enable LEASE messages (leasing code)\n"
+"\t\tBit 8 (0x100) will enable DP messages (displayport code)");
 module_param_named(debug, drm_debug, int, 0600);
 
 static DEFINE_SPINLOCK(drm_minor_lock);
@@ -99,8 +101,6 @@ static struct drm_minor **drm_minor_get_slot(struct drm_device *dev,
 		return &dev->primary;
 	case DRM_MINOR_RENDER:
 		return &dev->render;
-	case DRM_MINOR_CONTROL:
-		return &dev->control;
 	default:
 		BUG();
 	}
@@ -371,13 +371,6 @@ EXPORT_SYMBOL(drm_dev_exit);
  */
 void drm_dev_unplug(struct drm_device *dev)
 {
-	drm_dev_unregister(dev);
-
-	mutex_lock(&drm_global_mutex);
-	if (dev->open_count == 0)
-		drm_dev_put(dev);
-	mutex_unlock(&drm_global_mutex);
-
 	/*
 	 * After synchronizing any critical read section is guaranteed to see
 	 * the new value of ->unplugged, and any critical section which might
@@ -386,6 +379,13 @@ void drm_dev_unplug(struct drm_device *dev)
 	 */
 	dev->unplugged = true;
 	synchronize_srcu(&drm_unplug_srcu);
+
+	drm_dev_unregister(dev);
+
+	mutex_lock(&drm_global_mutex);
+	if (dev->open_count == 0)
+		drm_dev_put(dev);
+	mutex_unlock(&drm_global_mutex);
 }
 EXPORT_SYMBOL(drm_dev_unplug);
 
@@ -507,6 +507,8 @@ int drm_dev_init(struct drm_device *dev,
 	dev->driver = driver;
 
 	INIT_LIST_HEAD(&dev->filelist);
+	INIT_LIST_HEAD(&dev->filelist_internal);
+	INIT_LIST_HEAD(&dev->clientlist);
 	INIT_LIST_HEAD(&dev->ctxlist);
 	INIT_LIST_HEAD(&dev->vmalist);
 	INIT_LIST_HEAD(&dev->maplist);
@@ -516,6 +518,7 @@ int drm_dev_init(struct drm_device *dev,
 	spin_lock_init(&dev->event_lock);
 	mutex_init(&dev->struct_mutex);
 	mutex_init(&dev->filelist_mutex);
+	mutex_init(&dev->clientlist_mutex);
 	mutex_init(&dev->ctxlist_mutex);
 	mutex_init(&dev->master_mutex);
 
@@ -567,11 +570,11 @@ err_ctxbitmap:
 err_minors:
 	drm_minor_free(dev, DRM_MINOR_PRIMARY);
 	drm_minor_free(dev, DRM_MINOR_RENDER);
-	drm_minor_free(dev, DRM_MINOR_CONTROL);
 	drm_fs_inode_free(dev->anon_inode);
 err_free:
 	mutex_destroy(&dev->master_mutex);
 	mutex_destroy(&dev->ctxlist_mutex);
+	mutex_destroy(&dev->clientlist_mutex);
 	mutex_destroy(&dev->filelist_mutex);
 	mutex_destroy(&dev->struct_mutex);
 	return ret;
@@ -603,10 +606,10 @@ void drm_dev_fini(struct drm_device *dev)
 
 	drm_minor_free(dev, DRM_MINOR_PRIMARY);
 	drm_minor_free(dev, DRM_MINOR_RENDER);
-	drm_minor_free(dev, DRM_MINOR_CONTROL);
 
 	mutex_destroy(&dev->master_mutex);
 	mutex_destroy(&dev->ctxlist_mutex);
+	mutex_destroy(&dev->clientlist_mutex);
 	mutex_destroy(&dev->filelist_mutex);
 	mutex_destroy(&dev->struct_mutex);
 	kfree(dev->unique);
@@ -760,7 +763,7 @@ static void remove_compat_control_link(struct drm_device *dev)
 	if (!minor)
 		return;
 
-	name = kasprintf(GFP_KERNEL, "controlD%d", minor->index);
+	name = kasprintf(GFP_KERNEL, "controlD%d", minor->index + 64);
 	if (!name)
 		return;
 
@@ -796,10 +799,6 @@ int drm_dev_register(struct drm_device *dev, unsigned long flags)
 
 	mutex_lock(&drm_global_mutex);
 
-	ret = drm_minor_register(dev, DRM_MINOR_CONTROL);
-	if (ret)
-		goto err_minors;
-
 	ret = drm_minor_register(dev, DRM_MINOR_RENDER);
 	if (ret)
 		goto err_minors;
@@ -837,7 +836,6 @@ err_minors:
 	remove_compat_control_link(dev);
 	drm_minor_unregister(dev, DRM_MINOR_PRIMARY);
 	drm_minor_unregister(dev, DRM_MINOR_RENDER);
-	drm_minor_unregister(dev, DRM_MINOR_CONTROL);
 out_unlock:
 	mutex_unlock(&drm_global_mutex);
 	return ret;
@@ -867,6 +865,8 @@ void drm_dev_unregister(struct drm_device *dev)
 
 	dev->registered = false;
 
+	drm_client_dev_unregister(dev);
+
 	if (drm_core_check_feature(dev, DRIVER_MODESET))
 		drm_modeset_unregister_all(dev);
 
@@ -882,7 +882,6 @@ void drm_dev_unregister(struct drm_device *dev)
 	remove_compat_control_link(dev);
 	drm_minor_unregister(dev, DRM_MINOR_PRIMARY);
 	drm_minor_unregister(dev, DRM_MINOR_RENDER);
-	drm_minor_unregister(dev, DRM_MINOR_CONTROL);
 }
 EXPORT_SYMBOL(drm_dev_unregister);
 
diff --git a/drivers/gpu/drm/drm_dumb_buffers.c b/drivers/gpu/drm/drm_dumb_buffers.c
index 39ac15ce4702..81dfdd33753a 100644
--- a/drivers/gpu/drm/drm_dumb_buffers.c
+++ b/drivers/gpu/drm/drm_dumb_buffers.c
@@ -53,10 +53,10 @@
  * a hardware-specific ioctl to allocate suitable buffer objects.
  */
 
-int drm_mode_create_dumb_ioctl(struct drm_device *dev,
-			       void *data, struct drm_file *file_priv)
+int drm_mode_create_dumb(struct drm_device *dev,
+			 struct drm_mode_create_dumb *args,
+			 struct drm_file *file_priv)
 {
-	struct drm_mode_create_dumb *args = data;
 	u32 cpp, stride, size;
 
 	if (!dev->driver->dumb_create)
@@ -65,12 +65,13 @@ int drm_mode_create_dumb_ioctl(struct drm_device *dev,
 		return -EINVAL;
 
 	/* overflow checks for 32bit size calculations */
-	/* NOTE: DIV_ROUND_UP() can overflow */
+	if (args->bpp > U32_MAX - 8)
+		return -EINVAL;
 	cpp = DIV_ROUND_UP(args->bpp, 8);
-	if (!cpp || cpp > 0xffffffffU / args->width)
+	if (cpp > U32_MAX / args->width)
 		return -EINVAL;
 	stride = cpp * args->width;
-	if (args->height > 0xffffffffU / stride)
+	if (args->height > U32_MAX / stride)
 		return -EINVAL;
 
 	/* test for wrap-around */
@@ -91,6 +92,12 @@ int drm_mode_create_dumb_ioctl(struct drm_device *dev,
 	return dev->driver->dumb_create(file_priv, dev, args);
 }
 
+int drm_mode_create_dumb_ioctl(struct drm_device *dev,
+			       void *data, struct drm_file *file_priv)
+{
+	return drm_mode_create_dumb(dev, data, file_priv);
+}
+
 /**
  * drm_mode_mmap_dumb_ioctl - create an mmap offset for a dumb backing storage buffer
  * @dev: DRM device
@@ -122,17 +129,22 @@ int drm_mode_mmap_dumb_ioctl(struct drm_device *dev,
 					       &args->offset);
 }
 
-int drm_mode_destroy_dumb_ioctl(struct drm_device *dev,
-				void *data, struct drm_file *file_priv)
+int drm_mode_destroy_dumb(struct drm_device *dev, u32 handle,
+			  struct drm_file *file_priv)
 {
-	struct drm_mode_destroy_dumb *args = data;
-
 	if (!dev->driver->dumb_create)
 		return -ENOSYS;
 
 	if (dev->driver->dumb_destroy)
-		return dev->driver->dumb_destroy(file_priv, dev, args->handle);
+		return dev->driver->dumb_destroy(file_priv, dev, handle);
 	else
-		return drm_gem_dumb_destroy(file_priv, dev, args->handle);
+		return drm_gem_dumb_destroy(file_priv, dev, handle);
 }
 
+int drm_mode_destroy_dumb_ioctl(struct drm_device *dev,
+				void *data, struct drm_file *file_priv)
+{
+	struct drm_mode_destroy_dumb *args = data;
+
+	return drm_mode_destroy_dumb(dev, args->handle, file_priv);
+}
diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c
index 08d33b48b14a..5dc742b27ca0 100644
--- a/drivers/gpu/drm/drm_edid.c
+++ b/drivers/gpu/drm/drm_edid.c
@@ -163,8 +163,9 @@ static const struct edid_quirk {
 	/* Rotel RSX-1058 forwards sink's EDID but only does HDMI 1.1*/
 	{ "ETR", 13896, EDID_QUIRK_FORCE_8BPC },
 
-	/* HTC Vive VR Headset */
+	/* HTC Vive and Vive Pro VR Headsets */
 	{ "HVR", 0xaa01, EDID_QUIRK_NON_DESKTOP },
+	{ "HVR", 0xaa02, EDID_QUIRK_NON_DESKTOP },
 
 	/* Oculus Rift DK1, DK2, and CV1 VR Headsets */
 	{ "OVR", 0x0001, EDID_QUIRK_NON_DESKTOP },
@@ -687,562 +688,562 @@ static const struct minimode extra_modes[] = {
 static const struct drm_display_mode edid_cea_modes[] = {
 	/* 0 - dummy, VICs start at 1 */
 	{ },
-	/* 1 - 640x480@60Hz */
+	/* 1 - 640x480@60Hz 4:3 */
 	{ DRM_MODE("640x480", DRM_MODE_TYPE_DRIVER, 25175, 640, 656,
 		   752, 800, 0, 480, 490, 492, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 2 - 720x480@60Hz */
+	/* 2 - 720x480@60Hz 4:3 */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 27000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 3 - 720x480@60Hz */
+	/* 3 - 720x480@60Hz 16:9 */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 27000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 4 - 1280x720@60Hz */
+	/* 4 - 1280x720@60Hz 16:9 */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 1390,
 		   1430, 1650, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 5 - 1920x1080i@60Hz */
+	/* 5 - 1920x1080i@60Hz 16:9 */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1094, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
-			DRM_MODE_FLAG_INTERLACE),
+		   DRM_MODE_FLAG_INTERLACE),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 6 - 720(1440)x480i@60Hz */
+	/* 6 - 720(1440)x480i@60Hz 4:3 */
 	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 13500, 720, 739,
 		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 7 - 720(1440)x480i@60Hz */
+	/* 7 - 720(1440)x480i@60Hz 16:9 */
 	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 13500, 720, 739,
 		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 8 - 720(1440)x240@60Hz */
+	/* 8 - 720(1440)x240@60Hz 4:3 */
 	{ DRM_MODE("720x240", DRM_MODE_TYPE_DRIVER, 13500, 720, 739,
 		   801, 858, 0, 240, 244, 247, 262, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 9 - 720(1440)x240@60Hz */
+	/* 9 - 720(1440)x240@60Hz 16:9 */
 	{ DRM_MODE("720x240", DRM_MODE_TYPE_DRIVER, 13500, 720, 739,
 		   801, 858, 0, 240, 244, 247, 262, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 10 - 2880x480i@60Hz */
+	/* 10 - 2880x480i@60Hz 4:3 */
 	{ DRM_MODE("2880x480i", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2956,
 		   3204, 3432, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE),
+		   DRM_MODE_FLAG_INTERLACE),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 11 - 2880x480i@60Hz */
+	/* 11 - 2880x480i@60Hz 16:9 */
 	{ DRM_MODE("2880x480i", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2956,
 		   3204, 3432, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE),
+		   DRM_MODE_FLAG_INTERLACE),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 12 - 2880x240@60Hz */
+	/* 12 - 2880x240@60Hz 4:3 */
 	{ DRM_MODE("2880x240", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2956,
 		   3204, 3432, 0, 240, 244, 247, 262, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 13 - 2880x240@60Hz */
+	/* 13 - 2880x240@60Hz 16:9 */
 	{ DRM_MODE("2880x240", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2956,
 		   3204, 3432, 0, 240, 244, 247, 262, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 14 - 1440x480@60Hz */
+	/* 14 - 1440x480@60Hz 4:3 */
 	{ DRM_MODE("1440x480", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1472,
 		   1596, 1716, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 15 - 1440x480@60Hz */
+	/* 15 - 1440x480@60Hz 16:9 */
 	{ DRM_MODE("1440x480", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1472,
 		   1596, 1716, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 16 - 1920x1080@60Hz */
+	/* 16 - 1920x1080@60Hz 16:9 */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 17 - 720x576@50Hz */
+	/* 17 - 720x576@50Hz 4:3 */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 27000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 18 - 720x576@50Hz */
+	/* 18 - 720x576@50Hz 16:9 */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 27000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 19 - 1280x720@50Hz */
+	/* 19 - 1280x720@50Hz 16:9 */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 1720,
 		   1760, 1980, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 20 - 1920x1080i@50Hz */
+	/* 20 - 1920x1080i@50Hz 16:9 */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1094, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
-			DRM_MODE_FLAG_INTERLACE),
+		   DRM_MODE_FLAG_INTERLACE),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 21 - 720(1440)x576i@50Hz */
+	/* 21 - 720(1440)x576i@50Hz 4:3 */
 	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 13500, 720, 732,
 		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 22 - 720(1440)x576i@50Hz */
+	/* 22 - 720(1440)x576i@50Hz 16:9 */
 	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 13500, 720, 732,
 		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 23 - 720(1440)x288@50Hz */
+	/* 23 - 720(1440)x288@50Hz 4:3 */
 	{ DRM_MODE("720x288", DRM_MODE_TYPE_DRIVER, 13500, 720, 732,
 		   795, 864, 0, 288, 290, 293, 312, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 24 - 720(1440)x288@50Hz */
+	/* 24 - 720(1440)x288@50Hz 16:9 */
 	{ DRM_MODE("720x288", DRM_MODE_TYPE_DRIVER, 13500, 720, 732,
 		   795, 864, 0, 288, 290, 293, 312, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 25 - 2880x576i@50Hz */
+	/* 25 - 2880x576i@50Hz 4:3 */
 	{ DRM_MODE("2880x576i", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2928,
 		   3180, 3456, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE),
+		   DRM_MODE_FLAG_INTERLACE),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 26 - 2880x576i@50Hz */
+	/* 26 - 2880x576i@50Hz 16:9 */
 	{ DRM_MODE("2880x576i", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2928,
 		   3180, 3456, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE),
+		   DRM_MODE_FLAG_INTERLACE),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 27 - 2880x288@50Hz */
+	/* 27 - 2880x288@50Hz 4:3 */
 	{ DRM_MODE("2880x288", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2928,
 		   3180, 3456, 0, 288, 290, 293, 312, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 28 - 2880x288@50Hz */
+	/* 28 - 2880x288@50Hz 16:9 */
 	{ DRM_MODE("2880x288", DRM_MODE_TYPE_DRIVER, 54000, 2880, 2928,
 		   3180, 3456, 0, 288, 290, 293, 312, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 29 - 1440x576@50Hz */
+	/* 29 - 1440x576@50Hz 4:3 */
 	{ DRM_MODE("1440x576", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1464,
 		   1592, 1728, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 30 - 1440x576@50Hz */
+	/* 30 - 1440x576@50Hz 16:9 */
 	{ DRM_MODE("1440x576", DRM_MODE_TYPE_DRIVER, 54000, 1440, 1464,
 		   1592, 1728, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 31 - 1920x1080@50Hz */
+	/* 31 - 1920x1080@50Hz 16:9 */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 32 - 1920x1080@24Hz */
+	/* 32 - 1920x1080@24Hz 16:9 */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2558,
 		   2602, 2750, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 24, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 33 - 1920x1080@25Hz */
+	/* 33 - 1920x1080@25Hz 16:9 */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 25, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 34 - 1920x1080@30Hz */
+	/* 34 - 1920x1080@30Hz 16:9 */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 30, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 35 - 2880x480@60Hz */
+	/* 35 - 2880x480@60Hz 4:3 */
 	{ DRM_MODE("2880x480", DRM_MODE_TYPE_DRIVER, 108000, 2880, 2944,
 		   3192, 3432, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 36 - 2880x480@60Hz */
+	/* 36 - 2880x480@60Hz 16:9 */
 	{ DRM_MODE("2880x480", DRM_MODE_TYPE_DRIVER, 108000, 2880, 2944,
 		   3192, 3432, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 37 - 2880x576@50Hz */
+	/* 37 - 2880x576@50Hz 4:3 */
 	{ DRM_MODE("2880x576", DRM_MODE_TYPE_DRIVER, 108000, 2880, 2928,
 		   3184, 3456, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 38 - 2880x576@50Hz */
+	/* 38 - 2880x576@50Hz 16:9 */
 	{ DRM_MODE("2880x576", DRM_MODE_TYPE_DRIVER, 108000, 2880, 2928,
 		   3184, 3456, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 39 - 1920x1080i@50Hz */
+	/* 39 - 1920x1080i@50Hz 16:9 */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 72000, 1920, 1952,
 		   2120, 2304, 0, 1080, 1126, 1136, 1250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE),
+		   DRM_MODE_FLAG_INTERLACE),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 40 - 1920x1080i@100Hz */
+	/* 40 - 1920x1080i@100Hz 16:9 */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1094, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
-			DRM_MODE_FLAG_INTERLACE),
+		   DRM_MODE_FLAG_INTERLACE),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 41 - 1280x720@100Hz */
+	/* 41 - 1280x720@100Hz 16:9 */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 148500, 1280, 1720,
 		   1760, 1980, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 42 - 720x576@100Hz */
+	/* 42 - 720x576@100Hz 4:3 */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 54000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 43 - 720x576@100Hz */
+	/* 43 - 720x576@100Hz 16:9 */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 54000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 44 - 720(1440)x576i@100Hz */
+	/* 44 - 720(1440)x576i@100Hz 4:3 */
 	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 27000, 720, 732,
 		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 45 - 720(1440)x576i@100Hz */
+	/* 45 - 720(1440)x576i@100Hz 16:9 */
 	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 27000, 720, 732,
 		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 46 - 1920x1080i@120Hz */
+	/* 46 - 1920x1080i@120Hz 16:9 */
 	{ DRM_MODE("1920x1080i", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1094, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC |
-			DRM_MODE_FLAG_INTERLACE),
+		   DRM_MODE_FLAG_INTERLACE),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 47 - 1280x720@120Hz */
+	/* 47 - 1280x720@120Hz 16:9 */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 148500, 1280, 1390,
 		   1430, 1650, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 48 - 720x480@120Hz */
+	/* 48 - 720x480@120Hz 4:3 */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 54000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 49 - 720x480@120Hz */
+	/* 49 - 720x480@120Hz 16:9 */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 54000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 50 - 720(1440)x480i@120Hz */
+	/* 50 - 720(1440)x480i@120Hz 4:3 */
 	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 27000, 720, 739,
 		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 51 - 720(1440)x480i@120Hz */
+	/* 51 - 720(1440)x480i@120Hz 16:9 */
 	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 27000, 720, 739,
 		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 52 - 720x576@200Hz */
+	/* 52 - 720x576@200Hz 4:3 */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 108000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 53 - 720x576@200Hz */
+	/* 53 - 720x576@200Hz 16:9 */
 	{ DRM_MODE("720x576", DRM_MODE_TYPE_DRIVER, 108000, 720, 732,
 		   796, 864, 0, 576, 581, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 54 - 720(1440)x576i@200Hz */
+	/* 54 - 720(1440)x576i@200Hz 4:3 */
 	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 54000, 720, 732,
 		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 55 - 720(1440)x576i@200Hz */
+	/* 55 - 720(1440)x576i@200Hz 16:9 */
 	{ DRM_MODE("720x576i", DRM_MODE_TYPE_DRIVER, 54000, 720, 732,
 		   795, 864, 0, 576, 580, 586, 625, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 200, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 56 - 720x480@240Hz */
+	/* 56 - 720x480@240Hz 4:3 */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 108000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 57 - 720x480@240Hz */
+	/* 57 - 720x480@240Hz 16:9 */
 	{ DRM_MODE("720x480", DRM_MODE_TYPE_DRIVER, 108000, 720, 736,
 		   798, 858, 0, 480, 489, 495, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC),
 	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 58 - 720(1440)x480i@240Hz */
+	/* 58 - 720(1440)x480i@240Hz 4:3 */
 	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 54000, 720, 739,
 		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_4_3, },
-	/* 59 - 720(1440)x480i@240Hz */
+	/* 59 - 720(1440)x480i@240Hz 16:9 */
 	{ DRM_MODE("720x480i", DRM_MODE_TYPE_DRIVER, 54000, 720, 739,
 		   801, 858, 0, 480, 488, 494, 525, 0,
 		   DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC |
-			DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
+		   DRM_MODE_FLAG_INTERLACE | DRM_MODE_FLAG_DBLCLK),
 	  .vrefresh = 240, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 60 - 1280x720@24Hz */
+	/* 60 - 1280x720@24Hz 16:9 */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 59400, 1280, 3040,
 		   3080, 3300, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 24, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 61 - 1280x720@25Hz */
+	/* 61 - 1280x720@25Hz 16:9 */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 3700,
 		   3740, 3960, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 25, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 62 - 1280x720@30Hz */
+	/* 62 - 1280x720@30Hz 16:9 */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 3040,
 		   3080, 3300, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 30, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 63 - 1920x1080@120Hz */
+	/* 63 - 1920x1080@120Hz 16:9 */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 297000, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	 .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 64 - 1920x1080@100Hz */
+	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
+	/* 64 - 1920x1080@100Hz 16:9 */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 297000, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
-	 .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 65 - 1280x720@24Hz */
+	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
+	/* 65 - 1280x720@24Hz 64:27 */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 59400, 1280, 3040,
 		   3080, 3300, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 24, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 66 - 1280x720@25Hz */
+	/* 66 - 1280x720@25Hz 64:27 */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 3700,
 		   3740, 3960, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 25, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 67 - 1280x720@30Hz */
+	/* 67 - 1280x720@30Hz 64:27 */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 3040,
 		   3080, 3300, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 30, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 68 - 1280x720@50Hz */
+	/* 68 - 1280x720@50Hz 64:27 */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 1720,
 		   1760, 1980, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 69 - 1280x720@60Hz */
+	/* 69 - 1280x720@60Hz 64:27 */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 74250, 1280, 1390,
 		   1430, 1650, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 70 - 1280x720@100Hz */
+	/* 70 - 1280x720@100Hz 64:27 */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 148500, 1280, 1720,
 		   1760, 1980, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 71 - 1280x720@120Hz */
+	/* 71 - 1280x720@120Hz 64:27 */
 	{ DRM_MODE("1280x720", DRM_MODE_TYPE_DRIVER, 148500, 1280, 1390,
 		   1430, 1650, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 72 - 1920x1080@24Hz */
+	/* 72 - 1920x1080@24Hz 64:27 */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2558,
 		   2602, 2750, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 24, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 73 - 1920x1080@25Hz */
+	/* 73 - 1920x1080@25Hz 64:27 */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 25, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 74 - 1920x1080@30Hz */
+	/* 74 - 1920x1080@30Hz 64:27 */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 74250, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 30, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 75 - 1920x1080@50Hz */
+	/* 75 - 1920x1080@50Hz 64:27 */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 76 - 1920x1080@60Hz */
+	/* 76 - 1920x1080@60Hz 64:27 */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 148500, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 77 - 1920x1080@100Hz */
+	/* 77 - 1920x1080@100Hz 64:27 */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 297000, 1920, 2448,
 		   2492, 2640, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 78 - 1920x1080@120Hz */
+	/* 78 - 1920x1080@120Hz 64:27 */
 	{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 297000, 1920, 2008,
 		   2052, 2200, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 79 - 1680x720@24Hz */
+	/* 79 - 1680x720@24Hz 64:27 */
 	{ DRM_MODE("1680x720", DRM_MODE_TYPE_DRIVER, 59400, 1680, 3040,
 		   3080, 3300, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 24, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 80 - 1680x720@25Hz */
+	/* 80 - 1680x720@25Hz 64:27 */
 	{ DRM_MODE("1680x720", DRM_MODE_TYPE_DRIVER, 59400, 1680, 2908,
 		   2948, 3168, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 25, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 81 - 1680x720@30Hz */
+	/* 81 - 1680x720@30Hz 64:27 */
 	{ DRM_MODE("1680x720", DRM_MODE_TYPE_DRIVER, 59400, 1680, 2380,
 		   2420, 2640, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 30, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 82 - 1680x720@50Hz */
+	/* 82 - 1680x720@50Hz 64:27 */
 	{ DRM_MODE("1680x720", DRM_MODE_TYPE_DRIVER, 82500, 1680, 1940,
 		   1980, 2200, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 83 - 1680x720@60Hz */
+	/* 83 - 1680x720@60Hz 64:27 */
 	{ DRM_MODE("1680x720", DRM_MODE_TYPE_DRIVER, 99000, 1680, 1940,
 		   1980, 2200, 0, 720, 725, 730, 750, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 84 - 1680x720@100Hz */
+	/* 84 - 1680x720@100Hz 64:27 */
 	{ DRM_MODE("1680x720", DRM_MODE_TYPE_DRIVER, 165000, 1680, 1740,
 		   1780, 2000, 0, 720, 725, 730, 825, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 85 - 1680x720@120Hz */
+	/* 85 - 1680x720@120Hz 64:27 */
 	{ DRM_MODE("1680x720", DRM_MODE_TYPE_DRIVER, 198000, 1680, 1740,
 		   1780, 2000, 0, 720, 725, 730, 825, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 86 - 2560x1080@24Hz */
+	/* 86 - 2560x1080@24Hz 64:27 */
 	{ DRM_MODE("2560x1080", DRM_MODE_TYPE_DRIVER, 99000, 2560, 3558,
 		   3602, 3750, 0, 1080, 1084, 1089, 1100, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 24, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 87 - 2560x1080@25Hz */
+	/* 87 - 2560x1080@25Hz 64:27 */
 	{ DRM_MODE("2560x1080", DRM_MODE_TYPE_DRIVER, 90000, 2560, 3008,
 		   3052, 3200, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 25, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 88 - 2560x1080@30Hz */
+	/* 88 - 2560x1080@30Hz 64:27 */
 	{ DRM_MODE("2560x1080", DRM_MODE_TYPE_DRIVER, 118800, 2560, 3328,
 		   3372, 3520, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 30, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 89 - 2560x1080@50Hz */
+	/* 89 - 2560x1080@50Hz 64:27 */
 	{ DRM_MODE("2560x1080", DRM_MODE_TYPE_DRIVER, 185625, 2560, 3108,
 		   3152, 3300, 0, 1080, 1084, 1089, 1125, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 90 - 2560x1080@60Hz */
+	/* 90 - 2560x1080@60Hz 64:27 */
 	{ DRM_MODE("2560x1080", DRM_MODE_TYPE_DRIVER, 198000, 2560, 2808,
 		   2852, 3000, 0, 1080, 1084, 1089, 1100, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 91 - 2560x1080@100Hz */
+	/* 91 - 2560x1080@100Hz 64:27 */
 	{ DRM_MODE("2560x1080", DRM_MODE_TYPE_DRIVER, 371250, 2560, 2778,
 		   2822, 2970, 0, 1080, 1084, 1089, 1250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 100, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 92 - 2560x1080@120Hz */
+	/* 92 - 2560x1080@120Hz 64:27 */
 	{ DRM_MODE("2560x1080", DRM_MODE_TYPE_DRIVER, 495000, 2560, 3108,
 		   3152, 3300, 0, 1080, 1084, 1089, 1250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 120, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 93 - 3840x2160p@24Hz 16:9 */
+	/* 93 - 3840x2160@24Hz 16:9 */
 	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 297000, 3840, 5116,
 		   5204, 5500, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 24, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 94 - 3840x2160p@25Hz 16:9 */
+	/* 94 - 3840x2160@25Hz 16:9 */
 	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 297000, 3840, 4896,
 		   4984, 5280, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 25, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 95 - 3840x2160p@30Hz 16:9 */
+	/* 95 - 3840x2160@30Hz 16:9 */
 	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 297000, 3840, 4016,
 		   4104, 4400, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 30, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 96 - 3840x2160p@50Hz 16:9 */
+	/* 96 - 3840x2160@50Hz 16:9 */
 	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 594000, 3840, 4896,
 		   4984, 5280, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 97 - 3840x2160p@60Hz 16:9 */
+	/* 97 - 3840x2160@60Hz 16:9 */
 	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 594000, 3840, 4016,
 		   4104, 4400, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_16_9, },
-	/* 98 - 4096x2160p@24Hz 256:135 */
+	/* 98 - 4096x2160@24Hz 256:135 */
 	{ DRM_MODE("4096x2160", DRM_MODE_TYPE_DRIVER, 297000, 4096, 5116,
 		   5204, 5500, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 24, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_256_135, },
-	/* 99 - 4096x2160p@25Hz 256:135 */
+	/* 99 - 4096x2160@25Hz 256:135 */
 	{ DRM_MODE("4096x2160", DRM_MODE_TYPE_DRIVER, 297000, 4096, 5064,
 		   5152, 5280, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 25, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_256_135, },
-	/* 100 - 4096x2160p@30Hz 256:135 */
+	/* 100 - 4096x2160@30Hz 256:135 */
 	{ DRM_MODE("4096x2160", DRM_MODE_TYPE_DRIVER, 297000, 4096, 4184,
 		   4272, 4400, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 30, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_256_135, },
-	/* 101 - 4096x2160p@50Hz 256:135 */
+	/* 101 - 4096x2160@50Hz 256:135 */
 	{ DRM_MODE("4096x2160", DRM_MODE_TYPE_DRIVER, 594000, 4096, 5064,
 		   5152, 5280, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_256_135, },
-	/* 102 - 4096x2160p@60Hz 256:135 */
+	/* 102 - 4096x2160@60Hz 256:135 */
 	{ DRM_MODE("4096x2160", DRM_MODE_TYPE_DRIVER, 594000, 4096, 4184,
 		   4272, 4400, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 60, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_256_135, },
-	/* 103 - 3840x2160p@24Hz 64:27 */
+	/* 103 - 3840x2160@24Hz 64:27 */
 	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 297000, 3840, 5116,
 		   5204, 5500, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 24, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 104 - 3840x2160p@25Hz 64:27 */
+	/* 104 - 3840x2160@25Hz 64:27 */
 	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 297000, 3840, 4896,
 		   4984, 5280, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 25, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 105 - 3840x2160p@30Hz 64:27 */
+	/* 105 - 3840x2160@30Hz 64:27 */
 	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 297000, 3840, 4016,
 		   4104, 4400, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 30, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 106 - 3840x2160p@50Hz 64:27 */
+	/* 106 - 3840x2160@50Hz 64:27 */
 	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 594000, 3840, 4896,
 		   4984, 5280, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
 	  .vrefresh = 50, .picture_aspect_ratio = HDMI_PICTURE_ASPECT_64_27, },
-	/* 107 - 3840x2160p@60Hz 64:27 */
+	/* 107 - 3840x2160@60Hz 64:27 */
 	{ DRM_MODE("3840x2160", DRM_MODE_TYPE_DRIVER, 594000, 3840, 4016,
 		   4104, 4400, 0, 2160, 2168, 2178, 2250, 0,
 		   DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC),
@@ -1633,7 +1634,8 @@ struct edid *drm_do_get_edid(struct drm_connector *connector,
 		edid[EDID_LENGTH-1] += edid[0x7e] - valid_extensions;
 		edid[0x7e] = valid_extensions;
 
-		new = kmalloc((valid_extensions + 1) * EDID_LENGTH, GFP_KERNEL);
+		new = kmalloc_array(valid_extensions + 1, EDID_LENGTH,
+				    GFP_KERNEL);
 		if (!new)
 			goto out;
 
@@ -2930,11 +2932,15 @@ cea_mode_alternate_timings(u8 vic, struct drm_display_mode *mode)
 static u8 drm_match_cea_mode_clock_tolerance(const struct drm_display_mode *to_match,
 					     unsigned int clock_tolerance)
 {
+	unsigned int match_flags = DRM_MODE_MATCH_TIMINGS | DRM_MODE_MATCH_FLAGS;
 	u8 vic;
 
 	if (!to_match->clock)
 		return 0;
 
+	if (to_match->picture_aspect_ratio)
+		match_flags |= DRM_MODE_MATCH_ASPECT_RATIO;
+
 	for (vic = 1; vic < ARRAY_SIZE(edid_cea_modes); vic++) {
 		struct drm_display_mode cea_mode = edid_cea_modes[vic];
 		unsigned int clock1, clock2;
@@ -2948,7 +2954,7 @@ static u8 drm_match_cea_mode_clock_tolerance(const struct drm_display_mode *to_m
 			continue;
 
 		do {
-			if (drm_mode_equal_no_clocks_no_stereo(to_match, &cea_mode))
+			if (drm_mode_match(to_match, &cea_mode, match_flags))
 				return vic;
 		} while (cea_mode_alternate_timings(vic, &cea_mode));
 	}
@@ -2965,11 +2971,15 @@ static u8 drm_match_cea_mode_clock_tolerance(const struct drm_display_mode *to_m
  */
 u8 drm_match_cea_mode(const struct drm_display_mode *to_match)
 {
+	unsigned int match_flags = DRM_MODE_MATCH_TIMINGS | DRM_MODE_MATCH_FLAGS;
 	u8 vic;
 
 	if (!to_match->clock)
 		return 0;
 
+	if (to_match->picture_aspect_ratio)
+		match_flags |= DRM_MODE_MATCH_ASPECT_RATIO;
+
 	for (vic = 1; vic < ARRAY_SIZE(edid_cea_modes); vic++) {
 		struct drm_display_mode cea_mode = edid_cea_modes[vic];
 		unsigned int clock1, clock2;
@@ -2983,7 +2993,7 @@ u8 drm_match_cea_mode(const struct drm_display_mode *to_match)
 			continue;
 
 		do {
-			if (drm_mode_equal_no_clocks_no_stereo(to_match, &cea_mode))
+			if (drm_mode_match(to_match, &cea_mode, match_flags))
 				return vic;
 		} while (cea_mode_alternate_timings(vic, &cea_mode));
 	}
@@ -3030,6 +3040,7 @@ hdmi_mode_alternate_clock(const struct drm_display_mode *hdmi_mode)
 static u8 drm_match_hdmi_mode_clock_tolerance(const struct drm_display_mode *to_match,
 					      unsigned int clock_tolerance)
 {
+	unsigned int match_flags = DRM_MODE_MATCH_TIMINGS | DRM_MODE_MATCH_FLAGS;
 	u8 vic;
 
 	if (!to_match->clock)
@@ -3047,7 +3058,7 @@ static u8 drm_match_hdmi_mode_clock_tolerance(const struct drm_display_mode *to_
 		    abs(to_match->clock - clock2) > clock_tolerance)
 			continue;
 
-		if (drm_mode_equal_no_clocks(to_match, hdmi_mode))
+		if (drm_mode_match(to_match, hdmi_mode, match_flags))
 			return vic;
 	}
 
@@ -3064,6 +3075,7 @@ static u8 drm_match_hdmi_mode_clock_tolerance(const struct drm_display_mode *to_
  */
 static u8 drm_match_hdmi_mode(const struct drm_display_mode *to_match)
 {
+	unsigned int match_flags = DRM_MODE_MATCH_TIMINGS | DRM_MODE_MATCH_FLAGS;
 	u8 vic;
 
 	if (!to_match->clock)
@@ -3079,7 +3091,7 @@ static u8 drm_match_hdmi_mode(const struct drm_display_mode *to_match)
 
 		if ((KHZ2PICOS(to_match->clock) == KHZ2PICOS(clock1) ||
 		     KHZ2PICOS(to_match->clock) == KHZ2PICOS(clock2)) &&
-		    drm_mode_equal_no_clocks_no_stereo(to_match, hdmi_mode))
+		    drm_mode_match(to_match, hdmi_mode, match_flags))
 			return vic;
 	}
 	return 0;
@@ -4823,6 +4835,7 @@ drm_hdmi_avi_infoframe_from_display_mode(struct hdmi_avi_infoframe *frame,
 					 const struct drm_display_mode *mode,
 					 bool is_hdmi2_sink)
 {
+	enum hdmi_picture_aspect picture_aspect;
 	int err;
 
 	if (!frame || !mode)
@@ -4862,16 +4875,34 @@ drm_hdmi_avi_infoframe_from_display_mode(struct hdmi_avi_infoframe *frame,
 	frame->picture_aspect = HDMI_PICTURE_ASPECT_NONE;
 
 	/*
+	 * As some drivers don't support atomic, we can't use connector state.
+	 * So just initialize the frame with default values, just the same way
+	 * as it's done with other properties here.
+	 */
+	frame->content_type = HDMI_CONTENT_TYPE_GRAPHICS;
+	frame->itc = 0;
+
+	/*
 	 * Populate picture aspect ratio from either
 	 * user input (if specified) or from the CEA mode list.
 	 */
-	if (mode->picture_aspect_ratio == HDMI_PICTURE_ASPECT_4_3 ||
-		mode->picture_aspect_ratio == HDMI_PICTURE_ASPECT_16_9)
-		frame->picture_aspect = mode->picture_aspect_ratio;
-	else if (frame->video_code > 0)
-		frame->picture_aspect = drm_get_cea_aspect_ratio(
-						frame->video_code);
+	picture_aspect = mode->picture_aspect_ratio;
+	if (picture_aspect == HDMI_PICTURE_ASPECT_NONE)
+		picture_aspect = drm_get_cea_aspect_ratio(frame->video_code);
+
+	/*
+	 * The infoframe can't convey anything but none, 4:3
+	 * and 16:9, so if the user has asked for anything else
+	 * we can only satisfy it by specifying the right VIC.
+	 */
+	if (picture_aspect > HDMI_PICTURE_ASPECT_16_9) {
+		if (picture_aspect !=
+		    drm_get_cea_aspect_ratio(frame->video_code))
+			return -EINVAL;
+		picture_aspect = HDMI_PICTURE_ASPECT_NONE;
+	}
 
+	frame->picture_aspect = picture_aspect;
 	frame->active_aspect = HDMI_ACTIVE_ASPECT_PICTURE;
 	frame->scan_mode = HDMI_SCAN_MODE_UNDERSCAN;
 
diff --git a/drivers/gpu/drm/drm_fb_cma_helper.c b/drivers/gpu/drm/drm_fb_cma_helper.c
index 186d00adfb5f..9da36a6271d3 100644
--- a/drivers/gpu/drm/drm_fb_cma_helper.c
+++ b/drivers/gpu/drm/drm_fb_cma_helper.c
@@ -18,6 +18,7 @@
  */
 
 #include <drm/drmP.h>
+#include <drm/drm_client.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_framebuffer.h>
 #include <drm/drm_gem_cma_helper.h>
@@ -26,11 +27,8 @@
 #include <drm/drm_print.h>
 #include <linux/module.h>
 
-#define DEFAULT_FBDEFIO_DELAY_MS 50
-
 struct drm_fbdev_cma {
 	struct drm_fb_helper	fb_helper;
-	const struct drm_framebuffer_funcs *fb_funcs;
 };
 
 /**
@@ -44,36 +42,6 @@ struct drm_fbdev_cma {
  *
  * An fbdev framebuffer backed by cma is also available by calling
  * drm_fb_cma_fbdev_init(). drm_fb_cma_fbdev_fini() tears it down.
- * If the &drm_framebuffer_funcs.dirty callback is set, fb_deferred_io will be
- * set up automatically. &drm_framebuffer_funcs.dirty is called by
- * drm_fb_helper_deferred_io() in process context (&struct delayed_work).
- *
- * Example fbdev deferred io code::
- *
- *     static int driver_fb_dirty(struct drm_framebuffer *fb,
- *                                struct drm_file *file_priv,
- *                                unsigned flags, unsigned color,
- *                                struct drm_clip_rect *clips,
- *                                unsigned num_clips)
- *     {
- *         struct drm_gem_cma_object *cma = drm_fb_cma_get_gem_obj(fb, 0);
- *         ... push changes ...
- *         return 0;
- *     }
- *
- *     static struct drm_framebuffer_funcs driver_fb_funcs = {
- *         .destroy       = drm_gem_fb_destroy,
- *         .create_handle = drm_gem_fb_create_handle,
- *         .dirty         = driver_fb_dirty,
- *     };
- *
- * Initialize::
- *
- *     fbdev = drm_fb_cma_fbdev_init_with_funcs(dev, 16,
- *                                           dev->mode_config.num_crtc,
- *                                           dev->mode_config.num_connector,
- *                                           &driver_fb_funcs);
- *
  */
 
 static inline struct drm_fbdev_cma *to_fbdev_cma(struct drm_fb_helper *helper)
@@ -131,236 +99,28 @@ dma_addr_t drm_fb_cma_get_gem_addr(struct drm_framebuffer *fb,
 }
 EXPORT_SYMBOL_GPL(drm_fb_cma_get_gem_addr);
 
-static int drm_fb_cma_mmap(struct fb_info *info, struct vm_area_struct *vma)
-{
-	return dma_mmap_writecombine(info->device, vma, info->screen_base,
-				     info->fix.smem_start, info->fix.smem_len);
-}
-
-static struct fb_ops drm_fbdev_cma_ops = {
-	.owner		= THIS_MODULE,
-	DRM_FB_HELPER_DEFAULT_OPS,
-	.fb_fillrect	= drm_fb_helper_sys_fillrect,
-	.fb_copyarea	= drm_fb_helper_sys_copyarea,
-	.fb_imageblit	= drm_fb_helper_sys_imageblit,
-	.fb_mmap	= drm_fb_cma_mmap,
-};
-
-static int drm_fbdev_cma_deferred_io_mmap(struct fb_info *info,
-					  struct vm_area_struct *vma)
-{
-	fb_deferred_io_mmap(info, vma);
-	vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot);
-
-	return 0;
-}
-
-static int drm_fbdev_cma_defio_init(struct fb_info *fbi,
-				    struct drm_gem_cma_object *cma_obj)
-{
-	struct fb_deferred_io *fbdefio;
-	struct fb_ops *fbops;
-
-	/*
-	 * Per device structures are needed because:
-	 * fbops: fb_deferred_io_cleanup() clears fbops.fb_mmap
-	 * fbdefio: individual delays
-	 */
-	fbdefio = kzalloc(sizeof(*fbdefio), GFP_KERNEL);
-	fbops = kzalloc(sizeof(*fbops), GFP_KERNEL);
-	if (!fbdefio || !fbops) {
-		kfree(fbdefio);
-		kfree(fbops);
-		return -ENOMEM;
-	}
-
-	/* can't be offset from vaddr since dirty() uses cma_obj */
-	fbi->screen_buffer = cma_obj->vaddr;
-	/* fb_deferred_io_fault() needs a physical address */
-	fbi->fix.smem_start = page_to_phys(virt_to_page(fbi->screen_buffer));
-
-	*fbops = *fbi->fbops;
-	fbi->fbops = fbops;
-
-	fbdefio->delay = msecs_to_jiffies(DEFAULT_FBDEFIO_DELAY_MS);
-	fbdefio->deferred_io = drm_fb_helper_deferred_io;
-	fbi->fbdefio = fbdefio;
-	fb_deferred_io_init(fbi);
-	fbi->fbops->fb_mmap = drm_fbdev_cma_deferred_io_mmap;
-
-	return 0;
-}
-
-static void drm_fbdev_cma_defio_fini(struct fb_info *fbi)
-{
-	if (!fbi->fbdefio)
-		return;
-
-	fb_deferred_io_cleanup(fbi);
-	kfree(fbi->fbdefio);
-	kfree(fbi->fbops);
-}
-
-static int
-drm_fbdev_cma_create(struct drm_fb_helper *helper,
-	struct drm_fb_helper_surface_size *sizes)
-{
-	struct drm_fbdev_cma *fbdev_cma = to_fbdev_cma(helper);
-	struct drm_device *dev = helper->dev;
-	struct drm_gem_cma_object *obj;
-	struct drm_framebuffer *fb;
-	unsigned int bytes_per_pixel;
-	unsigned long offset;
-	struct fb_info *fbi;
-	size_t size;
-	int ret;
-
-	DRM_DEBUG_KMS("surface width(%d), height(%d) and bpp(%d)\n",
-			sizes->surface_width, sizes->surface_height,
-			sizes->surface_bpp);
-
-	bytes_per_pixel = DIV_ROUND_UP(sizes->surface_bpp, 8);
-	size = sizes->surface_width * sizes->surface_height * bytes_per_pixel;
-	obj = drm_gem_cma_create(dev, size);
-	if (IS_ERR(obj))
-		return -ENOMEM;
-
-	fbi = drm_fb_helper_alloc_fbi(helper);
-	if (IS_ERR(fbi)) {
-		ret = PTR_ERR(fbi);
-		goto err_gem_free_object;
-	}
-
-	fb = drm_gem_fbdev_fb_create(dev, sizes, 0, &obj->base,
-				     fbdev_cma->fb_funcs);
-	if (IS_ERR(fb)) {
-		dev_err(dev->dev, "Failed to allocate DRM framebuffer.\n");
-		ret = PTR_ERR(fb);
-		goto err_fb_info_destroy;
-	}
-
-	helper->fb = fb;
-
-	fbi->par = helper;
-	fbi->flags = FBINFO_FLAG_DEFAULT;
-	fbi->fbops = &drm_fbdev_cma_ops;
-
-	drm_fb_helper_fill_fix(fbi, fb->pitches[0], fb->format->depth);
-	drm_fb_helper_fill_var(fbi, helper, sizes->fb_width, sizes->fb_height);
-
-	offset = fbi->var.xoffset * bytes_per_pixel;
-	offset += fbi->var.yoffset * fb->pitches[0];
-
-	dev->mode_config.fb_base = (resource_size_t)obj->paddr;
-	fbi->screen_base = obj->vaddr + offset;
-	fbi->fix.smem_start = (unsigned long)(obj->paddr + offset);
-	fbi->screen_size = size;
-	fbi->fix.smem_len = size;
-
-	if (fb->funcs->dirty) {
-		ret = drm_fbdev_cma_defio_init(fbi, obj);
-		if (ret)
-			goto err_cma_destroy;
-	}
-
-	return 0;
-
-err_cma_destroy:
-	drm_framebuffer_remove(fb);
-err_fb_info_destroy:
-	drm_fb_helper_fini(helper);
-err_gem_free_object:
-	drm_gem_object_put_unlocked(&obj->base);
-	return ret;
-}
-
-static const struct drm_fb_helper_funcs drm_fb_cma_helper_funcs = {
-	.fb_probe = drm_fbdev_cma_create,
-};
-
 /**
- * drm_fb_cma_fbdev_init_with_funcs() - Allocate and initialize fbdev emulation
+ * drm_fb_cma_fbdev_init() - Allocate and initialize fbdev emulation
  * @dev: DRM device
  * @preferred_bpp: Preferred bits per pixel for the device.
  *                 @dev->mode_config.preferred_depth is used if this is zero.
  * @max_conn_count: Maximum number of connectors.
  *                  @dev->mode_config.num_connector is used if this is zero.
- * @funcs: Framebuffer functions, in particular a custom dirty() callback.
- *         Can be NULL.
  *
  * Returns:
  * Zero on success or negative error code on failure.
  */
-int drm_fb_cma_fbdev_init_with_funcs(struct drm_device *dev,
-	unsigned int preferred_bpp, unsigned int max_conn_count,
-	const struct drm_framebuffer_funcs *funcs)
+int drm_fb_cma_fbdev_init(struct drm_device *dev, unsigned int preferred_bpp,
+			  unsigned int max_conn_count)
 {
 	struct drm_fbdev_cma *fbdev_cma;
-	struct drm_fb_helper *fb_helper;
-	int ret;
-
-	if (!preferred_bpp)
-		preferred_bpp = dev->mode_config.preferred_depth;
-	if (!preferred_bpp)
-		preferred_bpp = 32;
-
-	if (!max_conn_count)
-		max_conn_count = dev->mode_config.num_connector;
-
-	fbdev_cma = kzalloc(sizeof(*fbdev_cma), GFP_KERNEL);
-	if (!fbdev_cma)
-		return -ENOMEM;
 
-	fbdev_cma->fb_funcs = funcs;
-	fb_helper = &fbdev_cma->fb_helper;
-
-	drm_fb_helper_prepare(dev, fb_helper, &drm_fb_cma_helper_funcs);
-
-	ret = drm_fb_helper_init(dev, fb_helper, max_conn_count);
-	if (ret < 0) {
-		DRM_DEV_ERROR(dev->dev, "Failed to initialize fbdev helper.\n");
-		goto err_free;
-	}
-
-	ret = drm_fb_helper_single_add_all_connectors(fb_helper);
-	if (ret < 0) {
-		DRM_DEV_ERROR(dev->dev, "Failed to add connectors.\n");
-		goto err_drm_fb_helper_fini;
-	}
-
-	ret = drm_fb_helper_initial_config(fb_helper, preferred_bpp);
-	if (ret < 0) {
-		DRM_DEV_ERROR(dev->dev, "Failed to set fbdev configuration.\n");
-		goto err_drm_fb_helper_fini;
-	}
+	/* dev->fb_helper will indirectly point to fbdev_cma after this call */
+	fbdev_cma = drm_fbdev_cma_init(dev, preferred_bpp, max_conn_count);
+	if (IS_ERR(fbdev_cma))
+		return PTR_ERR(fbdev_cma);
 
 	return 0;
-
-err_drm_fb_helper_fini:
-	drm_fb_helper_fini(fb_helper);
-err_free:
-	kfree(fbdev_cma);
-
-	return ret;
-}
-EXPORT_SYMBOL_GPL(drm_fb_cma_fbdev_init_with_funcs);
-
-/**
- * drm_fb_cma_fbdev_init() - Allocate and initialize fbdev emulation
- * @dev: DRM device
- * @preferred_bpp: Preferred bits per pixel for the device.
- *                 @dev->mode_config.preferred_depth is used if this is zero.
- * @max_conn_count: Maximum number of connectors.
- *                  @dev->mode_config.num_connector is used if this is zero.
- *
- * Returns:
- * Zero on success or negative error code on failure.
- */
-int drm_fb_cma_fbdev_init(struct drm_device *dev, unsigned int preferred_bpp,
-			  unsigned int max_conn_count)
-{
-	return drm_fb_cma_fbdev_init_with_funcs(dev, preferred_bpp,
-						max_conn_count, NULL);
 }
 EXPORT_SYMBOL_GPL(drm_fb_cma_fbdev_init);
 
@@ -370,104 +130,54 @@ EXPORT_SYMBOL_GPL(drm_fb_cma_fbdev_init);
  */
 void drm_fb_cma_fbdev_fini(struct drm_device *dev)
 {
-	struct drm_fb_helper *fb_helper = dev->fb_helper;
-
-	if (!fb_helper)
-		return;
-
-	/* Unregister if it hasn't been done already */
-	if (fb_helper->fbdev && fb_helper->fbdev->dev)
-		drm_fb_helper_unregister_fbi(fb_helper);
-
-	if (fb_helper->fbdev)
-		drm_fbdev_cma_defio_fini(fb_helper->fbdev);
-
-	if (fb_helper->fb)
-		drm_framebuffer_remove(fb_helper->fb);
-
-	drm_fb_helper_fini(fb_helper);
-	kfree(to_fbdev_cma(fb_helper));
+	if (dev->fb_helper)
+		drm_fbdev_cma_fini(to_fbdev_cma(dev->fb_helper));
 }
 EXPORT_SYMBOL_GPL(drm_fb_cma_fbdev_fini);
 
+static const struct drm_fb_helper_funcs drm_fb_cma_helper_funcs = {
+	.fb_probe = drm_fb_helper_generic_probe,
+};
+
 /**
- * drm_fbdev_cma_init_with_funcs() - Allocate and initializes a drm_fbdev_cma struct
+ * drm_fbdev_cma_init() - Allocate and initializes a drm_fbdev_cma struct
  * @dev: DRM device
  * @preferred_bpp: Preferred bits per pixel for the device
  * @max_conn_count: Maximum number of connectors
- * @funcs: fb helper functions, in particular a custom dirty() callback
  *
  * Returns a newly allocated drm_fbdev_cma struct or a ERR_PTR.
  */
-struct drm_fbdev_cma *drm_fbdev_cma_init_with_funcs(struct drm_device *dev,
-	unsigned int preferred_bpp, unsigned int max_conn_count,
-	const struct drm_framebuffer_funcs *funcs)
+struct drm_fbdev_cma *drm_fbdev_cma_init(struct drm_device *dev,
+	unsigned int preferred_bpp, unsigned int max_conn_count)
 {
 	struct drm_fbdev_cma *fbdev_cma;
-	struct drm_fb_helper *helper;
+	struct drm_fb_helper *fb_helper;
 	int ret;
 
 	fbdev_cma = kzalloc(sizeof(*fbdev_cma), GFP_KERNEL);
-	if (!fbdev_cma) {
-		dev_err(dev->dev, "Failed to allocate drm fbdev.\n");
+	if (!fbdev_cma)
 		return ERR_PTR(-ENOMEM);
-	}
-	fbdev_cma->fb_funcs = funcs;
 
-	helper = &fbdev_cma->fb_helper;
-
-	drm_fb_helper_prepare(dev, helper, &drm_fb_cma_helper_funcs);
+	fb_helper = &fbdev_cma->fb_helper;
 
-	ret = drm_fb_helper_init(dev, helper, max_conn_count);
-	if (ret < 0) {
-		dev_err(dev->dev, "Failed to initialize drm fb helper.\n");
+	ret = drm_client_new(dev, &fb_helper->client, "fbdev", NULL);
+	if (ret)
 		goto err_free;
-	}
-
-	ret = drm_fb_helper_single_add_all_connectors(helper);
-	if (ret < 0) {
-		dev_err(dev->dev, "Failed to add connectors.\n");
-		goto err_drm_fb_helper_fini;
-
-	}
 
-	ret = drm_fb_helper_initial_config(helper, preferred_bpp);
-	if (ret < 0) {
-		dev_err(dev->dev, "Failed to set initial hw configuration.\n");
-		goto err_drm_fb_helper_fini;
-	}
+	ret = drm_fb_helper_fbdev_setup(dev, fb_helper, &drm_fb_cma_helper_funcs,
+					preferred_bpp, max_conn_count);
+	if (ret)
+		goto err_client_put;
 
 	return fbdev_cma;
 
-err_drm_fb_helper_fini:
-	drm_fb_helper_fini(helper);
+err_client_put:
+	drm_client_release(&fb_helper->client);
 err_free:
 	kfree(fbdev_cma);
 
 	return ERR_PTR(ret);
 }
-EXPORT_SYMBOL_GPL(drm_fbdev_cma_init_with_funcs);
-
-static const struct drm_framebuffer_funcs drm_fb_cma_funcs = {
-	.destroy	= drm_gem_fb_destroy,
-	.create_handle	= drm_gem_fb_create_handle,
-};
-
-/**
- * drm_fbdev_cma_init() - Allocate and initializes a drm_fbdev_cma struct
- * @dev: DRM device
- * @preferred_bpp: Preferred bits per pixel for the device
- * @max_conn_count: Maximum number of connectors
- *
- * Returns a newly allocated drm_fbdev_cma struct or a ERR_PTR.
- */
-struct drm_fbdev_cma *drm_fbdev_cma_init(struct drm_device *dev,
-	unsigned int preferred_bpp, unsigned int max_conn_count)
-{
-	return drm_fbdev_cma_init_with_funcs(dev, preferred_bpp,
-					     max_conn_count,
-					     &drm_fb_cma_funcs);
-}
 EXPORT_SYMBOL_GPL(drm_fbdev_cma_init);
 
 /**
@@ -477,14 +187,7 @@ EXPORT_SYMBOL_GPL(drm_fbdev_cma_init);
 void drm_fbdev_cma_fini(struct drm_fbdev_cma *fbdev_cma)
 {
 	drm_fb_helper_unregister_fbi(&fbdev_cma->fb_helper);
-	if (fbdev_cma->fb_helper.fbdev)
-		drm_fbdev_cma_defio_fini(fbdev_cma->fb_helper.fbdev);
-
-	if (fbdev_cma->fb_helper.fb)
-		drm_framebuffer_remove(fbdev_cma->fb_helper.fb);
-
-	drm_fb_helper_fini(&fbdev_cma->fb_helper);
-	kfree(fbdev_cma);
+	/* All resources have now been freed by drm_fbdev_fb_destroy() */
 }
 EXPORT_SYMBOL_GPL(drm_fbdev_cma_fini);
 
diff --git a/drivers/gpu/drm/drm_fb_helper.c b/drivers/gpu/drm/drm_fb_helper.c
index 0646b108030b..4b0dd20bccb8 100644
--- a/drivers/gpu/drm/drm_fb_helper.c
+++ b/drivers/gpu/drm/drm_fb_helper.c
@@ -30,6 +30,7 @@
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
 #include <linux/console.h>
+#include <linux/dma-buf.h>
 #include <linux/kernel.h>
 #include <linux/sysrq.h>
 #include <linux/slab.h>
@@ -66,6 +67,9 @@ static DEFINE_MUTEX(kernel_fb_helper_lock);
  * helper functions used by many drivers to implement the kernel mode setting
  * interfaces.
  *
+ * Drivers that support a dumb buffer with a virtual address and mmap support,
+ * should try out the generic fbdev emulation using drm_fbdev_generic_setup().
+ *
  * Setup fbdev emulation by calling drm_fb_helper_fbdev_setup() and tear it
  * down by calling drm_fb_helper_fbdev_teardown().
  *
@@ -368,7 +372,6 @@ static int restore_fbdev_mode_atomic(struct drm_fb_helper *fb_helper, bool activ
 	struct drm_plane *plane;
 	struct drm_atomic_state *state;
 	int i, ret;
-	unsigned int plane_mask;
 	struct drm_modeset_acquire_ctx ctx;
 
 	drm_modeset_acquire_init(&ctx, 0);
@@ -381,7 +384,6 @@ static int restore_fbdev_mode_atomic(struct drm_fb_helper *fb_helper, bool activ
 
 	state->acquire_ctx = &ctx;
 retry:
-	plane_mask = 0;
 	drm_for_each_plane(plane, dev) {
 		plane_state = drm_atomic_get_plane_state(state, plane);
 		if (IS_ERR(plane_state)) {
@@ -391,9 +393,6 @@ retry:
 
 		plane_state->rotation = DRM_MODE_ROTATE_0;
 
-		plane->old_fb = plane->fb;
-		plane_mask |= 1 << drm_plane_index(plane);
-
 		/* disable non-primary: */
 		if (plane->type == DRM_PLANE_TYPE_PRIMARY)
 			continue;
@@ -430,8 +429,6 @@ retry:
 	ret = drm_atomic_commit(state);
 
 out_state:
-	drm_atomic_clean_old_fb(dev, plane_mask, ret);
-
 	if (ret == -EDEADLK)
 		goto backoff;
 
@@ -745,6 +742,24 @@ static void drm_fb_helper_resume_worker(struct work_struct *work)
 	console_unlock();
 }
 
+static void drm_fb_helper_dirty_blit_real(struct drm_fb_helper *fb_helper,
+					  struct drm_clip_rect *clip)
+{
+	struct drm_framebuffer *fb = fb_helper->fb;
+	unsigned int cpp = drm_format_plane_cpp(fb->format->format, 0);
+	size_t offset = clip->y1 * fb->pitches[0] + clip->x1 * cpp;
+	void *src = fb_helper->fbdev->screen_buffer + offset;
+	void *dst = fb_helper->buffer->vaddr + offset;
+	size_t len = (clip->x2 - clip->x1) * cpp;
+	unsigned int y;
+
+	for (y = clip->y1; y < clip->y2; y++) {
+		memcpy(dst, src, len);
+		src += fb->pitches[0];
+		dst += fb->pitches[0];
+	}
+}
+
 static void drm_fb_helper_dirty_work(struct work_struct *work)
 {
 	struct drm_fb_helper *helper = container_of(work, struct drm_fb_helper,
@@ -760,8 +775,12 @@ static void drm_fb_helper_dirty_work(struct work_struct *work)
 	spin_unlock_irqrestore(&helper->dirty_lock, flags);
 
 	/* call dirty callback only when it has been really touched */
-	if (clip_copy.x1 < clip_copy.x2 && clip_copy.y1 < clip_copy.y2)
+	if (clip_copy.x1 < clip_copy.x2 && clip_copy.y1 < clip_copy.y2) {
+		/* Generic fbdev uses a shadow buffer */
+		if (helper->buffer)
+			drm_fb_helper_dirty_blit_real(helper, &clip_copy);
 		helper->fb->funcs->dirty(helper->fb, NULL, 0, 0, &clip_copy, 1);
+	}
 }
 
 /**
@@ -1164,7 +1183,7 @@ EXPORT_SYMBOL(drm_fb_helper_sys_imageblit);
  * @info: fbdev registered by the helper
  * @rect: info about rectangle to fill
  *
- * A wrapper around cfb_imageblit implemented by fbdev core
+ * A wrapper around cfb_fillrect implemented by fbdev core
  */
 void drm_fb_helper_cfb_fillrect(struct fb_info *info,
 				const struct fb_fillrect *rect)
@@ -2183,7 +2202,11 @@ static bool drm_target_cloned(struct drm_fb_helper *fb_helper,
 		for (j = 0; j < i; j++) {
 			if (!enabled[j])
 				continue;
-			if (!drm_mode_equal(modes[j], modes[i]))
+			if (!drm_mode_match(modes[j], modes[i],
+					    DRM_MODE_MATCH_TIMINGS |
+					    DRM_MODE_MATCH_CLOCK |
+					    DRM_MODE_MATCH_FLAGS |
+					    DRM_MODE_MATCH_3D_FLAGS))
 				can_clone = false;
 		}
 	}
@@ -2203,7 +2226,11 @@ static bool drm_target_cloned(struct drm_fb_helper *fb_helper,
 
 		fb_helper_conn = fb_helper->connector_info[i];
 		list_for_each_entry(mode, &fb_helper_conn->connector->modes, head) {
-			if (drm_mode_equal(mode, dmt_mode))
+			if (drm_mode_match(mode, dmt_mode,
+					   DRM_MODE_MATCH_TIMINGS |
+					   DRM_MODE_MATCH_CLOCK |
+					   DRM_MODE_MATCH_FLAGS |
+					   DRM_MODE_MATCH_3D_FLAGS))
 				modes[i] = mode;
 		}
 		if (!modes[i])
@@ -2322,6 +2349,20 @@ retry:
 	return true;
 }
 
+static bool connector_has_possible_crtc(struct drm_connector *connector,
+					struct drm_crtc *crtc)
+{
+	struct drm_encoder *encoder;
+	int i;
+
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
+		if (encoder->possible_crtcs & drm_crtc_mask(crtc))
+			return true;
+	}
+
+	return false;
+}
+
 static int drm_pick_crtcs(struct drm_fb_helper *fb_helper,
 			  struct drm_fb_helper_crtc **best_crtcs,
 			  struct drm_display_mode **modes,
@@ -2330,7 +2371,6 @@ static int drm_pick_crtcs(struct drm_fb_helper *fb_helper,
 	int c, o;
 	struct drm_connector *connector;
 	const struct drm_connector_helper_funcs *connector_funcs;
-	struct drm_encoder *encoder;
 	int my_score, best_score, score;
 	struct drm_fb_helper_crtc **crtcs, *crtc;
 	struct drm_fb_helper_connector *fb_helper_conn;
@@ -2362,27 +2402,14 @@ static int drm_pick_crtcs(struct drm_fb_helper *fb_helper,
 	connector_funcs = connector->helper_private;
 
 	/*
-	 * If the DRM device implements atomic hooks and ->best_encoder() is
-	 * NULL we fallback to the default drm_atomic_helper_best_encoder()
-	 * helper.
-	 */
-	if (drm_drv_uses_atomic_modeset(fb_helper->dev) &&
-	    !connector_funcs->best_encoder)
-		encoder = drm_atomic_helper_best_encoder(connector);
-	else
-		encoder = connector_funcs->best_encoder(connector);
-
-	if (!encoder)
-		goto out;
-
-	/*
 	 * select a crtc for this connector and then attempt to configure
 	 * remaining connectors
 	 */
 	for (c = 0; c < fb_helper->crtc_count; c++) {
 		crtc = &fb_helper->crtc_info[c];
 
-		if ((encoder->possible_crtcs & (1 << c)) == 0)
+		if (!connector_has_possible_crtc(connector,
+						 crtc->mode_set.crtc))
 			continue;
 
 		for (o = 0; o < n; o++)
@@ -2409,7 +2436,7 @@ static int drm_pick_crtcs(struct drm_fb_helper *fb_helper,
 			       sizeof(struct drm_fb_helper_crtc *));
 		}
 	}
-out:
+
 	kfree(crtcs);
 	return best_score;
 }
@@ -2920,6 +2947,294 @@ void drm_fb_helper_output_poll_changed(struct drm_device *dev)
 }
 EXPORT_SYMBOL(drm_fb_helper_output_poll_changed);
 
+/* @user: 1=userspace, 0=fbcon */
+static int drm_fbdev_fb_open(struct fb_info *info, int user)
+{
+	struct drm_fb_helper *fb_helper = info->par;
+
+	if (!try_module_get(fb_helper->dev->driver->fops->owner))
+		return -ENODEV;
+
+	return 0;
+}
+
+static int drm_fbdev_fb_release(struct fb_info *info, int user)
+{
+	struct drm_fb_helper *fb_helper = info->par;
+
+	module_put(fb_helper->dev->driver->fops->owner);
+
+	return 0;
+}
+
+/*
+ * fb_ops.fb_destroy is called by the last put_fb_info() call at the end of
+ * unregister_framebuffer() or fb_release().
+ */
+static void drm_fbdev_fb_destroy(struct fb_info *info)
+{
+	struct drm_fb_helper *fb_helper = info->par;
+	struct fb_info *fbi = fb_helper->fbdev;
+	struct fb_ops *fbops = NULL;
+	void *shadow = NULL;
+
+	if (fbi->fbdefio) {
+		fb_deferred_io_cleanup(fbi);
+		shadow = fbi->screen_buffer;
+		fbops = fbi->fbops;
+	}
+
+	drm_fb_helper_fini(fb_helper);
+
+	if (shadow) {
+		vfree(shadow);
+		kfree(fbops);
+	}
+
+	drm_client_framebuffer_delete(fb_helper->buffer);
+	/*
+	 * FIXME:
+	 * Remove conditional when all CMA drivers have been moved over to using
+	 * drm_fbdev_generic_setup().
+	 */
+	if (fb_helper->client.funcs) {
+		drm_client_release(&fb_helper->client);
+		kfree(fb_helper);
+	}
+}
+
+static int drm_fbdev_fb_mmap(struct fb_info *info, struct vm_area_struct *vma)
+{
+	struct drm_fb_helper *fb_helper = info->par;
+
+	if (fb_helper->dev->driver->gem_prime_mmap)
+		return fb_helper->dev->driver->gem_prime_mmap(fb_helper->buffer->gem, vma);
+	else
+		return -ENODEV;
+}
+
+static struct fb_ops drm_fbdev_fb_ops = {
+	.owner		= THIS_MODULE,
+	DRM_FB_HELPER_DEFAULT_OPS,
+	.fb_open	= drm_fbdev_fb_open,
+	.fb_release	= drm_fbdev_fb_release,
+	.fb_destroy	= drm_fbdev_fb_destroy,
+	.fb_mmap	= drm_fbdev_fb_mmap,
+	.fb_read	= drm_fb_helper_sys_read,
+	.fb_write	= drm_fb_helper_sys_write,
+	.fb_fillrect	= drm_fb_helper_sys_fillrect,
+	.fb_copyarea	= drm_fb_helper_sys_copyarea,
+	.fb_imageblit	= drm_fb_helper_sys_imageblit,
+};
+
+static struct fb_deferred_io drm_fbdev_defio = {
+	.delay		= HZ / 20,
+	.deferred_io	= drm_fb_helper_deferred_io,
+};
+
+/**
+ * drm_fb_helper_generic_probe - Generic fbdev emulation probe helper
+ * @fb_helper: fbdev helper structure
+ * @sizes: describes fbdev size and scanout surface size
+ *
+ * This function uses the client API to crate a framebuffer backed by a dumb buffer.
+ *
+ * The _sys_ versions are used for &fb_ops.fb_read, fb_write, fb_fillrect,
+ * fb_copyarea, fb_imageblit.
+ *
+ * Returns:
+ * Zero on success or negative error code on failure.
+ */
+int drm_fb_helper_generic_probe(struct drm_fb_helper *fb_helper,
+				struct drm_fb_helper_surface_size *sizes)
+{
+	struct drm_client_dev *client = &fb_helper->client;
+	struct drm_client_buffer *buffer;
+	struct drm_framebuffer *fb;
+	struct fb_info *fbi;
+	u32 format;
+	int ret;
+
+	DRM_DEBUG_KMS("surface width(%d), height(%d) and bpp(%d)\n",
+		      sizes->surface_width, sizes->surface_height,
+		      sizes->surface_bpp);
+
+	format = drm_mode_legacy_fb_format(sizes->surface_bpp, sizes->surface_depth);
+	buffer = drm_client_framebuffer_create(client, sizes->surface_width,
+					       sizes->surface_height, format);
+	if (IS_ERR(buffer))
+		return PTR_ERR(buffer);
+
+	fb_helper->buffer = buffer;
+	fb_helper->fb = buffer->fb;
+	fb = buffer->fb;
+
+	fbi = drm_fb_helper_alloc_fbi(fb_helper);
+	if (IS_ERR(fbi)) {
+		ret = PTR_ERR(fbi);
+		goto err_free_buffer;
+	}
+
+	fbi->par = fb_helper;
+	fbi->fbops = &drm_fbdev_fb_ops;
+	fbi->screen_size = fb->height * fb->pitches[0];
+	fbi->fix.smem_len = fbi->screen_size;
+	fbi->screen_buffer = buffer->vaddr;
+	strcpy(fbi->fix.id, "DRM emulated");
+
+	drm_fb_helper_fill_fix(fbi, fb->pitches[0], fb->format->depth);
+	drm_fb_helper_fill_var(fbi, fb_helper, sizes->fb_width, sizes->fb_height);
+
+	if (fb->funcs->dirty) {
+		struct fb_ops *fbops;
+		void *shadow;
+
+		/*
+		 * fb_deferred_io_cleanup() clears &fbops->fb_mmap so a per
+		 * instance version is necessary.
+		 */
+		fbops = kzalloc(sizeof(*fbops), GFP_KERNEL);
+		shadow = vzalloc(fbi->screen_size);
+		if (!fbops || !shadow) {
+			kfree(fbops);
+			vfree(shadow);
+			ret = -ENOMEM;
+			goto err_fb_info_destroy;
+		}
+
+		*fbops = *fbi->fbops;
+		fbi->fbops = fbops;
+		fbi->screen_buffer = shadow;
+		fbi->fbdefio = &drm_fbdev_defio;
+
+		fb_deferred_io_init(fbi);
+	}
+
+	return 0;
+
+err_fb_info_destroy:
+	drm_fb_helper_fini(fb_helper);
+err_free_buffer:
+	drm_client_framebuffer_delete(buffer);
+
+	return ret;
+}
+EXPORT_SYMBOL(drm_fb_helper_generic_probe);
+
+static const struct drm_fb_helper_funcs drm_fb_helper_generic_funcs = {
+	.fb_probe = drm_fb_helper_generic_probe,
+};
+
+static void drm_fbdev_client_unregister(struct drm_client_dev *client)
+{
+	struct drm_fb_helper *fb_helper = drm_fb_helper_from_client(client);
+
+	if (fb_helper->fbdev) {
+		drm_fb_helper_unregister_fbi(fb_helper);
+		/* drm_fbdev_fb_destroy() takes care of cleanup */
+		return;
+	}
+
+	/* Did drm_fb_helper_fbdev_setup() run? */
+	if (fb_helper->dev)
+		drm_fb_helper_fini(fb_helper);
+
+	drm_client_release(client);
+	kfree(fb_helper);
+}
+
+static int drm_fbdev_client_restore(struct drm_client_dev *client)
+{
+	struct drm_fb_helper *fb_helper = drm_fb_helper_from_client(client);
+
+	drm_fb_helper_restore_fbdev_mode_unlocked(fb_helper);
+
+	return 0;
+}
+
+static int drm_fbdev_client_hotplug(struct drm_client_dev *client)
+{
+	struct drm_fb_helper *fb_helper = drm_fb_helper_from_client(client);
+	struct drm_device *dev = client->dev;
+	int ret;
+
+	/* If drm_fb_helper_fbdev_setup() failed, we only try once */
+	if (!fb_helper->dev && fb_helper->funcs)
+		return 0;
+
+	if (dev->fb_helper)
+		return drm_fb_helper_hotplug_event(dev->fb_helper);
+
+	if (!dev->mode_config.num_connector)
+		return 0;
+
+	ret = drm_fb_helper_fbdev_setup(dev, fb_helper, &drm_fb_helper_generic_funcs,
+					fb_helper->preferred_bpp, 0);
+	if (ret) {
+		fb_helper->dev = NULL;
+		fb_helper->fbdev = NULL;
+		return ret;
+	}
+
+	return 0;
+}
+
+static const struct drm_client_funcs drm_fbdev_client_funcs = {
+	.owner		= THIS_MODULE,
+	.unregister	= drm_fbdev_client_unregister,
+	.restore	= drm_fbdev_client_restore,
+	.hotplug	= drm_fbdev_client_hotplug,
+};
+
+/**
+ * drm_fb_helper_generic_fbdev_setup() - Setup generic fbdev emulation
+ * @dev: DRM device
+ * @preferred_bpp: Preferred bits per pixel for the device.
+ *                 @dev->mode_config.preferred_depth is used if this is zero.
+ *
+ * This function sets up generic fbdev emulation for drivers that supports
+ * dumb buffers with a virtual address and that can be mmap'ed.
+ *
+ * Restore, hotplug events and teardown are all taken care of. Drivers that do
+ * suspend/resume need to call drm_fb_helper_set_suspend_unlocked() themselves.
+ * Simple drivers might use drm_mode_config_helper_suspend().
+ *
+ * Drivers that set the dirty callback on their framebuffer will get a shadow
+ * fbdev buffer that is blitted onto the real buffer. This is done in order to
+ * make deferred I/O work with all kinds of buffers.
+ *
+ * This function is safe to call even when there are no connectors present.
+ * Setup will be retried on the next hotplug event.
+ *
+ * Returns:
+ * Zero on success or negative error code on failure.
+ */
+int drm_fbdev_generic_setup(struct drm_device *dev, unsigned int preferred_bpp)
+{
+	struct drm_fb_helper *fb_helper;
+	int ret;
+
+	if (!drm_fbdev_emulation)
+		return 0;
+
+	fb_helper = kzalloc(sizeof(*fb_helper), GFP_KERNEL);
+	if (!fb_helper)
+		return -ENOMEM;
+
+	ret = drm_client_new(dev, &fb_helper->client, "fbdev", &drm_fbdev_client_funcs);
+	if (ret) {
+		kfree(fb_helper);
+		return ret;
+	}
+
+	fb_helper->preferred_bpp = preferred_bpp;
+
+	drm_fbdev_client_hotplug(&fb_helper->client);
+
+	return 0;
+}
+EXPORT_SYMBOL(drm_fbdev_generic_setup);
+
 /* The Kconfig DRM_KMS_HELPER selects FRAMEBUFFER_CONSOLE (if !EXPERT)
  * but the module doesn't depend on any fb console symbols.  At least
  * attempt to load fbcon to avoid leaving the system without a usable console.
diff --git a/drivers/gpu/drm/drm_file.c b/drivers/gpu/drm/drm_file.c
index e394799979a6..ffa8dc35515f 100644
--- a/drivers/gpu/drm/drm_file.c
+++ b/drivers/gpu/drm/drm_file.c
@@ -35,6 +35,7 @@
 #include <linux/slab.h>
 #include <linux/module.h>
 
+#include <drm/drm_client.h>
 #include <drm/drm_file.h>
 #include <drm/drmP.h>
 
@@ -101,6 +102,166 @@ DEFINE_MUTEX(drm_global_mutex);
 
 static int drm_open_helper(struct file *filp, struct drm_minor *minor);
 
+/**
+ * drm_file_alloc - allocate file context
+ * @minor: minor to allocate on
+ *
+ * This allocates a new DRM file context. It is not linked into any context and
+ * can be used by the caller freely. Note that the context keeps a pointer to
+ * @minor, so it must be freed before @minor is.
+ *
+ * RETURNS:
+ * Pointer to newly allocated context, ERR_PTR on failure.
+ */
+struct drm_file *drm_file_alloc(struct drm_minor *minor)
+{
+	struct drm_device *dev = minor->dev;
+	struct drm_file *file;
+	int ret;
+
+	file = kzalloc(sizeof(*file), GFP_KERNEL);
+	if (!file)
+		return ERR_PTR(-ENOMEM);
+
+	file->pid = get_pid(task_pid(current));
+	file->minor = minor;
+
+	/* for compatibility root is always authenticated */
+	file->authenticated = capable(CAP_SYS_ADMIN);
+	file->lock_count = 0;
+
+	INIT_LIST_HEAD(&file->lhead);
+	INIT_LIST_HEAD(&file->fbs);
+	mutex_init(&file->fbs_lock);
+	INIT_LIST_HEAD(&file->blobs);
+	INIT_LIST_HEAD(&file->pending_event_list);
+	INIT_LIST_HEAD(&file->event_list);
+	init_waitqueue_head(&file->event_wait);
+	file->event_space = 4096; /* set aside 4k for event buffer */
+
+	mutex_init(&file->event_read_lock);
+
+	if (drm_core_check_feature(dev, DRIVER_GEM))
+		drm_gem_open(dev, file);
+
+	if (drm_core_check_feature(dev, DRIVER_SYNCOBJ))
+		drm_syncobj_open(file);
+
+	if (drm_core_check_feature(dev, DRIVER_PRIME))
+		drm_prime_init_file_private(&file->prime);
+
+	if (dev->driver->open) {
+		ret = dev->driver->open(dev, file);
+		if (ret < 0)
+			goto out_prime_destroy;
+	}
+
+	return file;
+
+out_prime_destroy:
+	if (drm_core_check_feature(dev, DRIVER_PRIME))
+		drm_prime_destroy_file_private(&file->prime);
+	if (drm_core_check_feature(dev, DRIVER_SYNCOBJ))
+		drm_syncobj_release(file);
+	if (drm_core_check_feature(dev, DRIVER_GEM))
+		drm_gem_release(dev, file);
+	put_pid(file->pid);
+	kfree(file);
+
+	return ERR_PTR(ret);
+}
+
+static void drm_events_release(struct drm_file *file_priv)
+{
+	struct drm_device *dev = file_priv->minor->dev;
+	struct drm_pending_event *e, *et;
+	unsigned long flags;
+
+	spin_lock_irqsave(&dev->event_lock, flags);
+
+	/* Unlink pending events */
+	list_for_each_entry_safe(e, et, &file_priv->pending_event_list,
+				 pending_link) {
+		list_del(&e->pending_link);
+		e->file_priv = NULL;
+	}
+
+	/* Remove unconsumed events */
+	list_for_each_entry_safe(e, et, &file_priv->event_list, link) {
+		list_del(&e->link);
+		kfree(e);
+	}
+
+	spin_unlock_irqrestore(&dev->event_lock, flags);
+}
+
+/**
+ * drm_file_free - free file context
+ * @file: context to free, or NULL
+ *
+ * This destroys and deallocates a DRM file context previously allocated via
+ * drm_file_alloc(). The caller must make sure to unlink it from any contexts
+ * before calling this.
+ *
+ * If NULL is passed, this is a no-op.
+ *
+ * RETURNS:
+ * 0 on success, or error code on failure.
+ */
+void drm_file_free(struct drm_file *file)
+{
+	struct drm_device *dev;
+
+	if (!file)
+		return;
+
+	dev = file->minor->dev;
+
+	DRM_DEBUG("pid = %d, device = 0x%lx, open_count = %d\n",
+		  task_pid_nr(current),
+		  (long)old_encode_dev(file->minor->kdev->devt),
+		  dev->open_count);
+
+	if (drm_core_check_feature(dev, DRIVER_LEGACY) &&
+	    dev->driver->preclose)
+		dev->driver->preclose(dev, file);
+
+	if (drm_core_check_feature(dev, DRIVER_LEGACY))
+		drm_legacy_lock_release(dev, file->filp);
+
+	if (drm_core_check_feature(dev, DRIVER_HAVE_DMA))
+		drm_legacy_reclaim_buffers(dev, file);
+
+	drm_events_release(file);
+
+	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
+		drm_fb_release(file);
+		drm_property_destroy_user_blobs(dev, file);
+	}
+
+	if (drm_core_check_feature(dev, DRIVER_SYNCOBJ))
+		drm_syncobj_release(file);
+
+	if (drm_core_check_feature(dev, DRIVER_GEM))
+		drm_gem_release(dev, file);
+
+	drm_legacy_ctxbitmap_flush(dev, file);
+
+	if (drm_is_primary_client(file))
+		drm_master_release(file);
+
+	if (dev->driver->postclose)
+		dev->driver->postclose(dev, file);
+
+	if (drm_core_check_feature(dev, DRIVER_PRIME))
+		drm_prime_destroy_file_private(&file->prime);
+
+	WARN_ON(!list_empty(&file->event_list));
+
+	put_pid(file->pid);
+	kfree(file);
+}
+
 static int drm_setup(struct drm_device * dev)
 {
 	int ret;
@@ -207,51 +368,22 @@ static int drm_open_helper(struct file *filp, struct drm_minor *minor)
 
 	DRM_DEBUG("pid = %d, minor = %d\n", task_pid_nr(current), minor->index);
 
-	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
-	if (!priv)
-		return -ENOMEM;
-
-	filp->private_data = priv;
-	priv->filp = filp;
-	priv->pid = get_pid(task_pid(current));
-	priv->minor = minor;
-
-	/* for compatibility root is always authenticated */
-	priv->authenticated = capable(CAP_SYS_ADMIN);
-	priv->lock_count = 0;
-
-	INIT_LIST_HEAD(&priv->lhead);
-	INIT_LIST_HEAD(&priv->fbs);
-	mutex_init(&priv->fbs_lock);
-	INIT_LIST_HEAD(&priv->blobs);
-	INIT_LIST_HEAD(&priv->pending_event_list);
-	INIT_LIST_HEAD(&priv->event_list);
-	init_waitqueue_head(&priv->event_wait);
-	priv->event_space = 4096; /* set aside 4k for event buffer */
-
-	mutex_init(&priv->event_read_lock);
-
-	if (drm_core_check_feature(dev, DRIVER_GEM))
-		drm_gem_open(dev, priv);
-
-	if (drm_core_check_feature(dev, DRIVER_SYNCOBJ))
-		drm_syncobj_open(priv);
-
-	if (drm_core_check_feature(dev, DRIVER_PRIME))
-		drm_prime_init_file_private(&priv->prime);
-
-	if (dev->driver->open) {
-		ret = dev->driver->open(dev, priv);
-		if (ret < 0)
-			goto out_prime_destroy;
-	}
+	priv = drm_file_alloc(minor);
+	if (IS_ERR(priv))
+		return PTR_ERR(priv);
 
 	if (drm_is_primary_client(priv)) {
 		ret = drm_master_open(priv);
-		if (ret)
-			goto out_close;
+		if (ret) {
+			drm_file_free(priv);
+			return ret;
+		}
 	}
 
+	filp->private_data = priv;
+	filp->f_mode |= FMODE_UNSIGNED_OFFSET;
+	priv->filp = filp;
+
 	mutex_lock(&dev->filelist_mutex);
 	list_add(&priv->lhead, &dev->filelist);
 	mutex_unlock(&dev->filelist_mutex);
@@ -277,45 +409,6 @@ static int drm_open_helper(struct file *filp, struct drm_minor *minor)
 #endif
 
 	return 0;
-
-out_close:
-	if (dev->driver->postclose)
-		dev->driver->postclose(dev, priv);
-out_prime_destroy:
-	if (drm_core_check_feature(dev, DRIVER_PRIME))
-		drm_prime_destroy_file_private(&priv->prime);
-	if (drm_core_check_feature(dev, DRIVER_SYNCOBJ))
-		drm_syncobj_release(priv);
-	if (drm_core_check_feature(dev, DRIVER_GEM))
-		drm_gem_release(dev, priv);
-	put_pid(priv->pid);
-	kfree(priv);
-	filp->private_data = NULL;
-	return ret;
-}
-
-static void drm_events_release(struct drm_file *file_priv)
-{
-	struct drm_device *dev = file_priv->minor->dev;
-	struct drm_pending_event *e, *et;
-	unsigned long flags;
-
-	spin_lock_irqsave(&dev->event_lock, flags);
-
-	/* Unlink pending events */
-	list_for_each_entry_safe(e, et, &file_priv->pending_event_list,
-				 pending_link) {
-		list_del(&e->pending_link);
-		e->file_priv = NULL;
-	}
-
-	/* Remove unconsumed events */
-	list_for_each_entry_safe(e, et, &file_priv->event_list, link) {
-		list_del(&e->link);
-		kfree(e);
-	}
-
-	spin_unlock_irqrestore(&dev->event_lock, flags);
 }
 
 static void drm_legacy_dev_reinit(struct drm_device *dev)
@@ -352,6 +445,8 @@ void drm_lastclose(struct drm_device * dev)
 
 	if (drm_core_check_feature(dev, DRIVER_LEGACY))
 		drm_legacy_dev_reinit(dev);
+
+	drm_client_dev_restore(dev);
 }
 
 /**
@@ -382,57 +477,7 @@ int drm_release(struct inode *inode, struct file *filp)
 	list_del(&file_priv->lhead);
 	mutex_unlock(&dev->filelist_mutex);
 
-	if (drm_core_check_feature(dev, DRIVER_LEGACY) &&
-	    dev->driver->preclose)
-		dev->driver->preclose(dev, file_priv);
-
-	/* ========================================================
-	 * Begin inline drm_release
-	 */
-
-	DRM_DEBUG("pid = %d, device = 0x%lx, open_count = %d\n",
-		  task_pid_nr(current),
-		  (long)old_encode_dev(file_priv->minor->kdev->devt),
-		  dev->open_count);
-
-	if (drm_core_check_feature(dev, DRIVER_LEGACY))
-		drm_legacy_lock_release(dev, filp);
-
-	if (drm_core_check_feature(dev, DRIVER_HAVE_DMA))
-		drm_legacy_reclaim_buffers(dev, file_priv);
-
-	drm_events_release(file_priv);
-
-	if (drm_core_check_feature(dev, DRIVER_MODESET)) {
-		drm_fb_release(file_priv);
-		drm_property_destroy_user_blobs(dev, file_priv);
-	}
-
-	if (drm_core_check_feature(dev, DRIVER_SYNCOBJ))
-		drm_syncobj_release(file_priv);
-
-	if (drm_core_check_feature(dev, DRIVER_GEM))
-		drm_gem_release(dev, file_priv);
-
-	drm_legacy_ctxbitmap_flush(dev, file_priv);
-
-	if (drm_is_primary_client(file_priv))
-		drm_master_release(file_priv);
-
-	if (dev->driver->postclose)
-		dev->driver->postclose(dev, file_priv);
-
-	if (drm_core_check_feature(dev, DRIVER_PRIME))
-		drm_prime_destroy_file_private(&file_priv->prime);
-
-	WARN_ON(!list_empty(&file_priv->event_list));
-
-	put_pid(file_priv->pid);
-	kfree(file_priv);
-
-	/* ========================================================
-	 * End inline drm_release
-	 */
+	drm_file_free(file_priv);
 
 	if (!--dev->open_count) {
 		drm_lastclose(dev);
diff --git a/drivers/gpu/drm/drm_fourcc.c b/drivers/gpu/drm/drm_fourcc.c
index 5ca6395cd4d3..35c1e2742c27 100644
--- a/drivers/gpu/drm/drm_fourcc.c
+++ b/drivers/gpu/drm/drm_fourcc.c
@@ -152,27 +152,27 @@ const struct drm_format_info *__drm_format_info(u32 format)
 		{ .format = DRM_FORMAT_XBGR8888_A8,	.depth = 32, .num_planes = 2, .cpp = { 4, 1, 0 }, .hsub = 1, .vsub = 1, .has_alpha = true },
 		{ .format = DRM_FORMAT_RGBX8888_A8,	.depth = 32, .num_planes = 2, .cpp = { 4, 1, 0 }, .hsub = 1, .vsub = 1, .has_alpha = true },
 		{ .format = DRM_FORMAT_BGRX8888_A8,	.depth = 32, .num_planes = 2, .cpp = { 4, 1, 0 }, .hsub = 1, .vsub = 1, .has_alpha = true },
-		{ .format = DRM_FORMAT_YUV410,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 4, .vsub = 4 },
-		{ .format = DRM_FORMAT_YVU410,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 4, .vsub = 4 },
-		{ .format = DRM_FORMAT_YUV411,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 4, .vsub = 1 },
-		{ .format = DRM_FORMAT_YVU411,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 4, .vsub = 1 },
-		{ .format = DRM_FORMAT_YUV420,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 2, .vsub = 2 },
-		{ .format = DRM_FORMAT_YVU420,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 2, .vsub = 2 },
-		{ .format = DRM_FORMAT_YUV422,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 2, .vsub = 1 },
-		{ .format = DRM_FORMAT_YVU422,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 2, .vsub = 1 },
-		{ .format = DRM_FORMAT_YUV444,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 1, .vsub = 1 },
-		{ .format = DRM_FORMAT_YVU444,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 1, .vsub = 1 },
-		{ .format = DRM_FORMAT_NV12,		.depth = 0,  .num_planes = 2, .cpp = { 1, 2, 0 }, .hsub = 2, .vsub = 2 },
-		{ .format = DRM_FORMAT_NV21,		.depth = 0,  .num_planes = 2, .cpp = { 1, 2, 0 }, .hsub = 2, .vsub = 2 },
-		{ .format = DRM_FORMAT_NV16,		.depth = 0,  .num_planes = 2, .cpp = { 1, 2, 0 }, .hsub = 2, .vsub = 1 },
-		{ .format = DRM_FORMAT_NV61,		.depth = 0,  .num_planes = 2, .cpp = { 1, 2, 0 }, .hsub = 2, .vsub = 1 },
-		{ .format = DRM_FORMAT_NV24,		.depth = 0,  .num_planes = 2, .cpp = { 1, 2, 0 }, .hsub = 1, .vsub = 1 },
-		{ .format = DRM_FORMAT_NV42,		.depth = 0,  .num_planes = 2, .cpp = { 1, 2, 0 }, .hsub = 1, .vsub = 1 },
-		{ .format = DRM_FORMAT_YUYV,		.depth = 0,  .num_planes = 1, .cpp = { 2, 0, 0 }, .hsub = 2, .vsub = 1 },
-		{ .format = DRM_FORMAT_YVYU,		.depth = 0,  .num_planes = 1, .cpp = { 2, 0, 0 }, .hsub = 2, .vsub = 1 },
-		{ .format = DRM_FORMAT_UYVY,		.depth = 0,  .num_planes = 1, .cpp = { 2, 0, 0 }, .hsub = 2, .vsub = 1 },
-		{ .format = DRM_FORMAT_VYUY,		.depth = 0,  .num_planes = 1, .cpp = { 2, 0, 0 }, .hsub = 2, .vsub = 1 },
-		{ .format = DRM_FORMAT_AYUV,		.depth = 0,  .num_planes = 1, .cpp = { 4, 0, 0 }, .hsub = 1, .vsub = 1, .has_alpha = true },
+		{ .format = DRM_FORMAT_YUV410,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 4, .vsub = 4, .is_yuv = true },
+		{ .format = DRM_FORMAT_YVU410,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 4, .vsub = 4, .is_yuv = true },
+		{ .format = DRM_FORMAT_YUV411,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 4, .vsub = 1, .is_yuv = true },
+		{ .format = DRM_FORMAT_YVU411,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 4, .vsub = 1, .is_yuv = true },
+		{ .format = DRM_FORMAT_YUV420,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 2, .vsub = 2, .is_yuv = true },
+		{ .format = DRM_FORMAT_YVU420,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 2, .vsub = 2, .is_yuv = true },
+		{ .format = DRM_FORMAT_YUV422,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 2, .vsub = 1, .is_yuv = true },
+		{ .format = DRM_FORMAT_YVU422,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 2, .vsub = 1, .is_yuv = true },
+		{ .format = DRM_FORMAT_YUV444,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 1, .vsub = 1, .is_yuv = true },
+		{ .format = DRM_FORMAT_YVU444,		.depth = 0,  .num_planes = 3, .cpp = { 1, 1, 1 }, .hsub = 1, .vsub = 1, .is_yuv = true },
+		{ .format = DRM_FORMAT_NV12,		.depth = 0,  .num_planes = 2, .cpp = { 1, 2, 0 }, .hsub = 2, .vsub = 2, .is_yuv = true },
+		{ .format = DRM_FORMAT_NV21,		.depth = 0,  .num_planes = 2, .cpp = { 1, 2, 0 }, .hsub = 2, .vsub = 2, .is_yuv = true },
+		{ .format = DRM_FORMAT_NV16,		.depth = 0,  .num_planes = 2, .cpp = { 1, 2, 0 }, .hsub = 2, .vsub = 1, .is_yuv = true },
+		{ .format = DRM_FORMAT_NV61,		.depth = 0,  .num_planes = 2, .cpp = { 1, 2, 0 }, .hsub = 2, .vsub = 1, .is_yuv = true },
+		{ .format = DRM_FORMAT_NV24,		.depth = 0,  .num_planes = 2, .cpp = { 1, 2, 0 }, .hsub = 1, .vsub = 1, .is_yuv = true },
+		{ .format = DRM_FORMAT_NV42,		.depth = 0,  .num_planes = 2, .cpp = { 1, 2, 0 }, .hsub = 1, .vsub = 1, .is_yuv = true },
+		{ .format = DRM_FORMAT_YUYV,		.depth = 0,  .num_planes = 1, .cpp = { 2, 0, 0 }, .hsub = 2, .vsub = 1, .is_yuv = true },
+		{ .format = DRM_FORMAT_YVYU,		.depth = 0,  .num_planes = 1, .cpp = { 2, 0, 0 }, .hsub = 2, .vsub = 1, .is_yuv = true },
+		{ .format = DRM_FORMAT_UYVY,		.depth = 0,  .num_planes = 1, .cpp = { 2, 0, 0 }, .hsub = 2, .vsub = 1, .is_yuv = true },
+		{ .format = DRM_FORMAT_VYUY,		.depth = 0,  .num_planes = 1, .cpp = { 2, 0, 0 }, .hsub = 2, .vsub = 1, .is_yuv = true },
+		{ .format = DRM_FORMAT_AYUV,		.depth = 0,  .num_planes = 1, .cpp = { 4, 0, 0 }, .hsub = 1, .vsub = 1, .has_alpha = true, .is_yuv = true },
 	};
 
 	unsigned int i;
diff --git a/drivers/gpu/drm/drm_framebuffer.c b/drivers/gpu/drm/drm_framebuffer.c
index 8c4d32adcc17..781af1d42d76 100644
--- a/drivers/gpu/drm/drm_framebuffer.c
+++ b/drivers/gpu/drm/drm_framebuffer.c
@@ -95,21 +95,20 @@ int drm_framebuffer_check_src_coords(uint32_t src_x, uint32_t src_y,
 /**
  * drm_mode_addfb - add an FB to the graphics configuration
  * @dev: drm device for the ioctl
- * @data: data pointer for the ioctl
- * @file_priv: drm file for the ioctl call
+ * @or: pointer to request structure
+ * @file_priv: drm file
  *
  * Add a new FB to the specified CRTC, given a user request. This is the
  * original addfb ioctl which only supported RGB formats.
  *
- * Called by the user via ioctl.
+ * Called by the user via ioctl, or by an in-kernel client.
  *
  * Returns:
  * Zero on success, negative errno on failure.
  */
-int drm_mode_addfb(struct drm_device *dev,
-		   void *data, struct drm_file *file_priv)
+int drm_mode_addfb(struct drm_device *dev, struct drm_mode_fb_cmd *or,
+		   struct drm_file *file_priv)
 {
-	struct drm_mode_fb_cmd *or = data;
 	struct drm_mode_fb_cmd2 r = {};
 	int ret;
 
@@ -134,6 +133,12 @@ int drm_mode_addfb(struct drm_device *dev,
 	return 0;
 }
 
+int drm_mode_addfb_ioctl(struct drm_device *dev,
+			 void *data, struct drm_file *file_priv)
+{
+	return drm_mode_addfb(dev, data, file_priv);
+}
+
 static int fb_plane_width(int width,
 			  const struct drm_format_info *format, int plane)
 {
@@ -367,29 +372,28 @@ static void drm_mode_rmfb_work_fn(struct work_struct *w)
 
 /**
  * drm_mode_rmfb - remove an FB from the configuration
- * @dev: drm device for the ioctl
- * @data: data pointer for the ioctl
- * @file_priv: drm file for the ioctl call
+ * @dev: drm device
+ * @fb_id: id of framebuffer to remove
+ * @file_priv: drm file
  *
- * Remove the FB specified by the user.
+ * Remove the specified FB.
  *
- * Called by the user via ioctl.
+ * Called by the user via ioctl, or by an in-kernel client.
  *
  * Returns:
  * Zero on success, negative errno on failure.
  */
-int drm_mode_rmfb(struct drm_device *dev,
-		   void *data, struct drm_file *file_priv)
+int drm_mode_rmfb(struct drm_device *dev, u32 fb_id,
+		  struct drm_file *file_priv)
 {
 	struct drm_framebuffer *fb = NULL;
 	struct drm_framebuffer *fbl = NULL;
-	uint32_t *id = data;
 	int found = 0;
 
 	if (!drm_core_check_feature(dev, DRIVER_MODESET))
 		return -EINVAL;
 
-	fb = drm_framebuffer_lookup(dev, file_priv, *id);
+	fb = drm_framebuffer_lookup(dev, file_priv, fb_id);
 	if (!fb)
 		return -ENOENT;
 
@@ -435,6 +439,14 @@ fail_unref:
 	return -ENOENT;
 }
 
+int drm_mode_rmfb_ioctl(struct drm_device *dev,
+			void *data, struct drm_file *file_priv)
+{
+	uint32_t *fb_id = data;
+
+	return drm_mode_rmfb(dev, *fb_id, file_priv);
+}
+
 /**
  * drm_mode_getfb - get FB info
  * @dev: drm device for the ioctl
@@ -484,8 +496,7 @@ int drm_mode_getfb(struct drm_device *dev,
 	 * backwards-compatibility reasons, we cannot make GET_FB() privileged,
 	 * so just return an invalid handle for non-masters.
 	 */
-	if (!drm_is_current_master(file_priv) && !capable(CAP_SYS_ADMIN) &&
-	    !drm_is_control_client(file_priv)) {
+	if (!drm_is_current_master(file_priv) && !capable(CAP_SYS_ADMIN)) {
 		r->handle = 0;
 		ret = 0;
 		goto out;
@@ -836,9 +847,7 @@ retry:
 		if (ret)
 			goto unlock;
 
-		plane_mask |= BIT(drm_plane_index(plane));
-
-		plane->old_fb = plane->fb;
+		plane_mask |= drm_plane_mask(plane);
 	}
 
 	/* This list is only filled when disable_crtcs is set. */
@@ -853,9 +862,6 @@ retry:
 		ret = drm_atomic_commit(state);
 
 unlock:
-	if (plane_mask)
-		drm_atomic_clean_old_fb(dev, plane_mask, ret);
-
 	if (ret == -EDEADLK) {
 		drm_atomic_state_clear(state);
 		drm_modeset_backoff(&ctx);
diff --git a/drivers/gpu/drm/drm_gem_framebuffer_helper.c b/drivers/gpu/drm/drm_gem_framebuffer_helper.c
index acfbc0641a06..2810d4131411 100644
--- a/drivers/gpu/drm/drm_gem_framebuffer_helper.c
+++ b/drivers/gpu/drm/drm_gem_framebuffer_helper.c
@@ -253,7 +253,7 @@ int drm_gem_fb_prepare_fb(struct drm_plane *plane,
 	struct dma_buf *dma_buf;
 	struct dma_fence *fence;
 
-	if (plane->state->fb == state->fb || !state->fb)
+	if (!state->fb)
 		return 0;
 
 	dma_buf = drm_gem_fb_get_obj(state->fb, 0)->dma_buf;
diff --git a/drivers/gpu/drm/drm_global.c b/drivers/gpu/drm/drm_global.c
index b2dc21e33ae0..5799e2782dd1 100644
--- a/drivers/gpu/drm/drm_global.c
+++ b/drivers/gpu/drm/drm_global.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
  * Copyright 2008-2009 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/drm_hashtab.c b/drivers/gpu/drm/drm_hashtab.c
index dae18e58e79b..c92b00d42ece 100644
--- a/drivers/gpu/drm/drm_hashtab.c
+++ b/drivers/gpu/drm/drm_hashtab.c
@@ -47,7 +47,7 @@ int drm_ht_create(struct drm_open_hash *ht, unsigned int order)
 	if (size <= PAGE_SIZE / sizeof(*ht->table))
 		ht->table = kcalloc(size, sizeof(*ht->table), GFP_KERNEL);
 	else
-		ht->table = vzalloc(size*sizeof(*ht->table));
+		ht->table = vzalloc(array_size(size, sizeof(*ht->table)));
 	if (!ht->table) {
 		DRM_ERROR("Out of memory for hash table\n");
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/drm_internal.h b/drivers/gpu/drm/drm_internal.h
index b72242e93ea4..40179c5fc6b8 100644
--- a/drivers/gpu/drm/drm_internal.h
+++ b/drivers/gpu/drm/drm_internal.h
@@ -26,6 +26,8 @@
 
 /* drm_file.c */
 extern struct mutex drm_global_mutex;
+struct drm_file *drm_file_alloc(struct drm_minor *minor);
+void drm_file_free(struct drm_file *file);
 void drm_lastclose(struct drm_device *dev);
 
 /* drm_pci.c */
diff --git a/drivers/gpu/drm/drm_ioc32.c b/drivers/gpu/drm/drm_ioc32.c
index f8e96e648acf..67b1fca39aa6 100644
--- a/drivers/gpu/drm/drm_ioc32.c
+++ b/drivers/gpu/drm/drm_ioc32.c
@@ -105,7 +105,7 @@ static int compat_drm_version(struct file *file, unsigned int cmd,
 		.desc = compat_ptr(v32.desc),
 	};
 	err = drm_ioctl_kernel(file, drm_version, &v,
-			DRM_UNLOCKED|DRM_RENDER_ALLOW|DRM_CONTROL_ALLOW);
+			       DRM_UNLOCKED|DRM_RENDER_ALLOW);
 	if (err)
 		return err;
 
@@ -885,7 +885,7 @@ static int compat_drm_mode_addfb2(struct file *file, unsigned int cmd,
 		return -EFAULT;
 
 	err = drm_ioctl_kernel(file, drm_mode_addfb2, &req64,
-				DRM_CONTROL_ALLOW|DRM_UNLOCKED);
+			       DRM_UNLOCKED);
 	if (err)
 		return err;
 
diff --git a/drivers/gpu/drm/drm_ioctl.c b/drivers/gpu/drm/drm_ioctl.c
index af782911c505..ea10e9a26aad 100644
--- a/drivers/gpu/drm/drm_ioctl.c
+++ b/drivers/gpu/drm/drm_ioctl.c
@@ -324,6 +324,22 @@ drm_setclientcap(struct drm_device *dev, void *data, struct drm_file *file_priv)
 			return -EINVAL;
 		file_priv->atomic = req->value;
 		file_priv->universal_planes = req->value;
+		/*
+		 * No atomic user-space blows up on aspect ratio mode bits.
+		 */
+		file_priv->aspect_ratio_allowed = req->value;
+		break;
+	case DRM_CLIENT_CAP_ASPECT_RATIO:
+		if (req->value > 1)
+			return -EINVAL;
+		file_priv->aspect_ratio_allowed = req->value;
+		break;
+	case DRM_CLIENT_CAP_WRITEBACK_CONNECTORS:
+		if (!file_priv->atomic)
+			return -EINVAL;
+		if (req->value > 1)
+			return -EINVAL;
+		file_priv->writeback_connectors = req->value;
 		break;
 	default:
 		return -EINVAL;
@@ -510,13 +526,7 @@ int drm_ioctl_permit(u32 flags, struct drm_file *file_priv)
 
 	/* MASTER is only for master or control clients */
 	if (unlikely((flags & DRM_MASTER) &&
-		     !drm_is_current_master(file_priv) &&
-		     !drm_is_control_client(file_priv)))
-		return -EACCES;
-
-	/* Control clients must be explicitly allowed */
-	if (unlikely(!(flags & DRM_CONTROL_ALLOW) &&
-		     drm_is_control_client(file_priv)))
+		     !drm_is_current_master(file_priv)))
 		return -EACCES;
 
 	/* Render clients must be explicitly allowed */
@@ -539,7 +549,7 @@ EXPORT_SYMBOL(drm_ioctl_permit);
 /* Ioctl table */
 static const struct drm_ioctl_desc drm_ioctls[] = {
 	DRM_IOCTL_DEF(DRM_IOCTL_VERSION, drm_version,
-		      DRM_UNLOCKED|DRM_RENDER_ALLOW|DRM_CONTROL_ALLOW),
+		      DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF(DRM_IOCTL_GET_UNIQUE, drm_getunique, DRM_UNLOCKED),
 	DRM_IOCTL_DEF(DRM_IOCTL_GET_MAGIC, drm_getmagic, DRM_UNLOCKED),
 	DRM_IOCTL_DEF(DRM_IOCTL_IRQ_BUSID, drm_irq_by_busid, DRM_MASTER|DRM_ROOT_ONLY),
@@ -613,41 +623,41 @@ static const struct drm_ioctl_desc drm_ioctls[] = {
 	DRM_IOCTL_DEF(DRM_IOCTL_GEM_FLINK, drm_gem_flink_ioctl, DRM_AUTH|DRM_UNLOCKED),
 	DRM_IOCTL_DEF(DRM_IOCTL_GEM_OPEN, drm_gem_open_ioctl, DRM_AUTH|DRM_UNLOCKED),
 
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETRESOURCES, drm_mode_getresources, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETRESOURCES, drm_mode_getresources, DRM_UNLOCKED),
 
 	DRM_IOCTL_DEF(DRM_IOCTL_PRIME_HANDLE_TO_FD, drm_prime_handle_to_fd_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF(DRM_IOCTL_PRIME_FD_TO_HANDLE, drm_prime_fd_to_handle_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW),
 
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPLANERESOURCES, drm_mode_getplane_res, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETCRTC, drm_mode_getcrtc, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETCRTC, drm_mode_setcrtc, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPLANE, drm_mode_getplane, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETPLANE, drm_mode_setplane, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CURSOR, drm_mode_cursor_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPLANERESOURCES, drm_mode_getplane_res, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETCRTC, drm_mode_getcrtc, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETCRTC, drm_mode_setcrtc, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPLANE, drm_mode_getplane, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETPLANE, drm_mode_setplane, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CURSOR, drm_mode_cursor_ioctl, DRM_MASTER|DRM_UNLOCKED),
 	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETGAMMA, drm_mode_gamma_get_ioctl, DRM_UNLOCKED),
 	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETGAMMA, drm_mode_gamma_set_ioctl, DRM_MASTER|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETENCODER, drm_mode_getencoder, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETCONNECTOR, drm_mode_getconnector, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ATTACHMODE, drm_noop, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DETACHMODE, drm_noop, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPROPERTY, drm_mode_getproperty_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETPROPERTY, drm_mode_connector_property_set_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPROPBLOB, drm_mode_getblob_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETFB, drm_mode_getfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ADDFB, drm_mode_addfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ADDFB2, drm_mode_addfb2, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_RMFB, drm_mode_rmfb, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_PAGE_FLIP, drm_mode_page_flip_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DIRTYFB, drm_mode_dirtyfb_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CREATE_DUMB, drm_mode_create_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_MAP_DUMB, drm_mode_mmap_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DESTROY_DUMB, drm_mode_destroy_dumb_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_OBJ_GETPROPERTIES, drm_mode_obj_get_properties_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_OBJ_SETPROPERTY, drm_mode_obj_set_property_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CURSOR2, drm_mode_cursor2_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ATOMIC, drm_mode_atomic_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CREATEPROPBLOB, drm_mode_createblob_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DESTROYPROPBLOB, drm_mode_destroyblob_ioctl, DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETENCODER, drm_mode_getencoder, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETCONNECTOR, drm_mode_getconnector, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ATTACHMODE, drm_noop, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DETACHMODE, drm_noop, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPROPERTY, drm_mode_getproperty_ioctl, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_SETPROPERTY, drm_connector_property_set_ioctl, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETPROPBLOB, drm_mode_getblob_ioctl, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GETFB, drm_mode_getfb, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ADDFB, drm_mode_addfb_ioctl, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ADDFB2, drm_mode_addfb2, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_RMFB, drm_mode_rmfb_ioctl, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_PAGE_FLIP, drm_mode_page_flip_ioctl, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DIRTYFB, drm_mode_dirtyfb_ioctl, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CREATE_DUMB, drm_mode_create_dumb_ioctl, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_MAP_DUMB, drm_mode_mmap_dumb_ioctl, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DESTROY_DUMB, drm_mode_destroy_dumb_ioctl, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_OBJ_GETPROPERTIES, drm_mode_obj_get_properties_ioctl, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_OBJ_SETPROPERTY, drm_mode_obj_set_property_ioctl, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CURSOR2, drm_mode_cursor2_ioctl, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_ATOMIC, drm_mode_atomic_ioctl, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CREATEPROPBLOB, drm_mode_createblob_ioctl, DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_DESTROYPROPBLOB, drm_mode_destroyblob_ioctl, DRM_UNLOCKED),
 
 	DRM_IOCTL_DEF(DRM_IOCTL_SYNCOBJ_CREATE, drm_syncobj_create_ioctl,
 		      DRM_UNLOCKED|DRM_RENDER_ALLOW),
@@ -665,10 +675,10 @@ static const struct drm_ioctl_desc drm_ioctls[] = {
 		      DRM_UNLOCKED|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF(DRM_IOCTL_CRTC_GET_SEQUENCE, drm_crtc_get_sequence_ioctl, DRM_UNLOCKED),
 	DRM_IOCTL_DEF(DRM_IOCTL_CRTC_QUEUE_SEQUENCE, drm_crtc_queue_sequence_ioctl, DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CREATE_LEASE, drm_mode_create_lease_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_LIST_LESSEES, drm_mode_list_lessees_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GET_LEASE, drm_mode_get_lease_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
-	DRM_IOCTL_DEF(DRM_IOCTL_MODE_REVOKE_LEASE, drm_mode_revoke_lease_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_CREATE_LEASE, drm_mode_create_lease_ioctl, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_LIST_LESSEES, drm_mode_list_lessees_ioctl, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_GET_LEASE, drm_mode_get_lease_ioctl, DRM_MASTER|DRM_UNLOCKED),
+	DRM_IOCTL_DEF(DRM_IOCTL_MODE_REVOKE_LEASE, drm_mode_revoke_lease_ioctl, DRM_MASTER|DRM_UNLOCKED),
 };
 
 #define DRM_CORE_IOCTL_COUNT	ARRAY_SIZE( drm_ioctls )
diff --git a/drivers/gpu/drm/drm_memory.c b/drivers/gpu/drm/drm_memory.c
index 3c54044214db..d69e4fc1ee77 100644
--- a/drivers/gpu/drm/drm_memory.c
+++ b/drivers/gpu/drm/drm_memory.c
@@ -80,7 +80,7 @@ static void *agp_remap(unsigned long offset, unsigned long size,
 	 * page-table instead (that's probably faster anyhow...).
 	 */
 	/* note: use vmalloc() because num_pages could be large... */
-	page_map = vmalloc(num_pages * sizeof(struct page *));
+	page_map = vmalloc(array_size(num_pages, sizeof(struct page *)));
 	if (!page_map)
 		return NULL;
 
diff --git a/drivers/gpu/drm/drm_mm.c b/drivers/gpu/drm/drm_mm.c
index 3166026a1874..3cc5fbd78ee2 100644
--- a/drivers/gpu/drm/drm_mm.c
+++ b/drivers/gpu/drm/drm_mm.c
@@ -239,6 +239,32 @@ static void drm_mm_interval_tree_add_node(struct drm_mm_node *hole_node,
 #define HOLE_SIZE(NODE) ((NODE)->hole_size)
 #define HOLE_ADDR(NODE) (__drm_mm_hole_node_start(NODE))
 
+static u64 rb_to_hole_size(struct rb_node *rb)
+{
+	return rb_entry(rb, struct drm_mm_node, rb_hole_size)->hole_size;
+}
+
+static void insert_hole_size(struct rb_root_cached *root,
+			     struct drm_mm_node *node)
+{
+	struct rb_node **link = &root->rb_root.rb_node, *rb = NULL;
+	u64 x = node->hole_size;
+	bool first = true;
+
+	while (*link) {
+		rb = *link;
+		if (x > rb_to_hole_size(rb)) {
+			link = &rb->rb_left;
+		} else {
+			link = &rb->rb_right;
+			first = false;
+		}
+	}
+
+	rb_link_node(&node->rb_hole_size, rb, link);
+	rb_insert_color_cached(&node->rb_hole_size, root, first);
+}
+
 static void add_hole(struct drm_mm_node *node)
 {
 	struct drm_mm *mm = node->mm;
@@ -247,7 +273,7 @@ static void add_hole(struct drm_mm_node *node)
 		__drm_mm_hole_node_end(node) - __drm_mm_hole_node_start(node);
 	DRM_MM_BUG_ON(!drm_mm_hole_follows(node));
 
-	RB_INSERT(mm->holes_size, rb_hole_size, HOLE_SIZE);
+	insert_hole_size(&mm->holes_size, node);
 	RB_INSERT(mm->holes_addr, rb_hole_addr, HOLE_ADDR);
 
 	list_add(&node->hole_stack, &mm->hole_stack);
@@ -258,7 +284,7 @@ static void rm_hole(struct drm_mm_node *node)
 	DRM_MM_BUG_ON(!drm_mm_hole_follows(node));
 
 	list_del(&node->hole_stack);
-	rb_erase(&node->rb_hole_size, &node->mm->holes_size);
+	rb_erase_cached(&node->rb_hole_size, &node->mm->holes_size);
 	rb_erase(&node->rb_hole_addr, &node->mm->holes_addr);
 	node->hole_size = 0;
 
@@ -282,38 +308,39 @@ static inline u64 rb_hole_size(struct rb_node *rb)
 
 static struct drm_mm_node *best_hole(struct drm_mm *mm, u64 size)
 {
-	struct rb_node *best = NULL;
-	struct rb_node **link = &mm->holes_size.rb_node;
+	struct rb_node *rb = mm->holes_size.rb_root.rb_node;
+	struct drm_mm_node *best = NULL;
 
-	while (*link) {
-		struct rb_node *rb = *link;
+	do {
+		struct drm_mm_node *node =
+			rb_entry(rb, struct drm_mm_node, rb_hole_size);
 
-		if (size <= rb_hole_size(rb)) {
-			link = &rb->rb_left;
-			best = rb;
+		if (size <= node->hole_size) {
+			best = node;
+			rb = rb->rb_right;
 		} else {
-			link = &rb->rb_right;
+			rb = rb->rb_left;
 		}
-	}
+	} while (rb);
 
-	return rb_hole_size_to_node(best);
+	return best;
 }
 
 static struct drm_mm_node *find_hole(struct drm_mm *mm, u64 addr)
 {
+	struct rb_node *rb = mm->holes_addr.rb_node;
 	struct drm_mm_node *node = NULL;
-	struct rb_node **link = &mm->holes_addr.rb_node;
 
-	while (*link) {
+	while (rb) {
 		u64 hole_start;
 
-		node = rb_hole_addr_to_node(*link);
+		node = rb_hole_addr_to_node(rb);
 		hole_start = __drm_mm_hole_node_start(node);
 
 		if (addr < hole_start)
-			link = &node->rb_hole_addr.rb_left;
+			rb = node->rb_hole_addr.rb_left;
 		else if (addr > hole_start + node->hole_size)
-			link = &node->rb_hole_addr.rb_right;
+			rb = node->rb_hole_addr.rb_right;
 		else
 			break;
 	}
@@ -326,9 +353,6 @@ first_hole(struct drm_mm *mm,
 	   u64 start, u64 end, u64 size,
 	   enum drm_mm_insert_mode mode)
 {
-	if (RB_EMPTY_ROOT(&mm->holes_size))
-		return NULL;
-
 	switch (mode) {
 	default:
 	case DRM_MM_INSERT_BEST:
@@ -355,7 +379,7 @@ next_hole(struct drm_mm *mm,
 	switch (mode) {
 	default:
 	case DRM_MM_INSERT_BEST:
-		return rb_hole_size_to_node(rb_next(&node->rb_hole_size));
+		return rb_hole_size_to_node(rb_prev(&node->rb_hole_size));
 
 	case DRM_MM_INSERT_LOW:
 		return rb_hole_addr_to_node(rb_next(&node->rb_hole_addr));
@@ -426,6 +450,11 @@ int drm_mm_reserve_node(struct drm_mm *mm, struct drm_mm_node *node)
 }
 EXPORT_SYMBOL(drm_mm_reserve_node);
 
+static u64 rb_to_hole_size_or_zero(struct rb_node *rb)
+{
+	return rb ? rb_to_hole_size(rb) : 0;
+}
+
 /**
  * drm_mm_insert_node_in_range - ranged search for space and insert @node
  * @mm: drm_mm to allocate from
@@ -451,18 +480,26 @@ int drm_mm_insert_node_in_range(struct drm_mm * const mm,
 {
 	struct drm_mm_node *hole;
 	u64 remainder_mask;
+	bool once;
 
 	DRM_MM_BUG_ON(range_start >= range_end);
 
 	if (unlikely(size == 0 || range_end - range_start < size))
 		return -ENOSPC;
 
+	if (rb_to_hole_size_or_zero(rb_first_cached(&mm->holes_size)) < size)
+		return -ENOSPC;
+
 	if (alignment <= 1)
 		alignment = 0;
 
+	once = mode & DRM_MM_INSERT_ONCE;
+	mode &= ~DRM_MM_INSERT_ONCE;
+
 	remainder_mask = is_power_of_2(alignment) ? alignment - 1 : 0;
-	for (hole = first_hole(mm, range_start, range_end, size, mode); hole;
-	     hole = next_hole(mm, hole, mode)) {
+	for (hole = first_hole(mm, range_start, range_end, size, mode);
+	     hole;
+	     hole = once ? NULL : next_hole(mm, hole, mode)) {
 		u64 hole_start = __drm_mm_hole_node_start(hole);
 		u64 hole_end = hole_start + hole->hole_size;
 		u64 adj_start, adj_end;
@@ -587,9 +624,9 @@ void drm_mm_replace_node(struct drm_mm_node *old, struct drm_mm_node *new)
 
 	if (drm_mm_hole_follows(old)) {
 		list_replace(&old->hole_stack, &new->hole_stack);
-		rb_replace_node(&old->rb_hole_size,
-				&new->rb_hole_size,
-				&mm->holes_size);
+		rb_replace_node_cached(&old->rb_hole_size,
+				       &new->rb_hole_size,
+				       &mm->holes_size);
 		rb_replace_node(&old->rb_hole_addr,
 				&new->rb_hole_addr,
 				&mm->holes_addr);
@@ -885,7 +922,7 @@ void drm_mm_init(struct drm_mm *mm, u64 start, u64 size)
 
 	INIT_LIST_HEAD(&mm->hole_stack);
 	mm->interval_tree = RB_ROOT_CACHED;
-	mm->holes_size = RB_ROOT;
+	mm->holes_size = RB_ROOT_CACHED;
 	mm->holes_addr = RB_ROOT;
 
 	/* Clever trick to avoid a special case in the free hole tracking. */
diff --git a/drivers/gpu/drm/drm_mode_config.c b/drivers/gpu/drm/drm_mode_config.c
index e5c653357024..21e353bd3948 100644
--- a/drivers/gpu/drm/drm_mode_config.c
+++ b/drivers/gpu/drm/drm_mode_config.c
@@ -145,6 +145,11 @@ int drm_mode_getresources(struct drm_device *dev, void *data,
 	count = 0;
 	connector_id = u64_to_user_ptr(card_res->connector_id_ptr);
 	drm_for_each_connector_iter(connector, &conn_iter) {
+		/* only expose writeback connectors if userspace understands them */
+		if (!file_priv->writeback_connectors &&
+		    (connector->connector_type == DRM_MODE_CONNECTOR_WRITEBACK))
+			continue;
+
 		if (drm_lease_held(file_priv, connector->base.id)) {
 			if (count < card_res->count_connectors &&
 			    put_user(connector->base.id, connector_id + count)) {
diff --git a/drivers/gpu/drm/drm_mode_object.c b/drivers/gpu/drm/drm_mode_object.c
index ce4d2fb32810..fcb0ab0abb75 100644
--- a/drivers/gpu/drm/drm_mode_object.c
+++ b/drivers/gpu/drm/drm_mode_object.c
@@ -433,8 +433,7 @@ static int set_property_legacy(struct drm_mode_object *obj,
 	drm_modeset_lock_all(dev);
 	switch (obj->type) {
 	case DRM_MODE_OBJECT_CONNECTOR:
-		ret = drm_mode_connector_set_obj_prop(obj, prop,
-						      prop_value);
+		ret = drm_connector_set_obj_prop(obj, prop, prop_value);
 		break;
 	case DRM_MODE_OBJECT_CRTC:
 		ret = drm_mode_crtc_set_obj_prop(obj, prop, prop_value);
diff --git a/drivers/gpu/drm/drm_modes.c b/drivers/gpu/drm/drm_modes.c
index e82b61e08f8c..02db9ac82d7a 100644
--- a/drivers/gpu/drm/drm_modes.c
+++ b/drivers/gpu/drm/drm_modes.c
@@ -659,10 +659,12 @@ EXPORT_SYMBOL_GPL(drm_display_mode_to_videomode);
  * drm_bus_flags_from_videomode - extract information about pixelclk and
  * DE polarity from videomode and store it in a separate variable
  * @vm: videomode structure to use
- * @bus_flags: information about pixelclk and DE polarity will be stored here
+ * @bus_flags: information about pixelclk, sync and DE polarity will be stored
+ * here
  *
- * Sets DRM_BUS_FLAG_DE_(LOW|HIGH) and DRM_BUS_FLAG_PIXDATA_(POS|NEG)EDGE
- * in @bus_flags according to DISPLAY_FLAGS found in @vm
+ * Sets DRM_BUS_FLAG_DE_(LOW|HIGH),  DRM_BUS_FLAG_PIXDATA_(POS|NEG)EDGE and
+ * DISPLAY_FLAGS_SYNC_(POS|NEG)EDGE in @bus_flags according to DISPLAY_FLAGS
+ * found in @vm
  */
 void drm_bus_flags_from_videomode(const struct videomode *vm, u32 *bus_flags)
 {
@@ -672,6 +674,11 @@ void drm_bus_flags_from_videomode(const struct videomode *vm, u32 *bus_flags)
 	if (vm->flags & DISPLAY_FLAGS_PIXDATA_NEGEDGE)
 		*bus_flags |= DRM_BUS_FLAG_PIXDATA_NEGEDGE;
 
+	if (vm->flags & DISPLAY_FLAGS_SYNC_POSEDGE)
+		*bus_flags |= DRM_BUS_FLAG_SYNC_POSEDGE;
+	if (vm->flags & DISPLAY_FLAGS_SYNC_NEGEDGE)
+		*bus_flags |= DRM_BUS_FLAG_SYNC_NEGEDGE;
+
 	if (vm->flags & DISPLAY_FLAGS_DE_LOW)
 		*bus_flags |= DRM_BUS_FLAG_DE_LOW;
 	if (vm->flags & DISPLAY_FLAGS_DE_HIGH)
@@ -684,7 +691,7 @@ EXPORT_SYMBOL_GPL(drm_bus_flags_from_videomode);
  * of_get_drm_display_mode - get a drm_display_mode from devicetree
  * @np: device_node with the timing specification
  * @dmode: will be set to the return value
- * @bus_flags: information about pixelclk and DE polarity
+ * @bus_flags: information about pixelclk, sync and DE polarity
  * @index: index into the list of display timings in devicetree
  *
  * This function is expensive and should only be used, if only one mode is to be
@@ -939,17 +946,68 @@ struct drm_display_mode *drm_mode_duplicate(struct drm_device *dev,
 }
 EXPORT_SYMBOL(drm_mode_duplicate);
 
+static bool drm_mode_match_timings(const struct drm_display_mode *mode1,
+				   const struct drm_display_mode *mode2)
+{
+	return mode1->hdisplay == mode2->hdisplay &&
+		mode1->hsync_start == mode2->hsync_start &&
+		mode1->hsync_end == mode2->hsync_end &&
+		mode1->htotal == mode2->htotal &&
+		mode1->hskew == mode2->hskew &&
+		mode1->vdisplay == mode2->vdisplay &&
+		mode1->vsync_start == mode2->vsync_start &&
+		mode1->vsync_end == mode2->vsync_end &&
+		mode1->vtotal == mode2->vtotal &&
+		mode1->vscan == mode2->vscan;
+}
+
+static bool drm_mode_match_clock(const struct drm_display_mode *mode1,
+				  const struct drm_display_mode *mode2)
+{
+	/*
+	 * do clock check convert to PICOS
+	 * so fb modes get matched the same
+	 */
+	if (mode1->clock && mode2->clock)
+		return KHZ2PICOS(mode1->clock) == KHZ2PICOS(mode2->clock);
+	else
+		return mode1->clock == mode2->clock;
+}
+
+static bool drm_mode_match_flags(const struct drm_display_mode *mode1,
+				 const struct drm_display_mode *mode2)
+{
+	return (mode1->flags & ~DRM_MODE_FLAG_3D_MASK) ==
+		(mode2->flags & ~DRM_MODE_FLAG_3D_MASK);
+}
+
+static bool drm_mode_match_3d_flags(const struct drm_display_mode *mode1,
+				    const struct drm_display_mode *mode2)
+{
+	return (mode1->flags & DRM_MODE_FLAG_3D_MASK) ==
+		(mode2->flags & DRM_MODE_FLAG_3D_MASK);
+}
+
+static bool drm_mode_match_aspect_ratio(const struct drm_display_mode *mode1,
+					const struct drm_display_mode *mode2)
+{
+	return mode1->picture_aspect_ratio == mode2->picture_aspect_ratio;
+}
+
 /**
- * drm_mode_equal - test modes for equality
+ * drm_mode_match - test modes for (partial) equality
  * @mode1: first mode
  * @mode2: second mode
+ * @match_flags: which parts need to match (DRM_MODE_MATCH_*)
  *
  * Check to see if @mode1 and @mode2 are equivalent.
  *
  * Returns:
- * True if the modes are equal, false otherwise.
+ * True if the modes are (partially) equal, false otherwise.
  */
-bool drm_mode_equal(const struct drm_display_mode *mode1, const struct drm_display_mode *mode2)
+bool drm_mode_match(const struct drm_display_mode *mode1,
+		    const struct drm_display_mode *mode2,
+		    unsigned int match_flags)
 {
 	if (!mode1 && !mode2)
 		return true;
@@ -957,15 +1015,49 @@ bool drm_mode_equal(const struct drm_display_mode *mode1, const struct drm_displ
 	if (!mode1 || !mode2)
 		return false;
 
-	/* do clock check convert to PICOS so fb modes get matched
-	 * the same */
-	if (mode1->clock && mode2->clock) {
-		if (KHZ2PICOS(mode1->clock) != KHZ2PICOS(mode2->clock))
-			return false;
-	} else if (mode1->clock != mode2->clock)
+	if (match_flags & DRM_MODE_MATCH_TIMINGS &&
+	    !drm_mode_match_timings(mode1, mode2))
+		return false;
+
+	if (match_flags & DRM_MODE_MATCH_CLOCK &&
+	    !drm_mode_match_clock(mode1, mode2))
+		return false;
+
+	if (match_flags & DRM_MODE_MATCH_FLAGS &&
+	    !drm_mode_match_flags(mode1, mode2))
+		return false;
+
+	if (match_flags & DRM_MODE_MATCH_3D_FLAGS &&
+	    !drm_mode_match_3d_flags(mode1, mode2))
+		return false;
+
+	if (match_flags & DRM_MODE_MATCH_ASPECT_RATIO &&
+	    !drm_mode_match_aspect_ratio(mode1, mode2))
 		return false;
 
-	return drm_mode_equal_no_clocks(mode1, mode2);
+	return true;
+}
+EXPORT_SYMBOL(drm_mode_match);
+
+/**
+ * drm_mode_equal - test modes for equality
+ * @mode1: first mode
+ * @mode2: second mode
+ *
+ * Check to see if @mode1 and @mode2 are equivalent.
+ *
+ * Returns:
+ * True if the modes are equal, false otherwise.
+ */
+bool drm_mode_equal(const struct drm_display_mode *mode1,
+		    const struct drm_display_mode *mode2)
+{
+	return drm_mode_match(mode1, mode2,
+			      DRM_MODE_MATCH_TIMINGS |
+			      DRM_MODE_MATCH_CLOCK |
+			      DRM_MODE_MATCH_FLAGS |
+			      DRM_MODE_MATCH_3D_FLAGS|
+			      DRM_MODE_MATCH_ASPECT_RATIO);
 }
 EXPORT_SYMBOL(drm_mode_equal);
 
@@ -980,13 +1072,13 @@ EXPORT_SYMBOL(drm_mode_equal);
  * Returns:
  * True if the modes are equal, false otherwise.
  */
-bool drm_mode_equal_no_clocks(const struct drm_display_mode *mode1, const struct drm_display_mode *mode2)
+bool drm_mode_equal_no_clocks(const struct drm_display_mode *mode1,
+			      const struct drm_display_mode *mode2)
 {
-	if ((mode1->flags & DRM_MODE_FLAG_3D_MASK) !=
-	    (mode2->flags & DRM_MODE_FLAG_3D_MASK))
-		return false;
-
-	return drm_mode_equal_no_clocks_no_stereo(mode1, mode2);
+	return drm_mode_match(mode1, mode2,
+			      DRM_MODE_MATCH_TIMINGS |
+			      DRM_MODE_MATCH_FLAGS |
+			      DRM_MODE_MATCH_3D_FLAGS);
 }
 EXPORT_SYMBOL(drm_mode_equal_no_clocks);
 
@@ -1004,21 +1096,9 @@ EXPORT_SYMBOL(drm_mode_equal_no_clocks);
 bool drm_mode_equal_no_clocks_no_stereo(const struct drm_display_mode *mode1,
 					const struct drm_display_mode *mode2)
 {
-	if (mode1->hdisplay == mode2->hdisplay &&
-	    mode1->hsync_start == mode2->hsync_start &&
-	    mode1->hsync_end == mode2->hsync_end &&
-	    mode1->htotal == mode2->htotal &&
-	    mode1->hskew == mode2->hskew &&
-	    mode1->vdisplay == mode2->vdisplay &&
-	    mode1->vsync_start == mode2->vsync_start &&
-	    mode1->vsync_end == mode2->vsync_end &&
-	    mode1->vtotal == mode2->vtotal &&
-	    mode1->vscan == mode2->vscan &&
-	    (mode1->flags & ~DRM_MODE_FLAG_3D_MASK) ==
-	     (mode2->flags & ~DRM_MODE_FLAG_3D_MASK))
-		return true;
-
-	return false;
+	return drm_mode_match(mode1, mode2,
+			      DRM_MODE_MATCH_TIMINGS |
+			      DRM_MODE_MATCH_FLAGS);
 }
 EXPORT_SYMBOL(drm_mode_equal_no_clocks_no_stereo);
 
@@ -1184,7 +1264,7 @@ static const char * const drm_mode_status_names[] = {
 
 #undef MODE_STATUS
 
-static const char *drm_get_mode_status_name(enum drm_mode_status status)
+const char *drm_get_mode_status_name(enum drm_mode_status status)
 {
 	int index = status + 3;
 
@@ -1273,7 +1353,7 @@ void drm_mode_sort(struct list_head *mode_list)
 EXPORT_SYMBOL(drm_mode_sort);
 
 /**
- * drm_mode_connector_list_update - update the mode list for the connector
+ * drm_connector_list_update - update the mode list for the connector
  * @connector: the connector to update
  *
  * This moves the modes from the @connector probed_modes list
@@ -1283,7 +1363,7 @@ EXPORT_SYMBOL(drm_mode_sort);
  * This is just a helper functions doesn't validate any modes itself and also
  * doesn't prune any invalid modes. Callers need to do that themselves.
  */
-void drm_mode_connector_list_update(struct drm_connector *connector)
+void drm_connector_list_update(struct drm_connector *connector)
 {
 	struct drm_display_mode *pmode, *pt;
 
@@ -1332,7 +1412,7 @@ void drm_mode_connector_list_update(struct drm_connector *connector)
 		}
 	}
 }
-EXPORT_SYMBOL(drm_mode_connector_list_update);
+EXPORT_SYMBOL(drm_connector_list_update);
 
 /**
  * drm_mode_parse_command_line_for_connector - parse command line modeline for connector
@@ -1575,6 +1655,26 @@ void drm_mode_convert_to_umode(struct drm_mode_modeinfo *out,
 	out->vrefresh = in->vrefresh;
 	out->flags = in->flags;
 	out->type = in->type;
+
+	switch (in->picture_aspect_ratio) {
+	case HDMI_PICTURE_ASPECT_4_3:
+		out->flags |= DRM_MODE_FLAG_PIC_AR_4_3;
+		break;
+	case HDMI_PICTURE_ASPECT_16_9:
+		out->flags |= DRM_MODE_FLAG_PIC_AR_16_9;
+		break;
+	case HDMI_PICTURE_ASPECT_64_27:
+		out->flags |= DRM_MODE_FLAG_PIC_AR_64_27;
+		break;
+	case HDMI_PICTURE_ASPECT_256_135:
+		out->flags |= DRM_MODE_FLAG_PIC_AR_256_135;
+		break;
+	case HDMI_PICTURE_ASPECT_RESERVED:
+	default:
+		out->flags |= DRM_MODE_FLAG_PIC_AR_NONE;
+		break;
+	}
+
 	strncpy(out->name, in->name, DRM_DISPLAY_MODE_LEN);
 	out->name[DRM_DISPLAY_MODE_LEN-1] = 0;
 }
@@ -1621,6 +1721,30 @@ int drm_mode_convert_umode(struct drm_device *dev,
 	strncpy(out->name, in->name, DRM_DISPLAY_MODE_LEN);
 	out->name[DRM_DISPLAY_MODE_LEN-1] = 0;
 
+	/* Clearing picture aspect ratio bits from out flags,
+	 * as the aspect-ratio information is not stored in
+	 * flags for kernel-mode, but in picture_aspect_ratio.
+	 */
+	out->flags &= ~DRM_MODE_FLAG_PIC_AR_MASK;
+
+	switch (in->flags & DRM_MODE_FLAG_PIC_AR_MASK) {
+	case DRM_MODE_FLAG_PIC_AR_4_3:
+		out->picture_aspect_ratio |= HDMI_PICTURE_ASPECT_4_3;
+		break;
+	case DRM_MODE_FLAG_PIC_AR_16_9:
+		out->picture_aspect_ratio |= HDMI_PICTURE_ASPECT_16_9;
+		break;
+	case DRM_MODE_FLAG_PIC_AR_64_27:
+		out->picture_aspect_ratio |= HDMI_PICTURE_ASPECT_64_27;
+		break;
+	case DRM_MODE_FLAG_PIC_AR_256_135:
+		out->picture_aspect_ratio |= HDMI_PICTURE_ASPECT_256_135;
+		break;
+	default:
+		out->picture_aspect_ratio = HDMI_PICTURE_ASPECT_NONE;
+		break;
+	}
+
 	out->status = drm_mode_validate_driver(dev, out);
 	if (out->status != MODE_OK)
 		return -EINVAL;
diff --git a/drivers/gpu/drm/drm_of.c b/drivers/gpu/drm/drm_of.c
index 1fe122461298..2763a5ec845b 100644
--- a/drivers/gpu/drm/drm_of.c
+++ b/drivers/gpu/drm/drm_of.c
@@ -9,21 +9,28 @@
 #include <drm/drm_panel.h>
 #include <drm/drm_of.h>
 
+/**
+ * DOC: overview
+ *
+ * A set of helper functions to aid DRM drivers in parsing standard DT
+ * properties.
+ */
+
 static void drm_release_of(struct device *dev, void *data)
 {
 	of_node_put(data);
 }
 
 /**
- * drm_crtc_port_mask - find the mask of a registered CRTC by port OF node
+ * drm_of_crtc_port_mask - find the mask of a registered CRTC by port OF node
  * @dev: DRM device
  * @port: port OF node
  *
  * Given a port OF node, return the possible mask of the corresponding
  * CRTC within a device's list of CRTCs.  Returns zero if not found.
  */
-static uint32_t drm_crtc_port_mask(struct drm_device *dev,
-				   struct device_node *port)
+uint32_t drm_of_crtc_port_mask(struct drm_device *dev,
+			    struct device_node *port)
 {
 	unsigned int index = 0;
 	struct drm_crtc *tmp;
@@ -37,6 +44,7 @@ static uint32_t drm_crtc_port_mask(struct drm_device *dev,
 
 	return 0;
 }
+EXPORT_SYMBOL(drm_of_crtc_port_mask);
 
 /**
  * drm_of_find_possible_crtcs - find the possible CRTCs for an encoder port
@@ -62,7 +70,7 @@ uint32_t drm_of_find_possible_crtcs(struct drm_device *dev,
 			return 0;
 		}
 
-		possible_crtcs |= drm_crtc_port_mask(dev, remote_port);
+		possible_crtcs |= drm_of_crtc_port_mask(dev, remote_port);
 
 		of_node_put(remote_port);
 	}
@@ -93,7 +101,7 @@ EXPORT_SYMBOL_GPL(drm_of_component_match_add);
  * drm_of_component_probe - Generic probe function for a component based master
  * @dev: master device containing the OF node
  * @compare_of: compare function used for matching components
- * @master_ops: component master ops to be used
+ * @m_ops: component master ops to be used
  *
  * Parse the platform device OF node and bind all the components associated
  * with the master. Interface ports are added before the encoders in order to
@@ -238,10 +246,17 @@ int drm_of_find_panel_or_bridge(const struct device_node *np,
 	if (!remote)
 		return -ENODEV;
 
+	if (!of_device_is_available(remote)) {
+		of_node_put(remote);
+		return -ENODEV;
+	}
+
 	if (panel) {
 		*panel = of_drm_find_panel(remote);
-		if (*panel)
+		if (!IS_ERR(*panel))
 			ret = 0;
+		else
+			*panel = NULL;
 	}
 
 	/* No panel found yet, check for a bridge next. */
diff --git a/drivers/gpu/drm/drm_panel.c b/drivers/gpu/drm/drm_panel.c
index 308d442a531b..b902361dee6e 100644
--- a/drivers/gpu/drm/drm_panel.c
+++ b/drivers/gpu/drm/drm_panel.c
@@ -24,6 +24,7 @@
 #include <linux/err.h>
 #include <linux/module.h>
 
+#include <drm/drm_device.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_panel.h>
 
@@ -94,6 +95,9 @@ EXPORT_SYMBOL(drm_panel_remove);
  *
  * An error is returned if the panel is already attached to another connector.
  *
+ * When unloading, the driver should detach from the panel by calling
+ * drm_panel_detach().
+ *
  * Return: 0 on success or a negative error code on failure.
  */
 int drm_panel_attach(struct drm_panel *panel, struct drm_connector *connector)
@@ -101,6 +105,13 @@ int drm_panel_attach(struct drm_panel *panel, struct drm_connector *connector)
 	if (panel->connector)
 		return -EBUSY;
 
+	panel->link = device_link_add(connector->dev->dev, panel->dev, 0);
+	if (!panel->link) {
+		dev_err(panel->dev, "failed to link panel to %s\n",
+			dev_name(connector->dev->dev));
+		return -EINVAL;
+	}
+
 	panel->connector = connector;
 	panel->drm = connector->dev;
 
@@ -115,10 +126,15 @@ EXPORT_SYMBOL(drm_panel_attach);
  * Detaches a panel from the connector it is attached to. If a panel is not
  * attached to any connector this is effectively a no-op.
  *
+ * This function should not be called by the panel device itself. It
+ * is only for the drm device that called drm_panel_attach().
+ *
  * Return: 0 on success or a negative error code on failure.
  */
 int drm_panel_detach(struct drm_panel *panel)
 {
+	device_link_del(panel->link);
+
 	panel->connector = NULL;
 	panel->drm = NULL;
 
@@ -135,12 +151,19 @@ EXPORT_SYMBOL(drm_panel_detach);
  * tree node. If a matching panel is found, return a pointer to it.
  *
  * Return: A pointer to the panel registered for the specified device tree
- * node or NULL if no panel matching the device tree node can be found.
+ * node or an ERR_PTR() if no panel matching the device tree node can be found.
+ * Possible error codes returned by this function:
+ * - EPROBE_DEFER: the panel device has not been probed yet, and the caller
+ *   should retry later
+ * - ENODEV: the device is not available (status != "okay" or "ok")
  */
 struct drm_panel *of_drm_find_panel(const struct device_node *np)
 {
 	struct drm_panel *panel;
 
+	if (!of_device_is_available(np))
+		return ERR_PTR(-ENODEV);
+
 	mutex_lock(&panel_lock);
 
 	list_for_each_entry(panel, &panel_list, list) {
@@ -151,7 +174,7 @@ struct drm_panel *of_drm_find_panel(const struct device_node *np)
 	}
 
 	mutex_unlock(&panel_lock);
-	return NULL;
+	return ERR_PTR(-EPROBE_DEFER);
 }
 EXPORT_SYMBOL(of_drm_find_panel);
 #endif
diff --git a/drivers/gpu/drm/drm_panel_orientation_quirks.c b/drivers/gpu/drm/drm_panel_orientation_quirks.c
index caebddda8bce..fe9c6c731e87 100644
--- a/drivers/gpu/drm/drm_panel_orientation_quirks.c
+++ b/drivers/gpu/drm/drm_panel_orientation_quirks.c
@@ -172,10 +172,9 @@ int drm_get_panel_orientation_quirk(int width, int height)
 		if (!bios_date)
 			continue;
 
-		for (i = 0; data->bios_dates[i]; i++) {
-			if (!strcmp(data->bios_dates[i], bios_date))
-				return data->orientation;
-		}
+		i = match_string(data->bios_dates, -1, bios_date);
+		if (i >= 0)
+			return data->orientation;
 	}
 
 	return DRM_MODE_PANEL_ORIENTATION_UNKNOWN;
diff --git a/drivers/gpu/drm/drm_pci.c b/drivers/gpu/drm/drm_pci.c
index 4db9c515b74f..896e42a34895 100644
--- a/drivers/gpu/drm/drm_pci.c
+++ b/drivers/gpu/drm/drm_pci.c
@@ -326,64 +326,6 @@ int drm_legacy_pci_init(struct drm_driver *driver, struct pci_driver *pdriver)
 }
 EXPORT_SYMBOL(drm_legacy_pci_init);
 
-int drm_pcie_get_speed_cap_mask(struct drm_device *dev, u32 *mask)
-{
-	struct pci_dev *root;
-	u32 lnkcap, lnkcap2;
-
-	*mask = 0;
-	if (!dev->pdev)
-		return -EINVAL;
-
-	root = dev->pdev->bus->self;
-
-	/* we've been informed via and serverworks don't make the cut */
-	if (root->vendor == PCI_VENDOR_ID_VIA ||
-	    root->vendor == PCI_VENDOR_ID_SERVERWORKS)
-		return -EINVAL;
-
-	pcie_capability_read_dword(root, PCI_EXP_LNKCAP, &lnkcap);
-	pcie_capability_read_dword(root, PCI_EXP_LNKCAP2, &lnkcap2);
-
-	if (lnkcap2) {	/* PCIe r3.0-compliant */
-		if (lnkcap2 & PCI_EXP_LNKCAP2_SLS_2_5GB)
-			*mask |= DRM_PCIE_SPEED_25;
-		if (lnkcap2 & PCI_EXP_LNKCAP2_SLS_5_0GB)
-			*mask |= DRM_PCIE_SPEED_50;
-		if (lnkcap2 & PCI_EXP_LNKCAP2_SLS_8_0GB)
-			*mask |= DRM_PCIE_SPEED_80;
-	} else {	/* pre-r3.0 */
-		if (lnkcap & PCI_EXP_LNKCAP_SLS_2_5GB)
-			*mask |= DRM_PCIE_SPEED_25;
-		if (lnkcap & PCI_EXP_LNKCAP_SLS_5_0GB)
-			*mask |= (DRM_PCIE_SPEED_25 | DRM_PCIE_SPEED_50);
-	}
-
-	DRM_INFO("probing gen 2 caps for device %x:%x = %x/%x\n", root->vendor, root->device, lnkcap, lnkcap2);
-	return 0;
-}
-EXPORT_SYMBOL(drm_pcie_get_speed_cap_mask);
-
-int drm_pcie_get_max_link_width(struct drm_device *dev, u32 *mlw)
-{
-	struct pci_dev *root;
-	u32 lnkcap;
-
-	*mlw = 0;
-	if (!dev->pdev)
-		return -EINVAL;
-
-	root = dev->pdev->bus->self;
-
-	pcie_capability_read_dword(root, PCI_EXP_LNKCAP, &lnkcap);
-
-	*mlw = (lnkcap & PCI_EXP_LNKCAP_MLW) >> 4;
-
-	DRM_INFO("probing mlw for device %x:%x = %x\n", root->vendor, root->device, lnkcap);
-	return 0;
-}
-EXPORT_SYMBOL(drm_pcie_get_max_link_width);
-
 #else
 
 void drm_pci_agp_destroy(struct drm_device *dev) {}
diff --git a/drivers/gpu/drm/drm_plane.c b/drivers/gpu/drm/drm_plane.c
index 035054455301..6153cbda239f 100644
--- a/drivers/gpu/drm/drm_plane.c
+++ b/drivers/gpu/drm/drm_plane.c
@@ -177,6 +177,10 @@ int drm_universal_plane_init(struct drm_device *dev, struct drm_plane *plane,
 	if (WARN_ON(config->num_total_plane >= 32))
 		return -EINVAL;
 
+	WARN_ON(drm_drv_uses_atomic_modeset(dev) &&
+		(!funcs->atomic_destroy_state ||
+		 !funcs->atomic_duplicate_state));
+
 	ret = drm_mode_object_add(dev, &plane->base, DRM_MODE_OBJECT_PLANE);
 	if (ret)
 		return ret;
@@ -561,19 +565,66 @@ int drm_plane_check_pixel_format(struct drm_plane *plane,
 	if (i == plane->format_count)
 		return -EINVAL;
 
-	if (!plane->modifier_count)
-		return 0;
+	if (plane->funcs->format_mod_supported) {
+		if (!plane->funcs->format_mod_supported(plane, format, modifier))
+			return -EINVAL;
+	} else {
+		if (!plane->modifier_count)
+			return 0;
 
-	for (i = 0; i < plane->modifier_count; i++) {
-		if (modifier == plane->modifiers[i])
-			break;
+		for (i = 0; i < plane->modifier_count; i++) {
+			if (modifier == plane->modifiers[i])
+				break;
+		}
+		if (i == plane->modifier_count)
+			return -EINVAL;
 	}
-	if (i == plane->modifier_count)
-		return -EINVAL;
 
-	if (plane->funcs->format_mod_supported &&
-	    !plane->funcs->format_mod_supported(plane, format, modifier))
+	return 0;
+}
+
+static int __setplane_check(struct drm_plane *plane,
+			    struct drm_crtc *crtc,
+			    struct drm_framebuffer *fb,
+			    int32_t crtc_x, int32_t crtc_y,
+			    uint32_t crtc_w, uint32_t crtc_h,
+			    uint32_t src_x, uint32_t src_y,
+			    uint32_t src_w, uint32_t src_h)
+{
+	int ret;
+
+	/* Check whether this plane is usable on this CRTC */
+	if (!(plane->possible_crtcs & drm_crtc_mask(crtc))) {
+		DRM_DEBUG_KMS("Invalid crtc for plane\n");
 		return -EINVAL;
+	}
+
+	/* Check whether this plane supports the fb pixel format. */
+	ret = drm_plane_check_pixel_format(plane, fb->format->format,
+					   fb->modifier);
+	if (ret) {
+		struct drm_format_name_buf format_name;
+
+		DRM_DEBUG_KMS("Invalid pixel format %s, modifier 0x%llx\n",
+			      drm_get_format_name(fb->format->format,
+						  &format_name),
+			      fb->modifier);
+		return ret;
+	}
+
+	/* Give drivers some help against integer overflows */
+	if (crtc_w > INT_MAX ||
+	    crtc_x > INT_MAX - (int32_t) crtc_w ||
+	    crtc_h > INT_MAX ||
+	    crtc_y > INT_MAX - (int32_t) crtc_h) {
+		DRM_DEBUG_KMS("Invalid CRTC coordinates %ux%u+%d+%d\n",
+			      crtc_w, crtc_h, crtc_x, crtc_y);
+		return -ERANGE;
+	}
+
+	ret = drm_framebuffer_check_src_coords(src_x, src_y, src_w, src_h, fb);
+	if (ret)
+		return ret;
 
 	return 0;
 }
@@ -598,6 +649,8 @@ static int __setplane_internal(struct drm_plane *plane,
 {
 	int ret = 0;
 
+	WARN_ON(drm_drv_uses_atomic_modeset(plane->dev));
+
 	/* No fb means shut it down */
 	if (!fb) {
 		plane->old_fb = plane->fb;
@@ -611,37 +664,9 @@ static int __setplane_internal(struct drm_plane *plane,
 		goto out;
 	}
 
-	/* Check whether this plane is usable on this CRTC */
-	if (!(plane->possible_crtcs & drm_crtc_mask(crtc))) {
-		DRM_DEBUG_KMS("Invalid crtc for plane\n");
-		ret = -EINVAL;
-		goto out;
-	}
-
-	/* Check whether this plane supports the fb pixel format. */
-	ret = drm_plane_check_pixel_format(plane, fb->format->format,
-					   fb->modifier);
-	if (ret) {
-		struct drm_format_name_buf format_name;
-		DRM_DEBUG_KMS("Invalid pixel format %s, modifier 0x%llx\n",
-			      drm_get_format_name(fb->format->format,
-						  &format_name),
-			      fb->modifier);
-		goto out;
-	}
-
-	/* Give drivers some help against integer overflows */
-	if (crtc_w > INT_MAX ||
-	    crtc_x > INT_MAX - (int32_t) crtc_w ||
-	    crtc_h > INT_MAX ||
-	    crtc_y > INT_MAX - (int32_t) crtc_h) {
-		DRM_DEBUG_KMS("Invalid CRTC coordinates %ux%u+%d+%d\n",
-			      crtc_w, crtc_h, crtc_x, crtc_y);
-		ret = -ERANGE;
-		goto out;
-	}
-
-	ret = drm_framebuffer_check_src_coords(src_x, src_y, src_w, src_h, fb);
+	ret = __setplane_check(plane, crtc, fb,
+			       crtc_x, crtc_y, crtc_w, crtc_h,
+			       src_x, src_y, src_w, src_h);
 	if (ret)
 		goto out;
 
@@ -665,6 +690,41 @@ out:
 	return ret;
 }
 
+static int __setplane_atomic(struct drm_plane *plane,
+			     struct drm_crtc *crtc,
+			     struct drm_framebuffer *fb,
+			     int32_t crtc_x, int32_t crtc_y,
+			     uint32_t crtc_w, uint32_t crtc_h,
+			     uint32_t src_x, uint32_t src_y,
+			     uint32_t src_w, uint32_t src_h,
+			     struct drm_modeset_acquire_ctx *ctx)
+{
+	int ret;
+
+	WARN_ON(!drm_drv_uses_atomic_modeset(plane->dev));
+
+	/* No fb means shut it down */
+	if (!fb)
+		return plane->funcs->disable_plane(plane, ctx);
+
+	/*
+	 * FIXME: This is redundant with drm_atomic_plane_check(),
+	 * but the legacy cursor/"async" .update_plane() tricks
+	 * don't call that so we still need this here. Should remove
+	 * this when all .update_plane() implementations have been
+	 * fixed to call drm_atomic_plane_check().
+	 */
+	ret = __setplane_check(plane, crtc, fb,
+			       crtc_x, crtc_y, crtc_w, crtc_h,
+			       src_x, src_y, src_w, src_h);
+	if (ret)
+		return ret;
+
+	return plane->funcs->update_plane(plane, crtc, fb,
+					  crtc_x, crtc_y, crtc_w, crtc_h,
+					  src_x, src_y, src_w, src_h, ctx);
+}
+
 static int setplane_internal(struct drm_plane *plane,
 			     struct drm_crtc *crtc,
 			     struct drm_framebuffer *fb,
@@ -682,9 +742,15 @@ retry:
 	ret = drm_modeset_lock_all_ctx(plane->dev, &ctx);
 	if (ret)
 		goto fail;
-	ret = __setplane_internal(plane, crtc, fb,
-				  crtc_x, crtc_y, crtc_w, crtc_h,
-				  src_x, src_y, src_w, src_h, &ctx);
+
+	if (drm_drv_uses_atomic_modeset(plane->dev))
+		ret = __setplane_atomic(plane, crtc, fb,
+					crtc_x, crtc_y, crtc_w, crtc_h,
+					src_x, src_y, src_w, src_h, &ctx);
+	else
+		ret = __setplane_internal(plane, crtc, fb,
+					  crtc_x, crtc_y, crtc_w, crtc_h,
+					  src_x, src_y, src_w, src_h, &ctx);
 
 fail:
 	if (ret == -EDEADLK) {
@@ -816,9 +882,14 @@ static int drm_mode_cursor_universal(struct drm_crtc *crtc,
 		src_h = fb->height << 16;
 	}
 
-	ret = __setplane_internal(plane, crtc, fb,
-				  crtc_x, crtc_y, crtc_w, crtc_h,
-				  0, 0, src_w, src_h, ctx);
+	if (drm_drv_uses_atomic_modeset(dev))
+		ret = __setplane_atomic(plane, crtc, fb,
+					crtc_x, crtc_y, crtc_w, crtc_h,
+					0, 0, src_w, src_h, ctx);
+	else
+		ret = __setplane_internal(plane, crtc, fb,
+					  crtc_x, crtc_y, crtc_w, crtc_h,
+					  0, 0, src_w, src_h, ctx);
 
 	if (fb)
 		drm_framebuffer_put(fb);
@@ -1092,8 +1163,10 @@ retry:
 		/* Keep the old fb, don't unref it. */
 		plane->old_fb = NULL;
 	} else {
-		plane->fb = fb;
-		drm_framebuffer_get(fb);
+		if (!plane->state) {
+			plane->fb = fb;
+			drm_framebuffer_get(fb);
+		}
 	}
 
 out:
diff --git a/drivers/gpu/drm/drm_plane_helper.c b/drivers/gpu/drm/drm_plane_helper.c
index f88f68161519..621f17643bb0 100644
--- a/drivers/gpu/drm/drm_plane_helper.c
+++ b/drivers/gpu/drm/drm_plane_helper.c
@@ -440,6 +440,7 @@ out:
  * @src_y: y offset of @fb for panning
  * @src_w: width of source rectangle in @fb
  * @src_h: height of source rectangle in @fb
+ * @ctx: lock acquire context, not used here
  *
  * Provides a default plane update handler using the atomic plane update
  * functions. It is fully left to the driver to check plane constraints and
@@ -455,7 +456,8 @@ int drm_plane_helper_update(struct drm_plane *plane, struct drm_crtc *crtc,
 			    int crtc_x, int crtc_y,
 			    unsigned int crtc_w, unsigned int crtc_h,
 			    uint32_t src_x, uint32_t src_y,
-			    uint32_t src_w, uint32_t src_h)
+			    uint32_t src_w, uint32_t src_h,
+			    struct drm_modeset_acquire_ctx *ctx)
 {
 	struct drm_plane_state *plane_state;
 
@@ -489,6 +491,7 @@ EXPORT_SYMBOL(drm_plane_helper_update);
 /**
  * drm_plane_helper_disable() - Transitional helper for plane disable
  * @plane: plane to disable
+ * @ctx: lock acquire context, not used here
  *
  * Provides a default plane disable handler using the atomic plane update
  * functions. It is fully left to the driver to check plane constraints and
@@ -499,9 +502,11 @@ EXPORT_SYMBOL(drm_plane_helper_update);
  * RETURNS:
  * Zero on success, error code on failure
  */
-int drm_plane_helper_disable(struct drm_plane *plane)
+int drm_plane_helper_disable(struct drm_plane *plane,
+			     struct drm_modeset_acquire_ctx *ctx)
 {
 	struct drm_plane_state *plane_state;
+	struct drm_framebuffer *old_fb;
 
 	/* crtc helpers love to call disable functions for already disabled hw
 	 * functions. So cope with that. */
@@ -521,8 +526,9 @@ int drm_plane_helper_disable(struct drm_plane *plane)
 	plane_state->plane = plane;
 
 	plane_state->crtc = NULL;
+	old_fb = plane_state->fb;
 	drm_atomic_set_fb_for_plane(plane_state, NULL);
 
-	return drm_plane_helper_commit(plane, plane_state, plane->fb);
+	return drm_plane_helper_commit(plane, plane_state, old_fb);
 }
 EXPORT_SYMBOL(drm_plane_helper_disable);
diff --git a/drivers/gpu/drm/drm_prime.c b/drivers/gpu/drm/drm_prime.c
index caf675e3e692..186db2e4c57a 100644
--- a/drivers/gpu/drm/drm_prime.c
+++ b/drivers/gpu/drm/drm_prime.c
@@ -186,7 +186,6 @@ static int drm_prime_lookup_buf_handle(struct drm_prime_file_private *prime_fpri
 /**
  * drm_gem_map_attach - dma_buf attach implementation for GEM
  * @dma_buf: buffer to attach device to
- * @target_dev: not used
  * @attach: buffer attachment data
  *
  * Allocates &drm_prime_attachment and calls &drm_driver.gem_prime_pin for
@@ -195,7 +194,7 @@ static int drm_prime_lookup_buf_handle(struct drm_prime_file_private *prime_fpri
  *
  * Returns 0 on success, negative error code on failure.
  */
-int drm_gem_map_attach(struct dma_buf *dma_buf, struct device *target_dev,
+int drm_gem_map_attach(struct dma_buf *dma_buf,
 		       struct dma_buf_attachment *attach)
 {
 	struct drm_prime_attachment *prime_attach;
@@ -409,7 +408,10 @@ void *drm_gem_dmabuf_vmap(struct dma_buf *dma_buf)
 	struct drm_gem_object *obj = dma_buf->priv;
 	struct drm_device *dev = obj->dev;
 
-	return dev->driver->gem_prime_vmap(obj);
+	if (dev->driver->gem_prime_vmap)
+		return dev->driver->gem_prime_vmap(obj);
+	else
+		return NULL;
 }
 EXPORT_SYMBOL(drm_gem_dmabuf_vmap);
 
@@ -426,40 +428,12 @@ void drm_gem_dmabuf_vunmap(struct dma_buf *dma_buf, void *vaddr)
 	struct drm_gem_object *obj = dma_buf->priv;
 	struct drm_device *dev = obj->dev;
 
-	dev->driver->gem_prime_vunmap(obj, vaddr);
+	if (dev->driver->gem_prime_vunmap)
+		dev->driver->gem_prime_vunmap(obj, vaddr);
 }
 EXPORT_SYMBOL(drm_gem_dmabuf_vunmap);
 
 /**
- * drm_gem_dmabuf_kmap_atomic - map_atomic implementation for GEM
- * @dma_buf: buffer to be mapped
- * @page_num: page number within the buffer
- *
- * Not implemented. This can be used as the &dma_buf_ops.map_atomic callback.
- */
-void *drm_gem_dmabuf_kmap_atomic(struct dma_buf *dma_buf,
-				 unsigned long page_num)
-{
-	return NULL;
-}
-EXPORT_SYMBOL(drm_gem_dmabuf_kmap_atomic);
-
-/**
- * drm_gem_dmabuf_kunmap_atomic - unmap_atomic implementation for GEM
- * @dma_buf: buffer to be unmapped
- * @page_num: page number within the buffer
- * @addr: virtual address of the buffer
- *
- * Not implemented. This can be used as the &dma_buf_ops.unmap_atomic callback.
- */
-void drm_gem_dmabuf_kunmap_atomic(struct dma_buf *dma_buf,
-				  unsigned long page_num, void *addr)
-{
-
-}
-EXPORT_SYMBOL(drm_gem_dmabuf_kunmap_atomic);
-
-/**
  * drm_gem_dmabuf_kmap - map implementation for GEM
  * @dma_buf: buffer to be mapped
  * @page_num: page number within the buffer
@@ -516,9 +490,7 @@ static const struct dma_buf_ops drm_gem_prime_dmabuf_ops =  {
 	.unmap_dma_buf = drm_gem_unmap_dma_buf,
 	.release = drm_gem_dmabuf_release,
 	.map = drm_gem_dmabuf_kmap,
-	.map_atomic = drm_gem_dmabuf_kmap_atomic,
 	.unmap = drm_gem_dmabuf_kunmap,
-	.unmap_atomic = drm_gem_dmabuf_kunmap_atomic,
 	.mmap = drm_gem_dmabuf_mmap,
 	.vmap = drm_gem_dmabuf_vmap,
 	.vunmap = drm_gem_dmabuf_vunmap,
diff --git a/drivers/gpu/drm/drm_probe_helper.c b/drivers/gpu/drm/drm_probe_helper.c
index 527743394150..a1bb157bfdfa 100644
--- a/drivers/gpu/drm/drm_probe_helper.c
+++ b/drivers/gpu/drm/drm_probe_helper.c
@@ -33,6 +33,7 @@
 #include <linux/moduleparam.h>
 
 #include <drm/drmP.h>
+#include <drm/drm_client.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_fourcc.h>
 #include <drm/drm_crtc_helper.h>
@@ -88,9 +89,9 @@ drm_mode_validate_pipeline(struct drm_display_mode *mode,
 			    struct drm_connector *connector)
 {
 	struct drm_device *dev = connector->dev;
-	uint32_t *ids = connector->encoder_ids;
 	enum drm_mode_status ret = MODE_OK;
-	unsigned int i;
+	struct drm_encoder *encoder;
+	int i;
 
 	/* Step 1: Validate against connector */
 	ret = drm_connector_mode_valid(connector, mode);
@@ -98,13 +99,9 @@ drm_mode_validate_pipeline(struct drm_display_mode *mode,
 		return ret;
 
 	/* Step 2: Validate against encoders and crtcs */
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		struct drm_encoder *encoder = drm_encoder_find(dev, NULL, ids[i]);
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		struct drm_crtc *crtc;
 
-		if (!encoder)
-			continue;
-
 		ret = drm_encoder_mode_valid(encoder, mode);
 		if (ret != MODE_OK) {
 			/* No point in continuing for crtc check as this encoder
@@ -363,7 +360,7 @@ EXPORT_SYMBOL(drm_helper_probe_detect);
  *    using the VESA GTF/CVT formulas.
  *
  * 3. Modes are moved from the probed_modes list to the modes list. Potential
- *    duplicates are merged together (see drm_mode_connector_list_update()).
+ *    duplicates are merged together (see drm_connector_list_update()).
  *    After this step the probed_modes list will be empty again.
  *
  * 4. Any non-stale mode on the modes list then undergoes validation
@@ -475,7 +472,7 @@ retry:
 	if (connector->status == connector_status_disconnected) {
 		DRM_DEBUG_KMS("[CONNECTOR:%d:%s] disconnected\n",
 			connector->base.id, connector->name);
-		drm_mode_connector_update_edid_property(connector, NULL);
+		drm_connector_update_edid_property(connector, NULL);
 		verbose_prune = false;
 		goto prune;
 	}
@@ -488,7 +485,7 @@ retry:
 	if (count == 0)
 		goto prune;
 
-	drm_mode_connector_list_update(connector);
+	drm_connector_list_update(connector);
 
 	if (connector->interlace_allowed)
 		mode_flags |= DRM_MODE_FLAG_INTERLACE;
@@ -563,6 +560,8 @@ void drm_kms_helper_hotplug_event(struct drm_device *dev)
 	drm_sysfs_hotplug_event(dev);
 	if (dev->mode_config.funcs->output_poll_changed)
 		dev->mode_config.funcs->output_poll_changed(dev);
+
+	drm_client_dev_hotplug(dev);
 }
 EXPORT_SYMBOL(drm_kms_helper_hotplug_event);
 
diff --git a/drivers/gpu/drm/drm_property.c b/drivers/gpu/drm/drm_property.c
index 8f4672daac7f..1f8031e30f53 100644
--- a/drivers/gpu/drm/drm_property.c
+++ b/drivers/gpu/drm/drm_property.c
@@ -169,9 +169,9 @@ struct drm_property *drm_property_create_enum(struct drm_device *dev,
 		return NULL;
 
 	for (i = 0; i < num_values; i++) {
-		ret = drm_property_add_enum(property, i,
-				      props[i].type,
-				      props[i].name);
+		ret = drm_property_add_enum(property,
+					    props[i].type,
+					    props[i].name);
 		if (ret) {
 			drm_property_destroy(dev, property);
 			return NULL;
@@ -209,7 +209,7 @@ struct drm_property *drm_property_create_bitmask(struct drm_device *dev,
 						 uint64_t supported_bits)
 {
 	struct drm_property *property;
-	int i, ret, index = 0;
+	int i, ret;
 	int num_values = hweight64(supported_bits);
 
 	flags |= DRM_MODE_PROP_BITMASK;
@@ -221,14 +221,9 @@ struct drm_property *drm_property_create_bitmask(struct drm_device *dev,
 		if (!(supported_bits & (1ULL << props[i].type)))
 			continue;
 
-		if (WARN_ON(index >= num_values)) {
-			drm_property_destroy(dev, property);
-			return NULL;
-		}
-
-		ret = drm_property_add_enum(property, index++,
-				      props[i].type,
-				      props[i].name);
+		ret = drm_property_add_enum(property,
+					    props[i].type,
+					    props[i].name);
 		if (ret) {
 			drm_property_destroy(dev, property);
 			return NULL;
@@ -376,7 +371,6 @@ EXPORT_SYMBOL(drm_property_create_bool);
 /**
  * drm_property_add_enum - add a possible value to an enumeration property
  * @property: enumeration property to change
- * @index: index of the new enumeration
  * @value: value of the new enumeration
  * @name: symbolic name of the new enumeration
  *
@@ -388,10 +382,11 @@ EXPORT_SYMBOL(drm_property_create_bool);
  * Returns:
  * Zero on success, error code on failure.
  */
-int drm_property_add_enum(struct drm_property *property, int index,
+int drm_property_add_enum(struct drm_property *property,
 			  uint64_t value, const char *name)
 {
 	struct drm_property_enum *prop_enum;
+	int index = 0;
 
 	if (WARN_ON(strlen(name) >= DRM_PROP_NAME_LEN))
 		return -EINVAL;
@@ -411,8 +406,12 @@ int drm_property_add_enum(struct drm_property *property, int index,
 	list_for_each_entry(prop_enum, &property->enum_list, head) {
 		if (WARN_ON(prop_enum->value == value))
 			return -EINVAL;
+		index++;
 	}
 
+	if (WARN_ON(index >= property->num_values))
+		return -EINVAL;
+
 	prop_enum = kzalloc(sizeof(struct drm_property_enum), GFP_KERNEL);
 	if (!prop_enum)
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/drm_rect.c b/drivers/gpu/drm/drm_rect.c
index 9817c1445ba9..8c057829b804 100644
--- a/drivers/gpu/drm/drm_rect.c
+++ b/drivers/gpu/drm/drm_rect.c
@@ -50,13 +50,25 @@ bool drm_rect_intersect(struct drm_rect *r1, const struct drm_rect *r2)
 }
 EXPORT_SYMBOL(drm_rect_intersect);
 
+static u32 clip_scaled(u32 src, u32 dst, u32 clip)
+{
+	u64 tmp = mul_u32_u32(src, dst - clip);
+
+	/*
+	 * Round toward 1.0 when clipping so that we don't accidentally
+	 * change upscaling to downscaling or vice versa.
+	 */
+	if (src < (dst << 16))
+		return DIV_ROUND_UP_ULL(tmp, dst);
+	else
+		return DIV_ROUND_DOWN_ULL(tmp, dst);
+}
+
 /**
  * drm_rect_clip_scaled - perform a scaled clip operation
  * @src: source window rectangle
  * @dst: destination window rectangle
  * @clip: clip rectangle
- * @hscale: horizontal scaling factor
- * @vscale: vertical scaling factor
  *
  * Clip rectangle @dst by rectangle @clip. Clip rectangle @src by the
  * same amounts multiplied by @hscale and @vscale.
@@ -66,33 +78,44 @@ EXPORT_SYMBOL(drm_rect_intersect);
  * %false otherwise
  */
 bool drm_rect_clip_scaled(struct drm_rect *src, struct drm_rect *dst,
-			  const struct drm_rect *clip,
-			  int hscale, int vscale)
+			  const struct drm_rect *clip)
 {
 	int diff;
 
 	diff = clip->x1 - dst->x1;
 	if (diff > 0) {
-		int64_t tmp = src->x1 + (int64_t) diff * hscale;
-		src->x1 = clamp_t(int64_t, tmp, INT_MIN, INT_MAX);
+		u32 new_src_w = clip_scaled(drm_rect_width(src),
+					    drm_rect_width(dst), diff);
+
+		src->x1 = clamp_t(int64_t, src->x2 - new_src_w, INT_MIN, INT_MAX);
+		dst->x1 = clip->x1;
 	}
 	diff = clip->y1 - dst->y1;
 	if (diff > 0) {
-		int64_t tmp = src->y1 + (int64_t) diff * vscale;
-		src->y1 = clamp_t(int64_t, tmp, INT_MIN, INT_MAX);
+		u32 new_src_h = clip_scaled(drm_rect_height(src),
+					    drm_rect_height(dst), diff);
+
+		src->y1 = clamp_t(int64_t, src->y2 - new_src_h, INT_MIN, INT_MAX);
+		dst->y1 = clip->y1;
 	}
 	diff = dst->x2 - clip->x2;
 	if (diff > 0) {
-		int64_t tmp = src->x2 - (int64_t) diff * hscale;
-		src->x2 = clamp_t(int64_t, tmp, INT_MIN, INT_MAX);
+		u32 new_src_w = clip_scaled(drm_rect_width(src),
+					    drm_rect_width(dst), diff);
+
+		src->x2 = clamp_t(int64_t, src->x1 + new_src_w, INT_MIN, INT_MAX);
+		dst->x2 = clip->x2;
 	}
 	diff = dst->y2 - clip->y2;
 	if (diff > 0) {
-		int64_t tmp = src->y2 - (int64_t) diff * vscale;
-		src->y2 = clamp_t(int64_t, tmp, INT_MIN, INT_MAX);
+		u32 new_src_h = clip_scaled(drm_rect_height(src),
+					    drm_rect_height(dst), diff);
+
+		src->y2 = clamp_t(int64_t, src->y1 + new_src_h, INT_MIN, INT_MAX);
+		dst->y2 = clip->y2;
 	}
 
-	return drm_rect_intersect(dst, clip);
+	return drm_rect_visible(dst);
 }
 EXPORT_SYMBOL(drm_rect_clip_scaled);
 
@@ -106,7 +129,10 @@ static int drm_calc_scale(int src, int dst)
 	if (dst == 0)
 		return 0;
 
-	scale = src / dst;
+	if (src > (dst << 16))
+		return DIV_ROUND_UP(src, dst);
+	else
+		scale = src / dst;
 
 	return scale;
 }
@@ -121,6 +147,10 @@ static int drm_calc_scale(int src, int dst)
  * Calculate the horizontal scaling factor as
  * (@src width) / (@dst width).
  *
+ * If the scale is below 1 << 16, round down. If the scale is above
+ * 1 << 16, round up. This will calculate the scale with the most
+ * pessimistic limit calculation.
+ *
  * RETURNS:
  * The horizontal scaling factor, or errno of out of limits.
  */
@@ -152,6 +182,10 @@ EXPORT_SYMBOL(drm_rect_calc_hscale);
  * Calculate the vertical scaling factor as
  * (@src height) / (@dst height).
  *
+ * If the scale is below 1 << 16, round down. If the scale is above
+ * 1 << 16, round up. This will calculate the scale with the most
+ * pessimistic limit calculation.
+ *
  * RETURNS:
  * The vertical scaling factor, or errno of out of limits.
  */
@@ -189,6 +223,10 @@ EXPORT_SYMBOL(drm_rect_calc_vscale);
  * If the calculated scaling factor is above @max_vscale,
  * decrease the height of rectangle @src to compensate.
  *
+ * If the scale is below 1 << 16, round down. If the scale is above
+ * 1 << 16, round up. This will calculate the scale with the most
+ * pessimistic limit calculation.
+ *
  * RETURNS:
  * The horizontal scaling factor.
  */
@@ -239,6 +277,10 @@ EXPORT_SYMBOL(drm_rect_calc_hscale_relaxed);
  * If the calculated scaling factor is above @max_vscale,
  * decrease the height of rectangle @src to compensate.
  *
+ * If the scale is below 1 << 16, round down. If the scale is above
+ * 1 << 16, round up. This will calculate the scale with the most
+ * pessimistic limit calculation.
+ *
  * RETURNS:
  * The vertical scaling factor.
  */
@@ -373,8 +415,8 @@ EXPORT_SYMBOL(drm_rect_rotate);
  * them when doing a rotatation and its inverse.
  * That is, if you do ::
  *
- *     DRM_MODE_PROP_ROTATE(&r, width, height, rotation);
- *     DRM_MODE_ROTATE_inv(&r, width, height, rotation);
+ *     drm_rect_rotate(&r, width, height, rotation);
+ *     drm_rect_rotate_inv(&r, width, height, rotation);
  *
  * you will always get back the original rectangle.
  */
diff --git a/drivers/gpu/drm/drm_simple_kms_helper.c b/drivers/gpu/drm/drm_simple_kms_helper.c
index 7a00455ca568..51fa978f0d23 100644
--- a/drivers/gpu/drm/drm_simple_kms_helper.c
+++ b/drivers/gpu/drm/drm_simple_kms_helper.c
@@ -52,7 +52,7 @@ static int drm_simple_kms_crtc_check(struct drm_crtc *crtc,
 				     struct drm_crtc_state *state)
 {
 	bool has_primary = state->plane_mask &
-			   BIT(drm_plane_index(crtc->primary));
+			   drm_plane_mask(crtc->primary);
 
 	/* We always want to have an active plane with an active CRTC */
 	if (has_primary != state->enable)
@@ -281,13 +281,13 @@ int drm_simple_display_pipe_init(struct drm_device *dev,
 	if (ret)
 		return ret;
 
-	encoder->possible_crtcs = 1 << drm_crtc_index(crtc);
+	encoder->possible_crtcs = drm_crtc_mask(crtc);
 	ret = drm_encoder_init(dev, encoder, &drm_simple_kms_encoder_funcs,
 			       DRM_MODE_ENCODER_NONE, NULL);
 	if (ret || !connector)
 		return ret;
 
-	return drm_mode_connector_attach_encoder(connector, encoder);
+	return drm_connector_attach_encoder(connector, encoder);
 }
 EXPORT_SYMBOL(drm_simple_display_pipe_init);
 
diff --git a/drivers/gpu/drm/drm_syncobj.c b/drivers/gpu/drm/drm_syncobj.c
index d4f4ce484529..adb3cb27d31e 100644
--- a/drivers/gpu/drm/drm_syncobj.c
+++ b/drivers/gpu/drm/drm_syncobj.c
@@ -207,7 +207,6 @@ static const struct dma_fence_ops drm_syncobj_null_fence_ops = {
 	.get_driver_name = drm_syncobj_null_fence_get_name,
 	.get_timeline_name = drm_syncobj_null_fence_get_name,
 	.enable_signaling = drm_syncobj_null_fence_enable_signaling,
-	.wait = dma_fence_default_wait,
 	.release = NULL,
 };
 
diff --git a/drivers/gpu/drm/drm_sysfs.c b/drivers/gpu/drm/drm_sysfs.c
index 1c5b5ce1fd7f..b3c1daad1169 100644
--- a/drivers/gpu/drm/drm_sysfs.c
+++ b/drivers/gpu/drm/drm_sysfs.c
@@ -331,9 +331,7 @@ struct device *drm_sysfs_minor_alloc(struct drm_minor *minor)
 	struct device *kdev;
 	int r;
 
-	if (minor->type == DRM_MINOR_CONTROL)
-		minor_str = "controlD%d";
-	else if (minor->type == DRM_MINOR_RENDER)
+	if (minor->type == DRM_MINOR_RENDER)
 		minor_str = "renderD%d";
 	else
 		minor_str = "card%d";
diff --git a/drivers/gpu/drm/drm_vm.c b/drivers/gpu/drm/drm_vm.c
index 2660543ad86a..c3301046dfaa 100644
--- a/drivers/gpu/drm/drm_vm.c
+++ b/drivers/gpu/drm/drm_vm.c
@@ -100,7 +100,7 @@ static pgprot_t drm_dma_prot(uint32_t map_type, struct vm_area_struct *vma)
  * map, get the page, increment the use count and return it.
  */
 #if IS_ENABLED(CONFIG_AGP)
-static int drm_vm_fault(struct vm_fault *vmf)
+static vm_fault_t drm_vm_fault(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 	struct drm_file *priv = vma->vm_file->private_data;
@@ -173,7 +173,7 @@ vm_fault_error:
 	return VM_FAULT_SIGBUS;	/* Disallow mremap */
 }
 #else
-static int drm_vm_fault(struct vm_fault *vmf)
+static vm_fault_t drm_vm_fault(struct vm_fault *vmf)
 {
 	return VM_FAULT_SIGBUS;
 }
@@ -189,7 +189,7 @@ static int drm_vm_fault(struct vm_fault *vmf)
  * Get the mapping, find the real physical page to map, get the page, and
  * return it.
  */
-static int drm_vm_shm_fault(struct vm_fault *vmf)
+static vm_fault_t drm_vm_shm_fault(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 	struct drm_local_map *map = vma->vm_private_data;
@@ -291,7 +291,7 @@ static void drm_vm_shm_close(struct vm_area_struct *vma)
  *
  * Determine the page number from the page offset and get it from drm_device_dma::pagelist.
  */
-static int drm_vm_dma_fault(struct vm_fault *vmf)
+static vm_fault_t drm_vm_dma_fault(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 	struct drm_file *priv = vma->vm_file->private_data;
@@ -326,7 +326,7 @@ static int drm_vm_dma_fault(struct vm_fault *vmf)
  *
  * Determine the map offset from the page offset and get it from drm_sg_mem::pagelist.
  */
-static int drm_vm_sg_fault(struct vm_fault *vmf)
+static vm_fault_t drm_vm_sg_fault(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 	struct drm_local_map *map = vma->vm_private_data;
diff --git a/drivers/gpu/drm/drm_vma_manager.c b/drivers/gpu/drm/drm_vma_manager.c
index 23c749c05b5a..a6b2fe36b025 100644
--- a/drivers/gpu/drm/drm_vma_manager.c
+++ b/drivers/gpu/drm/drm_vma_manager.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /*
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
  * Copyright (c) 2012 David Airlie <airlied@linux.ie>
diff --git a/drivers/gpu/drm/drm_writeback.c b/drivers/gpu/drm/drm_writeback.c
new file mode 100644
index 000000000000..c20e6fe00cb3
--- /dev/null
+++ b/drivers/gpu/drm/drm_writeback.c
@@ -0,0 +1,353 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * (C) COPYRIGHT 2016 ARM Limited. All rights reserved.
+ * Author: Brian Starkey <brian.starkey@arm.com>
+ *
+ * This program is free software and is provided to you under the terms of the
+ * GNU General Public License version 2 as published by the Free Software
+ * Foundation, and any use by you of this program is subject to the terms
+ * of such GNU licence.
+ */
+
+#include <drm/drm_crtc.h>
+#include <drm/drm_modeset_helper_vtables.h>
+#include <drm/drm_property.h>
+#include <drm/drm_writeback.h>
+#include <drm/drmP.h>
+#include <linux/dma-fence.h>
+
+/**
+ * DOC: overview
+ *
+ * Writeback connectors are used to expose hardware which can write the output
+ * from a CRTC to a memory buffer. They are used and act similarly to other
+ * types of connectors, with some important differences:
+ *
+ * * Writeback connectors don't provide a way to output visually to the user.
+ *
+ * * Writeback connectors are visible to userspace only when the client sets
+ *   DRM_CLIENT_CAP_WRITEBACK_CONNECTORS.
+ *
+ * * Writeback connectors don't have EDID.
+ *
+ * A framebuffer may only be attached to a writeback connector when the
+ * connector is attached to a CRTC. The WRITEBACK_FB_ID property which sets the
+ * framebuffer applies only to a single commit (see below). A framebuffer may
+ * not be attached while the CRTC is off.
+ *
+ * Unlike with planes, when a writeback framebuffer is removed by userspace DRM
+ * makes no attempt to remove it from active use by the connector. This is
+ * because no method is provided to abort a writeback operation, and in any
+ * case making a new commit whilst a writeback is ongoing is undefined (see
+ * WRITEBACK_OUT_FENCE_PTR below). As soon as the current writeback is finished,
+ * the framebuffer will automatically no longer be in active use. As it will
+ * also have already been removed from the framebuffer list, there will be no
+ * way for any userspace application to retrieve a reference to it in the
+ * intervening period.
+ *
+ * Writeback connectors have some additional properties, which userspace
+ * can use to query and control them:
+ *
+ *  "WRITEBACK_FB_ID":
+ *	Write-only object property storing a DRM_MODE_OBJECT_FB: it stores the
+ *	framebuffer to be written by the writeback connector. This property is
+ *	similar to the FB_ID property on planes, but will always read as zero
+ *	and is not preserved across commits.
+ *	Userspace must set this property to an output buffer every time it
+ *	wishes the buffer to get filled.
+ *
+ *  "WRITEBACK_PIXEL_FORMATS":
+ *	Immutable blob property to store the supported pixel formats table. The
+ *	data is an array of u32 DRM_FORMAT_* fourcc values.
+ *	Userspace can use this blob to find out what pixel formats are supported
+ *	by the connector's writeback engine.
+ *
+ *  "WRITEBACK_OUT_FENCE_PTR":
+ *	Userspace can use this property to provide a pointer for the kernel to
+ *	fill with a sync_file file descriptor, which will signal once the
+ *	writeback is finished. The value should be the address of a 32-bit
+ *	signed integer, cast to a u64.
+ *	Userspace should wait for this fence to signal before making another
+ *	commit affecting any of the same CRTCs, Planes or Connectors.
+ *	**Failure to do so will result in undefined behaviour.**
+ *	For this reason it is strongly recommended that all userspace
+ *	applications making use of writeback connectors *always* retrieve an
+ *	out-fence for the commit and use it appropriately.
+ *	From userspace, this property will always read as zero.
+ */
+
+#define fence_to_wb_connector(x) container_of(x->lock, \
+					      struct drm_writeback_connector, \
+					      fence_lock)
+
+static const char *drm_writeback_fence_get_driver_name(struct dma_fence *fence)
+{
+	struct drm_writeback_connector *wb_connector =
+		fence_to_wb_connector(fence);
+
+	return wb_connector->base.dev->driver->name;
+}
+
+static const char *
+drm_writeback_fence_get_timeline_name(struct dma_fence *fence)
+{
+	struct drm_writeback_connector *wb_connector =
+		fence_to_wb_connector(fence);
+
+	return wb_connector->timeline_name;
+}
+
+static bool drm_writeback_fence_enable_signaling(struct dma_fence *fence)
+{
+	return true;
+}
+
+static const struct dma_fence_ops drm_writeback_fence_ops = {
+	.get_driver_name = drm_writeback_fence_get_driver_name,
+	.get_timeline_name = drm_writeback_fence_get_timeline_name,
+	.enable_signaling = drm_writeback_fence_enable_signaling,
+	.wait = dma_fence_default_wait,
+};
+
+static int create_writeback_properties(struct drm_device *dev)
+{
+	struct drm_property *prop;
+
+	if (!dev->mode_config.writeback_fb_id_property) {
+		prop = drm_property_create_object(dev, DRM_MODE_PROP_ATOMIC,
+						  "WRITEBACK_FB_ID",
+						  DRM_MODE_OBJECT_FB);
+		if (!prop)
+			return -ENOMEM;
+		dev->mode_config.writeback_fb_id_property = prop;
+	}
+
+	if (!dev->mode_config.writeback_pixel_formats_property) {
+		prop = drm_property_create(dev, DRM_MODE_PROP_BLOB |
+					   DRM_MODE_PROP_ATOMIC |
+					   DRM_MODE_PROP_IMMUTABLE,
+					   "WRITEBACK_PIXEL_FORMATS", 0);
+		if (!prop)
+			return -ENOMEM;
+		dev->mode_config.writeback_pixel_formats_property = prop;
+	}
+
+	if (!dev->mode_config.writeback_out_fence_ptr_property) {
+		prop = drm_property_create_range(dev, DRM_MODE_PROP_ATOMIC,
+						 "WRITEBACK_OUT_FENCE_PTR", 0,
+						 U64_MAX);
+		if (!prop)
+			return -ENOMEM;
+		dev->mode_config.writeback_out_fence_ptr_property = prop;
+	}
+
+	return 0;
+}
+
+static const struct drm_encoder_funcs drm_writeback_encoder_funcs = {
+	.destroy = drm_encoder_cleanup,
+};
+
+/**
+ * drm_writeback_connector_init - Initialize a writeback connector and its properties
+ * @dev: DRM device
+ * @wb_connector: Writeback connector to initialize
+ * @con_funcs: Connector funcs vtable
+ * @enc_helper_funcs: Encoder helper funcs vtable to be used by the internal encoder
+ * @formats: Array of supported pixel formats for the writeback engine
+ * @n_formats: Length of the formats array
+ *
+ * This function creates the writeback-connector-specific properties if they
+ * have not been already created, initializes the connector as
+ * type DRM_MODE_CONNECTOR_WRITEBACK, and correctly initializes the property
+ * values. It will also create an internal encoder associated with the
+ * drm_writeback_connector and set it to use the @enc_helper_funcs vtable for
+ * the encoder helper.
+ *
+ * Drivers should always use this function instead of drm_connector_init() to
+ * set up writeback connectors.
+ *
+ * Returns: 0 on success, or a negative error code
+ */
+int drm_writeback_connector_init(struct drm_device *dev,
+				 struct drm_writeback_connector *wb_connector,
+				 const struct drm_connector_funcs *con_funcs,
+				 const struct drm_encoder_helper_funcs *enc_helper_funcs,
+				 const u32 *formats, int n_formats)
+{
+	struct drm_property_blob *blob;
+	struct drm_connector *connector = &wb_connector->base;
+	struct drm_mode_config *config = &dev->mode_config;
+	int ret = create_writeback_properties(dev);
+
+	if (ret != 0)
+		return ret;
+
+	blob = drm_property_create_blob(dev, n_formats * sizeof(*formats),
+					formats);
+	if (IS_ERR(blob))
+		return PTR_ERR(blob);
+
+	drm_encoder_helper_add(&wb_connector->encoder, enc_helper_funcs);
+	ret = drm_encoder_init(dev, &wb_connector->encoder,
+			       &drm_writeback_encoder_funcs,
+			       DRM_MODE_ENCODER_VIRTUAL, NULL);
+	if (ret)
+		goto fail;
+
+	connector->interlace_allowed = 0;
+
+	ret = drm_connector_init(dev, connector, con_funcs,
+				 DRM_MODE_CONNECTOR_WRITEBACK);
+	if (ret)
+		goto connector_fail;
+
+	ret = drm_connector_attach_encoder(connector,
+						&wb_connector->encoder);
+	if (ret)
+		goto attach_fail;
+
+	INIT_LIST_HEAD(&wb_connector->job_queue);
+	spin_lock_init(&wb_connector->job_lock);
+
+	wb_connector->fence_context = dma_fence_context_alloc(1);
+	spin_lock_init(&wb_connector->fence_lock);
+	snprintf(wb_connector->timeline_name,
+		 sizeof(wb_connector->timeline_name),
+		 "CONNECTOR:%d-%s", connector->base.id, connector->name);
+
+	drm_object_attach_property(&connector->base,
+				   config->writeback_out_fence_ptr_property, 0);
+
+	drm_object_attach_property(&connector->base,
+				   config->writeback_fb_id_property, 0);
+
+	drm_object_attach_property(&connector->base,
+				   config->writeback_pixel_formats_property,
+				   blob->base.id);
+	wb_connector->pixel_formats_blob_ptr = blob;
+
+	return 0;
+
+attach_fail:
+	drm_connector_cleanup(connector);
+connector_fail:
+	drm_encoder_cleanup(&wb_connector->encoder);
+fail:
+	drm_property_blob_put(blob);
+	return ret;
+}
+EXPORT_SYMBOL(drm_writeback_connector_init);
+
+/**
+ * drm_writeback_queue_job - Queue a writeback job for later signalling
+ * @wb_connector: The writeback connector to queue a job on
+ * @job: The job to queue
+ *
+ * This function adds a job to the job_queue for a writeback connector. It
+ * should be considered to take ownership of the writeback job, and so any other
+ * references to the job must be cleared after calling this function.
+ *
+ * Drivers must ensure that for a given writeback connector, jobs are queued in
+ * exactly the same order as they will be completed by the hardware (and
+ * signaled via drm_writeback_signal_completion).
+ *
+ * For every call to drm_writeback_queue_job() there must be exactly one call to
+ * drm_writeback_signal_completion()
+ *
+ * See also: drm_writeback_signal_completion()
+ */
+void drm_writeback_queue_job(struct drm_writeback_connector *wb_connector,
+			     struct drm_writeback_job *job)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&wb_connector->job_lock, flags);
+	list_add_tail(&job->list_entry, &wb_connector->job_queue);
+	spin_unlock_irqrestore(&wb_connector->job_lock, flags);
+}
+EXPORT_SYMBOL(drm_writeback_queue_job);
+
+/*
+ * @cleanup_work: deferred cleanup of a writeback job
+ *
+ * The job cannot be cleaned up directly in drm_writeback_signal_completion,
+ * because it may be called in interrupt context. Dropping the framebuffer
+ * reference can sleep, and so the cleanup is deferred to a workqueue.
+ */
+static void cleanup_work(struct work_struct *work)
+{
+	struct drm_writeback_job *job = container_of(work,
+						     struct drm_writeback_job,
+						     cleanup_work);
+	drm_framebuffer_put(job->fb);
+	kfree(job);
+}
+
+
+/**
+ * drm_writeback_signal_completion - Signal the completion of a writeback job
+ * @wb_connector: The writeback connector whose job is complete
+ * @status: Status code to set in the writeback out_fence (0 for success)
+ *
+ * Drivers should call this to signal the completion of a previously queued
+ * writeback job. It should be called as soon as possible after the hardware
+ * has finished writing, and may be called from interrupt context.
+ * It is the driver's responsibility to ensure that for a given connector, the
+ * hardware completes writeback jobs in the same order as they are queued.
+ *
+ * Unless the driver is holding its own reference to the framebuffer, it must
+ * not be accessed after calling this function.
+ *
+ * See also: drm_writeback_queue_job()
+ */
+void
+drm_writeback_signal_completion(struct drm_writeback_connector *wb_connector,
+				int status)
+{
+	unsigned long flags;
+	struct drm_writeback_job *job;
+
+	spin_lock_irqsave(&wb_connector->job_lock, flags);
+	job = list_first_entry_or_null(&wb_connector->job_queue,
+				       struct drm_writeback_job,
+				       list_entry);
+	if (job) {
+		list_del(&job->list_entry);
+		if (job->out_fence) {
+			if (status)
+				dma_fence_set_error(job->out_fence, status);
+			dma_fence_signal(job->out_fence);
+			dma_fence_put(job->out_fence);
+		}
+	}
+	spin_unlock_irqrestore(&wb_connector->job_lock, flags);
+
+	if (WARN_ON(!job))
+		return;
+
+	INIT_WORK(&job->cleanup_work, cleanup_work);
+	queue_work(system_long_wq, &job->cleanup_work);
+}
+EXPORT_SYMBOL(drm_writeback_signal_completion);
+
+struct dma_fence *
+drm_writeback_get_out_fence(struct drm_writeback_connector *wb_connector)
+{
+	struct dma_fence *fence;
+
+	if (WARN_ON(wb_connector->base.connector_type !=
+		    DRM_MODE_CONNECTOR_WRITEBACK))
+		return NULL;
+
+	fence = kzalloc(sizeof(*fence), GFP_KERNEL);
+	if (!fence)
+		return NULL;
+
+	dma_fence_init(fence, &drm_writeback_fence_ops,
+		       &wb_connector->fence_lock, wb_connector->fence_context,
+		       ++wb_connector->fence_seqno);
+
+	return fence;
+}
+EXPORT_SYMBOL(drm_writeback_get_out_fence);
diff --git a/drivers/gpu/drm/etnaviv/Kconfig b/drivers/gpu/drm/etnaviv/Kconfig
index e5bfeca361bd..041a77e400d4 100644
--- a/drivers/gpu/drm/etnaviv/Kconfig
+++ b/drivers/gpu/drm/etnaviv/Kconfig
@@ -22,11 +22,3 @@ config DRM_ETNAVIV_THERMAL
 	help
 	  Compile in support for thermal throttling.
 	  Say Y unless you want to risk burning your SoC.
-
-config DRM_ETNAVIV_REGISTER_LOGGING
-	bool "enable ETNAVIV register logging"
-	depends on DRM_ETNAVIV
-	help
-	  Compile in support for logging register reads/writes in a format
-	  that can be parsed by envytools demsm tool.  If enabled, register
-	  logging can be switched on via etnaviv.reglog=y module param.
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_buffer.c b/drivers/gpu/drm/etnaviv/etnaviv_buffer.c
index bfc6d4aa3b7c..7fea74861a87 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_buffer.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_buffer.c
@@ -1,18 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2014 Etnaviv Project
- * Author: Christian Gmeiner <christian.gmeiner@gmail.com>
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2014-2018 Etnaviv Project
  */
 
 #include "etnaviv_cmdbuf.h"
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_cmd_parser.c b/drivers/gpu/drm/etnaviv/etnaviv_cmd_parser.c
index 68e6d3772ad8..b106e8b288ad 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_cmd_parser.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_cmd_parser.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #include <linux/kernel.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.c b/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.c
index 3746827f45eb..a3c44f145c1d 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2017 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2017-2018 Etnaviv Project
  */
 
 #include <drm/drm_mm.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.h b/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.h
index ddc3f7ea169c..acb68c698363 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_cmdbuf.h
@@ -1,17 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
  * Copyright (C) 2017 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
 #ifndef __ETNAVIV_CMDBUF_H__
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.c b/drivers/gpu/drm/etnaviv/etnaviv_drv.c
index ab50090d066c..36414ba56b22 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_drv.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #include <linux/component.h>
@@ -25,57 +14,6 @@
 #include "etnaviv_mmu.h"
 #include "etnaviv_perfmon.h"
 
-#ifdef CONFIG_DRM_ETNAVIV_REGISTER_LOGGING
-static bool reglog;
-MODULE_PARM_DESC(reglog, "Enable register read/write logging");
-module_param(reglog, bool, 0600);
-#else
-#define reglog 0
-#endif
-
-void __iomem *etnaviv_ioremap(struct platform_device *pdev, const char *name,
-		const char *dbgname)
-{
-	struct resource *res;
-	void __iomem *ptr;
-
-	if (name)
-		res = platform_get_resource_byname(pdev, IORESOURCE_MEM, name);
-	else
-		res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-
-	ptr = devm_ioremap_resource(&pdev->dev, res);
-	if (IS_ERR(ptr)) {
-		dev_err(&pdev->dev, "failed to ioremap %s: %ld\n", name,
-			PTR_ERR(ptr));
-		return ptr;
-	}
-
-	if (reglog)
-		dev_printk(KERN_DEBUG, &pdev->dev, "IO:region %s 0x%p %08zx\n",
-			   dbgname, ptr, (size_t)resource_size(res));
-
-	return ptr;
-}
-
-void etnaviv_writel(u32 data, void __iomem *addr)
-{
-	if (reglog)
-		printk(KERN_DEBUG "IO:W %p %08x\n", addr, data);
-
-	writel(data, addr);
-}
-
-u32 etnaviv_readl(const void __iomem *addr)
-{
-	u32 val = readl(addr);
-
-	if (reglog)
-		printk(KERN_DEBUG "IO:R %p %08x\n", addr, val);
-
-	return val;
-}
-
 /*
  * DRM operations:
  */
@@ -111,12 +49,12 @@ static int etnaviv_open(struct drm_device *dev, struct drm_file *file)
 
 	for (i = 0; i < ETNA_MAX_PIPES; i++) {
 		struct etnaviv_gpu *gpu = priv->gpu[i];
+		struct drm_sched_rq *rq;
 
 		if (gpu) {
-			drm_sched_entity_init(&gpu->sched,
-				&ctx->sched_entity[i],
-				&gpu->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL],
-				32, NULL);
+			rq = &gpu->sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
+			drm_sched_entity_init(&ctx->sched_entity[i],
+					      &rq, 1, NULL);
 			}
 	}
 
@@ -140,8 +78,8 @@ static void etnaviv_postclose(struct drm_device *dev, struct drm_file *file)
 				gpu->lastctx = NULL;
 			mutex_unlock(&gpu->lock);
 
-			drm_sched_entity_fini(&gpu->sched,
-					      &ctx->sched_entity[i]);
+			drm_sched_entity_destroy(&gpu->sched,
+						&ctx->sched_entity[i]);
 		}
 	}
 
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.h b/drivers/gpu/drm/etnaviv/etnaviv_drv.h
index ddb17ee565e9..d36c7bbe66db 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_drv.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.h
@@ -1,17 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #ifndef __ETNAVIV_DRV_H__
@@ -26,6 +15,7 @@
 #include <linux/pm_runtime.h>
 #include <linux/slab.h>
 #include <linux/list.h>
+#include <linux/time64.h>
 #include <linux/types.h>
 #include <linux/sizes.h>
 
@@ -101,11 +91,6 @@ void etnaviv_gem_describe_objects(struct etnaviv_drm_private *priv,
 	struct seq_file *m);
 #endif
 
-void __iomem *etnaviv_ioremap(struct platform_device *pdev, const char *name,
-		const char *dbgname);
-void etnaviv_writel(u32 data, void __iomem *addr);
-u32 etnaviv_readl(const void __iomem *addr);
-
 #define DBG(fmt, ...) DRM_DEBUG(fmt"\n", ##__VA_ARGS__)
 #define VERB(fmt, ...) if (0) DRM_DEBUG(fmt"\n", ##__VA_ARGS__)
 
@@ -132,19 +117,27 @@ static inline bool fence_after_eq(u32 a, u32 b)
 	return (s32)(a - b) >= 0;
 }
 
+/*
+ * Etnaviv timeouts are specified wrt CLOCK_MONOTONIC, not jiffies.
+ * We need to calculate the timeout in terms of number of jiffies
+ * between the specified timeout and the current CLOCK_MONOTONIC time.
+ */
 static inline unsigned long etnaviv_timeout_to_jiffies(
 	const struct timespec *timeout)
 {
-	unsigned long timeout_jiffies = timespec_to_jiffies(timeout);
-	unsigned long start_jiffies = jiffies;
-	unsigned long remaining_jiffies;
+	struct timespec64 ts, to;
+
+	to = timespec_to_timespec64(*timeout);
+
+	ktime_get_ts64(&ts);
+
+	/* timeouts before "now" have already expired */
+	if (timespec64_compare(&to, &ts) <= 0)
+		return 0;
 
-	if (time_after(start_jiffies, timeout_jiffies))
-		remaining_jiffies = 0;
-	else
-		remaining_jiffies = timeout_jiffies - start_jiffies;
+	ts = timespec64_sub(to, ts);
 
-	return remaining_jiffies;
+	return timespec64_to_jiffies(&ts);
 }
 
 #endif /* __ETNAVIV_DRV_H__ */
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_dump.c b/drivers/gpu/drm/etnaviv/etnaviv_dump.c
index 48aef6cf6a42..9146e30e24a6 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_dump.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_dump.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #include <linux/devcoredump.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_dump.h b/drivers/gpu/drm/etnaviv/etnaviv_dump.h
index 97f2f8db9133..2d916c2667ee 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_dump.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_dump.h
@@ -1,20 +1,8 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
  * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
- *
- * Etnaviv devcoredump file definitions
  */
+
 #ifndef ETNAVIV_DUMP_H
 #define ETNAVIV_DUMP_H
 
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem.c b/drivers/gpu/drm/etnaviv/etnaviv_gem.c
index fcc969fa0e69..209ef1274b80 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gem.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gem.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #include <linux/spinlock.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem.h b/drivers/gpu/drm/etnaviv/etnaviv_gem.h
index 93e696fcc14f..76079c2291f8 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gem.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gem.h
@@ -1,17 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #ifndef __ETNAVIV_GEM_H__
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c b/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c
index 5704305d41e6..0566171f8df2 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c
@@ -1,18 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2013 Red Hat
- * Author: Rob Clark <robdclark@gmail.com>
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2014-2018 Etnaviv Project
  */
 
 #include <linux/dma-buf.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c
index 8a88799bf79b..19b09a59e30e 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #include <linux/component.h>
@@ -1038,11 +1027,6 @@ static const char *etnaviv_fence_get_timeline_name(struct dma_fence *fence)
 	return dev_name(f->gpu->dev);
 }
 
-static bool etnaviv_fence_enable_signaling(struct dma_fence *fence)
-{
-	return true;
-}
-
 static bool etnaviv_fence_signaled(struct dma_fence *fence)
 {
 	struct etnaviv_fence *f = to_etnaviv_fence(fence);
@@ -1060,9 +1044,7 @@ static void etnaviv_fence_release(struct dma_fence *fence)
 static const struct dma_fence_ops etnaviv_fence_ops = {
 	.get_driver_name = etnaviv_fence_get_driver_name,
 	.get_timeline_name = etnaviv_fence_get_timeline_name,
-	.enable_signaling = etnaviv_fence_enable_signaling,
 	.signaled = etnaviv_fence_signaled,
-	.wait = dma_fence_default_wait,
 	.release = etnaviv_fence_release,
 };
 
@@ -1735,6 +1717,7 @@ static int etnaviv_gpu_platform_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
 	struct etnaviv_gpu *gpu;
+	struct resource *res;
 	int err;
 
 	gpu = devm_kzalloc(dev, sizeof(*gpu), GFP_KERNEL);
@@ -1746,7 +1729,8 @@ static int etnaviv_gpu_platform_probe(struct platform_device *pdev)
 	mutex_init(&gpu->fence_idr_lock);
 
 	/* Map registers: */
-	gpu->mmio = etnaviv_ioremap(pdev, NULL, dev_name(gpu->dev));
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	gpu->mmio = devm_ioremap_resource(&pdev->dev, res);
 	if (IS_ERR(gpu->mmio))
 		return PTR_ERR(gpu->mmio);
 
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.h b/drivers/gpu/drm/etnaviv/etnaviv_gpu.h
index 3c3005501846..dd430f0f8ff5 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.h
@@ -1,17 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #ifndef __ETNAVIV_GPU_H__
@@ -161,12 +150,12 @@ struct etnaviv_gpu {
 
 static inline void gpu_write(struct etnaviv_gpu *gpu, u32 reg, u32 data)
 {
-	etnaviv_writel(data, gpu->mmio + reg);
+	writel(data, gpu->mmio + reg);
 }
 
 static inline u32 gpu_read(struct etnaviv_gpu *gpu, u32 reg)
 {
-	return etnaviv_readl(gpu->mmio + reg);
+	return readl(gpu->mmio + reg);
 }
 
 static inline bool fence_completed(struct etnaviv_gpu *gpu, u32 fence)
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_hwdb.c b/drivers/gpu/drm/etnaviv/etnaviv_hwdb.c
index ea08bb38caaf..39b463db76c9 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_hwdb.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_hwdb.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Copyright (C) 2018 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
 #include "etnaviv_gpu.h"
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_iommu.c b/drivers/gpu/drm/etnaviv/etnaviv_iommu.c
index 4b9b11ca6f03..b163bdbcb880 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_iommu.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_iommu.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2014 Christian Gmeiner <christian.gmeiner@gmail.com>
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2014-2018 Etnaviv Project
  */
 
 #include <linux/platform_device.h>
@@ -47,11 +36,10 @@ static int __etnaviv_iommu_init(struct etnaviv_iommuv1_domain *etnaviv_domain)
 	u32 *p;
 	int i;
 
-	etnaviv_domain->base.bad_page_cpu = dma_alloc_coherent(
-						etnaviv_domain->base.dev,
-						SZ_4K,
-						&etnaviv_domain->base.bad_page_dma,
-						GFP_KERNEL);
+	etnaviv_domain->base.bad_page_cpu =
+			dma_alloc_wc(etnaviv_domain->base.dev, SZ_4K,
+				     &etnaviv_domain->base.bad_page_dma,
+				     GFP_KERNEL);
 	if (!etnaviv_domain->base.bad_page_cpu)
 		return -ENOMEM;
 
@@ -59,14 +47,14 @@ static int __etnaviv_iommu_init(struct etnaviv_iommuv1_domain *etnaviv_domain)
 	for (i = 0; i < SZ_4K / 4; i++)
 		*p++ = 0xdead55aa;
 
-	etnaviv_domain->pgtable_cpu =
-			dma_alloc_coherent(etnaviv_domain->base.dev, PT_SIZE,
-					   &etnaviv_domain->pgtable_dma,
-					   GFP_KERNEL);
+	etnaviv_domain->pgtable_cpu = dma_alloc_wc(etnaviv_domain->base.dev,
+						   PT_SIZE,
+						   &etnaviv_domain->pgtable_dma,
+						   GFP_KERNEL);
 	if (!etnaviv_domain->pgtable_cpu) {
-		dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-				  etnaviv_domain->base.bad_page_cpu,
-				  etnaviv_domain->base.bad_page_dma);
+		dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+			    etnaviv_domain->base.bad_page_cpu,
+			    etnaviv_domain->base.bad_page_dma);
 		return -ENOMEM;
 	}
 
@@ -81,13 +69,12 @@ static void etnaviv_iommuv1_domain_free(struct etnaviv_iommu_domain *domain)
 	struct etnaviv_iommuv1_domain *etnaviv_domain =
 			to_etnaviv_domain(domain);
 
-	dma_free_coherent(etnaviv_domain->base.dev, PT_SIZE,
-			  etnaviv_domain->pgtable_cpu,
-			  etnaviv_domain->pgtable_dma);
+	dma_free_wc(etnaviv_domain->base.dev, PT_SIZE,
+		    etnaviv_domain->pgtable_cpu, etnaviv_domain->pgtable_dma);
 
-	dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-			  etnaviv_domain->base.bad_page_cpu,
-			  etnaviv_domain->base.bad_page_dma);
+	dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+		    etnaviv_domain->base.bad_page_cpu,
+		    etnaviv_domain->base.bad_page_dma);
 
 	kfree(etnaviv_domain);
 }
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_iommu.h b/drivers/gpu/drm/etnaviv/etnaviv_iommu.h
index 01d59bf70d78..b279404ce91a 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_iommu.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_iommu.h
@@ -1,17 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
- * Copyright (C) 2014 Christian Gmeiner <christian.gmeiner@gmail.com>
-  *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2014-2018 Etnaviv Project
  */
 
 #ifndef __ETNAVIV_IOMMU_H__
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c b/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
index 9752dbd5d28b..71fbc1f96cb6 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_iommu_v2.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2016 Etnaviv Project
-  *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2016-2018 Etnaviv Project
  */
 
 #include <linux/platform_device.h>
@@ -47,8 +36,8 @@ struct etnaviv_iommuv2_domain {
 	u32 *mtlb_cpu;
 	dma_addr_t mtlb_dma;
 	/* S(lave) TLB aka second level pagetable */
-	u32 *stlb_cpu[1024];
-	dma_addr_t stlb_dma[1024];
+	u32 *stlb_cpu[MMUv2_MAX_STLB_ENTRIES];
+	dma_addr_t stlb_dma[MMUv2_MAX_STLB_ENTRIES];
 };
 
 static struct etnaviv_iommuv2_domain *
@@ -57,24 +46,54 @@ to_etnaviv_domain(struct etnaviv_iommu_domain *domain)
 	return container_of(domain, struct etnaviv_iommuv2_domain, base);
 }
 
+static int
+etnaviv_iommuv2_ensure_stlb(struct etnaviv_iommuv2_domain *etnaviv_domain,
+			    int stlb)
+{
+	if (etnaviv_domain->stlb_cpu[stlb])
+		return 0;
+
+	etnaviv_domain->stlb_cpu[stlb] =
+			dma_alloc_wc(etnaviv_domain->base.dev, SZ_4K,
+				     &etnaviv_domain->stlb_dma[stlb],
+				     GFP_KERNEL);
+
+	if (!etnaviv_domain->stlb_cpu[stlb])
+		return -ENOMEM;
+
+	memset32(etnaviv_domain->stlb_cpu[stlb], MMUv2_PTE_EXCEPTION,
+		 SZ_4K / sizeof(u32));
+
+	etnaviv_domain->mtlb_cpu[stlb] = etnaviv_domain->stlb_dma[stlb] |
+						      MMUv2_PTE_PRESENT;
+	return 0;
+}
+
 static int etnaviv_iommuv2_map(struct etnaviv_iommu_domain *domain,
 			       unsigned long iova, phys_addr_t paddr,
 			       size_t size, int prot)
 {
 	struct etnaviv_iommuv2_domain *etnaviv_domain =
 			to_etnaviv_domain(domain);
-	int mtlb_entry, stlb_entry;
-	u32 entry = (u32)paddr | MMUv2_PTE_PRESENT;
+	int mtlb_entry, stlb_entry, ret;
+	u32 entry = lower_32_bits(paddr) | MMUv2_PTE_PRESENT;
 
 	if (size != SZ_4K)
 		return -EINVAL;
 
+	if (IS_ENABLED(CONFIG_PHYS_ADDR_T_64BIT))
+		entry |= (upper_32_bits(paddr) & 0xff) << 4;
+
 	if (prot & ETNAVIV_PROT_WRITE)
 		entry |= MMUv2_PTE_WRITEABLE;
 
 	mtlb_entry = (iova & MMUv2_MTLB_MASK) >> MMUv2_MTLB_SHIFT;
 	stlb_entry = (iova & MMUv2_STLB_MASK) >> MMUv2_STLB_SHIFT;
 
+	ret = etnaviv_iommuv2_ensure_stlb(etnaviv_domain, mtlb_entry);
+	if (ret)
+		return ret;
+
 	etnaviv_domain->stlb_cpu[mtlb_entry][stlb_entry] = entry;
 
 	return 0;
@@ -101,14 +120,13 @@ static size_t etnaviv_iommuv2_unmap(struct etnaviv_iommu_domain *domain,
 static int etnaviv_iommuv2_init(struct etnaviv_iommuv2_domain *etnaviv_domain)
 {
 	u32 *p;
-	int ret, i, j;
+	int ret, i;
 
 	/* allocate scratch page */
-	etnaviv_domain->base.bad_page_cpu = dma_alloc_coherent(
-						etnaviv_domain->base.dev,
-						SZ_4K,
-						&etnaviv_domain->base.bad_page_dma,
-						GFP_KERNEL);
+	etnaviv_domain->base.bad_page_cpu =
+			dma_alloc_wc(etnaviv_domain->base.dev, SZ_4K,
+				     &etnaviv_domain->base.bad_page_dma,
+				     GFP_KERNEL);
 	if (!etnaviv_domain->base.bad_page_cpu) {
 		ret = -ENOMEM;
 		goto fail_mem;
@@ -117,67 +135,40 @@ static int etnaviv_iommuv2_init(struct etnaviv_iommuv2_domain *etnaviv_domain)
 	for (i = 0; i < SZ_4K / 4; i++)
 		*p++ = 0xdead55aa;
 
-	etnaviv_domain->pta_cpu = dma_alloc_coherent(etnaviv_domain->base.dev,
-						     SZ_4K,
-						     &etnaviv_domain->pta_dma,
-						     GFP_KERNEL);
+	etnaviv_domain->pta_cpu = dma_alloc_wc(etnaviv_domain->base.dev,
+					       SZ_4K, &etnaviv_domain->pta_dma,
+					       GFP_KERNEL);
 	if (!etnaviv_domain->pta_cpu) {
 		ret = -ENOMEM;
 		goto fail_mem;
 	}
 
-	etnaviv_domain->mtlb_cpu = dma_alloc_coherent(etnaviv_domain->base.dev,
-						  SZ_4K,
-						  &etnaviv_domain->mtlb_dma,
-						  GFP_KERNEL);
+	etnaviv_domain->mtlb_cpu = dma_alloc_wc(etnaviv_domain->base.dev,
+						SZ_4K, &etnaviv_domain->mtlb_dma,
+						GFP_KERNEL);
 	if (!etnaviv_domain->mtlb_cpu) {
 		ret = -ENOMEM;
 		goto fail_mem;
 	}
 
-	/* pre-populate STLB pages (may want to switch to on-demand later) */
-	for (i = 0; i < MMUv2_MAX_STLB_ENTRIES; i++) {
-		etnaviv_domain->stlb_cpu[i] =
-				dma_alloc_coherent(etnaviv_domain->base.dev,
-						   SZ_4K,
-						   &etnaviv_domain->stlb_dma[i],
-						   GFP_KERNEL);
-		if (!etnaviv_domain->stlb_cpu[i]) {
-			ret = -ENOMEM;
-			goto fail_mem;
-		}
-		p = etnaviv_domain->stlb_cpu[i];
-		for (j = 0; j < SZ_4K / 4; j++)
-			*p++ = MMUv2_PTE_EXCEPTION;
-
-		etnaviv_domain->mtlb_cpu[i] = etnaviv_domain->stlb_dma[i] |
-					      MMUv2_PTE_PRESENT;
-	}
+	memset32(etnaviv_domain->mtlb_cpu, MMUv2_PTE_EXCEPTION,
+		 MMUv2_MAX_STLB_ENTRIES);
 
 	return 0;
 
 fail_mem:
 	if (etnaviv_domain->base.bad_page_cpu)
-		dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-				  etnaviv_domain->base.bad_page_cpu,
-				  etnaviv_domain->base.bad_page_dma);
+		dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+			    etnaviv_domain->base.bad_page_cpu,
+			    etnaviv_domain->base.bad_page_dma);
 
 	if (etnaviv_domain->pta_cpu)
-		dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-				  etnaviv_domain->pta_cpu,
-				  etnaviv_domain->pta_dma);
+		dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+			    etnaviv_domain->pta_cpu, etnaviv_domain->pta_dma);
 
 	if (etnaviv_domain->mtlb_cpu)
-		dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-				  etnaviv_domain->mtlb_cpu,
-				  etnaviv_domain->mtlb_dma);
-
-	for (i = 0; i < MMUv2_MAX_STLB_ENTRIES; i++) {
-		if (etnaviv_domain->stlb_cpu[i])
-			dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-					  etnaviv_domain->stlb_cpu[i],
-					  etnaviv_domain->stlb_dma[i]);
-	}
+		dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+			    etnaviv_domain->mtlb_cpu, etnaviv_domain->mtlb_dma);
 
 	return ret;
 }
@@ -188,23 +179,21 @@ static void etnaviv_iommuv2_domain_free(struct etnaviv_iommu_domain *domain)
 			to_etnaviv_domain(domain);
 	int i;
 
-	dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-			  etnaviv_domain->base.bad_page_cpu,
-			  etnaviv_domain->base.bad_page_dma);
+	dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+		    etnaviv_domain->base.bad_page_cpu,
+		    etnaviv_domain->base.bad_page_dma);
 
-	dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-			  etnaviv_domain->pta_cpu,
-			  etnaviv_domain->pta_dma);
+	dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+		    etnaviv_domain->pta_cpu, etnaviv_domain->pta_dma);
 
-	dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-			  etnaviv_domain->mtlb_cpu,
-			  etnaviv_domain->mtlb_dma);
+	dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+		    etnaviv_domain->mtlb_cpu, etnaviv_domain->mtlb_dma);
 
 	for (i = 0; i < MMUv2_MAX_STLB_ENTRIES; i++) {
 		if (etnaviv_domain->stlb_cpu[i])
-			dma_free_coherent(etnaviv_domain->base.dev, SZ_4K,
-					  etnaviv_domain->stlb_cpu[i],
-					  etnaviv_domain->stlb_dma[i]);
+			dma_free_wc(etnaviv_domain->base.dev, SZ_4K,
+				    etnaviv_domain->stlb_cpu[i],
+				    etnaviv_domain->stlb_dma[i]);
 	}
 
 	vfree(etnaviv_domain);
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_mmu.c b/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
index 49e049713a52..8069f9f36a2e 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_mmu.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #include "common.xml.h"
@@ -162,22 +151,10 @@ static int etnaviv_iommu_find_iova(struct etnaviv_iommu *mmu,
 		bool found;
 
 		ret = drm_mm_insert_node_in_range(&mmu->mm, node,
-						  size, 0, 0,
-						  mmu->last_iova, U64_MAX,
-						  mode);
+						  size, 0, 0, 0, U64_MAX, mode);
 		if (ret != -ENOSPC)
 			break;
 
-		/*
-		 * If we did not search from the start of the MMU region,
-		 * try again in case there are free slots.
-		 */
-		if (mmu->last_iova) {
-			mmu->last_iova = 0;
-			mmu->need_flush = true;
-			continue;
-		}
-
 		/* Try to retire some entries */
 		drm_mm_scan_init(&scan, &mmu->mm, size, 0, 0, mode);
 
@@ -274,7 +251,6 @@ int etnaviv_iommu_map_gem(struct etnaviv_iommu *mmu,
 	if (ret < 0)
 		goto unlock;
 
-	mmu->last_iova = node->start + etnaviv_obj->base.size;
 	mapping->iova = node->start;
 	ret = etnaviv_iommu_map(mmu, node->start, sgt, etnaviv_obj->base.size,
 				ETNAVIV_PROT_READ | ETNAVIV_PROT_WRITE);
@@ -381,7 +357,6 @@ int etnaviv_iommu_get_suballoc_va(struct etnaviv_gpu *gpu, dma_addr_t paddr,
 			mutex_unlock(&mmu->lock);
 			return ret;
 		}
-		mmu->last_iova = vram_node->start + size;
 		gpu->mmu->need_flush = true;
 		mutex_unlock(&mmu->lock);
 
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_mmu.h b/drivers/gpu/drm/etnaviv/etnaviv_mmu.h
index ab603f5166b1..a0db17ffb686 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_mmu.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_mmu.h
@@ -1,17 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
- * Copyright (C) 2015 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
+ * Copyright (C) 2015-2018 Etnaviv Project
  */
 
 #ifndef __ETNAVIV_MMU_H__
@@ -59,7 +48,6 @@ struct etnaviv_iommu {
 	struct mutex lock;
 	struct list_head mappings;
 	struct drm_mm mm;
-	u32 last_iova;
 	bool need_flush;
 };
 
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c b/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
index 26dddfc41aac..9980d81a26e3 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_perfmon.c
@@ -1,18 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Copyright (C) 2017 Etnaviv Project
  * Copyright (C) 2017 Zodiac Inflight Innovations
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
 #include "etnaviv_gpu.h"
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_perfmon.h b/drivers/gpu/drm/etnaviv/etnaviv_perfmon.h
index c1653c64ab6b..4a9d508f6e10 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_perfmon.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_perfmon.h
@@ -1,18 +1,7 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
  * Copyright (C) 2017 Etnaviv Project
  * Copyright (C) 2017 Zodiac Inflight Innovations
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
 #ifndef __ETNAVIV_PERFMON_H__
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_sched.c b/drivers/gpu/drm/etnaviv/etnaviv_sched.c
index 6cf0775dbcd7..a74eb57af15b 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_sched.c
+++ b/drivers/gpu/drm/etnaviv/etnaviv_sched.c
@@ -1,17 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0
 /*
  * Copyright (C) 2017 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
 #include <linux/kthread.h>
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_sched.h b/drivers/gpu/drm/etnaviv/etnaviv_sched.h
index 097635fa78ae..c0a6796e22c9 100644
--- a/drivers/gpu/drm/etnaviv/etnaviv_sched.h
+++ b/drivers/gpu/drm/etnaviv/etnaviv_sched.h
@@ -1,17 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 */
 /*
  * Copyright (C) 2017 Etnaviv Project
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License version 2 as published by
- * the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program.  If not, see <http://www.gnu.org/licenses/>.
  */
 
 #ifndef __ETNAVIV_SCHED_H__
diff --git a/drivers/gpu/drm/exynos/Kconfig b/drivers/gpu/drm/exynos/Kconfig
index 735ce47688f9..208bc27be3cc 100644
--- a/drivers/gpu/drm/exynos/Kconfig
+++ b/drivers/gpu/drm/exynos/Kconfig
@@ -1,6 +1,6 @@
 config DRM_EXYNOS
 	tristate "DRM Support for Samsung SoC EXYNOS Series"
-	depends on OF && DRM && (ARCH_S3C64XX || ARCH_EXYNOS || ARCH_MULTIPLATFORM)
+	depends on OF && DRM && (ARCH_S3C64XX || ARCH_S5PV210 || ARCH_EXYNOS || ARCH_MULTIPLATFORM)
 	select DRM_KMS_HELPER
 	select VIDEOMODE_HELPERS
 	select SND_SOC_HDMI_CODEC if SND_SOC
@@ -95,21 +95,31 @@ config DRM_EXYNOS_G2D
 	help
 	  Choose this option if you want to use Exynos G2D for DRM.
 
+config DRM_EXYNOS_IPP
+	bool
+
 config DRM_EXYNOS_FIMC
 	bool "FIMC"
-	depends on BROKEN && MFD_SYSCON
+	select DRM_EXYNOS_IPP
 	help
 	  Choose this option if you want to use Exynos FIMC for DRM.
 
 config DRM_EXYNOS_ROTATOR
 	bool "Rotator"
-	depends on BROKEN
+	select DRM_EXYNOS_IPP
 	help
 	  Choose this option if you want to use Exynos Rotator for DRM.
 
+config DRM_EXYNOS_SCALER
+	bool "Scaler"
+	select DRM_EXYNOS_IPP
+	help
+	  Choose this option if you want to use Exynos Scaler for DRM.
+
 config DRM_EXYNOS_GSC
 	bool "GScaler"
-	depends on BROKEN && ARCH_EXYNOS5 && VIDEO_SAMSUNG_EXYNOS_GSC=n
+	depends on VIDEO_SAMSUNG_EXYNOS_GSC=n
+	select DRM_EXYNOS_IPP
 	help
 	  Choose this option if you want to use Exynos GSC for DRM.
 
diff --git a/drivers/gpu/drm/exynos/Makefile b/drivers/gpu/drm/exynos/Makefile
index a51c5459bb13..3b323f1e0475 100644
--- a/drivers/gpu/drm/exynos/Makefile
+++ b/drivers/gpu/drm/exynos/Makefile
@@ -18,8 +18,10 @@ exynosdrm-$(CONFIG_DRM_EXYNOS_MIXER)	+= exynos_mixer.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_HDMI)	+= exynos_hdmi.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_VIDI)	+= exynos_drm_vidi.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_G2D)	+= exynos_drm_g2d.o
+exynosdrm-$(CONFIG_DRM_EXYNOS_IPP)	+= exynos_drm_ipp.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_FIMC)	+= exynos_drm_fimc.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_ROTATOR)	+= exynos_drm_rotator.o
+exynosdrm-$(CONFIG_DRM_EXYNOS_SCALER)	+= exynos_drm_scaler.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_GSC)	+= exynos_drm_gsc.o
 exynosdrm-$(CONFIG_DRM_EXYNOS_MIC)     += exynos_drm_mic.o
 
diff --git a/drivers/gpu/drm/exynos/exynos5433_drm_decon.c b/drivers/gpu/drm/exynos/exynos5433_drm_decon.c
index 1c330f2a7a5d..82c95c34447f 100644
--- a/drivers/gpu/drm/exynos/exynos5433_drm_decon.c
+++ b/drivers/gpu/drm/exynos/exynos5433_drm_decon.c
@@ -31,7 +31,10 @@
 #define DSD_CFG_MUX 0x1004
 #define DSD_CFG_MUX_TE_UNMASK_GLOBAL BIT(13)
 
-#define WINDOWS_NR	3
+#define WINDOWS_NR	5
+#define PRIMARY_WIN	2
+#define CURSON_WIN	4
+
 #define MIN_FB_WIDTH_FOR_16WORD_BURST	128
 
 #define I80_HW_TRG	(1 << 0)
@@ -43,6 +46,9 @@ static const char * const decon_clks_name[] = {
 	"aclk_smmu_decon0x",
 	"aclk_xiu_decon0x",
 	"pclk_smmu_decon0x",
+	"aclk_smmu_decon1x",
+	"aclk_xiu_decon1x",
+	"pclk_smmu_decon1x",
 	"sclk_decon_vclk",
 	"sclk_decon_eclk",
 };
@@ -74,9 +80,8 @@ static const uint32_t decon_formats[] = {
 };
 
 static const enum drm_plane_type decon_win_types[WINDOWS_NR] = {
-	DRM_PLANE_TYPE_PRIMARY,
-	DRM_PLANE_TYPE_OVERLAY,
-	DRM_PLANE_TYPE_CURSOR,
+	[PRIMARY_WIN] = DRM_PLANE_TYPE_PRIMARY,
+	[CURSON_WIN] = DRM_PLANE_TYPE_CURSOR,
 };
 
 static inline void decon_set_bits(struct decon_context *ctx, u32 reg, u32 mask,
@@ -552,12 +557,10 @@ static int decon_bind(struct device *dev, struct device *master, void *data)
 	drm_dev->max_vblank_count = 0xffffffff;
 
 	for (win = ctx->first_win; win < WINDOWS_NR; win++) {
-		int tmp = (win == ctx->first_win) ? 0 : win;
-
 		ctx->configs[win].pixel_formats = decon_formats;
 		ctx->configs[win].num_pixel_formats = ARRAY_SIZE(decon_formats);
-		ctx->configs[win].zpos = win;
-		ctx->configs[win].type = decon_win_types[tmp];
+		ctx->configs[win].zpos = win - ctx->first_win;
+		ctx->configs[win].type = decon_win_types[win];
 
 		ret = exynos_plane_init(drm_dev, &ctx->planes[win], win,
 					&ctx->configs[win]);
@@ -565,7 +568,7 @@ static int decon_bind(struct device *dev, struct device *master, void *data)
 			return ret;
 	}
 
-	exynos_plane = &ctx->planes[ctx->first_win];
+	exynos_plane = &ctx->planes[PRIMARY_WIN];
 	out_type = (ctx->out_type & IFTYPE_HDMI) ? EXYNOS_DISPLAY_TYPE_HDMI
 						  : EXYNOS_DISPLAY_TYPE_LCD;
 	ctx->crtc = exynos_drm_crtc_create(drm_dev, &exynos_plane->base,
diff --git a/drivers/gpu/drm/exynos/exynos_dp.c b/drivers/gpu/drm/exynos/exynos_dp.c
index 86330f396784..af7ab1ceb50f 100644
--- a/drivers/gpu/drm/exynos/exynos_dp.c
+++ b/drivers/gpu/drm/exynos/exynos_dp.c
@@ -232,9 +232,11 @@ static int exynos_dp_probe(struct platform_device *pdev)
 	np = of_parse_phandle(dev->of_node, "panel", 0);
 	if (np) {
 		dp->plat_data.panel = of_drm_find_panel(np);
+
 		of_node_put(np);
-		if (!dp->plat_data.panel)
-			return -EPROBE_DEFER;
+		if (IS_ERR(dp->plat_data.panel))
+			return PTR_ERR(dp->plat_data.panel);
+
 		goto out;
 	}
 
diff --git a/drivers/gpu/drm/exynos/exynos_drm_crtc.c b/drivers/gpu/drm/exynos/exynos_drm_crtc.c
index dc01342e759a..eea90251808f 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_crtc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_crtc.c
@@ -228,7 +228,7 @@ struct exynos_drm_crtc *exynos_drm_crtc_get_by_type(struct drm_device *drm_dev,
 		if (to_exynos_crtc(crtc)->type == out_type)
 			return to_exynos_crtc(crtc);
 
-	return ERR_PTR(-EPERM);
+	return ERR_PTR(-ENODEV);
 }
 
 int exynos_drm_set_possible_crtcs(struct drm_encoder *encoder,
diff --git a/drivers/gpu/drm/exynos/exynos_drm_dpi.c b/drivers/gpu/drm/exynos/exynos_drm_dpi.c
index 66945e0dc57f..2f0babb67c51 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_dpi.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_dpi.c
@@ -113,7 +113,7 @@ static int exynos_dpi_create_connector(struct drm_encoder *encoder)
 	}
 
 	drm_connector_helper_add(connector, &exynos_dpi_connector_helper_funcs);
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	return 0;
 }
@@ -240,8 +240,8 @@ struct drm_encoder *exynos_dpi_probe(struct device *dev)
 
 	if (ctx->panel_node) {
 		ctx->panel = of_drm_find_panel(ctx->panel_node);
-		if (!ctx->panel)
-			return ERR_PTR(-EPROBE_DEFER);
+		if (IS_ERR(ctx->panel))
+			return ERR_CAST(ctx->panel);
 	}
 
 	return &ctx->encoder;
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.c b/drivers/gpu/drm/exynos/exynos_drm_drv.c
index 39284bb7c2c2..a81b4a5e24a7 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.c
@@ -27,15 +27,23 @@
 #include "exynos_drm_fb.h"
 #include "exynos_drm_gem.h"
 #include "exynos_drm_plane.h"
+#include "exynos_drm_ipp.h"
 #include "exynos_drm_vidi.h"
 #include "exynos_drm_g2d.h"
 #include "exynos_drm_iommu.h"
 
 #define DRIVER_NAME	"exynos"
 #define DRIVER_DESC	"Samsung SoC DRM"
-#define DRIVER_DATE	"20110530"
+#define DRIVER_DATE	"20180330"
+
+/*
+ * Interface history:
+ *
+ * 1.0 - Original version
+ * 1.1 - Upgrade IPP driver to version 2.0
+ */
 #define DRIVER_MAJOR	1
-#define DRIVER_MINOR	0
+#define DRIVER_MINOR	1
 
 static int exynos_drm_open(struct drm_device *dev, struct drm_file *file)
 {
@@ -88,6 +96,16 @@ static const struct drm_ioctl_desc exynos_ioctls[] = {
 			DRM_AUTH | DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(EXYNOS_G2D_EXEC, exynos_g2d_exec_ioctl,
 			DRM_AUTH | DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(EXYNOS_IPP_GET_RESOURCES,
+			exynos_drm_ipp_get_res_ioctl,
+			DRM_AUTH | DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(EXYNOS_IPP_GET_CAPS, exynos_drm_ipp_get_caps_ioctl,
+			DRM_AUTH | DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(EXYNOS_IPP_GET_LIMITS,
+			exynos_drm_ipp_get_limits_ioctl,
+			DRM_AUTH | DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(EXYNOS_IPP_COMMIT, exynos_drm_ipp_commit_ioctl,
+			DRM_AUTH | DRM_RENDER_ALLOW),
 };
 
 static const struct file_operations exynos_drm_driver_fops = {
@@ -184,6 +202,7 @@ struct exynos_drm_driver_info {
 #define DRM_COMPONENT_DRIVER	BIT(0)	/* supports component framework */
 #define DRM_VIRTUAL_DEVICE	BIT(1)	/* create virtual platform device */
 #define DRM_DMA_DEVICE		BIT(2)	/* can be used for dma allocations */
+#define DRM_FIMC_DEVICE		BIT(3)	/* devices shared with V4L2 subsystem */
 
 #define DRV_PTR(drv, cond) (IS_ENABLED(cond) ? &drv : NULL)
 
@@ -223,10 +242,16 @@ static struct exynos_drm_driver_info exynos_drm_drivers[] = {
 		DRV_PTR(g2d_driver, CONFIG_DRM_EXYNOS_G2D),
 	}, {
 		DRV_PTR(fimc_driver, CONFIG_DRM_EXYNOS_FIMC),
+		DRM_COMPONENT_DRIVER | DRM_FIMC_DEVICE,
 	}, {
 		DRV_PTR(rotator_driver, CONFIG_DRM_EXYNOS_ROTATOR),
+		DRM_COMPONENT_DRIVER
+	}, {
+		DRV_PTR(scaler_driver, CONFIG_DRM_EXYNOS_SCALER),
+		DRM_COMPONENT_DRIVER
 	}, {
 		DRV_PTR(gsc_driver, CONFIG_DRM_EXYNOS_GSC),
+		DRM_COMPONENT_DRIVER
 	}, {
 		&exynos_drm_platform_driver,
 		DRM_VIRTUAL_DEVICE
@@ -254,7 +279,11 @@ static struct component_match *exynos_drm_match_add(struct device *dev)
 					    &info->driver->driver,
 					    (void *)platform_bus_type.match))) {
 			put_device(p);
-			component_match_add(dev, &match, compare_dev, d);
+
+			if (!(info->flags & DRM_FIMC_DEVICE) ||
+			    exynos_drm_check_fimc_device(d) == 0)
+				component_match_add(dev, &match,
+						    compare_dev, d);
 			p = d;
 		}
 		put_device(p);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_drv.h b/drivers/gpu/drm/exynos/exynos_drm_drv.h
index 075957cb6ba1..0f6d079a55c9 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_drv.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_drv.h
@@ -273,6 +273,15 @@ static inline int exynos_dpi_bind(struct drm_device *dev,
 }
 #endif
 
+#ifdef CONFIG_DRM_EXYNOS_FIMC
+int exynos_drm_check_fimc_device(struct device *dev);
+#else
+static inline int exynos_drm_check_fimc_device(struct device *dev)
+{
+	return 0;
+}
+#endif
+
 int exynos_atomic_commit(struct drm_device *dev, struct drm_atomic_state *state,
 			 bool nonblock);
 
@@ -288,6 +297,7 @@ extern struct platform_driver vidi_driver;
 extern struct platform_driver g2d_driver;
 extern struct platform_driver fimc_driver;
 extern struct platform_driver rotator_driver;
+extern struct platform_driver scaler_driver;
 extern struct platform_driver gsc_driver;
 extern struct platform_driver mic_driver;
 #endif
diff --git a/drivers/gpu/drm/exynos/exynos_drm_dsi.c b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
index 7904ffa9abfb..a1ed6146a3b5 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_dsi.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_dsi.c
@@ -270,7 +270,6 @@ struct exynos_dsi {
 	u32 lanes;
 	u32 mode_flags;
 	u32 format;
-	struct videomode vm;
 
 	int state;
 	struct drm_property *brightness;
@@ -881,30 +880,30 @@ static int exynos_dsi_init_link(struct exynos_dsi *dsi)
 
 static void exynos_dsi_set_display_mode(struct exynos_dsi *dsi)
 {
-	struct videomode *vm = &dsi->vm;
+	struct drm_display_mode *m = &dsi->encoder.crtc->state->adjusted_mode;
 	unsigned int num_bits_resol = dsi->driver_data->num_bits_resol;
 	u32 reg;
 
 	if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO) {
 		reg = DSIM_CMD_ALLOW(0xf)
-			| DSIM_STABLE_VFP(vm->vfront_porch)
-			| DSIM_MAIN_VBP(vm->vback_porch);
+			| DSIM_STABLE_VFP(m->vsync_start - m->vdisplay)
+			| DSIM_MAIN_VBP(m->vtotal - m->vsync_end);
 		exynos_dsi_write(dsi, DSIM_MVPORCH_REG, reg);
 
-		reg = DSIM_MAIN_HFP(vm->hfront_porch)
-			| DSIM_MAIN_HBP(vm->hback_porch);
+		reg = DSIM_MAIN_HFP(m->hsync_start - m->hdisplay)
+			| DSIM_MAIN_HBP(m->htotal - m->hsync_end);
 		exynos_dsi_write(dsi, DSIM_MHPORCH_REG, reg);
 
-		reg = DSIM_MAIN_VSA(vm->vsync_len)
-			| DSIM_MAIN_HSA(vm->hsync_len);
+		reg = DSIM_MAIN_VSA(m->vsync_end - m->vsync_start)
+			| DSIM_MAIN_HSA(m->hsync_end - m->hsync_start);
 		exynos_dsi_write(dsi, DSIM_MSYNC_REG, reg);
 	}
-	reg =  DSIM_MAIN_HRESOL(vm->hactive, num_bits_resol) |
-		DSIM_MAIN_VRESOL(vm->vactive, num_bits_resol);
+	reg =  DSIM_MAIN_HRESOL(m->hdisplay, num_bits_resol) |
+		DSIM_MAIN_VRESOL(m->vdisplay, num_bits_resol);
 
 	exynos_dsi_write(dsi, DSIM_MDRESOL_REG, reg);
 
-	dev_dbg(dsi->dev, "LCD size = %dx%d\n", vm->hactive, vm->vactive);
+	dev_dbg(dsi->dev, "LCD size = %dx%d\n", m->hdisplay, m->vdisplay);
 }
 
 static void exynos_dsi_set_display_enable(struct exynos_dsi *dsi, bool enable)
@@ -1265,15 +1264,15 @@ static irqreturn_t exynos_dsi_irq(int irq, void *dev_id)
 
 	if (status & DSIM_INT_SW_RST_RELEASE) {
 		u32 mask = ~(DSIM_INT_RX_DONE | DSIM_INT_SFR_FIFO_EMPTY |
-			DSIM_INT_SFR_HDR_FIFO_EMPTY | DSIM_INT_FRAME_DONE |
-			DSIM_INT_RX_ECC_ERR | DSIM_INT_SW_RST_RELEASE);
+			DSIM_INT_SFR_HDR_FIFO_EMPTY | DSIM_INT_RX_ECC_ERR |
+			DSIM_INT_SW_RST_RELEASE);
 		exynos_dsi_write(dsi, DSIM_INTMSK_REG, mask);
 		complete(&dsi->completed);
 		return IRQ_HANDLED;
 	}
 
 	if (!(status & (DSIM_INT_RX_DONE | DSIM_INT_SFR_FIFO_EMPTY |
-			DSIM_INT_FRAME_DONE | DSIM_INT_PLL_STABLE)))
+			DSIM_INT_PLL_STABLE)))
 		return IRQ_HANDLED;
 
 	if (exynos_dsi_transfer_finish(dsi))
@@ -1480,31 +1479,12 @@ static int exynos_dsi_create_connector(struct drm_encoder *encoder)
 
 	connector->status = connector_status_disconnected;
 	drm_connector_helper_add(connector, &exynos_dsi_connector_helper_funcs);
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	return 0;
 }
 
-static void exynos_dsi_mode_set(struct drm_encoder *encoder,
-				struct drm_display_mode *mode,
-				struct drm_display_mode *adjusted_mode)
-{
-	struct exynos_dsi *dsi = encoder_to_dsi(encoder);
-	struct videomode *vm = &dsi->vm;
-	struct drm_display_mode *m = adjusted_mode;
-
-	vm->hactive = m->hdisplay;
-	vm->vactive = m->vdisplay;
-	vm->vfront_porch = m->vsync_start - m->vdisplay;
-	vm->vback_porch = m->vtotal - m->vsync_end;
-	vm->vsync_len = m->vsync_end - m->vsync_start;
-	vm->hfront_porch = m->hsync_start - m->hdisplay;
-	vm->hback_porch = m->htotal - m->hsync_end;
-	vm->hsync_len = m->hsync_end - m->hsync_start;
-}
-
 static const struct drm_encoder_helper_funcs exynos_dsi_encoder_helper_funcs = {
-	.mode_set = exynos_dsi_mode_set,
 	.enable = exynos_dsi_enable,
 	.disable = exynos_dsi_disable,
 };
@@ -1539,6 +1519,9 @@ static int exynos_dsi_host_attach(struct mipi_dsi_host *host,
 	dsi->format = device->format;
 	dsi->mode_flags = device->mode_flags;
 	dsi->panel = of_drm_find_panel(device->dev.of_node);
+	if (IS_ERR(dsi->panel))
+		dsi->panel = NULL;
+
 	if (dsi->panel) {
 		drm_panel_attach(dsi->panel, &dsi->connector);
 		dsi->connector.status = connector_status_connected;
@@ -1743,8 +1726,8 @@ static int exynos_dsi_probe(struct platform_device *pdev)
 		return -EPROBE_DEFER;
 	}
 
-	dsi->clks = devm_kzalloc(dev,
-			sizeof(*dsi->clks) * dsi->driver_data->num_clks,
+	dsi->clks = devm_kcalloc(dev,
+			dsi->driver_data->num_clks, sizeof(*dsi->clks),
 			GFP_KERNEL);
 	if (!dsi->clks)
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimc.c b/drivers/gpu/drm/exynos/exynos_drm_fimc.c
index 5b18b5c5fdf2..6127ef25acd6 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fimc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fimc.c
@@ -12,6 +12,7 @@
  *
  */
 #include <linux/kernel.h>
+#include <linux/component.h>
 #include <linux/platform_device.h>
 #include <linux/mfd/syscon.h>
 #include <linux/regmap.h>
@@ -24,8 +25,8 @@
 #include <drm/exynos_drm.h>
 #include "regs-fimc.h"
 #include "exynos_drm_drv.h"
+#include "exynos_drm_iommu.h"
 #include "exynos_drm_ipp.h"
-#include "exynos_drm_fimc.h"
 
 /*
  * FIMC stands for Fully Interactive Mobile Camera and
@@ -33,23 +34,6 @@
  * input DMA reads image data from the memory.
  * output DMA writes image data to memory.
  * FIMC supports image rotation and image effect functions.
- *
- * M2M operation : supports crop/scale/rotation/csc so on.
- * Memory ----> FIMC H/W ----> Memory.
- * Writeback operation : supports cloned screen with FIMD.
- * FIMD ----> FIMC H/W ----> Memory.
- * Output operation : supports direct display using local path.
- * Memory ----> FIMC H/W ----> FIMD.
- */
-
-/*
- * TODO
- * 1. check suspend/resume api if needed.
- * 2. need to check use case platform_device_id.
- * 3. check src/dst size with, height.
- * 4. added check_prepare api for right register.
- * 5. need to add supported list in prop_list.
- * 6. check prescaler/scaler optimization.
  */
 
 #define FIMC_MAX_DEVS	4
@@ -59,29 +43,19 @@
 #define FIMC_BUF_STOP	1
 #define FIMC_BUF_START	2
 #define FIMC_WIDTH_ITU_709	1280
-#define FIMC_REFRESH_MAX	60
-#define FIMC_REFRESH_MIN	12
-#define FIMC_CROP_MAX	8192
-#define FIMC_CROP_MIN	32
-#define FIMC_SCALE_MAX	4224
-#define FIMC_SCALE_MIN	32
+#define FIMC_AUTOSUSPEND_DELAY	2000
+
+static unsigned int fimc_mask = 0xc;
+module_param_named(fimc_devs, fimc_mask, uint, 0644);
+MODULE_PARM_DESC(fimc_devs, "Alias mask for assigning FIMC devices to Exynos DRM");
 
 #define get_fimc_context(dev)	platform_get_drvdata(to_platform_device(dev))
-#define get_ctx_from_ippdrv(ippdrv)	container_of(ippdrv,\
-					struct fimc_context, ippdrv);
-enum fimc_wb {
-	FIMC_WB_NONE,
-	FIMC_WB_A,
-	FIMC_WB_B,
-};
 
 enum {
 	FIMC_CLK_LCLK,
 	FIMC_CLK_GATE,
 	FIMC_CLK_WB_A,
 	FIMC_CLK_WB_B,
-	FIMC_CLK_MUX,
-	FIMC_CLK_PARENT,
 	FIMC_CLKS_MAX
 };
 
@@ -90,12 +64,8 @@ static const char * const fimc_clock_names[] = {
 	[FIMC_CLK_GATE]   = "fimc",
 	[FIMC_CLK_WB_A]   = "pxl_async0",
 	[FIMC_CLK_WB_B]   = "pxl_async1",
-	[FIMC_CLK_MUX]    = "mux",
-	[FIMC_CLK_PARENT] = "parent",
 };
 
-#define FIMC_DEFAULT_LCLK_FREQUENCY 133000000UL
-
 /*
  * A structure of scaler.
  *
@@ -107,7 +77,7 @@ static const char * const fimc_clock_names[] = {
  * @vratio: vertical ratio.
  */
 struct fimc_scaler {
-	bool	range;
+	bool range;
 	bool bypass;
 	bool up_h;
 	bool up_v;
@@ -116,56 +86,32 @@ struct fimc_scaler {
 };
 
 /*
- * A structure of scaler capability.
- *
- * find user manual table 43-1.
- * @in_hori: scaler input horizontal size.
- * @bypass: scaler bypass mode.
- * @dst_h_wo_rot: target horizontal size without output rotation.
- * @dst_h_rot: target horizontal size with output rotation.
- * @rl_w_wo_rot: real width without input rotation.
- * @rl_h_rot: real height without output rotation.
- */
-struct fimc_capability {
-	/* scaler */
-	u32	in_hori;
-	u32	bypass;
-	/* output rotator */
-	u32	dst_h_wo_rot;
-	u32	dst_h_rot;
-	/* input rotator */
-	u32	rl_w_wo_rot;
-	u32	rl_h_rot;
-};
-
-/*
  * A structure of fimc context.
  *
- * @ippdrv: prepare initialization using ippdrv.
  * @regs_res: register resources.
  * @regs: memory mapped io registers.
  * @lock: locking of operations.
  * @clocks: fimc clocks.
- * @clk_frequency: LCLK clock frequency.
- * @sysreg: handle to SYSREG block regmap.
  * @sc: scaler infomations.
  * @pol: porarity of writeback.
  * @id: fimc id.
  * @irq: irq number.
- * @suspended: qos operations.
  */
 struct fimc_context {
-	struct exynos_drm_ippdrv	ippdrv;
+	struct exynos_drm_ipp ipp;
+	struct drm_device *drm_dev;
+	struct device	*dev;
+	struct exynos_drm_ipp_task	*task;
+	struct exynos_drm_ipp_formats	*formats;
+	unsigned int			num_formats;
+
 	struct resource	*regs_res;
 	void __iomem	*regs;
 	spinlock_t	lock;
 	struct clk	*clocks[FIMC_CLKS_MAX];
-	u32		clk_frequency;
-	struct regmap	*sysreg;
 	struct fimc_scaler	sc;
 	int	id;
 	int	irq;
-	bool	suspended;
 };
 
 static u32 fimc_read(struct fimc_context *ctx, u32 reg)
@@ -217,19 +163,10 @@ static void fimc_sw_reset(struct fimc_context *ctx)
 	fimc_write(ctx, 0x0, EXYNOS_CIFCNTSEQ);
 }
 
-static int fimc_set_camblk_fimd0_wb(struct fimc_context *ctx)
-{
-	return regmap_update_bits(ctx->sysreg, SYSREG_CAMERA_BLK,
-				  SYSREG_FIMD0WB_DEST_MASK,
-				  ctx->id << SYSREG_FIMD0WB_DEST_SHIFT);
-}
-
-static void fimc_set_type_ctrl(struct fimc_context *ctx, enum fimc_wb wb)
+static void fimc_set_type_ctrl(struct fimc_context *ctx)
 {
 	u32 cfg;
 
-	DRM_DEBUG_KMS("wb[%d]\n", wb);
-
 	cfg = fimc_read(ctx, EXYNOS_CIGCTRL);
 	cfg &= ~(EXYNOS_CIGCTRL_TESTPATTERN_MASK |
 		EXYNOS_CIGCTRL_SELCAM_ITU_MASK |
@@ -238,23 +175,10 @@ static void fimc_set_type_ctrl(struct fimc_context *ctx, enum fimc_wb wb)
 		EXYNOS_CIGCTRL_SELWB_CAMIF_MASK |
 		EXYNOS_CIGCTRL_SELWRITEBACK_MASK);
 
-	switch (wb) {
-	case FIMC_WB_A:
-		cfg |= (EXYNOS_CIGCTRL_SELWRITEBACK_A |
-			EXYNOS_CIGCTRL_SELWB_CAMIF_WRITEBACK);
-		break;
-	case FIMC_WB_B:
-		cfg |= (EXYNOS_CIGCTRL_SELWRITEBACK_B |
-			EXYNOS_CIGCTRL_SELWB_CAMIF_WRITEBACK);
-		break;
-	case FIMC_WB_NONE:
-	default:
-		cfg |= (EXYNOS_CIGCTRL_SELCAM_ITU_A |
-			EXYNOS_CIGCTRL_SELWRITEBACK_A |
-			EXYNOS_CIGCTRL_SELCAM_MIPI_A |
-			EXYNOS_CIGCTRL_SELCAM_FIMC_ITU);
-		break;
-	}
+	cfg |= (EXYNOS_CIGCTRL_SELCAM_ITU_A |
+		EXYNOS_CIGCTRL_SELWRITEBACK_A |
+		EXYNOS_CIGCTRL_SELCAM_MIPI_A |
+		EXYNOS_CIGCTRL_SELCAM_FIMC_ITU);
 
 	fimc_write(ctx, cfg, EXYNOS_CIGCTRL);
 }
@@ -296,7 +220,6 @@ static void fimc_clear_irq(struct fimc_context *ctx)
 
 static bool fimc_check_ovf(struct fimc_context *ctx)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 status, flag;
 
 	status = fimc_read(ctx, EXYNOS_CISTATUS);
@@ -310,7 +233,7 @@ static bool fimc_check_ovf(struct fimc_context *ctx)
 			EXYNOS_CIWDOFST_CLROVFIY | EXYNOS_CIWDOFST_CLROVFICB |
 			EXYNOS_CIWDOFST_CLROVFICR);
 
-		dev_err(ippdrv->dev, "occurred overflow at %d, status 0x%x.\n",
+		dev_err(ctx->dev, "occurred overflow at %d, status 0x%x.\n",
 			ctx->id, status);
 		return true;
 	}
@@ -376,10 +299,8 @@ static void fimc_handle_lastend(struct fimc_context *ctx, bool enable)
 	fimc_write(ctx, cfg, EXYNOS_CIOCTRL);
 }
 
-
-static int fimc_src_set_fmt_order(struct fimc_context *ctx, u32 fmt)
+static void fimc_src_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg;
 
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
@@ -392,12 +313,12 @@ static int fimc_src_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 	case DRM_FORMAT_RGB565:
 		cfg |= EXYNOS_CISCCTRL_INRGB_FMT_RGB565;
 		fimc_write(ctx, cfg, EXYNOS_CISCCTRL);
-		return 0;
+		return;
 	case DRM_FORMAT_RGB888:
 	case DRM_FORMAT_XRGB8888:
 		cfg |= EXYNOS_CISCCTRL_INRGB_FMT_RGB888;
 		fimc_write(ctx, cfg, EXYNOS_CISCCTRL);
-		return 0;
+		return;
 	default:
 		/* bypass */
 		break;
@@ -438,20 +359,13 @@ static int fimc_src_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 		cfg |= (EXYNOS_MSCTRL_ORDER2P_LSB_CBCR |
 			EXYNOS_MSCTRL_C_INT_IN_2PLANE);
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid source yuv order 0x%x.\n", fmt);
-		return -EINVAL;
 	}
 
 	fimc_write(ctx, cfg, EXYNOS_MSCTRL);
-
-	return 0;
 }
 
-static int fimc_src_set_fmt(struct device *dev, u32 fmt)
+static void fimc_src_set_fmt(struct fimc_context *ctx, u32 fmt, bool tiled)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg;
 
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
@@ -485,9 +399,6 @@ static int fimc_src_set_fmt(struct device *dev, u32 fmt)
 	case DRM_FORMAT_NV21:
 		cfg |= EXYNOS_MSCTRL_INFORMAT_YCBCR420;
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid source format 0x%x.\n", fmt);
-		return -EINVAL;
 	}
 
 	fimc_write(ctx, cfg, EXYNOS_MSCTRL);
@@ -495,22 +406,22 @@ static int fimc_src_set_fmt(struct device *dev, u32 fmt)
 	cfg = fimc_read(ctx, EXYNOS_CIDMAPARAM);
 	cfg &= ~EXYNOS_CIDMAPARAM_R_MODE_MASK;
 
-	cfg |= EXYNOS_CIDMAPARAM_R_MODE_LINEAR;
+	if (tiled)
+		cfg |= EXYNOS_CIDMAPARAM_R_MODE_64X32;
+	else
+		cfg |= EXYNOS_CIDMAPARAM_R_MODE_LINEAR;
 
 	fimc_write(ctx, cfg, EXYNOS_CIDMAPARAM);
 
-	return fimc_src_set_fmt_order(ctx, fmt);
+	fimc_src_set_fmt_order(ctx, fmt);
 }
 
-static int fimc_src_set_transf(struct device *dev,
-		enum drm_exynos_degree degree,
-		enum drm_exynos_flip flip, bool *swap)
+static void fimc_src_set_transf(struct fimc_context *ctx, unsigned int rotation)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
+	unsigned int degree = rotation & DRM_MODE_ROTATE_MASK;
 	u32 cfg1, cfg2;
 
-	DRM_DEBUG_KMS("degree[%d]flip[0x%x]\n", degree, flip);
+	DRM_DEBUG_KMS("rotation[%x]\n", rotation);
 
 	cfg1 = fimc_read(ctx, EXYNOS_MSCTRL);
 	cfg1 &= ~(EXYNOS_MSCTRL_FLIP_X_MIRROR |
@@ -520,61 +431,56 @@ static int fimc_src_set_transf(struct device *dev,
 	cfg2 &= ~EXYNOS_CITRGFMT_INROT90_CLOCKWISE;
 
 	switch (degree) {
-	case EXYNOS_DRM_DEGREE_0:
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+	case DRM_MODE_ROTATE_0:
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg1 |= EXYNOS_MSCTRL_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg1 |= EXYNOS_MSCTRL_FLIP_Y_MIRROR;
 		break;
-	case EXYNOS_DRM_DEGREE_90:
+	case DRM_MODE_ROTATE_90:
 		cfg2 |= EXYNOS_CITRGFMT_INROT90_CLOCKWISE;
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg1 |= EXYNOS_MSCTRL_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg1 |= EXYNOS_MSCTRL_FLIP_Y_MIRROR;
 		break;
-	case EXYNOS_DRM_DEGREE_180:
+	case DRM_MODE_ROTATE_180:
 		cfg1 |= (EXYNOS_MSCTRL_FLIP_X_MIRROR |
 			EXYNOS_MSCTRL_FLIP_Y_MIRROR);
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg1 &= ~EXYNOS_MSCTRL_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg1 &= ~EXYNOS_MSCTRL_FLIP_Y_MIRROR;
 		break;
-	case EXYNOS_DRM_DEGREE_270:
+	case DRM_MODE_ROTATE_270:
 		cfg1 |= (EXYNOS_MSCTRL_FLIP_X_MIRROR |
 			EXYNOS_MSCTRL_FLIP_Y_MIRROR);
 		cfg2 |= EXYNOS_CITRGFMT_INROT90_CLOCKWISE;
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg1 &= ~EXYNOS_MSCTRL_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg1 &= ~EXYNOS_MSCTRL_FLIP_Y_MIRROR;
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid degree value %d.\n", degree);
-		return -EINVAL;
 	}
 
 	fimc_write(ctx, cfg1, EXYNOS_MSCTRL);
 	fimc_write(ctx, cfg2, EXYNOS_CITRGFMT);
-	*swap = (cfg2 & EXYNOS_CITRGFMT_INROT90_CLOCKWISE) ? 1 : 0;
-
-	return 0;
 }
 
-static int fimc_set_window(struct fimc_context *ctx,
-		struct drm_exynos_pos *pos, struct drm_exynos_sz *sz)
+static void fimc_set_window(struct fimc_context *ctx,
+			    struct exynos_drm_ipp_buffer *buf)
 {
 	u32 cfg, h1, h2, v1, v2;
 
 	/* cropped image */
-	h1 = pos->x;
-	h2 = sz->hsize - pos->w - pos->x;
-	v1 = pos->y;
-	v2 = sz->vsize - pos->h - pos->y;
+	h1 = buf->rect.x;
+	h2 = buf->buf.width - buf->rect.w - buf->rect.x;
+	v1 = buf->rect.y;
+	v2 = buf->buf.height - buf->rect.h - buf->rect.y;
 
 	DRM_DEBUG_KMS("x[%d]y[%d]w[%d]h[%d]hsize[%d]vsize[%d]\n",
-		pos->x, pos->y, pos->w, pos->h, sz->hsize, sz->vsize);
+		buf->rect.x, buf->rect.y, buf->rect.w, buf->rect.h,
+		buf->buf.width, buf->buf.height);
 	DRM_DEBUG_KMS("h1[%d]h2[%d]v1[%d]v2[%d]\n", h1, h2, v1, v2);
 
 	/*
@@ -592,42 +498,30 @@ static int fimc_set_window(struct fimc_context *ctx,
 	cfg = (EXYNOS_CIWDOFST2_WINHOROFST2(h2) |
 		EXYNOS_CIWDOFST2_WINVEROFST2(v2));
 	fimc_write(ctx, cfg, EXYNOS_CIWDOFST2);
-
-	return 0;
 }
 
-static int fimc_src_set_size(struct device *dev, int swap,
-		struct drm_exynos_pos *pos, struct drm_exynos_sz *sz)
+static void fimc_src_set_size(struct fimc_context *ctx,
+			      struct exynos_drm_ipp_buffer *buf)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct drm_exynos_pos img_pos = *pos;
-	struct drm_exynos_sz img_sz = *sz;
 	u32 cfg;
 
-	DRM_DEBUG_KMS("swap[%d]hsize[%d]vsize[%d]\n",
-		swap, sz->hsize, sz->vsize);
+	DRM_DEBUG_KMS("hsize[%d]vsize[%d]\n", buf->buf.width, buf->buf.height);
 
 	/* original size */
-	cfg = (EXYNOS_ORGISIZE_HORIZONTAL(img_sz.hsize) |
-		EXYNOS_ORGISIZE_VERTICAL(img_sz.vsize));
+	cfg = (EXYNOS_ORGISIZE_HORIZONTAL(buf->buf.width) |
+		EXYNOS_ORGISIZE_VERTICAL(buf->buf.height));
 
 	fimc_write(ctx, cfg, EXYNOS_ORGISIZE);
 
-	DRM_DEBUG_KMS("x[%d]y[%d]w[%d]h[%d]\n", pos->x, pos->y, pos->w, pos->h);
-
-	if (swap) {
-		img_pos.w = pos->h;
-		img_pos.h = pos->w;
-		img_sz.hsize = sz->vsize;
-		img_sz.vsize = sz->hsize;
-	}
+	DRM_DEBUG_KMS("x[%d]y[%d]w[%d]h[%d]\n", buf->rect.x, buf->rect.y,
+		buf->rect.w, buf->rect.h);
 
 	/* set input DMA image size */
 	cfg = fimc_read(ctx, EXYNOS_CIREAL_ISIZE);
 	cfg &= ~(EXYNOS_CIREAL_ISIZE_HEIGHT_MASK |
 		EXYNOS_CIREAL_ISIZE_WIDTH_MASK);
-	cfg |= (EXYNOS_CIREAL_ISIZE_WIDTH(img_pos.w) |
-		EXYNOS_CIREAL_ISIZE_HEIGHT(img_pos.h));
+	cfg |= (EXYNOS_CIREAL_ISIZE_WIDTH(buf->rect.w) |
+		EXYNOS_CIREAL_ISIZE_HEIGHT(buf->rect.h));
 	fimc_write(ctx, cfg, EXYNOS_CIREAL_ISIZE);
 
 	/*
@@ -635,91 +529,34 @@ static int fimc_src_set_size(struct device *dev, int swap,
 	 * for now, we support only ITU601 8 bit mode
 	 */
 	cfg = (EXYNOS_CISRCFMT_ITU601_8BIT |
-		EXYNOS_CISRCFMT_SOURCEHSIZE(img_sz.hsize) |
-		EXYNOS_CISRCFMT_SOURCEVSIZE(img_sz.vsize));
+		EXYNOS_CISRCFMT_SOURCEHSIZE(buf->buf.width) |
+		EXYNOS_CISRCFMT_SOURCEVSIZE(buf->buf.height));
 	fimc_write(ctx, cfg, EXYNOS_CISRCFMT);
 
 	/* offset Y(RGB), Cb, Cr */
-	cfg = (EXYNOS_CIIYOFF_HORIZONTAL(img_pos.x) |
-		EXYNOS_CIIYOFF_VERTICAL(img_pos.y));
+	cfg = (EXYNOS_CIIYOFF_HORIZONTAL(buf->rect.x) |
+		EXYNOS_CIIYOFF_VERTICAL(buf->rect.y));
 	fimc_write(ctx, cfg, EXYNOS_CIIYOFF);
-	cfg = (EXYNOS_CIICBOFF_HORIZONTAL(img_pos.x) |
-		EXYNOS_CIICBOFF_VERTICAL(img_pos.y));
+	cfg = (EXYNOS_CIICBOFF_HORIZONTAL(buf->rect.x) |
+		EXYNOS_CIICBOFF_VERTICAL(buf->rect.y));
 	fimc_write(ctx, cfg, EXYNOS_CIICBOFF);
-	cfg = (EXYNOS_CIICROFF_HORIZONTAL(img_pos.x) |
-		EXYNOS_CIICROFF_VERTICAL(img_pos.y));
+	cfg = (EXYNOS_CIICROFF_HORIZONTAL(buf->rect.x) |
+		EXYNOS_CIICROFF_VERTICAL(buf->rect.y));
 	fimc_write(ctx, cfg, EXYNOS_CIICROFF);
 
-	return fimc_set_window(ctx, &img_pos, &img_sz);
+	fimc_set_window(ctx, buf);
 }
 
-static int fimc_src_set_addr(struct device *dev,
-		struct drm_exynos_ipp_buf_info *buf_info, u32 buf_id,
-		enum drm_exynos_ipp_buf_type buf_type)
+static void fimc_src_set_addr(struct fimc_context *ctx,
+			      struct exynos_drm_ipp_buffer *buf)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_property *property;
-	struct drm_exynos_ipp_config *config;
-
-	if (!c_node) {
-		DRM_ERROR("failed to get c_node.\n");
-		return -EINVAL;
-	}
-
-	property = &c_node->property;
-
-	DRM_DEBUG_KMS("prop_id[%d]buf_id[%d]buf_type[%d]\n",
-		property->prop_id, buf_id, buf_type);
-
-	if (buf_id > FIMC_MAX_SRC) {
-		dev_info(ippdrv->dev, "invalid buf_id %d.\n", buf_id);
-		return -ENOMEM;
-	}
-
-	/* address register set */
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		config = &property->config[EXYNOS_DRM_OPS_SRC];
-		fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_Y],
-			EXYNOS_CIIYSA0);
-
-		if (config->fmt == DRM_FORMAT_YVU420) {
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
-				EXYNOS_CIICBSA0);
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
-				EXYNOS_CIICRSA0);
-		} else {
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
-				EXYNOS_CIICBSA0);
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
-				EXYNOS_CIICRSA0);
-		}
-		break;
-	case IPP_BUF_DEQUEUE:
-		fimc_write(ctx, 0x0, EXYNOS_CIIYSA0);
-		fimc_write(ctx, 0x0, EXYNOS_CIICBSA0);
-		fimc_write(ctx, 0x0, EXYNOS_CIICRSA0);
-		break;
-	default:
-		/* bypass */
-		break;
-	}
-
-	return 0;
+	fimc_write(ctx, buf->dma_addr[0], EXYNOS_CIIYSA(0));
+	fimc_write(ctx, buf->dma_addr[1], EXYNOS_CIICBSA(0));
+	fimc_write(ctx, buf->dma_addr[2], EXYNOS_CIICRSA(0));
 }
 
-static struct exynos_drm_ipp_ops fimc_src_ops = {
-	.set_fmt = fimc_src_set_fmt,
-	.set_transf = fimc_src_set_transf,
-	.set_size = fimc_src_set_size,
-	.set_addr = fimc_src_set_addr,
-};
-
-static int fimc_dst_set_fmt_order(struct fimc_context *ctx, u32 fmt)
+static void fimc_dst_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg;
 
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
@@ -732,11 +569,11 @@ static int fimc_dst_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 	case DRM_FORMAT_RGB565:
 		cfg |= EXYNOS_CISCCTRL_OUTRGB_FMT_RGB565;
 		fimc_write(ctx, cfg, EXYNOS_CISCCTRL);
-		return 0;
+		return;
 	case DRM_FORMAT_RGB888:
 		cfg |= EXYNOS_CISCCTRL_OUTRGB_FMT_RGB888;
 		fimc_write(ctx, cfg, EXYNOS_CISCCTRL);
-		return 0;
+		return;
 	case DRM_FORMAT_XRGB8888:
 		cfg |= (EXYNOS_CISCCTRL_OUTRGB_FMT_RGB888 |
 			EXYNOS_CISCCTRL_EXTRGB_EXTENSION);
@@ -784,20 +621,13 @@ static int fimc_dst_set_fmt_order(struct fimc_context *ctx, u32 fmt)
 		cfg |= EXYNOS_CIOCTRL_ORDER2P_LSB_CBCR;
 		cfg |= EXYNOS_CIOCTRL_YCBCR_2PLANE;
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid target yuv order 0x%x.\n", fmt);
-		return -EINVAL;
 	}
 
 	fimc_write(ctx, cfg, EXYNOS_CIOCTRL);
-
-	return 0;
 }
 
-static int fimc_dst_set_fmt(struct device *dev, u32 fmt)
+static void fimc_dst_set_fmt(struct fimc_context *ctx, u32 fmt, bool tiled)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg;
 
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
@@ -837,10 +667,6 @@ static int fimc_dst_set_fmt(struct device *dev, u32 fmt)
 		case DRM_FORMAT_NV21:
 			cfg |= EXYNOS_CITRGFMT_OUTFORMAT_YCBCR420;
 			break;
-		default:
-			dev_err(ippdrv->dev, "invalid target format 0x%x.\n",
-				fmt);
-			return -EINVAL;
 		}
 
 		fimc_write(ctx, cfg, EXYNOS_CITRGFMT);
@@ -849,73 +675,67 @@ static int fimc_dst_set_fmt(struct device *dev, u32 fmt)
 	cfg = fimc_read(ctx, EXYNOS_CIDMAPARAM);
 	cfg &= ~EXYNOS_CIDMAPARAM_W_MODE_MASK;
 
-	cfg |= EXYNOS_CIDMAPARAM_W_MODE_LINEAR;
+	if (tiled)
+		cfg |= EXYNOS_CIDMAPARAM_W_MODE_64X32;
+	else
+		cfg |= EXYNOS_CIDMAPARAM_W_MODE_LINEAR;
 
 	fimc_write(ctx, cfg, EXYNOS_CIDMAPARAM);
 
-	return fimc_dst_set_fmt_order(ctx, fmt);
+	fimc_dst_set_fmt_order(ctx, fmt);
 }
 
-static int fimc_dst_set_transf(struct device *dev,
-		enum drm_exynos_degree degree,
-		enum drm_exynos_flip flip, bool *swap)
+static void fimc_dst_set_transf(struct fimc_context *ctx, unsigned int rotation)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
+	unsigned int degree = rotation & DRM_MODE_ROTATE_MASK;
 	u32 cfg;
 
-	DRM_DEBUG_KMS("degree[%d]flip[0x%x]\n", degree, flip);
+	DRM_DEBUG_KMS("rotation[0x%x]\n", rotation);
 
 	cfg = fimc_read(ctx, EXYNOS_CITRGFMT);
 	cfg &= ~EXYNOS_CITRGFMT_FLIP_MASK;
 	cfg &= ~EXYNOS_CITRGFMT_OUTROT90_CLOCKWISE;
 
 	switch (degree) {
-	case EXYNOS_DRM_DEGREE_0:
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+	case DRM_MODE_ROTATE_0:
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg |= EXYNOS_CITRGFMT_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg |= EXYNOS_CITRGFMT_FLIP_Y_MIRROR;
 		break;
-	case EXYNOS_DRM_DEGREE_90:
+	case DRM_MODE_ROTATE_90:
 		cfg |= EXYNOS_CITRGFMT_OUTROT90_CLOCKWISE;
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg |= EXYNOS_CITRGFMT_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg |= EXYNOS_CITRGFMT_FLIP_Y_MIRROR;
 		break;
-	case EXYNOS_DRM_DEGREE_180:
+	case DRM_MODE_ROTATE_180:
 		cfg |= (EXYNOS_CITRGFMT_FLIP_X_MIRROR |
 			EXYNOS_CITRGFMT_FLIP_Y_MIRROR);
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg &= ~EXYNOS_CITRGFMT_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg &= ~EXYNOS_CITRGFMT_FLIP_Y_MIRROR;
 		break;
-	case EXYNOS_DRM_DEGREE_270:
+	case DRM_MODE_ROTATE_270:
 		cfg |= (EXYNOS_CITRGFMT_OUTROT90_CLOCKWISE |
 			EXYNOS_CITRGFMT_FLIP_X_MIRROR |
 			EXYNOS_CITRGFMT_FLIP_Y_MIRROR);
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg &= ~EXYNOS_CITRGFMT_FLIP_X_MIRROR;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg &= ~EXYNOS_CITRGFMT_FLIP_Y_MIRROR;
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid degree value %d.\n", degree);
-		return -EINVAL;
 	}
 
 	fimc_write(ctx, cfg, EXYNOS_CITRGFMT);
-	*swap = (cfg & EXYNOS_CITRGFMT_OUTROT90_CLOCKWISE) ? 1 : 0;
-
-	return 0;
 }
 
 static int fimc_set_prescaler(struct fimc_context *ctx, struct fimc_scaler *sc,
-		struct drm_exynos_pos *src, struct drm_exynos_pos *dst)
+			      struct drm_exynos_ipp_task_rect *src,
+			      struct drm_exynos_ipp_task_rect *dst)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg, cfg_ext, shfactor;
 	u32 pre_dst_width, pre_dst_height;
 	u32 hfactor, vfactor;
@@ -942,13 +762,13 @@ static int fimc_set_prescaler(struct fimc_context *ctx, struct fimc_scaler *sc,
 	/* fimc_ippdrv_check_property assures that dividers are not null */
 	hfactor = fls(src_w / dst_w / 2);
 	if (hfactor > FIMC_SHFACTOR / 2) {
-		dev_err(ippdrv->dev, "failed to get ratio horizontal.\n");
+		dev_err(ctx->dev, "failed to get ratio horizontal.\n");
 		return -EINVAL;
 	}
 
 	vfactor = fls(src_h / dst_h / 2);
 	if (vfactor > FIMC_SHFACTOR / 2) {
-		dev_err(ippdrv->dev, "failed to get ratio vertical.\n");
+		dev_err(ctx->dev, "failed to get ratio vertical.\n");
 		return -EINVAL;
 	}
 
@@ -1019,83 +839,77 @@ static void fimc_set_scaler(struct fimc_context *ctx, struct fimc_scaler *sc)
 	fimc_write(ctx, cfg_ext, EXYNOS_CIEXTEN);
 }
 
-static int fimc_dst_set_size(struct device *dev, int swap,
-		struct drm_exynos_pos *pos, struct drm_exynos_sz *sz)
+static void fimc_dst_set_size(struct fimc_context *ctx,
+			     struct exynos_drm_ipp_buffer *buf)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct drm_exynos_pos img_pos = *pos;
-	struct drm_exynos_sz img_sz = *sz;
-	u32 cfg;
+	u32 cfg, cfg_ext;
 
-	DRM_DEBUG_KMS("swap[%d]hsize[%d]vsize[%d]\n",
-		swap, sz->hsize, sz->vsize);
+	DRM_DEBUG_KMS("hsize[%d]vsize[%d]\n", buf->buf.width, buf->buf.height);
 
 	/* original size */
-	cfg = (EXYNOS_ORGOSIZE_HORIZONTAL(img_sz.hsize) |
-		EXYNOS_ORGOSIZE_VERTICAL(img_sz.vsize));
+	cfg = (EXYNOS_ORGOSIZE_HORIZONTAL(buf->buf.width) |
+		EXYNOS_ORGOSIZE_VERTICAL(buf->buf.height));
 
 	fimc_write(ctx, cfg, EXYNOS_ORGOSIZE);
 
-	DRM_DEBUG_KMS("x[%d]y[%d]w[%d]h[%d]\n", pos->x, pos->y, pos->w, pos->h);
+	DRM_DEBUG_KMS("x[%d]y[%d]w[%d]h[%d]\n", buf->rect.x, buf->rect.y,
+		buf->rect.w, buf->rect.h);
 
 	/* CSC ITU */
 	cfg = fimc_read(ctx, EXYNOS_CIGCTRL);
 	cfg &= ~EXYNOS_CIGCTRL_CSC_MASK;
 
-	if (sz->hsize >= FIMC_WIDTH_ITU_709)
+	if (buf->buf.width >= FIMC_WIDTH_ITU_709)
 		cfg |= EXYNOS_CIGCTRL_CSC_ITU709;
 	else
 		cfg |= EXYNOS_CIGCTRL_CSC_ITU601;
 
 	fimc_write(ctx, cfg, EXYNOS_CIGCTRL);
 
-	if (swap) {
-		img_pos.w = pos->h;
-		img_pos.h = pos->w;
-		img_sz.hsize = sz->vsize;
-		img_sz.vsize = sz->hsize;
-	}
+	cfg_ext = fimc_read(ctx, EXYNOS_CITRGFMT);
 
 	/* target image size */
 	cfg = fimc_read(ctx, EXYNOS_CITRGFMT);
 	cfg &= ~(EXYNOS_CITRGFMT_TARGETH_MASK |
 		EXYNOS_CITRGFMT_TARGETV_MASK);
-	cfg |= (EXYNOS_CITRGFMT_TARGETHSIZE(img_pos.w) |
-		EXYNOS_CITRGFMT_TARGETVSIZE(img_pos.h));
+	if (cfg_ext & EXYNOS_CITRGFMT_OUTROT90_CLOCKWISE)
+		cfg |= (EXYNOS_CITRGFMT_TARGETHSIZE(buf->rect.h) |
+			EXYNOS_CITRGFMT_TARGETVSIZE(buf->rect.w));
+	else
+		cfg |= (EXYNOS_CITRGFMT_TARGETHSIZE(buf->rect.w) |
+			EXYNOS_CITRGFMT_TARGETVSIZE(buf->rect.h));
 	fimc_write(ctx, cfg, EXYNOS_CITRGFMT);
 
 	/* target area */
-	cfg = EXYNOS_CITAREA_TARGET_AREA(img_pos.w * img_pos.h);
+	cfg = EXYNOS_CITAREA_TARGET_AREA(buf->rect.w * buf->rect.h);
 	fimc_write(ctx, cfg, EXYNOS_CITAREA);
 
 	/* offset Y(RGB), Cb, Cr */
-	cfg = (EXYNOS_CIOYOFF_HORIZONTAL(img_pos.x) |
-		EXYNOS_CIOYOFF_VERTICAL(img_pos.y));
+	cfg = (EXYNOS_CIOYOFF_HORIZONTAL(buf->rect.x) |
+		EXYNOS_CIOYOFF_VERTICAL(buf->rect.y));
 	fimc_write(ctx, cfg, EXYNOS_CIOYOFF);
-	cfg = (EXYNOS_CIOCBOFF_HORIZONTAL(img_pos.x) |
-		EXYNOS_CIOCBOFF_VERTICAL(img_pos.y));
+	cfg = (EXYNOS_CIOCBOFF_HORIZONTAL(buf->rect.x) |
+		EXYNOS_CIOCBOFF_VERTICAL(buf->rect.y));
 	fimc_write(ctx, cfg, EXYNOS_CIOCBOFF);
-	cfg = (EXYNOS_CIOCROFF_HORIZONTAL(img_pos.x) |
-		EXYNOS_CIOCROFF_VERTICAL(img_pos.y));
+	cfg = (EXYNOS_CIOCROFF_HORIZONTAL(buf->rect.x) |
+		EXYNOS_CIOCROFF_VERTICAL(buf->rect.y));
 	fimc_write(ctx, cfg, EXYNOS_CIOCROFF);
-
-	return 0;
 }
 
 static void fimc_dst_set_buf_seq(struct fimc_context *ctx, u32 buf_id,
-		enum drm_exynos_ipp_buf_type buf_type)
+		bool enqueue)
 {
 	unsigned long flags;
 	u32 buf_num;
 	u32 cfg;
 
-	DRM_DEBUG_KMS("buf_id[%d]buf_type[%d]\n", buf_id, buf_type);
+	DRM_DEBUG_KMS("buf_id[%d]enqueu[%d]\n", buf_id, enqueue);
 
 	spin_lock_irqsave(&ctx->lock, flags);
 
 	cfg = fimc_read(ctx, EXYNOS_CIFCNTSEQ);
 
-	if (buf_type == IPP_BUF_ENQUEUE)
+	if (enqueue)
 		cfg |= (1 << buf_id);
 	else
 		cfg &= ~(1 << buf_id);
@@ -1104,88 +918,29 @@ static void fimc_dst_set_buf_seq(struct fimc_context *ctx, u32 buf_id,
 
 	buf_num = hweight32(cfg);
 
-	if (buf_type == IPP_BUF_ENQUEUE && buf_num >= FIMC_BUF_START)
+	if (enqueue && buf_num >= FIMC_BUF_START)
 		fimc_mask_irq(ctx, true);
-	else if (buf_type == IPP_BUF_DEQUEUE && buf_num <= FIMC_BUF_STOP)
+	else if (!enqueue && buf_num <= FIMC_BUF_STOP)
 		fimc_mask_irq(ctx, false);
 
 	spin_unlock_irqrestore(&ctx->lock, flags);
 }
 
-static int fimc_dst_set_addr(struct device *dev,
-		struct drm_exynos_ipp_buf_info *buf_info, u32 buf_id,
-		enum drm_exynos_ipp_buf_type buf_type)
+static void fimc_dst_set_addr(struct fimc_context *ctx,
+			     struct exynos_drm_ipp_buffer *buf)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_property *property;
-	struct drm_exynos_ipp_config *config;
-
-	if (!c_node) {
-		DRM_ERROR("failed to get c_node.\n");
-		return -EINVAL;
-	}
-
-	property = &c_node->property;
-
-	DRM_DEBUG_KMS("prop_id[%d]buf_id[%d]buf_type[%d]\n",
-		property->prop_id, buf_id, buf_type);
+	fimc_write(ctx, buf->dma_addr[0], EXYNOS_CIOYSA(0));
+	fimc_write(ctx, buf->dma_addr[1], EXYNOS_CIOCBSA(0));
+	fimc_write(ctx, buf->dma_addr[2], EXYNOS_CIOCRSA(0));
 
-	if (buf_id > FIMC_MAX_DST) {
-		dev_info(ippdrv->dev, "invalid buf_id %d.\n", buf_id);
-		return -ENOMEM;
-	}
-
-	/* address register set */
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		config = &property->config[EXYNOS_DRM_OPS_DST];
-
-		fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_Y],
-			EXYNOS_CIOYSA(buf_id));
-
-		if (config->fmt == DRM_FORMAT_YVU420) {
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
-				EXYNOS_CIOCBSA(buf_id));
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
-				EXYNOS_CIOCRSA(buf_id));
-		} else {
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CB],
-				EXYNOS_CIOCBSA(buf_id));
-			fimc_write(ctx, buf_info->base[EXYNOS_DRM_PLANAR_CR],
-				EXYNOS_CIOCRSA(buf_id));
-		}
-		break;
-	case IPP_BUF_DEQUEUE:
-		fimc_write(ctx, 0x0, EXYNOS_CIOYSA(buf_id));
-		fimc_write(ctx, 0x0, EXYNOS_CIOCBSA(buf_id));
-		fimc_write(ctx, 0x0, EXYNOS_CIOCRSA(buf_id));
-		break;
-	default:
-		/* bypass */
-		break;
-	}
-
-	fimc_dst_set_buf_seq(ctx, buf_id, buf_type);
-
-	return 0;
+	fimc_dst_set_buf_seq(ctx, 0, true);
 }
 
-static struct exynos_drm_ipp_ops fimc_dst_ops = {
-	.set_fmt = fimc_dst_set_fmt,
-	.set_transf = fimc_dst_set_transf,
-	.set_size = fimc_dst_set_size,
-	.set_addr = fimc_dst_set_addr,
-};
+static void fimc_stop(struct fimc_context *ctx);
 
 static irqreturn_t fimc_irq_handler(int irq, void *dev_id)
 {
 	struct fimc_context *ctx = dev_id;
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_event_work *event_work =
-		c_node->event_work;
 	int buf_id;
 
 	DRM_DEBUG_KMS("fimc id[%d]\n", ctx->id);
@@ -1203,170 +958,19 @@ static irqreturn_t fimc_irq_handler(int irq, void *dev_id)
 
 	DRM_DEBUG_KMS("buf_id[%d]\n", buf_id);
 
-	fimc_dst_set_buf_seq(ctx, buf_id, IPP_BUF_DEQUEUE);
-
-	event_work->ippdrv = ippdrv;
-	event_work->buf_id[EXYNOS_DRM_OPS_DST] = buf_id;
-	queue_work(ippdrv->event_workq, &event_work->work);
-
-	return IRQ_HANDLED;
-}
-
-static int fimc_init_prop_list(struct exynos_drm_ippdrv *ippdrv)
-{
-	struct drm_exynos_ipp_prop_list *prop_list = &ippdrv->prop_list;
-
-	prop_list->version = 1;
-	prop_list->writeback = 1;
-	prop_list->refresh_min = FIMC_REFRESH_MIN;
-	prop_list->refresh_max = FIMC_REFRESH_MAX;
-	prop_list->flip = (1 << EXYNOS_DRM_FLIP_NONE) |
-				(1 << EXYNOS_DRM_FLIP_VERTICAL) |
-				(1 << EXYNOS_DRM_FLIP_HORIZONTAL);
-	prop_list->degree = (1 << EXYNOS_DRM_DEGREE_0) |
-				(1 << EXYNOS_DRM_DEGREE_90) |
-				(1 << EXYNOS_DRM_DEGREE_180) |
-				(1 << EXYNOS_DRM_DEGREE_270);
-	prop_list->csc = 1;
-	prop_list->crop = 1;
-	prop_list->crop_max.hsize = FIMC_CROP_MAX;
-	prop_list->crop_max.vsize = FIMC_CROP_MAX;
-	prop_list->crop_min.hsize = FIMC_CROP_MIN;
-	prop_list->crop_min.vsize = FIMC_CROP_MIN;
-	prop_list->scale = 1;
-	prop_list->scale_max.hsize = FIMC_SCALE_MAX;
-	prop_list->scale_max.vsize = FIMC_SCALE_MAX;
-	prop_list->scale_min.hsize = FIMC_SCALE_MIN;
-	prop_list->scale_min.vsize = FIMC_SCALE_MIN;
-
-	return 0;
-}
-
-static inline bool fimc_check_drm_flip(enum drm_exynos_flip flip)
-{
-	switch (flip) {
-	case EXYNOS_DRM_FLIP_NONE:
-	case EXYNOS_DRM_FLIP_VERTICAL:
-	case EXYNOS_DRM_FLIP_HORIZONTAL:
-	case EXYNOS_DRM_FLIP_BOTH:
-		return true;
-	default:
-		DRM_DEBUG_KMS("invalid flip\n");
-		return false;
-	}
-}
-
-static int fimc_ippdrv_check_property(struct device *dev,
-		struct drm_exynos_ipp_property *property)
-{
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_prop_list *pp = &ippdrv->prop_list;
-	struct drm_exynos_ipp_config *config;
-	struct drm_exynos_pos *pos;
-	struct drm_exynos_sz *sz;
-	bool swap;
-	int i;
-
-	for_each_ipp_ops(i) {
-		if ((i == EXYNOS_DRM_OPS_SRC) &&
-			(property->cmd == IPP_CMD_WB))
-			continue;
-
-		config = &property->config[i];
-		pos = &config->pos;
-		sz = &config->sz;
-
-		/* check for flip */
-		if (!fimc_check_drm_flip(config->flip)) {
-			DRM_ERROR("invalid flip.\n");
-			goto err_property;
-		}
-
-		/* check for degree */
-		switch (config->degree) {
-		case EXYNOS_DRM_DEGREE_90:
-		case EXYNOS_DRM_DEGREE_270:
-			swap = true;
-			break;
-		case EXYNOS_DRM_DEGREE_0:
-		case EXYNOS_DRM_DEGREE_180:
-			swap = false;
-			break;
-		default:
-			DRM_ERROR("invalid degree.\n");
-			goto err_property;
-		}
-
-		/* check for buffer bound */
-		if ((pos->x + pos->w > sz->hsize) ||
-			(pos->y + pos->h > sz->vsize)) {
-			DRM_ERROR("out of buf bound.\n");
-			goto err_property;
-		}
+	if (ctx->task) {
+		struct exynos_drm_ipp_task *task = ctx->task;
 
-		/* check for crop */
-		if ((i == EXYNOS_DRM_OPS_SRC) && (pp->crop)) {
-			if (swap) {
-				if ((pos->h < pp->crop_min.hsize) ||
-					(sz->vsize > pp->crop_max.hsize) ||
-					(pos->w < pp->crop_min.vsize) ||
-					(sz->hsize > pp->crop_max.vsize)) {
-					DRM_ERROR("out of crop size.\n");
-					goto err_property;
-				}
-			} else {
-				if ((pos->w < pp->crop_min.hsize) ||
-					(sz->hsize > pp->crop_max.hsize) ||
-					(pos->h < pp->crop_min.vsize) ||
-					(sz->vsize > pp->crop_max.vsize)) {
-					DRM_ERROR("out of crop size.\n");
-					goto err_property;
-				}
-			}
-		}
-
-		/* check for scale */
-		if ((i == EXYNOS_DRM_OPS_DST) && (pp->scale)) {
-			if (swap) {
-				if ((pos->h < pp->scale_min.hsize) ||
-					(sz->vsize > pp->scale_max.hsize) ||
-					(pos->w < pp->scale_min.vsize) ||
-					(sz->hsize > pp->scale_max.vsize)) {
-					DRM_ERROR("out of scale size.\n");
-					goto err_property;
-				}
-			} else {
-				if ((pos->w < pp->scale_min.hsize) ||
-					(sz->hsize > pp->scale_max.hsize) ||
-					(pos->h < pp->scale_min.vsize) ||
-					(sz->vsize > pp->scale_max.vsize)) {
-					DRM_ERROR("out of scale size.\n");
-					goto err_property;
-				}
-			}
-		}
+		ctx->task = NULL;
+		pm_runtime_mark_last_busy(ctx->dev);
+		pm_runtime_put_autosuspend(ctx->dev);
+		exynos_drm_ipp_task_done(task, 0);
 	}
 
-	return 0;
+	fimc_dst_set_buf_seq(ctx, buf_id, false);
+	fimc_stop(ctx);
 
-err_property:
-	for_each_ipp_ops(i) {
-		if ((i == EXYNOS_DRM_OPS_SRC) &&
-			(property->cmd == IPP_CMD_WB))
-			continue;
-
-		config = &property->config[i];
-		pos = &config->pos;
-		sz = &config->sz;
-
-		DRM_ERROR("[%s]f[%d]r[%d]pos[%d %d %d %d]sz[%d %d]\n",
-			i ? "dst" : "src", config->flip, config->degree,
-			pos->x, pos->y, pos->w, pos->h,
-			sz->hsize, sz->vsize);
-	}
-
-	return -EINVAL;
+	return IRQ_HANDLED;
 }
 
 static void fimc_clear_addr(struct fimc_context *ctx)
@@ -1386,10 +990,8 @@ static void fimc_clear_addr(struct fimc_context *ctx)
 	}
 }
 
-static int fimc_ippdrv_reset(struct device *dev)
+static void fimc_reset(struct fimc_context *ctx)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-
 	/* reset h/w block */
 	fimc_sw_reset(ctx);
 
@@ -1397,82 +999,26 @@ static int fimc_ippdrv_reset(struct device *dev)
 	memset(&ctx->sc, 0x0, sizeof(ctx->sc));
 
 	fimc_clear_addr(ctx);
-
-	return 0;
 }
 
-static int fimc_ippdrv_start(struct device *dev, enum drm_exynos_ipp_cmd cmd)
+static void fimc_start(struct fimc_context *ctx)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_property *property;
-	struct drm_exynos_ipp_config *config;
-	struct drm_exynos_pos	img_pos[EXYNOS_DRM_OPS_MAX];
-	struct drm_exynos_ipp_set_wb set_wb;
-	int ret, i;
 	u32 cfg0, cfg1;
 
-	DRM_DEBUG_KMS("cmd[%d]\n", cmd);
-
-	if (!c_node) {
-		DRM_ERROR("failed to get c_node.\n");
-		return -EINVAL;
-	}
-
-	property = &c_node->property;
-
 	fimc_mask_irq(ctx, true);
 
-	for_each_ipp_ops(i) {
-		config = &property->config[i];
-		img_pos[i] = config->pos;
-	}
-
-	ret = fimc_set_prescaler(ctx, &ctx->sc,
-		&img_pos[EXYNOS_DRM_OPS_SRC],
-		&img_pos[EXYNOS_DRM_OPS_DST]);
-	if (ret) {
-		dev_err(dev, "failed to set prescaler.\n");
-		return ret;
-	}
-
-	/* If set ture, we can save jpeg about screen */
+	/* If set true, we can save jpeg about screen */
 	fimc_handle_jpeg(ctx, false);
 	fimc_set_scaler(ctx, &ctx->sc);
 
-	switch (cmd) {
-	case IPP_CMD_M2M:
-		fimc_set_type_ctrl(ctx, FIMC_WB_NONE);
-		fimc_handle_lastend(ctx, false);
-
-		/* setup dma */
-		cfg0 = fimc_read(ctx, EXYNOS_MSCTRL);
-		cfg0 &= ~EXYNOS_MSCTRL_INPUT_MASK;
-		cfg0 |= EXYNOS_MSCTRL_INPUT_MEMORY;
-		fimc_write(ctx, cfg0, EXYNOS_MSCTRL);
-		break;
-	case IPP_CMD_WB:
-		fimc_set_type_ctrl(ctx, FIMC_WB_A);
-		fimc_handle_lastend(ctx, true);
-
-		/* setup FIMD */
-		ret = fimc_set_camblk_fimd0_wb(ctx);
-		if (ret < 0) {
-			dev_err(dev, "camblk setup failed.\n");
-			return ret;
-		}
+	fimc_set_type_ctrl(ctx);
+	fimc_handle_lastend(ctx, false);
 
-		set_wb.enable = 1;
-		set_wb.refresh = property->refresh_rate;
-		exynos_drm_ippnb_send_event(IPP_SET_WRITEBACK, (void *)&set_wb);
-		break;
-	case IPP_CMD_OUTPUT:
-	default:
-		ret = -EINVAL;
-		dev_err(dev, "invalid operations.\n");
-		return ret;
-	}
+	/* setup dma */
+	cfg0 = fimc_read(ctx, EXYNOS_MSCTRL);
+	cfg0 &= ~EXYNOS_MSCTRL_INPUT_MASK;
+	cfg0 |= EXYNOS_MSCTRL_INPUT_MEMORY;
+	fimc_write(ctx, cfg0, EXYNOS_MSCTRL);
 
 	/* Reset status */
 	fimc_write(ctx, 0x0, EXYNOS_CISTATUS);
@@ -1498,36 +1044,18 @@ static int fimc_ippdrv_start(struct device *dev, enum drm_exynos_ipp_cmd cmd)
 
 	fimc_clear_bits(ctx, EXYNOS_CIOCTRL, EXYNOS_CIOCTRL_WEAVE_MASK);
 
-	if (cmd == IPP_CMD_M2M)
-		fimc_set_bits(ctx, EXYNOS_MSCTRL, EXYNOS_MSCTRL_ENVID);
-
-	return 0;
+	fimc_set_bits(ctx, EXYNOS_MSCTRL, EXYNOS_MSCTRL_ENVID);
 }
 
-static void fimc_ippdrv_stop(struct device *dev, enum drm_exynos_ipp_cmd cmd)
+static void fimc_stop(struct fimc_context *ctx)
 {
-	struct fimc_context *ctx = get_fimc_context(dev);
-	struct drm_exynos_ipp_set_wb set_wb = {0, 0};
 	u32 cfg;
 
-	DRM_DEBUG_KMS("cmd[%d]\n", cmd);
-
-	switch (cmd) {
-	case IPP_CMD_M2M:
-		/* Source clear */
-		cfg = fimc_read(ctx, EXYNOS_MSCTRL);
-		cfg &= ~EXYNOS_MSCTRL_INPUT_MASK;
-		cfg &= ~EXYNOS_MSCTRL_ENVID;
-		fimc_write(ctx, cfg, EXYNOS_MSCTRL);
-		break;
-	case IPP_CMD_WB:
-		exynos_drm_ippnb_send_event(IPP_SET_WRITEBACK, (void *)&set_wb);
-		break;
-	case IPP_CMD_OUTPUT:
-	default:
-		dev_err(dev, "invalid operations.\n");
-		break;
-	}
+	/* Source clear */
+	cfg = fimc_read(ctx, EXYNOS_MSCTRL);
+	cfg &= ~EXYNOS_MSCTRL_INPUT_MASK;
+	cfg &= ~EXYNOS_MSCTRL_ENVID;
+	fimc_write(ctx, cfg, EXYNOS_MSCTRL);
 
 	fimc_mask_irq(ctx, false);
 
@@ -1545,6 +1073,87 @@ static void fimc_ippdrv_stop(struct device *dev, enum drm_exynos_ipp_cmd cmd)
 	fimc_set_bits(ctx, EXYNOS_CIGCTRL, EXYNOS_CIGCTRL_IRQ_END_DISABLE);
 }
 
+static int fimc_commit(struct exynos_drm_ipp *ipp,
+			  struct exynos_drm_ipp_task *task)
+{
+	struct fimc_context *ctx =
+			container_of(ipp, struct fimc_context, ipp);
+
+	pm_runtime_get_sync(ctx->dev);
+	ctx->task = task;
+
+	fimc_src_set_fmt(ctx, task->src.buf.fourcc, task->src.buf.modifier);
+	fimc_src_set_size(ctx, &task->src);
+	fimc_src_set_transf(ctx, DRM_MODE_ROTATE_0);
+	fimc_src_set_addr(ctx, &task->src);
+	fimc_dst_set_fmt(ctx, task->dst.buf.fourcc, task->dst.buf.modifier);
+	fimc_dst_set_transf(ctx, task->transform.rotation);
+	fimc_dst_set_size(ctx, &task->dst);
+	fimc_dst_set_addr(ctx, &task->dst);
+	fimc_set_prescaler(ctx, &ctx->sc, &task->src.rect, &task->dst.rect);
+	fimc_start(ctx);
+
+	return 0;
+}
+
+static void fimc_abort(struct exynos_drm_ipp *ipp,
+			  struct exynos_drm_ipp_task *task)
+{
+	struct fimc_context *ctx =
+			container_of(ipp, struct fimc_context, ipp);
+
+	fimc_reset(ctx);
+
+	if (ctx->task) {
+		struct exynos_drm_ipp_task *task = ctx->task;
+
+		ctx->task = NULL;
+		pm_runtime_mark_last_busy(ctx->dev);
+		pm_runtime_put_autosuspend(ctx->dev);
+		exynos_drm_ipp_task_done(task, -EIO);
+	}
+}
+
+static struct exynos_drm_ipp_funcs ipp_funcs = {
+	.commit = fimc_commit,
+	.abort = fimc_abort,
+};
+
+static int fimc_bind(struct device *dev, struct device *master, void *data)
+{
+	struct fimc_context *ctx = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &ctx->ipp;
+
+	ctx->drm_dev = drm_dev;
+	drm_iommu_attach_device(drm_dev, dev);
+
+	exynos_drm_ipp_register(drm_dev, ipp, &ipp_funcs,
+			DRM_EXYNOS_IPP_CAP_CROP | DRM_EXYNOS_IPP_CAP_ROTATE |
+			DRM_EXYNOS_IPP_CAP_SCALE | DRM_EXYNOS_IPP_CAP_CONVERT,
+			ctx->formats, ctx->num_formats, "fimc");
+
+	dev_info(dev, "The exynos fimc has been probed successfully\n");
+
+	return 0;
+}
+
+static void fimc_unbind(struct device *dev, struct device *master,
+			void *data)
+{
+	struct fimc_context *ctx = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &ctx->ipp;
+
+	exynos_drm_ipp_unregister(drm_dev, ipp);
+	drm_iommu_detach_device(drm_dev, dev);
+}
+
+static const struct component_ops fimc_component_ops = {
+	.bind	= fimc_bind,
+	.unbind = fimc_unbind,
+};
+
 static void fimc_put_clocks(struct fimc_context *ctx)
 {
 	int i;
@@ -1559,7 +1168,7 @@ static void fimc_put_clocks(struct fimc_context *ctx)
 
 static int fimc_setup_clocks(struct fimc_context *ctx)
 {
-	struct device *fimc_dev = ctx->ippdrv.dev;
+	struct device *fimc_dev = ctx->dev;
 	struct device *dev;
 	int ret, i;
 
@@ -1574,8 +1183,6 @@ static int fimc_setup_clocks(struct fimc_context *ctx)
 
 		ctx->clocks[i] = clk_get(dev, fimc_clock_names[i]);
 		if (IS_ERR(ctx->clocks[i])) {
-			if (i >= FIMC_CLK_MUX)
-				break;
 			ret = PTR_ERR(ctx->clocks[i]);
 			dev_err(fimc_dev, "failed to get clock: %s\n",
 						fimc_clock_names[i]);
@@ -1583,20 +1190,6 @@ static int fimc_setup_clocks(struct fimc_context *ctx)
 		}
 	}
 
-	/* Optional FIMC LCLK parent clock setting */
-	if (!IS_ERR(ctx->clocks[FIMC_CLK_PARENT])) {
-		ret = clk_set_parent(ctx->clocks[FIMC_CLK_MUX],
-				     ctx->clocks[FIMC_CLK_PARENT]);
-		if (ret < 0) {
-			dev_err(fimc_dev, "failed to set parent.\n");
-			goto e_clk_free;
-		}
-	}
-
-	ret = clk_set_rate(ctx->clocks[FIMC_CLK_LCLK], ctx->clk_frequency);
-	if (ret < 0)
-		goto e_clk_free;
-
 	ret = clk_prepare_enable(ctx->clocks[FIMC_CLK_LCLK]);
 	if (!ret)
 		return ret;
@@ -1605,57 +1198,119 @@ e_clk_free:
 	return ret;
 }
 
-static int fimc_parse_dt(struct fimc_context *ctx)
+int exynos_drm_check_fimc_device(struct device *dev)
 {
-	struct device_node *node = ctx->ippdrv.dev->of_node;
+	int id = of_alias_get_id(dev->of_node, "fimc");
 
-	/* Handle only devices that support the LCD Writeback data path */
-	if (!of_property_read_bool(node, "samsung,lcd-wb"))
-		return -ENODEV;
+	if (id >= 0 && (BIT(id) & fimc_mask))
+		return 0;
+	return -ENODEV;
+}
 
-	if (of_property_read_u32(node, "clock-frequency",
-					&ctx->clk_frequency))
-		ctx->clk_frequency = FIMC_DEFAULT_LCLK_FREQUENCY;
+static const unsigned int fimc_formats[] = {
+	DRM_FORMAT_XRGB8888, DRM_FORMAT_RGB565,
+	DRM_FORMAT_NV12, DRM_FORMAT_NV16, DRM_FORMAT_NV21, DRM_FORMAT_NV61,
+	DRM_FORMAT_UYVY, DRM_FORMAT_VYUY, DRM_FORMAT_YUYV, DRM_FORMAT_YVYU,
+	DRM_FORMAT_YUV420, DRM_FORMAT_YVU420, DRM_FORMAT_YUV422,
+	DRM_FORMAT_YUV444,
+};
 
-	ctx->id = of_alias_get_id(node, "fimc");
+static const unsigned int fimc_tiled_formats[] = {
+	DRM_FORMAT_NV12, DRM_FORMAT_NV21,
+};
 
-	if (ctx->id < 0) {
-		dev_err(ctx->ippdrv.dev, "failed to get node alias id.\n");
-		return -EINVAL;
-	}
+static const struct drm_exynos_ipp_limit fimc_4210_limits_v1[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 16, 8192, 8 }, .v = { 16, 8192, 2 }) },
+	{ IPP_SIZE_LIMIT(AREA, .h = { 16, 4224, 2 }, .v = { 16, 0, 2 }) },
+	{ IPP_SIZE_LIMIT(ROTATED, .h = { 128, 1920 }, .v = { 128, 0 }) },
+	{ IPP_SCALE_LIMIT(.h = { (1 << 16) / 64, (1 << 16) * 64 },
+			  .v = { (1 << 16) / 64, (1 << 16) * 64 }) },
+};
 
-	return 0;
-}
+static const struct drm_exynos_ipp_limit fimc_4210_limits_v2[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 16, 8192, 8 }, .v = { 16, 8192, 2 }) },
+	{ IPP_SIZE_LIMIT(AREA, .h = { 16, 1920, 2 }, .v = { 16, 0, 2 }) },
+	{ IPP_SIZE_LIMIT(ROTATED, .h = { 128, 1366 }, .v = { 128, 0 }) },
+	{ IPP_SCALE_LIMIT(.h = { (1 << 16) / 64, (1 << 16) * 64 },
+			  .v = { (1 << 16) / 64, (1 << 16) * 64 }) },
+};
+
+static const struct drm_exynos_ipp_limit fimc_4210_limits_tiled_v1[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 128, 1920, 128 }, .v = { 32, 1920, 32 }) },
+	{ IPP_SIZE_LIMIT(AREA, .h = { 128, 1920, 2 }, .v = { 128, 0, 2 }) },
+	{ IPP_SCALE_LIMIT(.h = { (1 << 16) / 64, (1 << 16) * 64 },
+			  .v = { (1 << 16) / 64, (1 << 16) * 64 }) },
+};
+
+static const struct drm_exynos_ipp_limit fimc_4210_limits_tiled_v2[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 128, 1920, 128 }, .v = { 32, 1920, 32 }) },
+	{ IPP_SIZE_LIMIT(AREA, .h = { 128, 1366, 2 }, .v = { 128, 0, 2 }) },
+	{ IPP_SCALE_LIMIT(.h = { (1 << 16) / 64, (1 << 16) * 64 },
+			  .v = { (1 << 16) / 64, (1 << 16) * 64 }) },
+};
 
 static int fimc_probe(struct platform_device *pdev)
 {
+	const struct drm_exynos_ipp_limit *limits;
+	struct exynos_drm_ipp_formats *formats;
 	struct device *dev = &pdev->dev;
 	struct fimc_context *ctx;
 	struct resource *res;
-	struct exynos_drm_ippdrv *ippdrv;
 	int ret;
+	int i, j, num_limits, num_formats;
 
-	if (!dev->of_node) {
-		dev_err(dev, "device tree node not found.\n");
+	if (exynos_drm_check_fimc_device(dev) != 0)
 		return -ENODEV;
-	}
 
 	ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
 	if (!ctx)
 		return -ENOMEM;
 
-	ctx->ippdrv.dev = dev;
+	ctx->dev = dev;
+	ctx->id = of_alias_get_id(dev->of_node, "fimc");
 
-	ret = fimc_parse_dt(ctx);
-	if (ret < 0)
-		return ret;
+	/* construct formats/limits array */
+	num_formats = ARRAY_SIZE(fimc_formats) + ARRAY_SIZE(fimc_tiled_formats);
+	formats = devm_kcalloc(dev, num_formats, sizeof(*formats),
+			       GFP_KERNEL);
+	if (!formats)
+		return -ENOMEM;
+
+	/* linear formats */
+	if (ctx->id < 3) {
+		limits = fimc_4210_limits_v1;
+		num_limits = ARRAY_SIZE(fimc_4210_limits_v1);
+	} else {
+		limits = fimc_4210_limits_v2;
+		num_limits = ARRAY_SIZE(fimc_4210_limits_v2);
+	}
+	for (i = 0; i < ARRAY_SIZE(fimc_formats); i++) {
+		formats[i].fourcc = fimc_formats[i];
+		formats[i].type = DRM_EXYNOS_IPP_FORMAT_SOURCE |
+				  DRM_EXYNOS_IPP_FORMAT_DESTINATION;
+		formats[i].limits = limits;
+		formats[i].num_limits = num_limits;
+	}
 
-	ctx->sysreg = syscon_regmap_lookup_by_phandle(dev->of_node,
-						"samsung,sysreg");
-	if (IS_ERR(ctx->sysreg)) {
-		dev_err(dev, "syscon regmap lookup failed.\n");
-		return PTR_ERR(ctx->sysreg);
+	/* tiled formats */
+	if (ctx->id < 3) {
+		limits = fimc_4210_limits_tiled_v1;
+		num_limits = ARRAY_SIZE(fimc_4210_limits_tiled_v1);
+	} else {
+		limits = fimc_4210_limits_tiled_v2;
+		num_limits = ARRAY_SIZE(fimc_4210_limits_tiled_v2);
 	}
+	for (j = i, i = 0; i < ARRAY_SIZE(fimc_tiled_formats); j++, i++) {
+		formats[j].fourcc = fimc_tiled_formats[i];
+		formats[j].modifier = DRM_FORMAT_MOD_SAMSUNG_64_32_TILE;
+		formats[j].type = DRM_EXYNOS_IPP_FORMAT_SOURCE |
+				  DRM_EXYNOS_IPP_FORMAT_DESTINATION;
+		formats[j].limits = limits;
+		formats[j].num_limits = num_limits;
+	}
+
+	ctx->formats = formats;
+	ctx->num_formats = num_formats;
 
 	/* resource memory */
 	ctx->regs_res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
@@ -1670,9 +1325,8 @@ static int fimc_probe(struct platform_device *pdev)
 		return -ENOENT;
 	}
 
-	ctx->irq = res->start;
-	ret = devm_request_threaded_irq(dev, ctx->irq, NULL, fimc_irq_handler,
-		IRQF_ONESHOT, "drm_fimc", ctx);
+	ret = devm_request_irq(dev, res->start, fimc_irq_handler,
+		0, dev_name(dev), ctx);
 	if (ret < 0) {
 		dev_err(dev, "failed to request irq.\n");
 		return ret;
@@ -1682,39 +1336,24 @@ static int fimc_probe(struct platform_device *pdev)
 	if (ret < 0)
 		return ret;
 
-	ippdrv = &ctx->ippdrv;
-	ippdrv->ops[EXYNOS_DRM_OPS_SRC] = &fimc_src_ops;
-	ippdrv->ops[EXYNOS_DRM_OPS_DST] = &fimc_dst_ops;
-	ippdrv->check_property = fimc_ippdrv_check_property;
-	ippdrv->reset = fimc_ippdrv_reset;
-	ippdrv->start = fimc_ippdrv_start;
-	ippdrv->stop = fimc_ippdrv_stop;
-	ret = fimc_init_prop_list(ippdrv);
-	if (ret < 0) {
-		dev_err(dev, "failed to init property list.\n");
-		goto err_put_clk;
-	}
-
-	DRM_DEBUG_KMS("id[%d]ippdrv[%pK]\n", ctx->id, ippdrv);
-
 	spin_lock_init(&ctx->lock);
 	platform_set_drvdata(pdev, ctx);
 
+	pm_runtime_use_autosuspend(dev);
+	pm_runtime_set_autosuspend_delay(dev, FIMC_AUTOSUSPEND_DELAY);
 	pm_runtime_enable(dev);
 
-	ret = exynos_drm_ippdrv_register(ippdrv);
-	if (ret < 0) {
-		dev_err(dev, "failed to register drm fimc device.\n");
+	ret = component_add(dev, &fimc_component_ops);
+	if (ret)
 		goto err_pm_dis;
-	}
 
 	dev_info(dev, "drm fimc registered successfully.\n");
 
 	return 0;
 
 err_pm_dis:
+	pm_runtime_dont_use_autosuspend(dev);
 	pm_runtime_disable(dev);
-err_put_clk:
 	fimc_put_clocks(ctx);
 
 	return ret;
@@ -1724,42 +1363,24 @@ static int fimc_remove(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
 	struct fimc_context *ctx = get_fimc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 
-	exynos_drm_ippdrv_unregister(ippdrv);
+	component_del(dev, &fimc_component_ops);
+	pm_runtime_dont_use_autosuspend(dev);
+	pm_runtime_disable(dev);
 
 	fimc_put_clocks(ctx);
-	pm_runtime_set_suspended(dev);
-	pm_runtime_disable(dev);
 
 	return 0;
 }
 
 #ifdef CONFIG_PM
-static int fimc_clk_ctrl(struct fimc_context *ctx, bool enable)
-{
-	DRM_DEBUG_KMS("enable[%d]\n", enable);
-
-	if (enable) {
-		clk_prepare_enable(ctx->clocks[FIMC_CLK_GATE]);
-		clk_prepare_enable(ctx->clocks[FIMC_CLK_WB_A]);
-		ctx->suspended = false;
-	} else {
-		clk_disable_unprepare(ctx->clocks[FIMC_CLK_GATE]);
-		clk_disable_unprepare(ctx->clocks[FIMC_CLK_WB_A]);
-		ctx->suspended = true;
-	}
-
-	return 0;
-}
-
 static int fimc_runtime_suspend(struct device *dev)
 {
 	struct fimc_context *ctx = get_fimc_context(dev);
 
 	DRM_DEBUG_KMS("id[%d]\n", ctx->id);
-
-	return  fimc_clk_ctrl(ctx, false);
+	clk_disable_unprepare(ctx->clocks[FIMC_CLK_GATE]);
+	return 0;
 }
 
 static int fimc_runtime_resume(struct device *dev)
@@ -1767,8 +1388,7 @@ static int fimc_runtime_resume(struct device *dev)
 	struct fimc_context *ctx = get_fimc_context(dev);
 
 	DRM_DEBUG_KMS("id[%d]\n", ctx->id);
-
-	return  fimc_clk_ctrl(ctx, true);
+	return clk_prepare_enable(ctx->clocks[FIMC_CLK_GATE]);
 }
 #endif
 
@@ -1795,4 +1415,3 @@ struct platform_driver fimc_driver = {
 		.pm	= &fimc_pm_ops,
 	},
 };
-
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimc.h b/drivers/gpu/drm/exynos/exynos_drm_fimc.h
deleted file mode 100644
index 127a424c5fdf..000000000000
--- a/drivers/gpu/drm/exynos/exynos_drm_fimc.h
+++ /dev/null
@@ -1,23 +0,0 @@
-/*
- * Copyright (c) 2012 Samsung Electronics Co., Ltd.
- *
- * Authors:
- *	Eunchul Kim <chulspro.kim@samsung.com>
- *	Jinyoung Jeon <jy0.jeon@samsung.com>
- *	Sangmin Lee <lsmin.lee@samsung.com>
- *
- * This program is free software; you can redistribute  it and/or modify it
- * under  the terms of  the GNU General  Public License as published by the
- * Free Software Foundation;  either version 2 of the  License, or (at your
- * option) any later version.
- */
-
-#ifndef _EXYNOS_DRM_FIMC_H_
-#define _EXYNOS_DRM_FIMC_H_
-
-/*
- * TODO
- * FIMD output interface notifier callback.
- */
-
-#endif /* _EXYNOS_DRM_FIMC_H_ */
diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimd.c b/drivers/gpu/drm/exynos/exynos_drm_fimd.c
index d42ae2bc3e56..01b1570d0c3a 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_fimd.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_fimd.c
@@ -121,6 +121,12 @@ static struct fimd_driver_data s3c64xx_fimd_driver_data = {
 	.has_limited_fmt = 1,
 };
 
+static struct fimd_driver_data s5pv210_fimd_driver_data = {
+	.timing_base = 0x0,
+	.has_shadowcon = 1,
+	.has_clksel = 1,
+};
+
 static struct fimd_driver_data exynos3_fimd_driver_data = {
 	.timing_base = 0x20000,
 	.lcdblk_offset = 0x210,
@@ -193,6 +199,8 @@ struct fimd_context {
 static const struct of_device_id fimd_driver_dt_match[] = {
 	{ .compatible = "samsung,s3c6400-fimd",
 	  .data = &s3c64xx_fimd_driver_data },
+	{ .compatible = "samsung,s5pv210-fimd",
+	  .data = &s5pv210_fimd_driver_data },
 	{ .compatible = "samsung,exynos3250-fimd",
 	  .data = &exynos3_fimd_driver_data },
 	{ .compatible = "samsung,exynos4210-fimd",
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.c b/drivers/gpu/drm/exynos/exynos_drm_gem.c
index 11cc01b47bc0..6e1494fa71b4 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.c
@@ -431,37 +431,24 @@ int exynos_drm_gem_dumb_create(struct drm_file *file_priv,
 	return 0;
 }
 
-int exynos_drm_gem_fault(struct vm_fault *vmf)
+vm_fault_t exynos_drm_gem_fault(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 	struct drm_gem_object *obj = vma->vm_private_data;
 	struct exynos_drm_gem *exynos_gem = to_exynos_gem(obj);
 	unsigned long pfn;
 	pgoff_t page_offset;
-	int ret;
 
 	page_offset = (vmf->address - vma->vm_start) >> PAGE_SHIFT;
 
 	if (page_offset >= (exynos_gem->size >> PAGE_SHIFT)) {
 		DRM_ERROR("invalid page offset\n");
-		ret = -EINVAL;
-		goto out;
+		return VM_FAULT_SIGBUS;
 	}
 
 	pfn = page_to_pfn(exynos_gem->pages[page_offset]);
-	ret = vm_insert_mixed(vma, vmf->address, __pfn_to_pfn_t(pfn, PFN_DEV));
-
-out:
-	switch (ret) {
-	case 0:
-	case -ERESTARTSYS:
-	case -EINTR:
-		return VM_FAULT_NOPAGE;
-	case -ENOMEM:
-		return VM_FAULT_OOM;
-	default:
-		return VM_FAULT_SIGBUS;
-	}
+	return vmf_insert_mixed(vma, vmf->address,
+			__pfn_to_pfn_t(pfn, PFN_DEV));
 }
 
 static int exynos_drm_gem_mmap_obj(struct drm_gem_object *obj,
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gem.h b/drivers/gpu/drm/exynos/exynos_drm_gem.h
index 5a4c7de80f65..9057d7f1d6ed 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gem.h
+++ b/drivers/gpu/drm/exynos/exynos_drm_gem.h
@@ -13,6 +13,7 @@
 #define _EXYNOS_DRM_GEM_H_
 
 #include <drm/drm_gem.h>
+#include <linux/mm_types.h>
 
 #define to_exynos_gem(x)	container_of(x, struct exynos_drm_gem, base)
 
@@ -111,7 +112,7 @@ int exynos_drm_gem_dumb_create(struct drm_file *file_priv,
 			       struct drm_mode_create_dumb *args);
 
 /* page fault handler and mmap fault address(virtual) to physical memory. */
-int exynos_drm_gem_fault(struct vm_fault *vmf);
+vm_fault_t exynos_drm_gem_fault(struct vm_fault *vmf);
 
 /* set vm_flags and we can change the vm attribute to other one at here. */
 int exynos_drm_gem_mmap(struct file *filp, struct vm_area_struct *vma);
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gsc.c b/drivers/gpu/drm/exynos/exynos_drm_gsc.c
index 0506b2b17ac1..35ac66730563 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_gsc.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_gsc.c
@@ -12,18 +12,20 @@
  *
  */
 #include <linux/kernel.h>
+#include <linux/component.h>
 #include <linux/platform_device.h>
 #include <linux/clk.h>
 #include <linux/pm_runtime.h>
 #include <linux/mfd/syscon.h>
+#include <linux/of_device.h>
 #include <linux/regmap.h>
 
 #include <drm/drmP.h>
 #include <drm/exynos_drm.h>
 #include "regs-gsc.h"
 #include "exynos_drm_drv.h"
+#include "exynos_drm_iommu.h"
 #include "exynos_drm_ipp.h"
-#include "exynos_drm_gsc.h"
 
 /*
  * GSC stands for General SCaler and
@@ -31,26 +33,10 @@
  * input DMA reads image data from the memory.
  * output DMA writes image data to memory.
  * GSC supports image rotation and image effect functions.
- *
- * M2M operation : supports crop/scale/rotation/csc so on.
- * Memory ----> GSC H/W ----> Memory.
- * Writeback operation : supports cloned screen with FIMD.
- * FIMD ----> GSC H/W ----> Memory.
- * Output operation : supports direct display using local path.
- * Memory ----> GSC H/W ----> FIMD, Mixer.
  */
 
-/*
- * TODO
- * 1. check suspend/resume api if needed.
- * 2. need to check use case platform_device_id.
- * 3. check src/dst size with, height.
- * 4. added check_prepare api for right register.
- * 5. need to add supported list in prop_list.
- * 6. check prescaler/scaler optimization.
- */
 
-#define GSC_MAX_DEVS	4
+#define GSC_MAX_CLOCKS	8
 #define GSC_MAX_SRC		4
 #define GSC_MAX_DST		16
 #define GSC_RESET_TIMEOUT	50
@@ -65,8 +51,6 @@
 #define GSC_SC_DOWN_RATIO_4_8		131072
 #define GSC_SC_DOWN_RATIO_3_8		174762
 #define GSC_SC_DOWN_RATIO_2_8		262144
-#define GSC_REFRESH_MIN	12
-#define GSC_REFRESH_MAX	60
 #define GSC_CROP_MAX	8192
 #define GSC_CROP_MIN	32
 #define GSC_SCALE_MAX	4224
@@ -77,10 +61,9 @@
 #define GSC_COEF_H_8T	8
 #define GSC_COEF_V_4T	4
 #define GSC_COEF_DEPTH	3
+#define GSC_AUTOSUSPEND_DELAY		2000
 
 #define get_gsc_context(dev)	platform_get_drvdata(to_platform_device(dev))
-#define get_ctx_from_ippdrv(ippdrv)	container_of(ippdrv,\
-					struct gsc_context, ippdrv);
 #define gsc_read(offset)		readl(ctx->regs + (offset))
 #define gsc_write(cfg, offset)	writel(cfg, ctx->regs + (offset))
 
@@ -104,50 +87,47 @@ struct gsc_scaler {
 };
 
 /*
- * A structure of scaler capability.
- *
- * find user manual 49.2 features.
- * @tile_w: tile mode or rotation width.
- * @tile_h: tile mode or rotation height.
- * @w: other cases width.
- * @h: other cases height.
- */
-struct gsc_capability {
-	/* tile or rotation */
-	u32	tile_w;
-	u32	tile_h;
-	/* other cases */
-	u32	w;
-	u32	h;
-};
-
-/*
  * A structure of gsc context.
  *
- * @ippdrv: prepare initialization using ippdrv.
  * @regs_res: register resources.
  * @regs: memory mapped io registers.
- * @sysreg: handle to SYSREG block regmap.
- * @lock: locking of operations.
  * @gsc_clk: gsc gate clock.
  * @sc: scaler infomations.
  * @id: gsc id.
  * @irq: irq number.
  * @rotation: supports rotation of src.
- * @suspended: qos operations.
  */
 struct gsc_context {
-	struct exynos_drm_ippdrv	ippdrv;
+	struct exynos_drm_ipp ipp;
+	struct drm_device *drm_dev;
+	struct device	*dev;
+	struct exynos_drm_ipp_task	*task;
+	struct exynos_drm_ipp_formats	*formats;
+	unsigned int			num_formats;
+
 	struct resource	*regs_res;
 	void __iomem	*regs;
-	struct regmap	*sysreg;
-	struct mutex	lock;
-	struct clk	*gsc_clk;
+	const char	**clk_names;
+	struct clk	*clocks[GSC_MAX_CLOCKS];
+	int		num_clocks;
 	struct gsc_scaler	sc;
 	int	id;
 	int	irq;
 	bool	rotation;
-	bool	suspended;
+};
+
+/**
+ * struct gsc_driverdata - per device type driver data for init time.
+ *
+ * @limits: picture size limits array
+ * @clk_names: names of clocks needed by this variant
+ * @num_clocks: the number of clocks needed by this variant
+ */
+struct gsc_driverdata {
+	const struct drm_exynos_ipp_limit *limits;
+	int		num_limits;
+	const char	*clk_names[GSC_MAX_CLOCKS];
+	int		num_clocks;
 };
 
 /* 8-tap Filter Coefficient */
@@ -438,25 +418,6 @@ static int gsc_sw_reset(struct gsc_context *ctx)
 	return 0;
 }
 
-static void gsc_set_gscblk_fimd_wb(struct gsc_context *ctx, bool enable)
-{
-	unsigned int gscblk_cfg;
-
-	if (!ctx->sysreg)
-		return;
-
-	regmap_read(ctx->sysreg, SYSREG_GSCBLK_CFG1, &gscblk_cfg);
-
-	if (enable)
-		gscblk_cfg |= GSC_BLK_DISP1WB_DEST(ctx->id) |
-				GSC_BLK_GSCL_WB_IN_SRC_SEL(ctx->id) |
-				GSC_BLK_SW_RESET_WB_DEST(ctx->id);
-	else
-		gscblk_cfg |= GSC_BLK_PXLASYNC_LO_MASK_WB(ctx->id);
-
-	regmap_write(ctx->sysreg, SYSREG_GSCBLK_CFG1, gscblk_cfg);
-}
-
 static void gsc_handle_irq(struct gsc_context *ctx, bool enable,
 		bool overflow, bool done)
 {
@@ -487,10 +448,8 @@ static void gsc_handle_irq(struct gsc_context *ctx, bool enable,
 }
 
 
-static int gsc_src_set_fmt(struct device *dev, u32 fmt)
+static void gsc_src_set_fmt(struct gsc_context *ctx, u32 fmt)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg;
 
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
@@ -506,6 +465,7 @@ static int gsc_src_set_fmt(struct device *dev, u32 fmt)
 		cfg |= GSC_IN_RGB565;
 		break;
 	case DRM_FORMAT_XRGB8888:
+	case DRM_FORMAT_ARGB8888:
 		cfg |= GSC_IN_XRGB8888;
 		break;
 	case DRM_FORMAT_BGRX8888:
@@ -548,115 +508,84 @@ static int gsc_src_set_fmt(struct device *dev, u32 fmt)
 		cfg |= (GSC_IN_CHROMA_ORDER_CBCR |
 			GSC_IN_YUV420_2P);
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid target yuv order 0x%x.\n", fmt);
-		return -EINVAL;
 	}
 
 	gsc_write(cfg, GSC_IN_CON);
-
-	return 0;
 }
 
-static int gsc_src_set_transf(struct device *dev,
-		enum drm_exynos_degree degree,
-		enum drm_exynos_flip flip, bool *swap)
+static void gsc_src_set_transf(struct gsc_context *ctx, unsigned int rotation)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
+	unsigned int degree = rotation & DRM_MODE_ROTATE_MASK;
 	u32 cfg;
 
-	DRM_DEBUG_KMS("degree[%d]flip[0x%x]\n", degree, flip);
-
 	cfg = gsc_read(GSC_IN_CON);
 	cfg &= ~GSC_IN_ROT_MASK;
 
 	switch (degree) {
-	case EXYNOS_DRM_DEGREE_0:
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+	case DRM_MODE_ROTATE_0:
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg |= GSC_IN_ROT_XFLIP;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg |= GSC_IN_ROT_YFLIP;
 		break;
-	case EXYNOS_DRM_DEGREE_90:
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
-			cfg |= GSC_IN_ROT_90_XFLIP;
-		else if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
-			cfg |= GSC_IN_ROT_90_YFLIP;
-		else
-			cfg |= GSC_IN_ROT_90;
+	case DRM_MODE_ROTATE_90:
+		cfg |= GSC_IN_ROT_90;
+		if (rotation & DRM_MODE_REFLECT_Y)
+			cfg |= GSC_IN_ROT_XFLIP;
+		if (rotation & DRM_MODE_REFLECT_X)
+			cfg |= GSC_IN_ROT_YFLIP;
 		break;
-	case EXYNOS_DRM_DEGREE_180:
+	case DRM_MODE_ROTATE_180:
 		cfg |= GSC_IN_ROT_180;
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg &= ~GSC_IN_ROT_XFLIP;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg &= ~GSC_IN_ROT_YFLIP;
 		break;
-	case EXYNOS_DRM_DEGREE_270:
+	case DRM_MODE_ROTATE_270:
 		cfg |= GSC_IN_ROT_270;
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
+		if (rotation & DRM_MODE_REFLECT_Y)
 			cfg &= ~GSC_IN_ROT_XFLIP;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
+		if (rotation & DRM_MODE_REFLECT_X)
 			cfg &= ~GSC_IN_ROT_YFLIP;
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid degree value %d.\n", degree);
-		return -EINVAL;
 	}
 
 	gsc_write(cfg, GSC_IN_CON);
 
 	ctx->rotation = (cfg & GSC_IN_ROT_90) ? 1 : 0;
-	*swap = ctx->rotation;
-
-	return 0;
 }
 
-static int gsc_src_set_size(struct device *dev, int swap,
-		struct drm_exynos_pos *pos, struct drm_exynos_sz *sz)
+static void gsc_src_set_size(struct gsc_context *ctx,
+			     struct exynos_drm_ipp_buffer *buf)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct drm_exynos_pos img_pos = *pos;
 	struct gsc_scaler *sc = &ctx->sc;
 	u32 cfg;
 
-	DRM_DEBUG_KMS("swap[%d]x[%d]y[%d]w[%d]h[%d]\n",
-		swap, pos->x, pos->y, pos->w, pos->h);
-
-	if (swap) {
-		img_pos.w = pos->h;
-		img_pos.h = pos->w;
-	}
-
 	/* pixel offset */
-	cfg = (GSC_SRCIMG_OFFSET_X(img_pos.x) |
-		GSC_SRCIMG_OFFSET_Y(img_pos.y));
+	cfg = (GSC_SRCIMG_OFFSET_X(buf->rect.x) |
+		GSC_SRCIMG_OFFSET_Y(buf->rect.y));
 	gsc_write(cfg, GSC_SRCIMG_OFFSET);
 
 	/* cropped size */
-	cfg = (GSC_CROPPED_WIDTH(img_pos.w) |
-		GSC_CROPPED_HEIGHT(img_pos.h));
+	cfg = (GSC_CROPPED_WIDTH(buf->rect.w) |
+		GSC_CROPPED_HEIGHT(buf->rect.h));
 	gsc_write(cfg, GSC_CROPPED_SIZE);
 
-	DRM_DEBUG_KMS("hsize[%d]vsize[%d]\n", sz->hsize, sz->vsize);
-
 	/* original size */
 	cfg = gsc_read(GSC_SRCIMG_SIZE);
 	cfg &= ~(GSC_SRCIMG_HEIGHT_MASK |
 		GSC_SRCIMG_WIDTH_MASK);
 
-	cfg |= (GSC_SRCIMG_WIDTH(sz->hsize) |
-		GSC_SRCIMG_HEIGHT(sz->vsize));
+	cfg |= (GSC_SRCIMG_WIDTH(buf->buf.width) |
+		GSC_SRCIMG_HEIGHT(buf->buf.height));
 
 	gsc_write(cfg, GSC_SRCIMG_SIZE);
 
 	cfg = gsc_read(GSC_IN_CON);
 	cfg &= ~GSC_IN_RGB_TYPE_MASK;
 
-	DRM_DEBUG_KMS("width[%d]range[%d]\n", pos->w, sc->range);
-
-	if (pos->w >= GSC_WIDTH_ITU_709)
+	if (buf->rect.w >= GSC_WIDTH_ITU_709)
 		if (sc->range)
 			cfg |= GSC_IN_RGB_HD_WIDE;
 		else
@@ -668,103 +597,39 @@ static int gsc_src_set_size(struct device *dev, int swap,
 			cfg |= GSC_IN_RGB_SD_NARROW;
 
 	gsc_write(cfg, GSC_IN_CON);
-
-	return 0;
 }
 
-static int gsc_src_set_buf_seq(struct gsc_context *ctx, u32 buf_id,
-		enum drm_exynos_ipp_buf_type buf_type)
+static void gsc_src_set_buf_seq(struct gsc_context *ctx, u32 buf_id,
+			       bool enqueue)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	bool masked;
+	bool masked = !enqueue;
 	u32 cfg;
 	u32 mask = 0x00000001 << buf_id;
 
-	DRM_DEBUG_KMS("buf_id[%d]buf_type[%d]\n", buf_id, buf_type);
-
 	/* mask register set */
 	cfg = gsc_read(GSC_IN_BASE_ADDR_Y_MASK);
 
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		masked = false;
-		break;
-	case IPP_BUF_DEQUEUE:
-		masked = true;
-		break;
-	default:
-		dev_err(ippdrv->dev, "invalid buf ctrl parameter.\n");
-		return -EINVAL;
-	}
-
 	/* sequence id */
 	cfg &= ~mask;
 	cfg |= masked << buf_id;
 	gsc_write(cfg, GSC_IN_BASE_ADDR_Y_MASK);
 	gsc_write(cfg, GSC_IN_BASE_ADDR_CB_MASK);
 	gsc_write(cfg, GSC_IN_BASE_ADDR_CR_MASK);
-
-	return 0;
 }
 
-static int gsc_src_set_addr(struct device *dev,
-		struct drm_exynos_ipp_buf_info *buf_info, u32 buf_id,
-		enum drm_exynos_ipp_buf_type buf_type)
+static void gsc_src_set_addr(struct gsc_context *ctx, u32 buf_id,
+			    struct exynos_drm_ipp_buffer *buf)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_property *property;
-
-	if (!c_node) {
-		DRM_ERROR("failed to get c_node.\n");
-		return -EFAULT;
-	}
-
-	property = &c_node->property;
-
-	DRM_DEBUG_KMS("prop_id[%d]buf_id[%d]buf_type[%d]\n",
-		property->prop_id, buf_id, buf_type);
-
-	if (buf_id > GSC_MAX_SRC) {
-		dev_info(ippdrv->dev, "invalid buf_id %d.\n", buf_id);
-		return -EINVAL;
-	}
-
 	/* address register set */
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		gsc_write(buf_info->base[EXYNOS_DRM_PLANAR_Y],
-			GSC_IN_BASE_ADDR_Y(buf_id));
-		gsc_write(buf_info->base[EXYNOS_DRM_PLANAR_CB],
-			GSC_IN_BASE_ADDR_CB(buf_id));
-		gsc_write(buf_info->base[EXYNOS_DRM_PLANAR_CR],
-			GSC_IN_BASE_ADDR_CR(buf_id));
-		break;
-	case IPP_BUF_DEQUEUE:
-		gsc_write(0x0, GSC_IN_BASE_ADDR_Y(buf_id));
-		gsc_write(0x0, GSC_IN_BASE_ADDR_CB(buf_id));
-		gsc_write(0x0, GSC_IN_BASE_ADDR_CR(buf_id));
-		break;
-	default:
-		/* bypass */
-		break;
-	}
+	gsc_write(buf->dma_addr[0], GSC_IN_BASE_ADDR_Y(buf_id));
+	gsc_write(buf->dma_addr[1], GSC_IN_BASE_ADDR_CB(buf_id));
+	gsc_write(buf->dma_addr[2], GSC_IN_BASE_ADDR_CR(buf_id));
 
-	return gsc_src_set_buf_seq(ctx, buf_id, buf_type);
+	gsc_src_set_buf_seq(ctx, buf_id, true);
 }
 
-static struct exynos_drm_ipp_ops gsc_src_ops = {
-	.set_fmt = gsc_src_set_fmt,
-	.set_transf = gsc_src_set_transf,
-	.set_size = gsc_src_set_size,
-	.set_addr = gsc_src_set_addr,
-};
-
-static int gsc_dst_set_fmt(struct device *dev, u32 fmt)
+static void gsc_dst_set_fmt(struct gsc_context *ctx, u32 fmt)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg;
 
 	DRM_DEBUG_KMS("fmt[0x%x]\n", fmt);
@@ -779,8 +644,9 @@ static int gsc_dst_set_fmt(struct device *dev, u32 fmt)
 	case DRM_FORMAT_RGB565:
 		cfg |= GSC_OUT_RGB565;
 		break;
+	case DRM_FORMAT_ARGB8888:
 	case DRM_FORMAT_XRGB8888:
-		cfg |= GSC_OUT_XRGB8888;
+		cfg |= (GSC_OUT_XRGB8888 | GSC_OUT_GLOBAL_ALPHA(0xff));
 		break;
 	case DRM_FORMAT_BGRX8888:
 		cfg |= (GSC_OUT_XRGB8888 | GSC_OUT_RB_SWAP);
@@ -819,69 +685,9 @@ static int gsc_dst_set_fmt(struct device *dev, u32 fmt)
 		cfg |= (GSC_OUT_CHROMA_ORDER_CBCR |
 			GSC_OUT_YUV420_2P);
 		break;
-	default:
-		dev_err(ippdrv->dev, "invalid target yuv order 0x%x.\n", fmt);
-		return -EINVAL;
 	}
 
 	gsc_write(cfg, GSC_OUT_CON);
-
-	return 0;
-}
-
-static int gsc_dst_set_transf(struct device *dev,
-		enum drm_exynos_degree degree,
-		enum drm_exynos_flip flip, bool *swap)
-{
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	u32 cfg;
-
-	DRM_DEBUG_KMS("degree[%d]flip[0x%x]\n", degree, flip);
-
-	cfg = gsc_read(GSC_IN_CON);
-	cfg &= ~GSC_IN_ROT_MASK;
-
-	switch (degree) {
-	case EXYNOS_DRM_DEGREE_0:
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
-			cfg |= GSC_IN_ROT_XFLIP;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
-			cfg |= GSC_IN_ROT_YFLIP;
-		break;
-	case EXYNOS_DRM_DEGREE_90:
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
-			cfg |= GSC_IN_ROT_90_XFLIP;
-		else if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
-			cfg |= GSC_IN_ROT_90_YFLIP;
-		else
-			cfg |= GSC_IN_ROT_90;
-		break;
-	case EXYNOS_DRM_DEGREE_180:
-		cfg |= GSC_IN_ROT_180;
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
-			cfg &= ~GSC_IN_ROT_XFLIP;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
-			cfg &= ~GSC_IN_ROT_YFLIP;
-		break;
-	case EXYNOS_DRM_DEGREE_270:
-		cfg |= GSC_IN_ROT_270;
-		if (flip & EXYNOS_DRM_FLIP_VERTICAL)
-			cfg &= ~GSC_IN_ROT_XFLIP;
-		if (flip & EXYNOS_DRM_FLIP_HORIZONTAL)
-			cfg &= ~GSC_IN_ROT_YFLIP;
-		break;
-	default:
-		dev_err(ippdrv->dev, "invalid degree value %d.\n", degree);
-		return -EINVAL;
-	}
-
-	gsc_write(cfg, GSC_IN_CON);
-
-	ctx->rotation = (cfg & GSC_IN_ROT_90) ? 1 : 0;
-	*swap = ctx->rotation;
-
-	return 0;
 }
 
 static int gsc_get_ratio_shift(u32 src, u32 dst, u32 *ratio)
@@ -919,9 +725,9 @@ static void gsc_get_prescaler_shfactor(u32 hratio, u32 vratio, u32 *shfactor)
 }
 
 static int gsc_set_prescaler(struct gsc_context *ctx, struct gsc_scaler *sc,
-		struct drm_exynos_pos *src, struct drm_exynos_pos *dst)
+			     struct drm_exynos_ipp_task_rect *src,
+			     struct drm_exynos_ipp_task_rect *dst)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 	u32 cfg;
 	u32 src_w, src_h, dst_w, dst_h;
 	int ret = 0;
@@ -939,13 +745,13 @@ static int gsc_set_prescaler(struct gsc_context *ctx, struct gsc_scaler *sc,
 
 	ret = gsc_get_ratio_shift(src_w, dst_w, &sc->pre_hratio);
 	if (ret) {
-		dev_err(ippdrv->dev, "failed to get ratio horizontal.\n");
+		dev_err(ctx->dev, "failed to get ratio horizontal.\n");
 		return ret;
 	}
 
 	ret = gsc_get_ratio_shift(src_h, dst_h, &sc->pre_vratio);
 	if (ret) {
-		dev_err(ippdrv->dev, "failed to get ratio vertical.\n");
+		dev_err(ctx->dev, "failed to get ratio vertical.\n");
 		return ret;
 	}
 
@@ -1039,47 +845,37 @@ static void gsc_set_scaler(struct gsc_context *ctx, struct gsc_scaler *sc)
 	gsc_write(cfg, GSC_MAIN_V_RATIO);
 }
 
-static int gsc_dst_set_size(struct device *dev, int swap,
-		struct drm_exynos_pos *pos, struct drm_exynos_sz *sz)
+static void gsc_dst_set_size(struct gsc_context *ctx,
+			     struct exynos_drm_ipp_buffer *buf)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct drm_exynos_pos img_pos = *pos;
 	struct gsc_scaler *sc = &ctx->sc;
 	u32 cfg;
 
-	DRM_DEBUG_KMS("swap[%d]x[%d]y[%d]w[%d]h[%d]\n",
-		swap, pos->x, pos->y, pos->w, pos->h);
-
-	if (swap) {
-		img_pos.w = pos->h;
-		img_pos.h = pos->w;
-	}
-
 	/* pixel offset */
-	cfg = (GSC_DSTIMG_OFFSET_X(pos->x) |
-		GSC_DSTIMG_OFFSET_Y(pos->y));
+	cfg = (GSC_DSTIMG_OFFSET_X(buf->rect.x) |
+		GSC_DSTIMG_OFFSET_Y(buf->rect.y));
 	gsc_write(cfg, GSC_DSTIMG_OFFSET);
 
 	/* scaled size */
-	cfg = (GSC_SCALED_WIDTH(img_pos.w) | GSC_SCALED_HEIGHT(img_pos.h));
+	if (ctx->rotation)
+		cfg = (GSC_SCALED_WIDTH(buf->rect.h) |
+		       GSC_SCALED_HEIGHT(buf->rect.w));
+	else
+		cfg = (GSC_SCALED_WIDTH(buf->rect.w) |
+		       GSC_SCALED_HEIGHT(buf->rect.h));
 	gsc_write(cfg, GSC_SCALED_SIZE);
 
-	DRM_DEBUG_KMS("hsize[%d]vsize[%d]\n", sz->hsize, sz->vsize);
-
 	/* original size */
 	cfg = gsc_read(GSC_DSTIMG_SIZE);
-	cfg &= ~(GSC_DSTIMG_HEIGHT_MASK |
-		GSC_DSTIMG_WIDTH_MASK);
-	cfg |= (GSC_DSTIMG_WIDTH(sz->hsize) |
-		GSC_DSTIMG_HEIGHT(sz->vsize));
+	cfg &= ~(GSC_DSTIMG_HEIGHT_MASK | GSC_DSTIMG_WIDTH_MASK);
+	cfg |= GSC_DSTIMG_WIDTH(buf->buf.width) |
+	       GSC_DSTIMG_HEIGHT(buf->buf.height);
 	gsc_write(cfg, GSC_DSTIMG_SIZE);
 
 	cfg = gsc_read(GSC_OUT_CON);
 	cfg &= ~GSC_OUT_RGB_TYPE_MASK;
 
-	DRM_DEBUG_KMS("width[%d]range[%d]\n", pos->w, sc->range);
-
-	if (pos->w >= GSC_WIDTH_ITU_709)
+	if (buf->rect.w >= GSC_WIDTH_ITU_709)
 		if (sc->range)
 			cfg |= GSC_OUT_RGB_HD_WIDE;
 		else
@@ -1091,8 +887,6 @@ static int gsc_dst_set_size(struct device *dev, int swap,
 			cfg |= GSC_OUT_RGB_SD_NARROW;
 
 	gsc_write(cfg, GSC_OUT_CON);
-
-	return 0;
 }
 
 static int gsc_dst_get_buf_seq(struct gsc_context *ctx)
@@ -1111,35 +905,16 @@ static int gsc_dst_get_buf_seq(struct gsc_context *ctx)
 	return buf_num;
 }
 
-static int gsc_dst_set_buf_seq(struct gsc_context *ctx, u32 buf_id,
-		enum drm_exynos_ipp_buf_type buf_type)
+static void gsc_dst_set_buf_seq(struct gsc_context *ctx, u32 buf_id,
+				bool enqueue)
 {
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	bool masked;
+	bool masked = !enqueue;
 	u32 cfg;
 	u32 mask = 0x00000001 << buf_id;
-	int ret = 0;
-
-	DRM_DEBUG_KMS("buf_id[%d]buf_type[%d]\n", buf_id, buf_type);
-
-	mutex_lock(&ctx->lock);
 
 	/* mask register set */
 	cfg = gsc_read(GSC_OUT_BASE_ADDR_Y_MASK);
 
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		masked = false;
-		break;
-	case IPP_BUF_DEQUEUE:
-		masked = true;
-		break;
-	default:
-		dev_err(ippdrv->dev, "invalid buf ctrl parameter.\n");
-		ret =  -EINVAL;
-		goto err_unlock;
-	}
-
 	/* sequence id */
 	cfg &= ~mask;
 	cfg |= masked << buf_id;
@@ -1148,94 +923,29 @@ static int gsc_dst_set_buf_seq(struct gsc_context *ctx, u32 buf_id,
 	gsc_write(cfg, GSC_OUT_BASE_ADDR_CR_MASK);
 
 	/* interrupt enable */
-	if (buf_type == IPP_BUF_ENQUEUE &&
-	    gsc_dst_get_buf_seq(ctx) >= GSC_BUF_START)
+	if (enqueue && gsc_dst_get_buf_seq(ctx) >= GSC_BUF_START)
 		gsc_handle_irq(ctx, true, false, true);
 
 	/* interrupt disable */
-	if (buf_type == IPP_BUF_DEQUEUE &&
-	    gsc_dst_get_buf_seq(ctx) <= GSC_BUF_STOP)
+	if (!enqueue && gsc_dst_get_buf_seq(ctx) <= GSC_BUF_STOP)
 		gsc_handle_irq(ctx, false, false, true);
-
-err_unlock:
-	mutex_unlock(&ctx->lock);
-	return ret;
 }
 
-static int gsc_dst_set_addr(struct device *dev,
-		struct drm_exynos_ipp_buf_info *buf_info, u32 buf_id,
-		enum drm_exynos_ipp_buf_type buf_type)
+static void gsc_dst_set_addr(struct gsc_context *ctx,
+			     u32 buf_id, struct exynos_drm_ipp_buffer *buf)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_property *property;
-
-	if (!c_node) {
-		DRM_ERROR("failed to get c_node.\n");
-		return -EFAULT;
-	}
-
-	property = &c_node->property;
-
-	DRM_DEBUG_KMS("prop_id[%d]buf_id[%d]buf_type[%d]\n",
-		property->prop_id, buf_id, buf_type);
-
-	if (buf_id > GSC_MAX_DST) {
-		dev_info(ippdrv->dev, "invalid buf_id %d.\n", buf_id);
-		return -EINVAL;
-	}
-
 	/* address register set */
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		gsc_write(buf_info->base[EXYNOS_DRM_PLANAR_Y],
-			GSC_OUT_BASE_ADDR_Y(buf_id));
-		gsc_write(buf_info->base[EXYNOS_DRM_PLANAR_CB],
-			GSC_OUT_BASE_ADDR_CB(buf_id));
-		gsc_write(buf_info->base[EXYNOS_DRM_PLANAR_CR],
-			GSC_OUT_BASE_ADDR_CR(buf_id));
-		break;
-	case IPP_BUF_DEQUEUE:
-		gsc_write(0x0, GSC_OUT_BASE_ADDR_Y(buf_id));
-		gsc_write(0x0, GSC_OUT_BASE_ADDR_CB(buf_id));
-		gsc_write(0x0, GSC_OUT_BASE_ADDR_CR(buf_id));
-		break;
-	default:
-		/* bypass */
-		break;
-	}
+	gsc_write(buf->dma_addr[0], GSC_OUT_BASE_ADDR_Y(buf_id));
+	gsc_write(buf->dma_addr[1], GSC_OUT_BASE_ADDR_CB(buf_id));
+	gsc_write(buf->dma_addr[2], GSC_OUT_BASE_ADDR_CR(buf_id));
 
-	return gsc_dst_set_buf_seq(ctx, buf_id, buf_type);
-}
-
-static struct exynos_drm_ipp_ops gsc_dst_ops = {
-	.set_fmt = gsc_dst_set_fmt,
-	.set_transf = gsc_dst_set_transf,
-	.set_size = gsc_dst_set_size,
-	.set_addr = gsc_dst_set_addr,
-};
-
-static int gsc_clk_ctrl(struct gsc_context *ctx, bool enable)
-{
-	DRM_DEBUG_KMS("enable[%d]\n", enable);
-
-	if (enable) {
-		clk_prepare_enable(ctx->gsc_clk);
-		ctx->suspended = false;
-	} else {
-		clk_disable_unprepare(ctx->gsc_clk);
-		ctx->suspended = true;
-	}
-
-	return 0;
+	gsc_dst_set_buf_seq(ctx, buf_id, true);
 }
 
 static int gsc_get_src_buf_index(struct gsc_context *ctx)
 {
 	u32 cfg, curr_index, i;
 	u32 buf_id = GSC_MAX_SRC;
-	int ret;
 
 	DRM_DEBUG_KMS("gsc id[%d]\n", ctx->id);
 
@@ -1249,19 +959,15 @@ static int gsc_get_src_buf_index(struct gsc_context *ctx)
 		}
 	}
 
+	DRM_DEBUG_KMS("cfg[0x%x]curr_index[%d]buf_id[%d]\n", cfg,
+		curr_index, buf_id);
+
 	if (buf_id == GSC_MAX_SRC) {
 		DRM_ERROR("failed to get in buffer index.\n");
 		return -EINVAL;
 	}
 
-	ret = gsc_src_set_buf_seq(ctx, buf_id, IPP_BUF_DEQUEUE);
-	if (ret < 0) {
-		DRM_ERROR("failed to dequeue.\n");
-		return ret;
-	}
-
-	DRM_DEBUG_KMS("cfg[0x%x]curr_index[%d]buf_id[%d]\n", cfg,
-		curr_index, buf_id);
+	gsc_src_set_buf_seq(ctx, buf_id, false);
 
 	return buf_id;
 }
@@ -1270,7 +976,6 @@ static int gsc_get_dst_buf_index(struct gsc_context *ctx)
 {
 	u32 cfg, curr_index, i;
 	u32 buf_id = GSC_MAX_DST;
-	int ret;
 
 	DRM_DEBUG_KMS("gsc id[%d]\n", ctx->id);
 
@@ -1289,11 +994,7 @@ static int gsc_get_dst_buf_index(struct gsc_context *ctx)
 		return -EINVAL;
 	}
 
-	ret = gsc_dst_set_buf_seq(ctx, buf_id, IPP_BUF_DEQUEUE);
-	if (ret < 0) {
-		DRM_ERROR("failed to dequeue.\n");
-		return ret;
-	}
+	gsc_dst_set_buf_seq(ctx, buf_id, false);
 
 	DRM_DEBUG_KMS("cfg[0x%x]curr_index[%d]buf_id[%d]\n", cfg,
 		curr_index, buf_id);
@@ -1304,215 +1005,55 @@ static int gsc_get_dst_buf_index(struct gsc_context *ctx)
 static irqreturn_t gsc_irq_handler(int irq, void *dev_id)
 {
 	struct gsc_context *ctx = dev_id;
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_event_work *event_work =
-		c_node->event_work;
 	u32 status;
-	int buf_id[EXYNOS_DRM_OPS_MAX];
+	int err = 0;
 
 	DRM_DEBUG_KMS("gsc id[%d]\n", ctx->id);
 
 	status = gsc_read(GSC_IRQ);
 	if (status & GSC_IRQ_STATUS_OR_IRQ) {
-		dev_err(ippdrv->dev, "occurred overflow at %d, status 0x%x.\n",
+		dev_err(ctx->dev, "occurred overflow at %d, status 0x%x.\n",
 			ctx->id, status);
-		return IRQ_NONE;
+		err = -EINVAL;
 	}
 
 	if (status & GSC_IRQ_STATUS_OR_FRM_DONE) {
-		dev_dbg(ippdrv->dev, "occurred frame done at %d, status 0x%x.\n",
-			ctx->id, status);
-
-		buf_id[EXYNOS_DRM_OPS_SRC] = gsc_get_src_buf_index(ctx);
-		if (buf_id[EXYNOS_DRM_OPS_SRC] < 0)
-			return IRQ_HANDLED;
-
-		buf_id[EXYNOS_DRM_OPS_DST] = gsc_get_dst_buf_index(ctx);
-		if (buf_id[EXYNOS_DRM_OPS_DST] < 0)
-			return IRQ_HANDLED;
-
-		DRM_DEBUG_KMS("buf_id_src[%d]buf_id_dst[%d]\n",
-			buf_id[EXYNOS_DRM_OPS_SRC], buf_id[EXYNOS_DRM_OPS_DST]);
-
-		event_work->ippdrv = ippdrv;
-		event_work->buf_id[EXYNOS_DRM_OPS_SRC] =
-			buf_id[EXYNOS_DRM_OPS_SRC];
-		event_work->buf_id[EXYNOS_DRM_OPS_DST] =
-			buf_id[EXYNOS_DRM_OPS_DST];
-		queue_work(ippdrv->event_workq, &event_work->work);
-	}
-
-	return IRQ_HANDLED;
-}
-
-static int gsc_init_prop_list(struct exynos_drm_ippdrv *ippdrv)
-{
-	struct drm_exynos_ipp_prop_list *prop_list = &ippdrv->prop_list;
-
-	prop_list->version = 1;
-	prop_list->writeback = 1;
-	prop_list->refresh_min = GSC_REFRESH_MIN;
-	prop_list->refresh_max = GSC_REFRESH_MAX;
-	prop_list->flip = (1 << EXYNOS_DRM_FLIP_VERTICAL) |
-				(1 << EXYNOS_DRM_FLIP_HORIZONTAL);
-	prop_list->degree = (1 << EXYNOS_DRM_DEGREE_0) |
-				(1 << EXYNOS_DRM_DEGREE_90) |
-				(1 << EXYNOS_DRM_DEGREE_180) |
-				(1 << EXYNOS_DRM_DEGREE_270);
-	prop_list->csc = 1;
-	prop_list->crop = 1;
-	prop_list->crop_max.hsize = GSC_CROP_MAX;
-	prop_list->crop_max.vsize = GSC_CROP_MAX;
-	prop_list->crop_min.hsize = GSC_CROP_MIN;
-	prop_list->crop_min.vsize = GSC_CROP_MIN;
-	prop_list->scale = 1;
-	prop_list->scale_max.hsize = GSC_SCALE_MAX;
-	prop_list->scale_max.vsize = GSC_SCALE_MAX;
-	prop_list->scale_min.hsize = GSC_SCALE_MIN;
-	prop_list->scale_min.vsize = GSC_SCALE_MIN;
-
-	return 0;
-}
-
-static inline bool gsc_check_drm_flip(enum drm_exynos_flip flip)
-{
-	switch (flip) {
-	case EXYNOS_DRM_FLIP_NONE:
-	case EXYNOS_DRM_FLIP_VERTICAL:
-	case EXYNOS_DRM_FLIP_HORIZONTAL:
-	case EXYNOS_DRM_FLIP_BOTH:
-		return true;
-	default:
-		DRM_DEBUG_KMS("invalid flip\n");
-		return false;
-	}
-}
-
-static int gsc_ippdrv_check_property(struct device *dev,
-		struct drm_exynos_ipp_property *property)
-{
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_prop_list *pp = &ippdrv->prop_list;
-	struct drm_exynos_ipp_config *config;
-	struct drm_exynos_pos *pos;
-	struct drm_exynos_sz *sz;
-	bool swap;
-	int i;
-
-	for_each_ipp_ops(i) {
-		if ((i == EXYNOS_DRM_OPS_SRC) &&
-			(property->cmd == IPP_CMD_WB))
-			continue;
+		int src_buf_id, dst_buf_id;
 
-		config = &property->config[i];
-		pos = &config->pos;
-		sz = &config->sz;
-
-		/* check for flip */
-		if (!gsc_check_drm_flip(config->flip)) {
-			DRM_ERROR("invalid flip.\n");
-			goto err_property;
-		}
-
-		/* check for degree */
-		switch (config->degree) {
-		case EXYNOS_DRM_DEGREE_90:
-		case EXYNOS_DRM_DEGREE_270:
-			swap = true;
-			break;
-		case EXYNOS_DRM_DEGREE_0:
-		case EXYNOS_DRM_DEGREE_180:
-			swap = false;
-			break;
-		default:
-			DRM_ERROR("invalid degree.\n");
-			goto err_property;
-		}
+		dev_dbg(ctx->dev, "occurred frame done at %d, status 0x%x.\n",
+			ctx->id, status);
 
-		/* check for buffer bound */
-		if ((pos->x + pos->w > sz->hsize) ||
-			(pos->y + pos->h > sz->vsize)) {
-			DRM_ERROR("out of buf bound.\n");
-			goto err_property;
-		}
+		src_buf_id = gsc_get_src_buf_index(ctx);
+		dst_buf_id = gsc_get_dst_buf_index(ctx);
 
-		/* check for crop */
-		if ((i == EXYNOS_DRM_OPS_SRC) && (pp->crop)) {
-			if (swap) {
-				if ((pos->h < pp->crop_min.hsize) ||
-					(sz->vsize > pp->crop_max.hsize) ||
-					(pos->w < pp->crop_min.vsize) ||
-					(sz->hsize > pp->crop_max.vsize)) {
-					DRM_ERROR("out of crop size.\n");
-					goto err_property;
-				}
-			} else {
-				if ((pos->w < pp->crop_min.hsize) ||
-					(sz->hsize > pp->crop_max.hsize) ||
-					(pos->h < pp->crop_min.vsize) ||
-					(sz->vsize > pp->crop_max.vsize)) {
-					DRM_ERROR("out of crop size.\n");
-					goto err_property;
-				}
-			}
-		}
+		DRM_DEBUG_KMS("buf_id_src[%d]buf_id_dst[%d]\n",	src_buf_id,
+			      dst_buf_id);
 
-		/* check for scale */
-		if ((i == EXYNOS_DRM_OPS_DST) && (pp->scale)) {
-			if (swap) {
-				if ((pos->h < pp->scale_min.hsize) ||
-					(sz->vsize > pp->scale_max.hsize) ||
-					(pos->w < pp->scale_min.vsize) ||
-					(sz->hsize > pp->scale_max.vsize)) {
-					DRM_ERROR("out of scale size.\n");
-					goto err_property;
-				}
-			} else {
-				if ((pos->w < pp->scale_min.hsize) ||
-					(sz->hsize > pp->scale_max.hsize) ||
-					(pos->h < pp->scale_min.vsize) ||
-					(sz->vsize > pp->scale_max.vsize)) {
-					DRM_ERROR("out of scale size.\n");
-					goto err_property;
-				}
-			}
-		}
+		if (src_buf_id < 0 || dst_buf_id < 0)
+			err = -EINVAL;
 	}
 
-	return 0;
-
-err_property:
-	for_each_ipp_ops(i) {
-		if ((i == EXYNOS_DRM_OPS_SRC) &&
-			(property->cmd == IPP_CMD_WB))
-			continue;
+	if (ctx->task) {
+		struct exynos_drm_ipp_task *task = ctx->task;
 
-		config = &property->config[i];
-		pos = &config->pos;
-		sz = &config->sz;
-
-		DRM_ERROR("[%s]f[%d]r[%d]pos[%d %d %d %d]sz[%d %d]\n",
-			i ? "dst" : "src", config->flip, config->degree,
-			pos->x, pos->y, pos->w, pos->h,
-			sz->hsize, sz->vsize);
+		ctx->task = NULL;
+		pm_runtime_mark_last_busy(ctx->dev);
+		pm_runtime_put_autosuspend(ctx->dev);
+		exynos_drm_ipp_task_done(task, err);
 	}
 
-	return -EINVAL;
+	return IRQ_HANDLED;
 }
 
-
-static int gsc_ippdrv_reset(struct device *dev)
+static int gsc_reset(struct gsc_context *ctx)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
 	struct gsc_scaler *sc = &ctx->sc;
 	int ret;
 
 	/* reset h/w block */
 	ret = gsc_sw_reset(ctx);
 	if (ret < 0) {
-		dev_err(dev, "failed to reset hardware.\n");
+		dev_err(ctx->dev, "failed to reset hardware.\n");
 		return ret;
 	}
 
@@ -1523,166 +1064,173 @@ static int gsc_ippdrv_reset(struct device *dev)
 	return 0;
 }
 
-static int gsc_ippdrv_start(struct device *dev, enum drm_exynos_ipp_cmd cmd)
+static void gsc_start(struct gsc_context *ctx)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_property *property;
-	struct drm_exynos_ipp_config *config;
-	struct drm_exynos_pos	img_pos[EXYNOS_DRM_OPS_MAX];
-	struct drm_exynos_ipp_set_wb set_wb;
 	u32 cfg;
-	int ret, i;
-
-	DRM_DEBUG_KMS("cmd[%d]\n", cmd);
-
-	if (!c_node) {
-		DRM_ERROR("failed to get c_node.\n");
-		return -EINVAL;
-	}
-
-	property = &c_node->property;
 
 	gsc_handle_irq(ctx, true, false, true);
 
-	for_each_ipp_ops(i) {
-		config = &property->config[i];
-		img_pos[i] = config->pos;
-	}
+	/* enable one shot */
+	cfg = gsc_read(GSC_ENABLE);
+	cfg &= ~(GSC_ENABLE_ON_CLEAR_MASK |
+		GSC_ENABLE_CLK_GATE_MODE_MASK);
+	cfg |= GSC_ENABLE_ON_CLEAR_ONESHOT;
+	gsc_write(cfg, GSC_ENABLE);
 
-	switch (cmd) {
-	case IPP_CMD_M2M:
-		/* enable one shot */
-		cfg = gsc_read(GSC_ENABLE);
-		cfg &= ~(GSC_ENABLE_ON_CLEAR_MASK |
-			GSC_ENABLE_CLK_GATE_MODE_MASK);
-		cfg |= GSC_ENABLE_ON_CLEAR_ONESHOT;
-		gsc_write(cfg, GSC_ENABLE);
-
-		/* src dma memory */
-		cfg = gsc_read(GSC_IN_CON);
-		cfg &= ~(GSC_IN_PATH_MASK | GSC_IN_LOCAL_SEL_MASK);
-		cfg |= GSC_IN_PATH_MEMORY;
-		gsc_write(cfg, GSC_IN_CON);
-
-		/* dst dma memory */
-		cfg = gsc_read(GSC_OUT_CON);
-		cfg |= GSC_OUT_PATH_MEMORY;
-		gsc_write(cfg, GSC_OUT_CON);
-		break;
-	case IPP_CMD_WB:
-		set_wb.enable = 1;
-		set_wb.refresh = property->refresh_rate;
-		gsc_set_gscblk_fimd_wb(ctx, set_wb.enable);
-		exynos_drm_ippnb_send_event(IPP_SET_WRITEBACK, (void *)&set_wb);
-
-		/* src local path */
-		cfg = gsc_read(GSC_IN_CON);
-		cfg &= ~(GSC_IN_PATH_MASK | GSC_IN_LOCAL_SEL_MASK);
-		cfg |= (GSC_IN_PATH_LOCAL | GSC_IN_LOCAL_FIMD_WB);
-		gsc_write(cfg, GSC_IN_CON);
-
-		/* dst dma memory */
-		cfg = gsc_read(GSC_OUT_CON);
-		cfg |= GSC_OUT_PATH_MEMORY;
-		gsc_write(cfg, GSC_OUT_CON);
-		break;
-	case IPP_CMD_OUTPUT:
-		/* src dma memory */
-		cfg = gsc_read(GSC_IN_CON);
-		cfg &= ~(GSC_IN_PATH_MASK | GSC_IN_LOCAL_SEL_MASK);
-		cfg |= GSC_IN_PATH_MEMORY;
-		gsc_write(cfg, GSC_IN_CON);
-
-		/* dst local path */
-		cfg = gsc_read(GSC_OUT_CON);
-		cfg |= GSC_OUT_PATH_MEMORY;
-		gsc_write(cfg, GSC_OUT_CON);
-		break;
-	default:
-		ret = -EINVAL;
-		dev_err(dev, "invalid operations.\n");
-		return ret;
-	}
+	/* src dma memory */
+	cfg = gsc_read(GSC_IN_CON);
+	cfg &= ~(GSC_IN_PATH_MASK | GSC_IN_LOCAL_SEL_MASK);
+	cfg |= GSC_IN_PATH_MEMORY;
+	gsc_write(cfg, GSC_IN_CON);
 
-	ret = gsc_set_prescaler(ctx, &ctx->sc,
-		&img_pos[EXYNOS_DRM_OPS_SRC],
-		&img_pos[EXYNOS_DRM_OPS_DST]);
-	if (ret) {
-		dev_err(dev, "failed to set prescaler.\n");
-		return ret;
-	}
+	/* dst dma memory */
+	cfg = gsc_read(GSC_OUT_CON);
+	cfg |= GSC_OUT_PATH_MEMORY;
+	gsc_write(cfg, GSC_OUT_CON);
 
 	gsc_set_scaler(ctx, &ctx->sc);
 
 	cfg = gsc_read(GSC_ENABLE);
 	cfg |= GSC_ENABLE_ON;
 	gsc_write(cfg, GSC_ENABLE);
+}
+
+static int gsc_commit(struct exynos_drm_ipp *ipp,
+			  struct exynos_drm_ipp_task *task)
+{
+	struct gsc_context *ctx = container_of(ipp, struct gsc_context, ipp);
+	int ret;
+
+	pm_runtime_get_sync(ctx->dev);
+	ctx->task = task;
+
+	ret = gsc_reset(ctx);
+	if (ret) {
+		pm_runtime_put_autosuspend(ctx->dev);
+		ctx->task = NULL;
+		return ret;
+	}
+
+	gsc_src_set_fmt(ctx, task->src.buf.fourcc);
+	gsc_src_set_transf(ctx, task->transform.rotation);
+	gsc_src_set_size(ctx, &task->src);
+	gsc_src_set_addr(ctx, 0, &task->src);
+	gsc_dst_set_fmt(ctx, task->dst.buf.fourcc);
+	gsc_dst_set_size(ctx, &task->dst);
+	gsc_dst_set_addr(ctx, 0, &task->dst);
+	gsc_set_prescaler(ctx, &ctx->sc, &task->src.rect, &task->dst.rect);
+	gsc_start(ctx);
 
 	return 0;
 }
 
-static void gsc_ippdrv_stop(struct device *dev, enum drm_exynos_ipp_cmd cmd)
+static void gsc_abort(struct exynos_drm_ipp *ipp,
+			  struct exynos_drm_ipp_task *task)
 {
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct drm_exynos_ipp_set_wb set_wb = {0, 0};
-	u32 cfg;
+	struct gsc_context *ctx =
+			container_of(ipp, struct gsc_context, ipp);
 
-	DRM_DEBUG_KMS("cmd[%d]\n", cmd);
+	gsc_reset(ctx);
+	if (ctx->task) {
+		struct exynos_drm_ipp_task *task = ctx->task;
 
-	switch (cmd) {
-	case IPP_CMD_M2M:
-		/* bypass */
-		break;
-	case IPP_CMD_WB:
-		gsc_set_gscblk_fimd_wb(ctx, set_wb.enable);
-		exynos_drm_ippnb_send_event(IPP_SET_WRITEBACK, (void *)&set_wb);
-		break;
-	case IPP_CMD_OUTPUT:
-	default:
-		dev_err(dev, "invalid operations.\n");
-		break;
+		ctx->task = NULL;
+		pm_runtime_mark_last_busy(ctx->dev);
+		pm_runtime_put_autosuspend(ctx->dev);
+		exynos_drm_ipp_task_done(task, -EIO);
 	}
+}
 
-	gsc_handle_irq(ctx, false, false, true);
+static struct exynos_drm_ipp_funcs ipp_funcs = {
+	.commit = gsc_commit,
+	.abort = gsc_abort,
+};
 
-	/* reset sequence */
-	gsc_write(0xff, GSC_OUT_BASE_ADDR_Y_MASK);
-	gsc_write(0xff, GSC_OUT_BASE_ADDR_CB_MASK);
-	gsc_write(0xff, GSC_OUT_BASE_ADDR_CR_MASK);
+static int gsc_bind(struct device *dev, struct device *master, void *data)
+{
+	struct gsc_context *ctx = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &ctx->ipp;
 
-	cfg = gsc_read(GSC_ENABLE);
-	cfg &= ~GSC_ENABLE_ON;
-	gsc_write(cfg, GSC_ENABLE);
+	ctx->drm_dev = drm_dev;
+	drm_iommu_attach_device(drm_dev, dev);
+
+	exynos_drm_ipp_register(drm_dev, ipp, &ipp_funcs,
+			DRM_EXYNOS_IPP_CAP_CROP | DRM_EXYNOS_IPP_CAP_ROTATE |
+			DRM_EXYNOS_IPP_CAP_SCALE | DRM_EXYNOS_IPP_CAP_CONVERT,
+			ctx->formats, ctx->num_formats, "gsc");
+
+	dev_info(dev, "The exynos gscaler has been probed successfully\n");
+
+	return 0;
+}
+
+static void gsc_unbind(struct device *dev, struct device *master,
+			void *data)
+{
+	struct gsc_context *ctx = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &ctx->ipp;
+
+	exynos_drm_ipp_unregister(drm_dev, ipp);
+	drm_iommu_detach_device(drm_dev, dev);
 }
 
+static const struct component_ops gsc_component_ops = {
+	.bind	= gsc_bind,
+	.unbind = gsc_unbind,
+};
+
+static const unsigned int gsc_formats[] = {
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XRGB8888, DRM_FORMAT_RGB565, DRM_FORMAT_BGRX8888,
+	DRM_FORMAT_NV12, DRM_FORMAT_NV16, DRM_FORMAT_NV21, DRM_FORMAT_NV61,
+	DRM_FORMAT_UYVY, DRM_FORMAT_VYUY, DRM_FORMAT_YUYV, DRM_FORMAT_YVYU,
+	DRM_FORMAT_YUV420, DRM_FORMAT_YVU420, DRM_FORMAT_YUV422,
+};
+
 static int gsc_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
+	struct gsc_driverdata *driver_data;
+	struct exynos_drm_ipp_formats *formats;
 	struct gsc_context *ctx;
 	struct resource *res;
-	struct exynos_drm_ippdrv *ippdrv;
-	int ret;
+	int ret, i;
 
 	ctx = devm_kzalloc(dev, sizeof(*ctx), GFP_KERNEL);
 	if (!ctx)
 		return -ENOMEM;
 
-	if (dev->of_node) {
-		ctx->sysreg = syscon_regmap_lookup_by_phandle(dev->of_node,
-							"samsung,sysreg");
-		if (IS_ERR(ctx->sysreg)) {
-			dev_warn(dev, "failed to get system register.\n");
-			ctx->sysreg = NULL;
-		}
+	formats = devm_kcalloc(dev,
+			       ARRAY_SIZE(gsc_formats), sizeof(*formats),
+			       GFP_KERNEL);
+	if (!formats)
+		return -ENOMEM;
+
+	driver_data = (struct gsc_driverdata *)of_device_get_match_data(dev);
+	ctx->dev = dev;
+	ctx->num_clocks = driver_data->num_clocks;
+	ctx->clk_names = driver_data->clk_names;
+
+	for (i = 0; i < ARRAY_SIZE(gsc_formats); i++) {
+		formats[i].fourcc = gsc_formats[i];
+		formats[i].type = DRM_EXYNOS_IPP_FORMAT_SOURCE |
+				  DRM_EXYNOS_IPP_FORMAT_DESTINATION;
+		formats[i].limits = driver_data->limits;
+		formats[i].num_limits = driver_data->num_limits;
 	}
+	ctx->formats = formats;
+	ctx->num_formats = ARRAY_SIZE(gsc_formats);
 
 	/* clock control */
-	ctx->gsc_clk = devm_clk_get(dev, "gscl");
-	if (IS_ERR(ctx->gsc_clk)) {
-		dev_err(dev, "failed to get gsc clock.\n");
-		return PTR_ERR(ctx->gsc_clk);
+	for (i = 0; i < ctx->num_clocks; i++) {
+		ctx->clocks[i] = devm_clk_get(dev, ctx->clk_names[i]);
+		if (IS_ERR(ctx->clocks[i])) {
+			dev_err(dev, "failed to get clock: %s\n",
+				ctx->clk_names[i]);
+			return PTR_ERR(ctx->clocks[i]);
+		}
 	}
 
 	/* resource memory */
@@ -1699,8 +1247,8 @@ static int gsc_probe(struct platform_device *pdev)
 	}
 
 	ctx->irq = res->start;
-	ret = devm_request_threaded_irq(dev, ctx->irq, NULL, gsc_irq_handler,
-		IRQF_ONESHOT, "drm_gsc", ctx);
+	ret = devm_request_irq(dev, ctx->irq, gsc_irq_handler, 0,
+			       dev_name(dev), ctx);
 	if (ret < 0) {
 		dev_err(dev, "failed to request irq.\n");
 		return ret;
@@ -1709,38 +1257,22 @@ static int gsc_probe(struct platform_device *pdev)
 	/* context initailization */
 	ctx->id = pdev->id;
 
-	ippdrv = &ctx->ippdrv;
-	ippdrv->dev = dev;
-	ippdrv->ops[EXYNOS_DRM_OPS_SRC] = &gsc_src_ops;
-	ippdrv->ops[EXYNOS_DRM_OPS_DST] = &gsc_dst_ops;
-	ippdrv->check_property = gsc_ippdrv_check_property;
-	ippdrv->reset = gsc_ippdrv_reset;
-	ippdrv->start = gsc_ippdrv_start;
-	ippdrv->stop = gsc_ippdrv_stop;
-	ret = gsc_init_prop_list(ippdrv);
-	if (ret < 0) {
-		dev_err(dev, "failed to init property list.\n");
-		return ret;
-	}
-
-	DRM_DEBUG_KMS("id[%d]ippdrv[%pK]\n", ctx->id, ippdrv);
-
-	mutex_init(&ctx->lock);
 	platform_set_drvdata(pdev, ctx);
 
+	pm_runtime_use_autosuspend(dev);
+	pm_runtime_set_autosuspend_delay(dev, GSC_AUTOSUSPEND_DELAY);
 	pm_runtime_enable(dev);
 
-	ret = exynos_drm_ippdrv_register(ippdrv);
-	if (ret < 0) {
-		dev_err(dev, "failed to register drm gsc device.\n");
-		goto err_ippdrv_register;
-	}
+	ret = component_add(dev, &gsc_component_ops);
+	if (ret)
+		goto err_pm_dis;
 
 	dev_info(dev, "drm gsc registered successfully.\n");
 
 	return 0;
 
-err_ippdrv_register:
+err_pm_dis:
+	pm_runtime_dont_use_autosuspend(dev);
 	pm_runtime_disable(dev);
 	return ret;
 }
@@ -1748,13 +1280,8 @@ err_ippdrv_register:
 static int gsc_remove(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
-	struct gsc_context *ctx = get_gsc_context(dev);
-	struct exynos_drm_ippdrv *ippdrv = &ctx->ippdrv;
 
-	exynos_drm_ippdrv_unregister(ippdrv);
-	mutex_destroy(&ctx->lock);
-
-	pm_runtime_set_suspended(dev);
+	pm_runtime_dont_use_autosuspend(dev);
 	pm_runtime_disable(dev);
 
 	return 0;
@@ -1763,19 +1290,32 @@ static int gsc_remove(struct platform_device *pdev)
 static int __maybe_unused gsc_runtime_suspend(struct device *dev)
 {
 	struct gsc_context *ctx = get_gsc_context(dev);
+	int i;
 
 	DRM_DEBUG_KMS("id[%d]\n", ctx->id);
 
-	return  gsc_clk_ctrl(ctx, false);
+	for (i = ctx->num_clocks - 1; i >= 0; i--)
+		clk_disable_unprepare(ctx->clocks[i]);
+
+	return 0;
 }
 
 static int __maybe_unused gsc_runtime_resume(struct device *dev)
 {
 	struct gsc_context *ctx = get_gsc_context(dev);
+	int i, ret;
 
 	DRM_DEBUG_KMS("id[%d]\n", ctx->id);
 
-	return  gsc_clk_ctrl(ctx, true);
+	for (i = 0; i < ctx->num_clocks; i++) {
+		ret = clk_prepare_enable(ctx->clocks[i]);
+		if (ret) {
+			while (--i > 0)
+				clk_disable_unprepare(ctx->clocks[i]);
+			return ret;
+		}
+	}
+	return 0;
 }
 
 static const struct dev_pm_ops gsc_pm_ops = {
@@ -1784,9 +1324,66 @@ static const struct dev_pm_ops gsc_pm_ops = {
 	SET_RUNTIME_PM_OPS(gsc_runtime_suspend, gsc_runtime_resume, NULL)
 };
 
+static const struct drm_exynos_ipp_limit gsc_5250_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 32, 4800, 8 }, .v = { 16, 3344, 8 }) },
+	{ IPP_SIZE_LIMIT(AREA, .h = { 16, 4800, 2 }, .v = { 8, 3344, 2 }) },
+	{ IPP_SIZE_LIMIT(ROTATED, .h = { 32, 2048 }, .v = { 16, 2048 }) },
+	{ IPP_SCALE_LIMIT(.h = { (1 << 16) / 16, (1 << 16) * 8 },
+			  .v = { (1 << 16) / 16, (1 << 16) * 8 }) },
+};
+
+static const struct drm_exynos_ipp_limit gsc_5420_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 32, 4800, 8 }, .v = { 16, 3344, 8 }) },
+	{ IPP_SIZE_LIMIT(AREA, .h = { 16, 4800, 2 }, .v = { 8, 3344, 2 }) },
+	{ IPP_SIZE_LIMIT(ROTATED, .h = { 16, 2016 }, .v = { 8, 2016 }) },
+	{ IPP_SCALE_LIMIT(.h = { (1 << 16) / 16, (1 << 16) * 8 },
+			  .v = { (1 << 16) / 16, (1 << 16) * 8 }) },
+};
+
+static const struct drm_exynos_ipp_limit gsc_5433_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 32, 8191, 2 }, .v = { 16, 8191, 2 }) },
+	{ IPP_SIZE_LIMIT(AREA, .h = { 16, 4800, 1 }, .v = { 8, 3344, 1 }) },
+	{ IPP_SIZE_LIMIT(ROTATED, .h = { 32, 2047 }, .v = { 8, 8191 }) },
+	{ IPP_SCALE_LIMIT(.h = { (1 << 16) / 16, (1 << 16) * 8 },
+			  .v = { (1 << 16) / 16, (1 << 16) * 8 }) },
+};
+
+static struct gsc_driverdata gsc_exynos5250_drvdata = {
+	.clk_names = {"gscl"},
+	.num_clocks = 1,
+	.limits = gsc_5250_limits,
+	.num_limits = ARRAY_SIZE(gsc_5250_limits),
+};
+
+static struct gsc_driverdata gsc_exynos5420_drvdata = {
+	.clk_names = {"gscl"},
+	.num_clocks = 1,
+	.limits = gsc_5420_limits,
+	.num_limits = ARRAY_SIZE(gsc_5420_limits),
+};
+
+static struct gsc_driverdata gsc_exynos5433_drvdata = {
+	.clk_names = {"pclk", "aclk", "aclk_xiu", "aclk_gsclbend"},
+	.num_clocks = 4,
+	.limits = gsc_5433_limits,
+	.num_limits = ARRAY_SIZE(gsc_5433_limits),
+};
+
 static const struct of_device_id exynos_drm_gsc_of_match[] = {
-	{ .compatible = "samsung,exynos5-gsc" },
-	{ },
+	{
+		.compatible = "samsung,exynos5-gsc",
+		.data = &gsc_exynos5250_drvdata,
+	}, {
+		.compatible = "samsung,exynos5250-gsc",
+		.data = &gsc_exynos5250_drvdata,
+	}, {
+		.compatible = "samsung,exynos5420-gsc",
+		.data = &gsc_exynos5420_drvdata,
+	}, {
+		.compatible = "samsung,exynos5433-gsc",
+		.data = &gsc_exynos5433_drvdata,
+	}, {
+	},
 };
 MODULE_DEVICE_TABLE(of, exynos_drm_gsc_of_match);
 
@@ -1800,4 +1397,3 @@ struct platform_driver gsc_driver = {
 		.of_match_table = of_match_ptr(exynos_drm_gsc_of_match),
 	},
 };
-
diff --git a/drivers/gpu/drm/exynos/exynos_drm_gsc.h b/drivers/gpu/drm/exynos/exynos_drm_gsc.h
deleted file mode 100644
index 29ec1c5efcf2..000000000000
--- a/drivers/gpu/drm/exynos/exynos_drm_gsc.h
+++ /dev/null
@@ -1,24 +0,0 @@
-/*
- * Copyright (c) 2012 Samsung Electronics Co., Ltd.
- *
- * Authors:
- *	Eunchul Kim <chulspro.kim@samsung.com>
- *	Jinyoung Jeon <jy0.jeon@samsung.com>
- *	Sangmin Lee <lsmin.lee@samsung.com>
- *
- * This program is free software; you can redistribute  it and/or modify it
- * under  the terms of  the GNU General  Public License as published by the
- * Free Software Foundation;  either version 2 of the  License, or (at your
- * option) any later version.
- */
-
-#ifndef _EXYNOS_DRM_GSC_H_
-#define _EXYNOS_DRM_GSC_H_
-
-/*
- * TODO
- * FIMD output interface notifier callback.
- * Mixer output interface notifier callback.
- */
-
-#endif /* _EXYNOS_DRM_GSC_H_ */
diff --git a/drivers/gpu/drm/exynos/exynos_drm_ipp.c b/drivers/gpu/drm/exynos/exynos_drm_ipp.c
new file mode 100644
index 000000000000..26374e58c557
--- /dev/null
+++ b/drivers/gpu/drm/exynos/exynos_drm_ipp.c
@@ -0,0 +1,916 @@
+/*
+ * Copyright (C) 2017 Samsung Electronics Co.Ltd
+ * Authors:
+ *	Marek Szyprowski <m.szyprowski@samsung.com>
+ *
+ * Exynos DRM Image Post Processing (IPP) related functions
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ */
+
+
+#include <drm/drmP.h>
+#include <drm/drm_mode.h>
+#include <uapi/drm/exynos_drm.h>
+
+#include "exynos_drm_drv.h"
+#include "exynos_drm_gem.h"
+#include "exynos_drm_ipp.h"
+
+static int num_ipp;
+static LIST_HEAD(ipp_list);
+
+/**
+ * exynos_drm_ipp_register - Register a new picture processor hardware module
+ * @dev: DRM device
+ * @ipp: ipp module to init
+ * @funcs: callbacks for the new ipp object
+ * @caps: bitmask of ipp capabilities (%DRM_EXYNOS_IPP_CAP_*)
+ * @formats: array of supported formats
+ * @num_formats: size of the supported formats array
+ * @name: name (for debugging purposes)
+ *
+ * Initializes a ipp module.
+ *
+ * Returns:
+ * Zero on success, error code on failure.
+ */
+int exynos_drm_ipp_register(struct drm_device *dev, struct exynos_drm_ipp *ipp,
+		const struct exynos_drm_ipp_funcs *funcs, unsigned int caps,
+		const struct exynos_drm_ipp_formats *formats,
+		unsigned int num_formats, const char *name)
+{
+	WARN_ON(!ipp);
+	WARN_ON(!funcs);
+	WARN_ON(!formats);
+	WARN_ON(!num_formats);
+
+	spin_lock_init(&ipp->lock);
+	INIT_LIST_HEAD(&ipp->todo_list);
+	init_waitqueue_head(&ipp->done_wq);
+	ipp->dev = dev;
+	ipp->funcs = funcs;
+	ipp->capabilities = caps;
+	ipp->name = name;
+	ipp->formats = formats;
+	ipp->num_formats = num_formats;
+
+	/* ipp_list modification is serialized by component framework */
+	list_add_tail(&ipp->head, &ipp_list);
+	ipp->id = num_ipp++;
+
+	DRM_DEBUG_DRIVER("Registered ipp %d\n", ipp->id);
+
+	return 0;
+}
+
+/**
+ * exynos_drm_ipp_unregister - Unregister the picture processor module
+ * @dev: DRM device
+ * @ipp: ipp module
+ */
+void exynos_drm_ipp_unregister(struct drm_device *dev,
+			       struct exynos_drm_ipp *ipp)
+{
+	WARN_ON(ipp->task);
+	WARN_ON(!list_empty(&ipp->todo_list));
+	list_del(&ipp->head);
+}
+
+/**
+ * exynos_drm_ipp_ioctl_get_res_ioctl - enumerate all ipp modules
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * Construct a list of ipp ids.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int exynos_drm_ipp_get_res_ioctl(struct drm_device *dev, void *data,
+				 struct drm_file *file_priv)
+{
+	struct drm_exynos_ioctl_ipp_get_res *resp = data;
+	struct exynos_drm_ipp *ipp;
+	uint32_t __user *ipp_ptr = (uint32_t __user *)
+						(unsigned long)resp->ipp_id_ptr;
+	unsigned int count = num_ipp, copied = 0;
+
+	/*
+	 * This ioctl is called twice, once to determine how much space is
+	 * needed, and the 2nd time to fill it.
+	 */
+	if (count && resp->count_ipps >= count) {
+		list_for_each_entry(ipp, &ipp_list, head) {
+			if (put_user(ipp->id, ipp_ptr + copied))
+				return -EFAULT;
+			copied++;
+		}
+	}
+	resp->count_ipps = count;
+
+	return 0;
+}
+
+static inline struct exynos_drm_ipp *__ipp_get(uint32_t id)
+{
+	struct exynos_drm_ipp *ipp;
+
+	list_for_each_entry(ipp, &ipp_list, head)
+		if (ipp->id == id)
+			return ipp;
+	return NULL;
+}
+
+/**
+ * exynos_drm_ipp_ioctl_get_caps - get ipp module capabilities and formats
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * Construct a structure describing ipp module capabilities.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int exynos_drm_ipp_get_caps_ioctl(struct drm_device *dev, void *data,
+				  struct drm_file *file_priv)
+{
+	struct drm_exynos_ioctl_ipp_get_caps *resp = data;
+	void __user *ptr = (void __user *)(unsigned long)resp->formats_ptr;
+	struct exynos_drm_ipp *ipp;
+	int i;
+
+	ipp = __ipp_get(resp->ipp_id);
+	if (!ipp)
+		return -ENOENT;
+
+	resp->ipp_id = ipp->id;
+	resp->capabilities = ipp->capabilities;
+
+	/*
+	 * This ioctl is called twice, once to determine how much space is
+	 * needed, and the 2nd time to fill it.
+	 */
+	if (resp->formats_count >= ipp->num_formats) {
+		for (i = 0; i < ipp->num_formats; i++) {
+			struct drm_exynos_ipp_format tmp = {
+				.fourcc = ipp->formats[i].fourcc,
+				.type = ipp->formats[i].type,
+				.modifier = ipp->formats[i].modifier,
+			};
+
+			if (copy_to_user(ptr, &tmp, sizeof(tmp)))
+				return -EFAULT;
+			ptr += sizeof(tmp);
+		}
+	}
+	resp->formats_count = ipp->num_formats;
+
+	return 0;
+}
+
+static inline const struct exynos_drm_ipp_formats *__ipp_format_get(
+				struct exynos_drm_ipp *ipp, uint32_t fourcc,
+				uint64_t mod, unsigned int type)
+{
+	int i;
+
+	for (i = 0; i < ipp->num_formats; i++) {
+		if ((ipp->formats[i].type & type) &&
+		    ipp->formats[i].fourcc == fourcc &&
+		    ipp->formats[i].modifier == mod)
+			return &ipp->formats[i];
+	}
+	return NULL;
+}
+
+/**
+ * exynos_drm_ipp_get_limits_ioctl - get ipp module limits
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * Construct a structure describing ipp module limitations for provided
+ * picture format.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int exynos_drm_ipp_get_limits_ioctl(struct drm_device *dev, void *data,
+				    struct drm_file *file_priv)
+{
+	struct drm_exynos_ioctl_ipp_get_limits *resp = data;
+	void __user *ptr = (void __user *)(unsigned long)resp->limits_ptr;
+	const struct exynos_drm_ipp_formats *format;
+	struct exynos_drm_ipp *ipp;
+
+	if (resp->type != DRM_EXYNOS_IPP_FORMAT_SOURCE &&
+	    resp->type != DRM_EXYNOS_IPP_FORMAT_DESTINATION)
+		return -EINVAL;
+
+	ipp = __ipp_get(resp->ipp_id);
+	if (!ipp)
+		return -ENOENT;
+
+	format = __ipp_format_get(ipp, resp->fourcc, resp->modifier,
+				  resp->type);
+	if (!format)
+		return -EINVAL;
+
+	/*
+	 * This ioctl is called twice, once to determine how much space is
+	 * needed, and the 2nd time to fill it.
+	 */
+	if (format->num_limits && resp->limits_count >= format->num_limits)
+		if (copy_to_user((void __user *)ptr, format->limits,
+				 sizeof(*format->limits) * format->num_limits))
+			return -EFAULT;
+	resp->limits_count = format->num_limits;
+
+	return 0;
+}
+
+struct drm_pending_exynos_ipp_event {
+	struct drm_pending_event base;
+	struct drm_exynos_ipp_event event;
+};
+
+static inline struct exynos_drm_ipp_task *
+			exynos_drm_ipp_task_alloc(struct exynos_drm_ipp *ipp)
+{
+	struct exynos_drm_ipp_task *task;
+
+	task = kzalloc(sizeof(*task), GFP_KERNEL);
+	if (!task)
+		return NULL;
+
+	task->dev = ipp->dev;
+	task->ipp = ipp;
+
+	/* some defaults */
+	task->src.rect.w = task->dst.rect.w = UINT_MAX;
+	task->src.rect.h = task->dst.rect.h = UINT_MAX;
+	task->transform.rotation = DRM_MODE_ROTATE_0;
+
+	DRM_DEBUG_DRIVER("Allocated task %pK\n", task);
+
+	return task;
+}
+
+static const struct exynos_drm_param_map {
+	unsigned int id;
+	unsigned int size;
+	unsigned int offset;
+} exynos_drm_ipp_params_maps[] = {
+	{
+		DRM_EXYNOS_IPP_TASK_BUFFER | DRM_EXYNOS_IPP_TASK_TYPE_SOURCE,
+		sizeof(struct drm_exynos_ipp_task_buffer),
+		offsetof(struct exynos_drm_ipp_task, src.buf),
+	}, {
+		DRM_EXYNOS_IPP_TASK_BUFFER |
+			DRM_EXYNOS_IPP_TASK_TYPE_DESTINATION,
+		sizeof(struct drm_exynos_ipp_task_buffer),
+		offsetof(struct exynos_drm_ipp_task, dst.buf),
+	}, {
+		DRM_EXYNOS_IPP_TASK_RECTANGLE | DRM_EXYNOS_IPP_TASK_TYPE_SOURCE,
+		sizeof(struct drm_exynos_ipp_task_rect),
+		offsetof(struct exynos_drm_ipp_task, src.rect),
+	}, {
+		DRM_EXYNOS_IPP_TASK_RECTANGLE |
+			DRM_EXYNOS_IPP_TASK_TYPE_DESTINATION,
+		sizeof(struct drm_exynos_ipp_task_rect),
+		offsetof(struct exynos_drm_ipp_task, dst.rect),
+	}, {
+		DRM_EXYNOS_IPP_TASK_TRANSFORM,
+		sizeof(struct drm_exynos_ipp_task_transform),
+		offsetof(struct exynos_drm_ipp_task, transform),
+	}, {
+		DRM_EXYNOS_IPP_TASK_ALPHA,
+		sizeof(struct drm_exynos_ipp_task_alpha),
+		offsetof(struct exynos_drm_ipp_task, alpha),
+	},
+};
+
+static int exynos_drm_ipp_task_set(struct exynos_drm_ipp_task *task,
+				   struct drm_exynos_ioctl_ipp_commit *arg)
+{
+	const struct exynos_drm_param_map *map = exynos_drm_ipp_params_maps;
+	void __user *params = (void __user *)(unsigned long)arg->params_ptr;
+	unsigned int size = arg->params_size;
+	uint32_t id;
+	int i;
+
+	while (size) {
+		if (get_user(id, (uint32_t __user *)params))
+			return -EFAULT;
+
+		for (i = 0; i < ARRAY_SIZE(exynos_drm_ipp_params_maps); i++)
+			if (map[i].id == id)
+				break;
+		if (i == ARRAY_SIZE(exynos_drm_ipp_params_maps) ||
+		    map[i].size > size)
+			return -EINVAL;
+
+		if (copy_from_user((void *)task + map[i].offset, params,
+				   map[i].size))
+			return -EFAULT;
+
+		params += map[i].size;
+		size -= map[i].size;
+	}
+
+	DRM_DEBUG_DRIVER("Got task %pK configuration from userspace\n", task);
+	return 0;
+}
+
+static int exynos_drm_ipp_task_setup_buffer(struct exynos_drm_ipp_buffer *buf,
+					    struct drm_file *filp)
+{
+	int ret = 0;
+	int i;
+
+	/* basic checks */
+	if (buf->buf.width == 0 || buf->buf.height == 0)
+		return -EINVAL;
+	buf->format = drm_format_info(buf->buf.fourcc);
+	for (i = 0; i < buf->format->num_planes; i++) {
+		unsigned int width = (i == 0) ? buf->buf.width :
+			     DIV_ROUND_UP(buf->buf.width, buf->format->hsub);
+
+		if (buf->buf.pitch[i] == 0)
+			buf->buf.pitch[i] = width * buf->format->cpp[i];
+		if (buf->buf.pitch[i] < width * buf->format->cpp[i])
+			return -EINVAL;
+		if (!buf->buf.gem_id[i])
+			return -ENOENT;
+	}
+
+	/* pitch for additional planes must match */
+	if (buf->format->num_planes > 2 &&
+	    buf->buf.pitch[1] != buf->buf.pitch[2])
+		return -EINVAL;
+
+	/* get GEM buffers and check their size */
+	for (i = 0; i < buf->format->num_planes; i++) {
+		unsigned int height = (i == 0) ? buf->buf.height :
+			     DIV_ROUND_UP(buf->buf.height, buf->format->vsub);
+		unsigned long size = height * buf->buf.pitch[i];
+		struct drm_gem_object *obj = drm_gem_object_lookup(filp,
+							    buf->buf.gem_id[i]);
+		if (!obj) {
+			ret = -ENOENT;
+			goto gem_free;
+		}
+		buf->exynos_gem[i] = to_exynos_gem(obj);
+
+		if (size + buf->buf.offset[i] > buf->exynos_gem[i]->size) {
+			i++;
+			ret = -EINVAL;
+			goto gem_free;
+		}
+		buf->dma_addr[i] = buf->exynos_gem[i]->dma_addr +
+				   buf->buf.offset[i];
+	}
+
+	return 0;
+gem_free:
+	while (i--) {
+		drm_gem_object_put_unlocked(&buf->exynos_gem[i]->base);
+		buf->exynos_gem[i] = NULL;
+	}
+	return ret;
+}
+
+static void exynos_drm_ipp_task_release_buf(struct exynos_drm_ipp_buffer *buf)
+{
+	int i;
+
+	if (!buf->exynos_gem[0])
+		return;
+	for (i = 0; i < buf->format->num_planes; i++)
+		drm_gem_object_put_unlocked(&buf->exynos_gem[i]->base);
+}
+
+static void exynos_drm_ipp_task_free(struct exynos_drm_ipp *ipp,
+				 struct exynos_drm_ipp_task *task)
+{
+	DRM_DEBUG_DRIVER("Freeing task %pK\n", task);
+
+	exynos_drm_ipp_task_release_buf(&task->src);
+	exynos_drm_ipp_task_release_buf(&task->dst);
+	if (task->event)
+		drm_event_cancel_free(ipp->dev, &task->event->base);
+	kfree(task);
+}
+
+struct drm_ipp_limit {
+	struct drm_exynos_ipp_limit_val h;
+	struct drm_exynos_ipp_limit_val v;
+};
+
+enum drm_ipp_size_id {
+	IPP_LIMIT_BUFFER, IPP_LIMIT_AREA, IPP_LIMIT_ROTATED, IPP_LIMIT_MAX
+};
+
+static const enum drm_ipp_size_id limit_id_fallback[IPP_LIMIT_MAX][4] = {
+	[IPP_LIMIT_BUFFER]  = { DRM_EXYNOS_IPP_LIMIT_SIZE_BUFFER },
+	[IPP_LIMIT_AREA]    = { DRM_EXYNOS_IPP_LIMIT_SIZE_AREA,
+				DRM_EXYNOS_IPP_LIMIT_SIZE_BUFFER },
+	[IPP_LIMIT_ROTATED] = { DRM_EXYNOS_IPP_LIMIT_SIZE_ROTATED,
+				DRM_EXYNOS_IPP_LIMIT_SIZE_AREA,
+				DRM_EXYNOS_IPP_LIMIT_SIZE_BUFFER },
+};
+
+static inline void __limit_set_val(unsigned int *ptr, unsigned int val)
+{
+	if (!*ptr)
+		*ptr = val;
+}
+
+static void __get_size_limit(const struct drm_exynos_ipp_limit *limits,
+			     unsigned int num_limits, enum drm_ipp_size_id id,
+			     struct drm_ipp_limit *res)
+{
+	const struct drm_exynos_ipp_limit *l = limits;
+	int i = 0;
+
+	memset(res, 0, sizeof(*res));
+	for (i = 0; limit_id_fallback[id][i]; i++)
+		for (l = limits; l - limits < num_limits; l++) {
+			if (((l->type & DRM_EXYNOS_IPP_LIMIT_TYPE_MASK) !=
+			      DRM_EXYNOS_IPP_LIMIT_TYPE_SIZE) ||
+			    ((l->type & DRM_EXYNOS_IPP_LIMIT_SIZE_MASK) !=
+						     limit_id_fallback[id][i]))
+				continue;
+			__limit_set_val(&res->h.min, l->h.min);
+			__limit_set_val(&res->h.max, l->h.max);
+			__limit_set_val(&res->h.align, l->h.align);
+			__limit_set_val(&res->v.min, l->v.min);
+			__limit_set_val(&res->v.max, l->v.max);
+			__limit_set_val(&res->v.align, l->v.align);
+		}
+}
+
+static inline bool __align_check(unsigned int val, unsigned int align)
+{
+	if (align && (val & (align - 1))) {
+		DRM_DEBUG_DRIVER("Value %d exceeds HW limits (align %d)\n",
+				 val, align);
+		return false;
+	}
+	return true;
+}
+
+static inline bool __size_limit_check(unsigned int val,
+				 struct drm_exynos_ipp_limit_val *l)
+{
+	if ((l->min && val < l->min) || (l->max && val > l->max)) {
+		DRM_DEBUG_DRIVER("Value %d exceeds HW limits (min %d, max %d)\n",
+				 val, l->min, l->max);
+		return false;
+	}
+	return __align_check(val, l->align);
+}
+
+static int exynos_drm_ipp_check_size_limits(struct exynos_drm_ipp_buffer *buf,
+	const struct drm_exynos_ipp_limit *limits, unsigned int num_limits,
+	bool rotate, bool swap)
+{
+	enum drm_ipp_size_id id = rotate ? IPP_LIMIT_ROTATED : IPP_LIMIT_AREA;
+	struct drm_ipp_limit l;
+	struct drm_exynos_ipp_limit_val *lh = &l.h, *lv = &l.v;
+
+	if (!limits)
+		return 0;
+
+	__get_size_limit(limits, num_limits, IPP_LIMIT_BUFFER, &l);
+	if (!__size_limit_check(buf->buf.width, &l.h) ||
+	    !__size_limit_check(buf->buf.height, &l.v))
+		return -EINVAL;
+
+	if (swap) {
+		lv = &l.h;
+		lh = &l.v;
+	}
+	__get_size_limit(limits, num_limits, id, &l);
+	if (!__size_limit_check(buf->rect.w, lh) ||
+	    !__align_check(buf->rect.x, lh->align) ||
+	    !__size_limit_check(buf->rect.h, lv) ||
+	    !__align_check(buf->rect.y, lv->align))
+		return -EINVAL;
+
+	return 0;
+}
+
+static inline bool __scale_limit_check(unsigned int src, unsigned int dst,
+				       unsigned int min, unsigned int max)
+{
+	if ((max && (dst << 16) > src * max) ||
+	    (min && (dst << 16) < src * min)) {
+		DRM_DEBUG_DRIVER("Scale from %d to %d exceeds HW limits (ratio min %d.%05d, max %d.%05d)\n",
+			 src, dst,
+			 min >> 16, 100000 * (min & 0xffff) / (1 << 16),
+			 max >> 16, 100000 * (max & 0xffff) / (1 << 16));
+		return false;
+	}
+	return true;
+}
+
+static int exynos_drm_ipp_check_scale_limits(
+				struct drm_exynos_ipp_task_rect *src,
+				struct drm_exynos_ipp_task_rect *dst,
+				const struct drm_exynos_ipp_limit *limits,
+				unsigned int num_limits, bool swap)
+{
+	const struct drm_exynos_ipp_limit_val *lh, *lv;
+	int dw, dh;
+
+	for (; num_limits; limits++, num_limits--)
+		if ((limits->type & DRM_EXYNOS_IPP_LIMIT_TYPE_MASK) ==
+		    DRM_EXYNOS_IPP_LIMIT_TYPE_SCALE)
+			break;
+	if (!num_limits)
+		return 0;
+
+	lh = (!swap) ? &limits->h : &limits->v;
+	lv = (!swap) ? &limits->v : &limits->h;
+	dw = (!swap) ? dst->w : dst->h;
+	dh = (!swap) ? dst->h : dst->w;
+
+	if (!__scale_limit_check(src->w, dw, lh->min, lh->max) ||
+	    !__scale_limit_check(src->h, dh, lv->min, lv->max))
+		return -EINVAL;
+
+	return 0;
+}
+
+static int exynos_drm_ipp_task_check(struct exynos_drm_ipp_task *task)
+{
+	struct exynos_drm_ipp *ipp = task->ipp;
+	const struct exynos_drm_ipp_formats *src_fmt, *dst_fmt;
+	struct exynos_drm_ipp_buffer *src = &task->src, *dst = &task->dst;
+	unsigned int rotation = task->transform.rotation;
+	int ret = 0;
+	bool swap = drm_rotation_90_or_270(rotation);
+	bool rotate = (rotation != DRM_MODE_ROTATE_0);
+	bool scale = false;
+
+	DRM_DEBUG_DRIVER("Checking task %pK\n", task);
+
+	if (src->rect.w == UINT_MAX)
+		src->rect.w = src->buf.width;
+	if (src->rect.h == UINT_MAX)
+		src->rect.h = src->buf.height;
+	if (dst->rect.w == UINT_MAX)
+		dst->rect.w = dst->buf.width;
+	if (dst->rect.h == UINT_MAX)
+		dst->rect.h = dst->buf.height;
+
+	if (src->rect.x + src->rect.w > (src->buf.width) ||
+	    src->rect.y + src->rect.h > (src->buf.height) ||
+	    dst->rect.x + dst->rect.w > (dst->buf.width) ||
+	    dst->rect.y + dst->rect.h > (dst->buf.height)) {
+		DRM_DEBUG_DRIVER("Task %pK: defined area is outside provided buffers\n",
+				 task);
+		return -EINVAL;
+	}
+
+	if ((!swap && (src->rect.w != dst->rect.w ||
+		       src->rect.h != dst->rect.h)) ||
+	    (swap && (src->rect.w != dst->rect.h ||
+		      src->rect.h != dst->rect.w)))
+		scale = true;
+
+	if ((!(ipp->capabilities & DRM_EXYNOS_IPP_CAP_CROP) &&
+	     (src->rect.x || src->rect.y || dst->rect.x || dst->rect.y)) ||
+	    (!(ipp->capabilities & DRM_EXYNOS_IPP_CAP_ROTATE) && rotate) ||
+	    (!(ipp->capabilities & DRM_EXYNOS_IPP_CAP_SCALE) && scale) ||
+	    (!(ipp->capabilities & DRM_EXYNOS_IPP_CAP_CONVERT) &&
+	     src->buf.fourcc != dst->buf.fourcc)) {
+		DRM_DEBUG_DRIVER("Task %pK: hw capabilities exceeded\n", task);
+		return -EINVAL;
+	}
+
+	src_fmt = __ipp_format_get(ipp, src->buf.fourcc, src->buf.modifier,
+				   DRM_EXYNOS_IPP_FORMAT_SOURCE);
+	if (!src_fmt) {
+		DRM_DEBUG_DRIVER("Task %pK: src format not supported\n", task);
+		return -EINVAL;
+	}
+	ret = exynos_drm_ipp_check_size_limits(src, src_fmt->limits,
+					       src_fmt->num_limits,
+					       rotate, false);
+	if (ret)
+		return ret;
+	ret = exynos_drm_ipp_check_scale_limits(&src->rect, &dst->rect,
+						src_fmt->limits,
+						src_fmt->num_limits, swap);
+	if (ret)
+		return ret;
+
+	dst_fmt = __ipp_format_get(ipp, dst->buf.fourcc, dst->buf.modifier,
+				   DRM_EXYNOS_IPP_FORMAT_DESTINATION);
+	if (!dst_fmt) {
+		DRM_DEBUG_DRIVER("Task %pK: dst format not supported\n", task);
+		return -EINVAL;
+	}
+	ret = exynos_drm_ipp_check_size_limits(dst, dst_fmt->limits,
+					       dst_fmt->num_limits,
+					       false, swap);
+	if (ret)
+		return ret;
+	ret = exynos_drm_ipp_check_scale_limits(&src->rect, &dst->rect,
+						dst_fmt->limits,
+						dst_fmt->num_limits, swap);
+	if (ret)
+		return ret;
+
+	DRM_DEBUG_DRIVER("Task %pK: all checks done.\n", task);
+
+	return ret;
+}
+
+static int exynos_drm_ipp_task_setup_buffers(struct exynos_drm_ipp_task *task,
+				     struct drm_file *filp)
+{
+	struct exynos_drm_ipp_buffer *src = &task->src, *dst = &task->dst;
+	int ret = 0;
+
+	DRM_DEBUG_DRIVER("Setting buffer for task %pK\n", task);
+
+	ret = exynos_drm_ipp_task_setup_buffer(src, filp);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Task %pK: src buffer setup failed\n", task);
+		return ret;
+	}
+	ret = exynos_drm_ipp_task_setup_buffer(dst, filp);
+	if (ret) {
+		DRM_DEBUG_DRIVER("Task %pK: dst buffer setup failed\n", task);
+		return ret;
+	}
+
+	DRM_DEBUG_DRIVER("Task %pK: buffers prepared.\n", task);
+
+	return ret;
+}
+
+
+static int exynos_drm_ipp_event_create(struct exynos_drm_ipp_task *task,
+				 struct drm_file *file_priv, uint64_t user_data)
+{
+	struct drm_pending_exynos_ipp_event *e = NULL;
+	int ret;
+
+	e = kzalloc(sizeof(*e), GFP_KERNEL);
+	if (!e)
+		return -ENOMEM;
+
+	e->event.base.type = DRM_EXYNOS_IPP_EVENT;
+	e->event.base.length = sizeof(e->event);
+	e->event.user_data = user_data;
+
+	ret = drm_event_reserve_init(task->dev, file_priv, &e->base,
+				     &e->event.base);
+	if (ret)
+		goto free;
+
+	task->event = e;
+	return 0;
+free:
+	kfree(e);
+	return ret;
+}
+
+static void exynos_drm_ipp_event_send(struct exynos_drm_ipp_task *task)
+{
+	struct timespec64 now;
+
+	ktime_get_ts64(&now);
+	task->event->event.tv_sec = now.tv_sec;
+	task->event->event.tv_usec = now.tv_nsec / NSEC_PER_USEC;
+	task->event->event.sequence = atomic_inc_return(&task->ipp->sequence);
+
+	drm_send_event(task->dev, &task->event->base);
+}
+
+static int exynos_drm_ipp_task_cleanup(struct exynos_drm_ipp_task *task)
+{
+	int ret = task->ret;
+
+	if (ret == 0 && task->event) {
+		exynos_drm_ipp_event_send(task);
+		/* ensure event won't be canceled on task free */
+		task->event = NULL;
+	}
+
+	exynos_drm_ipp_task_free(task->ipp, task);
+	return ret;
+}
+
+static void exynos_drm_ipp_cleanup_work(struct work_struct *work)
+{
+	struct exynos_drm_ipp_task *task = container_of(work,
+				      struct exynos_drm_ipp_task, cleanup_work);
+
+	exynos_drm_ipp_task_cleanup(task);
+}
+
+static void exynos_drm_ipp_next_task(struct exynos_drm_ipp *ipp);
+
+/**
+ * exynos_drm_ipp_task_done - finish given task and set return code
+ * @task: ipp task to finish
+ * @ret: error code or 0 if operation has been performed successfully
+ */
+void exynos_drm_ipp_task_done(struct exynos_drm_ipp_task *task, int ret)
+{
+	struct exynos_drm_ipp *ipp = task->ipp;
+	unsigned long flags;
+
+	DRM_DEBUG_DRIVER("ipp: %d, task %pK done: %d\n", ipp->id, task, ret);
+
+	spin_lock_irqsave(&ipp->lock, flags);
+	if (ipp->task == task)
+		ipp->task = NULL;
+	task->flags |= DRM_EXYNOS_IPP_TASK_DONE;
+	task->ret = ret;
+	spin_unlock_irqrestore(&ipp->lock, flags);
+
+	exynos_drm_ipp_next_task(ipp);
+	wake_up(&ipp->done_wq);
+
+	if (task->flags & DRM_EXYNOS_IPP_TASK_ASYNC) {
+		INIT_WORK(&task->cleanup_work, exynos_drm_ipp_cleanup_work);
+		schedule_work(&task->cleanup_work);
+	}
+}
+
+static void exynos_drm_ipp_next_task(struct exynos_drm_ipp *ipp)
+{
+	struct exynos_drm_ipp_task *task;
+	unsigned long flags;
+	int ret;
+
+	DRM_DEBUG_DRIVER("ipp: %d, try to run new task\n", ipp->id);
+
+	spin_lock_irqsave(&ipp->lock, flags);
+
+	if (ipp->task || list_empty(&ipp->todo_list)) {
+		spin_unlock_irqrestore(&ipp->lock, flags);
+		return;
+	}
+
+	task = list_first_entry(&ipp->todo_list, struct exynos_drm_ipp_task,
+				head);
+	list_del_init(&task->head);
+	ipp->task = task;
+
+	spin_unlock_irqrestore(&ipp->lock, flags);
+
+	DRM_DEBUG_DRIVER("ipp: %d, selected task %pK to run\n", ipp->id, task);
+
+	ret = ipp->funcs->commit(ipp, task);
+	if (ret)
+		exynos_drm_ipp_task_done(task, ret);
+}
+
+static void exynos_drm_ipp_schedule_task(struct exynos_drm_ipp *ipp,
+					 struct exynos_drm_ipp_task *task)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&ipp->lock, flags);
+	list_add(&task->head, &ipp->todo_list);
+	spin_unlock_irqrestore(&ipp->lock, flags);
+
+	exynos_drm_ipp_next_task(ipp);
+}
+
+static void exynos_drm_ipp_task_abort(struct exynos_drm_ipp *ipp,
+				      struct exynos_drm_ipp_task *task)
+{
+	unsigned long flags;
+
+	spin_lock_irqsave(&ipp->lock, flags);
+	if (task->flags & DRM_EXYNOS_IPP_TASK_DONE) {
+		/* already completed task */
+		exynos_drm_ipp_task_cleanup(task);
+	} else if (ipp->task != task) {
+		/* task has not been scheduled for execution yet */
+		list_del_init(&task->head);
+		exynos_drm_ipp_task_cleanup(task);
+	} else {
+		/*
+		 * currently processed task, call abort() and perform
+		 * cleanup with async worker
+		 */
+		task->flags |= DRM_EXYNOS_IPP_TASK_ASYNC;
+		spin_unlock_irqrestore(&ipp->lock, flags);
+		if (ipp->funcs->abort)
+			ipp->funcs->abort(ipp, task);
+		return;
+	}
+	spin_unlock_irqrestore(&ipp->lock, flags);
+}
+
+/**
+ * exynos_drm_ipp_commit_ioctl - perform image processing operation
+ * @dev: DRM device
+ * @data: ioctl data
+ * @file_priv: DRM file info
+ *
+ * Construct a ipp task from the set of properties provided from the user
+ * and try to schedule it to framebuffer processor hardware.
+ *
+ * Called by the user via ioctl.
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
+int exynos_drm_ipp_commit_ioctl(struct drm_device *dev, void *data,
+				struct drm_file *file_priv)
+{
+	struct drm_exynos_ioctl_ipp_commit *arg = data;
+	struct exynos_drm_ipp *ipp;
+	struct exynos_drm_ipp_task *task;
+	int ret = 0;
+
+	if ((arg->flags & ~DRM_EXYNOS_IPP_FLAGS) || arg->reserved)
+		return -EINVAL;
+
+	/* can't test and expect an event at the same time */
+	if ((arg->flags & DRM_EXYNOS_IPP_FLAG_TEST_ONLY) &&
+			(arg->flags & DRM_EXYNOS_IPP_FLAG_EVENT))
+		return -EINVAL;
+
+	ipp = __ipp_get(arg->ipp_id);
+	if (!ipp)
+		return -ENOENT;
+
+	task = exynos_drm_ipp_task_alloc(ipp);
+	if (!task)
+		return -ENOMEM;
+
+	ret = exynos_drm_ipp_task_set(task, arg);
+	if (ret)
+		goto free;
+
+	ret = exynos_drm_ipp_task_check(task);
+	if (ret)
+		goto free;
+
+	ret = exynos_drm_ipp_task_setup_buffers(task, file_priv);
+	if (ret || arg->flags & DRM_EXYNOS_IPP_FLAG_TEST_ONLY)
+		goto free;
+
+	if (arg->flags & DRM_EXYNOS_IPP_FLAG_EVENT) {
+		ret = exynos_drm_ipp_event_create(task, file_priv,
+						 arg->user_data);
+		if (ret)
+			goto free;
+	}
+
+	/*
+	 * Queue task for processing on the hardware. task object will be
+	 * then freed after exynos_drm_ipp_task_done()
+	 */
+	if (arg->flags & DRM_EXYNOS_IPP_FLAG_NONBLOCK) {
+		DRM_DEBUG_DRIVER("ipp: %d, nonblocking processing task %pK\n",
+				 ipp->id, task);
+
+		task->flags |= DRM_EXYNOS_IPP_TASK_ASYNC;
+		exynos_drm_ipp_schedule_task(task->ipp, task);
+		ret = 0;
+	} else {
+		DRM_DEBUG_DRIVER("ipp: %d, processing task %pK\n", ipp->id,
+				 task);
+		exynos_drm_ipp_schedule_task(ipp, task);
+		ret = wait_event_interruptible(ipp->done_wq,
+					task->flags & DRM_EXYNOS_IPP_TASK_DONE);
+		if (ret)
+			exynos_drm_ipp_task_abort(ipp, task);
+		else
+			ret = exynos_drm_ipp_task_cleanup(task);
+	}
+	return ret;
+free:
+	exynos_drm_ipp_task_free(ipp, task);
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/exynos/exynos_drm_ipp.h b/drivers/gpu/drm/exynos/exynos_drm_ipp.h
new file mode 100644
index 000000000000..0b27d4a9bf94
--- /dev/null
+++ b/drivers/gpu/drm/exynos/exynos_drm_ipp.h
@@ -0,0 +1,175 @@
+/*
+ * Copyright (c) 2017 Samsung Electronics Co., Ltd.
+ *
+ * This program is free software; you can redistribute  it and/or modify it
+ * under  the terms of  the GNU General  Public License as published by the
+ * Free Software Foundation;  either version 2 of the  License, or (at your
+ * option) any later version.
+ */
+
+#ifndef _EXYNOS_DRM_IPP_H_
+#define _EXYNOS_DRM_IPP_H_
+
+#include <drm/drmP.h>
+
+struct exynos_drm_ipp;
+struct exynos_drm_ipp_task;
+
+/**
+ * struct exynos_drm_ipp_funcs - exynos_drm_ipp control functions
+ */
+struct exynos_drm_ipp_funcs {
+	/**
+	 * @commit:
+	 *
+	 * This is the main entry point to start framebuffer processing
+	 * in the hardware. The exynos_drm_ipp_task has been already validated.
+	 * This function must not wait until the device finishes processing.
+	 * When the driver finishes processing, it has to call
+	 * exynos_exynos_drm_ipp_task_done() function.
+	 *
+	 * RETURNS:
+	 *
+	 * 0 on success or negative error codes in case of failure.
+	 */
+	int (*commit)(struct exynos_drm_ipp *ipp,
+		      struct exynos_drm_ipp_task *task);
+
+	/**
+	 * @abort:
+	 *
+	 * Informs the driver that it has to abort the currently running
+	 * task as soon as possible (i.e. as soon as it can stop the device
+	 * safely), even if the task would not have been finished by then.
+	 * After the driver performs the necessary steps, it has to call
+	 * exynos_drm_ipp_task_done() (as if the task ended normally).
+	 * This function does not have to (and will usually not) wait
+	 * until the device enters a state when it can be stopped.
+	 */
+	void (*abort)(struct exynos_drm_ipp *ipp,
+		      struct exynos_drm_ipp_task *task);
+};
+
+/**
+ * struct exynos_drm_ipp - central picture processor module structure
+ */
+struct exynos_drm_ipp {
+	struct drm_device *dev;
+	struct list_head head;
+	unsigned int id;
+
+	const char *name;
+	const struct exynos_drm_ipp_funcs *funcs;
+	unsigned int capabilities;
+	const struct exynos_drm_ipp_formats *formats;
+	unsigned int num_formats;
+	atomic_t sequence;
+
+	spinlock_t lock;
+	struct exynos_drm_ipp_task *task;
+	struct list_head todo_list;
+	wait_queue_head_t done_wq;
+};
+
+struct exynos_drm_ipp_buffer {
+	struct drm_exynos_ipp_task_buffer buf;
+	struct drm_exynos_ipp_task_rect rect;
+
+	struct exynos_drm_gem *exynos_gem[MAX_FB_BUFFER];
+	const struct drm_format_info *format;
+	dma_addr_t dma_addr[MAX_FB_BUFFER];
+};
+
+/**
+ * struct exynos_drm_ipp_task - a structure describing transformation that
+ * has to be performed by the picture processor hardware module
+ */
+struct exynos_drm_ipp_task {
+	struct drm_device *dev;
+	struct exynos_drm_ipp *ipp;
+	struct list_head head;
+
+	struct exynos_drm_ipp_buffer src;
+	struct exynos_drm_ipp_buffer dst;
+
+	struct drm_exynos_ipp_task_transform transform;
+	struct drm_exynos_ipp_task_alpha alpha;
+
+	struct work_struct cleanup_work;
+	unsigned int flags;
+	int ret;
+
+	struct drm_pending_exynos_ipp_event *event;
+};
+
+#define DRM_EXYNOS_IPP_TASK_DONE	(1 << 0)
+#define DRM_EXYNOS_IPP_TASK_ASYNC	(1 << 1)
+
+struct exynos_drm_ipp_formats {
+	uint32_t fourcc;
+	uint32_t type;
+	uint64_t modifier;
+	const struct drm_exynos_ipp_limit *limits;
+	unsigned int num_limits;
+};
+
+/* helper macros to set exynos_drm_ipp_formats structure and limits*/
+#define IPP_SRCDST_MFORMAT(f, m, l) \
+	.fourcc = DRM_FORMAT_##f, .modifier = m, .limits = l, \
+	.num_limits = ARRAY_SIZE(l), \
+	.type = (DRM_EXYNOS_IPP_FORMAT_SOURCE | \
+		 DRM_EXYNOS_IPP_FORMAT_DESTINATION)
+
+#define IPP_SRCDST_FORMAT(f, l) IPP_SRCDST_MFORMAT(f, 0, l)
+
+#define IPP_SIZE_LIMIT(l, val...)	\
+	.type = (DRM_EXYNOS_IPP_LIMIT_TYPE_SIZE | \
+		 DRM_EXYNOS_IPP_LIMIT_SIZE_##l), val
+
+#define IPP_SCALE_LIMIT(val...)		\
+	.type = (DRM_EXYNOS_IPP_LIMIT_TYPE_SCALE), val
+
+int exynos_drm_ipp_register(struct drm_device *dev, struct exynos_drm_ipp *ipp,
+		const struct exynos_drm_ipp_funcs *funcs, unsigned int caps,
+		const struct exynos_drm_ipp_formats *formats,
+		unsigned int num_formats, const char *name);
+void exynos_drm_ipp_unregister(struct drm_device *dev,
+			       struct exynos_drm_ipp *ipp);
+
+void exynos_drm_ipp_task_done(struct exynos_drm_ipp_task *task, int ret);
+
+#ifdef CONFIG_DRM_EXYNOS_IPP
+int exynos_drm_ipp_get_res_ioctl(struct drm_device *dev, void *data,
+				 struct drm_file *file_priv);
+int exynos_drm_ipp_get_caps_ioctl(struct drm_device *dev, void *data,
+				  struct drm_file *file_priv);
+int exynos_drm_ipp_get_limits_ioctl(struct drm_device *dev, void *data,
+				    struct drm_file *file_priv);
+int exynos_drm_ipp_commit_ioctl(struct drm_device *dev,
+				void *data, struct drm_file *file_priv);
+#else
+static inline int exynos_drm_ipp_get_res_ioctl(struct drm_device *dev,
+	 void *data, struct drm_file *file_priv)
+{
+	struct drm_exynos_ioctl_ipp_get_res *resp = data;
+
+	resp->count_ipps = 0;
+	return 0;
+}
+static inline int exynos_drm_ipp_get_caps_ioctl(struct drm_device *dev,
+	 void *data, struct drm_file *file_priv)
+{
+	return -ENODEV;
+}
+static inline int exynos_drm_ipp_get_limits_ioctl(struct drm_device *dev,
+	 void *data, struct drm_file *file_priv)
+{
+	return -ENODEV;
+}
+static inline int exynos_drm_ipp_commit_ioctl(struct drm_device *dev,
+	 void *data, struct drm_file *file_priv)
+{
+	return -ENODEV;
+}
+#endif
+#endif
diff --git a/drivers/gpu/drm/exynos/exynos_drm_plane.c b/drivers/gpu/drm/exynos/exynos_drm_plane.c
index d2a90dae5c71..eb9915da7dec 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_plane.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_plane.c
@@ -263,8 +263,6 @@ static void exynos_plane_atomic_update(struct drm_plane *plane,
 	if (!state->crtc)
 		return;
 
-	plane->crtc = state->crtc;
-
 	if (exynos_crtc->ops->update_plane)
 		exynos_crtc->ops->update_plane(exynos_crtc, exynos_plane);
 }
@@ -289,13 +287,12 @@ static const struct drm_plane_helper_funcs plane_helper_funcs = {
 };
 
 static void exynos_plane_attach_zpos_property(struct drm_plane *plane,
-					      bool immutable)
+					      int zpos, bool immutable)
 {
-	/* FIXME */
 	if (immutable)
-		drm_plane_create_zpos_immutable_property(plane, 0);
+		drm_plane_create_zpos_immutable_property(plane, zpos);
 	else
-		drm_plane_create_zpos_property(plane, 0, 0, MAX_PLANE - 1);
+		drm_plane_create_zpos_property(plane, zpos, 0, MAX_PLANE - 1);
 }
 
 int exynos_plane_init(struct drm_device *dev,
@@ -320,7 +317,7 @@ int exynos_plane_init(struct drm_device *dev,
 	exynos_plane->index = index;
 	exynos_plane->config = config;
 
-	exynos_plane_attach_zpos_property(&exynos_plane->base,
+	exynos_plane_attach_zpos_property(&exynos_plane->base, config->zpos,
 			   !(config->capabilities & EXYNOS_DRM_PLANE_CAP_ZPOS));
 
 	return 0;
diff --git a/drivers/gpu/drm/exynos/exynos_drm_rotator.c b/drivers/gpu/drm/exynos/exynos_drm_rotator.c
index 79282a820ecc..1a76dd3d52e1 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_rotator.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_rotator.c
@@ -10,6 +10,7 @@
  */
 
 #include <linux/kernel.h>
+#include <linux/component.h>
 #include <linux/err.h>
 #include <linux/interrupt.h>
 #include <linux/io.h>
@@ -22,29 +23,18 @@
 #include <drm/exynos_drm.h>
 #include "regs-rotator.h"
 #include "exynos_drm_drv.h"
+#include "exynos_drm_iommu.h"
 #include "exynos_drm_ipp.h"
 
 /*
  * Rotator supports image crop/rotator and input/output DMA operations.
  * input DMA reads image data from the memory.
  * output DMA writes image data to memory.
- *
- * M2M operation : supports crop/scale/rotation/csc so on.
- * Memory ----> Rotator H/W ----> Memory.
  */
 
-/*
- * TODO
- * 1. check suspend/resume api if needed.
- * 2. need to check use case platform_device_id.
- * 3. check src/dst size with, height.
- * 4. need to add supported list in prop_list.
- */
+#define ROTATOR_AUTOSUSPEND_DELAY	2000
 
-#define get_rot_context(dev)	platform_get_drvdata(to_platform_device(dev))
-#define get_ctx_from_ippdrv(ippdrv)	container_of(ippdrv,\
-					struct rot_context, ippdrv);
-#define rot_read(offset)		readl(rot->regs + (offset))
+#define rot_read(offset)	readl(rot->regs + (offset))
 #define rot_write(cfg, offset)	writel(cfg, rot->regs + (offset))
 
 enum rot_irq_status {
@@ -52,54 +42,28 @@ enum rot_irq_status {
 	ROT_IRQ_STATUS_ILLEGAL	= 9,
 };
 
-/*
- * A structure of limitation.
- *
- * @min_w: minimum width.
- * @min_h: minimum height.
- * @max_w: maximum width.
- * @max_h: maximum height.
- * @align: align size.
- */
-struct rot_limit {
-	u32	min_w;
-	u32	min_h;
-	u32	max_w;
-	u32	max_h;
-	u32	align;
-};
-
-/*
- * A structure of limitation table.
- *
- * @ycbcr420_2p: case of YUV.
- * @rgb888: case of RGB.
- */
-struct rot_limit_table {
-	struct rot_limit	ycbcr420_2p;
-	struct rot_limit	rgb888;
+struct rot_variant {
+	const struct exynos_drm_ipp_formats *formats;
+	unsigned int	num_formats;
 };
 
 /*
  * A structure of rotator context.
  * @ippdrv: prepare initialization using ippdrv.
- * @regs_res: register resources.
  * @regs: memory mapped io registers.
  * @clock: rotator gate clock.
  * @limit_tbl: limitation of rotator.
  * @irq: irq number.
- * @cur_buf_id: current operation buffer id.
- * @suspended: suspended state.
  */
 struct rot_context {
-	struct exynos_drm_ippdrv	ippdrv;
-	struct resource	*regs_res;
+	struct exynos_drm_ipp ipp;
+	struct drm_device *drm_dev;
+	struct device	*dev;
 	void __iomem	*regs;
 	struct clk	*clock;
-	struct rot_limit_table	*limit_tbl;
-	int	irq;
-	int	cur_buf_id[EXYNOS_DRM_OPS_MAX];
-	bool	suspended;
+	const struct exynos_drm_ipp_formats *formats;
+	unsigned int	num_formats;
+	struct exynos_drm_ipp_task	*task;
 };
 
 static void rotator_reg_set_irq(struct rot_context *rot, bool enable)
@@ -114,15 +78,6 @@ static void rotator_reg_set_irq(struct rot_context *rot, bool enable)
 	rot_write(val, ROT_CONFIG);
 }
 
-static u32 rotator_reg_get_fmt(struct rot_context *rot)
-{
-	u32 val = rot_read(ROT_CONTROL);
-
-	val &= ROT_CONTROL_FMT_MASK;
-
-	return val;
-}
-
 static enum rot_irq_status rotator_reg_get_irq_status(struct rot_context *rot)
 {
 	u32 val = rot_read(ROT_STATUS);
@@ -138,9 +93,6 @@ static enum rot_irq_status rotator_reg_get_irq_status(struct rot_context *rot)
 static irqreturn_t rotator_irq_handler(int irq, void *arg)
 {
 	struct rot_context *rot = arg;
-	struct exynos_drm_ippdrv *ippdrv = &rot->ippdrv;
-	struct drm_exynos_ipp_cmd_node *c_node = ippdrv->c_node;
-	struct drm_exynos_ipp_event_work *event_work = c_node->event_work;
 	enum rot_irq_status irq_status;
 	u32 val;
 
@@ -152,56 +104,21 @@ static irqreturn_t rotator_irq_handler(int irq, void *arg)
 	val |= ROT_STATUS_IRQ_PENDING((u32)irq_status);
 	rot_write(val, ROT_STATUS);
 
-	if (irq_status == ROT_IRQ_STATUS_COMPLETE) {
-		event_work->ippdrv = ippdrv;
-		event_work->buf_id[EXYNOS_DRM_OPS_DST] =
-			rot->cur_buf_id[EXYNOS_DRM_OPS_DST];
-		queue_work(ippdrv->event_workq, &event_work->work);
-	} else {
-		DRM_ERROR("the SFR is set illegally\n");
+	if (rot->task) {
+		struct exynos_drm_ipp_task *task = rot->task;
+
+		rot->task = NULL;
+		pm_runtime_mark_last_busy(rot->dev);
+		pm_runtime_put_autosuspend(rot->dev);
+		exynos_drm_ipp_task_done(task,
+			irq_status == ROT_IRQ_STATUS_COMPLETE ? 0 : -EINVAL);
 	}
 
 	return IRQ_HANDLED;
 }
 
-static void rotator_align_size(struct rot_context *rot, u32 fmt, u32 *hsize,
-		u32 *vsize)
+static void rotator_src_set_fmt(struct rot_context *rot, u32 fmt)
 {
-	struct rot_limit_table *limit_tbl = rot->limit_tbl;
-	struct rot_limit *limit;
-	u32 mask, val;
-
-	/* Get size limit */
-	if (fmt == ROT_CONTROL_FMT_RGB888)
-		limit = &limit_tbl->rgb888;
-	else
-		limit = &limit_tbl->ycbcr420_2p;
-
-	/* Get mask for rounding to nearest aligned val */
-	mask = ~((1 << limit->align) - 1);
-
-	/* Set aligned width */
-	val = ROT_ALIGN(*hsize, limit->align, mask);
-	if (val < limit->min_w)
-		*hsize = ROT_MIN(limit->min_w, mask);
-	else if (val > limit->max_w)
-		*hsize = ROT_MAX(limit->max_w, mask);
-	else
-		*hsize = val;
-
-	/* Set aligned height */
-	val = ROT_ALIGN(*vsize, limit->align, mask);
-	if (val < limit->min_h)
-		*vsize = ROT_MIN(limit->min_h, mask);
-	else if (val > limit->max_h)
-		*vsize = ROT_MAX(limit->max_h, mask);
-	else
-		*vsize = val;
-}
-
-static int rotator_src_set_fmt(struct device *dev, u32 fmt)
-{
-	struct rot_context *rot = dev_get_drvdata(dev);
 	u32 val;
 
 	val = rot_read(ROT_CONTROL);
@@ -214,515 +131,176 @@ static int rotator_src_set_fmt(struct device *dev, u32 fmt)
 	case DRM_FORMAT_XRGB8888:
 		val |= ROT_CONTROL_FMT_RGB888;
 		break;
-	default:
-		DRM_ERROR("invalid image format\n");
-		return -EINVAL;
 	}
 
 	rot_write(val, ROT_CONTROL);
-
-	return 0;
 }
 
-static inline bool rotator_check_reg_fmt(u32 fmt)
+static void rotator_src_set_buf(struct rot_context *rot,
+				struct exynos_drm_ipp_buffer *buf)
 {
-	if ((fmt == ROT_CONTROL_FMT_YCBCR420_2P) ||
-	    (fmt == ROT_CONTROL_FMT_RGB888))
-		return true;
-
-	return false;
-}
-
-static int rotator_src_set_size(struct device *dev, int swap,
-		struct drm_exynos_pos *pos,
-		struct drm_exynos_sz *sz)
-{
-	struct rot_context *rot = dev_get_drvdata(dev);
-	u32 fmt, hsize, vsize;
 	u32 val;
 
-	/* Get format */
-	fmt = rotator_reg_get_fmt(rot);
-	if (!rotator_check_reg_fmt(fmt)) {
-		DRM_ERROR("invalid format.\n");
-		return -EINVAL;
-	}
-
-	/* Align buffer size */
-	hsize = sz->hsize;
-	vsize = sz->vsize;
-	rotator_align_size(rot, fmt, &hsize, &vsize);
-
 	/* Set buffer size configuration */
-	val = ROT_SET_BUF_SIZE_H(vsize) | ROT_SET_BUF_SIZE_W(hsize);
+	val = ROT_SET_BUF_SIZE_H(buf->buf.height) |
+	      ROT_SET_BUF_SIZE_W(buf->buf.pitch[0] / buf->format->cpp[0]);
 	rot_write(val, ROT_SRC_BUF_SIZE);
 
 	/* Set crop image position configuration */
-	val = ROT_CROP_POS_Y(pos->y) | ROT_CROP_POS_X(pos->x);
+	val = ROT_CROP_POS_Y(buf->rect.y) | ROT_CROP_POS_X(buf->rect.x);
 	rot_write(val, ROT_SRC_CROP_POS);
-	val = ROT_SRC_CROP_SIZE_H(pos->h) | ROT_SRC_CROP_SIZE_W(pos->w);
+	val = ROT_SRC_CROP_SIZE_H(buf->rect.h) |
+	      ROT_SRC_CROP_SIZE_W(buf->rect.w);
 	rot_write(val, ROT_SRC_CROP_SIZE);
 
-	return 0;
+	/* Set buffer DMA address */
+	rot_write(buf->dma_addr[0], ROT_SRC_BUF_ADDR(0));
+	rot_write(buf->dma_addr[1], ROT_SRC_BUF_ADDR(1));
 }
 
-static int rotator_src_set_addr(struct device *dev,
-		struct drm_exynos_ipp_buf_info *buf_info,
-		u32 buf_id, enum drm_exynos_ipp_buf_type buf_type)
+static void rotator_dst_set_transf(struct rot_context *rot,
+				   unsigned int rotation)
 {
-	struct rot_context *rot = dev_get_drvdata(dev);
-	dma_addr_t addr[EXYNOS_DRM_PLANAR_MAX];
-	u32 val, fmt, hsize, vsize;
-	int i;
-
-	/* Set current buf_id */
-	rot->cur_buf_id[EXYNOS_DRM_OPS_SRC] = buf_id;
-
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		/* Set address configuration */
-		for_each_ipp_planar(i)
-			addr[i] = buf_info->base[i];
-
-		/* Get format */
-		fmt = rotator_reg_get_fmt(rot);
-		if (!rotator_check_reg_fmt(fmt)) {
-			DRM_ERROR("invalid format.\n");
-			return -EINVAL;
-		}
-
-		/* Re-set cb planar for NV12 format */
-		if ((fmt == ROT_CONTROL_FMT_YCBCR420_2P) &&
-		    !addr[EXYNOS_DRM_PLANAR_CB]) {
-
-			val = rot_read(ROT_SRC_BUF_SIZE);
-			hsize = ROT_GET_BUF_SIZE_W(val);
-			vsize = ROT_GET_BUF_SIZE_H(val);
-
-			/* Set cb planar */
-			addr[EXYNOS_DRM_PLANAR_CB] =
-				addr[EXYNOS_DRM_PLANAR_Y] + hsize * vsize;
-		}
-
-		for_each_ipp_planar(i)
-			rot_write(addr[i], ROT_SRC_BUF_ADDR(i));
-		break;
-	case IPP_BUF_DEQUEUE:
-		for_each_ipp_planar(i)
-			rot_write(0x0, ROT_SRC_BUF_ADDR(i));
-		break;
-	default:
-		/* Nothing to do */
-		break;
-	}
-
-	return 0;
-}
-
-static int rotator_dst_set_transf(struct device *dev,
-		enum drm_exynos_degree degree,
-		enum drm_exynos_flip flip, bool *swap)
-{
-	struct rot_context *rot = dev_get_drvdata(dev);
 	u32 val;
 
 	/* Set transform configuration */
 	val = rot_read(ROT_CONTROL);
 	val &= ~ROT_CONTROL_FLIP_MASK;
 
-	switch (flip) {
-	case EXYNOS_DRM_FLIP_VERTICAL:
-		val |= ROT_CONTROL_FLIP_VERTICAL;
-		break;
-	case EXYNOS_DRM_FLIP_HORIZONTAL:
+	if (rotation & DRM_MODE_REFLECT_X)
 		val |= ROT_CONTROL_FLIP_HORIZONTAL;
-		break;
-	default:
-		/* Flip None */
-		break;
-	}
+	if (rotation & DRM_MODE_REFLECT_Y)
+		val |= ROT_CONTROL_FLIP_VERTICAL;
 
 	val &= ~ROT_CONTROL_ROT_MASK;
 
-	switch (degree) {
-	case EXYNOS_DRM_DEGREE_90:
+	if (rotation & DRM_MODE_ROTATE_90)
 		val |= ROT_CONTROL_ROT_90;
-		break;
-	case EXYNOS_DRM_DEGREE_180:
+	else if (rotation & DRM_MODE_ROTATE_180)
 		val |= ROT_CONTROL_ROT_180;
-		break;
-	case EXYNOS_DRM_DEGREE_270:
+	else if (rotation & DRM_MODE_ROTATE_270)
 		val |= ROT_CONTROL_ROT_270;
-		break;
-	default:
-		/* Rotation 0 Degree */
-		break;
-	}
 
 	rot_write(val, ROT_CONTROL);
-
-	/* Check degree for setting buffer size swap */
-	if ((degree == EXYNOS_DRM_DEGREE_90) ||
-	    (degree == EXYNOS_DRM_DEGREE_270))
-		*swap = true;
-	else
-		*swap = false;
-
-	return 0;
 }
 
-static int rotator_dst_set_size(struct device *dev, int swap,
-		struct drm_exynos_pos *pos,
-		struct drm_exynos_sz *sz)
+static void rotator_dst_set_buf(struct rot_context *rot,
+				struct exynos_drm_ipp_buffer *buf)
 {
-	struct rot_context *rot = dev_get_drvdata(dev);
-	u32 val, fmt, hsize, vsize;
-
-	/* Get format */
-	fmt = rotator_reg_get_fmt(rot);
-	if (!rotator_check_reg_fmt(fmt)) {
-		DRM_ERROR("invalid format.\n");
-		return -EINVAL;
-	}
-
-	/* Align buffer size */
-	hsize = sz->hsize;
-	vsize = sz->vsize;
-	rotator_align_size(rot, fmt, &hsize, &vsize);
+	u32 val;
 
 	/* Set buffer size configuration */
-	val = ROT_SET_BUF_SIZE_H(vsize) | ROT_SET_BUF_SIZE_W(hsize);
+	val = ROT_SET_BUF_SIZE_H(buf->buf.height) |
+	      ROT_SET_BUF_SIZE_W(buf->buf.pitch[0] / buf->format->cpp[0]);
 	rot_write(val, ROT_DST_BUF_SIZE);
 
 	/* Set crop image position configuration */
-	val = ROT_CROP_POS_Y(pos->y) | ROT_CROP_POS_X(pos->x);
+	val = ROT_CROP_POS_Y(buf->rect.y) | ROT_CROP_POS_X(buf->rect.x);
 	rot_write(val, ROT_DST_CROP_POS);
 
-	return 0;
+	/* Set buffer DMA address */
+	rot_write(buf->dma_addr[0], ROT_DST_BUF_ADDR(0));
+	rot_write(buf->dma_addr[1], ROT_DST_BUF_ADDR(1));
 }
 
-static int rotator_dst_set_addr(struct device *dev,
-		struct drm_exynos_ipp_buf_info *buf_info,
-		u32 buf_id, enum drm_exynos_ipp_buf_type buf_type)
+static void rotator_start(struct rot_context *rot)
 {
-	struct rot_context *rot = dev_get_drvdata(dev);
-	dma_addr_t addr[EXYNOS_DRM_PLANAR_MAX];
-	u32 val, fmt, hsize, vsize;
-	int i;
-
-	/* Set current buf_id */
-	rot->cur_buf_id[EXYNOS_DRM_OPS_DST] = buf_id;
-
-	switch (buf_type) {
-	case IPP_BUF_ENQUEUE:
-		/* Set address configuration */
-		for_each_ipp_planar(i)
-			addr[i] = buf_info->base[i];
-
-		/* Get format */
-		fmt = rotator_reg_get_fmt(rot);
-		if (!rotator_check_reg_fmt(fmt)) {
-			DRM_ERROR("invalid format.\n");
-			return -EINVAL;
-		}
-
-		/* Re-set cb planar for NV12 format */
-		if ((fmt == ROT_CONTROL_FMT_YCBCR420_2P) &&
-		    !addr[EXYNOS_DRM_PLANAR_CB]) {
-			/* Get buf size */
-			val = rot_read(ROT_DST_BUF_SIZE);
-
-			hsize = ROT_GET_BUF_SIZE_W(val);
-			vsize = ROT_GET_BUF_SIZE_H(val);
-
-			/* Set cb planar */
-			addr[EXYNOS_DRM_PLANAR_CB] =
-				addr[EXYNOS_DRM_PLANAR_Y] + hsize * vsize;
-		}
-
-		for_each_ipp_planar(i)
-			rot_write(addr[i], ROT_DST_BUF_ADDR(i));
-		break;
-	case IPP_BUF_DEQUEUE:
-		for_each_ipp_planar(i)
-			rot_write(0x0, ROT_DST_BUF_ADDR(i));
-		break;
-	default:
-		/* Nothing to do */
-		break;
-	}
+	u32 val;
 
-	return 0;
+	/* Set interrupt enable */
+	rotator_reg_set_irq(rot, true);
+
+	val = rot_read(ROT_CONTROL);
+	val |= ROT_CONTROL_START;
+	rot_write(val, ROT_CONTROL);
 }
 
-static struct exynos_drm_ipp_ops rot_src_ops = {
-	.set_fmt	=	rotator_src_set_fmt,
-	.set_size	=	rotator_src_set_size,
-	.set_addr	=	rotator_src_set_addr,
-};
+static int rotator_commit(struct exynos_drm_ipp *ipp,
+			  struct exynos_drm_ipp_task *task)
+{
+	struct rot_context *rot =
+			container_of(ipp, struct rot_context, ipp);
 
-static struct exynos_drm_ipp_ops rot_dst_ops = {
-	.set_transf	=	rotator_dst_set_transf,
-	.set_size	=	rotator_dst_set_size,
-	.set_addr	=	rotator_dst_set_addr,
-};
+	pm_runtime_get_sync(rot->dev);
+	rot->task = task;
 
-static int rotator_init_prop_list(struct exynos_drm_ippdrv *ippdrv)
-{
-	struct drm_exynos_ipp_prop_list *prop_list = &ippdrv->prop_list;
-
-	prop_list->version = 1;
-	prop_list->flip = (1 << EXYNOS_DRM_FLIP_VERTICAL) |
-				(1 << EXYNOS_DRM_FLIP_HORIZONTAL);
-	prop_list->degree = (1 << EXYNOS_DRM_DEGREE_0) |
-				(1 << EXYNOS_DRM_DEGREE_90) |
-				(1 << EXYNOS_DRM_DEGREE_180) |
-				(1 << EXYNOS_DRM_DEGREE_270);
-	prop_list->csc = 0;
-	prop_list->crop = 0;
-	prop_list->scale = 0;
+	rotator_src_set_fmt(rot, task->src.buf.fourcc);
+	rotator_src_set_buf(rot, &task->src);
+	rotator_dst_set_transf(rot, task->transform.rotation);
+	rotator_dst_set_buf(rot, &task->dst);
+	rotator_start(rot);
 
 	return 0;
 }
 
-static inline bool rotator_check_drm_fmt(u32 fmt)
-{
-	switch (fmt) {
-	case DRM_FORMAT_XRGB8888:
-	case DRM_FORMAT_NV12:
-		return true;
-	default:
-		DRM_DEBUG_KMS("not support format\n");
-		return false;
-	}
-}
-
-static inline bool rotator_check_drm_flip(enum drm_exynos_flip flip)
-{
-	switch (flip) {
-	case EXYNOS_DRM_FLIP_NONE:
-	case EXYNOS_DRM_FLIP_VERTICAL:
-	case EXYNOS_DRM_FLIP_HORIZONTAL:
-	case EXYNOS_DRM_FLIP_BOTH:
-		return true;
-	default:
-		DRM_DEBUG_KMS("invalid flip\n");
-		return false;
-	}
-}
+static const struct exynos_drm_ipp_funcs ipp_funcs = {
+	.commit = rotator_commit,
+};
 
-static int rotator_ippdrv_check_property(struct device *dev,
-		struct drm_exynos_ipp_property *property)
+static int rotator_bind(struct device *dev, struct device *master, void *data)
 {
-	struct drm_exynos_ipp_config *src_config =
-					&property->config[EXYNOS_DRM_OPS_SRC];
-	struct drm_exynos_ipp_config *dst_config =
-					&property->config[EXYNOS_DRM_OPS_DST];
-	struct drm_exynos_pos *src_pos = &src_config->pos;
-	struct drm_exynos_pos *dst_pos = &dst_config->pos;
-	struct drm_exynos_sz *src_sz = &src_config->sz;
-	struct drm_exynos_sz *dst_sz = &dst_config->sz;
-	bool swap = false;
-
-	/* Check format configuration */
-	if (src_config->fmt != dst_config->fmt) {
-		DRM_DEBUG_KMS("not support csc feature\n");
-		return -EINVAL;
-	}
-
-	if (!rotator_check_drm_fmt(dst_config->fmt)) {
-		DRM_DEBUG_KMS("invalid format\n");
-		return -EINVAL;
-	}
-
-	/* Check transform configuration */
-	if (src_config->degree != EXYNOS_DRM_DEGREE_0) {
-		DRM_DEBUG_KMS("not support source-side rotation\n");
-		return -EINVAL;
-	}
-
-	switch (dst_config->degree) {
-	case EXYNOS_DRM_DEGREE_90:
-	case EXYNOS_DRM_DEGREE_270:
-		swap = true;
-	case EXYNOS_DRM_DEGREE_0:
-	case EXYNOS_DRM_DEGREE_180:
-		/* No problem */
-		break;
-	default:
-		DRM_DEBUG_KMS("invalid degree\n");
-		return -EINVAL;
-	}
-
-	if (src_config->flip != EXYNOS_DRM_FLIP_NONE) {
-		DRM_DEBUG_KMS("not support source-side flip\n");
-		return -EINVAL;
-	}
+	struct rot_context *rot = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &rot->ipp;
 
-	if (!rotator_check_drm_flip(dst_config->flip)) {
-		DRM_DEBUG_KMS("invalid flip\n");
-		return -EINVAL;
-	}
+	rot->drm_dev = drm_dev;
+	drm_iommu_attach_device(drm_dev, dev);
 
-	/* Check size configuration */
-	if ((src_pos->x + src_pos->w > src_sz->hsize) ||
-		(src_pos->y + src_pos->h > src_sz->vsize)) {
-		DRM_DEBUG_KMS("out of source buffer bound\n");
-		return -EINVAL;
-	}
+	exynos_drm_ipp_register(drm_dev, ipp, &ipp_funcs,
+			   DRM_EXYNOS_IPP_CAP_CROP | DRM_EXYNOS_IPP_CAP_ROTATE,
+			   rot->formats, rot->num_formats, "rotator");
 
-	if (swap) {
-		if ((dst_pos->x + dst_pos->h > dst_sz->vsize) ||
-			(dst_pos->y + dst_pos->w > dst_sz->hsize)) {
-			DRM_DEBUG_KMS("out of destination buffer bound\n");
-			return -EINVAL;
-		}
-
-		if ((src_pos->w != dst_pos->h) || (src_pos->h != dst_pos->w)) {
-			DRM_DEBUG_KMS("not support scale feature\n");
-			return -EINVAL;
-		}
-	} else {
-		if ((dst_pos->x + dst_pos->w > dst_sz->hsize) ||
-			(dst_pos->y + dst_pos->h > dst_sz->vsize)) {
-			DRM_DEBUG_KMS("out of destination buffer bound\n");
-			return -EINVAL;
-		}
-
-		if ((src_pos->w != dst_pos->w) || (src_pos->h != dst_pos->h)) {
-			DRM_DEBUG_KMS("not support scale feature\n");
-			return -EINVAL;
-		}
-	}
+	dev_info(dev, "The exynos rotator has been probed successfully\n");
 
 	return 0;
 }
 
-static int rotator_ippdrv_start(struct device *dev, enum drm_exynos_ipp_cmd cmd)
+static void rotator_unbind(struct device *dev, struct device *master,
+			void *data)
 {
 	struct rot_context *rot = dev_get_drvdata(dev);
-	u32 val;
-
-	if (rot->suspended) {
-		DRM_ERROR("suspended state\n");
-		return -EPERM;
-	}
-
-	if (cmd != IPP_CMD_M2M) {
-		DRM_ERROR("not support cmd: %d\n", cmd);
-		return -EINVAL;
-	}
-
-	/* Set interrupt enable */
-	rotator_reg_set_irq(rot, true);
-
-	val = rot_read(ROT_CONTROL);
-	val |= ROT_CONTROL_START;
-
-	rot_write(val, ROT_CONTROL);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &rot->ipp;
 
-	return 0;
+	exynos_drm_ipp_unregister(drm_dev, ipp);
+	drm_iommu_detach_device(rot->drm_dev, rot->dev);
 }
 
-static struct rot_limit_table rot_limit_tbl_4210 = {
-	.ycbcr420_2p = {
-		.min_w = 32,
-		.min_h = 32,
-		.max_w = SZ_64K,
-		.max_h = SZ_64K,
-		.align = 3,
-	},
-	.rgb888 = {
-		.min_w = 8,
-		.min_h = 8,
-		.max_w = SZ_16K,
-		.max_h = SZ_16K,
-		.align = 2,
-	},
-};
-
-static struct rot_limit_table rot_limit_tbl_4x12 = {
-	.ycbcr420_2p = {
-		.min_w = 32,
-		.min_h = 32,
-		.max_w = SZ_32K,
-		.max_h = SZ_32K,
-		.align = 3,
-	},
-	.rgb888 = {
-		.min_w = 8,
-		.min_h = 8,
-		.max_w = SZ_8K,
-		.max_h = SZ_8K,
-		.align = 2,
-	},
+static const struct component_ops rotator_component_ops = {
+	.bind	= rotator_bind,
+	.unbind = rotator_unbind,
 };
 
-static struct rot_limit_table rot_limit_tbl_5250 = {
-	.ycbcr420_2p = {
-		.min_w = 32,
-		.min_h = 32,
-		.max_w = SZ_32K,
-		.max_h = SZ_32K,
-		.align = 3,
-	},
-	.rgb888 = {
-		.min_w = 8,
-		.min_h = 8,
-		.max_w = SZ_8K,
-		.max_h = SZ_8K,
-		.align = 1,
-	},
-};
-
-static const struct of_device_id exynos_rotator_match[] = {
-	{
-		.compatible = "samsung,exynos4210-rotator",
-		.data = &rot_limit_tbl_4210,
-	},
-	{
-		.compatible = "samsung,exynos4212-rotator",
-		.data = &rot_limit_tbl_4x12,
-	},
-	{
-		.compatible = "samsung,exynos5250-rotator",
-		.data = &rot_limit_tbl_5250,
-	},
-	{},
-};
-MODULE_DEVICE_TABLE(of, exynos_rotator_match);
-
 static int rotator_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
+	struct resource	*regs_res;
 	struct rot_context *rot;
-	struct exynos_drm_ippdrv *ippdrv;
+	const struct rot_variant *variant;
+	int irq;
 	int ret;
 
-	if (!dev->of_node) {
-		dev_err(dev, "cannot find of_node.\n");
-		return -ENODEV;
-	}
-
 	rot = devm_kzalloc(dev, sizeof(*rot), GFP_KERNEL);
 	if (!rot)
 		return -ENOMEM;
 
-	rot->limit_tbl = (struct rot_limit_table *)
-				of_device_get_match_data(dev);
-	rot->regs_res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
-	rot->regs = devm_ioremap_resource(dev, rot->regs_res);
+	variant = of_device_get_match_data(dev);
+	rot->formats = variant->formats;
+	rot->num_formats = variant->num_formats;
+	rot->dev = dev;
+	regs_res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	rot->regs = devm_ioremap_resource(dev, regs_res);
 	if (IS_ERR(rot->regs))
 		return PTR_ERR(rot->regs);
 
-	rot->irq = platform_get_irq(pdev, 0);
-	if (rot->irq < 0) {
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
 		dev_err(dev, "failed to get irq\n");
-		return rot->irq;
+		return irq;
 	}
 
-	ret = devm_request_threaded_irq(dev, rot->irq, NULL,
-			rotator_irq_handler, IRQF_ONESHOT, "drm_rotator", rot);
+	ret = devm_request_irq(dev, irq, rotator_irq_handler, 0, dev_name(dev),
+			       rot);
 	if (ret < 0) {
 		dev_err(dev, "failed to request irq\n");
 		return ret;
@@ -734,35 +312,19 @@ static int rotator_probe(struct platform_device *pdev)
 		return PTR_ERR(rot->clock);
 	}
 
+	pm_runtime_use_autosuspend(dev);
+	pm_runtime_set_autosuspend_delay(dev, ROTATOR_AUTOSUSPEND_DELAY);
 	pm_runtime_enable(dev);
-
-	ippdrv = &rot->ippdrv;
-	ippdrv->dev = dev;
-	ippdrv->ops[EXYNOS_DRM_OPS_SRC] = &rot_src_ops;
-	ippdrv->ops[EXYNOS_DRM_OPS_DST] = &rot_dst_ops;
-	ippdrv->check_property = rotator_ippdrv_check_property;
-	ippdrv->start = rotator_ippdrv_start;
-	ret = rotator_init_prop_list(ippdrv);
-	if (ret < 0) {
-		dev_err(dev, "failed to init property list.\n");
-		goto err_ippdrv_register;
-	}
-
-	DRM_DEBUG_KMS("ippdrv[%pK]\n", ippdrv);
-
 	platform_set_drvdata(pdev, rot);
 
-	ret = exynos_drm_ippdrv_register(ippdrv);
-	if (ret < 0) {
-		dev_err(dev, "failed to register drm rotator device\n");
-		goto err_ippdrv_register;
-	}
-
-	dev_info(dev, "The exynos rotator is probed successfully\n");
+	ret = component_add(dev, &rotator_component_ops);
+	if (ret)
+		goto err_component;
 
 	return 0;
 
-err_ippdrv_register:
+err_component:
+	pm_runtime_dont_use_autosuspend(dev);
 	pm_runtime_disable(dev);
 	return ret;
 }
@@ -770,45 +332,101 @@ err_ippdrv_register:
 static int rotator_remove(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
-	struct rot_context *rot = dev_get_drvdata(dev);
-	struct exynos_drm_ippdrv *ippdrv = &rot->ippdrv;
-
-	exynos_drm_ippdrv_unregister(ippdrv);
 
+	component_del(dev, &rotator_component_ops);
+	pm_runtime_dont_use_autosuspend(dev);
 	pm_runtime_disable(dev);
 
 	return 0;
 }
 
 #ifdef CONFIG_PM
-static int rotator_clk_crtl(struct rot_context *rot, bool enable)
-{
-	if (enable) {
-		clk_prepare_enable(rot->clock);
-		rot->suspended = false;
-	} else {
-		clk_disable_unprepare(rot->clock);
-		rot->suspended = true;
-	}
-
-	return 0;
-}
-
 static int rotator_runtime_suspend(struct device *dev)
 {
 	struct rot_context *rot = dev_get_drvdata(dev);
 
-	return  rotator_clk_crtl(rot, false);
+	clk_disable_unprepare(rot->clock);
+	return 0;
 }
 
 static int rotator_runtime_resume(struct device *dev)
 {
 	struct rot_context *rot = dev_get_drvdata(dev);
 
-	return  rotator_clk_crtl(rot, true);
+	return clk_prepare_enable(rot->clock);
 }
 #endif
 
+static const struct drm_exynos_ipp_limit rotator_4210_rbg888_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 8, SZ_16K }, .v = { 8, SZ_16K }) },
+	{ IPP_SIZE_LIMIT(AREA, .h.align = 4, .v.align = 4) },
+};
+
+static const struct drm_exynos_ipp_limit rotator_4412_rbg888_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 8, SZ_8K }, .v = { 8, SZ_8K }) },
+	{ IPP_SIZE_LIMIT(AREA, .h.align = 4, .v.align = 4) },
+};
+
+static const struct drm_exynos_ipp_limit rotator_5250_rbg888_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 8, SZ_8K }, .v = { 8, SZ_8K }) },
+	{ IPP_SIZE_LIMIT(AREA, .h.align = 2, .v.align = 2) },
+};
+
+static const struct drm_exynos_ipp_limit rotator_4210_yuv_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 32, SZ_64K }, .v = { 32, SZ_64K }) },
+	{ IPP_SIZE_LIMIT(AREA, .h.align = 8, .v.align = 8) },
+};
+
+static const struct drm_exynos_ipp_limit rotator_4412_yuv_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 32, SZ_32K }, .v = { 32, SZ_32K }) },
+	{ IPP_SIZE_LIMIT(AREA, .h.align = 8, .v.align = 8) },
+};
+
+static const struct exynos_drm_ipp_formats rotator_4210_formats[] = {
+	{ IPP_SRCDST_FORMAT(XRGB8888, rotator_4210_rbg888_limits) },
+	{ IPP_SRCDST_FORMAT(NV12, rotator_4210_yuv_limits) },
+};
+
+static const struct exynos_drm_ipp_formats rotator_4412_formats[] = {
+	{ IPP_SRCDST_FORMAT(XRGB8888, rotator_4412_rbg888_limits) },
+	{ IPP_SRCDST_FORMAT(NV12, rotator_4412_yuv_limits) },
+};
+
+static const struct exynos_drm_ipp_formats rotator_5250_formats[] = {
+	{ IPP_SRCDST_FORMAT(XRGB8888, rotator_5250_rbg888_limits) },
+	{ IPP_SRCDST_FORMAT(NV12, rotator_4412_yuv_limits) },
+};
+
+static const struct rot_variant rotator_4210_data = {
+	.formats = rotator_4210_formats,
+	.num_formats = ARRAY_SIZE(rotator_4210_formats),
+};
+
+static const struct rot_variant rotator_4412_data = {
+	.formats = rotator_4412_formats,
+	.num_formats = ARRAY_SIZE(rotator_4412_formats),
+};
+
+static const struct rot_variant rotator_5250_data = {
+	.formats = rotator_5250_formats,
+	.num_formats = ARRAY_SIZE(rotator_5250_formats),
+};
+
+static const struct of_device_id exynos_rotator_match[] = {
+	{
+		.compatible = "samsung,exynos4210-rotator",
+		.data = &rotator_4210_data,
+	}, {
+		.compatible = "samsung,exynos4212-rotator",
+		.data = &rotator_4412_data,
+	}, {
+		.compatible = "samsung,exynos5250-rotator",
+		.data = &rotator_5250_data,
+	}, {
+	},
+};
+MODULE_DEVICE_TABLE(of, exynos_rotator_match);
+
 static const struct dev_pm_ops rotator_pm_ops = {
 	SET_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend,
 				pm_runtime_force_resume)
@@ -820,7 +438,7 @@ struct platform_driver rotator_driver = {
 	.probe		= rotator_probe,
 	.remove		= rotator_remove,
 	.driver		= {
-		.name	= "exynos-rot",
+		.name	= "exynos-rotator",
 		.owner	= THIS_MODULE,
 		.pm	= &rotator_pm_ops,
 		.of_match_table = exynos_rotator_match,
diff --git a/drivers/gpu/drm/exynos/exynos_drm_scaler.c b/drivers/gpu/drm/exynos/exynos_drm_scaler.c
new file mode 100644
index 000000000000..91d4382343d0
--- /dev/null
+++ b/drivers/gpu/drm/exynos/exynos_drm_scaler.c
@@ -0,0 +1,694 @@
+/*
+ * Copyright (C) 2017 Samsung Electronics Co.Ltd
+ * Author:
+ *	Andrzej Pietrasiewicz <andrzej.p@samsung.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundationr
+ */
+
+#include <linux/kernel.h>
+#include <linux/component.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/platform_device.h>
+#include <linux/clk.h>
+#include <linux/of_device.h>
+#include <linux/pm_runtime.h>
+
+#include <drm/drmP.h>
+#include <drm/exynos_drm.h>
+#include "regs-scaler.h"
+#include "exynos_drm_fb.h"
+#include "exynos_drm_drv.h"
+#include "exynos_drm_iommu.h"
+#include "exynos_drm_ipp.h"
+
+#define scaler_read(offset)		readl(scaler->regs + (offset))
+#define scaler_write(cfg, offset)	writel(cfg, scaler->regs + (offset))
+#define SCALER_MAX_CLK			4
+#define SCALER_AUTOSUSPEND_DELAY	2000
+
+struct scaler_data {
+	const char	*clk_name[SCALER_MAX_CLK];
+	unsigned int	num_clk;
+	const struct exynos_drm_ipp_formats *formats;
+	unsigned int	num_formats;
+};
+
+struct scaler_context {
+	struct exynos_drm_ipp		ipp;
+	struct drm_device		*drm_dev;
+	struct device			*dev;
+	void __iomem			*regs;
+	struct clk			*clock[SCALER_MAX_CLK];
+	struct exynos_drm_ipp_task	*task;
+	const struct scaler_data	*scaler_data;
+};
+
+static u32 scaler_get_format(u32 drm_fmt)
+{
+	switch (drm_fmt) {
+	case DRM_FORMAT_NV21:
+		return SCALER_YUV420_2P_UV;
+	case DRM_FORMAT_NV12:
+		return SCALER_YUV420_2P_VU;
+	case DRM_FORMAT_YUV420:
+		return SCALER_YUV420_3P;
+	case DRM_FORMAT_YUYV:
+		return SCALER_YUV422_1P_YUYV;
+	case DRM_FORMAT_UYVY:
+		return SCALER_YUV422_1P_UYVY;
+	case DRM_FORMAT_YVYU:
+		return SCALER_YUV422_1P_YVYU;
+	case DRM_FORMAT_NV61:
+		return SCALER_YUV422_2P_UV;
+	case DRM_FORMAT_NV16:
+		return SCALER_YUV422_2P_VU;
+	case DRM_FORMAT_YUV422:
+		return SCALER_YUV422_3P;
+	case DRM_FORMAT_NV42:
+		return SCALER_YUV444_2P_UV;
+	case DRM_FORMAT_NV24:
+		return SCALER_YUV444_2P_VU;
+	case DRM_FORMAT_YUV444:
+		return SCALER_YUV444_3P;
+	case DRM_FORMAT_RGB565:
+		return SCALER_RGB_565;
+	case DRM_FORMAT_XRGB1555:
+		return SCALER_ARGB1555;
+	case DRM_FORMAT_ARGB1555:
+		return SCALER_ARGB1555;
+	case DRM_FORMAT_XRGB4444:
+		return SCALER_ARGB4444;
+	case DRM_FORMAT_ARGB4444:
+		return SCALER_ARGB4444;
+	case DRM_FORMAT_XRGB8888:
+		return SCALER_ARGB8888;
+	case DRM_FORMAT_ARGB8888:
+		return SCALER_ARGB8888;
+	case DRM_FORMAT_RGBX8888:
+		return SCALER_RGBA8888;
+	case DRM_FORMAT_RGBA8888:
+		return SCALER_RGBA8888;
+	default:
+		break;
+	}
+
+	return 0;
+}
+
+static inline void scaler_enable_int(struct scaler_context *scaler)
+{
+	u32 val;
+
+	val = SCALER_INT_EN_TIMEOUT |
+		SCALER_INT_EN_ILLEGAL_BLEND |
+		SCALER_INT_EN_ILLEGAL_RATIO |
+		SCALER_INT_EN_ILLEGAL_DST_HEIGHT |
+		SCALER_INT_EN_ILLEGAL_DST_WIDTH |
+		SCALER_INT_EN_ILLEGAL_DST_V_POS |
+		SCALER_INT_EN_ILLEGAL_DST_H_POS |
+		SCALER_INT_EN_ILLEGAL_DST_C_SPAN |
+		SCALER_INT_EN_ILLEGAL_DST_Y_SPAN |
+		SCALER_INT_EN_ILLEGAL_DST_CR_BASE |
+		SCALER_INT_EN_ILLEGAL_DST_CB_BASE |
+		SCALER_INT_EN_ILLEGAL_DST_Y_BASE |
+		SCALER_INT_EN_ILLEGAL_DST_COLOR |
+		SCALER_INT_EN_ILLEGAL_SRC_HEIGHT |
+		SCALER_INT_EN_ILLEGAL_SRC_WIDTH |
+		SCALER_INT_EN_ILLEGAL_SRC_CV_POS |
+		SCALER_INT_EN_ILLEGAL_SRC_CH_POS |
+		SCALER_INT_EN_ILLEGAL_SRC_YV_POS |
+		SCALER_INT_EN_ILLEGAL_SRC_YH_POS |
+		SCALER_INT_EN_ILLEGAL_DST_SPAN |
+		SCALER_INT_EN_ILLEGAL_SRC_Y_SPAN |
+		SCALER_INT_EN_ILLEGAL_SRC_CR_BASE |
+		SCALER_INT_EN_ILLEGAL_SRC_CB_BASE |
+		SCALER_INT_EN_ILLEGAL_SRC_Y_BASE |
+		SCALER_INT_EN_ILLEGAL_SRC_COLOR |
+		SCALER_INT_EN_FRAME_END;
+	scaler_write(val, SCALER_INT_EN);
+}
+
+static inline void scaler_set_src_fmt(struct scaler_context *scaler,
+	u32 src_fmt)
+{
+	u32 val;
+
+	val = SCALER_SRC_CFG_SET_COLOR_FORMAT(src_fmt);
+	scaler_write(val, SCALER_SRC_CFG);
+}
+
+static inline void scaler_set_src_base(struct scaler_context *scaler,
+	struct exynos_drm_ipp_buffer *src_buf)
+{
+	static unsigned int bases[] = {
+		SCALER_SRC_Y_BASE,
+		SCALER_SRC_CB_BASE,
+		SCALER_SRC_CR_BASE,
+	};
+	int i;
+
+	for (i = 0; i < src_buf->format->num_planes; ++i)
+		scaler_write(src_buf->dma_addr[i], bases[i]);
+}
+
+static inline void scaler_set_src_span(struct scaler_context *scaler,
+	struct exynos_drm_ipp_buffer *src_buf)
+{
+	u32 val;
+
+	val = SCALER_SRC_SPAN_SET_Y_SPAN(src_buf->buf.pitch[0] /
+		src_buf->format->cpp[0]);
+
+	if (src_buf->format->num_planes > 1)
+		val |= SCALER_SRC_SPAN_SET_C_SPAN(src_buf->buf.pitch[1]);
+
+	scaler_write(val, SCALER_SRC_SPAN);
+}
+
+static inline void scaler_set_src_luma_pos(struct scaler_context *scaler,
+	struct drm_exynos_ipp_task_rect *src_pos)
+{
+	u32 val;
+
+	val = SCALER_SRC_Y_POS_SET_YH_POS(src_pos->x << 2);
+	val |=  SCALER_SRC_Y_POS_SET_YV_POS(src_pos->y << 2);
+	scaler_write(val, SCALER_SRC_Y_POS);
+	scaler_write(val, SCALER_SRC_C_POS); /* ATTENTION! */
+}
+
+static inline void scaler_set_src_wh(struct scaler_context *scaler,
+	struct drm_exynos_ipp_task_rect *src_pos)
+{
+	u32 val;
+
+	val = SCALER_SRC_WH_SET_WIDTH(src_pos->w);
+	val |= SCALER_SRC_WH_SET_HEIGHT(src_pos->h);
+	scaler_write(val, SCALER_SRC_WH);
+}
+
+static inline void scaler_set_dst_fmt(struct scaler_context *scaler,
+	u32 dst_fmt)
+{
+	u32 val;
+
+	val = SCALER_DST_CFG_SET_COLOR_FORMAT(dst_fmt);
+	scaler_write(val, SCALER_DST_CFG);
+}
+
+static inline void scaler_set_dst_base(struct scaler_context *scaler,
+	struct exynos_drm_ipp_buffer *dst_buf)
+{
+	static unsigned int bases[] = {
+		SCALER_DST_Y_BASE,
+		SCALER_DST_CB_BASE,
+		SCALER_DST_CR_BASE,
+	};
+	int i;
+
+	for (i = 0; i < dst_buf->format->num_planes; ++i)
+		scaler_write(dst_buf->dma_addr[i], bases[i]);
+}
+
+static inline void scaler_set_dst_span(struct scaler_context *scaler,
+	struct exynos_drm_ipp_buffer *dst_buf)
+{
+	u32 val;
+
+	val = SCALER_DST_SPAN_SET_Y_SPAN(dst_buf->buf.pitch[0] /
+		dst_buf->format->cpp[0]);
+
+	if (dst_buf->format->num_planes > 1)
+		val |= SCALER_DST_SPAN_SET_C_SPAN(dst_buf->buf.pitch[1]);
+
+	scaler_write(val, SCALER_DST_SPAN);
+}
+
+static inline void scaler_set_dst_luma_pos(struct scaler_context *scaler,
+	struct drm_exynos_ipp_task_rect *dst_pos)
+{
+	u32 val;
+
+	val = SCALER_DST_WH_SET_WIDTH(dst_pos->w);
+	val |= SCALER_DST_WH_SET_HEIGHT(dst_pos->h);
+	scaler_write(val, SCALER_DST_WH);
+}
+
+static inline void scaler_set_dst_wh(struct scaler_context *scaler,
+	struct drm_exynos_ipp_task_rect *dst_pos)
+{
+	u32 val;
+
+	val = SCALER_DST_POS_SET_H_POS(dst_pos->x);
+	val |= SCALER_DST_POS_SET_V_POS(dst_pos->y);
+	scaler_write(val, SCALER_DST_POS);
+}
+
+static inline void scaler_set_hv_ratio(struct scaler_context *scaler,
+	unsigned int rotation,
+	struct drm_exynos_ipp_task_rect *src_pos,
+	struct drm_exynos_ipp_task_rect *dst_pos)
+{
+	u32 val, h_ratio, v_ratio;
+
+	if (drm_rotation_90_or_270(rotation)) {
+		h_ratio = (src_pos->h << 16) / dst_pos->w;
+		v_ratio = (src_pos->w << 16) / dst_pos->h;
+	} else {
+		h_ratio = (src_pos->w << 16) / dst_pos->w;
+		v_ratio = (src_pos->h << 16) / dst_pos->h;
+	}
+
+	val = SCALER_H_RATIO_SET(h_ratio);
+	scaler_write(val, SCALER_H_RATIO);
+
+	val = SCALER_V_RATIO_SET(v_ratio);
+	scaler_write(val, SCALER_V_RATIO);
+}
+
+static inline void scaler_set_rotation(struct scaler_context *scaler,
+	unsigned int rotation)
+{
+	u32 val = 0;
+
+	if (rotation & DRM_MODE_ROTATE_90)
+		val |= SCALER_ROT_CFG_SET_ROTMODE(SCALER_ROT_MODE_90);
+	else if (rotation & DRM_MODE_ROTATE_180)
+		val |= SCALER_ROT_CFG_SET_ROTMODE(SCALER_ROT_MODE_180);
+	else if (rotation & DRM_MODE_ROTATE_270)
+		val |= SCALER_ROT_CFG_SET_ROTMODE(SCALER_ROT_MODE_270);
+	if (rotation & DRM_MODE_REFLECT_X)
+		val |= SCALER_ROT_CFG_FLIP_X_EN;
+	if (rotation & DRM_MODE_REFLECT_Y)
+		val |= SCALER_ROT_CFG_FLIP_Y_EN;
+	scaler_write(val, SCALER_ROT_CFG);
+}
+
+static inline void scaler_set_csc(struct scaler_context *scaler,
+	const struct drm_format_info *fmt)
+{
+	static const u32 csc_mtx[2][3][3] = {
+		{ /* YCbCr to RGB */
+			{0x254, 0x000, 0x331},
+			{0x254, 0xf38, 0xe60},
+			{0x254, 0x409, 0x000},
+		},
+		{ /* RGB to YCbCr */
+			{0x084, 0x102, 0x032},
+			{0xfb4, 0xf6b, 0x0e1},
+			{0x0e1, 0xf44, 0xfdc},
+		},
+	};
+	int i, j, dir;
+
+	switch (fmt->format) {
+	case DRM_FORMAT_RGB565:
+	case DRM_FORMAT_XRGB1555:
+	case DRM_FORMAT_ARGB1555:
+	case DRM_FORMAT_XRGB4444:
+	case DRM_FORMAT_ARGB4444:
+	case DRM_FORMAT_XRGB8888:
+	case DRM_FORMAT_ARGB8888:
+	case DRM_FORMAT_RGBX8888:
+	case DRM_FORMAT_RGBA8888:
+		dir = 1;
+		break;
+	default:
+		dir = 0;
+	}
+
+	for (i = 0; i < 3; i++)
+		for (j = 0; j < 3; j++)
+			scaler_write(csc_mtx[dir][i][j], SCALER_CSC_COEF(j, i));
+}
+
+static inline void scaler_set_timer(struct scaler_context *scaler,
+	unsigned int timer, unsigned int divider)
+{
+	u32 val;
+
+	val = SCALER_TIMEOUT_CTRL_TIMER_ENABLE;
+	val |= SCALER_TIMEOUT_CTRL_SET_TIMER_VALUE(timer);
+	val |= SCALER_TIMEOUT_CTRL_SET_TIMER_DIV(divider);
+	scaler_write(val, SCALER_TIMEOUT_CTRL);
+}
+
+static inline void scaler_start_hw(struct scaler_context *scaler)
+{
+	scaler_write(SCALER_CFG_START_CMD, SCALER_CFG);
+}
+
+static int scaler_commit(struct exynos_drm_ipp *ipp,
+			  struct exynos_drm_ipp_task *task)
+{
+	struct scaler_context *scaler =
+			container_of(ipp, struct scaler_context, ipp);
+
+	u32 src_fmt = scaler_get_format(task->src.buf.fourcc);
+	struct drm_exynos_ipp_task_rect *src_pos = &task->src.rect;
+
+	u32 dst_fmt = scaler_get_format(task->dst.buf.fourcc);
+	struct drm_exynos_ipp_task_rect *dst_pos = &task->dst.rect;
+
+	scaler->task = task;
+
+	pm_runtime_get_sync(scaler->dev);
+
+	scaler_set_src_fmt(scaler, src_fmt);
+	scaler_set_src_base(scaler, &task->src);
+	scaler_set_src_span(scaler, &task->src);
+	scaler_set_src_luma_pos(scaler, src_pos);
+	scaler_set_src_wh(scaler, src_pos);
+
+	scaler_set_dst_fmt(scaler, dst_fmt);
+	scaler_set_dst_base(scaler, &task->dst);
+	scaler_set_dst_span(scaler, &task->dst);
+	scaler_set_dst_luma_pos(scaler, dst_pos);
+	scaler_set_dst_wh(scaler, dst_pos);
+
+	scaler_set_hv_ratio(scaler, task->transform.rotation, src_pos, dst_pos);
+	scaler_set_rotation(scaler, task->transform.rotation);
+
+	scaler_set_csc(scaler, task->src.format);
+
+	scaler_set_timer(scaler, 0xffff, 0xf);
+
+	scaler_enable_int(scaler);
+	scaler_start_hw(scaler);
+
+	return 0;
+}
+
+static struct exynos_drm_ipp_funcs ipp_funcs = {
+	.commit = scaler_commit,
+};
+
+static inline void scaler_disable_int(struct scaler_context *scaler)
+{
+	scaler_write(0, SCALER_INT_EN);
+}
+
+static inline u32 scaler_get_int_status(struct scaler_context *scaler)
+{
+	return scaler_read(SCALER_INT_STATUS);
+}
+
+static inline int scaler_task_done(u32 val)
+{
+	return val & SCALER_INT_STATUS_FRAME_END ? 0 : -EINVAL;
+}
+
+static irqreturn_t scaler_irq_handler(int irq, void *arg)
+{
+	struct scaler_context *scaler = arg;
+
+	u32 val = scaler_get_int_status(scaler);
+
+	scaler_disable_int(scaler);
+
+	if (scaler->task) {
+		struct exynos_drm_ipp_task *task = scaler->task;
+
+		scaler->task = NULL;
+		pm_runtime_mark_last_busy(scaler->dev);
+		pm_runtime_put_autosuspend(scaler->dev);
+		exynos_drm_ipp_task_done(task, scaler_task_done(val));
+	}
+
+	return IRQ_HANDLED;
+}
+
+static int scaler_bind(struct device *dev, struct device *master, void *data)
+{
+	struct scaler_context *scaler = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &scaler->ipp;
+
+	scaler->drm_dev = drm_dev;
+	drm_iommu_attach_device(drm_dev, dev);
+
+	exynos_drm_ipp_register(drm_dev, ipp, &ipp_funcs,
+			DRM_EXYNOS_IPP_CAP_CROP | DRM_EXYNOS_IPP_CAP_ROTATE |
+			DRM_EXYNOS_IPP_CAP_SCALE | DRM_EXYNOS_IPP_CAP_CONVERT,
+			scaler->scaler_data->formats,
+			scaler->scaler_data->num_formats, "scaler");
+
+	dev_info(dev, "The exynos scaler has been probed successfully\n");
+
+	return 0;
+}
+
+static void scaler_unbind(struct device *dev, struct device *master,
+			void *data)
+{
+	struct scaler_context *scaler = dev_get_drvdata(dev);
+	struct drm_device *drm_dev = data;
+	struct exynos_drm_ipp *ipp = &scaler->ipp;
+
+	exynos_drm_ipp_unregister(drm_dev, ipp);
+	drm_iommu_detach_device(scaler->drm_dev, scaler->dev);
+}
+
+static const struct component_ops scaler_component_ops = {
+	.bind	= scaler_bind,
+	.unbind = scaler_unbind,
+};
+
+static int scaler_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct resource	*regs_res;
+	struct scaler_context *scaler;
+	int irq;
+	int ret, i;
+
+	scaler = devm_kzalloc(dev, sizeof(*scaler), GFP_KERNEL);
+	if (!scaler)
+		return -ENOMEM;
+
+	scaler->scaler_data =
+		(struct scaler_data *)of_device_get_match_data(dev);
+
+	scaler->dev = dev;
+	regs_res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	scaler->regs = devm_ioremap_resource(dev, regs_res);
+	if (IS_ERR(scaler->regs))
+		return PTR_ERR(scaler->regs);
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0) {
+		dev_err(dev, "failed to get irq\n");
+		return irq;
+	}
+
+	ret = devm_request_threaded_irq(dev, irq, NULL,	scaler_irq_handler,
+					IRQF_ONESHOT, "drm_scaler", scaler);
+	if (ret < 0) {
+		dev_err(dev, "failed to request irq\n");
+		return ret;
+	}
+
+	for (i = 0; i < scaler->scaler_data->num_clk; ++i) {
+		scaler->clock[i] = devm_clk_get(dev,
+					      scaler->scaler_data->clk_name[i]);
+		if (IS_ERR(scaler->clock[i])) {
+			dev_err(dev, "failed to get clock\n");
+			return PTR_ERR(scaler->clock[i]);
+		}
+	}
+
+	pm_runtime_use_autosuspend(dev);
+	pm_runtime_set_autosuspend_delay(dev, SCALER_AUTOSUSPEND_DELAY);
+	pm_runtime_enable(dev);
+	platform_set_drvdata(pdev, scaler);
+
+	ret = component_add(dev, &scaler_component_ops);
+	if (ret)
+		goto err_ippdrv_register;
+
+	return 0;
+
+err_ippdrv_register:
+	pm_runtime_dont_use_autosuspend(dev);
+	pm_runtime_disable(dev);
+	return ret;
+}
+
+static int scaler_remove(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+
+	component_del(dev, &scaler_component_ops);
+	pm_runtime_dont_use_autosuspend(dev);
+	pm_runtime_disable(dev);
+
+	return 0;
+}
+
+#ifdef CONFIG_PM
+
+static int clk_disable_unprepare_wrapper(struct clk *clk)
+{
+	clk_disable_unprepare(clk);
+
+	return 0;
+}
+
+static int scaler_clk_ctrl(struct scaler_context *scaler, bool enable)
+{
+	int (*clk_fun)(struct clk *clk), i;
+
+	clk_fun = enable ? clk_prepare_enable : clk_disable_unprepare_wrapper;
+
+	for (i = 0; i < scaler->scaler_data->num_clk; ++i)
+		clk_fun(scaler->clock[i]);
+
+	return 0;
+}
+
+static int scaler_runtime_suspend(struct device *dev)
+{
+	struct scaler_context *scaler = dev_get_drvdata(dev);
+
+	return  scaler_clk_ctrl(scaler, false);
+}
+
+static int scaler_runtime_resume(struct device *dev)
+{
+	struct scaler_context *scaler = dev_get_drvdata(dev);
+
+	return  scaler_clk_ctrl(scaler, true);
+}
+#endif
+
+static const struct dev_pm_ops scaler_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(pm_runtime_force_suspend,
+				pm_runtime_force_resume)
+	SET_RUNTIME_PM_OPS(scaler_runtime_suspend, scaler_runtime_resume, NULL)
+};
+
+static const struct drm_exynos_ipp_limit scaler_5420_two_pixel_hv_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 16, SZ_8K }, .v = { 16, SZ_8K }) },
+	{ IPP_SIZE_LIMIT(AREA, .h.align = 2, .v.align = 2) },
+	{ IPP_SCALE_LIMIT(.h = { 65536 * 1 / 4, 65536 * 16 },
+			  .v = { 65536 * 1 / 4, 65536 * 16 }) },
+};
+
+static const struct drm_exynos_ipp_limit scaler_5420_two_pixel_h_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 16, SZ_8K }, .v = { 16, SZ_8K }) },
+	{ IPP_SIZE_LIMIT(AREA, .h.align = 2, .v.align = 1) },
+	{ IPP_SCALE_LIMIT(.h = { 65536 * 1 / 4, 65536 * 16 },
+			  .v = { 65536 * 1 / 4, 65536 * 16 }) },
+};
+
+static const struct drm_exynos_ipp_limit scaler_5420_one_pixel_limits[] = {
+	{ IPP_SIZE_LIMIT(BUFFER, .h = { 16, SZ_8K }, .v = { 16, SZ_8K }) },
+	{ IPP_SCALE_LIMIT(.h = { 65536 * 1 / 4, 65536 * 16 },
+			  .v = { 65536 * 1 / 4, 65536 * 16 }) },
+};
+
+static const struct exynos_drm_ipp_formats exynos5420_formats[] = {
+	/* SCALER_YUV420_2P_UV */
+	{ IPP_SRCDST_FORMAT(NV21, scaler_5420_two_pixel_hv_limits) },
+
+	/* SCALER_YUV420_2P_VU */
+	{ IPP_SRCDST_FORMAT(NV12, scaler_5420_two_pixel_hv_limits) },
+
+	/* SCALER_YUV420_3P */
+	{ IPP_SRCDST_FORMAT(YUV420, scaler_5420_two_pixel_hv_limits) },
+
+	/* SCALER_YUV422_1P_YUYV */
+	{ IPP_SRCDST_FORMAT(YUYV, scaler_5420_two_pixel_h_limits) },
+
+	/* SCALER_YUV422_1P_UYVY */
+	{ IPP_SRCDST_FORMAT(UYVY, scaler_5420_two_pixel_h_limits) },
+
+	/* SCALER_YUV422_1P_YVYU */
+	{ IPP_SRCDST_FORMAT(YVYU, scaler_5420_two_pixel_h_limits) },
+
+	/* SCALER_YUV422_2P_UV */
+	{ IPP_SRCDST_FORMAT(NV61, scaler_5420_two_pixel_h_limits) },
+
+	/* SCALER_YUV422_2P_VU */
+	{ IPP_SRCDST_FORMAT(NV16, scaler_5420_two_pixel_h_limits) },
+
+	/* SCALER_YUV422_3P */
+	{ IPP_SRCDST_FORMAT(YUV422, scaler_5420_two_pixel_h_limits) },
+
+	/* SCALER_YUV444_2P_UV */
+	{ IPP_SRCDST_FORMAT(NV42, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_YUV444_2P_VU */
+	{ IPP_SRCDST_FORMAT(NV24, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_YUV444_3P */
+	{ IPP_SRCDST_FORMAT(YUV444, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_RGB_565 */
+	{ IPP_SRCDST_FORMAT(RGB565, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_ARGB1555 */
+	{ IPP_SRCDST_FORMAT(XRGB1555, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_ARGB1555 */
+	{ IPP_SRCDST_FORMAT(ARGB1555, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_ARGB4444 */
+	{ IPP_SRCDST_FORMAT(XRGB4444, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_ARGB4444 */
+	{ IPP_SRCDST_FORMAT(ARGB4444, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_ARGB8888 */
+	{ IPP_SRCDST_FORMAT(XRGB8888, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_ARGB8888 */
+	{ IPP_SRCDST_FORMAT(ARGB8888, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_RGBA8888 */
+	{ IPP_SRCDST_FORMAT(RGBX8888, scaler_5420_one_pixel_limits) },
+
+	/* SCALER_RGBA8888 */
+	{ IPP_SRCDST_FORMAT(RGBA8888, scaler_5420_one_pixel_limits) },
+};
+
+static const struct scaler_data exynos5420_data = {
+	.clk_name	= {"mscl"},
+	.num_clk	= 1,
+	.formats	= exynos5420_formats,
+	.num_formats	= ARRAY_SIZE(exynos5420_formats),
+};
+
+static const struct scaler_data exynos5433_data = {
+	.clk_name	= {"pclk", "aclk", "aclk_xiu"},
+	.num_clk	= 3,
+	.formats	= exynos5420_formats, /* intentional */
+	.num_formats	= ARRAY_SIZE(exynos5420_formats),
+};
+
+static const struct of_device_id exynos_scaler_match[] = {
+	{
+		.compatible = "samsung,exynos5420-scaler",
+		.data = &exynos5420_data,
+	}, {
+		.compatible = "samsung,exynos5433-scaler",
+		.data = &exynos5433_data,
+	}, {
+	},
+};
+MODULE_DEVICE_TABLE(of, exynos_scaler_match);
+
+struct platform_driver scaler_driver = {
+	.probe		= scaler_probe,
+	.remove		= scaler_remove,
+	.driver		= {
+		.name	= "exynos-scaler",
+		.owner	= THIS_MODULE,
+		.pm	= &scaler_pm_ops,
+		.of_match_table = exynos_scaler_match,
+	},
+};
diff --git a/drivers/gpu/drm/exynos/exynos_drm_vidi.c b/drivers/gpu/drm/exynos/exynos_drm_vidi.c
index e6b0940b1ac2..19697c1362d8 100644
--- a/drivers/gpu/drm/exynos/exynos_drm_vidi.c
+++ b/drivers/gpu/drm/exynos/exynos_drm_vidi.c
@@ -319,7 +319,7 @@ static int vidi_get_modes(struct drm_connector *connector)
 		return -ENOMEM;
 	}
 
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 
 	return drm_add_edid_modes(connector, edid);
 }
@@ -344,7 +344,7 @@ static int vidi_create_connector(struct drm_encoder *encoder)
 	}
 
 	drm_connector_helper_add(connector, &vidi_connector_helper_funcs);
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/exynos/exynos_hdmi.c b/drivers/gpu/drm/exynos/exynos_hdmi.c
index abd84cbcf1c2..3a11c719a580 100644
--- a/drivers/gpu/drm/exynos/exynos_hdmi.c
+++ b/drivers/gpu/drm/exynos/exynos_hdmi.c
@@ -888,7 +888,7 @@ static int hdmi_get_modes(struct drm_connector *connector)
 		(hdata->dvi_mode ? "dvi monitor" : "hdmi monitor"),
 		edid->width_cm, edid->height_cm);
 
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 	cec_notifier_set_phys_addr_from_edid(hdata->notifier, edid);
 
 	ret = drm_add_edid_modes(connector, edid);
@@ -951,11 +951,9 @@ static int hdmi_create_connector(struct drm_encoder *encoder)
 	}
 
 	drm_connector_helper_add(connector, &hdmi_connector_helper_funcs);
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	if (hdata->bridge) {
-		encoder->bridge = hdata->bridge;
-		hdata->bridge->encoder = encoder;
 		ret = drm_bridge_attach(encoder, hdata->bridge, NULL);
 		if (ret)
 			DRM_ERROR("Failed to attach bridge\n");
@@ -1694,7 +1692,7 @@ static int hdmi_clk_init(struct hdmi_context *hdata)
 	if (!count)
 		return 0;
 
-	clks = devm_kzalloc(dev, sizeof(*clks) * count, GFP_KERNEL);
+	clks = devm_kcalloc(dev, count, sizeof(*clks), GFP_KERNEL);
 	if (!clks)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/exynos/exynos_mixer.c b/drivers/gpu/drm/exynos/exynos_mixer.c
index 257299ec95c4..272c79f5f5bf 100644
--- a/drivers/gpu/drm/exynos/exynos_mixer.c
+++ b/drivers/gpu/drm/exynos/exynos_mixer.c
@@ -473,7 +473,7 @@ static void vp_video_buffer(struct mixer_context *ctx,
 			chroma_addr[1] = chroma_addr[0] + 0x40;
 		} else {
 			luma_addr[1] = luma_addr[0] + fb->pitches[0];
-			chroma_addr[1] = chroma_addr[0] + fb->pitches[0];
+			chroma_addr[1] = chroma_addr[0] + fb->pitches[1];
 		}
 	} else {
 		luma_addr[1] = 0;
@@ -482,6 +482,7 @@ static void vp_video_buffer(struct mixer_context *ctx,
 
 	spin_lock_irqsave(&ctx->reg_slock, flags);
 
+	vp_reg_write(ctx, VP_SHADOW_UPDATE, 1);
 	/* interlace or progressive scan mode */
 	val = (test_bit(MXR_BIT_INTERLACE, &ctx->flags) ? ~0 : 0);
 	vp_reg_writemask(ctx, VP_MODE, val, VP_MODE_LINE_SKIP);
@@ -495,21 +496,23 @@ static void vp_video_buffer(struct mixer_context *ctx,
 	vp_reg_write(ctx, VP_IMG_SIZE_Y, VP_IMG_HSIZE(fb->pitches[0]) |
 		VP_IMG_VSIZE(fb->height));
 	/* chroma plane for NV12/NV21 is half the height of the luma plane */
-	vp_reg_write(ctx, VP_IMG_SIZE_C, VP_IMG_HSIZE(fb->pitches[0]) |
+	vp_reg_write(ctx, VP_IMG_SIZE_C, VP_IMG_HSIZE(fb->pitches[1]) |
 		VP_IMG_VSIZE(fb->height / 2));
 
 	vp_reg_write(ctx, VP_SRC_WIDTH, state->src.w);
-	vp_reg_write(ctx, VP_SRC_HEIGHT, state->src.h);
 	vp_reg_write(ctx, VP_SRC_H_POSITION,
 			VP_SRC_H_POSITION_VAL(state->src.x));
-	vp_reg_write(ctx, VP_SRC_V_POSITION, state->src.y);
-
 	vp_reg_write(ctx, VP_DST_WIDTH, state->crtc.w);
 	vp_reg_write(ctx, VP_DST_H_POSITION, state->crtc.x);
+
 	if (test_bit(MXR_BIT_INTERLACE, &ctx->flags)) {
+		vp_reg_write(ctx, VP_SRC_HEIGHT, state->src.h / 2);
+		vp_reg_write(ctx, VP_SRC_V_POSITION, state->src.y / 2);
 		vp_reg_write(ctx, VP_DST_HEIGHT, state->crtc.h / 2);
 		vp_reg_write(ctx, VP_DST_V_POSITION, state->crtc.y / 2);
 	} else {
+		vp_reg_write(ctx, VP_SRC_HEIGHT, state->src.h);
+		vp_reg_write(ctx, VP_SRC_V_POSITION, state->src.y);
 		vp_reg_write(ctx, VP_DST_HEIGHT, state->crtc.h);
 		vp_reg_write(ctx, VP_DST_V_POSITION, state->crtc.y);
 	}
@@ -699,6 +702,15 @@ static irqreturn_t mixer_irq_handler(int irq, void *arg)
 
 		/* interlace scan need to check shadow register */
 		if (test_bit(MXR_BIT_INTERLACE, &ctx->flags)) {
+			if (test_bit(MXR_BIT_VP_ENABLED, &ctx->flags) &&
+			    vp_reg_read(ctx, VP_SHADOW_UPDATE))
+				goto out;
+
+			base = mixer_reg_read(ctx, MXR_CFG);
+			shadow = mixer_reg_read(ctx, MXR_CFG_S);
+			if (base != shadow)
+				goto out;
+
 			base = mixer_reg_read(ctx, MXR_GRAPHIC_BASE(0));
 			shadow = mixer_reg_read(ctx, MXR_GRAPHIC_BASE_S(0));
 			if (base != shadow)
diff --git a/drivers/gpu/drm/exynos/regs-mixer.h b/drivers/gpu/drm/exynos/regs-mixer.h
index c311f571bdf9..189cfa2470a8 100644
--- a/drivers/gpu/drm/exynos/regs-mixer.h
+++ b/drivers/gpu/drm/exynos/regs-mixer.h
@@ -47,6 +47,7 @@
 #define MXR_MO				0x0304
 #define MXR_RESOLUTION			0x0310
 
+#define MXR_CFG_S			0x2004
 #define MXR_GRAPHIC0_BASE_S		0x2024
 #define MXR_GRAPHIC1_BASE_S		0x2044
 
diff --git a/drivers/gpu/drm/exynos/regs-scaler.h b/drivers/gpu/drm/exynos/regs-scaler.h
new file mode 100644
index 000000000000..fc7ccad75e74
--- /dev/null
+++ b/drivers/gpu/drm/exynos/regs-scaler.h
@@ -0,0 +1,426 @@
+/* drivers/gpu/drm/exynos/regs-scaler.h
+ *
+ * Copyright (c) 2017 Samsung Electronics Co., Ltd.
+ *		http://www.samsung.com/
+ * Author: Andrzej Pietrasiewicz <andrzej.p@samsung.com>
+ *
+ * Register definition file for Samsung scaler driver
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#ifndef EXYNOS_REGS_SCALER_H
+#define EXYNOS_REGS_SCALER_H
+
+/* Register part */
+
+/* Global setting */
+#define SCALER_STATUS			0x0	/* no shadow */
+#define SCALER_CFG			0x4
+
+/* Interrupt */
+#define SCALER_INT_EN			0x8	/* no shadow */
+#define SCALER_INT_STATUS		0xc	/* no shadow */
+
+/* SRC */
+#define SCALER_SRC_CFG			0x10
+#define SCALER_SRC_Y_BASE		0x14
+#define SCALER_SRC_CB_BASE		0x18
+#define SCALER_SRC_CR_BASE		0x294
+#define SCALER_SRC_SPAN			0x1c
+#define SCALER_SRC_Y_POS		0x20
+#define SCALER_SRC_WH			0x24
+#define SCALER_SRC_C_POS		0x28
+
+/* DST */
+#define SCALER_DST_CFG			0x30
+#define SCALER_DST_Y_BASE		0x34
+#define SCALER_DST_CB_BASE		0x38
+#define SCALER_DST_CR_BASE		0x298
+#define SCALER_DST_SPAN			0x3c
+#define SCALER_DST_WH			0x40
+#define SCALER_DST_POS			0x44
+
+/* Ratio */
+#define SCALER_H_RATIO			0x50
+#define SCALER_V_RATIO			0x54
+
+/* Rotation */
+#define SCALER_ROT_CFG			0x58
+
+/* Coefficient */
+/*
+ * YHCOEF_{x}{A|B|C|D}			CHCOEF_{x}{A|B|C|D}
+ *
+ *	A	B	C	D	A	B	C	D
+ * 0	60	64	68	6c	140	144	148	14c
+ * 1	70	74	78	7c	150	154	158	15c
+ * 2	80	84	88	8c	160	164	168	16c
+ * 3	90	94	98	9c	170	174	178	17c
+ * 4	a0	a4	a8	ac	180	184	188	18c
+ * 5	b0	b4	b8	bc	190	194	198	19c
+ * 6	c0	c4	c8	cc	1a0	1a4	1a8	1ac
+ * 7	d0	d4	d8	dc	1b0	1b4	1b8	1bc
+ * 8	e0	e4	e8	ec	1c0	1c4	1c8	1cc
+ *
+ *
+ * YVCOEF_{x}{A|B}			CVCOEF_{x}{A|B}
+ *
+ *	A	B			A	B
+ * 0	f0	f4			1d0	1d4
+ * 1	f8	fc			1d8	1dc
+ * 2	100	104			1e0	1e4
+ * 3	108	10c			1e8	1ec
+ * 4	110	114			1f0	1f4
+ * 5	118	11c			1f8	1fc
+ * 6	120	124			200	204
+ * 7	128	12c			208	20c
+ * 8	130	134			210	214
+ */
+#define _SCALER_HCOEF_DELTA(r, c)	((r) * 0x10 + (c) * 0x4)
+#define _SCALER_VCOEF_DELTA(r, c)	((r) * 0x8 + (c) * 0x4)
+
+#define SCALER_YHCOEF(r, c)		(0x60 + _SCALER_HCOEF_DELTA((r), (c)))
+#define SCALER_YVCOEF(r, c)		(0xf0 + _SCALER_VCOEF_DELTA((r), (c)))
+#define SCALER_CHCOEF(r, c)		(0x140 + _SCALER_HCOEF_DELTA((r), (c)))
+#define SCALER_CVCOEF(r, c)		(0x1d0 + _SCALER_VCOEF_DELTA((r), (c)))
+
+
+/* Color Space Conversion */
+#define SCALER_CSC_COEF(x, y)		(0x220 + (y) * 0xc + (x) * 0x4)
+
+/* Dithering */
+#define SCALER_DITH_CFG			0x250
+
+/* Version Number */
+#define SCALER_VER			0x260	/* no shadow */
+
+/* Cycle count and Timeout */
+#define SCALER_CYCLE_COUNT		0x278	/* no shadow */
+#define SCALER_TIMEOUT_CTRL		0x2c0	/* no shadow */
+#define SCALER_TIMEOUT_CNT		0x2c4	/* no shadow */
+
+/* Blending */
+#define SCALER_SRC_BLEND_COLOR		0x280
+#define SCALER_SRC_BLEND_ALPHA		0x284
+#define SCALER_DST_BLEND_COLOR		0x288
+#define SCALER_DST_BLEND_ALPHA		0x28c
+
+/* Color Fill */
+#define SCALER_FILL_COLOR		0x290
+
+/* Multiple Command Queue */
+#define SCALER_ADDR_Q_CONFIG		0x2a0	/* no shadow */
+#define SCALER_SRC_ADDR_Q_STATUS	0x2a4	/* no shadow */
+#define SCALER_SRC_ADDR_Q		0x2a8	/* no shadow */
+
+/* CRC */
+#define SCALER_CRC_COLOR00_10		0x2b0	/* no shadow */
+#define SCALER_CRC_COLOR20_30		0x2b4	/* no shadow */
+#define SCALER_CRC_COLOR01_11		0x2b8	/* no shadow */
+#define SCALER_CRC_COLOR21_31		0x2bc	/* no shadow */
+
+/* Shadow Registers */
+#define SCALER_SHADOW_OFFSET		0x1000
+
+
+/* Bit definition part */
+#define SCALER_MASK(hi_b, lo_b)		((1 << ((hi_b) - (lo_b) + 1)) - 1)
+#define SCALER_GET(reg, hi_b, lo_b)	\
+	(((reg) >> (lo_b)) & SCALER_MASK(hi_b, lo_b))
+#define SCALER_SET(val, hi_b, lo_b) \
+	(((val) & SCALER_MASK(hi_b, lo_b)) << lo_b)
+
+/* SCALER_STATUS */
+#define SCALER_STATUS_SCALER_RUNNING		(1 << 1)
+#define SCALER_STATUS_SCALER_READY_CLK_DOWN	(1 << 0)
+
+/* SCALER_CFG */
+#define SCALER_CFG_FILL_EN			(1 << 24)
+#define SCALER_CFG_BLEND_COLOR_DIVIDE_ALPHA_EN	(1 << 17)
+#define SCALER_CFG_BLEND_EN			(1 << 16)
+#define SCALER_CFG_CSC_Y_OFFSET_SRC_EN		(1 << 10)
+#define SCALER_CFG_CSC_Y_OFFSET_DST_EN		(1 << 9)
+#define SCALER_CFG_16_BURST_MODE		(1 << 8)
+#define SCALER_CFG_SOFT_RESET			(1 << 1)
+#define SCALER_CFG_START_CMD			(1 << 0)
+
+/* SCALER_INT_EN */
+#define SCALER_INT_EN_TIMEOUT			(1 << 31)
+#define SCALER_INT_EN_ILLEGAL_BLEND		(1 << 24)
+#define SCALER_INT_EN_ILLEGAL_RATIO		(1 << 23)
+#define SCALER_INT_EN_ILLEGAL_DST_HEIGHT	(1 << 22)
+#define SCALER_INT_EN_ILLEGAL_DST_WIDTH		(1 << 21)
+#define SCALER_INT_EN_ILLEGAL_DST_V_POS		(1 << 20)
+#define SCALER_INT_EN_ILLEGAL_DST_H_POS		(1 << 19)
+#define SCALER_INT_EN_ILLEGAL_DST_C_SPAN	(1 << 18)
+#define SCALER_INT_EN_ILLEGAL_DST_Y_SPAN	(1 << 17)
+#define SCALER_INT_EN_ILLEGAL_DST_CR_BASE	(1 << 16)
+#define SCALER_INT_EN_ILLEGAL_DST_CB_BASE	(1 << 15)
+#define SCALER_INT_EN_ILLEGAL_DST_Y_BASE	(1 << 14)
+#define SCALER_INT_EN_ILLEGAL_DST_COLOR		(1 << 13)
+#define SCALER_INT_EN_ILLEGAL_SRC_HEIGHT	(1 << 12)
+#define SCALER_INT_EN_ILLEGAL_SRC_WIDTH		(1 << 11)
+#define SCALER_INT_EN_ILLEGAL_SRC_CV_POS	(1 << 10)
+#define SCALER_INT_EN_ILLEGAL_SRC_CH_POS	(1 << 9)
+#define SCALER_INT_EN_ILLEGAL_SRC_YV_POS	(1 << 8)
+#define SCALER_INT_EN_ILLEGAL_SRC_YH_POS	(1 << 7)
+#define SCALER_INT_EN_ILLEGAL_DST_SPAN		(1 << 6)
+#define SCALER_INT_EN_ILLEGAL_SRC_Y_SPAN	(1 << 5)
+#define SCALER_INT_EN_ILLEGAL_SRC_CR_BASE	(1 << 4)
+#define SCALER_INT_EN_ILLEGAL_SRC_CB_BASE	(1 << 3)
+#define SCALER_INT_EN_ILLEGAL_SRC_Y_BASE	(1 << 2)
+#define SCALER_INT_EN_ILLEGAL_SRC_COLOR		(1 << 1)
+#define SCALER_INT_EN_FRAME_END			(1 << 0)
+
+/* SCALER_INT_STATUS */
+#define SCALER_INT_STATUS_TIMEOUT		(1 << 31)
+#define SCALER_INT_STATUS_ILLEGAL_BLEND		(1 << 24)
+#define SCALER_INT_STATUS_ILLEGAL_RATIO		(1 << 23)
+#define SCALER_INT_STATUS_ILLEGAL_DST_HEIGHT	(1 << 22)
+#define SCALER_INT_STATUS_ILLEGAL_DST_WIDTH	(1 << 21)
+#define SCALER_INT_STATUS_ILLEGAL_DST_V_POS	(1 << 20)
+#define SCALER_INT_STATUS_ILLEGAL_DST_H_POS	(1 << 19)
+#define SCALER_INT_STATUS_ILLEGAL_DST_C_SPAN	(1 << 18)
+#define SCALER_INT_STATUS_ILLEGAL_DST_Y_SPAN	(1 << 17)
+#define SCALER_INT_STATUS_ILLEGAL_DST_CR_BASE	(1 << 16)
+#define SCALER_INT_STATUS_ILLEGAL_DST_CB_BASE	(1 << 15)
+#define SCALER_INT_STATUS_ILLEGAL_DST_Y_BASE	(1 << 14)
+#define SCALER_INT_STATUS_ILLEGAL_DST_COLOR	(1 << 13)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_HEIGHT	(1 << 12)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_WIDTH	(1 << 11)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_CV_POS	(1 << 10)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_CH_POS	(1 << 9)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_YV_POS	(1 << 8)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_YH_POS	(1 << 7)
+#define SCALER_INT_STATUS_ILLEGAL_DST_SPAN	(1 << 6)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_Y_SPAN	(1 << 5)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_CR_BASE	(1 << 4)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_CB_BASE	(1 << 3)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_Y_BASE	(1 << 2)
+#define SCALER_INT_STATUS_ILLEGAL_SRC_COLOR	(1 << 1)
+#define SCALER_INT_STATUS_FRAME_END		(1 << 0)
+
+/* SCALER_SRC_CFG */
+#define SCALER_SRC_CFG_TILE_EN			(1 << 10)
+#define SCALER_SRC_CFG_GET_BYTE_SWAP(r)		SCALER_GET(r, 6, 5)
+#define SCALER_SRC_CFG_SET_BYTE_SWAP(v)		SCALER_SET(v, 6, 5)
+#define SCALER_SRC_CFG_GET_COLOR_FORMAT(r)	SCALER_GET(r, 4, 0)
+#define SCALER_SRC_CFG_SET_COLOR_FORMAT(v)	SCALER_SET(v, 4, 0)
+#define SCALER_YUV420_2P_UV			0
+#define SCALER_YUV422_2P_UV			2
+#define SCALER_YUV444_2P_UV			3
+#define SCALER_RGB_565				4
+#define SCALER_ARGB1555				5
+#define SCALER_ARGB8888				6
+#define SCALER_ARGB8888_PRE			7
+#define SCALER_YUV422_1P_YVYU			9
+#define SCALER_YUV422_1P_YUYV			10
+#define SCALER_YUV422_1P_UYVY			11
+#define SCALER_ARGB4444				12
+#define SCALER_L8A8				13
+#define SCALER_RGBA8888				14
+#define SCALER_L8				15
+#define SCALER_YUV420_2P_VU			16
+#define SCALER_YUV422_2P_VU			18
+#define SCALER_YUV444_2P_VU			19
+#define SCALER_YUV420_3P			20
+#define SCALER_YUV422_3P			22
+#define SCALER_YUV444_3P			23
+
+/* SCALER_SRC_SPAN */
+#define SCALER_SRC_SPAN_GET_C_SPAN(r)		SCALER_GET(r, 29, 16)
+#define SCALER_SRC_SPAN_SET_C_SPAN(v)		SCALER_SET(v, 29, 16)
+#define SCALER_SRC_SPAN_GET_Y_SPAN(r)		SCALER_GET(r, 13, 0)
+#define SCALER_SRC_SPAN_SET_Y_SPAN(v)		SCALER_SET(v, 13, 0)
+
+/* SCALER_SRC_Y_POS */
+#define SCALER_SRC_Y_POS_GET_YH_POS(r)		SCALER_GET(r, 31, 16)
+#define SCALER_SRC_Y_POS_SET_YH_POS(v)		SCALER_SET(v, 31, 16)
+#define SCALER_SRC_Y_POS_GET_YV_POS(r)		SCALER_GET(r, 15, 0)
+#define SCALER_SRC_Y_POS_SET_YV_POS(v)		SCALER_SET(v, 15, 0)
+
+/* SCALER_SRC_WH */
+#define SCALER_SRC_WH_GET_WIDTH(r)		SCALER_GET(r, 29, 16)
+#define SCALER_SRC_WH_SET_WIDTH(v)		SCALER_SET(v, 29, 16)
+#define SCALER_SRC_WH_GET_HEIGHT(r)		SCALER_GET(r, 13, 0)
+#define SCALER_SRC_WH_SET_HEIGHT(v)		SCALER_SET(v, 13, 0)
+
+/* SCALER_SRC_C_POS */
+#define SCALER_SRC_C_POS_GET_CH_POS(r)		SCALER_GET(r, 31, 16)
+#define SCALER_SRC_C_POS_SET_CH_POS(v)		SCALER_SET(v, 31, 16)
+#define SCALER_SRC_C_POS_GET_CV_POS(r)		SCALER_GET(r, 15, 0)
+#define SCALER_SRC_C_POS_SET_CV_POS(v)		SCALER_SET(v, 15, 0)
+
+/* SCALER_DST_CFG */
+#define SCALER_DST_CFG_GET_BYTE_SWAP(r)		SCALER_GET(r, 6, 5)
+#define SCALER_DST_CFG_SET_BYTE_SWAP(v)		SCALER_SET(v, 6, 5)
+#define SCALER_DST_CFG_GET_COLOR_FORMAT(r)	SCALER_GET(r, 4, 0)
+#define SCALER_DST_CFG_SET_COLOR_FORMAT(v)	SCALER_SET(v, 4, 0)
+
+/* SCALER_DST_SPAN */
+#define SCALER_DST_SPAN_GET_C_SPAN(r)		SCALER_GET(r, 29, 16)
+#define SCALER_DST_SPAN_SET_C_SPAN(v)		SCALER_SET(v, 29, 16)
+#define SCALER_DST_SPAN_GET_Y_SPAN(r)		SCALER_GET(r, 13, 0)
+#define SCALER_DST_SPAN_SET_Y_SPAN(v)		SCALER_SET(v, 13, 0)
+
+/* SCALER_DST_WH */
+#define SCALER_DST_WH_GET_WIDTH(r)		SCALER_GET(r, 29, 16)
+#define SCALER_DST_WH_SET_WIDTH(v)		SCALER_SET(v, 29, 16)
+#define SCALER_DST_WH_GET_HEIGHT(r)		SCALER_GET(r, 13, 0)
+#define SCALER_DST_WH_SET_HEIGHT(v)		SCALER_SET(v, 13, 0)
+
+/* SCALER_DST_POS */
+#define SCALER_DST_POS_GET_H_POS(r)		SCALER_GET(r, 29, 16)
+#define SCALER_DST_POS_SET_H_POS(v)		SCALER_SET(v, 29, 16)
+#define SCALER_DST_POS_GET_V_POS(r)		SCALER_GET(r, 13, 0)
+#define SCALER_DST_POS_SET_V_POS(v)		SCALER_SET(v, 13, 0)
+
+/* SCALER_H_RATIO */
+#define SCALER_H_RATIO_GET(r)			SCALER_GET(r, 18, 0)
+#define SCALER_H_RATIO_SET(v)			SCALER_SET(v, 18, 0)
+
+/* SCALER_V_RATIO */
+#define SCALER_V_RATIO_GET(r)			SCALER_GET(r, 18, 0)
+#define SCALER_V_RATIO_SET(v)			SCALER_SET(v, 18, 0)
+
+/* SCALER_ROT_CFG */
+#define SCALER_ROT_CFG_FLIP_X_EN		(1 << 3)
+#define SCALER_ROT_CFG_FLIP_Y_EN		(1 << 2)
+#define SCALER_ROT_CFG_GET_ROTMODE(r)		SCALER_GET(r, 1, 0)
+#define SCALER_ROT_CFG_SET_ROTMODE(v)		SCALER_SET(v, 1, 0)
+#define SCALER_ROT_MODE_90			1
+#define SCALER_ROT_MODE_180			2
+#define SCALER_ROT_MODE_270			3
+
+/* SCALER_HCOEF, SCALER_VCOEF */
+#define SCALER_COEF_SHIFT(i)			(16 * (1 - (i) % 2))
+#define SCALER_COEF_GET(r, i)			\
+				(((r) >> SCALER_COEF_SHIFT(i)) & 0x1ff)
+#define SCALER_COEF_SET(v, i)			\
+				(((v) & 0x1ff) << SCALER_COEF_SHIFT(i))
+
+/* SCALER_CSC_COEFxy */
+#define SCALER_CSC_COEF_GET(r)			SCALER_GET(r, 11, 0)
+#define SCALER_CSC_COEF_SET(v)			SCALER_SET(v, 11, 0)
+
+/* SCALER_DITH_CFG */
+#define SCALER_DITH_CFG_GET_R_TYPE(r)		SCALER_GET(r, 8, 6)
+#define SCALER_DITH_CFG_SET_R_TYPE(v)		SCALER_SET(v, 8, 6)
+#define SCALER_DITH_CFG_GET_G_TYPE(r)		SCALER_GET(r, 5, 3)
+#define SCALER_DITH_CFG_SET_G_TYPE(v)		SCALER_SET(v, 5, 3)
+#define SCALER_DITH_CFG_GET_B_TYPE(r)		SCALER_GET(r, 2, 0)
+#define SCALER_DITH_CFG_SET_B_TYPE(v)		SCALER_SET(v, 2, 0)
+
+/* SCALER_TIMEOUT_CTRL */
+#define SCALER_TIMEOUT_CTRL_GET_TIMER_VALUE(r)	SCALER_GET(r, 31, 16)
+#define SCALER_TIMEOUT_CTRL_SET_TIMER_VALUE(v)	SCALER_SET(v, 31, 16)
+#define SCALER_TIMEOUT_CTRL_GET_TIMER_DIV(r)	SCALER_GET(r, 7, 4)
+#define SCALER_TIMEOUT_CTRL_SET_TIMER_DIV(v)	SCALER_SET(v, 7, 4)
+#define SCALER_TIMEOUT_CTRL_TIMER_ENABLE	(1 << 0)
+
+/* SCALER_TIMEOUT_CNT */
+#define SCALER_TIMEOUT_CTRL_GET_TIMER_COUNT(r)	SCALER_GET(r, 31, 16)
+
+/* SCALER_SRC_BLEND_COLOR */
+#define SCALER_SRC_BLEND_COLOR_SEL_INV		(1 << 31)
+#define SCALER_SRC_BLEND_COLOR_GET_SEL(r)	SCALER_GET(r, 30, 29)
+#define SCALER_SRC_BLEND_COLOR_SET_SEL(v)	SCALER_SET(v, 30, 29)
+#define SCALER_SRC_BLEND_COLOR_OP_SEL_INV	(1 << 28)
+#define SCALER_SRC_BLEND_COLOR_GET_OP_SEL(r)	SCALER_GET(r, 27, 24)
+#define SCALER_SRC_BLEND_COLOR_SET_OP_SEL(v)	SCALER_SET(v, 27, 24)
+#define SCALER_SRC_BLEND_COLOR_GET_COLOR0(r)	SCALER_GET(r, 23, 16)
+#define SCALER_SRC_BLEND_COLOR_SET_COLOR0(v)	SCALER_SET(v, 23, 16)
+#define SCALER_SRC_BLEND_COLOR_GET_COLOR1(r)	SCALER_GET(r, 15, 8)
+#define SCALER_SRC_BLEND_COLOR_SET_COLOR1(v)	SCALER_SET(v, 15, 8)
+#define SCALER_SRC_BLEND_COLOR_GET_COLOR2(r)	SCALER_GET(r, 7, 0)
+#define SCALER_SRC_BLEND_COLOR_SET_COLOR2(v)	SCALER_SET(v, 7, 0)
+
+/* SCALER_SRC_BLEND_ALPHA */
+#define SCALER_SRC_BLEND_ALPHA_SEL_INV		(1 << 31)
+#define SCALER_SRC_BLEND_ALPHA_GET_SEL(r)	SCALER_GET(r, 30, 29)
+#define SCALER_SRC_BLEND_ALPHA_SET_SEL(v)	SCALER_SET(v, 30, 29)
+#define SCALER_SRC_BLEND_ALPHA_OP_SEL_INV	(1 << 28)
+#define SCALER_SRC_BLEND_ALPHA_GET_OP_SEL(r)	SCALER_GET(r, 27, 24)
+#define SCALER_SRC_BLEND_ALPHA_SET_OP_SEL(v)	SCALER_SET(v, 27, 24)
+#define SCALER_SRC_BLEND_ALPHA_GET_ALPHA(r)	SCALER_GET(r, 7, 0)
+#define SCALER_SRC_BLEND_ALPHA_SET_ALPHA(v)	SCALER_SET(v, 7, 0)
+
+/* SCALER_DST_BLEND_COLOR */
+#define SCALER_DST_BLEND_COLOR_SEL_INV		(1 << 31)
+#define SCALER_DST_BLEND_COLOR_GET_SEL(r)	SCALER_GET(r, 30, 29)
+#define SCALER_DST_BLEND_COLOR_SET_SEL(v)	SCALER_SET(v, 30, 29)
+#define SCALER_DST_BLEND_COLOR_OP_SEL_INV	(1 << 28)
+#define SCALER_DST_BLEND_COLOR_GET_OP_SEL(r)	SCALER_GET(r, 27, 24)
+#define SCALER_DST_BLEND_COLOR_SET_OP_SEL(v)	SCALER_SET(v, 27, 24)
+#define SCALER_DST_BLEND_COLOR_GET_COLOR0(r)	SCALER_GET(r, 23, 16)
+#define SCALER_DST_BLEND_COLOR_SET_COLOR0(v)	SCALER_SET(v, 23, 16)
+#define SCALER_DST_BLEND_COLOR_GET_COLOR1(r)	SCALER_GET(r, 15, 8)
+#define SCALER_DST_BLEND_COLOR_SET_COLOR1(v)	SCALER_SET(v, 15, 8)
+#define SCALER_DST_BLEND_COLOR_GET_COLOR2(r)	SCALER_GET(r, 7, 0)
+#define SCALER_DST_BLEND_COLOR_SET_COLOR2(v)	SCALER_SET(v, 7, 0)
+
+/* SCALER_DST_BLEND_ALPHA */
+#define SCALER_DST_BLEND_ALPHA_SEL_INV		(1 << 31)
+#define SCALER_DST_BLEND_ALPHA_GET_SEL(r)	SCALER_GET(r, 30, 29)
+#define SCALER_DST_BLEND_ALPHA_SET_SEL(v)	SCALER_SET(v, 30, 29)
+#define SCALER_DST_BLEND_ALPHA_OP_SEL_INV	(1 << 28)
+#define SCALER_DST_BLEND_ALPHA_GET_OP_SEL(r)	SCALER_GET(r, 27, 24)
+#define SCALER_DST_BLEND_ALPHA_SET_OP_SEL(v)	SCALER_SET(v, 27, 24)
+#define SCALER_DST_BLEND_ALPHA_GET_ALPHA(r)	SCALER_GET(r, 7, 0)
+#define SCALER_DST_BLEND_ALPHA_SET_ALPHA(v)	SCALER_SET(v, 7, 0)
+
+/* SCALER_FILL_COLOR */
+#define SCALER_FILL_COLOR_GET_ALPHA(r)		SCALER_GET(r, 31, 24)
+#define SCALER_FILL_COLOR_SET_ALPHA(v)		SCALER_SET(v, 31, 24)
+#define SCALER_FILL_COLOR_GET_FILL_COLOR0(r)	SCALER_GET(r, 23, 16)
+#define SCALER_FILL_COLOR_SET_FILL_COLOR0(v)	SCALER_SET(v, 23, 16)
+#define SCALER_FILL_COLOR_GET_FILL_COLOR1(r)	SCALER_GET(r, 15, 8)
+#define SCALER_FILL_COLOR_SET_FILL_COLOR1(v)	SCALER_SET(v, 15, 8)
+#define SCALER_FILL_COLOR_GET_FILL_COLOR2(r)	SCALER_GET(r, 7, 0)
+#define SCALER_FILL_COLOR_SET_FILL_COLOR2(v)	SCALER_SET(v, 7, 0)
+
+/* SCALER_ADDR_Q_CONFIG */
+#define SCALER_ADDR_Q_CONFIG_RST		(1 << 0)
+
+/* SCALER_SRC_ADDR_Q_STATUS */
+#define SCALER_SRC_ADDR_Q_STATUS_Y_FULL		(1 << 23)
+#define SCALER_SRC_ADDR_Q_STATUS_Y_EMPTY	(1 << 22)
+#define SCALER_SRC_ADDR_Q_STATUS_GET_Y_WR_IDX(r)	SCALER_GET(r, 21, 16)
+#define SCALER_SRC_ADDR_Q_STATUS_CB_FULL	(1 << 15)
+#define SCALER_SRC_ADDR_Q_STATUS_CB_EMPTY	(1 << 14)
+#define SCALER_SRC_ADDR_Q_STATUS_GET_CB_WR_IDX(r)	SCALER_GET(r, 13, 8)
+#define SCALER_SRC_ADDR_Q_STATUS_CR_FULL	(1 << 7)
+#define SCALER_SRC_ADDR_Q_STATUS_CR_EMPTY	(1 << 6)
+#define SCALER_SRC_ADDR_Q_STATUS_GET_CR_WR_IDX(r)	SCALER_GET(r, 5, 0)
+
+/* SCALER_DST_ADDR_Q_STATUS */
+#define SCALER_DST_ADDR_Q_STATUS_Y_FULL		(1 << 23)
+#define SCALER_DST_ADDR_Q_STATUS_Y_EMPTY	(1 << 22)
+#define SCALER_DST_ADDR_Q_STATUS_GET_Y_WR_IDX(r)	SCALER_GET(r, 21, 16)
+#define SCALER_DST_ADDR_Q_STATUS_CB_FULL	(1 << 15)
+#define SCALER_DST_ADDR_Q_STATUS_CB_EMPTY	(1 << 14)
+#define SCALER_DST_ADDR_Q_STATUS_GET_CB_WR_IDX(r)	SCALER_GET(r, 13, 8)
+#define SCALER_DST_ADDR_Q_STATUS_CR_FULL	(1 << 7)
+#define SCALER_DST_ADDR_Q_STATUS_CR_EMPTY	(1 << 6)
+#define SCALER_DST_ADDR_Q_STATUS_GET_CR_WR_IDX(r)	SCALER_GET(r, 5, 0)
+
+/* SCALER_CRC_COLOR00_10 */
+#define SCALER_CRC_COLOR00_10_GET_00(r)		SCALER_GET(r, 31, 16)
+#define SCALER_CRC_COLOR00_10_GET_10(r)		SCALER_GET(r, 15, 0)
+
+/* SCALER_CRC_COLOR20_30 */
+#define SCALER_CRC_COLOR20_30_GET_20(r)		SCALER_GET(r, 31, 16)
+#define SCALER_CRC_COLOR20_30_GET_30(r)		SCALER_GET(r, 15, 0)
+
+/* SCALER_CRC_COLOR01_11 */
+#define SCALER_CRC_COLOR01_11_GET_01(r)		SCALER_GET(r, 31, 16)
+#define SCALER_CRC_COLOR01_11_GET_11(r)		SCALER_GET(r, 15, 0)
+
+/* SCALER_CRC_COLOR21_31 */
+#define SCALER_CRC_COLOR21_31_GET_21(r)		SCALER_GET(r, 31, 16)
+#define SCALER_CRC_COLOR21_31_GET_31(r)		SCALER_GET(r, 15, 0)
+
+#endif /* EXYNOS_REGS_SCALER_H */
diff --git a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_rgb.c b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_rgb.c
index c54806d08dd7..2298ed2a9e1c 100644
--- a/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_rgb.c
+++ b/drivers/gpu/drm/fsl-dcu/fsl_dcu_drm_rgb.c
@@ -117,7 +117,7 @@ static int fsl_dcu_attach_panel(struct fsl_dcu_drm_device *fsl_dev,
 	if (ret < 0)
 		goto err_cleanup;
 
-	ret = drm_mode_connector_attach_encoder(connector, encoder);
+	ret = drm_connector_attach_encoder(connector, encoder);
 	if (ret < 0)
 		goto err_sysfs;
 
@@ -148,8 +148,9 @@ int fsl_dcu_create_outputs(struct fsl_dcu_drm_device *fsl_dev)
 	if (panel_node) {
 		fsl_dev->connector.panel = of_drm_find_panel(panel_node);
 		of_node_put(panel_node);
-		if (!fsl_dev->connector.panel)
-			return -EPROBE_DEFER;
+		if (IS_ERR(fsl_dev->connector.panel))
+			return PTR_ERR(fsl_dev->connector.panel);
+
 		return fsl_dcu_attach_panel(fsl_dev, fsl_dev->connector.panel);
 	}
 
diff --git a/drivers/gpu/drm/gma500/accel_2d.c b/drivers/gpu/drm/gma500/accel_2d.c
index c51d9259c7a7..204c8e452eb7 100644
--- a/drivers/gpu/drm/gma500/accel_2d.c
+++ b/drivers/gpu/drm/gma500/accel_2d.c
@@ -251,7 +251,7 @@ static void psbfb_copyarea_accel(struct fb_info *info,
 	if (!fb)
 		return;
 
-	offset = psbfb->gtt->offset;
+	offset = to_gtt_range(fb->obj[0])->offset;
 	stride = fb->pitches[0];
 
 	switch (fb->format->depth) {
diff --git a/drivers/gpu/drm/gma500/cdv_device.c b/drivers/gpu/drm/gma500/cdv_device.c
index 3a3bf752e03a..34b85767e4da 100644
--- a/drivers/gpu/drm/gma500/cdv_device.c
+++ b/drivers/gpu/drm/gma500/cdv_device.c
@@ -485,7 +485,7 @@ void cdv_intel_attach_force_audio_property(struct drm_connector *connector)
 			return;
 
 		for (i = 0; i < ARRAY_SIZE(force_audio_names); i++)
-			drm_property_add_enum(prop, i, i-1, force_audio_names[i]);
+			drm_property_add_enum(prop, i-1, force_audio_names[i]);
 
 		dev_priv->force_audio_property = prop;
 	}
@@ -514,7 +514,7 @@ void cdv_intel_attach_broadcast_rgb_property(struct drm_connector *connector)
 			return;
 
 		for (i = 0; i < ARRAY_SIZE(broadcast_rgb_names); i++)
-			drm_property_add_enum(prop, i, i, broadcast_rgb_names[i]);
+			drm_property_add_enum(prop, i, broadcast_rgb_names[i]);
 
 		dev_priv->broadcast_rgb_property = prop;
 	}
diff --git a/drivers/gpu/drm/gma500/cdv_intel_dp.c b/drivers/gpu/drm/gma500/cdv_intel_dp.c
index 5ea785f07ba8..90ed20083009 100644
--- a/drivers/gpu/drm/gma500/cdv_intel_dp.c
+++ b/drivers/gpu/drm/gma500/cdv_intel_dp.c
@@ -1770,7 +1770,7 @@ static int cdv_intel_dp_get_modes(struct drm_connector *connector)
 
 	edid = drm_get_edid(connector, &intel_dp->adapter);
 	if (edid) {
-		drm_mode_connector_update_edid_property(connector, edid);
+		drm_connector_update_edid_property(connector, edid);
 		ret = drm_add_edid_modes(connector, edid);
 		kfree(edid);
 	}
diff --git a/drivers/gpu/drm/gma500/cdv_intel_hdmi.c b/drivers/gpu/drm/gma500/cdv_intel_hdmi.c
index f0878998526a..4e4e4a66eaee 100644
--- a/drivers/gpu/drm/gma500/cdv_intel_hdmi.c
+++ b/drivers/gpu/drm/gma500/cdv_intel_hdmi.c
@@ -216,7 +216,7 @@ static int cdv_hdmi_get_modes(struct drm_connector *connector)
 
 	edid = drm_get_edid(connector, &gma_encoder->i2c_bus->adapter);
 	if (edid) {
-		drm_mode_connector_update_edid_property(connector, edid);
+		drm_connector_update_edid_property(connector, edid);
 		ret = drm_add_edid_modes(connector, edid);
 		kfree(edid);
 	}
diff --git a/drivers/gpu/drm/gma500/framebuffer.c b/drivers/gpu/drm/gma500/framebuffer.c
index cb0a2ae916e0..2f00a37684a2 100644
--- a/drivers/gpu/drm/gma500/framebuffer.c
+++ b/drivers/gpu/drm/gma500/framebuffer.c
@@ -33,6 +33,7 @@
 #include <drm/drm.h>
 #include <drm/drm_crtc.h>
 #include <drm/drm_fb_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 
 #include "psb_drv.h"
 #include "psb_intel_reg.h"
@@ -40,14 +41,9 @@
 #include "framebuffer.h"
 #include "gtt.h"
 
-static void psb_user_framebuffer_destroy(struct drm_framebuffer *fb);
-static int psb_user_framebuffer_create_handle(struct drm_framebuffer *fb,
-					      struct drm_file *file_priv,
-					      unsigned int *handle);
-
 static const struct drm_framebuffer_funcs psb_fb_funcs = {
-	.destroy = psb_user_framebuffer_destroy,
-	.create_handle = psb_user_framebuffer_create_handle,
+	.destroy = drm_gem_fb_destroy,
+	.create_handle = drm_gem_fb_create_handle,
 };
 
 #define CMAP_TOHW(_val, _width) ((((_val) << (_width)) + 0x7FFF - (_val)) >> 16)
@@ -96,17 +92,18 @@ static int psbfb_pan(struct fb_var_screeninfo *var, struct fb_info *info)
 	struct psb_fbdev *fbdev = info->par;
 	struct psb_framebuffer *psbfb = &fbdev->pfb;
 	struct drm_device *dev = psbfb->base.dev;
+	struct gtt_range *gtt = to_gtt_range(psbfb->base.obj[0]);
 
 	/*
 	 *	We have to poke our nose in here. The core fb code assumes
 	 *	panning is part of the hardware that can be invoked before
 	 *	the actual fb is mapped. In our case that isn't quite true.
 	 */
-	if (psbfb->gtt->npage) {
+	if (gtt->npage) {
 		/* GTT roll shifts in 4K pages, we need to shift the right
 		   number of pages */
 		int pages = info->fix.line_length >> 12;
-		psb_gtt_roll(dev, psbfb->gtt, var->yoffset * pages);
+		psb_gtt_roll(dev, gtt, var->yoffset * pages);
 	}
         return 0;
 }
@@ -117,13 +114,14 @@ static int psbfb_vm_fault(struct vm_fault *vmf)
 	struct psb_framebuffer *psbfb = vma->vm_private_data;
 	struct drm_device *dev = psbfb->base.dev;
 	struct drm_psb_private *dev_priv = dev->dev_private;
+	struct gtt_range *gtt = to_gtt_range(psbfb->base.obj[0]);
 	int page_num;
 	int i;
 	unsigned long address;
 	int ret;
 	unsigned long pfn;
 	unsigned long phys_addr = (unsigned long)dev_priv->stolen_base +
-				  psbfb->gtt->offset;
+				  gtt->offset;
 
 	page_num = vma_pages(vma);
 	address = vmf->address - (vmf->pgoff << PAGE_SHIFT);
@@ -246,7 +244,7 @@ static int psb_framebuffer_init(struct drm_device *dev,
 		return -EINVAL;
 
 	drm_helper_mode_fill_fb_struct(dev, &fb->base, mode_cmd);
-	fb->gtt = gt;
+	fb->base.obj[0] = &gt->gem;
 	ret = drm_framebuffer_init(dev, &fb->base, &psb_fb_funcs);
 	if (ret) {
 		dev_err(dev->dev, "framebuffer init failed: %d\n", ret);
@@ -518,8 +516,8 @@ static int psb_fbdev_destroy(struct drm_device *dev, struct psb_fbdev *fbdev)
 	drm_framebuffer_unregister_private(&psbfb->base);
 	drm_framebuffer_cleanup(&psbfb->base);
 
-	if (psbfb->gtt)
-		drm_gem_object_unreference_unlocked(&psbfb->gtt->gem);
+	if (psbfb->base.obj[0])
+		drm_gem_object_put_unlocked(psbfb->base.obj[0]);
 	return 0;
 }
 
@@ -576,44 +574,6 @@ static void psb_fbdev_fini(struct drm_device *dev)
 	dev_priv->fbdev = NULL;
 }
 
-/**
- *	psb_user_framebuffer_create_handle - add hamdle to a framebuffer
- *	@fb: framebuffer
- *	@file_priv: our DRM file
- *	@handle: returned handle
- *
- *	Our framebuffer object is a GTT range which also contains a GEM
- *	object. We need to turn it into a handle for userspace. GEM will do
- *	the work for us
- */
-static int psb_user_framebuffer_create_handle(struct drm_framebuffer *fb,
-					      struct drm_file *file_priv,
-					      unsigned int *handle)
-{
-	struct psb_framebuffer *psbfb = to_psb_fb(fb);
-	struct gtt_range *r = psbfb->gtt;
-	return drm_gem_handle_create(file_priv, &r->gem, handle);
-}
-
-/**
- *	psb_user_framebuffer_destroy	-	destruct user created fb
- *	@fb: framebuffer
- *
- *	User framebuffers are backed by GEM objects so all we have to do is
- *	clean up a bit and drop the reference, GEM will handle the fallout
- */
-static void psb_user_framebuffer_destroy(struct drm_framebuffer *fb)
-{
-	struct psb_framebuffer *psbfb = to_psb_fb(fb);
-	struct gtt_range *r = psbfb->gtt;
-
-	/* Let DRM do its clean up */
-	drm_framebuffer_cleanup(fb);
-	/*  We are no longer using the resource in GEM */
-	drm_gem_object_unreference_unlocked(&r->gem);
-	kfree(fb);
-}
-
 static const struct drm_mode_config_funcs psb_mode_funcs = {
 	.fb_create = psb_user_framebuffer_create,
 	.output_poll_changed = drm_fb_helper_output_poll_changed,
diff --git a/drivers/gpu/drm/gma500/framebuffer.h b/drivers/gpu/drm/gma500/framebuffer.h
index 395f20b07aab..23dc3c5f8f0d 100644
--- a/drivers/gpu/drm/gma500/framebuffer.h
+++ b/drivers/gpu/drm/gma500/framebuffer.h
@@ -31,7 +31,6 @@ struct psb_framebuffer {
 	struct drm_framebuffer base;
 	struct address_space *addr_space;
 	struct fb_info *fbdev;
-	struct gtt_range *gtt;
 };
 
 struct psb_fbdev {
diff --git a/drivers/gpu/drm/gma500/gem.c b/drivers/gpu/drm/gma500/gem.c
index 131239759a75..913bf4c256fa 100644
--- a/drivers/gpu/drm/gma500/gem.c
+++ b/drivers/gpu/drm/gma500/gem.c
@@ -93,7 +93,7 @@ int psb_gem_create(struct drm_file *file, struct drm_device *dev, u64 size,
 		return ret;
 	}
 	/* We have the initial and handle reference but need only one now */
-	drm_gem_object_unreference_unlocked(&r->gem);
+	drm_gem_object_put_unlocked(&r->gem);
 	*handlep = handle;
 	return 0;
 }
diff --git a/drivers/gpu/drm/gma500/gma_display.c b/drivers/gpu/drm/gma500/gma_display.c
index f3c48a2be71b..09c1161a7ac6 100644
--- a/drivers/gpu/drm/gma500/gma_display.c
+++ b/drivers/gpu/drm/gma500/gma_display.c
@@ -60,7 +60,7 @@ int gma_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 	struct drm_psb_private *dev_priv = dev->dev_private;
 	struct gma_crtc *gma_crtc = to_gma_crtc(crtc);
 	struct drm_framebuffer *fb = crtc->primary->fb;
-	struct psb_framebuffer *psbfb = to_psb_fb(fb);
+	struct gtt_range *gtt;
 	int pipe = gma_crtc->pipe;
 	const struct psb_offset *map = &dev_priv->regmap[pipe];
 	unsigned long start, offset;
@@ -76,12 +76,14 @@ int gma_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 		goto gma_pipe_cleaner;
 	}
 
+	gtt = to_gtt_range(fb->obj[0]);
+
 	/* We are displaying this buffer, make sure it is actually loaded
 	   into the GTT */
-	ret = psb_gtt_pin(psbfb->gtt);
+	ret = psb_gtt_pin(gtt);
 	if (ret < 0)
 		goto gma_pipe_set_base_exit;
-	start = psbfb->gtt->offset;
+	start = gtt->offset;
 	offset = y * fb->pitches[0] + x * fb->format->cpp[0];
 
 	REG_WRITE(map->stride, fb->pitches[0]);
@@ -129,7 +131,7 @@ int gma_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 gma_pipe_cleaner:
 	/* If there was a previous display we can now unpin it */
 	if (old_fb)
-		psb_gtt_unpin(to_psb_fb(old_fb)->gtt);
+		psb_gtt_unpin(to_gtt_range(old_fb->obj[0]));
 
 gma_pipe_set_base_exit:
 	gma_power_end(dev);
@@ -353,7 +355,7 @@ int gma_crtc_cursor_set(struct drm_crtc *crtc,
 			gt = container_of(gma_crtc->cursor_obj,
 					  struct gtt_range, gem);
 			psb_gtt_unpin(gt);
-			drm_gem_object_unreference_unlocked(gma_crtc->cursor_obj);
+			drm_gem_object_put_unlocked(gma_crtc->cursor_obj);
 			gma_crtc->cursor_obj = NULL;
 		}
 		return 0;
@@ -429,7 +431,7 @@ int gma_crtc_cursor_set(struct drm_crtc *crtc,
 	if (gma_crtc->cursor_obj) {
 		gt = container_of(gma_crtc->cursor_obj, struct gtt_range, gem);
 		psb_gtt_unpin(gt);
-		drm_gem_object_unreference_unlocked(gma_crtc->cursor_obj);
+		drm_gem_object_put_unlocked(gma_crtc->cursor_obj);
 	}
 
 	gma_crtc->cursor_obj = obj;
@@ -437,7 +439,7 @@ unlock:
 	return ret;
 
 unref_cursor:
-	drm_gem_object_unreference_unlocked(obj);
+	drm_gem_object_put_unlocked(obj);
 	return ret;
 }
 
@@ -491,7 +493,7 @@ void gma_crtc_disable(struct drm_crtc *crtc)
 	crtc_funcs->dpms(crtc, DRM_MODE_DPMS_OFF);
 
 	if (crtc->primary->fb) {
-		gt = to_psb_fb(crtc->primary->fb)->gtt;
+		gt = to_gtt_range(crtc->primary->fb->obj[0]);
 		psb_gtt_unpin(gt);
 	}
 }
@@ -663,7 +665,7 @@ void gma_connector_attach_encoder(struct gma_connector *connector,
 				  struct gma_encoder *encoder)
 {
 	connector->encoder = encoder;
-	drm_mode_connector_attach_encoder(&connector->base,
+	drm_connector_attach_encoder(&connector->base,
 					  &encoder->base);
 }
 
diff --git a/drivers/gpu/drm/gma500/gtt.h b/drivers/gpu/drm/gma500/gtt.h
index cdbb350c9d5d..cb0c3a2a1fd4 100644
--- a/drivers/gpu/drm/gma500/gtt.h
+++ b/drivers/gpu/drm/gma500/gtt.h
@@ -53,6 +53,8 @@ struct gtt_range {
 	int roll;			/* Roll applied to the GTT entries */
 };
 
+#define to_gtt_range(x) container_of(x, struct gtt_range, gem)
+
 extern struct gtt_range *psb_gtt_alloc_range(struct drm_device *dev, int len,
 					     const char *name, int backed,
 					     u32 align);
diff --git a/drivers/gpu/drm/gma500/intel_bios.h b/drivers/gpu/drm/gma500/intel_bios.h
index 978ae4b25e82..e0ccf1d19a4d 100644
--- a/drivers/gpu/drm/gma500/intel_bios.h
+++ b/drivers/gpu/drm/gma500/intel_bios.h
@@ -34,7 +34,7 @@ struct vbt_header {
 	u8 reserved0;
 	u32 bdb_offset;			/**< from beginning of VBT */
 	u32 aim_offset[4];		/**< from beginning of VBT */
-} __attribute__((packed));
+} __packed;
 
 
 struct bdb_header {
@@ -61,7 +61,7 @@ struct vbios_data {
 	u8 rsvd4; /* popup memory size */
 	u8 resize_pci_bios;
 	u8 rsvd5; /* is crt already on ddc2 */
-} __attribute__((packed));
+} __packed;
 
 /*
  * There are several types of BIOS data blocks (BDBs), each block has
@@ -133,7 +133,7 @@ struct bdb_general_features {
 	u8 dp_ssc_enb:1;	/* PCH attached eDP supports SSC */
 	u8 dp_ssc_freq:1;	/* SSC freq for PCH attached eDP */
 	u8 rsvd11:3; /* finish byte */
-} __attribute__((packed));
+} __packed;
 
 /* pre-915 */
 #define GPIO_PIN_DVI_LVDS	0x03 /* "DVI/LVDS DDC GPIO pins" */
@@ -213,7 +213,7 @@ struct child_device_config {
 	u8  dvo2_wiring;
 	u16 extended_type;
 	u8  dvo_function;
-} __attribute__((packed));
+} __packed;
 
 
 struct bdb_general_definitions {
@@ -256,7 +256,7 @@ struct bdb_lvds_options {
 	u8 lvds_edid:1;
 	u8 rsvd2:1;
 	u8 rsvd4;
-} __attribute__((packed));
+} __packed;
 
 struct bdb_lvds_backlight {
 	u8 type:2;
@@ -268,7 +268,7 @@ struct bdb_lvds_backlight {
 	u8 i2caddr;
 	u8 brightnesscmd;
 	/*FIXME: more...*/
-} __attribute__((packed));
+} __packed;
 
 /* LFP pointer table contains entries to the struct below */
 struct bdb_lvds_lfp_data_ptr {
@@ -278,12 +278,12 @@ struct bdb_lvds_lfp_data_ptr {
 	u8 dvo_table_size;
 	u16 panel_pnp_id_offset;
 	u8 pnp_table_size;
-} __attribute__((packed));
+} __packed;
 
 struct bdb_lvds_lfp_data_ptrs {
 	u8 lvds_entries; /* followed by one or more lvds_data_ptr structs */
 	struct bdb_lvds_lfp_data_ptr ptr[16];
-} __attribute__((packed));
+} __packed;
 
 /* LFP data has 3 blocks per entry */
 struct lvds_fp_timing {
@@ -300,7 +300,7 @@ struct lvds_fp_timing {
 	u32 pfit_reg;
 	u32 pfit_reg_val;
 	u16 terminator;
-} __attribute__((packed));
+} __packed;
 
 struct lvds_dvo_timing {
 	u16 clock;		/**< In 10khz */
@@ -328,7 +328,7 @@ struct lvds_dvo_timing {
 	u8 vsync_positive:1;
 	u8 hsync_positive:1;
 	u8 rsvd2:1;
-} __attribute__((packed));
+} __packed;
 
 struct lvds_pnp_id {
 	u16 mfg_name;
@@ -336,17 +336,17 @@ struct lvds_pnp_id {
 	u32 serial;
 	u8 mfg_week;
 	u8 mfg_year;
-} __attribute__((packed));
+} __packed;
 
 struct bdb_lvds_lfp_data_entry {
 	struct lvds_fp_timing fp_timing;
 	struct lvds_dvo_timing dvo_timing;
 	struct lvds_pnp_id pnp_id;
-} __attribute__((packed));
+} __packed;
 
 struct bdb_lvds_lfp_data {
 	struct bdb_lvds_lfp_data_entry data[16];
-} __attribute__((packed));
+} __packed;
 
 struct aimdb_header {
 	char signature[16];
@@ -354,12 +354,12 @@ struct aimdb_header {
 	u16 aimdb_version;
 	u16 aimdb_header_size;
 	u16 aimdb_size;
-} __attribute__((packed));
+} __packed;
 
 struct aimdb_block {
 	u8 aimdb_id;
 	u16 aimdb_size;
-} __attribute__((packed));
+} __packed;
 
 struct vch_panel_data {
 	u16 fp_timing_offset;
@@ -370,12 +370,12 @@ struct vch_panel_data {
 	u8 text_fitting_size;
 	u16 graphics_fitting_offset;
 	u8 graphics_fitting_size;
-} __attribute__((packed));
+} __packed;
 
 struct vch_bdb_22 {
 	struct aimdb_block aimdb_block;
 	struct vch_panel_data panels[16];
-} __attribute__((packed));
+} __packed;
 
 struct bdb_sdvo_lvds_options {
 	u8 panel_backlight;
@@ -391,7 +391,7 @@ struct bdb_sdvo_lvds_options {
 	u8 panel_misc_bits_2;
 	u8 panel_misc_bits_3;
 	u8 panel_misc_bits_4;
-} __attribute__((packed));
+} __packed;
 
 #define BDB_DRIVER_FEATURE_NO_LVDS		0
 #define BDB_DRIVER_FEATURE_INT_LVDS		1
@@ -436,7 +436,7 @@ struct bdb_driver_features {
 
 	u8 hdmi_termination;
 	u8 custom_vbt_version;
-} __attribute__((packed));
+} __packed;
 
 #define EDP_18BPP	0
 #define EDP_24BPP	1
diff --git a/drivers/gpu/drm/gma500/mdfld_dsi_dpi.c b/drivers/gpu/drm/gma500/mdfld_dsi_dpi.c
index a05c020602bd..d0bf5a1e94e8 100644
--- a/drivers/gpu/drm/gma500/mdfld_dsi_dpi.c
+++ b/drivers/gpu/drm/gma500/mdfld_dsi_dpi.c
@@ -999,7 +999,7 @@ struct mdfld_dsi_encoder *mdfld_dsi_dpi_init(struct drm_device *dev,
 				p_funcs->encoder_helper_funcs);
 
 	/*attach to given connector*/
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	/*set possible crtcs and clones*/
 	if (dsi_connector->pipe) {
diff --git a/drivers/gpu/drm/gma500/mdfld_intel_display.c b/drivers/gpu/drm/gma500/mdfld_intel_display.c
index 5c066448be5b..2b9fa0163dea 100644
--- a/drivers/gpu/drm/gma500/mdfld_intel_display.c
+++ b/drivers/gpu/drm/gma500/mdfld_intel_display.c
@@ -167,7 +167,6 @@ static int mdfld__intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 	struct drm_psb_private *dev_priv = dev->dev_private;
 	struct drm_framebuffer *fb = crtc->primary->fb;
 	struct gma_crtc *gma_crtc = to_gma_crtc(crtc);
-	struct psb_framebuffer *psbfb = to_psb_fb(fb);
 	int pipe = gma_crtc->pipe;
 	const struct psb_offset *map = &dev_priv->regmap[pipe];
 	unsigned long start, offset;
@@ -196,7 +195,7 @@ static int mdfld__intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 	if (!gma_power_begin(dev, true))
 		return 0;
 
-	start = psbfb->gtt->offset;
+	start = to_gtt_range(fb->obj[0])->offset;
 	offset = y * fb->pitches[0] + x * fb->format->cpp[0];
 
 	REG_WRITE(map->stride, fb->pitches[0]);
diff --git a/drivers/gpu/drm/gma500/mid_bios.c b/drivers/gpu/drm/gma500/mid_bios.c
index 7171b7475f58..237041a37532 100644
--- a/drivers/gpu/drm/gma500/mid_bios.c
+++ b/drivers/gpu/drm/gma500/mid_bios.c
@@ -239,7 +239,7 @@ static int mid_get_vbt_data_r10(struct drm_psb_private *dev_priv, u32 addr)
 	if (read_vbt_r10(addr, &vbt))
 		return -1;
 
-	gct = kmalloc(sizeof(*gct) * vbt.panel_count, GFP_KERNEL);
+	gct = kmalloc_array(vbt.panel_count, sizeof(*gct), GFP_KERNEL);
 	if (!gct)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/gma500/oaktrail_crtc.c b/drivers/gpu/drm/gma500/oaktrail_crtc.c
index 0fff269d3fe6..1b7fd6a9d8a5 100644
--- a/drivers/gpu/drm/gma500/oaktrail_crtc.c
+++ b/drivers/gpu/drm/gma500/oaktrail_crtc.c
@@ -600,7 +600,6 @@ static int oaktrail_pipe_set_base(struct drm_crtc *crtc,
 	struct drm_psb_private *dev_priv = dev->dev_private;
 	struct gma_crtc *gma_crtc = to_gma_crtc(crtc);
 	struct drm_framebuffer *fb = crtc->primary->fb;
-	struct psb_framebuffer *psbfb = to_psb_fb(fb);
 	int pipe = gma_crtc->pipe;
 	const struct psb_offset *map = &dev_priv->regmap[pipe];
 	unsigned long start, offset;
@@ -617,7 +616,7 @@ static int oaktrail_pipe_set_base(struct drm_crtc *crtc,
 	if (!gma_power_begin(dev, true))
 		return 0;
 
-	start = psbfb->gtt->offset;
+	start = to_gtt_range(fb->obj[0])->offset;
 	offset = y * fb->pitches[0] + x * fb->format->cpp[0];
 
 	REG_WRITE(map->stride, fb->pitches[0]);
diff --git a/drivers/gpu/drm/gma500/oaktrail_hdmi.c b/drivers/gpu/drm/gma500/oaktrail_hdmi.c
index 78566a80ad25..c6d72de1c054 100644
--- a/drivers/gpu/drm/gma500/oaktrail_hdmi.c
+++ b/drivers/gpu/drm/gma500/oaktrail_hdmi.c
@@ -578,7 +578,7 @@ static int oaktrail_hdmi_get_modes(struct drm_connector *connector)
 	}
 
 	if (edid) {
-		drm_mode_connector_update_edid_property(connector, edid);
+		drm_connector_update_edid_property(connector, edid);
 		ret = drm_add_edid_modes(connector, edid);
 	}
 	return ret;
diff --git a/drivers/gpu/drm/gma500/oaktrail_lvds.c b/drivers/gpu/drm/gma500/oaktrail_lvds.c
index e6943fef0611..83babb815a5d 100644
--- a/drivers/gpu/drm/gma500/oaktrail_lvds.c
+++ b/drivers/gpu/drm/gma500/oaktrail_lvds.c
@@ -376,7 +376,7 @@ void oaktrail_lvds_init(struct drm_device *dev,
 	 * preferred mode is the right one.
 	 */
 	if (edid) {
-		drm_mode_connector_update_edid_property(connector, edid);
+		drm_connector_update_edid_property(connector, edid);
 		drm_add_edid_modes(connector, edid);
 		kfree(edid);
 
diff --git a/drivers/gpu/drm/gma500/psb_intel_modes.c b/drivers/gpu/drm/gma500/psb_intel_modes.c
index e5360726d80b..fb4da3cd6681 100644
--- a/drivers/gpu/drm/gma500/psb_intel_modes.c
+++ b/drivers/gpu/drm/gma500/psb_intel_modes.c
@@ -66,7 +66,7 @@ int psb_intel_ddc_get_modes(struct drm_connector *connector,
 
 	edid = drm_get_edid(connector, adapter);
 	if (edid) {
-		drm_mode_connector_update_edid_property(connector, edid);
+		drm_connector_update_edid_property(connector, edid);
 		ret = drm_add_edid_modes(connector, edid);
 		kfree(edid);
 	}
diff --git a/drivers/gpu/drm/gma500/psb_intel_sdvo.c b/drivers/gpu/drm/gma500/psb_intel_sdvo.c
index 8dc2b19f913b..dd3cec0e3190 100644
--- a/drivers/gpu/drm/gma500/psb_intel_sdvo.c
+++ b/drivers/gpu/drm/gma500/psb_intel_sdvo.c
@@ -429,13 +429,20 @@ static const char *cmd_status_names[] = {
 	"Scaling not supported"
 };
 
+#define MAX_ARG_LEN 32
+
 static bool psb_intel_sdvo_write_cmd(struct psb_intel_sdvo *psb_intel_sdvo, u8 cmd,
 				 const void *args, int args_len)
 {
-	u8 buf[args_len*2 + 2], status;
-	struct i2c_msg msgs[args_len + 3];
+	u8 buf[MAX_ARG_LEN*2 + 2], status;
+	struct i2c_msg msgs[MAX_ARG_LEN + 3];
 	int i, ret;
 
+	if (args_len > MAX_ARG_LEN) {
+		DRM_ERROR("Need to increase arg length\n");
+		return false;
+	}
+
 	psb_intel_sdvo_debug_write(psb_intel_sdvo, cmd, args, args_len);
 
 	for (i = 0; i < args_len; i++) {
@@ -1465,7 +1472,7 @@ static void psb_intel_sdvo_get_ddc_modes(struct drm_connector *connector)
 		bool connector_is_digital = !!IS_TMDS(psb_intel_sdvo_connector);
 
 		if (connector_is_digital == monitor_is_digital) {
-			drm_mode_connector_update_edid_property(connector, edid);
+			drm_connector_update_edid_property(connector, edid);
 			drm_add_edid_modes(connector, edid);
 		}
 
@@ -2281,7 +2288,7 @@ static bool psb_intel_sdvo_tv_create_property(struct psb_intel_sdvo *psb_intel_s
 
 	for (i = 0; i < psb_intel_sdvo_connector->format_supported_num; i++)
 		drm_property_add_enum(
-				psb_intel_sdvo_connector->tv_format, i,
+				psb_intel_sdvo_connector->tv_format,
 				i, tv_format_names[psb_intel_sdvo_connector->tv_format_supported[i]]);
 
 	psb_intel_sdvo->tv_format_index = psb_intel_sdvo_connector->tv_format_supported[0];
diff --git a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_vdac.c b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_vdac.c
index d2f4749ebf8d..744956cea749 100644
--- a/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_vdac.c
+++ b/drivers/gpu/drm/hisilicon/hibmc/hibmc_drm_vdac.c
@@ -133,7 +133,7 @@ int hibmc_vdac_init(struct hibmc_drm_private *priv)
 	}
 
 	drm_encoder_helper_add(encoder, &hibmc_encoder_helper_funcs);
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c
index 2269be91f3e1..bb774202a5a1 100644
--- a/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c
+++ b/drivers/gpu/drm/hisilicon/kirin/kirin_drm_ade.c
@@ -859,7 +859,6 @@ static int ade_plane_atomic_check(struct drm_plane *plane,
 		return PTR_ERR(crtc_state);
 
 	if (src_w != crtc_w || src_h != crtc_h) {
-		DRM_ERROR("Scale not support!!!\n");
 		return -EINVAL;
 	}
 
diff --git a/drivers/gpu/drm/i2c/Kconfig b/drivers/gpu/drm/i2c/Kconfig
index a6c92beb410a..65d3acb61c03 100644
--- a/drivers/gpu/drm/i2c/Kconfig
+++ b/drivers/gpu/drm/i2c/Kconfig
@@ -22,8 +22,14 @@ config DRM_I2C_SIL164
 config DRM_I2C_NXP_TDA998X
 	tristate "NXP Semiconductors TDA998X HDMI encoder"
 	default m if DRM_TILCDC
+	select CEC_CORE if CEC_NOTIFIER
 	select SND_SOC_HDMI_CODEC if SND_SOC
 	help
 	  Support for NXP Semiconductors TDA998X HDMI encoders.
 
+config DRM_I2C_NXP_TDA9950
+	tristate "NXP Semiconductors TDA9950/TDA998X HDMI CEC"
+	select CEC_NOTIFIER
+	select CEC_CORE
+
 endmenu
diff --git a/drivers/gpu/drm/i2c/Makefile b/drivers/gpu/drm/i2c/Makefile
index b20100c18ffb..a962f6f08568 100644
--- a/drivers/gpu/drm/i2c/Makefile
+++ b/drivers/gpu/drm/i2c/Makefile
@@ -7,3 +7,4 @@ obj-$(CONFIG_DRM_I2C_SIL164) += sil164.o
 
 tda998x-y := tda998x_drv.o
 obj-$(CONFIG_DRM_I2C_NXP_TDA998X) += tda998x.o
+obj-$(CONFIG_DRM_I2C_NXP_TDA9950) += tda9950.o
diff --git a/drivers/gpu/drm/i2c/tda9950.c b/drivers/gpu/drm/i2c/tda9950.c
new file mode 100644
index 000000000000..3f7396caad48
--- /dev/null
+++ b/drivers/gpu/drm/i2c/tda9950.c
@@ -0,0 +1,509 @@
+/*
+ *  TDA9950 Consumer Electronics Control driver
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * The NXP TDA9950 implements the HDMI Consumer Electronics Control
+ * interface.  The host interface is similar to a mailbox: the data
+ * registers starting at REG_CDR0 are written to send a command to the
+ * internal CPU, and replies are read from these registers.
+ *
+ * As the data registers represent a mailbox, they must be accessed
+ * as a single I2C transaction.  See the TDA9950 data sheet for details.
+ */
+#include <linux/delay.h>
+#include <linux/i2c.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/platform_data/tda9950.h>
+#include <linux/slab.h>
+#include <drm/drm_edid.h>
+#include <media/cec.h>
+#include <media/cec-notifier.h>
+
+enum {
+	REG_CSR = 0x00,
+	CSR_BUSY = BIT(7),
+	CSR_INT  = BIT(6),
+	CSR_ERR  = BIT(5),
+
+	REG_CER = 0x01,
+
+	REG_CVR = 0x02,
+
+	REG_CCR = 0x03,
+	CCR_RESET = BIT(7),
+	CCR_ON    = BIT(6),
+
+	REG_ACKH = 0x04,
+	REG_ACKL = 0x05,
+
+	REG_CCONR = 0x06,
+	CCONR_ENABLE_ERROR = BIT(4),
+	CCONR_RETRY_MASK = 7,
+
+	REG_CDR0 = 0x07,
+
+	CDR1_REQ = 0x00,
+	CDR1_CNF = 0x01,
+	CDR1_IND = 0x81,
+	CDR1_ERR = 0x82,
+	CDR1_IER = 0x83,
+
+	CDR2_CNF_SUCCESS    = 0x00,
+	CDR2_CNF_OFF_STATE  = 0x80,
+	CDR2_CNF_BAD_REQ    = 0x81,
+	CDR2_CNF_CEC_ACCESS = 0x82,
+	CDR2_CNF_ARB_ERROR  = 0x83,
+	CDR2_CNF_BAD_TIMING = 0x84,
+	CDR2_CNF_NACK_ADDR  = 0x85,
+	CDR2_CNF_NACK_DATA  = 0x86,
+};
+
+struct tda9950_priv {
+	struct i2c_client *client;
+	struct device *hdmi;
+	struct cec_adapter *adap;
+	struct tda9950_glue *glue;
+	u16 addresses;
+	struct cec_msg rx_msg;
+	struct cec_notifier *notify;
+	bool open;
+};
+
+static int tda9950_write_range(struct i2c_client *client, u8 addr, u8 *p, int cnt)
+{
+	struct i2c_msg msg;
+	u8 buf[cnt + 1];
+	int ret;
+
+	buf[0] = addr;
+	memcpy(buf + 1, p, cnt);
+
+	msg.addr = client->addr;
+	msg.flags = 0;
+	msg.len = cnt + 1;
+	msg.buf = buf;
+
+	dev_dbg(&client->dev, "wr 0x%02x: %*ph\n", addr, cnt, p);
+
+	ret = i2c_transfer(client->adapter, &msg, 1);
+	if (ret < 0)
+		dev_err(&client->dev, "Error %d writing to cec:0x%x\n", ret, addr);
+	return ret < 0 ? ret : 0;
+}
+
+static void tda9950_write(struct i2c_client *client, u8 addr, u8 val)
+{
+	tda9950_write_range(client, addr, &val, 1);
+}
+
+static int tda9950_read_range(struct i2c_client *client, u8 addr, u8 *p, int cnt)
+{
+	struct i2c_msg msg[2];
+	int ret;
+
+	msg[0].addr = client->addr;
+	msg[0].flags = 0;
+	msg[0].len = 1;
+	msg[0].buf = &addr;
+	msg[1].addr = client->addr;
+	msg[1].flags = I2C_M_RD;
+	msg[1].len = cnt;
+	msg[1].buf = p;
+
+	ret = i2c_transfer(client->adapter, msg, 2);
+	if (ret < 0)
+		dev_err(&client->dev, "Error %d reading from cec:0x%x\n", ret, addr);
+
+	dev_dbg(&client->dev, "rd 0x%02x: %*ph\n", addr, cnt, p);
+
+	return ret;
+}
+
+static u8 tda9950_read(struct i2c_client *client, u8 addr)
+{
+	int ret;
+	u8 val;
+
+	ret = tda9950_read_range(client, addr, &val, 1);
+	if (ret < 0)
+		val = 0;
+
+	return val;
+}
+
+static irqreturn_t tda9950_irq(int irq, void *data)
+{
+	struct tda9950_priv *priv = data;
+	unsigned int tx_status;
+	u8 csr, cconr, buf[19];
+	u8 arb_lost_cnt, nack_cnt, err_cnt;
+
+	if (!priv->open)
+		return IRQ_NONE;
+
+	csr = tda9950_read(priv->client, REG_CSR);
+	if (!(csr & CSR_INT))
+		return IRQ_NONE;
+
+	cconr = tda9950_read(priv->client, REG_CCONR) & CCONR_RETRY_MASK;
+
+	tda9950_read_range(priv->client, REG_CDR0, buf, sizeof(buf));
+
+	/*
+	 * This should never happen: the data sheet says that there will
+	 * always be a valid message if the interrupt line is asserted.
+	 */
+	if (buf[0] == 0) {
+		dev_warn(&priv->client->dev, "interrupt pending, but no message?\n");
+		return IRQ_NONE;
+	}
+
+	switch (buf[1]) {
+	case CDR1_CNF: /* transmit result */
+		arb_lost_cnt = nack_cnt = err_cnt = 0;
+		switch (buf[2]) {
+		case CDR2_CNF_SUCCESS:
+			tx_status = CEC_TX_STATUS_OK;
+			break;
+
+		case CDR2_CNF_ARB_ERROR:
+			tx_status = CEC_TX_STATUS_ARB_LOST;
+			arb_lost_cnt = cconr;
+			break;
+
+		case CDR2_CNF_NACK_ADDR:
+			tx_status = CEC_TX_STATUS_NACK;
+			nack_cnt = cconr;
+			break;
+
+		default: /* some other error, refer to TDA9950 docs */
+			dev_err(&priv->client->dev, "CNF reply error 0x%02x\n",
+				buf[2]);
+			tx_status = CEC_TX_STATUS_ERROR;
+			err_cnt = cconr;
+			break;
+		}
+		/* TDA9950 executes all retries for us */
+		tx_status |= CEC_TX_STATUS_MAX_RETRIES;
+		cec_transmit_done(priv->adap, tx_status, arb_lost_cnt,
+				  nack_cnt, 0, err_cnt);
+		break;
+
+	case CDR1_IND:
+		priv->rx_msg.len = buf[0] - 2;
+		if (priv->rx_msg.len > CEC_MAX_MSG_SIZE)
+			priv->rx_msg.len = CEC_MAX_MSG_SIZE;
+
+		memcpy(priv->rx_msg.msg, buf + 2, priv->rx_msg.len);
+		cec_received_msg(priv->adap, &priv->rx_msg);
+		break;
+
+	default: /* unknown */
+		dev_err(&priv->client->dev, "unknown service id 0x%02x\n",
+			buf[1]);
+		break;
+	}
+
+	return IRQ_HANDLED;
+}
+
+static int tda9950_cec_transmit(struct cec_adapter *adap, u8 attempts,
+				u32 signal_free_time, struct cec_msg *msg)
+{
+	struct tda9950_priv *priv = adap->priv;
+	u8 buf[CEC_MAX_MSG_SIZE + 2];
+
+	buf[0] = 2 + msg->len;
+	buf[1] = CDR1_REQ;
+	memcpy(buf + 2, msg->msg, msg->len);
+
+	if (attempts > 5)
+		attempts = 5;
+
+	tda9950_write(priv->client, REG_CCONR, attempts);
+
+	return tda9950_write_range(priv->client, REG_CDR0, buf, 2 + msg->len);
+}
+
+static int tda9950_cec_adap_log_addr(struct cec_adapter *adap, u8 addr)
+{
+	struct tda9950_priv *priv = adap->priv;
+	u16 addresses;
+	u8 buf[2];
+
+	if (addr == CEC_LOG_ADDR_INVALID)
+		addresses = priv->addresses = 0;
+	else
+		addresses = priv->addresses |= BIT(addr);
+
+	/* TDA9950 doesn't want address 15 set */
+	addresses &= 0x7fff;
+	buf[0] = addresses >> 8;
+	buf[1] = addresses;
+
+	return tda9950_write_range(priv->client, REG_ACKH, buf, 2);
+}
+
+/*
+ * When operating as part of the TDA998x, we need additional handling
+ * to initialise and shut down the TDA9950 part of the device.  These
+ * two hooks are provided to allow the TDA998x code to perform those
+ * activities.
+ */
+static int tda9950_glue_open(struct tda9950_priv *priv)
+{
+	int ret = 0;
+
+	if (priv->glue && priv->glue->open)
+		ret = priv->glue->open(priv->glue->data);
+
+	priv->open = true;
+
+	return ret;
+}
+
+static void tda9950_glue_release(struct tda9950_priv *priv)
+{
+	priv->open = false;
+
+	if (priv->glue && priv->glue->release)
+		priv->glue->release(priv->glue->data);
+}
+
+static int tda9950_open(struct tda9950_priv *priv)
+{
+	struct i2c_client *client = priv->client;
+	int ret;
+
+	ret = tda9950_glue_open(priv);
+	if (ret)
+		return ret;
+
+	/* Reset the TDA9950, and wait 250ms for it to recover */
+	tda9950_write(client, REG_CCR, CCR_RESET);
+	msleep(250);
+
+	tda9950_cec_adap_log_addr(priv->adap, CEC_LOG_ADDR_INVALID);
+
+	/* Start the command processor */
+	tda9950_write(client, REG_CCR, CCR_ON);
+
+	return 0;
+}
+
+static void tda9950_release(struct tda9950_priv *priv)
+{
+	struct i2c_client *client = priv->client;
+	int timeout = 50;
+	u8 csr;
+
+	/* Stop the command processor */
+	tda9950_write(client, REG_CCR, 0);
+
+	/* Wait up to .5s for it to signal non-busy */
+	do {
+		csr = tda9950_read(client, REG_CSR);
+		if (!(csr & CSR_BUSY) || --timeout)
+			break;
+		msleep(10);
+	} while (1);
+
+	/* Warn the user that their IRQ may die if it's shared. */
+	if (csr & CSR_BUSY)
+		dev_warn(&client->dev, "command processor failed to stop, irq%d may die (csr=0x%02x)\n",
+			 client->irq, csr);
+
+	tda9950_glue_release(priv);
+}
+
+static int tda9950_cec_adap_enable(struct cec_adapter *adap, bool enable)
+{
+	struct tda9950_priv *priv = adap->priv;
+
+	if (!enable) {
+		tda9950_release(priv);
+		return 0;
+	} else {
+		return tda9950_open(priv);
+	}
+}
+
+static const struct cec_adap_ops tda9950_cec_ops = {
+	.adap_enable = tda9950_cec_adap_enable,
+	.adap_log_addr = tda9950_cec_adap_log_addr,
+	.adap_transmit = tda9950_cec_transmit,
+};
+
+/*
+ * When operating as part of the TDA998x, we need to claim additional
+ * resources.  These two hooks permit the management of those resources.
+ */
+static void tda9950_devm_glue_exit(void *data)
+{
+	struct tda9950_glue *glue = data;
+
+	if (glue && glue->exit)
+		glue->exit(glue->data);
+}
+
+static int tda9950_devm_glue_init(struct device *dev, struct tda9950_glue *glue)
+{
+	int ret;
+
+	if (glue && glue->init) {
+		ret = glue->init(glue->data);
+		if (ret)
+			return ret;
+	}
+
+	ret = devm_add_action(dev, tda9950_devm_glue_exit, glue);
+	if (ret)
+		tda9950_devm_glue_exit(glue);
+
+	return ret;
+}
+
+static void tda9950_cec_del(void *data)
+{
+	struct tda9950_priv *priv = data;
+
+	cec_delete_adapter(priv->adap);
+}
+
+static int tda9950_probe(struct i2c_client *client,
+			 const struct i2c_device_id *id)
+{
+	struct tda9950_glue *glue = client->dev.platform_data;
+	struct device *dev = &client->dev;
+	struct tda9950_priv *priv;
+	unsigned long irqflags;
+	int ret;
+	u8 cvr;
+
+	/*
+	 * We must have I2C functionality: our multi-byte accesses
+	 * must be performed as a single contiguous transaction.
+	 */
+	if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) {
+		dev_err(&client->dev,
+			"adapter does not support I2C functionality\n");
+		return -ENXIO;
+	}
+
+	/* We must have an interrupt to be functional. */
+	if (client->irq <= 0) {
+		dev_err(&client->dev, "driver requires an interrupt\n");
+		return -ENXIO;
+	}
+
+	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	priv->client = client;
+	priv->glue = glue;
+
+	i2c_set_clientdata(client, priv);
+
+	/*
+	 * If we're part of a TDA998x, we want the class devices to be
+	 * associated with the HDMI Tx so we have a tight relationship
+	 * between the HDMI interface and the CEC interface.
+	 */
+	priv->hdmi = dev;
+	if (glue && glue->parent)
+		priv->hdmi = glue->parent;
+
+	priv->adap = cec_allocate_adapter(&tda9950_cec_ops, priv, "tda9950",
+					  CEC_CAP_DEFAULTS,
+					  CEC_MAX_LOG_ADDRS);
+	if (IS_ERR(priv->adap))
+		return PTR_ERR(priv->adap);
+
+	ret = devm_add_action(dev, tda9950_cec_del, priv);
+	if (ret) {
+		cec_delete_adapter(priv->adap);
+		return ret;
+	}
+
+	ret = tda9950_devm_glue_init(dev, glue);
+	if (ret)
+		return ret;
+
+	ret = tda9950_glue_open(priv);
+	if (ret)
+		return ret;
+
+	cvr = tda9950_read(client, REG_CVR);
+
+	dev_info(&client->dev,
+		 "TDA9950 CEC interface, hardware version %u.%u\n",
+		 cvr >> 4, cvr & 15);
+
+	tda9950_glue_release(priv);
+
+	irqflags = IRQF_TRIGGER_FALLING;
+	if (glue)
+		irqflags = glue->irq_flags;
+
+	ret = devm_request_threaded_irq(dev, client->irq, NULL, tda9950_irq,
+					irqflags | IRQF_SHARED | IRQF_ONESHOT,
+					dev_name(&client->dev), priv);
+	if (ret < 0)
+		return ret;
+
+	priv->notify = cec_notifier_get(priv->hdmi);
+	if (!priv->notify)
+		return -ENOMEM;
+
+	ret = cec_register_adapter(priv->adap, priv->hdmi);
+	if (ret < 0) {
+		cec_notifier_put(priv->notify);
+		return ret;
+	}
+
+	/*
+	 * CEC documentation says we must not call cec_delete_adapter
+	 * after a successful call to cec_register_adapter().
+	 */
+	devm_remove_action(dev, tda9950_cec_del, priv);
+
+	cec_register_cec_notifier(priv->adap, priv->notify);
+
+	return 0;
+}
+
+static int tda9950_remove(struct i2c_client *client)
+{
+	struct tda9950_priv *priv = i2c_get_clientdata(client);
+
+	cec_unregister_adapter(priv->adap);
+	cec_notifier_put(priv->notify);
+
+	return 0;
+}
+
+static struct i2c_device_id tda9950_ids[] = {
+	{ "tda9950", 0 },
+	{ },
+};
+MODULE_DEVICE_TABLE(i2c, tda9950_ids);
+
+static struct i2c_driver tda9950_driver = {
+	.probe = tda9950_probe,
+	.remove = tda9950_remove,
+	.driver = {
+		.name = "tda9950",
+	},
+	.id_table = tda9950_ids,
+};
+
+module_i2c_driver(tda9950_driver);
+
+MODULE_AUTHOR("Russell King <rmk+kernel@armlinux.org.uk>");
+MODULE_DESCRIPTION("TDA9950/TDA998x Consumer Electronics Control Driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
index 421c8a72369e..eecdc327b9f8 100644
--- a/drivers/gpu/drm/i2c/tda998x_drv.c
+++ b/drivers/gpu/drm/i2c/tda998x_drv.c
@@ -16,8 +16,10 @@
  */
 
 #include <linux/component.h>
+#include <linux/gpio/consumer.h>
 #include <linux/hdmi.h>
 #include <linux/module.h>
+#include <linux/platform_data/tda9950.h>
 #include <linux/irq.h>
 #include <sound/asoundef.h>
 #include <sound/hdmi-codec.h>
@@ -29,6 +31,8 @@
 #include <drm/drm_of.h>
 #include <drm/i2c/tda998x.h>
 
+#include <media/cec-notifier.h>
+
 #define DBG(fmt, ...) DRM_DEBUG(fmt"\n", ##__VA_ARGS__)
 
 struct tda998x_audio_port {
@@ -55,6 +59,7 @@ struct tda998x_priv {
 	struct platform_device *audio_pdev;
 	struct mutex audio_mutex;
 
+	struct mutex edid_mutex;
 	wait_queue_head_t wq_edid;
 	volatile int wq_edid_wait;
 
@@ -67,6 +72,9 @@ struct tda998x_priv {
 	struct drm_connector connector;
 
 	struct tda998x_audio_port audio_port[2];
+	struct tda9950_glue cec_glue;
+	struct gpio_desc *calib;
+	struct cec_notifier *cec_notify;
 };
 
 #define conn_to_tda998x_priv(x) \
@@ -345,6 +353,12 @@ struct tda998x_priv {
 #define REG_CEC_INTSTATUS	  0xee		      /* read */
 # define CEC_INTSTATUS_CEC	  (1 << 0)
 # define CEC_INTSTATUS_HDMI	  (1 << 1)
+#define REG_CEC_CAL_XOSC_CTRL1    0xf2
+# define CEC_CAL_XOSC_CTRL1_ENA_CAL	BIT(0)
+#define REG_CEC_DES_FREQ2         0xf5
+# define CEC_DES_FREQ2_DIS_AUTOCAL BIT(7)
+#define REG_CEC_CLK               0xf6
+# define CEC_CLK_FRO              0x11
 #define REG_CEC_FRO_IM_CLK_CTRL   0xfb                /* read/write */
 # define CEC_FRO_IM_CLK_CTRL_GHOST_DIS (1 << 7)
 # define CEC_FRO_IM_CLK_CTRL_ENA_OTP   (1 << 6)
@@ -359,6 +373,7 @@ struct tda998x_priv {
 # define CEC_RXSHPDLEV_HPD        (1 << 1)
 
 #define REG_CEC_ENAMODS           0xff                /* read/write */
+# define CEC_ENAMODS_EN_CEC_CLK   (1 << 7)
 # define CEC_ENAMODS_DIS_FRO      (1 << 6)
 # define CEC_ENAMODS_DIS_CCLK     (1 << 5)
 # define CEC_ENAMODS_EN_RXSENS    (1 << 2)
@@ -417,6 +432,114 @@ cec_read(struct tda998x_priv *priv, u8 addr)
 	return val;
 }
 
+static void cec_enamods(struct tda998x_priv *priv, u8 mods, bool enable)
+{
+	int val = cec_read(priv, REG_CEC_ENAMODS);
+
+	if (val < 0)
+		return;
+
+	if (enable)
+		val |= mods;
+	else
+		val &= ~mods;
+
+	cec_write(priv, REG_CEC_ENAMODS, val);
+}
+
+static void tda998x_cec_set_calibration(struct tda998x_priv *priv, bool enable)
+{
+	if (enable) {
+		u8 val;
+
+		cec_write(priv, 0xf3, 0xc0);
+		cec_write(priv, 0xf4, 0xd4);
+
+		/* Enable automatic calibration mode */
+		val = cec_read(priv, REG_CEC_DES_FREQ2);
+		val &= ~CEC_DES_FREQ2_DIS_AUTOCAL;
+		cec_write(priv, REG_CEC_DES_FREQ2, val);
+
+		/* Enable free running oscillator */
+		cec_write(priv, REG_CEC_CLK, CEC_CLK_FRO);
+		cec_enamods(priv, CEC_ENAMODS_DIS_FRO, false);
+
+		cec_write(priv, REG_CEC_CAL_XOSC_CTRL1,
+			  CEC_CAL_XOSC_CTRL1_ENA_CAL);
+	} else {
+		cec_write(priv, REG_CEC_CAL_XOSC_CTRL1, 0);
+	}
+}
+
+/*
+ * Calibration for the internal oscillator: we need to set calibration mode,
+ * and then pulse the IRQ line low for a 10ms ± 1% period.
+ */
+static void tda998x_cec_calibration(struct tda998x_priv *priv)
+{
+	struct gpio_desc *calib = priv->calib;
+
+	mutex_lock(&priv->edid_mutex);
+	if (priv->hdmi->irq > 0)
+		disable_irq(priv->hdmi->irq);
+	gpiod_direction_output(calib, 1);
+	tda998x_cec_set_calibration(priv, true);
+
+	local_irq_disable();
+	gpiod_set_value(calib, 0);
+	mdelay(10);
+	gpiod_set_value(calib, 1);
+	local_irq_enable();
+
+	tda998x_cec_set_calibration(priv, false);
+	gpiod_direction_input(calib);
+	if (priv->hdmi->irq > 0)
+		enable_irq(priv->hdmi->irq);
+	mutex_unlock(&priv->edid_mutex);
+}
+
+static int tda998x_cec_hook_init(void *data)
+{
+	struct tda998x_priv *priv = data;
+	struct gpio_desc *calib;
+
+	calib = gpiod_get(&priv->hdmi->dev, "nxp,calib", GPIOD_ASIS);
+	if (IS_ERR(calib)) {
+		dev_warn(&priv->hdmi->dev, "failed to get calibration gpio: %ld\n",
+			 PTR_ERR(calib));
+		return PTR_ERR(calib);
+	}
+
+	priv->calib = calib;
+
+	return 0;
+}
+
+static void tda998x_cec_hook_exit(void *data)
+{
+	struct tda998x_priv *priv = data;
+
+	gpiod_put(priv->calib);
+	priv->calib = NULL;
+}
+
+static int tda998x_cec_hook_open(void *data)
+{
+	struct tda998x_priv *priv = data;
+
+	cec_enamods(priv, CEC_ENAMODS_EN_CEC_CLK | CEC_ENAMODS_EN_CEC, true);
+	tda998x_cec_calibration(priv);
+
+	return 0;
+}
+
+static void tda998x_cec_hook_release(void *data)
+{
+	struct tda998x_priv *priv = data;
+
+	cec_enamods(priv, CEC_ENAMODS_EN_CEC_CLK | CEC_ENAMODS_EN_CEC, false);
+}
+
 static int
 set_page(struct tda998x_priv *priv, u16 reg)
 {
@@ -466,13 +589,22 @@ out:
 	return ret;
 }
 
+#define MAX_WRITE_RANGE_BUF 32
+
 static void
 reg_write_range(struct tda998x_priv *priv, u16 reg, u8 *p, int cnt)
 {
 	struct i2c_client *client = priv->hdmi;
-	u8 buf[cnt+1];
+	/* This is the maximum size of the buffer passed in */
+	u8 buf[MAX_WRITE_RANGE_BUF + 1];
 	int ret;
 
+	if (cnt > MAX_WRITE_RANGE_BUF) {
+		dev_err(&client->dev, "Fixed write buffer too small (%d)\n",
+				MAX_WRITE_RANGE_BUF);
+		return;
+	}
+
 	buf[0] = REG2ADDR(reg);
 	memcpy(&buf[1], p, cnt);
 
@@ -657,10 +789,13 @@ static irqreturn_t tda998x_irq_thread(int irq, void *data)
 			sta, cec, lvl, flag0, flag1, flag2);
 
 		if (cec & CEC_RXSHPDINT_HPD) {
-			if (lvl & CEC_RXSHPDLEV_HPD)
+			if (lvl & CEC_RXSHPDLEV_HPD) {
 				tda998x_edid_delay_start(priv);
-			else
+			} else {
 				schedule_work(&priv->detect_work);
+				cec_notifier_set_phys_addr(priv->cec_notify,
+						   CEC_PHYS_ADDR_INVALID);
+			}
 
 			handled = true;
 		}
@@ -679,7 +814,7 @@ static void
 tda998x_write_if(struct tda998x_priv *priv, u8 bit, u16 addr,
 		 union hdmi_infoframe *frame)
 {
-	u8 buf[32];
+	u8 buf[MAX_WRITE_RANGE_BUF];
 	ssize_t len;
 
 	len = hdmi_infoframe_pack(frame, buf, sizeof(buf));
@@ -981,6 +1116,8 @@ static int tda998x_connector_fill_modes(struct drm_connector *connector,
 	if (connector->edid_blob_ptr) {
 		struct edid *edid = (void *)connector->edid_blob_ptr->data;
 
+		cec_notifier_set_phys_addr_from_edid(priv->cec_notify, edid);
+
 		priv->sink_has_audio = drm_detect_monitor_audio(edid);
 	} else {
 		priv->sink_has_audio = false;
@@ -1024,6 +1161,8 @@ static int read_edid_block(void *data, u8 *buf, unsigned int blk, size_t length)
 	offset = (blk & 1) ? 128 : 0;
 	segptr = blk / 2;
 
+	mutex_lock(&priv->edid_mutex);
+
 	reg_write(priv, REG_DDC_ADDR, 0xa0);
 	reg_write(priv, REG_DDC_OFFS, offset);
 	reg_write(priv, REG_DDC_SEGM_ADDR, 0x60);
@@ -1043,14 +1182,15 @@ static int read_edid_block(void *data, u8 *buf, unsigned int blk, size_t length)
 					msecs_to_jiffies(100));
 		if (i < 0) {
 			dev_err(&priv->hdmi->dev, "read edid wait err %d\n", i);
-			return i;
+			ret = i;
+			goto failed;
 		}
 	} else {
 		for (i = 100; i > 0; i--) {
 			msleep(1);
 			ret = reg_read(priv, REG_INT_FLAGS_2);
 			if (ret < 0)
-				return ret;
+				goto failed;
 			if (ret & INT_FLAGS_2_EDID_BLK_RD)
 				break;
 		}
@@ -1058,17 +1198,22 @@ static int read_edid_block(void *data, u8 *buf, unsigned int blk, size_t length)
 
 	if (i == 0) {
 		dev_err(&priv->hdmi->dev, "read edid timeout\n");
-		return -ETIMEDOUT;
+		ret = -ETIMEDOUT;
+		goto failed;
 	}
 
 	ret = reg_read_range(priv, REG_EDID_DATA_0, buf, length);
 	if (ret != length) {
 		dev_err(&priv->hdmi->dev, "failed to read edid block %d: %d\n",
 			blk, ret);
-		return ret;
+		goto failed;
 	}
 
-	return 0;
+	ret = 0;
+
+ failed:
+	mutex_unlock(&priv->edid_mutex);
+	return ret;
 }
 
 static int tda998x_connector_get_modes(struct drm_connector *connector)
@@ -1098,7 +1243,7 @@ static int tda998x_connector_get_modes(struct drm_connector *connector)
 		return 0;
 	}
 
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 	n = drm_add_edid_modes(connector, edid);
 
 	kfree(edid);
@@ -1156,7 +1301,7 @@ static int tda998x_connector_init(struct tda998x_priv *priv,
 	if (ret)
 		return ret;
 
-	drm_mode_connector_attach_encoder(&priv->connector, &priv->encoder);
+	drm_connector_attach_encoder(&priv->connector, &priv->encoder);
 
 	return 0;
 }
@@ -1423,6 +1568,9 @@ static void tda998x_destroy(struct tda998x_priv *priv)
 	cancel_work_sync(&priv->detect_work);
 
 	i2c_unregister_device(priv->cec);
+
+	if (priv->cec_notify)
+		cec_notifier_put(priv->cec_notify);
 }
 
 /* I2C driver functions */
@@ -1472,10 +1620,16 @@ static int tda998x_get_audio_ports(struct tda998x_priv *priv,
 static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 {
 	struct device_node *np = client->dev.of_node;
+	struct i2c_board_info cec_info;
 	u32 video;
 	int rev_lo, rev_hi, ret;
 
-	mutex_init(&priv->audio_mutex); /* Protect access from audio thread */
+	mutex_init(&priv->mutex);	/* protect the page access */
+	mutex_init(&priv->audio_mutex); /* protect access from audio thread */
+	mutex_init(&priv->edid_mutex);
+	init_waitqueue_head(&priv->edid_delay_waitq);
+	timer_setup(&priv->edid_delay_timer, tda998x_edid_delay_done, 0);
+	INIT_WORK(&priv->detect_work, tda998x_detect_work);
 
 	priv->vip_cntrl_0 = VIP_CNTRL_0_SWAP_A(2) | VIP_CNTRL_0_SWAP_B(3);
 	priv->vip_cntrl_1 = VIP_CNTRL_1_SWAP_C(0) | VIP_CNTRL_1_SWAP_D(1);
@@ -1485,14 +1639,6 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 	priv->cec_addr = 0x34 + (client->addr & 0x03);
 	priv->current_page = 0xff;
 	priv->hdmi = client;
-	priv->cec = i2c_new_dummy(client->adapter, priv->cec_addr);
-	if (!priv->cec)
-		return -ENODEV;
-
-	mutex_init(&priv->mutex);	/* protect the page access */
-	init_waitqueue_head(&priv->edid_delay_waitq);
-	timer_setup(&priv->edid_delay_timer, tda998x_edid_delay_done, 0);
-	INIT_WORK(&priv->detect_work, tda998x_detect_work);
 
 	/* wake up the device: */
 	cec_write(priv, REG_CEC_ENAMODS,
@@ -1502,10 +1648,15 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 
 	/* read version: */
 	rev_lo = reg_read(priv, REG_VERSION_LSB);
+	if (rev_lo < 0) {
+		dev_err(&client->dev, "failed to read version: %d\n", rev_lo);
+		return rev_lo;
+	}
+
 	rev_hi = reg_read(priv, REG_VERSION_MSB);
-	if (rev_lo < 0 || rev_hi < 0) {
-		ret = rev_lo < 0 ? rev_lo : rev_hi;
-		goto fail;
+	if (rev_hi < 0) {
+		dev_err(&client->dev, "failed to read version: %d\n", rev_hi);
+		return rev_hi;
 	}
 
 	priv->rev = rev_lo | rev_hi << 8;
@@ -1529,7 +1680,7 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 	default:
 		dev_err(&client->dev, "found unsupported device: %04x\n",
 			priv->rev);
-		goto fail;
+		return -ENXIO;
 	}
 
 	/* after reset, enable DDC: */
@@ -1545,6 +1696,15 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 	cec_write(priv, REG_CEC_FRO_IM_CLK_CTRL,
 			CEC_FRO_IM_CLK_CTRL_GHOST_DIS | CEC_FRO_IM_CLK_CTRL_IMCLK_SEL);
 
+	/* ensure interrupts are disabled */
+	cec_write(priv, REG_CEC_RXSHPDINTENA, 0);
+
+	/* clear pending interrupts */
+	cec_read(priv, REG_CEC_RXSHPDINT);
+	reg_read(priv, REG_INT_FLAGS_0);
+	reg_read(priv, REG_INT_FLAGS_1);
+	reg_read(priv, REG_INT_FLAGS_2);
+
 	/* initialize the optional IRQ */
 	if (client->irq) {
 		unsigned long irq_flags;
@@ -1552,13 +1712,11 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 		/* init read EDID waitqueue and HDP work */
 		init_waitqueue_head(&priv->wq_edid);
 
-		/* clear pending interrupts */
-		reg_read(priv, REG_INT_FLAGS_0);
-		reg_read(priv, REG_INT_FLAGS_1);
-		reg_read(priv, REG_INT_FLAGS_2);
-
 		irq_flags =
 			irqd_get_trigger_type(irq_get_irq_data(client->irq));
+
+		priv->cec_glue.irq_flags = irq_flags;
+
 		irq_flags |= IRQF_SHARED | IRQF_ONESHOT;
 		ret = request_threaded_irq(client->irq, NULL,
 					   tda998x_irq_thread, irq_flags,
@@ -1567,13 +1725,46 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 			dev_err(&client->dev,
 				"failed to request IRQ#%u: %d\n",
 				client->irq, ret);
-			goto fail;
+			goto err_irq;
 		}
 
 		/* enable HPD irq */
 		cec_write(priv, REG_CEC_RXSHPDINTENA, CEC_RXSHPDLEV_HPD);
 	}
 
+	priv->cec_notify = cec_notifier_get(&client->dev);
+	if (!priv->cec_notify) {
+		ret = -ENOMEM;
+		goto fail;
+	}
+
+	priv->cec_glue.parent = &client->dev;
+	priv->cec_glue.data = priv;
+	priv->cec_glue.init = tda998x_cec_hook_init;
+	priv->cec_glue.exit = tda998x_cec_hook_exit;
+	priv->cec_glue.open = tda998x_cec_hook_open;
+	priv->cec_glue.release = tda998x_cec_hook_release;
+
+	/*
+	 * Some TDA998x are actually two I2C devices merged onto one piece
+	 * of silicon: TDA9989 and TDA19989 combine the HDMI transmitter
+	 * with a slightly modified TDA9950 CEC device.  The CEC device
+	 * is at the TDA9950 address, with the address pins strapped across
+	 * to the TDA998x address pins.  Hence, it always has the same
+	 * offset.
+	 */
+	memset(&cec_info, 0, sizeof(cec_info));
+	strlcpy(cec_info.type, "tda9950", sizeof(cec_info.type));
+	cec_info.addr = priv->cec_addr;
+	cec_info.platform_data = &priv->cec_glue;
+	cec_info.irq = client->irq;
+
+	priv->cec = i2c_new_device(client->adapter, &cec_info);
+	if (!priv->cec) {
+		ret = -ENODEV;
+		goto fail;
+	}
+
 	/* enable EDID read irq: */
 	reg_set(priv, REG_INT_FLAGS_2, INT_FLAGS_2_EDID_BLK_RD);
 
@@ -1596,12 +1787,18 @@ static int tda998x_create(struct i2c_client *client, struct tda998x_priv *priv)
 		tda998x_audio_codec_init(priv, &client->dev);
 
 	return 0;
+
 fail:
 	/* if encoder_init fails, the encoder slave is never registered,
 	 * so cleanup here:
 	 */
 	i2c_unregister_device(priv->cec);
-	return -ENXIO;
+	if (priv->cec_notify)
+		cec_notifier_put(priv->cec_notify);
+	if (client->irq)
+		free_irq(client->irq, priv);
+err_irq:
+	return ret;
 }
 
 static void tda998x_encoder_prepare(struct drm_encoder *encoder)
diff --git a/drivers/gpu/drm/i810/i810_dma.c b/drivers/gpu/drm/i810/i810_dma.c
index 576a417690d4..3b378936f575 100644
--- a/drivers/gpu/drm/i810/i810_dma.c
+++ b/drivers/gpu/drm/i810/i810_dma.c
@@ -934,7 +934,7 @@ static int i810_dma_vertex(struct drm_device *dev, void *data,
 	DRM_DEBUG("idx %d used %d discard %d\n",
 		  vertex->idx, vertex->used, vertex->discard);
 
-	if (vertex->idx < 0 || vertex->idx > dma->buf_count)
+	if (vertex->idx < 0 || vertex->idx >= dma->buf_count)
 		return -EINVAL;
 
 	i810_dma_dispatch_vertex(dev,
diff --git a/drivers/gpu/drm/i915/Kconfig.debug b/drivers/gpu/drm/i915/Kconfig.debug
index 96b577f62752..459f8f88a34c 100644
--- a/drivers/gpu/drm/i915/Kconfig.debug
+++ b/drivers/gpu/drm/i915/Kconfig.debug
@@ -26,7 +26,7 @@ config DRM_I915_DEBUG
         select DRM_VGEM # used by igt/prime_vgem (dmabuf interop checks)
         select DRM_DEBUG_MM if DRM=y
         select STACKDEPOT if DRM=y # for DRM_DEBUG_MM
-	select DRM_DEBUG_MM_SELFTEST
+	select DRM_DEBUG_SELFTEST
 	select SW_SYNC # signaling validation framework (igt/syncobj*)
 	select DRM_I915_SW_FENCE_DEBUG_OBJECTS
 	select DRM_I915_SELFTEST
diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c
index 0dd88fe2e39a..0651e63b25fb 100644
--- a/drivers/gpu/drm/i915/gvt/cmd_parser.c
+++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c
@@ -2902,6 +2902,7 @@ static int init_cmd_table(struct intel_gvt *gvt)
 		if (info) {
 			gvt_err("%s %s duplicated\n", e->info->name,
 					info->name);
+			kfree(e);
 			return -EEXIST;
 		}
 
diff --git a/drivers/gpu/drm/i915/gvt/display.h b/drivers/gpu/drm/i915/gvt/display.h
index b46b86892d58..ea7c1c525b8c 100644
--- a/drivers/gpu/drm/i915/gvt/display.h
+++ b/drivers/gpu/drm/i915/gvt/display.h
@@ -67,7 +67,7 @@
 #define AUX_NATIVE_REPLY_NAK    (0x1 << 4)
 #define AUX_NATIVE_REPLY_DEFER  (0x2 << 4)
 
-#define AUX_BURST_SIZE          16
+#define AUX_BURST_SIZE          20
 
 /* DPCD addresses */
 #define DPCD_REV			0x000
diff --git a/drivers/gpu/drm/i915/gvt/gtt.c b/drivers/gpu/drm/i915/gvt/gtt.c
index 156ceeeb7446..39980dfbbebd 100644
--- a/drivers/gpu/drm/i915/gvt/gtt.c
+++ b/drivers/gpu/drm/i915/gvt/gtt.c
@@ -1894,8 +1894,9 @@ static struct intel_vgpu_mm *intel_vgpu_create_ggtt_mm(struct intel_vgpu *vgpu)
 	mm->type = INTEL_GVT_MM_GGTT;
 
 	nr_entries = gvt_ggtt_gm_sz(vgpu->gvt) >> I915_GTT_PAGE_SHIFT;
-	mm->ggtt_mm.virtual_ggtt = vzalloc(nr_entries *
-					vgpu->gvt->device_info.gtt_entry_size);
+	mm->ggtt_mm.virtual_ggtt =
+		vzalloc(array_size(nr_entries,
+				   vgpu->gvt->device_info.gtt_entry_size));
 	if (!mm->ggtt_mm.virtual_ggtt) {
 		vgpu_free_mm(mm);
 		return ERR_PTR(-ENOMEM);
diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c
index e2e252c67de8..6b50f850dc28 100644
--- a/drivers/gpu/drm/i915/gvt/handlers.c
+++ b/drivers/gpu/drm/i915/gvt/handlers.c
@@ -936,11 +936,14 @@ static int dp_aux_ch_ctl_mmio_write(struct intel_vgpu *vgpu,
 		}
 
 		/*
-		 * Write request format: (command + address) occupies
-		 * 3 bytes, followed by (len + 1) bytes of data.
+		 * Write request format: Headr (command + address + size) occupies
+		 * 4 bytes, followed by (len + 1) bytes of data. See details at
+		 * intel_dp_aux_transfer().
 		 */
-		if (WARN_ON((len + 4) > AUX_BURST_SIZE))
+		if ((len + 1 + 4) > AUX_BURST_SIZE) {
+			gvt_vgpu_err("dp_aux_header: len %d is too large\n", len);
 			return -EINVAL;
+		}
 
 		/* unpack data from vreg to buf */
 		for (t = 0; t < 4; t++) {
@@ -1004,8 +1007,10 @@ static int dp_aux_ch_ctl_mmio_write(struct intel_vgpu *vgpu,
 		/*
 		 * Read reply format: ACK (1 byte) plus (len + 1) bytes of data.
 		 */
-		if (WARN_ON((len + 2) > AUX_BURST_SIZE))
+		if ((len + 2) > AUX_BURST_SIZE) {
+			gvt_vgpu_err("dp_aux_header: len %d is too large\n", len);
 			return -EINVAL;
+		}
 
 		/* read from virtual DPCD to vreg */
 		/* first 4 bytes: [ACK][addr][addr+1][addr+2] */
diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c
index 685cb3de6dab..718ab307a500 100644
--- a/drivers/gpu/drm/i915/gvt/kvmgt.c
+++ b/drivers/gpu/drm/i915/gvt/kvmgt.c
@@ -185,6 +185,12 @@ static int gvt_dma_map_page(struct intel_vgpu *vgpu, unsigned long gfn,
 	if (ret)
 		return ret;
 
+	if (!pfn_valid(pfn)) {
+		gvt_vgpu_err("pfn 0x%lx is not mem backed\n", pfn);
+		vfio_unpin_pages(mdev_dev(vgpu->vdev.mdev), &gfn, 1);
+		return -EINVAL;
+	}
+
 	/* Setup DMA mapping. */
 	*dma_addr = dma_map_page(dev, page, 0, size, PCI_DMA_BIDIRECTIONAL);
 	ret = dma_mapping_error(dev, *dma_addr);
@@ -644,6 +650,17 @@ out:
 	return ret;
 }
 
+static void intel_vgpu_release_msi_eventfd_ctx(struct intel_vgpu *vgpu)
+{
+	struct eventfd_ctx *trigger;
+
+	trigger = vgpu->vdev.msi_trigger;
+	if (trigger) {
+		eventfd_ctx_put(trigger);
+		vgpu->vdev.msi_trigger = NULL;
+	}
+}
+
 static void __intel_vgpu_release(struct intel_vgpu *vgpu)
 {
 	struct kvmgt_guest_info *info;
@@ -668,6 +685,8 @@ static void __intel_vgpu_release(struct intel_vgpu *vgpu)
 	info = (struct kvmgt_guest_info *)vgpu->handle;
 	kvmgt_guest_exit(info);
 
+	intel_vgpu_release_msi_eventfd_ctx(vgpu);
+
 	vgpu->vdev.kvm = NULL;
 	vgpu->handle = 0;
 }
@@ -1048,7 +1067,8 @@ static int intel_vgpu_set_msi_trigger(struct intel_vgpu *vgpu,
 			return PTR_ERR(trigger);
 		}
 		vgpu->vdev.msi_trigger = trigger;
-	}
+	} else if ((flags & VFIO_IRQ_SET_DATA_NONE) && !count)
+		intel_vgpu_release_msi_eventfd_ctx(vgpu);
 
 	return 0;
 }
@@ -1653,6 +1673,18 @@ static int kvmgt_inject_msi(unsigned long handle, u32 addr, u16 data)
 	info = (struct kvmgt_guest_info *)handle;
 	vgpu = info->vgpu;
 
+	/*
+	 * When guest is poweroff, msi_trigger is set to NULL, but vgpu's
+	 * config and mmio register isn't restored to default during guest
+	 * poweroff. If this vgpu is still used in next vm, this vgpu's pipe
+	 * may be enabled, then once this vgpu is active, it will get inject
+	 * vblank interrupt request. But msi_trigger is null until msi is
+	 * enabled by guest. so if msi_trigger is null, success is still
+	 * returned and don't inject interrupt into guest.
+	 */
+	if (vgpu->vdev.msi_trigger == NULL)
+		return 0;
+
 	if (eventfd_signal(vgpu->vdev.msi_trigger, 1) == 1)
 		return 0;
 
diff --git a/drivers/gpu/drm/i915/gvt/mmio.c b/drivers/gpu/drm/i915/gvt/mmio.c
index 2be1be2cf49a..994366035364 100644
--- a/drivers/gpu/drm/i915/gvt/mmio.c
+++ b/drivers/gpu/drm/i915/gvt/mmio.c
@@ -267,7 +267,7 @@ int intel_vgpu_init_mmio(struct intel_vgpu *vgpu)
 {
 	const struct intel_gvt_device_info *info = &vgpu->gvt->device_info;
 
-	vgpu->mmio.vreg = vzalloc(info->mmio_size * 2);
+	vgpu->mmio.vreg = vzalloc(array_size(info->mmio_size, 2));
 	if (!vgpu->mmio.vreg)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/i915/gvt/vgpu.c b/drivers/gpu/drm/i915/gvt/vgpu.c
index aa063b275e81..f6fa916517c3 100644
--- a/drivers/gpu/drm/i915/gvt/vgpu.c
+++ b/drivers/gpu/drm/i915/gvt/vgpu.c
@@ -125,7 +125,7 @@ int intel_gvt_init_vgpu_types(struct intel_gvt *gvt)
 	high_avail = gvt_hidden_sz(gvt) - HOST_HIGH_GM_SIZE;
 	num_types = sizeof(vgpu_types) / sizeof(vgpu_types[0]);
 
-	gvt->types = kzalloc(num_types * sizeof(struct intel_vgpu_type),
+	gvt->types = kcalloc(num_types, sizeof(struct intel_vgpu_type),
 			     GFP_KERNEL);
 	if (!gvt->types)
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index 23e9a86cbc2a..18a45e7a3d7c 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -2847,10 +2847,10 @@ static const struct drm_ioctl_desc i915_ioctls[] = {
 	DRM_IOCTL_DEF_DRV(I915_GEM_GET_APERTURE, i915_gem_get_aperture_ioctl, DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GET_PIPE_FROM_CRTC_ID, intel_get_pipe_from_crtc_id_ioctl, 0),
 	DRM_IOCTL_DEF_DRV(I915_GEM_MADVISE, i915_gem_madvise_ioctl, DRM_RENDER_ALLOW),
-	DRM_IOCTL_DEF_DRV(I915_OVERLAY_PUT_IMAGE, intel_overlay_put_image_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW),
-	DRM_IOCTL_DEF_DRV(I915_OVERLAY_ATTRS, intel_overlay_attrs_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW),
-	DRM_IOCTL_DEF_DRV(I915_SET_SPRITE_COLORKEY, intel_sprite_set_colorkey_ioctl, DRM_MASTER|DRM_CONTROL_ALLOW),
-	DRM_IOCTL_DEF_DRV(I915_GET_SPRITE_COLORKEY, drm_noop, DRM_MASTER|DRM_CONTROL_ALLOW),
+	DRM_IOCTL_DEF_DRV(I915_OVERLAY_PUT_IMAGE, intel_overlay_put_image_ioctl, DRM_MASTER),
+	DRM_IOCTL_DEF_DRV(I915_OVERLAY_ATTRS, intel_overlay_attrs_ioctl, DRM_MASTER),
+	DRM_IOCTL_DEF_DRV(I915_SET_SPRITE_COLORKEY, intel_sprite_set_colorkey_ioctl, DRM_MASTER),
+	DRM_IOCTL_DEF_DRV(I915_GET_SPRITE_COLORKEY, drm_noop, DRM_MASTER),
 	DRM_IOCTL_DEF_DRV(I915_GEM_WAIT, i915_gem_wait_ioctl, DRM_AUTH|DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GEM_CONTEXT_CREATE, i915_gem_context_create_ioctl, DRM_RENDER_ALLOW),
 	DRM_IOCTL_DEF_DRV(I915_GEM_CONTEXT_DESTROY, i915_gem_context_destroy_ioctl, DRM_RENDER_ALLOW),
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index ef1aabb751c6..0f49f9988dfa 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2230,9 +2230,6 @@ static inline struct scatterlist *____sg_next(struct scatterlist *sg)
  **/
 static inline struct scatterlist *__sg_next(struct scatterlist *sg)
 {
-#ifdef CONFIG_DEBUG_SG
-	BUG_ON(sg->sg_magic != SG_MAGIC);
-#endif
 	return sg_is_last(sg) ? NULL : ____sg_next(sg);
 }
 
diff --git a/drivers/gpu/drm/i915/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/i915_gem_dmabuf.c
index 69a7aec49e84..82e2ca17a441 100644
--- a/drivers/gpu/drm/i915/i915_gem_dmabuf.c
+++ b/drivers/gpu/drm/i915/i915_gem_dmabuf.c
@@ -111,15 +111,6 @@ static void i915_gem_dmabuf_vunmap(struct dma_buf *dma_buf, void *vaddr)
 	i915_gem_object_unpin_map(obj);
 }
 
-static void *i915_gem_dmabuf_kmap_atomic(struct dma_buf *dma_buf, unsigned long page_num)
-{
-	return NULL;
-}
-
-static void i915_gem_dmabuf_kunmap_atomic(struct dma_buf *dma_buf, unsigned long page_num, void *addr)
-{
-
-}
 static void *i915_gem_dmabuf_kmap(struct dma_buf *dma_buf, unsigned long page_num)
 {
 	struct drm_i915_gem_object *obj = dma_buf_to_obj(dma_buf);
@@ -225,9 +216,7 @@ static const struct dma_buf_ops i915_dmabuf_ops =  {
 	.unmap_dma_buf = i915_gem_unmap_dma_buf,
 	.release = drm_gem_dmabuf_release,
 	.map = i915_gem_dmabuf_kmap,
-	.map_atomic = i915_gem_dmabuf_kmap_atomic,
 	.unmap = i915_gem_dmabuf_kunmap,
-	.unmap_atomic = i915_gem_dmabuf_kunmap_atomic,
 	.mmap = i915_gem_dmabuf_mmap,
 	.vmap = i915_gem_dmabuf_vmap,
 	.vunmap = i915_gem_dmabuf_vunmap,
diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c
index 81a2c340c091..1b476423bfab 100644
--- a/drivers/gpu/drm/i915/i915_gem_gtt.c
+++ b/drivers/gpu/drm/i915/i915_gem_gtt.c
@@ -4031,7 +4031,7 @@ int i915_gem_gtt_insert(struct i915_address_space *vm,
 
 	mode = DRM_MM_INSERT_BEST;
 	if (flags & PIN_HIGH)
-		mode = DRM_MM_INSERT_HIGH;
+		mode = DRM_MM_INSERT_HIGHEST;
 	if (flags & PIN_MAPPABLE)
 		mode = DRM_MM_INSERT_LOW;
 
@@ -4051,6 +4051,15 @@ int i915_gem_gtt_insert(struct i915_address_space *vm,
 	if (err != -ENOSPC)
 		return err;
 
+	if (mode & DRM_MM_INSERT_ONCE) {
+		err = drm_mm_insert_node_in_range(&vm->mm, node,
+						  size, alignment, color,
+						  start, end,
+						  DRM_MM_INSERT_BEST);
+		if (err != -ENOSPC)
+			return err;
+	}
+
 	if (flags & PIN_NOEVICT)
 		return -ENOSPC;
 
diff --git a/drivers/gpu/drm/i915/intel_atomic.c b/drivers/gpu/drm/i915/intel_atomic.c
index fee026ca449b..b04952bacf77 100644
--- a/drivers/gpu/drm/i915/intel_atomic.c
+++ b/drivers/gpu/drm/i915/intel_atomic.c
@@ -126,6 +126,7 @@ int intel_digital_connector_atomic_check(struct drm_connector *conn,
 	if (new_conn_state->force_audio != old_conn_state->force_audio ||
 	    new_conn_state->broadcast_rgb != old_conn_state->broadcast_rgb ||
 	    new_conn_state->base.picture_aspect_ratio != old_conn_state->base.picture_aspect_ratio ||
+	    new_conn_state->base.content_type != old_conn_state->base.content_type ||
 	    new_conn_state->base.scaling_mode != old_conn_state->base.scaling_mode)
 		crtc_state->mode_changed = true;
 
diff --git a/drivers/gpu/drm/i915/intel_atomic_plane.c b/drivers/gpu/drm/i915/intel_atomic_plane.c
index 69a18d4ff59d..dcba645cabb8 100644
--- a/drivers/gpu/drm/i915/intel_atomic_plane.c
+++ b/drivers/gpu/drm/i915/intel_atomic_plane.c
@@ -120,12 +120,6 @@ int intel_plane_atomic_check_with_state(const struct intel_crtc_state *old_crtc_
 		&crtc_state->base.adjusted_mode;
 	int ret;
 
-	/*
-	 * Both crtc and plane->crtc could be NULL if we're updating a
-	 * property while the plane is disabled.  We don't actually have
-	 * anything driver-specific we need to test in that case, so
-	 * just return success.
-	 */
 	if (!intel_state->base.crtc && !old_plane_state->base.crtc)
 		return 0;
 
@@ -209,12 +203,6 @@ static int intel_plane_atomic_check(struct drm_plane *plane,
 	const struct drm_crtc_state *old_crtc_state;
 	struct drm_crtc_state *new_crtc_state;
 
-	/*
-	 * Both crtc and plane->crtc could be NULL if we're updating a
-	 * property while the plane is disabled.  We don't actually have
-	 * anything driver-specific we need to test in that case, so
-	 * just return success.
-	 */
 	if (!crtc)
 		return 0;
 
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 8719c1a9d1ce..577b30dde45b 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -1022,7 +1022,7 @@ bool intel_crtc_active(struct intel_crtc *crtc)
 	 * We can ditch the adjusted_mode.crtc_clock check as soon
 	 * as Haswell has gained clock readout/fastboot support.
 	 *
-	 * We can ditch the crtc->primary->fb check as soon as we can
+	 * We can ditch the crtc->primary->state->fb check as soon as we can
 	 * properly reconstruct framebuffers.
 	 *
 	 * FIXME: The intel_crtc->active here should be switched to
@@ -2756,10 +2756,10 @@ intel_set_plane_visible(struct intel_crtc_state *crtc_state,
 
 	/* FIXME pre-g4x don't work like this */
 	if (visible) {
-		crtc_state->base.plane_mask |= BIT(drm_plane_index(&plane->base));
+		crtc_state->base.plane_mask |= drm_plane_mask(&plane->base);
 		crtc_state->active_planes |= BIT(plane->id);
 	} else {
-		crtc_state->base.plane_mask &= ~BIT(drm_plane_index(&plane->base));
+		crtc_state->base.plane_mask &= ~drm_plane_mask(&plane->base);
 		crtc_state->active_planes &= ~BIT(plane->id);
 	}
 
@@ -2882,9 +2882,8 @@ valid_fb:
 	if (i915_gem_object_is_tiled(obj))
 		dev_priv->preserve_bios_swizzle = true;
 
-	drm_framebuffer_get(fb);
-	primary->fb = primary->state->fb = fb;
-	primary->crtc = primary->state->crtc = &intel_crtc->base;
+	plane_state->fb = fb;
+	plane_state->crtc = &intel_crtc->base;
 
 	intel_set_plane_visible(to_intel_crtc_state(crtc_state),
 				to_intel_plane_state(plane_state),
@@ -3658,7 +3657,7 @@ u32 glk_plane_color_ctl(const struct intel_crtc_state *crtc_state,
 	plane_color_ctl |= PLANE_COLOR_PLANE_GAMMA_DISABLE;
 	plane_color_ctl |= glk_plane_color_ctl_alpha(fb->format->format);
 
-	if (intel_format_is_yuv(fb->format->format)) {
+	if (fb->format->is_yuv) {
 		if (plane_state->base.color_encoding == DRM_COLOR_YCBCR_BT709)
 			plane_color_ctl |= PLANE_COLOR_CSC_MODE_YUV709_TO_RGB709;
 		else
@@ -11899,7 +11898,7 @@ verify_single_dpll_state(struct drm_i915_private *dev_priv,
 			 struct drm_crtc_state *new_state)
 {
 	struct intel_dpll_hw_state dpll_hw_state;
-	unsigned crtc_mask;
+	unsigned int crtc_mask;
 	bool active;
 
 	memset(&dpll_hw_state, 0, sizeof(dpll_hw_state));
@@ -11926,7 +11925,7 @@ verify_single_dpll_state(struct drm_i915_private *dev_priv,
 		return;
 	}
 
-	crtc_mask = 1 << drm_crtc_index(crtc);
+	crtc_mask = drm_crtc_mask(crtc);
 
 	if (new_state->active)
 		I915_STATE_WARN(!(pll->active_mask & crtc_mask),
@@ -11961,7 +11960,7 @@ verify_shared_dpll_state(struct drm_device *dev, struct drm_crtc *crtc,
 
 	if (old_state->shared_dpll &&
 	    old_state->shared_dpll != new_state->shared_dpll) {
-		unsigned crtc_mask = 1 << drm_crtc_index(crtc);
+		unsigned int crtc_mask = drm_crtc_mask(crtc);
 		struct intel_shared_dpll *pll = old_state->shared_dpll;
 
 		I915_STATE_WARN(pll->active_mask & crtc_mask,
@@ -13305,8 +13304,17 @@ void intel_plane_destroy(struct drm_plane *plane)
 	kfree(to_intel_plane(plane));
 }
 
-static bool i8xx_mod_supported(uint32_t format, uint64_t modifier)
+static bool i8xx_plane_format_mod_supported(struct drm_plane *_plane,
+					    u32 format, u64 modifier)
 {
+	switch (modifier) {
+	case DRM_FORMAT_MOD_LINEAR:
+	case I915_FORMAT_MOD_X_TILED:
+		break;
+	default:
+		return false;
+	}
+
 	switch (format) {
 	case DRM_FORMAT_C8:
 	case DRM_FORMAT_RGB565:
@@ -13319,8 +13327,17 @@ static bool i8xx_mod_supported(uint32_t format, uint64_t modifier)
 	}
 }
 
-static bool i965_mod_supported(uint32_t format, uint64_t modifier)
+static bool i965_plane_format_mod_supported(struct drm_plane *_plane,
+					    u32 format, u64 modifier)
 {
+	switch (modifier) {
+	case DRM_FORMAT_MOD_LINEAR:
+	case I915_FORMAT_MOD_X_TILED:
+		break;
+	default:
+		return false;
+	}
+
 	switch (format) {
 	case DRM_FORMAT_C8:
 	case DRM_FORMAT_RGB565:
@@ -13335,8 +13352,26 @@ static bool i965_mod_supported(uint32_t format, uint64_t modifier)
 	}
 }
 
-static bool skl_mod_supported(uint32_t format, uint64_t modifier)
+static bool skl_plane_format_mod_supported(struct drm_plane *_plane,
+					   u32 format, u64 modifier)
 {
+	struct intel_plane *plane = to_intel_plane(_plane);
+
+	switch (modifier) {
+	case DRM_FORMAT_MOD_LINEAR:
+	case I915_FORMAT_MOD_X_TILED:
+	case I915_FORMAT_MOD_Y_TILED:
+	case I915_FORMAT_MOD_Yf_TILED:
+		break;
+	case I915_FORMAT_MOD_Y_TILED_CCS:
+	case I915_FORMAT_MOD_Yf_TILED_CCS:
+		if (!plane->has_ccs)
+			return false;
+		break;
+	default:
+		return false;
+	}
+
 	switch (format) {
 	case DRM_FORMAT_XRGB8888:
 	case DRM_FORMAT_XBGR8888:
@@ -13368,38 +13403,36 @@ static bool skl_mod_supported(uint32_t format, uint64_t modifier)
 	}
 }
 
-static bool intel_primary_plane_format_mod_supported(struct drm_plane *plane,
-						     uint32_t format,
-						     uint64_t modifier)
+static bool intel_cursor_format_mod_supported(struct drm_plane *_plane,
+					      u32 format, u64 modifier)
 {
-	struct drm_i915_private *dev_priv = to_i915(plane->dev);
-
-	if (WARN_ON(modifier == DRM_FORMAT_MOD_INVALID))
-		return false;
-
-	if ((modifier >> 56) != DRM_FORMAT_MOD_VENDOR_INTEL &&
-	    modifier != DRM_FORMAT_MOD_LINEAR)
-		return false;
-
-	if (INTEL_GEN(dev_priv) >= 9)
-		return skl_mod_supported(format, modifier);
-	else if (INTEL_GEN(dev_priv) >= 4)
-		return i965_mod_supported(format, modifier);
-	else
-		return i8xx_mod_supported(format, modifier);
+	return modifier == DRM_FORMAT_MOD_LINEAR &&
+		format == DRM_FORMAT_ARGB8888;
 }
 
-static bool intel_cursor_plane_format_mod_supported(struct drm_plane *plane,
-						    uint32_t format,
-						    uint64_t modifier)
-{
-	if (WARN_ON(modifier == DRM_FORMAT_MOD_INVALID))
-		return false;
+static struct drm_plane_funcs skl_plane_funcs = {
+	.update_plane = drm_atomic_helper_update_plane,
+	.disable_plane = drm_atomic_helper_disable_plane,
+	.destroy = intel_plane_destroy,
+	.atomic_get_property = intel_plane_atomic_get_property,
+	.atomic_set_property = intel_plane_atomic_set_property,
+	.atomic_duplicate_state = intel_plane_duplicate_state,
+	.atomic_destroy_state = intel_plane_destroy_state,
+	.format_mod_supported = skl_plane_format_mod_supported,
+};
 
-	return modifier == DRM_FORMAT_MOD_LINEAR && format == DRM_FORMAT_ARGB8888;
-}
+static struct drm_plane_funcs i965_plane_funcs = {
+	.update_plane = drm_atomic_helper_update_plane,
+	.disable_plane = drm_atomic_helper_disable_plane,
+	.destroy = intel_plane_destroy,
+	.atomic_get_property = intel_plane_atomic_get_property,
+	.atomic_set_property = intel_plane_atomic_set_property,
+	.atomic_duplicate_state = intel_plane_duplicate_state,
+	.atomic_destroy_state = intel_plane_destroy_state,
+	.format_mod_supported = i965_plane_format_mod_supported,
+};
 
-static struct drm_plane_funcs intel_plane_funcs = {
+static struct drm_plane_funcs i8xx_plane_funcs = {
 	.update_plane = drm_atomic_helper_update_plane,
 	.disable_plane = drm_atomic_helper_disable_plane,
 	.destroy = intel_plane_destroy,
@@ -13407,7 +13440,7 @@ static struct drm_plane_funcs intel_plane_funcs = {
 	.atomic_set_property = intel_plane_atomic_set_property,
 	.atomic_duplicate_state = intel_plane_duplicate_state,
 	.atomic_destroy_state = intel_plane_destroy_state,
-	.format_mod_supported = intel_primary_plane_format_mod_supported,
+	.format_mod_supported = i8xx_plane_format_mod_supported,
 };
 
 static int
@@ -13532,7 +13565,7 @@ static const struct drm_plane_funcs intel_cursor_plane_funcs = {
 	.atomic_set_property = intel_plane_atomic_set_property,
 	.atomic_duplicate_state = intel_plane_duplicate_state,
 	.atomic_destroy_state = intel_plane_destroy_state,
-	.format_mod_supported = intel_cursor_plane_format_mod_supported,
+	.format_mod_supported = intel_cursor_format_mod_supported,
 };
 
 static bool i9xx_plane_has_fbc(struct drm_i915_private *dev_priv,
@@ -13590,6 +13623,7 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 {
 	struct intel_plane *primary = NULL;
 	struct intel_plane_state *state = NULL;
+	const struct drm_plane_funcs *plane_funcs;
 	const uint32_t *intel_primary_formats;
 	unsigned int supported_rotations;
 	unsigned int num_formats;
@@ -13645,6 +13679,9 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 	primary->check_plane = intel_check_primary_plane;
 
 	if (INTEL_GEN(dev_priv) >= 9) {
+		primary->has_ccs = skl_plane_has_ccs(dev_priv, pipe,
+						     PLANE_PRIMARY);
+
 		if (skl_plane_has_planar(dev_priv, pipe, PLANE_PRIMARY)) {
 			intel_primary_formats = skl_pri_planar_formats;
 			num_formats = ARRAY_SIZE(skl_pri_planar_formats);
@@ -13653,7 +13690,7 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 			num_formats = ARRAY_SIZE(skl_primary_formats);
 		}
 
-		if (skl_plane_has_ccs(dev_priv, pipe, PLANE_PRIMARY))
+		if (primary->has_ccs)
 			modifiers = skl_format_modifiers_ccs;
 		else
 			modifiers = skl_format_modifiers_noccs;
@@ -13661,6 +13698,8 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 		primary->update_plane = skl_update_plane;
 		primary->disable_plane = skl_disable_plane;
 		primary->get_hw_state = skl_plane_get_hw_state;
+
+		plane_funcs = &skl_plane_funcs;
 	} else if (INTEL_GEN(dev_priv) >= 4) {
 		intel_primary_formats = i965_primary_formats;
 		num_formats = ARRAY_SIZE(i965_primary_formats);
@@ -13669,6 +13708,8 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 		primary->update_plane = i9xx_update_plane;
 		primary->disable_plane = i9xx_disable_plane;
 		primary->get_hw_state = i9xx_plane_get_hw_state;
+
+		plane_funcs = &i965_plane_funcs;
 	} else {
 		intel_primary_formats = i8xx_primary_formats;
 		num_formats = ARRAY_SIZE(i8xx_primary_formats);
@@ -13677,25 +13718,27 @@ intel_primary_plane_create(struct drm_i915_private *dev_priv, enum pipe pipe)
 		primary->update_plane = i9xx_update_plane;
 		primary->disable_plane = i9xx_disable_plane;
 		primary->get_hw_state = i9xx_plane_get_hw_state;
+
+		plane_funcs = &i8xx_plane_funcs;
 	}
 
 	if (INTEL_GEN(dev_priv) >= 9)
 		ret = drm_universal_plane_init(&dev_priv->drm, &primary->base,
-					       0, &intel_plane_funcs,
+					       0, plane_funcs,
 					       intel_primary_formats, num_formats,
 					       modifiers,
 					       DRM_PLANE_TYPE_PRIMARY,
 					       "plane 1%c", pipe_name(pipe));
 	else if (INTEL_GEN(dev_priv) >= 5 || IS_G4X(dev_priv))
 		ret = drm_universal_plane_init(&dev_priv->drm, &primary->base,
-					       0, &intel_plane_funcs,
+					       0, plane_funcs,
 					       intel_primary_formats, num_formats,
 					       modifiers,
 					       DRM_PLANE_TYPE_PRIMARY,
 					       "primary %c", pipe_name(pipe));
 	else
 		ret = drm_universal_plane_init(&dev_priv->drm, &primary->base,
-					       0, &intel_plane_funcs,
+					       0, plane_funcs,
 					       intel_primary_formats, num_formats,
 					       modifiers,
 					       DRM_PLANE_TYPE_PRIMARY,
@@ -15666,9 +15709,9 @@ static void intel_modeset_readout_hw_state(struct drm_device *dev)
 				 * rely on the connector_mask being accurate.
 				 */
 				encoder->base.crtc->state->connector_mask |=
-					1 << drm_connector_index(&connector->base);
+					drm_connector_mask(&connector->base);
 				encoder->base.crtc->state->encoder_mask |=
-					1 << drm_encoder_index(&encoder->base);
+					drm_encoder_mask(&encoder->base);
 			}
 
 		} else {
@@ -15968,8 +16011,7 @@ void intel_connector_attach_encoder(struct intel_connector *connector,
 				    struct intel_encoder *encoder)
 {
 	connector->encoder = encoder;
-	drm_mode_connector_attach_encoder(&connector->base,
-					  &encoder->base);
+	drm_connector_attach_encoder(&connector->base, &encoder->base);
 }
 
 /*
diff --git a/drivers/gpu/drm/i915/intel_display.h b/drivers/gpu/drm/i915/intel_display.h
index 7559dfc235a2..9292001cdd14 100644
--- a/drivers/gpu/drm/i915/intel_display.h
+++ b/drivers/gpu/drm/i915/intel_display.h
@@ -265,7 +265,7 @@ struct intel_link_m_n {
 			    &(dev)->mode_config.plane_list,		\
 			    base.head)					\
 		for_each_if((plane_mask) &				\
-			    BIT(drm_plane_index(&intel_plane->base)))
+			    drm_plane_mask(&intel_plane->base)))
 
 #define for_each_intel_plane_on_crtc(dev, intel_crtc, intel_plane)	\
 	list_for_each_entry(intel_plane,				\
@@ -282,7 +282,7 @@ struct intel_link_m_n {
 	list_for_each_entry(intel_crtc,					\
 			    &(dev)->mode_config.crtc_list,		\
 			    base.head)					\
-		for_each_if((crtc_mask) & BIT(drm_crtc_index(&intel_crtc->base)))
+		for_each_if((crtc_mask) & drm_crtc_mask(&intel_crtc->base))
 
 #define for_each_intel_encoder(dev, intel_encoder)		\
 	list_for_each_entry(intel_encoder,			\
diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c
index fee23cf93a2b..cd0f649b57a5 100644
--- a/drivers/gpu/drm/i915/intel_dp.c
+++ b/drivers/gpu/drm/i915/intel_dp.c
@@ -4333,6 +4333,9 @@ intel_dp_short_pulse(struct intel_dp *intel_dp)
 			DRM_DEBUG_DRIVER("CP or sink specific irq unhandled\n");
 	}
 
+	/* Handle CEC interrupts, if any */
+	drm_dp_cec_irq(&intel_dp->aux);
+
 	/* defer to the hotplug work for link retraining if needed */
 	if (intel_dp_needs_link_retrain(intel_dp))
 		return false;
@@ -4642,6 +4645,7 @@ intel_dp_set_edid(struct intel_dp *intel_dp)
 	intel_connector->detect_edid = edid;
 
 	intel_dp->has_audio = drm_detect_monitor_audio(edid);
+	drm_dp_cec_set_edid(&intel_dp->aux, edid);
 }
 
 static void
@@ -4649,6 +4653,7 @@ intel_dp_unset_edid(struct intel_dp *intel_dp)
 {
 	struct intel_connector *intel_connector = intel_dp->attached_connector;
 
+	drm_dp_cec_unset_edid(&intel_dp->aux);
 	kfree(intel_connector->detect_edid);
 	intel_connector->detect_edid = NULL;
 
@@ -4837,6 +4842,7 @@ static int
 intel_dp_connector_register(struct drm_connector *connector)
 {
 	struct intel_dp *intel_dp = intel_attached_dp(connector);
+	struct drm_device *dev = connector->dev;
 	int ret;
 
 	ret = intel_connector_register(connector);
@@ -4849,13 +4855,20 @@ intel_dp_connector_register(struct drm_connector *connector)
 		      intel_dp->aux.name, connector->kdev->kobj.name);
 
 	intel_dp->aux.dev = connector->kdev;
-	return drm_dp_aux_register(&intel_dp->aux);
+	ret = drm_dp_aux_register(&intel_dp->aux);
+	if (!ret)
+		drm_dp_cec_register_connector(&intel_dp->aux,
+					      connector->name, dev->dev);
+	return ret;
 }
 
 static void
 intel_dp_connector_unregister(struct drm_connector *connector)
 {
-	drm_dp_aux_unregister(&intel_attached_dp(connector)->aux);
+	struct intel_dp *intel_dp = intel_attached_dp(connector);
+
+	drm_dp_cec_unregister_connector(&intel_dp->aux);
+	drm_dp_aux_unregister(&intel_dp->aux);
 	intel_connector_unregister(connector);
 }
 
@@ -6051,7 +6064,7 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp,
 	edid = drm_get_edid(connector, &intel_dp->aux.ddc);
 	if (edid) {
 		if (drm_add_edid_modes(connector, edid)) {
-			drm_mode_connector_update_edid_property(connector,
+			drm_connector_update_edid_property(connector,
 								edid);
 		} else {
 			kfree(edid);
@@ -6140,8 +6153,8 @@ static void intel_dp_modeset_retry_work_fn(struct work_struct *work)
 	/* Set connector link status to BAD and send a Uevent to notify
 	 * userspace to do a modeset.
 	 */
-	drm_mode_connector_set_link_status_property(connector,
-						    DRM_MODE_LINK_STATUS_BAD);
+	drm_connector_set_link_status_property(connector,
+					       DRM_MODE_LINK_STATUS_BAD);
 	mutex_unlock(&connector->dev->mode_config.mutex);
 	/* Send Hotplug uevent so userspace can reprobe */
 	drm_kms_helper_hotplug_event(connector->dev);
diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c b/drivers/gpu/drm/i915/intel_dp_mst.c
index d88d0f5abdce..18c65f8e4fe8 100644
--- a/drivers/gpu/drm/i915/intel_dp_mst.c
+++ b/drivers/gpu/drm/i915/intel_dp_mst.c
@@ -402,20 +402,10 @@ static struct drm_encoder *intel_mst_atomic_best_encoder(struct drm_connector *c
 	return &intel_dp->mst_encoders[crtc->pipe]->base.base;
 }
 
-static struct drm_encoder *intel_mst_best_encoder(struct drm_connector *connector)
-{
-	struct intel_connector *intel_connector = to_intel_connector(connector);
-	struct intel_dp *intel_dp = intel_connector->mst_port;
-	if (!intel_dp)
-		return NULL;
-	return &intel_dp->mst_encoders[0]->base.base;
-}
-
 static const struct drm_connector_helper_funcs intel_dp_mst_connector_helper_funcs = {
 	.get_modes = intel_dp_mst_get_modes,
 	.mode_valid = intel_dp_mst_mode_valid,
 	.atomic_best_encoder = intel_mst_atomic_best_encoder,
-	.best_encoder = intel_mst_best_encoder,
 	.atomic_check = intel_dp_mst_atomic_check,
 };
 
@@ -475,8 +465,7 @@ static struct drm_connector *intel_dp_add_mst_connector(struct drm_dp_mst_topolo
 		struct drm_encoder *enc =
 			&intel_dp->mst_encoders[pipe]->base.base;
 
-		ret = drm_mode_connector_attach_encoder(&intel_connector->base,
-							enc);
+		ret = drm_connector_attach_encoder(&intel_connector->base, enc);
 		if (ret)
 			goto err;
 	}
@@ -484,7 +473,7 @@ static struct drm_connector *intel_dp_add_mst_connector(struct drm_dp_mst_topolo
 	drm_object_attach_property(&connector->base, dev->mode_config.path_property, 0);
 	drm_object_attach_property(&connector->base, dev->mode_config.tile_property, 0);
 
-	ret = drm_mode_connector_set_path_property(connector, pathprop);
+	ret = drm_connector_set_path_property(connector, pathprop);
 	if (ret)
 		goto err;
 
diff --git a/drivers/gpu/drm/i915/intel_dpll_mgr.c b/drivers/gpu/drm/i915/intel_dpll_mgr.c
index e046c4f668e0..7e5e6eb5dfe2 100644
--- a/drivers/gpu/drm/i915/intel_dpll_mgr.c
+++ b/drivers/gpu/drm/i915/intel_dpll_mgr.c
@@ -163,8 +163,8 @@ void intel_enable_shared_dpll(struct intel_crtc *crtc)
 	struct drm_device *dev = crtc->base.dev;
 	struct drm_i915_private *dev_priv = to_i915(dev);
 	struct intel_shared_dpll *pll = crtc->config->shared_dpll;
-	unsigned crtc_mask = 1 << drm_crtc_index(&crtc->base);
-	unsigned old_mask;
+	unsigned int crtc_mask = drm_crtc_mask(&crtc->base);
+	unsigned int old_mask;
 
 	if (WARN_ON(pll == NULL))
 		return;
@@ -207,7 +207,7 @@ void intel_disable_shared_dpll(struct intel_crtc *crtc)
 {
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_shared_dpll *pll = crtc->config->shared_dpll;
-	unsigned crtc_mask = 1 << drm_crtc_index(&crtc->base);
+	unsigned int crtc_mask = drm_crtc_mask(&crtc->base);
 
 	/* PCH only available on ILK+ */
 	if (INTEL_GEN(dev_priv) < 5)
diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h
index edcfd85e7b94..c275f91244a6 100644
--- a/drivers/gpu/drm/i915/intel_drv.h
+++ b/drivers/gpu/drm/i915/intel_drv.h
@@ -949,6 +949,7 @@ struct intel_plane {
 	enum pipe pipe;
 	bool can_scale;
 	bool has_fbc;
+	bool has_ccs;
 	int max_downscale;
 	uint32_t frontbuffer_bit;
 
@@ -2086,7 +2087,6 @@ bool intel_sdvo_init(struct drm_i915_private *dev_priv,
 
 
 /* intel_sprite.c */
-bool intel_format_is_yuv(u32 format);
 int intel_usecs_to_scanlines(const struct drm_display_mode *adjusted_mode,
 			     int usecs);
 struct intel_plane *intel_sprite_plane_create(struct drm_i915_private *dev_priv,
@@ -2102,7 +2102,6 @@ void skl_disable_plane(struct intel_plane *plane, struct intel_crtc *crtc);
 bool skl_plane_get_hw_state(struct intel_plane *plane, enum pipe *pipe);
 bool skl_plane_has_ccs(struct drm_i915_private *dev_priv,
 		       enum pipe pipe, enum plane_id plane_id);
-bool intel_format_is_yuv(uint32_t format);
 bool skl_plane_has_planar(struct drm_i915_private *dev_priv,
 			  enum pipe pipe, enum plane_id plane_id);
 
diff --git a/drivers/gpu/drm/i915/intel_hdcp.c b/drivers/gpu/drm/i915/intel_hdcp.c
index 2db5da550a1c..0cc6a861bcf8 100644
--- a/drivers/gpu/drm/i915/intel_hdcp.c
+++ b/drivers/gpu/drm/i915/intel_hdcp.c
@@ -429,7 +429,7 @@ int intel_hdcp_auth_downstream(struct intel_digital_port *intel_dig_port,
 	if (num_downstream == 0)
 		return -EINVAL;
 
-	ksv_fifo = kzalloc(num_downstream * DRM_HDCP_KSV_LEN, GFP_KERNEL);
+	ksv_fifo = kcalloc(DRM_HDCP_KSV_LEN, num_downstream, GFP_KERNEL);
 	if (!ksv_fifo)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c
index 98ad4313d9f2..8363fbd18ee8 100644
--- a/drivers/gpu/drm/i915/intel_hdmi.c
+++ b/drivers/gpu/drm/i915/intel_hdmi.c
@@ -467,7 +467,8 @@ static void intel_write_infoframe(struct drm_encoder *encoder,
 }
 
 static void intel_hdmi_set_avi_infoframe(struct drm_encoder *encoder,
-					 const struct intel_crtc_state *crtc_state)
+					 const struct intel_crtc_state *crtc_state,
+					 const struct drm_connector_state *conn_state)
 {
 	struct intel_hdmi *intel_hdmi = enc_to_intel_hdmi(encoder);
 	const struct drm_display_mode *adjusted_mode =
@@ -497,6 +498,9 @@ static void intel_hdmi_set_avi_infoframe(struct drm_encoder *encoder,
 					   intel_hdmi->rgb_quant_range_selectable,
 					   is_hdmi2_sink);
 
+	drm_hdmi_avi_infoframe_content_type(&frame.avi,
+					    conn_state);
+
 	/* TODO: handle pixel repetition for YCBCR420 outputs */
 	intel_write_infoframe(encoder, crtc_state, &frame);
 }
@@ -592,7 +596,7 @@ static void g4x_set_infoframes(struct drm_encoder *encoder,
 	I915_WRITE(reg, val);
 	POSTING_READ(reg);
 
-	intel_hdmi_set_avi_infoframe(encoder, crtc_state);
+	intel_hdmi_set_avi_infoframe(encoder, crtc_state, conn_state);
 	intel_hdmi_set_spd_infoframe(encoder, crtc_state);
 	intel_hdmi_set_hdmi_infoframe(encoder, crtc_state, conn_state);
 }
@@ -733,7 +737,7 @@ static void ibx_set_infoframes(struct drm_encoder *encoder,
 	I915_WRITE(reg, val);
 	POSTING_READ(reg);
 
-	intel_hdmi_set_avi_infoframe(encoder, crtc_state);
+	intel_hdmi_set_avi_infoframe(encoder, crtc_state, conn_state);
 	intel_hdmi_set_spd_infoframe(encoder, crtc_state);
 	intel_hdmi_set_hdmi_infoframe(encoder, crtc_state, conn_state);
 }
@@ -776,7 +780,7 @@ static void cpt_set_infoframes(struct drm_encoder *encoder,
 	I915_WRITE(reg, val);
 	POSTING_READ(reg);
 
-	intel_hdmi_set_avi_infoframe(encoder, crtc_state);
+	intel_hdmi_set_avi_infoframe(encoder, crtc_state, conn_state);
 	intel_hdmi_set_spd_infoframe(encoder, crtc_state);
 	intel_hdmi_set_hdmi_infoframe(encoder, crtc_state, conn_state);
 }
@@ -829,7 +833,7 @@ static void vlv_set_infoframes(struct drm_encoder *encoder,
 	I915_WRITE(reg, val);
 	POSTING_READ(reg);
 
-	intel_hdmi_set_avi_infoframe(encoder, crtc_state);
+	intel_hdmi_set_avi_infoframe(encoder, crtc_state, conn_state);
 	intel_hdmi_set_spd_infoframe(encoder, crtc_state);
 	intel_hdmi_set_hdmi_infoframe(encoder, crtc_state, conn_state);
 }
@@ -862,7 +866,7 @@ static void hsw_set_infoframes(struct drm_encoder *encoder,
 	I915_WRITE(reg, val);
 	POSTING_READ(reg);
 
-	intel_hdmi_set_avi_infoframe(encoder, crtc_state);
+	intel_hdmi_set_avi_infoframe(encoder, crtc_state, conn_state);
 	intel_hdmi_set_spd_infoframe(encoder, crtc_state);
 	intel_hdmi_set_hdmi_infoframe(encoder, crtc_state, conn_state);
 }
@@ -2092,6 +2096,7 @@ intel_hdmi_add_properties(struct intel_hdmi *intel_hdmi, struct drm_connector *c
 	intel_attach_force_audio_property(connector);
 	intel_attach_broadcast_rgb_property(connector);
 	intel_attach_aspect_ratio_property(connector);
+	drm_connector_attach_content_type_property(connector);
 	connector->state->picture_aspect_ratio = HDMI_PICTURE_ASPECT_NONE;
 }
 
diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c
index a35404119257..f9f3b0885ba5 100644
--- a/drivers/gpu/drm/i915/intel_lvds.c
+++ b/drivers/gpu/drm/i915/intel_lvds.c
@@ -999,7 +999,7 @@ void intel_lvds_init(struct drm_i915_private *dev_priv)
 				    intel_gmbus_get_adapter(dev_priv, pin));
 	if (edid) {
 		if (drm_add_edid_modes(connector, edid)) {
-			drm_mode_connector_update_edid_property(connector,
+			drm_connector_update_edid_property(connector,
 								edid);
 		} else {
 			kfree(edid);
diff --git a/drivers/gpu/drm/i915/intel_modes.c b/drivers/gpu/drm/i915/intel_modes.c
index b39846613e3c..ca44bf368e24 100644
--- a/drivers/gpu/drm/i915/intel_modes.c
+++ b/drivers/gpu/drm/i915/intel_modes.c
@@ -40,7 +40,7 @@ int intel_connector_update_modes(struct drm_connector *connector,
 {
 	int ret;
 
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 	ret = drm_add_edid_modes(connector, edid);
 
 	return ret;
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c
index 69bd7f697f6d..33faad3197fe 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.c
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.c
@@ -1023,6 +1023,8 @@ int intel_ring_pin(struct intel_ring *ring,
 		flags |= PIN_OFFSET_BIAS | offset_bias;
 	if (vma->obj->stolen)
 		flags |= PIN_MAPPABLE;
+	else
+		flags |= PIN_HIGH;
 
 	if (!(vma->flags & I915_VMA_GLOBAL_BIND)) {
 		if (flags & PIN_MAPPABLE || map == I915_MAP_WC)
diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c
index 5417c54f67da..812fe7b06f87 100644
--- a/drivers/gpu/drm/i915/intel_sdvo.c
+++ b/drivers/gpu/drm/i915/intel_sdvo.c
@@ -1911,7 +1911,7 @@ static void intel_sdvo_get_ddc_modes(struct drm_connector *connector)
 	if (edid != NULL) {
 		if (intel_sdvo_connector_matches_edid(to_intel_sdvo_connector(connector),
 						      edid)) {
-			drm_mode_connector_update_edid_property(connector, edid);
+			drm_connector_update_edid_property(connector, edid);
 			drm_add_edid_modes(connector, edid);
 		}
 
@@ -2798,9 +2798,8 @@ static bool intel_sdvo_tv_create_property(struct intel_sdvo *intel_sdvo,
 		return false;
 
 	for (i = 0; i < intel_sdvo_connector->format_supported_num; i++)
-		drm_property_add_enum(
-				intel_sdvo_connector->tv_format, i,
-				i, tv_format_names[intel_sdvo_connector->tv_format_supported[i]]);
+		drm_property_add_enum(intel_sdvo_connector->tv_format, i,
+				      tv_format_names[intel_sdvo_connector->tv_format_supported[i]]);
 
 	intel_sdvo_connector->base.base.state->tv.mode = intel_sdvo_connector->tv_format_supported[0];
 	drm_object_attach_property(&intel_sdvo_connector->base.base.base,
diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c
index 3a4a26dd770f..f7026e887fa9 100644
--- a/drivers/gpu/drm/i915/intel_sprite.c
+++ b/drivers/gpu/drm/i915/intel_sprite.c
@@ -41,20 +41,6 @@
 #include <drm/i915_drm.h>
 #include "i915_drv.h"
 
-bool intel_format_is_yuv(u32 format)
-{
-	switch (format) {
-	case DRM_FORMAT_YUYV:
-	case DRM_FORMAT_UYVY:
-	case DRM_FORMAT_VYUY:
-	case DRM_FORMAT_YVYU:
-	case DRM_FORMAT_NV12:
-		return true;
-	default:
-		return false;
-	}
-}
-
 int intel_usecs_to_scanlines(const struct drm_display_mode *adjusted_mode,
 			     int usecs)
 {
@@ -416,7 +402,7 @@ chv_update_csc(const struct intel_plane_state *plane_state)
 	const s16 *csc = csc_matrix[plane_state->base.color_encoding];
 
 	/* Seems RGB data bypasses the CSC always */
-	if (!intel_format_is_yuv(fb->format->format))
+	if (!fb->format->is_yuv)
 		return;
 
 	I915_WRITE_FW(SPCSCYGOFF(plane_id), SPCSC_OOFF(0) | SPCSC_IOFF(0));
@@ -451,7 +437,7 @@ vlv_update_clrc(const struct intel_plane_state *plane_state)
 	enum plane_id plane_id = plane->id;
 	int contrast, brightness, sh_scale, sh_sin, sh_cos;
 
-	if (intel_format_is_yuv(fb->format->format) &&
+	if (fb->format->is_yuv &&
 	    plane_state->base.color_range == DRM_COLOR_YCBCR_LIMITED_RANGE) {
 		/*
 		 * Expand limited range to full range:
@@ -978,22 +964,12 @@ intel_check_sprite_plane(struct intel_plane *plane,
 	struct drm_i915_private *dev_priv = to_i915(plane->base.dev);
 	struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc);
 	struct drm_framebuffer *fb = state->base.fb;
-	int crtc_x, crtc_y;
-	unsigned int crtc_w, crtc_h;
-	uint32_t src_x, src_y, src_w, src_h;
-	struct drm_rect *src = &state->base.src;
-	struct drm_rect *dst = &state->base.dst;
-	struct drm_rect clip = {};
 	int max_stride = INTEL_GEN(dev_priv) >= 9 ? 32768 : 16384;
-	int hscale, vscale;
 	int max_scale, min_scale;
 	bool can_scale;
 	int ret;
 	uint32_t pixel_format = 0;
 
-	*src = drm_plane_state_src(&state->base);
-	*dst = drm_plane_state_dest(&state->base);
-
 	if (!fb) {
 		state->base.visible = false;
 		return 0;
@@ -1032,64 +1008,19 @@ intel_check_sprite_plane(struct intel_plane *plane,
 		min_scale = plane->can_scale ? 1 : (1 << 16);
 	}
 
-	/*
-	 * FIXME the following code does a bunch of fuzzy adjustments to the
-	 * coordinates and sizes. We probably need some way to decide whether
-	 * more strict checking should be done instead.
-	 */
-	drm_rect_rotate(src, fb->width << 16, fb->height << 16,
-			state->base.rotation);
-
-	hscale = drm_rect_calc_hscale_relaxed(src, dst, min_scale, max_scale);
-	BUG_ON(hscale < 0);
-
-	vscale = drm_rect_calc_vscale_relaxed(src, dst, min_scale, max_scale);
-	BUG_ON(vscale < 0);
-
-	if (crtc_state->base.enable)
-		drm_mode_get_hv_timing(&crtc_state->base.mode,
-				       &clip.x2, &clip.y2);
-
-	state->base.visible = drm_rect_clip_scaled(src, dst, &clip, hscale, vscale);
-
-	crtc_x = dst->x1;
-	crtc_y = dst->y1;
-	crtc_w = drm_rect_width(dst);
-	crtc_h = drm_rect_height(dst);
+	ret = drm_atomic_helper_check_plane_state(&state->base,
+						  &crtc_state->base,
+						  min_scale, max_scale,
+						  true, true);
+	if (ret)
+		return ret;
 
 	if (state->base.visible) {
-		/* check again in case clipping clamped the results */
-		hscale = drm_rect_calc_hscale(src, dst, min_scale, max_scale);
-		if (hscale < 0) {
-			DRM_DEBUG_KMS("Horizontal scaling factor out of limits\n");
-			drm_rect_debug_print("src: ", src, true);
-			drm_rect_debug_print("dst: ", dst, false);
-
-			return hscale;
-		}
-
-		vscale = drm_rect_calc_vscale(src, dst, min_scale, max_scale);
-		if (vscale < 0) {
-			DRM_DEBUG_KMS("Vertical scaling factor out of limits\n");
-			drm_rect_debug_print("src: ", src, true);
-			drm_rect_debug_print("dst: ", dst, false);
-
-			return vscale;
-		}
-
-		/* Make the source viewport size an exact multiple of the scaling factors. */
-		drm_rect_adjust_size(src,
-				     drm_rect_width(dst) * hscale - drm_rect_width(src),
-				     drm_rect_height(dst) * vscale - drm_rect_height(src));
-
-		drm_rect_rotate_inv(src, fb->width << 16, fb->height << 16,
-				    state->base.rotation);
-
-		/* sanity check to make sure the src viewport wasn't enlarged */
-		WARN_ON(src->x1 < (int) state->base.src_x ||
-			src->y1 < (int) state->base.src_y ||
-			src->x2 > (int) state->base.src_x + state->base.src_w ||
-			src->y2 > (int) state->base.src_y + state->base.src_h);
+		struct drm_rect *src = &state->base.src;
+		struct drm_rect *dst = &state->base.dst;
+		unsigned int crtc_w = drm_rect_width(dst);
+		unsigned int crtc_h = drm_rect_height(dst);
+		uint32_t src_x, src_y, src_w, src_h;
 
 		/*
 		 * Hardware doesn't handle subpixel coordinates.
@@ -1102,59 +1033,40 @@ intel_check_sprite_plane(struct intel_plane *plane,
 		src_y = src->y1 >> 16;
 		src_h = drm_rect_height(src) >> 16;
 
-		if (intel_format_is_yuv(fb->format->format) &&
-		    fb->format->format != DRM_FORMAT_NV12) {
-			src_x &= ~1;
-			src_w &= ~1;
-
-			/*
-			 * Must keep src and dst the
-			 * same if we can't scale.
-			 */
-			if (!can_scale)
-				crtc_w &= ~1;
+		src->x1 = src_x << 16;
+		src->x2 = (src_x + src_w) << 16;
+		src->y1 = src_y << 16;
+		src->y2 = (src_y + src_h) << 16;
 
-			if (crtc_w == 0)
-				state->base.visible = false;
+		if (fb->format->is_yuv &&
+    		    fb->format->format != DRM_FORMAT_NV12 &&
+		    (src_x % 2 || src_w % 2)) {
+			DRM_DEBUG_KMS("src x/w (%u, %u) must be a multiple of 2 for YUV planes\n",
+				      src_x, src_w);
+			return -EINVAL;
 		}
-	}
-
-	/* Check size restrictions when scaling */
-	if (state->base.visible && (src_w != crtc_w || src_h != crtc_h)) {
-		unsigned int width_bytes;
-		int cpp = fb->format->cpp[0];
-
-		WARN_ON(!can_scale);
 
-		/* FIXME interlacing min height is 6 */
+		/* Check size restrictions when scaling */
+		if (src_w != crtc_w || src_h != crtc_h) {
+			unsigned int width_bytes;
+			int cpp = fb->format->cpp[0];
 
-		if (crtc_w < 3 || crtc_h < 3)
-			state->base.visible = false;
+			WARN_ON(!can_scale);
 
-		if (src_w < 3 || src_h < 3)
-			state->base.visible = false;
+			width_bytes = ((src_x * cpp) & 63) + src_w * cpp;
 
-		width_bytes = ((src_x * cpp) & 63) + src_w * cpp;
-
-		if (INTEL_GEN(dev_priv) < 9 && (src_w > 2048 || src_h > 2048 ||
-		    width_bytes > 4096 || fb->pitches[0] > 4096)) {
-			DRM_DEBUG_KMS("Source dimensions exceed hardware limits\n");
-			return -EINVAL;
+			/* FIXME interlacing min height is 6 */
+			if (INTEL_GEN(dev_priv) < 9 && (
+			     src_w < 3 || src_h < 3 ||
+			     src_w > 2048 || src_h > 2048 ||
+			     crtc_w < 3 || crtc_h < 3 ||
+			     width_bytes > 4096 || fb->pitches[0] > 4096)) {
+				DRM_DEBUG_KMS("Source dimensions exceed hardware limits\n");
+				return -EINVAL;
+			}
 		}
 	}
 
-	if (state->base.visible) {
-		src->x1 = src_x << 16;
-		src->x2 = (src_x + src_w) << 16;
-		src->y1 = src_y << 16;
-		src->y2 = (src_y + src_h) << 16;
-	}
-
-	dst->x1 = crtc_x;
-	dst->x2 = crtc_x + crtc_w;
-	dst->y1 = crtc_y;
-	dst->y2 = crtc_y + crtc_h;
-
 	if (INTEL_GEN(dev_priv) >= 9) {
 		ret = skl_check_plane_surface(crtc_state, state);
 		if (ret)
@@ -1385,8 +1297,17 @@ static const uint64_t skl_plane_format_modifiers_ccs[] = {
 	DRM_FORMAT_MOD_INVALID
 };
 
-static bool g4x_mod_supported(uint32_t format, uint64_t modifier)
+static bool g4x_sprite_format_mod_supported(struct drm_plane *_plane,
+					    u32 format, u64 modifier)
 {
+	switch (modifier) {
+	case DRM_FORMAT_MOD_LINEAR:
+	case I915_FORMAT_MOD_X_TILED:
+		break;
+	default:
+		return false;
+	}
+
 	switch (format) {
 	case DRM_FORMAT_XRGB8888:
 	case DRM_FORMAT_YUYV:
@@ -1402,8 +1323,17 @@ static bool g4x_mod_supported(uint32_t format, uint64_t modifier)
 	}
 }
 
-static bool snb_mod_supported(uint32_t format, uint64_t modifier)
+static bool snb_sprite_format_mod_supported(struct drm_plane *_plane,
+					    u32 format, u64 modifier)
 {
+	switch (modifier) {
+	case DRM_FORMAT_MOD_LINEAR:
+	case I915_FORMAT_MOD_X_TILED:
+		break;
+	default:
+		return false;
+	}
+
 	switch (format) {
 	case DRM_FORMAT_XRGB8888:
 	case DRM_FORMAT_XBGR8888:
@@ -1420,8 +1350,17 @@ static bool snb_mod_supported(uint32_t format, uint64_t modifier)
 	}
 }
 
-static bool vlv_mod_supported(uint32_t format, uint64_t modifier)
+static bool vlv_sprite_format_mod_supported(struct drm_plane *_plane,
+					    u32 format, u64 modifier)
 {
+	switch (modifier) {
+	case DRM_FORMAT_MOD_LINEAR:
+	case I915_FORMAT_MOD_X_TILED:
+		break;
+	default:
+		return false;
+	}
+
 	switch (format) {
 	case DRM_FORMAT_RGB565:
 	case DRM_FORMAT_ABGR8888:
@@ -1443,8 +1382,26 @@ static bool vlv_mod_supported(uint32_t format, uint64_t modifier)
 	}
 }
 
-static bool skl_mod_supported(uint32_t format, uint64_t modifier)
+static bool skl_plane_format_mod_supported(struct drm_plane *_plane,
+					   u32 format, u64 modifier)
 {
+	struct intel_plane *plane = to_intel_plane(_plane);
+
+	switch (modifier) {
+	case DRM_FORMAT_MOD_LINEAR:
+	case I915_FORMAT_MOD_X_TILED:
+	case I915_FORMAT_MOD_Y_TILED:
+	case I915_FORMAT_MOD_Yf_TILED:
+		break;
+	case I915_FORMAT_MOD_Y_TILED_CCS:
+	case I915_FORMAT_MOD_Yf_TILED_CCS:
+		if (!plane->has_ccs)
+			return false;
+		break;
+	default:
+		return false;
+	}
+
 	switch (format) {
 	case DRM_FORMAT_XRGB8888:
 	case DRM_FORMAT_XBGR8888:
@@ -1476,30 +1433,40 @@ static bool skl_mod_supported(uint32_t format, uint64_t modifier)
 	}
 }
 
-static bool intel_sprite_plane_format_mod_supported(struct drm_plane *plane,
-						    uint32_t format,
-						    uint64_t modifier)
-{
-	struct drm_i915_private *dev_priv = to_i915(plane->dev);
-
-	if (WARN_ON(modifier == DRM_FORMAT_MOD_INVALID))
-		return false;
+static const struct drm_plane_funcs g4x_sprite_funcs = {
+	.update_plane = drm_atomic_helper_update_plane,
+	.disable_plane = drm_atomic_helper_disable_plane,
+	.destroy = intel_plane_destroy,
+	.atomic_get_property = intel_plane_atomic_get_property,
+	.atomic_set_property = intel_plane_atomic_set_property,
+	.atomic_duplicate_state = intel_plane_duplicate_state,
+	.atomic_destroy_state = intel_plane_destroy_state,
+	.format_mod_supported = g4x_sprite_format_mod_supported,
+};
 
-	if ((modifier >> 56) != DRM_FORMAT_MOD_VENDOR_INTEL &&
-	    modifier != DRM_FORMAT_MOD_LINEAR)
-		return false;
+static const struct drm_plane_funcs snb_sprite_funcs = {
+	.update_plane = drm_atomic_helper_update_plane,
+	.disable_plane = drm_atomic_helper_disable_plane,
+	.destroy = intel_plane_destroy,
+	.atomic_get_property = intel_plane_atomic_get_property,
+	.atomic_set_property = intel_plane_atomic_set_property,
+	.atomic_duplicate_state = intel_plane_duplicate_state,
+	.atomic_destroy_state = intel_plane_destroy_state,
+	.format_mod_supported = snb_sprite_format_mod_supported,
+};
 
-	if (INTEL_GEN(dev_priv) >= 9)
-		return skl_mod_supported(format, modifier);
-	else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv))
-		return vlv_mod_supported(format, modifier);
-	else if (INTEL_GEN(dev_priv) >= 6)
-		return snb_mod_supported(format, modifier);
-	else
-		return g4x_mod_supported(format, modifier);
-}
+static const struct drm_plane_funcs vlv_sprite_funcs = {
+	.update_plane = drm_atomic_helper_update_plane,
+	.disable_plane = drm_atomic_helper_disable_plane,
+	.destroy = intel_plane_destroy,
+	.atomic_get_property = intel_plane_atomic_get_property,
+	.atomic_set_property = intel_plane_atomic_set_property,
+	.atomic_duplicate_state = intel_plane_duplicate_state,
+	.atomic_destroy_state = intel_plane_destroy_state,
+	.format_mod_supported = vlv_sprite_format_mod_supported,
+};
 
-static const struct drm_plane_funcs intel_sprite_plane_funcs = {
+static const struct drm_plane_funcs skl_plane_funcs = {
 	.update_plane = drm_atomic_helper_update_plane,
 	.disable_plane = drm_atomic_helper_disable_plane,
 	.destroy = intel_plane_destroy,
@@ -1507,7 +1474,7 @@ static const struct drm_plane_funcs intel_sprite_plane_funcs = {
 	.atomic_set_property = intel_plane_atomic_set_property,
 	.atomic_duplicate_state = intel_plane_duplicate_state,
 	.atomic_destroy_state = intel_plane_destroy_state,
-	.format_mod_supported = intel_sprite_plane_format_mod_supported,
+	.format_mod_supported = skl_plane_format_mod_supported,
 };
 
 bool skl_plane_has_ccs(struct drm_i915_private *dev_priv,
@@ -1533,6 +1500,7 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 {
 	struct intel_plane *intel_plane = NULL;
 	struct intel_plane_state *state = NULL;
+	const struct drm_plane_funcs *plane_funcs;
 	unsigned long possible_crtcs;
 	const uint32_t *plane_formats;
 	const uint64_t *modifiers;
@@ -1557,6 +1525,9 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 		intel_plane->can_scale = true;
 		state->scaler_id = -1;
 
+		intel_plane->has_ccs = skl_plane_has_ccs(dev_priv, pipe,
+							 PLANE_SPRITE0 + plane);
+
 		intel_plane->update_plane = skl_update_plane;
 		intel_plane->disable_plane = skl_disable_plane;
 		intel_plane->get_hw_state = skl_plane_get_hw_state;
@@ -1570,10 +1541,12 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 			num_plane_formats = ARRAY_SIZE(skl_plane_formats);
 		}
 
-		if (skl_plane_has_ccs(dev_priv, pipe, PLANE_SPRITE0 + plane))
+		if (intel_plane->has_ccs)
 			modifiers = skl_plane_format_modifiers_ccs;
 		else
 			modifiers = skl_plane_format_modifiers_noccs;
+
+		plane_funcs = &skl_plane_funcs;
 	} else if (IS_VALLEYVIEW(dev_priv) || IS_CHERRYVIEW(dev_priv)) {
 		intel_plane->can_scale = false;
 		intel_plane->max_downscale = 1;
@@ -1585,6 +1558,8 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 		plane_formats = vlv_plane_formats;
 		num_plane_formats = ARRAY_SIZE(vlv_plane_formats);
 		modifiers = i9xx_plane_format_modifiers;
+
+		plane_funcs = &vlv_sprite_funcs;
 	} else if (INTEL_GEN(dev_priv) >= 7) {
 		if (IS_IVYBRIDGE(dev_priv)) {
 			intel_plane->can_scale = true;
@@ -1601,6 +1576,8 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 		plane_formats = snb_plane_formats;
 		num_plane_formats = ARRAY_SIZE(snb_plane_formats);
 		modifiers = i9xx_plane_format_modifiers;
+
+		plane_funcs = &snb_sprite_funcs;
 	} else {
 		intel_plane->can_scale = true;
 		intel_plane->max_downscale = 16;
@@ -1613,9 +1590,13 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 		if (IS_GEN6(dev_priv)) {
 			plane_formats = snb_plane_formats;
 			num_plane_formats = ARRAY_SIZE(snb_plane_formats);
+
+			plane_funcs = &snb_sprite_funcs;
 		} else {
 			plane_formats = g4x_plane_formats;
 			num_plane_formats = ARRAY_SIZE(g4x_plane_formats);
+
+			plane_funcs = &g4x_sprite_funcs;
 		}
 	}
 
@@ -1642,14 +1623,14 @@ intel_sprite_plane_create(struct drm_i915_private *dev_priv,
 
 	if (INTEL_GEN(dev_priv) >= 9)
 		ret = drm_universal_plane_init(&dev_priv->drm, &intel_plane->base,
-					       possible_crtcs, &intel_sprite_plane_funcs,
+					       possible_crtcs, plane_funcs,
 					       plane_formats, num_plane_formats,
 					       modifiers,
 					       DRM_PLANE_TYPE_OVERLAY,
 					       "plane %d%c", plane + 2, pipe_name(pipe));
 	else
 		ret = drm_universal_plane_init(&dev_priv->drm, &intel_plane->base,
-					       possible_crtcs, &intel_sprite_plane_funcs,
+					       possible_crtcs, plane_funcs,
 					       plane_formats, num_plane_formats,
 					       modifiers,
 					       DRM_PLANE_TYPE_OVERLAY,
diff --git a/drivers/gpu/drm/i915/selftests/intel_uncore.c b/drivers/gpu/drm/i915/selftests/intel_uncore.c
index f76f2597df5c..47bc5b2ddb56 100644
--- a/drivers/gpu/drm/i915/selftests/intel_uncore.c
+++ b/drivers/gpu/drm/i915/selftests/intel_uncore.c
@@ -137,7 +137,7 @@ static int intel_uncore_check_forcewake_domains(struct drm_i915_private *dev_pri
 	if (!IS_ENABLED(CONFIG_DRM_I915_SELFTEST_BROKEN))
 		return 0;
 
-	valid = kzalloc(BITS_TO_LONGS(FW_RANGE) * sizeof(*valid),
+	valid = kcalloc(BITS_TO_LONGS(FW_RANGE), sizeof(*valid),
 			GFP_KERNEL);
 	if (!valid)
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/i915/selftests/mock_dmabuf.c b/drivers/gpu/drm/i915/selftests/mock_dmabuf.c
index 302f7d103635..ca682caf1062 100644
--- a/drivers/gpu/drm/i915/selftests/mock_dmabuf.c
+++ b/drivers/gpu/drm/i915/selftests/mock_dmabuf.c
@@ -94,18 +94,6 @@ static void mock_dmabuf_vunmap(struct dma_buf *dma_buf, void *vaddr)
 	vm_unmap_ram(vaddr, mock->npages);
 }
 
-static void *mock_dmabuf_kmap_atomic(struct dma_buf *dma_buf, unsigned long page_num)
-{
-	struct mock_dmabuf *mock = to_mock(dma_buf);
-
-	return kmap_atomic(mock->pages[page_num]);
-}
-
-static void mock_dmabuf_kunmap_atomic(struct dma_buf *dma_buf, unsigned long page_num, void *addr)
-{
-	kunmap_atomic(addr);
-}
-
 static void *mock_dmabuf_kmap(struct dma_buf *dma_buf, unsigned long page_num)
 {
 	struct mock_dmabuf *mock = to_mock(dma_buf);
@@ -130,9 +118,7 @@ static const struct dma_buf_ops mock_dmabuf_ops =  {
 	.unmap_dma_buf = mock_unmap_dma_buf,
 	.release = mock_dmabuf_release,
 	.map = mock_dmabuf_kmap,
-	.map_atomic = mock_dmabuf_kmap_atomic,
 	.unmap = mock_dmabuf_kunmap,
-	.unmap_atomic = mock_dmabuf_kunmap_atomic,
 	.mmap = mock_dmabuf_mmap,
 	.vmap = mock_dmabuf_vmap,
 	.vunmap = mock_dmabuf_vunmap,
diff --git a/drivers/gpu/drm/imx/imx-ldb.c b/drivers/gpu/drm/imx/imx-ldb.c
index 56dd7a9a8e25..7312beb6f1fc 100644
--- a/drivers/gpu/drm/imx/imx-ldb.c
+++ b/drivers/gpu/drm/imx/imx-ldb.c
@@ -143,7 +143,7 @@ static int imx_ldb_connector_get_modes(struct drm_connector *connector)
 		imx_ldb_ch->edid = drm_get_edid(connector, imx_ldb_ch->ddc);
 
 	if (imx_ldb_ch->edid) {
-		drm_mode_connector_update_edid_property(connector,
+		drm_connector_update_edid_property(connector,
 							imx_ldb_ch->edid);
 		num_modes = drm_add_edid_modes(connector, imx_ldb_ch->edid);
 	}
@@ -471,8 +471,7 @@ static int imx_ldb_register(struct drm_device *drm,
 		drm_connector_init(drm, &imx_ldb_ch->connector,
 				&imx_ldb_connector_funcs,
 				DRM_MODE_CONNECTOR_LVDS);
-		drm_mode_connector_attach_encoder(&imx_ldb_ch->connector,
-				encoder);
+		drm_connector_attach_encoder(&imx_ldb_ch->connector, encoder);
 	}
 
 	if (imx_ldb_ch->panel) {
diff --git a/drivers/gpu/drm/imx/imx-tve.c b/drivers/gpu/drm/imx/imx-tve.c
index bc27c2699464..cffd3310240e 100644
--- a/drivers/gpu/drm/imx/imx-tve.c
+++ b/drivers/gpu/drm/imx/imx-tve.c
@@ -235,7 +235,7 @@ static int imx_tve_connector_get_modes(struct drm_connector *connector)
 
 	edid = drm_get_edid(connector, tve->ddc);
 	if (edid) {
-		drm_mode_connector_update_edid_property(connector, edid);
+		drm_connector_update_edid_property(connector, edid);
 		ret = drm_add_edid_modes(connector, edid);
 		kfree(edid);
 	}
@@ -493,7 +493,7 @@ static int imx_tve_register(struct drm_device *drm, struct imx_tve *tve)
 	drm_connector_init(drm, &tve->connector, &imx_tve_connector_funcs,
 			   DRM_MODE_CONNECTOR_VGA);
 
-	drm_mode_connector_attach_encoder(&tve->connector, &tve->encoder);
+	drm_connector_attach_encoder(&tve->connector, &tve->encoder);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/imx/ipuv3-crtc.c b/drivers/gpu/drm/imx/ipuv3-crtc.c
index e83af0f2be86..21d002859ae0 100644
--- a/drivers/gpu/drm/imx/ipuv3-crtc.c
+++ b/drivers/gpu/drm/imx/ipuv3-crtc.c
@@ -213,7 +213,7 @@ static bool ipu_crtc_mode_fixup(struct drm_crtc *crtc,
 static int ipu_crtc_atomic_check(struct drm_crtc *crtc,
 				 struct drm_crtc_state *state)
 {
-	u32 primary_plane_mask = 1 << drm_plane_index(crtc->primary);
+	u32 primary_plane_mask = drm_plane_mask(crtc->primary);
 
 	if (state->active && (primary_plane_mask & state->plane_mask) == 0)
 		return -EINVAL;
diff --git a/drivers/gpu/drm/imx/parallel-display.c b/drivers/gpu/drm/imx/parallel-display.c
index aedecda9728a..aefd04e18f93 100644
--- a/drivers/gpu/drm/imx/parallel-display.c
+++ b/drivers/gpu/drm/imx/parallel-display.c
@@ -63,7 +63,7 @@ static int imx_pd_connector_get_modes(struct drm_connector *connector)
 	}
 
 	if (imxpd->edid) {
-		drm_mode_connector_update_edid_property(connector, imxpd->edid);
+		drm_connector_update_edid_property(connector, imxpd->edid);
 		num_modes = drm_add_edid_modes(connector, imxpd->edid);
 	}
 
@@ -197,7 +197,7 @@ static int imx_pd_register(struct drm_device *drm,
 			return ret;
 		}
 	} else {
-		drm_mode_connector_attach_encoder(&imxpd->connector, encoder);
+		drm_connector_attach_encoder(&imxpd->connector, encoder);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/mediatek/Kconfig b/drivers/gpu/drm/mediatek/Kconfig
index 294de4549922..119ec0a21de2 100644
--- a/drivers/gpu/drm/mediatek/Kconfig
+++ b/drivers/gpu/drm/mediatek/Kconfig
@@ -11,6 +11,7 @@ config DRM_MEDIATEK
 	select DRM_PANEL
 	select MEMORY
 	select MTK_SMI
+	select VIDEOMODE_HELPERS
 	help
 	  Choose this option if you have a Mediatek SoCs.
 	  The module will be called mediatek-drm
diff --git a/drivers/gpu/drm/mediatek/mtk_dpi.c b/drivers/gpu/drm/mediatek/mtk_dpi.c
index e80a603e5fb0..6c0ea39d5739 100644
--- a/drivers/gpu/drm/mediatek/mtk_dpi.c
+++ b/drivers/gpu/drm/mediatek/mtk_dpi.c
@@ -22,6 +22,7 @@
 #include <linux/interrupt.h>
 #include <linux/types.h>
 #include <linux/clk.h>
+#include <video/videomode.h>
 
 #include "mtk_dpi_regs.h"
 #include "mtk_drm_ddp_comp.h"
@@ -429,34 +430,35 @@ static int mtk_dpi_set_display_mode(struct mtk_dpi *dpi,
 	struct mtk_dpi_sync_param vsync_leven = { 0 };
 	struct mtk_dpi_sync_param vsync_rodd = { 0 };
 	struct mtk_dpi_sync_param vsync_reven = { 0 };
-	unsigned long pix_rate;
+	struct videomode vm = { 0 };
 	unsigned long pll_rate;
 	unsigned int factor;
 
 	/* let pll_rate can fix the valid range of tvdpll (1G~2GHz) */
-	pix_rate = 1000UL * mode->clock;
+
 	if (mode->clock <= 27000)
-		factor = 16 * 3;
+		factor = 3 << 4;
 	else if (mode->clock <= 84000)
-		factor = 8 * 3;
+		factor = 3 << 3;
 	else if (mode->clock <= 167000)
-		factor = 4 * 3;
+		factor = 3 << 2;
 	else
-		factor = 2 * 3;
-	pll_rate = pix_rate * factor;
+		factor = 3 << 1;
+	drm_display_mode_to_videomode(mode, &vm);
+	pll_rate = vm.pixelclock * factor;
 
 	dev_dbg(dpi->dev, "Want PLL %lu Hz, pixel clock %lu Hz\n",
-		pll_rate, pix_rate);
+		pll_rate, vm.pixelclock);
 
 	clk_set_rate(dpi->tvd_clk, pll_rate);
 	pll_rate = clk_get_rate(dpi->tvd_clk);
 
-	pix_rate = pll_rate / factor;
-	clk_set_rate(dpi->pixel_clk, pix_rate);
-	pix_rate = clk_get_rate(dpi->pixel_clk);
+	vm.pixelclock = pll_rate / factor;
+	clk_set_rate(dpi->pixel_clk, vm.pixelclock);
+	vm.pixelclock = clk_get_rate(dpi->pixel_clk);
 
 	dev_dbg(dpi->dev, "Got  PLL %lu Hz, pixel clock %lu Hz\n",
-		pll_rate, pix_rate);
+		pll_rate, vm.pixelclock);
 
 	limit.c_bottom = 0x0010;
 	limit.c_top = 0x0FE0;
@@ -465,33 +467,31 @@ static int mtk_dpi_set_display_mode(struct mtk_dpi *dpi,
 
 	dpi_pol.ck_pol = MTK_DPI_POLARITY_FALLING;
 	dpi_pol.de_pol = MTK_DPI_POLARITY_RISING;
-	dpi_pol.hsync_pol = mode->flags & DRM_MODE_FLAG_PHSYNC ?
+	dpi_pol.hsync_pol = vm.flags & DISPLAY_FLAGS_HSYNC_HIGH ?
 			    MTK_DPI_POLARITY_FALLING : MTK_DPI_POLARITY_RISING;
-	dpi_pol.vsync_pol = mode->flags & DRM_MODE_FLAG_PVSYNC ?
+	dpi_pol.vsync_pol = vm.flags & DISPLAY_FLAGS_VSYNC_HIGH ?
 			    MTK_DPI_POLARITY_FALLING : MTK_DPI_POLARITY_RISING;
-
-	hsync.sync_width = mode->hsync_end - mode->hsync_start;
-	hsync.back_porch = mode->htotal - mode->hsync_end;
-	hsync.front_porch = mode->hsync_start - mode->hdisplay;
+	hsync.sync_width = vm.hsync_len;
+	hsync.back_porch = vm.hback_porch;
+	hsync.front_porch = vm.hfront_porch;
 	hsync.shift_half_line = false;
-
-	vsync_lodd.sync_width = mode->vsync_end - mode->vsync_start;
-	vsync_lodd.back_porch = mode->vtotal - mode->vsync_end;
-	vsync_lodd.front_porch = mode->vsync_start - mode->vdisplay;
+	vsync_lodd.sync_width = vm.vsync_len;
+	vsync_lodd.back_porch = vm.vback_porch;
+	vsync_lodd.front_porch = vm.vfront_porch;
 	vsync_lodd.shift_half_line = false;
 
-	if (mode->flags & DRM_MODE_FLAG_INTERLACE &&
+	if (vm.flags & DISPLAY_FLAGS_INTERLACED &&
 	    mode->flags & DRM_MODE_FLAG_3D_MASK) {
 		vsync_leven = vsync_lodd;
 		vsync_rodd = vsync_lodd;
 		vsync_reven = vsync_lodd;
 		vsync_leven.shift_half_line = true;
 		vsync_reven.shift_half_line = true;
-	} else if (mode->flags & DRM_MODE_FLAG_INTERLACE &&
+	} else if (vm.flags & DISPLAY_FLAGS_INTERLACED &&
 		   !(mode->flags & DRM_MODE_FLAG_3D_MASK)) {
 		vsync_leven = vsync_lodd;
 		vsync_leven.shift_half_line = true;
-	} else if (!(mode->flags & DRM_MODE_FLAG_INTERLACE) &&
+	} else if (!(vm.flags & DISPLAY_FLAGS_INTERLACED) &&
 		   mode->flags & DRM_MODE_FLAG_3D_MASK) {
 		vsync_rodd = vsync_lodd;
 	}
@@ -505,12 +505,12 @@ static int mtk_dpi_set_display_mode(struct mtk_dpi *dpi,
 	mtk_dpi_config_vsync_reven(dpi, &vsync_reven);
 
 	mtk_dpi_config_3d(dpi, !!(mode->flags & DRM_MODE_FLAG_3D_MASK));
-	mtk_dpi_config_interface(dpi, !!(mode->flags &
-					 DRM_MODE_FLAG_INTERLACE));
-	if (mode->flags & DRM_MODE_FLAG_INTERLACE)
-		mtk_dpi_config_fb_size(dpi, mode->hdisplay, mode->vdisplay / 2);
+	mtk_dpi_config_interface(dpi, !!(vm.flags &
+					 DISPLAY_FLAGS_INTERLACED));
+	if (vm.flags & DISPLAY_FLAGS_INTERLACED)
+		mtk_dpi_config_fb_size(dpi, vm.hactive, vm.vactive >> 1);
 	else
-		mtk_dpi_config_fb_size(dpi, mode->hdisplay, mode->vdisplay);
+		mtk_dpi_config_fb_size(dpi, vm.hactive, vm.vactive);
 
 	mtk_dpi_config_channel_limit(dpi, &limit);
 	mtk_dpi_config_bit_num(dpi, dpi->bit_num);
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
index 658b8dd45b83..2d6aa150a9ff 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c
@@ -539,6 +539,9 @@ int mtk_drm_crtc_create(struct drm_device *drm_dev,
 	int ret;
 	int i;
 
+	if (!path)
+		return 0;
+
 	for (i = 0; i < path_len; i++) {
 		enum mtk_ddp_comp_id comp_id = path[i];
 		struct device_node *node;
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp.c b/drivers/gpu/drm/mediatek/mtk_drm_ddp.c
index 8130f3dab661..87e4191c250e 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_ddp.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp.c
@@ -28,8 +28,12 @@
 #define DISP_REG_CONFIG_DISP_UFOE_MOUT_EN	0x050
 #define DISP_REG_CONFIG_DISP_COLOR0_SEL_IN	0x084
 #define DISP_REG_CONFIG_DISP_COLOR1_SEL_IN	0x088
+#define DISP_REG_CONFIG_DSIE_SEL_IN		0x0a4
+#define DISP_REG_CONFIG_DSIO_SEL_IN		0x0a8
 #define DISP_REG_CONFIG_DPI_SEL_IN		0x0ac
-#define DISP_REG_CONFIG_DISP_RDMA1_MOUT_EN	0x0c8
+#define DISP_REG_CONFIG_DISP_RDMA2_SOUT		0x0b8
+#define DISP_REG_CONFIG_DISP_RDMA0_SOUT_EN	0x0c4
+#define DISP_REG_CONFIG_DISP_RDMA1_SOUT_EN	0x0c8
 #define DISP_REG_CONFIG_MMSYS_CG_CON0		0x100
 
 #define DISP_REG_CONFIG_DISP_OVL_MOUT_EN	0x030
@@ -41,45 +45,89 @@
 #define DISP_REG_MUTEX_RST(n)	(0x28 + 0x20 * (n))
 #define DISP_REG_MUTEX_MOD(n)	(0x2c + 0x20 * (n))
 #define DISP_REG_MUTEX_SOF(n)	(0x30 + 0x20 * (n))
+#define DISP_REG_MUTEX_MOD2(n)	(0x34 + 0x20 * (n))
 
 #define INT_MUTEX				BIT(1)
 
-#define MT8173_MUTEX_MOD_DISP_OVL0		BIT(11)
-#define MT8173_MUTEX_MOD_DISP_OVL1		BIT(12)
-#define MT8173_MUTEX_MOD_DISP_RDMA0		BIT(13)
-#define MT8173_MUTEX_MOD_DISP_RDMA1		BIT(14)
-#define MT8173_MUTEX_MOD_DISP_RDMA2		BIT(15)
-#define MT8173_MUTEX_MOD_DISP_WDMA0		BIT(16)
-#define MT8173_MUTEX_MOD_DISP_WDMA1		BIT(17)
-#define MT8173_MUTEX_MOD_DISP_COLOR0		BIT(18)
-#define MT8173_MUTEX_MOD_DISP_COLOR1		BIT(19)
-#define MT8173_MUTEX_MOD_DISP_AAL		BIT(20)
-#define MT8173_MUTEX_MOD_DISP_GAMMA		BIT(21)
-#define MT8173_MUTEX_MOD_DISP_UFOE		BIT(22)
-#define MT8173_MUTEX_MOD_DISP_PWM0		BIT(23)
-#define MT8173_MUTEX_MOD_DISP_PWM1		BIT(24)
-#define MT8173_MUTEX_MOD_DISP_OD		BIT(25)
-
-#define MT2701_MUTEX_MOD_DISP_OVL		BIT(3)
-#define MT2701_MUTEX_MOD_DISP_WDMA		BIT(6)
-#define MT2701_MUTEX_MOD_DISP_COLOR		BIT(7)
-#define MT2701_MUTEX_MOD_DISP_BLS		BIT(9)
-#define MT2701_MUTEX_MOD_DISP_RDMA0		BIT(10)
-#define MT2701_MUTEX_MOD_DISP_RDMA1		BIT(12)
+#define MT8173_MUTEX_MOD_DISP_OVL0		11
+#define MT8173_MUTEX_MOD_DISP_OVL1		12
+#define MT8173_MUTEX_MOD_DISP_RDMA0		13
+#define MT8173_MUTEX_MOD_DISP_RDMA1		14
+#define MT8173_MUTEX_MOD_DISP_RDMA2		15
+#define MT8173_MUTEX_MOD_DISP_WDMA0		16
+#define MT8173_MUTEX_MOD_DISP_WDMA1		17
+#define MT8173_MUTEX_MOD_DISP_COLOR0		18
+#define MT8173_MUTEX_MOD_DISP_COLOR1		19
+#define MT8173_MUTEX_MOD_DISP_AAL		20
+#define MT8173_MUTEX_MOD_DISP_GAMMA		21
+#define MT8173_MUTEX_MOD_DISP_UFOE		22
+#define MT8173_MUTEX_MOD_DISP_PWM0		23
+#define MT8173_MUTEX_MOD_DISP_PWM1		24
+#define MT8173_MUTEX_MOD_DISP_OD		25
+
+#define MT2712_MUTEX_MOD_DISP_PWM2		10
+#define MT2712_MUTEX_MOD_DISP_OVL0		11
+#define MT2712_MUTEX_MOD_DISP_OVL1		12
+#define MT2712_MUTEX_MOD_DISP_RDMA0		13
+#define MT2712_MUTEX_MOD_DISP_RDMA1		14
+#define MT2712_MUTEX_MOD_DISP_RDMA2		15
+#define MT2712_MUTEX_MOD_DISP_WDMA0		16
+#define MT2712_MUTEX_MOD_DISP_WDMA1		17
+#define MT2712_MUTEX_MOD_DISP_COLOR0		18
+#define MT2712_MUTEX_MOD_DISP_COLOR1		19
+#define MT2712_MUTEX_MOD_DISP_AAL0		20
+#define MT2712_MUTEX_MOD_DISP_UFOE		22
+#define MT2712_MUTEX_MOD_DISP_PWM0		23
+#define MT2712_MUTEX_MOD_DISP_PWM1		24
+#define MT2712_MUTEX_MOD_DISP_OD0		25
+#define MT2712_MUTEX_MOD2_DISP_AAL1		33
+#define MT2712_MUTEX_MOD2_DISP_OD1		34
+
+#define MT2701_MUTEX_MOD_DISP_OVL		3
+#define MT2701_MUTEX_MOD_DISP_WDMA		6
+#define MT2701_MUTEX_MOD_DISP_COLOR		7
+#define MT2701_MUTEX_MOD_DISP_BLS		9
+#define MT2701_MUTEX_MOD_DISP_RDMA0		10
+#define MT2701_MUTEX_MOD_DISP_RDMA1		12
 
 #define MUTEX_SOF_SINGLE_MODE		0
 #define MUTEX_SOF_DSI0			1
 #define MUTEX_SOF_DSI1			2
 #define MUTEX_SOF_DPI0			3
+#define MUTEX_SOF_DPI1			4
+#define MUTEX_SOF_DSI2			5
+#define MUTEX_SOF_DSI3			6
 
 #define OVL0_MOUT_EN_COLOR0		0x1
 #define OD_MOUT_EN_RDMA0		0x1
+#define OD1_MOUT_EN_RDMA1		BIT(16)
 #define UFOE_MOUT_EN_DSI0		0x1
 #define COLOR0_SEL_IN_OVL0		0x1
 #define OVL1_MOUT_EN_COLOR1		0x1
 #define GAMMA_MOUT_EN_RDMA1		0x1
-#define RDMA1_MOUT_DPI0			0x2
+#define RDMA0_SOUT_DPI0			0x2
+#define RDMA0_SOUT_DSI2			0x4
+#define RDMA0_SOUT_DSI3			0x5
+#define RDMA1_SOUT_DPI0			0x2
+#define RDMA1_SOUT_DPI1			0x3
+#define RDMA1_SOUT_DSI1			0x1
+#define RDMA1_SOUT_DSI2			0x4
+#define RDMA1_SOUT_DSI3			0x5
+#define RDMA2_SOUT_DPI0			0x2
+#define RDMA2_SOUT_DPI1			0x3
+#define RDMA2_SOUT_DSI1			0x1
+#define RDMA2_SOUT_DSI2			0x4
+#define RDMA2_SOUT_DSI3			0x5
 #define DPI0_SEL_IN_RDMA1		0x1
+#define DPI0_SEL_IN_RDMA2		0x3
+#define DPI1_SEL_IN_RDMA1		(0x1 << 8)
+#define DPI1_SEL_IN_RDMA2		(0x3 << 8)
+#define DSI1_SEL_IN_RDMA1		0x1
+#define DSI1_SEL_IN_RDMA2		0x4
+#define DSI2_SEL_IN_RDMA1		(0x1 << 16)
+#define DSI2_SEL_IN_RDMA2		(0x4 << 16)
+#define DSI3_SEL_IN_RDMA1		(0x1 << 16)
+#define DSI3_SEL_IN_RDMA2		(0x4 << 16)
 #define COLOR1_SEL_IN_OVL1		0x1
 
 #define OVL_MOUT_EN_RDMA		0x1
@@ -108,12 +156,32 @@ static const unsigned int mt2701_mutex_mod[DDP_COMPONENT_ID_MAX] = {
 	[DDP_COMPONENT_WDMA0] = MT2701_MUTEX_MOD_DISP_WDMA,
 };
 
+static const unsigned int mt2712_mutex_mod[DDP_COMPONENT_ID_MAX] = {
+	[DDP_COMPONENT_AAL0] = MT2712_MUTEX_MOD_DISP_AAL0,
+	[DDP_COMPONENT_AAL1] = MT2712_MUTEX_MOD2_DISP_AAL1,
+	[DDP_COMPONENT_COLOR0] = MT2712_MUTEX_MOD_DISP_COLOR0,
+	[DDP_COMPONENT_COLOR1] = MT2712_MUTEX_MOD_DISP_COLOR1,
+	[DDP_COMPONENT_OD0] = MT2712_MUTEX_MOD_DISP_OD0,
+	[DDP_COMPONENT_OD1] = MT2712_MUTEX_MOD2_DISP_OD1,
+	[DDP_COMPONENT_OVL0] = MT2712_MUTEX_MOD_DISP_OVL0,
+	[DDP_COMPONENT_OVL1] = MT2712_MUTEX_MOD_DISP_OVL1,
+	[DDP_COMPONENT_PWM0] = MT2712_MUTEX_MOD_DISP_PWM0,
+	[DDP_COMPONENT_PWM1] = MT2712_MUTEX_MOD_DISP_PWM1,
+	[DDP_COMPONENT_PWM2] = MT2712_MUTEX_MOD_DISP_PWM2,
+	[DDP_COMPONENT_RDMA0] = MT2712_MUTEX_MOD_DISP_RDMA0,
+	[DDP_COMPONENT_RDMA1] = MT2712_MUTEX_MOD_DISP_RDMA1,
+	[DDP_COMPONENT_RDMA2] = MT2712_MUTEX_MOD_DISP_RDMA2,
+	[DDP_COMPONENT_UFOE] = MT2712_MUTEX_MOD_DISP_UFOE,
+	[DDP_COMPONENT_WDMA0] = MT2712_MUTEX_MOD_DISP_WDMA0,
+	[DDP_COMPONENT_WDMA1] = MT2712_MUTEX_MOD_DISP_WDMA1,
+};
+
 static const unsigned int mt8173_mutex_mod[DDP_COMPONENT_ID_MAX] = {
-	[DDP_COMPONENT_AAL] = MT8173_MUTEX_MOD_DISP_AAL,
+	[DDP_COMPONENT_AAL0] = MT8173_MUTEX_MOD_DISP_AAL,
 	[DDP_COMPONENT_COLOR0] = MT8173_MUTEX_MOD_DISP_COLOR0,
 	[DDP_COMPONENT_COLOR1] = MT8173_MUTEX_MOD_DISP_COLOR1,
 	[DDP_COMPONENT_GAMMA] = MT8173_MUTEX_MOD_DISP_GAMMA,
-	[DDP_COMPONENT_OD] = MT8173_MUTEX_MOD_DISP_OD,
+	[DDP_COMPONENT_OD0] = MT8173_MUTEX_MOD_DISP_OD,
 	[DDP_COMPONENT_OVL0] = MT8173_MUTEX_MOD_DISP_OVL0,
 	[DDP_COMPONENT_OVL1] = MT8173_MUTEX_MOD_DISP_OVL1,
 	[DDP_COMPONENT_PWM0] = MT8173_MUTEX_MOD_DISP_PWM0,
@@ -138,7 +206,7 @@ static unsigned int mtk_ddp_mout_en(enum mtk_ddp_comp_id cur,
 	} else if (cur == DDP_COMPONENT_OVL0 && next == DDP_COMPONENT_RDMA0) {
 		*addr = DISP_REG_CONFIG_DISP_OVL_MOUT_EN;
 		value = OVL_MOUT_EN_RDMA;
-	} else if (cur == DDP_COMPONENT_OD && next == DDP_COMPONENT_RDMA0) {
+	} else if (cur == DDP_COMPONENT_OD0 && next == DDP_COMPONENT_RDMA0) {
 		*addr = DISP_REG_CONFIG_DISP_OD_MOUT_EN;
 		value = OD_MOUT_EN_RDMA0;
 	} else if (cur == DDP_COMPONENT_UFOE && next == DDP_COMPONENT_DSI0) {
@@ -150,9 +218,48 @@ static unsigned int mtk_ddp_mout_en(enum mtk_ddp_comp_id cur,
 	} else if (cur == DDP_COMPONENT_GAMMA && next == DDP_COMPONENT_RDMA1) {
 		*addr = DISP_REG_CONFIG_DISP_GAMMA_MOUT_EN;
 		value = GAMMA_MOUT_EN_RDMA1;
+	} else if (cur == DDP_COMPONENT_OD1 && next == DDP_COMPONENT_RDMA1) {
+		*addr = DISP_REG_CONFIG_DISP_OD_MOUT_EN;
+		value = OD1_MOUT_EN_RDMA1;
+	} else if (cur == DDP_COMPONENT_RDMA0 && next == DDP_COMPONENT_DPI0) {
+		*addr = DISP_REG_CONFIG_DISP_RDMA0_SOUT_EN;
+		value = RDMA0_SOUT_DPI0;
+	} else if (cur == DDP_COMPONENT_RDMA0 && next == DDP_COMPONENT_DSI2) {
+		*addr = DISP_REG_CONFIG_DISP_RDMA0_SOUT_EN;
+		value = RDMA0_SOUT_DSI2;
+	} else if (cur == DDP_COMPONENT_RDMA0 && next == DDP_COMPONENT_DSI3) {
+		*addr = DISP_REG_CONFIG_DISP_RDMA0_SOUT_EN;
+		value = RDMA0_SOUT_DSI3;
+	} else if (cur == DDP_COMPONENT_RDMA1 && next == DDP_COMPONENT_DSI1) {
+		*addr = DISP_REG_CONFIG_DISP_RDMA1_SOUT_EN;
+		value = RDMA1_SOUT_DSI1;
+	} else if (cur == DDP_COMPONENT_RDMA1 && next == DDP_COMPONENT_DSI2) {
+		*addr = DISP_REG_CONFIG_DISP_RDMA1_SOUT_EN;
+		value = RDMA1_SOUT_DSI2;
+	} else if (cur == DDP_COMPONENT_RDMA1 && next == DDP_COMPONENT_DSI3) {
+		*addr = DISP_REG_CONFIG_DISP_RDMA1_SOUT_EN;
+		value = RDMA1_SOUT_DSI3;
 	} else if (cur == DDP_COMPONENT_RDMA1 && next == DDP_COMPONENT_DPI0) {
-		*addr = DISP_REG_CONFIG_DISP_RDMA1_MOUT_EN;
-		value = RDMA1_MOUT_DPI0;
+		*addr = DISP_REG_CONFIG_DISP_RDMA1_SOUT_EN;
+		value = RDMA1_SOUT_DPI0;
+	} else if (cur == DDP_COMPONENT_RDMA1 && next == DDP_COMPONENT_DPI1) {
+		*addr = DISP_REG_CONFIG_DISP_RDMA1_SOUT_EN;
+		value = RDMA1_SOUT_DPI1;
+	} else if (cur == DDP_COMPONENT_RDMA2 && next == DDP_COMPONENT_DPI0) {
+		*addr = DISP_REG_CONFIG_DISP_RDMA2_SOUT;
+		value = RDMA2_SOUT_DPI0;
+	} else if (cur == DDP_COMPONENT_RDMA2 && next == DDP_COMPONENT_DPI1) {
+		*addr = DISP_REG_CONFIG_DISP_RDMA2_SOUT;
+		value = RDMA2_SOUT_DPI1;
+	} else if (cur == DDP_COMPONENT_RDMA2 && next == DDP_COMPONENT_DSI1) {
+		*addr = DISP_REG_CONFIG_DISP_RDMA2_SOUT;
+		value = RDMA2_SOUT_DSI1;
+	} else if (cur == DDP_COMPONENT_RDMA2 && next == DDP_COMPONENT_DSI2) {
+		*addr = DISP_REG_CONFIG_DISP_RDMA2_SOUT;
+		value = RDMA2_SOUT_DSI2;
+	} else if (cur == DDP_COMPONENT_RDMA2 && next == DDP_COMPONENT_DSI3) {
+		*addr = DISP_REG_CONFIG_DISP_RDMA2_SOUT;
+		value = RDMA2_SOUT_DSI3;
 	} else {
 		value = 0;
 	}
@@ -172,6 +279,33 @@ static unsigned int mtk_ddp_sel_in(enum mtk_ddp_comp_id cur,
 	} else if (cur == DDP_COMPONENT_RDMA1 && next == DDP_COMPONENT_DPI0) {
 		*addr = DISP_REG_CONFIG_DPI_SEL_IN;
 		value = DPI0_SEL_IN_RDMA1;
+	} else if (cur == DDP_COMPONENT_RDMA1 && next == DDP_COMPONENT_DPI1) {
+		*addr = DISP_REG_CONFIG_DPI_SEL_IN;
+		value = DPI1_SEL_IN_RDMA1;
+	} else if (cur == DDP_COMPONENT_RDMA1 && next == DDP_COMPONENT_DSI1) {
+		*addr = DISP_REG_CONFIG_DSIO_SEL_IN;
+		value = DSI1_SEL_IN_RDMA1;
+	} else if (cur == DDP_COMPONENT_RDMA1 && next == DDP_COMPONENT_DSI2) {
+		*addr = DISP_REG_CONFIG_DSIE_SEL_IN;
+		value = DSI2_SEL_IN_RDMA1;
+	} else if (cur == DDP_COMPONENT_RDMA1 && next == DDP_COMPONENT_DSI3) {
+		*addr = DISP_REG_CONFIG_DSIO_SEL_IN;
+		value = DSI3_SEL_IN_RDMA1;
+	} else if (cur == DDP_COMPONENT_RDMA2 && next == DDP_COMPONENT_DPI0) {
+		*addr = DISP_REG_CONFIG_DPI_SEL_IN;
+		value = DPI0_SEL_IN_RDMA2;
+	} else if (cur == DDP_COMPONENT_RDMA2 && next == DDP_COMPONENT_DPI1) {
+		*addr = DISP_REG_CONFIG_DPI_SEL_IN;
+		value = DPI1_SEL_IN_RDMA2;
+	} else if (cur == DDP_COMPONENT_RDMA2 && next == DDP_COMPONENT_DSI1) {
+		*addr = DISP_REG_CONFIG_DSIE_SEL_IN;
+		value = DSI1_SEL_IN_RDMA2;
+	} else if (cur == DDP_COMPONENT_RDMA2 && next == DDP_COMPONENT_DSI2) {
+		*addr = DISP_REG_CONFIG_DSIE_SEL_IN;
+		value = DSI2_SEL_IN_RDMA2;
+	} else if (cur == DDP_COMPONENT_RDMA2 && next == DDP_COMPONENT_DSI3) {
+		*addr = DISP_REG_CONFIG_DSIE_SEL_IN;
+		value = DSI3_SEL_IN_RDMA2;
 	} else if (cur == DDP_COMPONENT_OVL1 && next == DDP_COMPONENT_COLOR1) {
 		*addr = DISP_REG_CONFIG_DISP_COLOR1_SEL_IN;
 		value = COLOR1_SEL_IN_OVL1;
@@ -278,6 +412,7 @@ void mtk_disp_mutex_add_comp(struct mtk_disp_mutex *mutex,
 	struct mtk_ddp *ddp = container_of(mutex, struct mtk_ddp,
 					   mutex[mutex->id]);
 	unsigned int reg;
+	unsigned int offset;
 
 	WARN_ON(&ddp->mutex[mutex->id] != mutex);
 
@@ -288,13 +423,30 @@ void mtk_disp_mutex_add_comp(struct mtk_disp_mutex *mutex,
 	case DDP_COMPONENT_DSI1:
 		reg = MUTEX_SOF_DSI0;
 		break;
+	case DDP_COMPONENT_DSI2:
+		reg = MUTEX_SOF_DSI2;
+		break;
+	case DDP_COMPONENT_DSI3:
+		reg = MUTEX_SOF_DSI3;
+		break;
 	case DDP_COMPONENT_DPI0:
 		reg = MUTEX_SOF_DPI0;
 		break;
+	case DDP_COMPONENT_DPI1:
+		reg = MUTEX_SOF_DPI1;
+		break;
 	default:
-		reg = readl_relaxed(ddp->regs + DISP_REG_MUTEX_MOD(mutex->id));
-		reg |= ddp->mutex_mod[id];
-		writel_relaxed(reg, ddp->regs + DISP_REG_MUTEX_MOD(mutex->id));
+		if (ddp->mutex_mod[id] < 32) {
+			offset = DISP_REG_MUTEX_MOD(mutex->id);
+			reg = readl_relaxed(ddp->regs + offset);
+			reg |= 1 << ddp->mutex_mod[id];
+			writel_relaxed(reg, ddp->regs + offset);
+		} else {
+			offset = DISP_REG_MUTEX_MOD2(mutex->id);
+			reg = readl_relaxed(ddp->regs + offset);
+			reg |= 1 << (ddp->mutex_mod[id] - 32);
+			writel_relaxed(reg, ddp->regs + offset);
+		}
 		return;
 	}
 
@@ -307,20 +459,32 @@ void mtk_disp_mutex_remove_comp(struct mtk_disp_mutex *mutex,
 	struct mtk_ddp *ddp = container_of(mutex, struct mtk_ddp,
 					   mutex[mutex->id]);
 	unsigned int reg;
+	unsigned int offset;
 
 	WARN_ON(&ddp->mutex[mutex->id] != mutex);
 
 	switch (id) {
 	case DDP_COMPONENT_DSI0:
 	case DDP_COMPONENT_DSI1:
+	case DDP_COMPONENT_DSI2:
+	case DDP_COMPONENT_DSI3:
 	case DDP_COMPONENT_DPI0:
+	case DDP_COMPONENT_DPI1:
 		writel_relaxed(MUTEX_SOF_SINGLE_MODE,
 			       ddp->regs + DISP_REG_MUTEX_SOF(mutex->id));
 		break;
 	default:
-		reg = readl_relaxed(ddp->regs + DISP_REG_MUTEX_MOD(mutex->id));
-		reg &= ~(ddp->mutex_mod[id]);
-		writel_relaxed(reg, ddp->regs + DISP_REG_MUTEX_MOD(mutex->id));
+		if (ddp->mutex_mod[id] < 32) {
+			offset = DISP_REG_MUTEX_MOD(mutex->id);
+			reg = readl_relaxed(ddp->regs + offset);
+			reg &= ~(1 << ddp->mutex_mod[id]);
+			writel_relaxed(reg, ddp->regs + offset);
+		} else {
+			offset = DISP_REG_MUTEX_MOD2(mutex->id);
+			reg = readl_relaxed(ddp->regs + offset);
+			reg &= ~(1 << (ddp->mutex_mod[id] - 32));
+			writel_relaxed(reg, ddp->regs + offset);
+		}
 		break;
 	}
 }
@@ -407,6 +571,7 @@ static int mtk_ddp_remove(struct platform_device *pdev)
 
 static const struct of_device_id ddp_driver_dt_match[] = {
 	{ .compatible = "mediatek,mt2701-disp-mutex", .data = mt2701_mutex_mod},
+	{ .compatible = "mediatek,mt2712-disp-mutex", .data = mt2712_mutex_mod},
 	{ .compatible = "mediatek,mt8173-disp-mutex", .data = mt8173_mutex_mod},
 	{},
 };
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
index 4672317e3ad1..ff974d82a4a6 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c
@@ -218,18 +218,25 @@ struct mtk_ddp_comp_match {
 };
 
 static const struct mtk_ddp_comp_match mtk_ddp_matches[DDP_COMPONENT_ID_MAX] = {
-	[DDP_COMPONENT_AAL]	= { MTK_DISP_AAL,	0, &ddp_aal },
+	[DDP_COMPONENT_AAL0]	= { MTK_DISP_AAL,	0, &ddp_aal },
+	[DDP_COMPONENT_AAL1]	= { MTK_DISP_AAL,	1, &ddp_aal },
 	[DDP_COMPONENT_BLS]	= { MTK_DISP_BLS,	0, NULL },
 	[DDP_COMPONENT_COLOR0]	= { MTK_DISP_COLOR,	0, NULL },
 	[DDP_COMPONENT_COLOR1]	= { MTK_DISP_COLOR,	1, NULL },
 	[DDP_COMPONENT_DPI0]	= { MTK_DPI,		0, NULL },
+	[DDP_COMPONENT_DPI1]	= { MTK_DPI,		1, NULL },
 	[DDP_COMPONENT_DSI0]	= { MTK_DSI,		0, NULL },
 	[DDP_COMPONENT_DSI1]	= { MTK_DSI,		1, NULL },
+	[DDP_COMPONENT_DSI2]	= { MTK_DSI,		2, NULL },
+	[DDP_COMPONENT_DSI3]	= { MTK_DSI,		3, NULL },
 	[DDP_COMPONENT_GAMMA]	= { MTK_DISP_GAMMA,	0, &ddp_gamma },
-	[DDP_COMPONENT_OD]	= { MTK_DISP_OD,	0, &ddp_od },
+	[DDP_COMPONENT_OD0]	= { MTK_DISP_OD,	0, &ddp_od },
+	[DDP_COMPONENT_OD1]	= { MTK_DISP_OD,	1, &ddp_od },
 	[DDP_COMPONENT_OVL0]	= { MTK_DISP_OVL,	0, NULL },
 	[DDP_COMPONENT_OVL1]	= { MTK_DISP_OVL,	1, NULL },
 	[DDP_COMPONENT_PWM0]	= { MTK_DISP_PWM,	0, NULL },
+	[DDP_COMPONENT_PWM1]	= { MTK_DISP_PWM,	1, NULL },
+	[DDP_COMPONENT_PWM2]	= { MTK_DISP_PWM,	2, NULL },
 	[DDP_COMPONENT_RDMA0]	= { MTK_DISP_RDMA,	0, NULL },
 	[DDP_COMPONENT_RDMA1]	= { MTK_DISP_RDMA,	1, NULL },
 	[DDP_COMPONENT_RDMA2]	= { MTK_DISP_RDMA,	2, NULL },
@@ -271,7 +278,11 @@ int mtk_ddp_comp_init(struct device *dev, struct device_node *node,
 
 	if (comp_id == DDP_COMPONENT_BLS ||
 	    comp_id == DDP_COMPONENT_DPI0 ||
+	    comp_id == DDP_COMPONENT_DPI1 ||
 	    comp_id == DDP_COMPONENT_DSI0 ||
+	    comp_id == DDP_COMPONENT_DSI1 ||
+	    comp_id == DDP_COMPONENT_DSI2 ||
+	    comp_id == DDP_COMPONENT_DSI3 ||
 	    comp_id == DDP_COMPONENT_PWM0) {
 		comp->regs = NULL;
 		comp->clk = NULL;
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.h b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.h
index 0828cf8bf85c..7413ffeb3c9d 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.h
+++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.h
@@ -41,19 +41,25 @@ enum mtk_ddp_comp_type {
 };
 
 enum mtk_ddp_comp_id {
-	DDP_COMPONENT_AAL,
+	DDP_COMPONENT_AAL0,
+	DDP_COMPONENT_AAL1,
 	DDP_COMPONENT_BLS,
 	DDP_COMPONENT_COLOR0,
 	DDP_COMPONENT_COLOR1,
 	DDP_COMPONENT_DPI0,
+	DDP_COMPONENT_DPI1,
 	DDP_COMPONENT_DSI0,
 	DDP_COMPONENT_DSI1,
+	DDP_COMPONENT_DSI2,
+	DDP_COMPONENT_DSI3,
 	DDP_COMPONENT_GAMMA,
-	DDP_COMPONENT_OD,
+	DDP_COMPONENT_OD0,
+	DDP_COMPONENT_OD1,
 	DDP_COMPONENT_OVL0,
 	DDP_COMPONENT_OVL1,
 	DDP_COMPONENT_PWM0,
 	DDP_COMPONENT_PWM1,
+	DDP_COMPONENT_PWM2,
 	DDP_COMPONENT_RDMA0,
 	DDP_COMPONENT_RDMA1,
 	DDP_COMPONENT_RDMA2,
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_drv.c b/drivers/gpu/drm/mediatek/mtk_drm_drv.c
index a2ca90fc403c..39721119713b 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_drv.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_drv.c
@@ -146,11 +146,37 @@ static const enum mtk_ddp_comp_id mt2701_mtk_ddp_ext[] = {
 	DDP_COMPONENT_DPI0,
 };
 
+static const enum mtk_ddp_comp_id mt2712_mtk_ddp_main[] = {
+	DDP_COMPONENT_OVL0,
+	DDP_COMPONENT_COLOR0,
+	DDP_COMPONENT_AAL0,
+	DDP_COMPONENT_OD0,
+	DDP_COMPONENT_RDMA0,
+	DDP_COMPONENT_DPI0,
+	DDP_COMPONENT_PWM0,
+};
+
+static const enum mtk_ddp_comp_id mt2712_mtk_ddp_ext[] = {
+	DDP_COMPONENT_OVL1,
+	DDP_COMPONENT_COLOR1,
+	DDP_COMPONENT_AAL1,
+	DDP_COMPONENT_OD1,
+	DDP_COMPONENT_RDMA1,
+	DDP_COMPONENT_DPI1,
+	DDP_COMPONENT_PWM1,
+};
+
+static const enum mtk_ddp_comp_id mt2712_mtk_ddp_third[] = {
+	DDP_COMPONENT_RDMA2,
+	DDP_COMPONENT_DSI3,
+	DDP_COMPONENT_PWM2,
+};
+
 static const enum mtk_ddp_comp_id mt8173_mtk_ddp_main[] = {
 	DDP_COMPONENT_OVL0,
 	DDP_COMPONENT_COLOR0,
-	DDP_COMPONENT_AAL,
-	DDP_COMPONENT_OD,
+	DDP_COMPONENT_AAL0,
+	DDP_COMPONENT_OD0,
 	DDP_COMPONENT_RDMA0,
 	DDP_COMPONENT_UFOE,
 	DDP_COMPONENT_DSI0,
@@ -173,6 +199,15 @@ static const struct mtk_mmsys_driver_data mt2701_mmsys_driver_data = {
 	.shadow_register = true,
 };
 
+static const struct mtk_mmsys_driver_data mt2712_mmsys_driver_data = {
+	.main_path = mt2712_mtk_ddp_main,
+	.main_len = ARRAY_SIZE(mt2712_mtk_ddp_main),
+	.ext_path = mt2712_mtk_ddp_ext,
+	.ext_len = ARRAY_SIZE(mt2712_mtk_ddp_ext),
+	.third_path = mt2712_mtk_ddp_third,
+	.third_len = ARRAY_SIZE(mt2712_mtk_ddp_third),
+};
+
 static const struct mtk_mmsys_driver_data mt8173_mmsys_driver_data = {
 	.main_path = mt8173_mtk_ddp_main,
 	.main_len = ARRAY_SIZE(mt8173_mtk_ddp_main),
@@ -232,6 +267,11 @@ static int mtk_drm_kms_init(struct drm_device *drm)
 	if (ret < 0)
 		goto err_component_unbind;
 
+	ret = mtk_drm_crtc_create(drm, private->data->third_path,
+				  private->data->third_len);
+	if (ret < 0)
+		goto err_component_unbind;
+
 	/* Use OVL device for all DMA memory allocations */
 	np = private->comp_node[private->data->main_path[0]] ?:
 	     private->comp_node[private->data->ext_path[0]];
@@ -360,24 +400,44 @@ static const struct component_master_ops mtk_drm_ops = {
 };
 
 static const struct of_device_id mtk_ddp_comp_dt_ids[] = {
-	{ .compatible = "mediatek,mt2701-disp-ovl",   .data = (void *)MTK_DISP_OVL },
-	{ .compatible = "mediatek,mt8173-disp-ovl",   .data = (void *)MTK_DISP_OVL },
-	{ .compatible = "mediatek,mt2701-disp-rdma",  .data = (void *)MTK_DISP_RDMA },
-	{ .compatible = "mediatek,mt8173-disp-rdma",  .data = (void *)MTK_DISP_RDMA },
-	{ .compatible = "mediatek,mt8173-disp-wdma",  .data = (void *)MTK_DISP_WDMA },
-	{ .compatible = "mediatek,mt2701-disp-color", .data = (void *)MTK_DISP_COLOR },
-	{ .compatible = "mediatek,mt8173-disp-color", .data = (void *)MTK_DISP_COLOR },
-	{ .compatible = "mediatek,mt8173-disp-aal",   .data = (void *)MTK_DISP_AAL},
-	{ .compatible = "mediatek,mt8173-disp-gamma", .data = (void *)MTK_DISP_GAMMA, },
-	{ .compatible = "mediatek,mt8173-disp-ufoe",  .data = (void *)MTK_DISP_UFOE },
-	{ .compatible = "mediatek,mt2701-dsi",	      .data = (void *)MTK_DSI },
-	{ .compatible = "mediatek,mt8173-dsi",        .data = (void *)MTK_DSI },
-	{ .compatible = "mediatek,mt8173-dpi",        .data = (void *)MTK_DPI },
-	{ .compatible = "mediatek,mt2701-disp-mutex", .data = (void *)MTK_DISP_MUTEX },
-	{ .compatible = "mediatek,mt8173-disp-mutex", .data = (void *)MTK_DISP_MUTEX },
-	{ .compatible = "mediatek,mt2701-disp-pwm",   .data = (void *)MTK_DISP_BLS },
-	{ .compatible = "mediatek,mt8173-disp-pwm",   .data = (void *)MTK_DISP_PWM },
-	{ .compatible = "mediatek,mt8173-disp-od",    .data = (void *)MTK_DISP_OD },
+	{ .compatible = "mediatek,mt2701-disp-ovl",
+	  .data = (void *)MTK_DISP_OVL },
+	{ .compatible = "mediatek,mt8173-disp-ovl",
+	  .data = (void *)MTK_DISP_OVL },
+	{ .compatible = "mediatek,mt2701-disp-rdma",
+	  .data = (void *)MTK_DISP_RDMA },
+	{ .compatible = "mediatek,mt8173-disp-rdma",
+	  .data = (void *)MTK_DISP_RDMA },
+	{ .compatible = "mediatek,mt8173-disp-wdma",
+	  .data = (void *)MTK_DISP_WDMA },
+	{ .compatible = "mediatek,mt2701-disp-color",
+	  .data = (void *)MTK_DISP_COLOR },
+	{ .compatible = "mediatek,mt8173-disp-color",
+	  .data = (void *)MTK_DISP_COLOR },
+	{ .compatible = "mediatek,mt8173-disp-aal",
+	  .data = (void *)MTK_DISP_AAL},
+	{ .compatible = "mediatek,mt8173-disp-gamma",
+	  .data = (void *)MTK_DISP_GAMMA, },
+	{ .compatible = "mediatek,mt8173-disp-ufoe",
+	  .data = (void *)MTK_DISP_UFOE },
+	{ .compatible = "mediatek,mt2701-dsi",
+	  .data = (void *)MTK_DSI },
+	{ .compatible = "mediatek,mt8173-dsi",
+	  .data = (void *)MTK_DSI },
+	{ .compatible = "mediatek,mt8173-dpi",
+	  .data = (void *)MTK_DPI },
+	{ .compatible = "mediatek,mt2701-disp-mutex",
+	  .data = (void *)MTK_DISP_MUTEX },
+	{ .compatible = "mediatek,mt2712-disp-mutex",
+	  .data = (void *)MTK_DISP_MUTEX },
+	{ .compatible = "mediatek,mt8173-disp-mutex",
+	  .data = (void *)MTK_DISP_MUTEX },
+	{ .compatible = "mediatek,mt2701-disp-pwm",
+	  .data = (void *)MTK_DISP_BLS },
+	{ .compatible = "mediatek,mt8173-disp-pwm",
+	  .data = (void *)MTK_DISP_PWM },
+	{ .compatible = "mediatek,mt8173-disp-od",
+	  .data = (void *)MTK_DISP_OD },
 	{ }
 };
 
@@ -552,6 +612,8 @@ static SIMPLE_DEV_PM_OPS(mtk_drm_pm_ops, mtk_drm_sys_suspend,
 static const struct of_device_id mtk_drm_of_ids[] = {
 	{ .compatible = "mediatek,mt2701-mmsys",
 	  .data = &mt2701_mmsys_driver_data},
+	{ .compatible = "mediatek,mt2712-mmsys",
+	  .data = &mt2712_mmsys_driver_data},
 	{ .compatible = "mediatek,mt8173-mmsys",
 	  .data = &mt8173_mmsys_driver_data},
 	{ }
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_drv.h b/drivers/gpu/drm/mediatek/mtk_drm_drv.h
index c3378c452c0a..ecc00ca3221d 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_drv.h
+++ b/drivers/gpu/drm/mediatek/mtk_drm_drv.h
@@ -17,7 +17,7 @@
 #include <linux/io.h>
 #include "mtk_drm_ddp_comp.h"
 
-#define MAX_CRTC	2
+#define MAX_CRTC	3
 #define MAX_CONNECTOR	2
 
 struct device;
@@ -33,6 +33,9 @@ struct mtk_mmsys_driver_data {
 	unsigned int main_len;
 	const enum mtk_ddp_comp_id *ext_path;
 	unsigned int ext_len;
+	const enum mtk_ddp_comp_id *third_path;
+	unsigned int third_len;
+
 	bool shadow_register;
 };
 
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_fb.c b/drivers/gpu/drm/mediatek/mtk_drm_fb.c
index 0d8d506695f9..be5f6f1daf55 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_fb.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_fb.c
@@ -15,6 +15,7 @@
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_gem.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 #include <linux/dma-buf.h>
 #include <linux/reservation.h>
 
@@ -22,78 +23,37 @@
 #include "mtk_drm_fb.h"
 #include "mtk_drm_gem.h"
 
-/*
- * mtk specific framebuffer structure.
- *
- * @fb: drm framebuffer object.
- * @gem_obj: array of gem objects.
- */
-struct mtk_drm_fb {
-	struct drm_framebuffer	base;
-	/* For now we only support a single plane */
-	struct drm_gem_object	*gem_obj;
-};
-
-#define to_mtk_fb(x) container_of(x, struct mtk_drm_fb, base)
-
-struct drm_gem_object *mtk_fb_get_gem_obj(struct drm_framebuffer *fb)
-{
-	struct mtk_drm_fb *mtk_fb = to_mtk_fb(fb);
-
-	return mtk_fb->gem_obj;
-}
-
-static int mtk_drm_fb_create_handle(struct drm_framebuffer *fb,
-				    struct drm_file *file_priv,
-				    unsigned int *handle)
-{
-	struct mtk_drm_fb *mtk_fb = to_mtk_fb(fb);
-
-	return drm_gem_handle_create(file_priv, mtk_fb->gem_obj, handle);
-}
-
-static void mtk_drm_fb_destroy(struct drm_framebuffer *fb)
-{
-	struct mtk_drm_fb *mtk_fb = to_mtk_fb(fb);
-
-	drm_framebuffer_cleanup(fb);
-
-	drm_gem_object_put_unlocked(mtk_fb->gem_obj);
-
-	kfree(mtk_fb);
-}
-
 static const struct drm_framebuffer_funcs mtk_drm_fb_funcs = {
-	.create_handle = mtk_drm_fb_create_handle,
-	.destroy = mtk_drm_fb_destroy,
+	.create_handle = drm_gem_fb_create_handle,
+	.destroy = drm_gem_fb_destroy,
 };
 
-static struct mtk_drm_fb *mtk_drm_framebuffer_init(struct drm_device *dev,
+static struct drm_framebuffer *mtk_drm_framebuffer_init(struct drm_device *dev,
 					const struct drm_mode_fb_cmd2 *mode,
 					struct drm_gem_object *obj)
 {
-	struct mtk_drm_fb *mtk_fb;
+	struct drm_framebuffer *fb;
 	int ret;
 
 	if (drm_format_num_planes(mode->pixel_format) != 1)
 		return ERR_PTR(-EINVAL);
 
-	mtk_fb = kzalloc(sizeof(*mtk_fb), GFP_KERNEL);
-	if (!mtk_fb)
+	fb = kzalloc(sizeof(*fb), GFP_KERNEL);
+	if (!fb)
 		return ERR_PTR(-ENOMEM);
 
-	drm_helper_mode_fill_fb_struct(dev, &mtk_fb->base, mode);
+	drm_helper_mode_fill_fb_struct(dev, fb, mode);
 
-	mtk_fb->gem_obj = obj;
+	fb->obj[0] = obj;
 
-	ret = drm_framebuffer_init(dev, &mtk_fb->base, &mtk_drm_fb_funcs);
+	ret = drm_framebuffer_init(dev, fb, &mtk_drm_fb_funcs);
 	if (ret) {
 		DRM_ERROR("failed to initialize framebuffer\n");
-		kfree(mtk_fb);
+		kfree(fb);
 		return ERR_PTR(ret);
 	}
 
-	return mtk_fb;
+	return fb;
 }
 
 /*
@@ -110,7 +70,7 @@ int mtk_fb_wait(struct drm_framebuffer *fb)
 	if (!fb)
 		return 0;
 
-	gem = mtk_fb_get_gem_obj(fb);
+	gem = fb->obj[0];
 	if (!gem || !gem->dma_buf || !gem->dma_buf->resv)
 		return 0;
 
@@ -128,7 +88,7 @@ struct drm_framebuffer *mtk_drm_mode_fb_create(struct drm_device *dev,
 					       struct drm_file *file,
 					       const struct drm_mode_fb_cmd2 *cmd)
 {
-	struct mtk_drm_fb *mtk_fb;
+	struct drm_framebuffer *fb;
 	struct drm_gem_object *gem;
 	unsigned int width = cmd->width;
 	unsigned int height = cmd->height;
@@ -151,13 +111,13 @@ struct drm_framebuffer *mtk_drm_mode_fb_create(struct drm_device *dev,
 		goto unreference;
 	}
 
-	mtk_fb = mtk_drm_framebuffer_init(dev, cmd, gem);
-	if (IS_ERR(mtk_fb)) {
-		ret = PTR_ERR(mtk_fb);
+	fb = mtk_drm_framebuffer_init(dev, cmd, gem);
+	if (IS_ERR(fb)) {
+		ret = PTR_ERR(fb);
 		goto unreference;
 	}
 
-	return &mtk_fb->base;
+	return fb;
 
 unreference:
 	drm_gem_object_put_unlocked(gem);
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_fb.h b/drivers/gpu/drm/mediatek/mtk_drm_fb.h
index 9b2ae345a4e9..7f976b196a15 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_fb.h
+++ b/drivers/gpu/drm/mediatek/mtk_drm_fb.h
@@ -14,7 +14,6 @@
 #ifndef MTK_DRM_FB_H
 #define MTK_DRM_FB_H
 
-struct drm_gem_object *mtk_fb_get_gem_obj(struct drm_framebuffer *fb);
 int mtk_fb_wait(struct drm_framebuffer *fb);
 struct drm_framebuffer *mtk_drm_mode_fb_create(struct drm_device *dev,
 					       struct drm_file *file,
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_gem.c b/drivers/gpu/drm/mediatek/mtk_drm_gem.c
index f595ac816b55..259b7b0de1d2 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_gem.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_gem.c
@@ -220,7 +220,7 @@ struct drm_gem_object *mtk_gem_prime_import_sg_table(struct drm_device *dev,
 	mtk_gem = mtk_drm_gem_init(dev, attach->dmabuf->size);
 
 	if (IS_ERR(mtk_gem))
-		return ERR_PTR(PTR_ERR(mtk_gem));
+		return ERR_CAST(mtk_gem);
 
 	expected = sg_dma_address(sg->sgl);
 	for_each_sg(sg->sgl, s, sg->nents, i) {
diff --git a/drivers/gpu/drm/mediatek/mtk_drm_plane.c b/drivers/gpu/drm/mediatek/mtk_drm_plane.c
index 2f4b0ffee598..f7e6aa1b5b7d 100644
--- a/drivers/gpu/drm/mediatek/mtk_drm_plane.c
+++ b/drivers/gpu/drm/mediatek/mtk_drm_plane.c
@@ -95,11 +95,6 @@ static int mtk_plane_atomic_check(struct drm_plane *plane,
 	if (!fb)
 		return 0;
 
-	if (!mtk_fb_get_gem_obj(fb)) {
-		DRM_DEBUG_KMS("buffer is null\n");
-		return -EFAULT;
-	}
-
 	if (!state->crtc)
 		return 0;
 
@@ -127,7 +122,7 @@ static void mtk_plane_atomic_update(struct drm_plane *plane,
 	if (!crtc || WARN_ON(!fb))
 		return;
 
-	gem = mtk_fb_get_gem_obj(fb);
+	gem = fb->obj[0];
 	mtk_gem = to_mtk_gem_obj(gem);
 	addr = mtk_gem->dma_addr;
 	pitch = fb->pitches[0];
diff --git a/drivers/gpu/drm/mediatek/mtk_dsi.c b/drivers/gpu/drm/mediatek/mtk_dsi.c
index 7e5e24c2152a..66df1b177959 100644
--- a/drivers/gpu/drm/mediatek/mtk_dsi.c
+++ b/drivers/gpu/drm/mediatek/mtk_dsi.c
@@ -551,13 +551,12 @@ static int mtk_dsi_poweron(struct mtk_dsi *dsi)
 	}
 
 	/**
-	 * vm.pixelclock is in kHz, pixel_clock unit is Hz, so multiply by 1000
 	 * htotal_time = htotal * byte_per_pixel / num_lanes
 	 * overhead_time = lpx + hs_prepare + hs_zero + hs_trail + hs_exit
 	 * mipi_ratio = (htotal_time + overhead_time) / htotal_time
 	 * data_rate = pixel_clock * bit_per_pixel * mipi_ratio / num_lanes;
 	 */
-	pixel_clock = dsi->vm.pixelclock * 1000;
+	pixel_clock = dsi->vm.pixelclock;
 	htotal = dsi->vm.hactive + dsi->vm.hback_porch + dsi->vm.hfront_porch +
 			dsi->vm.hsync_len;
 	htotal_bits = htotal * bit_per_pixel;
@@ -725,16 +724,7 @@ static void mtk_dsi_encoder_mode_set(struct drm_encoder *encoder,
 {
 	struct mtk_dsi *dsi = encoder_to_dsi(encoder);
 
-	dsi->vm.pixelclock = adjusted->clock;
-	dsi->vm.hactive = adjusted->hdisplay;
-	dsi->vm.hback_porch = adjusted->htotal - adjusted->hsync_end;
-	dsi->vm.hfront_porch = adjusted->hsync_start - adjusted->hdisplay;
-	dsi->vm.hsync_len = adjusted->hsync_end - adjusted->hsync_start;
-
-	dsi->vm.vactive = adjusted->vdisplay;
-	dsi->vm.vback_porch = adjusted->vtotal - adjusted->vsync_end;
-	dsi->vm.vfront_porch = adjusted->vsync_start - adjusted->vdisplay;
-	dsi->vm.vsync_len = adjusted->vsync_end - adjusted->vsync_start;
+	drm_display_mode_to_videomode(adjusted, &dsi->vm);
 }
 
 static void mtk_dsi_encoder_disable(struct drm_encoder *encoder)
@@ -792,7 +782,7 @@ static int mtk_dsi_create_connector(struct drm_device *drm, struct mtk_dsi *dsi)
 	drm_connector_helper_add(&dsi->conn, &mtk_dsi_connector_helper_funcs);
 
 	dsi->conn.dpms = DRM_MODE_DPMS_OFF;
-	drm_mode_connector_attach_encoder(&dsi->conn, &dsi->encoder);
+	drm_connector_attach_encoder(&dsi->conn, &dsi->encoder);
 
 	if (dsi->panel) {
 		ret = drm_panel_attach(dsi->panel, &dsi->conn);
diff --git a/drivers/gpu/drm/mediatek/mtk_hdmi.c b/drivers/gpu/drm/mediatek/mtk_hdmi.c
index 59a11026dceb..2d45d1dd9554 100644
--- a/drivers/gpu/drm/mediatek/mtk_hdmi.c
+++ b/drivers/gpu/drm/mediatek/mtk_hdmi.c
@@ -1220,7 +1220,7 @@ static int mtk_hdmi_conn_get_modes(struct drm_connector *conn)
 
 	hdmi->dvi_mode = !drm_detect_monitor_audio(edid);
 
-	drm_mode_connector_update_edid_property(conn, edid);
+	drm_connector_update_edid_property(conn, edid);
 
 	ret = drm_add_edid_modes(conn, edid);
 	kfree(edid);
@@ -1306,7 +1306,7 @@ static int mtk_hdmi_bridge_attach(struct drm_bridge *bridge)
 	hdmi->conn.interlace_allowed = true;
 	hdmi->conn.doublescan_allowed = false;
 
-	ret = drm_mode_connector_attach_encoder(&hdmi->conn,
+	ret = drm_connector_attach_encoder(&hdmi->conn,
 						bridge->encoder);
 	if (ret) {
 		dev_err(hdmi->dev,
diff --git a/drivers/gpu/drm/meson/meson_drv.c b/drivers/gpu/drm/meson/meson_drv.c
index 32b1a6cdecfc..d3443125e661 100644
--- a/drivers/gpu/drm/meson/meson_drv.c
+++ b/drivers/gpu/drm/meson/meson_drv.c
@@ -197,8 +197,10 @@ static int meson_drv_bind_master(struct device *dev, bool has_components)
 	priv->io_base = regs;
 
 	res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "hhi");
-	if (!res)
-		return -EINVAL;
+	if (!res) {
+		ret = -EINVAL;
+		goto free_drm;
+	}
 	/* Simply ioremap since it may be a shared register zone */
 	regs = devm_ioremap(dev, res->start, resource_size(res));
 	if (!regs) {
@@ -215,8 +217,10 @@ static int meson_drv_bind_master(struct device *dev, bool has_components)
 	}
 
 	res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "dmc");
-	if (!res)
-		return -EINVAL;
+	if (!res) {
+		ret = -EINVAL;
+		goto free_drm;
+	}
 	/* Simply ioremap since it may be a shared register zone */
 	regs = devm_ioremap(dev, res->start, resource_size(res));
 	if (!regs) {
diff --git a/drivers/gpu/drm/meson/meson_dw_hdmi.c b/drivers/gpu/drm/meson/meson_dw_hdmi.c
index a393095aac1a..df7247cd93f9 100644
--- a/drivers/gpu/drm/meson/meson_dw_hdmi.c
+++ b/drivers/gpu/drm/meson/meson_dw_hdmi.c
@@ -329,6 +329,12 @@ static void dw_hdmi_set_vclk(struct meson_dw_hdmi *dw_hdmi,
 
 	vclk_freq = mode->clock;
 
+	if (!vic) {
+		meson_vclk_setup(priv, MESON_VCLK_TARGET_DMT, vclk_freq,
+				 vclk_freq, vclk_freq, false);
+		return;
+	}
+
 	if (mode->flags & DRM_MODE_FLAG_DBLCLK)
 		vclk_freq *= 2;
 
@@ -529,7 +535,7 @@ static irqreturn_t dw_hdmi_top_thread_irq(int irq, void *dev_id)
 		if (stat & HDMITX_TOP_INTR_HPD_RISE)
 			hpd_connected = true;
 
-		dw_hdmi_setup_rx_sense(dw_hdmi->dev, hpd_connected,
+		dw_hdmi_setup_rx_sense(dw_hdmi->hdmi, hpd_connected,
 				       hpd_connected);
 
 		drm_helper_hpd_irq_event(dw_hdmi->encoder.dev);
@@ -542,10 +548,12 @@ static enum drm_mode_status
 dw_hdmi_mode_valid(struct drm_connector *connector,
 		   const struct drm_display_mode *mode)
 {
+	struct meson_drm *priv = connector->dev->dev_private;
 	unsigned int vclk_freq;
 	unsigned int venc_freq;
 	unsigned int hdmi_freq;
 	int vic = drm_match_cea_mode(mode);
+	enum drm_mode_status status;
 
 	DRM_DEBUG_DRIVER("Modeline %d:\"%s\" %d %d %d %d %d %d %d %d %d %d 0x%x 0x%x\n",
 		mode->base.id, mode->name, mode->vrefresh, mode->clock,
@@ -556,8 +564,11 @@ dw_hdmi_mode_valid(struct drm_connector *connector,
 
 	/* Check against non-VIC supported modes */
 	if (!vic) {
-		if (!meson_venc_hdmi_supported_mode(mode))
-			return MODE_BAD;
+		status = meson_venc_hdmi_supported_mode(mode);
+		if (status != MODE_OK)
+			return status;
+
+		return meson_vclk_dmt_supported_freq(priv, mode->clock);
 	/* Check against supported VIC modes */
 	} else if (!meson_venc_hdmi_supported_vic(vic))
 		return MODE_BAD;
@@ -583,16 +594,11 @@ dw_hdmi_mode_valid(struct drm_connector *connector,
 	dev_dbg(connector->dev->dev, "%s: vclk:%d venc=%d hdmi=%d\n", __func__,
 		vclk_freq, venc_freq, hdmi_freq);
 
-	/* Finally filter by configurable vclk frequencies */
+	/* Finally filter by configurable vclk frequencies for VIC modes */
 	switch (vclk_freq) {
-	case 25175:
-	case 40000:
 	case 54000:
-	case 65000:
 	case 74250:
-	case 108000:
 	case 148500:
-	case 162000:
 	case 297000:
 	case 594000:
 		return MODE_OK;
diff --git a/drivers/gpu/drm/meson/meson_vclk.c b/drivers/gpu/drm/meson/meson_vclk.c
index f0511220317f..ae5473257f72 100644
--- a/drivers/gpu/drm/meson/meson_vclk.c
+++ b/drivers/gpu/drm/meson/meson_vclk.c
@@ -320,32 +320,23 @@ static void meson_venci_cvbs_clock_config(struct meson_drm *priv)
 				CTS_VDAC_EN, CTS_VDAC_EN);
 }
 
-
+enum {
 /* PLL	O1 O2 O3 VP DV     EN TX */
 /* 4320 /4 /4 /1 /5 /1  => /2 /2 */
-#define MESON_VCLK_HDMI_ENCI_54000	1
+	MESON_VCLK_HDMI_ENCI_54000 = 1,
 /* 4320 /4 /4 /1 /5 /1  => /1 /2 */
-#define MESON_VCLK_HDMI_DDR_54000	2
+	MESON_VCLK_HDMI_DDR_54000,
 /* 2970 /4 /1 /1 /5 /1  => /1 /2 */
-#define MESON_VCLK_HDMI_DDR_148500	3
-/* 4028 /4 /4 /1 /5 /2  => /1 /1 */
-#define MESON_VCLK_HDMI_25175		4
-/* 3200 /4 /2 /1 /5 /2  => /1 /1 */
-#define MESON_VCLK_HDMI_40000		5
-/* 5200 /4 /2 /1 /5 /2  => /1 /1 */
-#define MESON_VCLK_HDMI_65000		6
+	MESON_VCLK_HDMI_DDR_148500,
 /* 2970 /2 /2 /2 /5 /1  => /1 /1 */
-#define MESON_VCLK_HDMI_74250		7
-/* 4320 /4 /1 /1 /5 /2  => /1 /1 */
-#define MESON_VCLK_HDMI_108000		8
+	MESON_VCLK_HDMI_74250,
 /* 2970 /1 /2 /2 /5 /1  => /1 /1 */
-#define MESON_VCLK_HDMI_148500		9
-/* 3240 /2 /1 /1 /5 /2  => /1 /1 */
-#define MESON_VCLK_HDMI_162000		10
+	MESON_VCLK_HDMI_148500,
 /* 2970 /1 /1 /1 /5 /2  => /1 /1 */
-#define MESON_VCLK_HDMI_297000		11
+	MESON_VCLK_HDMI_297000,
 /* 5940 /1 /1 /2 /5 /1  => /1 /1 */
-#define MESON_VCLK_HDMI_594000		12
+	MESON_VCLK_HDMI_594000
+};
 
 struct meson_vclk_params {
 	unsigned int pll_base_freq;
@@ -411,46 +402,6 @@ struct meson_vclk_params {
 		.vid_pll_div = VID_PLL_DIV_5,
 		.vclk_div = 1,
 	},
-	[MESON_VCLK_HDMI_25175] = {
-		.pll_base_freq = 4028000,
-		.pll_od1 = 4,
-		.pll_od2 = 4,
-		.pll_od3 = 1,
-		.vid_pll_div = VID_PLL_DIV_5,
-		.vclk_div = 2,
-	},
-	[MESON_VCLK_HDMI_40000] = {
-		.pll_base_freq = 3200000,
-		.pll_od1 = 4,
-		.pll_od2 = 2,
-		.pll_od3 = 1,
-		.vid_pll_div = VID_PLL_DIV_5,
-		.vclk_div = 2,
-	},
-	[MESON_VCLK_HDMI_65000] = {
-		.pll_base_freq = 5200000,
-		.pll_od1 = 4,
-		.pll_od2 = 2,
-		.pll_od3 = 1,
-		.vid_pll_div = VID_PLL_DIV_5,
-		.vclk_div = 2,
-	},
-	[MESON_VCLK_HDMI_108000] = {
-		.pll_base_freq = 4320000,
-		.pll_od1 = 4,
-		.pll_od2 = 1,
-		.pll_od3 = 1,
-		.vid_pll_div = VID_PLL_DIV_5,
-		.vclk_div = 2,
-	},
-	[MESON_VCLK_HDMI_162000] = {
-		.pll_base_freq = 3240000,
-		.pll_od1 = 2,
-		.pll_od2 = 1,
-		.pll_od3 = 1,
-		.vid_pll_div = VID_PLL_DIV_5,
-		.vclk_div = 2,
-	},
 };
 
 static inline unsigned int pll_od_to_reg(unsigned int od)
@@ -470,358 +421,217 @@ static inline unsigned int pll_od_to_reg(unsigned int od)
 	return 0;
 }
 
-void meson_hdmi_pll_set(struct meson_drm *priv,
-			unsigned int base,
-			unsigned int od1,
-			unsigned int od2,
-			unsigned int od3)
+void meson_hdmi_pll_set_params(struct meson_drm *priv, unsigned int m,
+			       unsigned int frac, unsigned int od1,
+			       unsigned int od2, unsigned int od3)
 {
 	unsigned int val;
 
 	if (meson_vpu_is_compatible(priv, "amlogic,meson-gxbb-vpu")) {
-		switch (base) {
-		case 2970000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x5800023d);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x00000000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x0d5c5091);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x801da72c);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x71486980);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x00000e55);
-
-			/* Enable and unreset */
-			regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL,
-						0x7 << 28, 0x4 << 28);
-
-			/* Poll for lock bit */
-			regmap_read_poll_timeout(priv->hhi, HHI_HDMI_PLL_CNTL,
-					val, (val & HDMI_PLL_LOCK), 10, 0);
-
-			/* div_frac */
-			regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL2,
-						0xFFFF,  0x4e00);
-			break;
-
-		case 3200000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x58000242);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x00000000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x0d5c5091);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x801da72c);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x71486980);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x00000e55);
-
-			/* unreset */
-			regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL,
-						BIT(28), 0);
-
-			/* Poll for lock bit */
-			regmap_read_poll_timeout(priv->hhi, HHI_HDMI_PLL_CNTL,
-					val, (val & HDMI_PLL_LOCK), 10, 0);
-
-			/* div_frac */
-			regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL2,
-						0xFFFF,  0x4aab);
-			break;
-
-		case 3240000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x58000243);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x00000000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x0d5c5091);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x801da72c);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x71486980);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x00000e55);
-
-			/* unreset */
-			regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL,
-						BIT(28), 0);
-
-			/* Poll for lock bit */
-			regmap_read_poll_timeout(priv->hhi, HHI_HDMI_PLL_CNTL,
-					val, (val & HDMI_PLL_LOCK), 10, 0);
-
-			/* div_frac */
-			regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL2,
-						0xFFFF,  0x4800);
-			break;
-
-		case 3865000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x58000250);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x00000000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x0d5c5091);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x801da72c);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x71486980);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x00000e55);
-
-			/* unreset */
-			regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL,
-						BIT(28), 0);
-
-			/* Poll for lock bit */
-			regmap_read_poll_timeout(priv->hhi, HHI_HDMI_PLL_CNTL,
-					val, (val & HDMI_PLL_LOCK), 10, 0);
-
-			/* div_frac */
-			regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL2,
-						0xFFFF,  0x4855);
-			break;
-
-		case 4028000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x58000253);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x00000000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x0d5c5091);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x801da72c);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x71486980);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x00000e55);
-
-			/* unreset */
-			regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL,
-						BIT(28), 0);
-
-			/* Poll for lock bit */
-			regmap_read_poll_timeout(priv->hhi, HHI_HDMI_PLL_CNTL,
-					val, (val & HDMI_PLL_LOCK), 10, 0);
-
-			/* div_frac */
-			regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL2,
-						0xFFFF,  0x4eab);
-			break;
-
-		case 4320000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x5800025a);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x00000000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x0d5c5091);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x801da72c);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x71486980);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x00000e55);
-
-			/* unreset */
-			regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL,
-						BIT(28), 0);
-
-			/* Poll for lock bit */
-			regmap_read_poll_timeout(priv->hhi, HHI_HDMI_PLL_CNTL,
-					val, (val & HDMI_PLL_LOCK), 10, 0);
-			break;
+		regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x58000200 | m);
+		if (frac)
+			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2,
+				     0x00004000 | frac);
+		else
+			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2,
+				     0x00000000);
+		regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x0d5c5091);
+		regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x801da72c);
+		regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x71486980);
+		regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x00000e55);
 
-		case 5940000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x5800027b);
-			regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL2,
-						0xFFFF,  0x4c00);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x135c5091);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x801da72c);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x71486980);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x00000e55);
-
-			/* unreset */
-			regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL,
-						BIT(28), 0);
-
-			/* Poll for lock bit */
-			regmap_read_poll_timeout(priv->hhi, HHI_HDMI_PLL_CNTL,
-					val, (val & HDMI_PLL_LOCK), 10, 0);
-			break;
+		/* Enable and unreset */
+		regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL,
+				   0x7 << 28, 0x4 << 28);
 
-		case 5200000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x5800026c);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x00000000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x135c5091);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x801da72c);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x71486980);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x00000e55);
-
-			/* unreset */
-			regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL,
-						BIT(28), 0);
-
-			/* Poll for lock bit */
-			regmap_read_poll_timeout(priv->hhi, HHI_HDMI_PLL_CNTL,
-					val, (val & HDMI_PLL_LOCK), 10, 0);
-			break;
-		};
+		/* Poll for lock bit */
+		regmap_read_poll_timeout(priv->hhi, HHI_HDMI_PLL_CNTL,
+					 val, (val & HDMI_PLL_LOCK), 10, 0);
 	} else if (meson_vpu_is_compatible(priv, "amlogic,meson-gxm-vpu") ||
 		   meson_vpu_is_compatible(priv, "amlogic,meson-gxl-vpu")) {
-		switch (base) {
-		case 2970000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x4000027b);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x800cb300);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x860f30c4);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x0c8e0000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x001fa729);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x01a31500);
-			break;
-
-		case 3200000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x40000285);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x800cb155);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x860f30c4);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x0c8e0000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x001fa729);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x01a31500);
-			break;
-
-		case 3240000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x40000287);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x800cb000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x860f30c4);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x0c8e0000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x001fa729);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x01a31500);
-			break;
-
-		case 3865000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x400002a1);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x800cb02b);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x860f30c4);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x0c8e0000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x001fa729);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x01a31500);
-			break;
-
-		case 4028000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x400002a7);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x800cb355);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x860f30c4);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x0c8e0000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x001fa729);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x01a31500);
-			break;
-
-		case 4320000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x400002b4);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x800cb000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x860f30c4);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x0c8e0000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x001fa729);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x01a31500);
-			break;
-
-		case 5940000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x400002f7);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x800cb200);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x860f30c4);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x0c8e0000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x001fa729);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x01a31500);
-			break;
-
-		case 5200000:
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x400002d8);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x800cb2ab);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x860f30c4);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x0c8e0000);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x001fa729);
-			regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x01a31500);
-			break;
-
-		};
+		regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL, 0x40000200 | m);
+		regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL2, 0x800cb000 | frac);
+		regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL3, 0x860f30c4);
+		regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL4, 0x0c8e0000);
+		regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL5, 0x001fa729);
+		regmap_write(priv->hhi, HHI_HDMI_PLL_CNTL6, 0x01a31500);
 
 		/* Reset PLL */
 		regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL,
-					HDMI_PLL_RESET, HDMI_PLL_RESET);
+				HDMI_PLL_RESET, HDMI_PLL_RESET);
 		regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL,
-					HDMI_PLL_RESET, 0);
+				HDMI_PLL_RESET, 0);
 
 		/* Poll for lock bit */
 		regmap_read_poll_timeout(priv->hhi, HHI_HDMI_PLL_CNTL, val,
 				(val & HDMI_PLL_LOCK), 10, 0);
-	};
+	}
 
 	if (meson_vpu_is_compatible(priv, "amlogic,meson-gxbb-vpu"))
 		regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL2,
-				   3 << 16, pll_od_to_reg(od1) << 16);
+				3 << 16, pll_od_to_reg(od1) << 16);
 	else if (meson_vpu_is_compatible(priv, "amlogic,meson-gxm-vpu") ||
-		 meson_vpu_is_compatible(priv, "amlogic,meson-gxl-vpu"))
+			meson_vpu_is_compatible(priv, "amlogic,meson-gxl-vpu"))
 		regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL3,
-				   3 << 21, pll_od_to_reg(od1) << 21);
+				3 << 21, pll_od_to_reg(od1) << 21);
 
 	if (meson_vpu_is_compatible(priv, "amlogic,meson-gxbb-vpu"))
 		regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL2,
-				   3 << 22, pll_od_to_reg(od2) << 22);
+				3 << 22, pll_od_to_reg(od2) << 22);
 	else if (meson_vpu_is_compatible(priv, "amlogic,meson-gxm-vpu") ||
-		 meson_vpu_is_compatible(priv, "amlogic,meson-gxl-vpu"))
+			meson_vpu_is_compatible(priv, "amlogic,meson-gxl-vpu"))
 		regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL3,
-				   3 << 23, pll_od_to_reg(od2) << 23);
+				3 << 23, pll_od_to_reg(od2) << 23);
 
 	if (meson_vpu_is_compatible(priv, "amlogic,meson-gxbb-vpu"))
 		regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL2,
-				   3 << 18, pll_od_to_reg(od3) << 18);
+				3 << 18, pll_od_to_reg(od3) << 18);
 	else if (meson_vpu_is_compatible(priv, "amlogic,meson-gxm-vpu") ||
-		 meson_vpu_is_compatible(priv, "amlogic,meson-gxl-vpu"))
+			meson_vpu_is_compatible(priv, "amlogic,meson-gxl-vpu"))
 		regmap_update_bits(priv->hhi, HHI_HDMI_PLL_CNTL3,
-				   3 << 19, pll_od_to_reg(od3) << 19);
+				3 << 19, pll_od_to_reg(od3) << 19);
+
 }
 
-void meson_vclk_setup(struct meson_drm *priv, unsigned int target,
-		      unsigned int vclk_freq, unsigned int venc_freq,
-		      unsigned int dac_freq, bool hdmi_use_enci)
+#define XTAL_FREQ 24000
+
+static unsigned int meson_hdmi_pll_get_m(struct meson_drm *priv,
+					 unsigned int pll_freq)
 {
-	unsigned int freq;
-	unsigned int hdmi_tx_div;
-	unsigned int venc_div;
+	/* The GXBB PLL has a /2 pre-multiplier */
+	if (meson_vpu_is_compatible(priv, "amlogic,meson-gxbb-vpu"))
+		pll_freq /= 2;
 
-	if (target == MESON_VCLK_TARGET_CVBS) {
-		meson_venci_cvbs_clock_config(priv);
-		return;
+	return pll_freq / XTAL_FREQ;
+}
+
+#define HDMI_FRAC_MAX_GXBB	4096
+#define HDMI_FRAC_MAX_GXL	1024
+
+static unsigned int meson_hdmi_pll_get_frac(struct meson_drm *priv,
+					    unsigned int m,
+					    unsigned int pll_freq)
+{
+	unsigned int parent_freq = XTAL_FREQ;
+	unsigned int frac_max = HDMI_FRAC_MAX_GXL;
+	unsigned int frac_m;
+	unsigned int frac;
+
+	/* The GXBB PLL has a /2 pre-multiplier and a larger FRAC width */
+	if (meson_vpu_is_compatible(priv, "amlogic,meson-gxbb-vpu")) {
+		frac_max = HDMI_FRAC_MAX_GXBB;
+		parent_freq *= 2;
 	}
 
-	hdmi_tx_div = vclk_freq / dac_freq;
+	/* We can have a perfect match !*/
+	if (pll_freq / m == parent_freq &&
+	    pll_freq % m == 0)
+		return 0;
 
-	if (hdmi_tx_div == 0) {
-		pr_err("Fatal Error, invalid HDMI-TX freq %d\n",
-				dac_freq);
-		return;
+	frac = div_u64((u64)pll_freq * (u64)frac_max, parent_freq);
+	frac_m = m * frac_max;
+	if (frac_m > frac)
+		return frac_max;
+	frac -= frac_m;
+
+	return min((u16)frac, (u16)(frac_max - 1));
+}
+
+static bool meson_hdmi_pll_validate_params(struct meson_drm *priv,
+					   unsigned int m,
+					   unsigned int frac)
+{
+	if (meson_vpu_is_compatible(priv, "amlogic,meson-gxbb-vpu")) {
+		/* Empiric supported min/max dividers */
+		if (m < 53 || m > 123)
+			return false;
+		if (frac >= HDMI_FRAC_MAX_GXBB)
+			return false;
+	} else if (meson_vpu_is_compatible(priv, "amlogic,meson-gxm-vpu") ||
+		   meson_vpu_is_compatible(priv, "amlogic,meson-gxl-vpu")) {
+		/* Empiric supported min/max dividers */
+		if (m < 106 || m > 247)
+			return false;
+		if (frac >= HDMI_FRAC_MAX_GXL)
+			return false;
 	}
 
-	venc_div = vclk_freq / venc_freq;
+	return true;
+}
 
-	if (venc_div == 0) {
-		pr_err("Fatal Error, invalid HDMI venc freq %d\n",
-				venc_freq);
-		return;
+static bool meson_hdmi_pll_find_params(struct meson_drm *priv,
+				       unsigned int freq,
+				       unsigned int *m,
+				       unsigned int *frac,
+				       unsigned int *od)
+{
+	/* Cycle from /16 to /2 */
+	for (*od = 16 ; *od > 1 ; *od >>= 1) {
+		*m = meson_hdmi_pll_get_m(priv, freq * *od);
+		if (!*m)
+			continue;
+		*frac = meson_hdmi_pll_get_frac(priv, *m, freq * *od);
+
+		DRM_DEBUG_DRIVER("PLL params for %dkHz: m=%x frac=%x od=%d\n",
+				 freq, *m, *frac, *od);
+
+		if (meson_hdmi_pll_validate_params(priv, *m, *frac))
+			return true;
 	}
 
-	switch (vclk_freq) {
-	case 54000:
-		if (hdmi_use_enci)
-			freq = MESON_VCLK_HDMI_ENCI_54000;
-		else
-			freq = MESON_VCLK_HDMI_DDR_54000;
-		break;
-	case 25175:
-		freq = MESON_VCLK_HDMI_25175;
-		break;
-	case 40000:
-		freq = MESON_VCLK_HDMI_40000;
-		break;
-	case 65000:
-		freq = MESON_VCLK_HDMI_65000;
-		break;
-	case 74250:
-		freq = MESON_VCLK_HDMI_74250;
-		break;
-	case 108000:
-		freq = MESON_VCLK_HDMI_108000;
-		break;
-	case 148500:
-		if (dac_freq != 148500)
-			freq = MESON_VCLK_HDMI_DDR_148500;
-		else
-			freq = MESON_VCLK_HDMI_148500;
-		break;
-	case 162000:
-		freq = MESON_VCLK_HDMI_162000;
-		break;
-	case 297000:
-		freq = MESON_VCLK_HDMI_297000;
-		break;
-	case 594000:
-		freq = MESON_VCLK_HDMI_594000;
-		break;
-	default:
-		pr_err("Fatal Error, invalid HDMI vclk freq %d\n",
-			vclk_freq);
+	return false;
+}
+
+/* pll_freq is the frequency after the OD dividers */
+enum drm_mode_status
+meson_vclk_dmt_supported_freq(struct meson_drm *priv, unsigned int freq)
+{
+	unsigned int od, m, frac;
+
+	/* In DMT mode, path after PLL is always /10 */
+	freq *= 10;
+
+	if (meson_hdmi_pll_find_params(priv, freq, &m, &frac, &od))
+		return MODE_OK;
+
+	return MODE_CLOCK_RANGE;
+}
+EXPORT_SYMBOL_GPL(meson_vclk_dmt_supported_freq);
+
+/* pll_freq is the frequency after the OD dividers */
+static void meson_hdmi_pll_generic_set(struct meson_drm *priv,
+				       unsigned int pll_freq)
+{
+	unsigned int od, m, frac, od1, od2, od3;
+
+	if (meson_hdmi_pll_find_params(priv, pll_freq, &m, &frac, &od)) {
+		od3 = 1;
+		if (od < 4) {
+			od1 = 2;
+			od2 = 1;
+		} else {
+			od2 = od / 4;
+			od1 = od / od2;
+		}
+
+		DRM_DEBUG_DRIVER("PLL params for %dkHz: m=%x frac=%x od=%d/%d/%d\n",
+				 pll_freq, m, frac, od1, od2, od3);
+
+		meson_hdmi_pll_set_params(priv, m, frac, od1, od2, od3);
+
 		return;
 	}
 
+	DRM_ERROR("Fatal, unable to find parameters for PLL freq %d\n",
+		  pll_freq);
+}
+
+static void meson_vclk_set(struct meson_drm *priv, unsigned int pll_base_freq,
+			   unsigned int od1, unsigned int od2, unsigned int od3,
+			   unsigned int vid_pll_div, unsigned int vclk_div,
+			   unsigned int hdmi_tx_div, unsigned int venc_div,
+			   bool hdmi_use_enci)
+{
 	/* Set HDMI-TX sys clock */
 	regmap_update_bits(priv->hhi, HHI_HDMI_CLK_CNTL,
 			   CTS_HDMI_SYS_SEL_MASK, 0);
@@ -831,19 +641,49 @@ void meson_vclk_setup(struct meson_drm *priv, unsigned int target,
 			   CTS_HDMI_SYS_EN, CTS_HDMI_SYS_EN);
 
 	/* Set HDMI PLL rate */
-	meson_hdmi_pll_set(priv, params[freq].pll_base_freq,
-			   params[freq].pll_od1,
-			   params[freq].pll_od2,
-			   params[freq].pll_od3);
+	if (!od1 && !od2 && !od3) {
+		meson_hdmi_pll_generic_set(priv, pll_base_freq);
+	} else if (meson_vpu_is_compatible(priv, "amlogic,meson-gxbb-vpu")) {
+		switch (pll_base_freq) {
+		case 2970000:
+			meson_hdmi_pll_set_params(priv, 0x3d, 0xe00,
+						  od1, od2, od3);
+			break;
+		case 4320000:
+			meson_hdmi_pll_set_params(priv, 0x5a, 0,
+						  od1, od2, od3);
+			break;
+		case 5940000:
+			meson_hdmi_pll_set_params(priv, 0x7b, 0xc00,
+						  od1, od2, od3);
+			break;
+		}
+	} else if (meson_vpu_is_compatible(priv, "amlogic,meson-gxm-vpu") ||
+		   meson_vpu_is_compatible(priv, "amlogic,meson-gxl-vpu")) {
+		switch (pll_base_freq) {
+		case 2970000:
+			meson_hdmi_pll_set_params(priv, 0x7b, 0x300,
+						  od1, od2, od3);
+			break;
+		case 4320000:
+			meson_hdmi_pll_set_params(priv, 0xb4, 0,
+						  od1, od2, od3);
+			break;
+		case 5940000:
+			meson_hdmi_pll_set_params(priv, 0xf7, 0x200,
+						  od1, od2, od3);
+			break;
+		}
+	}
 
 	/* Setup vid_pll divider */
-	meson_vid_pll_set(priv, params[freq].vid_pll_div);
+	meson_vid_pll_set(priv, vid_pll_div);
 
 	/* Set VCLK div */
 	regmap_update_bits(priv->hhi, HHI_VID_CLK_CNTL,
 			   VCLK_SEL_MASK, 0);
 	regmap_update_bits(priv->hhi, HHI_VID_CLK_DIV,
-			   VCLK_DIV_MASK, params[freq].vclk_div - 1);
+			   VCLK_DIV_MASK, vclk_div - 1);
 
 	/* Set HDMI-TX source */
 	switch (hdmi_tx_div) {
@@ -981,4 +821,80 @@ void meson_vclk_setup(struct meson_drm *priv, unsigned int target,
 
 	regmap_update_bits(priv->hhi, HHI_VID_CLK_CNTL, VCLK_EN, VCLK_EN);
 }
+
+void meson_vclk_setup(struct meson_drm *priv, unsigned int target,
+		      unsigned int vclk_freq, unsigned int venc_freq,
+		      unsigned int dac_freq, bool hdmi_use_enci)
+{
+	unsigned int freq;
+	unsigned int hdmi_tx_div;
+	unsigned int venc_div;
+
+	if (target == MESON_VCLK_TARGET_CVBS) {
+		meson_venci_cvbs_clock_config(priv);
+		return;
+	} else if (target == MESON_VCLK_TARGET_DMT) {
+		/* The DMT clock path is fixed after the PLL:
+		 * - automatic PLL freq + OD management
+		 * - vid_pll_div = VID_PLL_DIV_5
+		 * - vclk_div = 2
+		 * - hdmi_tx_div = 1
+		 * - venc_div = 1
+		 * - encp encoder
+		 */
+		meson_vclk_set(priv, vclk_freq * 10, 0, 0, 0,
+			       VID_PLL_DIV_5, 2, 1, 1, false);
+		return;
+	}
+
+	hdmi_tx_div = vclk_freq / dac_freq;
+
+	if (hdmi_tx_div == 0) {
+		pr_err("Fatal Error, invalid HDMI-TX freq %d\n",
+		       dac_freq);
+		return;
+	}
+
+	venc_div = vclk_freq / venc_freq;
+
+	if (venc_div == 0) {
+		pr_err("Fatal Error, invalid HDMI venc freq %d\n",
+		       venc_freq);
+		return;
+	}
+
+	switch (vclk_freq) {
+	case 54000:
+		if (hdmi_use_enci)
+			freq = MESON_VCLK_HDMI_ENCI_54000;
+		else
+			freq = MESON_VCLK_HDMI_DDR_54000;
+		break;
+	case 74250:
+		freq = MESON_VCLK_HDMI_74250;
+		break;
+	case 148500:
+		if (dac_freq != 148500)
+			freq = MESON_VCLK_HDMI_DDR_148500;
+		else
+			freq = MESON_VCLK_HDMI_148500;
+		break;
+	case 297000:
+		freq = MESON_VCLK_HDMI_297000;
+		break;
+	case 594000:
+		freq = MESON_VCLK_HDMI_594000;
+		break;
+	default:
+		pr_err("Fatal Error, invalid HDMI vclk freq %d\n",
+		       vclk_freq);
+		return;
+	}
+
+	meson_vclk_set(priv, params[freq].pll_base_freq,
+		       params[freq].pll_od1, params[freq].pll_od2,
+		       params[freq].pll_od3, params[freq].vid_pll_div,
+		       params[freq].vclk_div, hdmi_tx_div, venc_div,
+		       hdmi_use_enci);
+}
 EXPORT_SYMBOL_GPL(meson_vclk_setup);
diff --git a/drivers/gpu/drm/meson/meson_vclk.h b/drivers/gpu/drm/meson/meson_vclk.h
index 0401b5213471..869fa3a3073e 100644
--- a/drivers/gpu/drm/meson/meson_vclk.h
+++ b/drivers/gpu/drm/meson/meson_vclk.h
@@ -24,11 +24,15 @@
 enum {
 	MESON_VCLK_TARGET_CVBS = 0,
 	MESON_VCLK_TARGET_HDMI = 1,
+	MESON_VCLK_TARGET_DMT = 2,
 };
 
 /* 27MHz is the CVBS Pixel Clock */
 #define MESON_VCLK_CVBS			27000
 
+enum drm_mode_status
+meson_vclk_dmt_supported_freq(struct meson_drm *priv, unsigned int freq);
+
 void meson_vclk_setup(struct meson_drm *priv, unsigned int target,
 		      unsigned int vclk_freq, unsigned int venc_freq,
 		      unsigned int dac_freq, bool hdmi_use_enci);
diff --git a/drivers/gpu/drm/meson/meson_venc.c b/drivers/gpu/drm/meson/meson_venc.c
index 6e2701389801..514245e69b38 100644
--- a/drivers/gpu/drm/meson/meson_venc.c
+++ b/drivers/gpu/drm/meson/meson_venc.c
@@ -697,314 +697,6 @@ union meson_hdmi_venc_mode meson_hdmi_encp_mode_1080p60 = {
 	},
 };
 
-union meson_hdmi_venc_mode meson_hdmi_encp_mode_640x480_60 = {
-	.encp = {
-		.dvi_settings = 0x21,
-		.video_mode = 0x4040,
-		.video_mode_adv = 0x18,
-		/* video_prog_mode */
-		/* video_sync_mode */
-		/* video_yc_dly */
-		/* video_rgb_ctrl */
-		/* video_filt_ctrl */
-		/* video_ofld_voav_ofst */
-		/* yfp1_htime */
-		/* yfp2_htime */
-		.max_pxcnt = 0x31f,
-		/* hspuls_begin */
-		/* hspuls_end */
-		/* hspuls_switch */
-		/* vspuls_begin */
-		/* vspuls_end */
-		/* vspuls_bline */
-		/* vspuls_eline */
-		.havon_begin = 0x90,
-		.havon_end = 0x30f,
-		.vavon_bline = 0x23,
-		.vavon_eline = 0x202,
-		/* eqpuls_begin */
-		/* eqpuls_end */
-		/* eqpuls_bline */
-		/* eqpuls_eline */
-		.hso_begin = 0,
-		.hso_end = 0x60,
-		.vso_begin = 0x1e,
-		.vso_end = 0x32,
-		.vso_bline = 0,
-		.vso_eline = 2,
-		.vso_eline_present = true,
-		/* sy_val */
-		/* sy2_val */
-		.max_lncnt = 0x20c,
-	},
-};
-
-union meson_hdmi_venc_mode meson_hdmi_encp_mode_800x600_60 = {
-	.encp = {
-		.dvi_settings = 0x21,
-		.video_mode = 0x4040,
-		.video_mode_adv = 0x18,
-		/* video_prog_mode */
-		/* video_sync_mode */
-		/* video_yc_dly */
-		/* video_rgb_ctrl */
-		/* video_filt_ctrl */
-		/* video_ofld_voav_ofst */
-		/* yfp1_htime */
-		/* yfp2_htime */
-		.max_pxcnt = 0x41f,
-		/* hspuls_begin */
-		/* hspuls_end */
-		/* hspuls_switch */
-		/* vspuls_begin */
-		/* vspuls_end */
-		/* vspuls_bline */
-		/* vspuls_eline */
-		.havon_begin = 0xD8,
-		.havon_end = 0x3f7,
-		.vavon_bline = 0x1b,
-		.vavon_eline = 0x272,
-		/* eqpuls_begin */
-		/* eqpuls_end */
-		/* eqpuls_bline */
-		/* eqpuls_eline */
-		.hso_begin = 0,
-		.hso_end = 0x80,
-		.vso_begin = 0x1e,
-		.vso_end = 0x32,
-		.vso_bline = 0,
-		.vso_eline = 4,
-		.vso_eline_present = true,
-		/* sy_val */
-		/* sy2_val */
-		.max_lncnt = 0x273,
-	},
-};
-
-union meson_hdmi_venc_mode meson_hdmi_encp_mode_1024x768_60 = {
-	.encp = {
-		.dvi_settings = 0x21,
-		.video_mode = 0x4040,
-		.video_mode_adv = 0x18,
-		/* video_prog_mode */
-		/* video_sync_mode */
-		/* video_yc_dly */
-		/* video_rgb_ctrl */
-		/* video_filt_ctrl */
-		/* video_ofld_voav_ofst */
-		/* yfp1_htime */
-		/* yfp2_htime */
-		.max_pxcnt = 1343,
-		/* hspuls_begin */
-		/* hspuls_end */
-		/* hspuls_switch */
-		/* vspuls_begin */
-		/* vspuls_end */
-		/* vspuls_bline */
-		/* vspuls_eline */
-		.havon_begin = 296,
-		.havon_end = 1319,
-		.vavon_bline = 35,
-		.vavon_eline = 802,
-		/* eqpuls_begin */
-		/* eqpuls_end */
-		/* eqpuls_bline */
-		/* eqpuls_eline */
-		.hso_begin = 0,
-		.hso_end = 136,
-		.vso_begin = 30,
-		.vso_end = 50,
-		.vso_bline = 0,
-		.vso_eline = 6,
-		.vso_eline_present = true,
-		/* sy_val */
-		/* sy2_val */
-		.max_lncnt = 805,
-	},
-};
-
-union meson_hdmi_venc_mode meson_hdmi_encp_mode_1152x864_75 = {
-	.encp = {
-		.dvi_settings = 0x21,
-		.video_mode = 0x4040,
-		.video_mode_adv = 0x18,
-		/* video_prog_mode */
-		/* video_sync_mode */
-		/* video_yc_dly */
-		/* video_rgb_ctrl */
-		/* video_filt_ctrl */
-		/* video_ofld_voav_ofst */
-		/* yfp1_htime */
-		/* yfp2_htime */
-		.max_pxcnt = 0x63f,
-		/* hspuls_begin */
-		/* hspuls_end */
-		/* hspuls_switch */
-		/* vspuls_begin */
-		/* vspuls_end */
-		/* vspuls_bline */
-		/* vspuls_eline */
-		.havon_begin = 0x180,
-		.havon_end = 0x5ff,
-		.vavon_bline = 0x23,
-		.vavon_eline = 0x382,
-		/* eqpuls_begin */
-		/* eqpuls_end */
-		/* eqpuls_bline */
-		/* eqpuls_eline */
-		.hso_begin = 0,
-		.hso_end = 0x80,
-		.vso_begin = 0x1e,
-		.vso_end = 0x32,
-		.vso_bline = 0,
-		.vso_eline = 3,
-		.vso_eline_present = true,
-		/* sy_val */
-		/* sy2_val */
-		.max_lncnt = 0x383,
-	},
-};
-
-union meson_hdmi_venc_mode meson_hdmi_encp_mode_1280x1024_60 = {
-	.encp = {
-		.dvi_settings = 0x21,
-		.video_mode = 0x4040,
-		.video_mode_adv = 0x18,
-		/* video_prog_mode */
-		/* video_sync_mode */
-		/* video_yc_dly */
-		/* video_rgb_ctrl */
-		/* video_filt_ctrl */
-		/* video_ofld_voav_ofst */
-		/* yfp1_htime */
-		/* yfp2_htime */
-		.max_pxcnt = 0x697,
-		/* hspuls_begin */
-		/* hspuls_end */
-		/* hspuls_switch */
-		/* vspuls_begin */
-		/* vspuls_end */
-		/* vspuls_bline */
-		/* vspuls_eline */
-		.havon_begin = 0x168,
-		.havon_end = 0x667,
-		.vavon_bline = 0x29,
-		.vavon_eline = 0x428,
-		/* eqpuls_begin */
-		/* eqpuls_end */
-		/* eqpuls_bline */
-		/* eqpuls_eline */
-		.hso_begin = 0,
-		.hso_end = 0x70,
-		.vso_begin = 0x1e,
-		.vso_end = 0x32,
-		.vso_bline = 0,
-		.vso_eline = 3,
-		.vso_eline_present = true,
-		/* sy_val */
-		/* sy2_val */
-		.max_lncnt = 0x429,
-	},
-};
-
-union meson_hdmi_venc_mode meson_hdmi_encp_mode_1600x1200_60 = {
-	.encp = {
-		.dvi_settings = 0x21,
-		.video_mode = 0x4040,
-		.video_mode_adv = 0x18,
-		/* video_prog_mode */
-		/* video_sync_mode */
-		/* video_yc_dly */
-		/* video_rgb_ctrl */
-		/* video_filt_ctrl */
-		/* video_ofld_voav_ofst */
-		/* yfp1_htime */
-		/* yfp2_htime */
-		.max_pxcnt = 0x86f,
-		/* hspuls_begin */
-		/* hspuls_end */
-		/* hspuls_switch */
-		/* vspuls_begin */
-		/* vspuls_end */
-		/* vspuls_bline */
-		/* vspuls_eline */
-		.havon_begin = 0x1f0,
-		.havon_end = 0x82f,
-		.vavon_bline = 0x31,
-		.vavon_eline = 0x4e0,
-		/* eqpuls_begin */
-		/* eqpuls_end */
-		/* eqpuls_bline */
-		/* eqpuls_eline */
-		.hso_begin = 0,
-		.hso_end = 0xc0,
-		.vso_begin = 0x1e,
-		.vso_end = 0x32,
-		.vso_bline = 0,
-		.vso_eline = 3,
-		.vso_eline_present = true,
-		/* sy_val */
-		/* sy2_val */
-		.max_lncnt = 0x4e1,
-	},
-};
-
-struct meson_hdmi_venc_dmt_mode {
-	struct drm_display_mode drm_mode;
-	union meson_hdmi_venc_mode *mode;
-} meson_hdmi_venc_dmt_modes[] = {
-	/* 640x480@60Hz */
-	{
-		{ DRM_MODE("640x480", DRM_MODE_TYPE_DRIVER, 25175, 640, 656,
-		  752, 800, 0, 480, 490, 492, 525, 0,
-		  DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC) },
-		&meson_hdmi_encp_mode_640x480_60,
-	},
-	/* 800x600@60Hz */
-	{
-		{ DRM_MODE("800x600", DRM_MODE_TYPE_DRIVER, 40000, 800, 840,
-		  968, 1056, 0, 600, 601, 605, 628, 0,
-		  DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC) },
-		&meson_hdmi_encp_mode_800x600_60,
-	},
-	/* 1024x768@60Hz */
-	{
-		{ DRM_MODE("1024x768", DRM_MODE_TYPE_DRIVER, 65000, 1024,
-		  1048, 1184, 1344, 0, 768, 771, 777, 806, 0,
-		  DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC) },
-		&meson_hdmi_encp_mode_1024x768_60,
-	},
-	/* 1152x864@75Hz */
-	{
-		{ DRM_MODE("1152x864", DRM_MODE_TYPE_DRIVER, 108000, 1152,
-		  1216, 1344, 1600, 0, 864, 865, 868, 900, 0,
-		  DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC) },
-		&meson_hdmi_encp_mode_1152x864_75,
-	},
-	/* 1280x1024@60Hz */
-	{
-		{ DRM_MODE("1280x1024", DRM_MODE_TYPE_DRIVER, 108000, 1280,
-		  1328, 1440, 1688, 0, 1024, 1025, 1028, 1066, 0,
-		  DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC) },
-		&meson_hdmi_encp_mode_1280x1024_60,
-	},
-	/* 1600x1200@60Hz */
-	{
-		{ DRM_MODE("1600x1200", DRM_MODE_TYPE_DRIVER, 162000, 1600,
-		  1664, 1856, 2160, 0, 1200, 1201, 1204, 1250, 0,
-		  DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC) },
-		&meson_hdmi_encp_mode_1600x1200_60,
-	},
-	/* 1920x1080@60Hz */
-	{
-		{ DRM_MODE("1920x1080", DRM_MODE_TYPE_DRIVER, 148500, 1920,
-		  2008, 2052, 2200, 0, 1080, 1084, 1089, 1125, 0,
-		  DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC) },
-		&meson_hdmi_encp_mode_1080p60
-	},
-	{ }, /* sentinel */
-};
-
 struct meson_hdmi_venc_vic_mode {
 	unsigned int vic;
 	union meson_hdmi_venc_mode *mode;
@@ -1044,17 +736,20 @@ static unsigned long modulo(unsigned long a, unsigned long b)
 		return a;
 }
 
-bool meson_venc_hdmi_supported_mode(const struct drm_display_mode *mode)
+enum drm_mode_status
+meson_venc_hdmi_supported_mode(const struct drm_display_mode *mode)
 {
-	struct meson_hdmi_venc_dmt_mode *vmode = meson_hdmi_venc_dmt_modes;
+	if (mode->flags & ~(DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_NHSYNC |
+			    DRM_MODE_FLAG_PVSYNC | DRM_MODE_FLAG_NVSYNC))
+		return MODE_BAD;
 
-	while (vmode->mode) {
-		if (drm_mode_equal(&vmode->drm_mode, mode))
-			return true;
-		vmode++;
-	}
+	if (mode->hdisplay < 640 || mode->hdisplay > 1920)
+		return MODE_BAD_HVALUE;
 
-	return false;
+	if (mode->vdisplay < 480 || mode->vdisplay > 1200)
+		return MODE_BAD_VVALUE;
+
+	return MODE_OK;
 }
 EXPORT_SYMBOL_GPL(meson_venc_hdmi_supported_mode);
 
@@ -1072,18 +767,29 @@ bool meson_venc_hdmi_supported_vic(int vic)
 }
 EXPORT_SYMBOL_GPL(meson_venc_hdmi_supported_vic);
 
-static union meson_hdmi_venc_mode
-*meson_venc_hdmi_get_dmt_vmode(const struct drm_display_mode *mode)
+void meson_venc_hdmi_get_dmt_vmode(const struct drm_display_mode *mode,
+				   union meson_hdmi_venc_mode *dmt_mode)
 {
-	struct meson_hdmi_venc_dmt_mode *vmode = meson_hdmi_venc_dmt_modes;
-
-	while (vmode->mode) {
-		if (drm_mode_equal(&vmode->drm_mode, mode))
-			return vmode->mode;
-		vmode++;
-	}
-
-	return NULL;
+	memset(dmt_mode, 0, sizeof(*dmt_mode));
+
+	dmt_mode->encp.dvi_settings = 0x21;
+	dmt_mode->encp.video_mode = 0x4040;
+	dmt_mode->encp.video_mode_adv = 0x18;
+	dmt_mode->encp.max_pxcnt = mode->htotal - 1;
+	dmt_mode->encp.havon_begin = mode->htotal - mode->hsync_start;
+	dmt_mode->encp.havon_end = dmt_mode->encp.havon_begin +
+				   mode->hdisplay - 1;
+	dmt_mode->encp.vavon_bline = mode->vtotal - mode->vsync_start;
+	dmt_mode->encp.vavon_eline = dmt_mode->encp.vavon_bline +
+				     mode->vdisplay - 1;
+	dmt_mode->encp.hso_begin = 0;
+	dmt_mode->encp.hso_end = mode->hsync_end - mode->hsync_start;
+	dmt_mode->encp.vso_begin = 30;
+	dmt_mode->encp.vso_end = 50;
+	dmt_mode->encp.vso_bline = 0;
+	dmt_mode->encp.vso_eline = mode->vsync_end - mode->vsync_start;
+	dmt_mode->encp.vso_eline_present = true;
+	dmt_mode->encp.max_lncnt = mode->vtotal - 1;
 }
 
 static union meson_hdmi_venc_mode *meson_venc_hdmi_get_vic_vmode(int vic)
@@ -1120,6 +826,7 @@ void meson_venc_hdmi_mode_set(struct meson_drm *priv, int vic,
 			      struct drm_display_mode *mode)
 {
 	union meson_hdmi_venc_mode *vmode = NULL;
+	union meson_hdmi_venc_mode vmode_dmt;
 	bool use_enci = false;
 	bool venc_repeat = false;
 	bool hdmi_repeat = false;
@@ -1147,14 +854,17 @@ void meson_venc_hdmi_mode_set(struct meson_drm *priv, int vic,
 	unsigned int sof_lines;
 	unsigned int vsync_lines;
 
-	if (meson_venc_hdmi_supported_vic(vic))
+	if (meson_venc_hdmi_supported_vic(vic)) {
 		vmode = meson_venc_hdmi_get_vic_vmode(vic);
-	else
-		vmode = meson_venc_hdmi_get_dmt_vmode(mode);
-	if (!vmode) {
-		dev_err(priv->dev, "%s: Fatal Error, unsupported mode "
-			DRM_MODE_FMT "\n", __func__, DRM_MODE_ARG(mode));
-		return;
+		if (!vmode) {
+			dev_err(priv->dev, "%s: Fatal Error, unsupported mode "
+				DRM_MODE_FMT "\n", __func__,
+				DRM_MODE_ARG(mode));
+			return;
+		}
+	} else {
+		meson_venc_hdmi_get_dmt_vmode(mode, &vmode_dmt);
+		vmode = &vmode_dmt;
 	}
 
 	/* Use VENCI for 480i and 576i and double HDMI pixels */
diff --git a/drivers/gpu/drm/meson/meson_venc.h b/drivers/gpu/drm/meson/meson_venc.h
index 7c18a36a0dd0..97eaebbfa0c4 100644
--- a/drivers/gpu/drm/meson/meson_venc.h
+++ b/drivers/gpu/drm/meson/meson_venc.h
@@ -58,7 +58,8 @@ struct meson_cvbs_enci_mode {
 };
 
 /* HDMI Clock parameters */
-bool meson_venc_hdmi_supported_mode(const struct drm_display_mode *mode);
+enum drm_mode_status
+meson_venc_hdmi_supported_mode(const struct drm_display_mode *mode);
 bool meson_venc_hdmi_supported_vic(int vic);
 bool meson_venc_hdmi_venc_repeat(int vic);
 
diff --git a/drivers/gpu/drm/meson/meson_venc_cvbs.c b/drivers/gpu/drm/meson/meson_venc_cvbs.c
index 79d95ca8a0c0..f7945bae3b4a 100644
--- a/drivers/gpu/drm/meson/meson_venc_cvbs.c
+++ b/drivers/gpu/drm/meson/meson_venc_cvbs.c
@@ -282,7 +282,7 @@ int meson_venc_cvbs_create(struct meson_drm *priv)
 
 	encoder->possible_crtcs = BIT(0);
 
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/mgag200/mgag200_mode.c b/drivers/gpu/drm/mgag200/mgag200_mode.c
index 8918539a19aa..acf7bfe68454 100644
--- a/drivers/gpu/drm/mgag200/mgag200_mode.c
+++ b/drivers/gpu/drm/mgag200/mgag200_mode.c
@@ -1553,7 +1553,7 @@ static int mga_vga_get_modes(struct drm_connector *connector)
 
 	edid = drm_get_edid(connector, &mga_connector->i2c->adapter);
 	if (edid) {
-		drm_mode_connector_update_edid_property(connector, edid);
+		drm_connector_update_edid_property(connector, edid);
 		ret = drm_add_edid_modes(connector, edid);
 		kfree(edid);
 	}
@@ -1747,7 +1747,7 @@ int mgag200_modeset_init(struct mga_device *mdev)
 		return -1;
 	}
 
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	ret = mgag200_fbdev_init(mdev);
 	if (ret) {
diff --git a/drivers/gpu/drm/msm/adreno/adreno_device.c b/drivers/gpu/drm/msm/adreno/adreno_device.c
index 8e0cb161754b..0ae5ace65462 100644
--- a/drivers/gpu/drm/msm/adreno/adreno_device.c
+++ b/drivers/gpu/drm/msm/adreno/adreno_device.c
@@ -168,7 +168,6 @@ struct msm_gpu *adreno_load_gpu(struct drm_device *dev)
 	if (gpu->funcs->debugfs_init) {
 		gpu->funcs->debugfs_init(gpu, dev->primary);
 		gpu->funcs->debugfs_init(gpu, dev->render);
-		gpu->funcs->debugfs_init(gpu, dev->control);
 	}
 #endif
 
diff --git a/drivers/gpu/drm/msm/disp/mdp4/mdp4_crtc.c b/drivers/gpu/drm/msm/disp/mdp4/mdp4_crtc.c
index b001699297c4..457c29dba4a1 100644
--- a/drivers/gpu/drm/msm/disp/mdp4/mdp4_crtc.c
+++ b/drivers/gpu/drm/msm/disp/mdp4/mdp4_crtc.c
@@ -201,7 +201,7 @@ static void blend_setup(struct drm_crtc *crtc)
 		int idx = idxs[pipe_id];
 		if (idx > 0) {
 			const struct mdp_format *format =
-					to_mdp_format(msm_framebuffer_format(plane->fb));
+					to_mdp_format(msm_framebuffer_format(plane->state->fb));
 			alpha[idx-1] = format->alpha_enable;
 		}
 	}
@@ -665,7 +665,6 @@ struct drm_crtc *mdp4_crtc_init(struct drm_device *dev,
 	drm_crtc_init_with_planes(dev, crtc, plane, NULL, &mdp4_crtc_funcs,
 				  NULL);
 	drm_crtc_helper_add(crtc, &mdp4_crtc_helper_funcs);
-	plane->crtc = crtc;
 
 	return crtc;
 }
diff --git a/drivers/gpu/drm/msm/disp/mdp4/mdp4_lcdc_encoder.c b/drivers/gpu/drm/msm/disp/mdp4/mdp4_lcdc_encoder.c
index 4a645926edb7..2bfb39082f54 100644
--- a/drivers/gpu/drm/msm/disp/mdp4/mdp4_lcdc_encoder.c
+++ b/drivers/gpu/drm/msm/disp/mdp4/mdp4_lcdc_encoder.c
@@ -341,7 +341,7 @@ static void mdp4_lcdc_encoder_disable(struct drm_encoder *encoder)
 	mdp4_write(mdp4_kms, REG_MDP4_LCDC_ENABLE, 0);
 
 	panel = of_drm_find_panel(mdp4_lcdc_encoder->panel_node);
-	if (panel) {
+	if (!IS_ERR(panel)) {
 		drm_panel_disable(panel);
 		drm_panel_unprepare(panel);
 	}
@@ -410,7 +410,7 @@ static void mdp4_lcdc_encoder_enable(struct drm_encoder *encoder)
 		dev_err(dev->dev, "failed to enable lcdc_clk: %d\n", ret);
 
 	panel = of_drm_find_panel(mdp4_lcdc_encoder->panel_node);
-	if (panel) {
+	if (!IS_ERR(panel)) {
 		drm_panel_prepare(panel);
 		drm_panel_enable(panel);
 	}
diff --git a/drivers/gpu/drm/msm/disp/mdp4/mdp4_lvds_connector.c b/drivers/gpu/drm/msm/disp/mdp4/mdp4_lvds_connector.c
index e3b1c86b7aae..5368e621999c 100644
--- a/drivers/gpu/drm/msm/disp/mdp4/mdp4_lvds_connector.c
+++ b/drivers/gpu/drm/msm/disp/mdp4/mdp4_lvds_connector.c
@@ -34,9 +34,12 @@ static enum drm_connector_status mdp4_lvds_connector_detect(
 	struct mdp4_lvds_connector *mdp4_lvds_connector =
 			to_mdp4_lvds_connector(connector);
 
-	if (!mdp4_lvds_connector->panel)
+	if (!mdp4_lvds_connector->panel) {
 		mdp4_lvds_connector->panel =
 			of_drm_find_panel(mdp4_lvds_connector->panel_node);
+		if (IS_ERR(mdp4_lvds_connector->panel))
+			mdp4_lvds_connector->panel = NULL;
+	}
 
 	return mdp4_lvds_connector->panel ?
 			connector_status_connected :
@@ -129,7 +132,7 @@ struct drm_connector *mdp4_lvds_connector_init(struct drm_device *dev,
 	connector->interlace_allowed = 0;
 	connector->doublescan_allowed = 0;
 
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	return connector;
 }
diff --git a/drivers/gpu/drm/msm/disp/mdp4/mdp4_plane.c b/drivers/gpu/drm/msm/disp/mdp4/mdp4_plane.c
index 7a1ad3af08e3..79ff653d8081 100644
--- a/drivers/gpu/drm/msm/disp/mdp4/mdp4_plane.c
+++ b/drivers/gpu/drm/msm/disp/mdp4/mdp4_plane.c
@@ -68,7 +68,7 @@ static void mdp4_plane_destroy(struct drm_plane *plane)
 {
 	struct mdp4_plane *mdp4_plane = to_mdp4_plane(plane);
 
-	drm_plane_helper_disable(plane);
+	drm_plane_helper_disable(plane, NULL);
 	drm_plane_cleanup(plane);
 
 	kfree(mdp4_plane);
@@ -98,21 +98,6 @@ static const struct drm_plane_funcs mdp4_plane_funcs = {
 		.atomic_destroy_state = drm_atomic_helper_plane_destroy_state,
 };
 
-static int mdp4_plane_prepare_fb(struct drm_plane *plane,
-				 struct drm_plane_state *new_state)
-{
-	struct mdp4_plane *mdp4_plane = to_mdp4_plane(plane);
-	struct mdp4_kms *mdp4_kms = get_kms(plane);
-	struct msm_kms *kms = &mdp4_kms->base.base;
-	struct drm_framebuffer *fb = new_state->fb;
-
-	if (!fb)
-		return 0;
-
-	DBG("%s: prepare: FB[%u]", mdp4_plane->name, fb->base.id);
-	return msm_framebuffer_prepare(fb, kms->aspace);
-}
-
 static void mdp4_plane_cleanup_fb(struct drm_plane *plane,
 				  struct drm_plane_state *old_state)
 {
@@ -152,7 +137,7 @@ static void mdp4_plane_atomic_update(struct drm_plane *plane,
 }
 
 static const struct drm_plane_helper_funcs mdp4_plane_helper_funcs = {
-		.prepare_fb = mdp4_plane_prepare_fb,
+		.prepare_fb = msm_atomic_prepare_fb,
 		.cleanup_fb = mdp4_plane_cleanup_fb,
 		.atomic_check = mdp4_plane_atomic_check,
 		.atomic_update = mdp4_plane_atomic_update,
@@ -182,8 +167,6 @@ static void mdp4_plane_set_scanout(struct drm_plane *plane,
 			msm_framebuffer_iova(fb, kms->aspace, 2));
 	mdp4_write(mdp4_kms, REG_MDP4_PIPE_SRCP3_BASE(pipe),
 			msm_framebuffer_iova(fb, kms->aspace, 3));
-
-	plane->fb = fb;
 }
 
 static void mdp4_write_csc_config(struct mdp4_kms *mdp4_kms,
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c
index 76b96081916f..24e00274844b 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_crtc.c
@@ -430,6 +430,7 @@ static void mdp5_crtc_atomic_disable(struct drm_crtc *crtc,
 	struct mdp5_crtc_state *mdp5_cstate = to_mdp5_crtc_state(crtc->state);
 	struct mdp5_kms *mdp5_kms = get_kms(crtc);
 	struct device *dev = &mdp5_kms->pdev->dev;
+	unsigned long flags;
 
 	DBG("%s", crtc->name);
 
@@ -445,6 +446,14 @@ static void mdp5_crtc_atomic_disable(struct drm_crtc *crtc,
 	mdp_irq_unregister(&mdp5_kms->base, &mdp5_crtc->err);
 	pm_runtime_put_sync(dev);
 
+	if (crtc->state->event && !crtc->state->active) {
+		WARN_ON(mdp5_crtc->event);
+		spin_lock_irqsave(&mdp5_kms->dev->event_lock, flags);
+		drm_crtc_send_vblank_event(crtc, crtc->state->event);
+		crtc->state->event = NULL;
+		spin_unlock_irqrestore(&mdp5_kms->dev->event_lock, flags);
+	}
+
 	mdp5_crtc->enabled = false;
 }
 
@@ -1198,7 +1207,6 @@ struct drm_crtc *mdp5_crtc_init(struct drm_device *dev,
 			"unref cursor", unref_cursor_worker);
 
 	drm_crtc_helper_add(crtc, &mdp5_crtc_helper_funcs);
-	plane->crtc = crtc;
 
 	return crtc;
 }
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c
index 6d8e3a9a6fc0..6e12e275deba 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.c
@@ -70,60 +70,110 @@ static int mdp5_hw_init(struct msm_kms *kms)
 	return 0;
 }
 
-struct mdp5_state *mdp5_get_state(struct drm_atomic_state *s)
+/* Global/shared object state funcs */
+
+/*
+ * This is a helper that returns the private state currently in operation.
+ * Note that this would return the "old_state" if called in the atomic check
+ * path, and the "new_state" after the atomic swap has been done.
+ */
+struct mdp5_global_state *
+mdp5_get_existing_global_state(struct mdp5_kms *mdp5_kms)
+{
+	return to_mdp5_global_state(mdp5_kms->glob_state.state);
+}
+
+/*
+ * This acquires the modeset lock set aside for global state, creates
+ * a new duplicated private object state.
+ */
+struct mdp5_global_state *mdp5_get_global_state(struct drm_atomic_state *s)
 {
 	struct msm_drm_private *priv = s->dev->dev_private;
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(priv->kms));
-	struct msm_kms_state *state = to_kms_state(s);
-	struct mdp5_state *new_state;
+	struct drm_private_state *priv_state;
 	int ret;
 
-	if (state->state)
-		return state->state;
-
-	ret = drm_modeset_lock(&mdp5_kms->state_lock, s->acquire_ctx);
+	ret = drm_modeset_lock(&mdp5_kms->glob_state_lock, s->acquire_ctx);
 	if (ret)
 		return ERR_PTR(ret);
 
-	new_state = kmalloc(sizeof(*mdp5_kms->state), GFP_KERNEL);
-	if (!new_state)
-		return ERR_PTR(-ENOMEM);
+	priv_state = drm_atomic_get_private_obj_state(s, &mdp5_kms->glob_state);
+	if (IS_ERR(priv_state))
+		return ERR_CAST(priv_state);
 
-	/* Copy state: */
-	new_state->hwpipe = mdp5_kms->state->hwpipe;
-	new_state->hwmixer = mdp5_kms->state->hwmixer;
-	if (mdp5_kms->smp)
-		new_state->smp = mdp5_kms->state->smp;
+	return to_mdp5_global_state(priv_state);
+}
+
+static struct drm_private_state *
+mdp5_global_duplicate_state(struct drm_private_obj *obj)
+{
+	struct mdp5_global_state *state;
+
+	state = kmemdup(obj->state, sizeof(*state), GFP_KERNEL);
+	if (!state)
+		return NULL;
 
-	state->state = new_state;
+	__drm_atomic_helper_private_obj_duplicate_state(obj, &state->base);
 
-	return new_state;
+	return &state->base;
 }
 
-static void mdp5_swap_state(struct msm_kms *kms, struct drm_atomic_state *state)
+static void mdp5_global_destroy_state(struct drm_private_obj *obj,
+				      struct drm_private_state *state)
 {
-	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
-	swap(to_kms_state(state)->state, mdp5_kms->state);
+	struct mdp5_global_state *mdp5_state = to_mdp5_global_state(state);
+
+	kfree(mdp5_state);
+}
+
+static const struct drm_private_state_funcs mdp5_global_state_funcs = {
+	.atomic_duplicate_state = mdp5_global_duplicate_state,
+	.atomic_destroy_state = mdp5_global_destroy_state,
+};
+
+static int mdp5_global_obj_init(struct mdp5_kms *mdp5_kms)
+{
+	struct mdp5_global_state *state;
+
+	drm_modeset_lock_init(&mdp5_kms->glob_state_lock);
+
+	state = kzalloc(sizeof(*state), GFP_KERNEL);
+	if (!state)
+		return -ENOMEM;
+
+	state->mdp5_kms = mdp5_kms;
+
+	drm_atomic_private_obj_init(&mdp5_kms->glob_state,
+				    &state->base,
+				    &mdp5_global_state_funcs);
+	return 0;
 }
 
 static void mdp5_prepare_commit(struct msm_kms *kms, struct drm_atomic_state *state)
 {
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
 	struct device *dev = &mdp5_kms->pdev->dev;
+	struct mdp5_global_state *global_state;
+
+	global_state = mdp5_get_existing_global_state(mdp5_kms);
 
 	pm_runtime_get_sync(dev);
 
 	if (mdp5_kms->smp)
-		mdp5_smp_prepare_commit(mdp5_kms->smp, &mdp5_kms->state->smp);
+		mdp5_smp_prepare_commit(mdp5_kms->smp, &global_state->smp);
 }
 
 static void mdp5_complete_commit(struct msm_kms *kms, struct drm_atomic_state *state)
 {
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(kms));
 	struct device *dev = &mdp5_kms->pdev->dev;
+	struct mdp5_global_state *global_state;
+
+	global_state = mdp5_get_existing_global_state(mdp5_kms);
 
 	if (mdp5_kms->smp)
-		mdp5_smp_complete_commit(mdp5_kms->smp, &mdp5_kms->state->smp);
+		mdp5_smp_complete_commit(mdp5_kms->smp, &global_state->smp);
 
 	pm_runtime_put_sync(dev);
 }
@@ -229,7 +279,6 @@ static const struct mdp_kms_funcs kms_funcs = {
 		.irq             = mdp5_irq,
 		.enable_vblank   = mdp5_enable_vblank,
 		.disable_vblank  = mdp5_disable_vblank,
-		.swap_state      = mdp5_swap_state,
 		.prepare_commit  = mdp5_prepare_commit,
 		.complete_commit = mdp5_complete_commit,
 		.wait_for_crtc_commit_done = mdp5_wait_for_crtc_commit_done,
@@ -727,7 +776,8 @@ static void mdp5_destroy(struct platform_device *pdev)
 	if (mdp5_kms->rpm_enabled)
 		pm_runtime_disable(&pdev->dev);
 
-	kfree(mdp5_kms->state);
+	drm_atomic_private_obj_fini(&mdp5_kms->glob_state);
+	drm_modeset_lock_fini(&mdp5_kms->glob_state_lock);
 }
 
 static int construct_pipes(struct mdp5_kms *mdp5_kms, int cnt,
@@ -880,12 +930,9 @@ static int mdp5_init(struct platform_device *pdev, struct drm_device *dev)
 	mdp5_kms->dev = dev;
 	mdp5_kms->pdev = pdev;
 
-	drm_modeset_lock_init(&mdp5_kms->state_lock);
-	mdp5_kms->state = kzalloc(sizeof(*mdp5_kms->state), GFP_KERNEL);
-	if (!mdp5_kms->state) {
-		ret = -ENOMEM;
+	ret = mdp5_global_obj_init(mdp5_kms);
+	if (ret)
 		goto fail;
-	}
 
 	mdp5_kms->mmio = msm_ioremap(pdev, "mdp_phys", "MDP5");
 	if (IS_ERR(mdp5_kms->mmio)) {
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h
index 425a03d213e5..854dfd30e829 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_kms.h
@@ -28,8 +28,6 @@
 #include "mdp5_ctl.h"
 #include "mdp5_smp.h"
 
-struct mdp5_state;
-
 struct mdp5_kms {
 	struct mdp_kms base;
 
@@ -49,11 +47,12 @@ struct mdp5_kms {
 	struct mdp5_cfg_handler *cfg;
 	uint32_t caps;	/* MDP capabilities (MDP_CAP_XXX bits) */
 
-	/**
-	 * Global atomic state.  Do not access directly, use mdp5_get_state()
+	/*
+	 * Global private object state, Do not access directly, use
+	 * mdp5_global_get_state()
 	 */
-	struct mdp5_state *state;
-	struct drm_modeset_lock state_lock;
+	struct drm_modeset_lock glob_state_lock;
+	struct drm_private_obj glob_state;
 
 	struct mdp5_smp *smp;
 	struct mdp5_ctl_manager *ctlm;
@@ -81,19 +80,23 @@ struct mdp5_kms {
 };
 #define to_mdp5_kms(x) container_of(x, struct mdp5_kms, base)
 
-/* Global atomic state for tracking resources that are shared across
+/* Global private object state for tracking resources that are shared across
  * multiple kms objects (planes/crtcs/etc).
- *
- * For atomic updates which require modifying global state,
  */
-struct mdp5_state {
+#define to_mdp5_global_state(x) container_of(x, struct mdp5_global_state, base)
+struct mdp5_global_state {
+	struct drm_private_state base;
+
+	struct drm_atomic_state *state;
+	struct mdp5_kms *mdp5_kms;
+
 	struct mdp5_hw_pipe_state hwpipe;
 	struct mdp5_hw_mixer_state hwmixer;
 	struct mdp5_smp_state smp;
 };
 
-struct mdp5_state *__must_check
-mdp5_get_state(struct drm_atomic_state *s);
+struct mdp5_global_state * mdp5_get_existing_global_state(struct mdp5_kms *mdp5_kms);
+struct mdp5_global_state *__must_check mdp5_get_global_state(struct drm_atomic_state *s);
 
 /* Atomic plane state.  Subclasses the base drm_plane_state in order to
  * track assigned hwpipe and hw specific state.
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_mixer.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_mixer.c
index 8a00991f03c7..113e6b569562 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_mixer.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_mixer.c
@@ -52,14 +52,14 @@ int mdp5_mixer_assign(struct drm_atomic_state *s, struct drm_crtc *crtc,
 {
 	struct msm_drm_private *priv = s->dev->dev_private;
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(priv->kms));
-	struct mdp5_state *state = mdp5_get_state(s);
+	struct mdp5_global_state *global_state = mdp5_get_global_state(s);
 	struct mdp5_hw_mixer_state *new_state;
 	int i;
 
-	if (IS_ERR(state))
-		return PTR_ERR(state);
+	if (IS_ERR(global_state))
+		return PTR_ERR(global_state);
 
-	new_state = &state->hwmixer;
+	new_state = &global_state->hwmixer;
 
 	for (i = 0; i < mdp5_kms->num_hwmixers; i++) {
 		struct mdp5_hw_mixer *cur = mdp5_kms->hwmixers[i];
@@ -129,8 +129,8 @@ int mdp5_mixer_assign(struct drm_atomic_state *s, struct drm_crtc *crtc,
 
 void mdp5_mixer_release(struct drm_atomic_state *s, struct mdp5_hw_mixer *mixer)
 {
-	struct mdp5_state *state = mdp5_get_state(s);
-	struct mdp5_hw_mixer_state *new_state = &state->hwmixer;
+	struct mdp5_global_state *global_state = mdp5_get_global_state(s);
+	struct mdp5_hw_mixer_state *new_state = &global_state->hwmixer;
 
 	if (!mixer)
 		return;
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_pipe.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_pipe.c
index ff52c49095f9..1ef26bc63163 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_pipe.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_pipe.c
@@ -24,17 +24,19 @@ int mdp5_pipe_assign(struct drm_atomic_state *s, struct drm_plane *plane,
 {
 	struct msm_drm_private *priv = s->dev->dev_private;
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(priv->kms));
-	struct mdp5_state *state;
+	struct mdp5_global_state *new_global_state, *old_global_state;
 	struct mdp5_hw_pipe_state *old_state, *new_state;
 	int i, j;
 
-	state = mdp5_get_state(s);
-	if (IS_ERR(state))
-		return PTR_ERR(state);
+	new_global_state = mdp5_get_global_state(s);
+	if (IS_ERR(new_global_state))
+		return PTR_ERR(new_global_state);
 
-	/* grab old_state after mdp5_get_state(), since now we hold lock: */
-	old_state = &mdp5_kms->state->hwpipe;
-	new_state = &state->hwpipe;
+	/* grab old_state after mdp5_get_global_state(), since now we hold lock: */
+	old_global_state = mdp5_get_existing_global_state(mdp5_kms);
+
+	old_state = &old_global_state->hwpipe;
+	new_state = &new_global_state->hwpipe;
 
 	for (i = 0; i < mdp5_kms->num_hwpipes; i++) {
 		struct mdp5_hw_pipe *cur = mdp5_kms->hwpipes[i];
@@ -107,7 +109,7 @@ int mdp5_pipe_assign(struct drm_atomic_state *s, struct drm_plane *plane,
 		WARN_ON(r_hwpipe);
 
 		DBG("%s: alloc SMP blocks", (*hwpipe)->name);
-		ret = mdp5_smp_assign(mdp5_kms->smp, &state->smp,
+		ret = mdp5_smp_assign(mdp5_kms->smp, &new_global_state->smp,
 				(*hwpipe)->pipe, blkcfg);
 		if (ret)
 			return -ENOMEM;
@@ -132,7 +134,7 @@ void mdp5_pipe_release(struct drm_atomic_state *s, struct mdp5_hw_pipe *hwpipe)
 {
 	struct msm_drm_private *priv = s->dev->dev_private;
 	struct mdp5_kms *mdp5_kms = to_mdp5_kms(to_mdp_kms(priv->kms));
-	struct mdp5_state *state = mdp5_get_state(s);
+	struct mdp5_global_state *state = mdp5_get_global_state(s);
 	struct mdp5_hw_pipe_state *new_state = &state->hwpipe;
 
 	if (!hwpipe)
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_plane.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_plane.c
index a9f31da7d45a..7d306c5acd09 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_plane.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_plane.c
@@ -46,7 +46,7 @@ static void mdp5_plane_destroy(struct drm_plane *plane)
 {
 	struct mdp5_plane *mdp5_plane = to_mdp5_plane(plane);
 
-	drm_plane_helper_disable(plane);
+	drm_plane_helper_disable(plane, NULL);
 	drm_plane_cleanup(plane);
 
 	kfree(mdp5_plane);
@@ -245,20 +245,6 @@ static const struct drm_plane_funcs mdp5_plane_funcs = {
 		.atomic_print_state = mdp5_plane_atomic_print_state,
 };
 
-static int mdp5_plane_prepare_fb(struct drm_plane *plane,
-				 struct drm_plane_state *new_state)
-{
-	struct mdp5_kms *mdp5_kms = get_kms(plane);
-	struct msm_kms *kms = &mdp5_kms->base.base;
-	struct drm_framebuffer *fb = new_state->fb;
-
-	if (!new_state->fb)
-		return 0;
-
-	DBG("%s: prepare: FB[%u]", plane->name, fb->base.id);
-	return msm_framebuffer_prepare(fb, kms->aspace);
-}
-
 static void mdp5_plane_cleanup_fb(struct drm_plane *plane,
 				  struct drm_plane_state *old_state)
 {
@@ -526,7 +512,7 @@ static void mdp5_plane_atomic_async_update(struct drm_plane *plane,
 	if (plane_enabled(new_state)) {
 		struct mdp5_ctl *ctl;
 		struct mdp5_pipeline *pipeline =
-					mdp5_crtc_get_pipeline(plane->crtc);
+					mdp5_crtc_get_pipeline(new_state->crtc);
 		int ret;
 
 		ret = mdp5_plane_mode_set(plane, new_state->crtc, new_state->fb,
@@ -543,7 +529,7 @@ static void mdp5_plane_atomic_async_update(struct drm_plane *plane,
 }
 
 static const struct drm_plane_helper_funcs mdp5_plane_helper_funcs = {
-		.prepare_fb = mdp5_plane_prepare_fb,
+		.prepare_fb = msm_atomic_prepare_fb,
 		.cleanup_fb = mdp5_plane_cleanup_fb,
 		.atomic_check = mdp5_plane_atomic_check,
 		.atomic_update = mdp5_plane_atomic_update,
@@ -1043,8 +1029,6 @@ static int mdp5_plane_mode_set(struct drm_plane *plane,
 				     src_img_w, src_img_h,
 				     src_x + src_w, src_y, src_w, src_h);
 
-	plane->fb = fb;
-
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/msm/disp/mdp5/mdp5_smp.c b/drivers/gpu/drm/msm/disp/mdp5/mdp5_smp.c
index ae4983d9d0a5..96c2b828dba4 100644
--- a/drivers/gpu/drm/msm/disp/mdp5/mdp5_smp.c
+++ b/drivers/gpu/drm/msm/disp/mdp5/mdp5_smp.c
@@ -340,17 +340,20 @@ void mdp5_smp_dump(struct mdp5_smp *smp, struct drm_printer *p)
 	struct mdp5_kms *mdp5_kms = get_kms(smp);
 	struct mdp5_hw_pipe_state *hwpstate;
 	struct mdp5_smp_state *state;
+	struct mdp5_global_state *global_state;
 	int total = 0, i, j;
 
 	drm_printf(p, "name\tinuse\tplane\n");
 	drm_printf(p, "----\t-----\t-----\n");
 
 	if (drm_can_sleep())
-		drm_modeset_lock(&mdp5_kms->state_lock, NULL);
+		drm_modeset_lock(&mdp5_kms->glob_state_lock, NULL);
+
+	global_state = mdp5_get_existing_global_state(mdp5_kms);
 
 	/* grab these *after* we hold the state_lock */
-	hwpstate = &mdp5_kms->state->hwpipe;
-	state = &mdp5_kms->state->smp;
+	hwpstate = &global_state->hwpipe;
+	state = &global_state->smp;
 
 	for (i = 0; i < mdp5_kms->num_hwpipes; i++) {
 		struct mdp5_hw_pipe *hwpipe = mdp5_kms->hwpipes[i];
@@ -374,7 +377,7 @@ void mdp5_smp_dump(struct mdp5_smp *smp, struct drm_printer *p)
 			bitmap_weight(state->state, smp->blk_cnt));
 
 	if (drm_can_sleep())
-		drm_modeset_unlock(&mdp5_kms->state_lock);
+		drm_modeset_unlock(&mdp5_kms->glob_state_lock);
 }
 
 void mdp5_smp_destroy(struct mdp5_smp *smp)
@@ -384,7 +387,8 @@ void mdp5_smp_destroy(struct mdp5_smp *smp)
 
 struct mdp5_smp *mdp5_smp_init(struct mdp5_kms *mdp5_kms, const struct mdp5_smp_block *cfg)
 {
-	struct mdp5_smp_state *state = &mdp5_kms->state->smp;
+	struct mdp5_smp_state *state;
+	struct mdp5_global_state *global_state;
 	struct mdp5_smp *smp = NULL;
 	int ret;
 
@@ -398,6 +402,9 @@ struct mdp5_smp *mdp5_smp_init(struct mdp5_kms *mdp5_kms, const struct mdp5_smp_
 	smp->blk_cnt = cfg->mmb_count;
 	smp->blk_size = cfg->mmb_size;
 
+	global_state = mdp5_get_existing_global_state(mdp5_kms);
+	state = &global_state->smp;
+
 	/* statically tied MMBs cannot be re-allocated: */
 	bitmap_copy(state->state, cfg->reserved_state, smp->blk_cnt);
 	memcpy(smp->reserved, cfg->reserved, sizeof(smp->reserved));
diff --git a/drivers/gpu/drm/msm/dsi/dsi_host.c b/drivers/gpu/drm/msm/dsi/dsi_host.c
index 8baba30d6c65..29841f440111 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_host.c
+++ b/drivers/gpu/drm/msm/dsi/dsi_host.c
@@ -1036,7 +1036,6 @@ static int dsi_tx_buf_alloc(struct msm_dsi_host *msm_host, int size)
 
 		ret = msm_gem_get_iova(msm_host->tx_gem_obj,
 				priv->kms->aspace, &iova);
-		mutex_unlock(&dev->struct_mutex);
 		if (ret) {
 			pr_err("%s: failed to get iova, %d\n", __func__, ret);
 			return ret;
@@ -1067,9 +1066,20 @@ static int dsi_tx_buf_alloc(struct msm_dsi_host *msm_host, int size)
 static void dsi_tx_buf_free(struct msm_dsi_host *msm_host)
 {
 	struct drm_device *dev = msm_host->dev;
+	struct msm_drm_private *priv;
 
+	/*
+	 * This is possible if we're tearing down before we've had a chance to
+	 * fully initialize. A very real possibility if our probe is deferred,
+	 * in which case we'll hit msm_dsi_host_destroy() without having run
+	 * through the dsi_tx_buf_alloc().
+	 */
+	if (!dev)
+		return;
+
+	priv = dev->dev_private;
 	if (msm_host->tx_gem_obj) {
-		msm_gem_put_iova(msm_host->tx_gem_obj, 0);
+		msm_gem_put_iova(msm_host->tx_gem_obj, priv->kms->aspace);
 		drm_gem_object_put_unlocked(msm_host->tx_gem_obj);
 		msm_host->tx_gem_obj = NULL;
 	}
@@ -1888,7 +1898,7 @@ int msm_dsi_host_register(struct mipi_dsi_host *host, bool check_defer)
 		 * output
 		 */
 		if (check_defer && msm_host->device_node) {
-			if (!of_drm_find_panel(msm_host->device_node))
+			if (IS_ERR(of_drm_find_panel(msm_host->device_node)))
 				if (!of_drm_find_bridge(msm_host->device_node))
 					return -EPROBE_DEFER;
 		}
diff --git a/drivers/gpu/drm/msm/dsi/dsi_manager.c b/drivers/gpu/drm/msm/dsi/dsi_manager.c
index 4cb1cb68878b..d5006d6923e0 100644
--- a/drivers/gpu/drm/msm/dsi/dsi_manager.c
+++ b/drivers/gpu/drm/msm/dsi/dsi_manager.c
@@ -393,7 +393,7 @@ static int dsi_mgr_connector_get_modes(struct drm_connector *connector)
 		ret = dsi_dual_connector_tile_init(connector, id);
 		if (ret)
 			return ret;
-		ret = drm_mode_connector_set_tile_property(connector);
+		ret = drm_connector_set_tile_property(connector);
 		if (ret) {
 			pr_err("%s: set tile property failed, %d\n",
 					__func__, ret);
@@ -684,7 +684,7 @@ struct drm_connector *msm_dsi_manager_connector_init(u8 id)
 	connector->interlace_allowed = 0;
 	connector->doublescan_allowed = 0;
 
-	drm_mode_connector_attach_encoder(connector, msm_dsi->encoder);
+	drm_connector_attach_encoder(connector, msm_dsi->encoder);
 
 	return connector;
 }
@@ -751,12 +751,8 @@ struct drm_connector *msm_dsi_manager_ext_bridge_init(u8 id)
 	connector_list = &dev->mode_config.connector_list;
 
 	list_for_each_entry(connector, connector_list, head) {
-		int i;
-
-		for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-			if (connector->encoder_ids[i] == encoder->base.id)
-				return connector;
-		}
+		if (drm_connector_has_possible_encoder(connector, encoder))
+			return connector;
 	}
 
 	return ERR_PTR(-ENODEV);
diff --git a/drivers/gpu/drm/msm/edp/edp_connector.c b/drivers/gpu/drm/msm/edp/edp_connector.c
index 6f3fc6b0f0a3..058ff92a0207 100644
--- a/drivers/gpu/drm/msm/edp/edp_connector.c
+++ b/drivers/gpu/drm/msm/edp/edp_connector.c
@@ -56,7 +56,7 @@ static int edp_connector_get_modes(struct drm_connector *connector)
 	if (ret)
 		return ret;
 
-	drm_mode_connector_update_edid_property(connector, drm_edid);
+	drm_connector_update_edid_property(connector, drm_edid);
 	if (drm_edid)
 		ret = drm_add_edid_modes(connector, drm_edid);
 
@@ -134,7 +134,7 @@ struct drm_connector *msm_edp_connector_init(struct msm_edp *edp)
 	connector->interlace_allowed = false;
 	connector->doublescan_allowed = false;
 
-	drm_mode_connector_attach_encoder(connector, edp->encoder);
+	drm_connector_attach_encoder(connector, edp->encoder);
 
 	return connector;
 }
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi.c b/drivers/gpu/drm/msm/hdmi/hdmi.c
index e63dc0fb55f8..c79659ca5706 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi.c
@@ -157,8 +157,10 @@ static struct hdmi *msm_hdmi_init(struct platform_device *pdev)
 		hdmi->qfprom_mmio = NULL;
 	}
 
-	hdmi->hpd_regs = devm_kzalloc(&pdev->dev, sizeof(hdmi->hpd_regs[0]) *
-			config->hpd_reg_cnt, GFP_KERNEL);
+	hdmi->hpd_regs = devm_kcalloc(&pdev->dev,
+				      config->hpd_reg_cnt,
+				      sizeof(hdmi->hpd_regs[0]),
+				      GFP_KERNEL);
 	if (!hdmi->hpd_regs) {
 		ret = -ENOMEM;
 		goto fail;
@@ -178,8 +180,10 @@ static struct hdmi *msm_hdmi_init(struct platform_device *pdev)
 		hdmi->hpd_regs[i] = reg;
 	}
 
-	hdmi->pwr_regs = devm_kzalloc(&pdev->dev, sizeof(hdmi->pwr_regs[0]) *
-			config->pwr_reg_cnt, GFP_KERNEL);
+	hdmi->pwr_regs = devm_kcalloc(&pdev->dev,
+				      config->pwr_reg_cnt,
+				      sizeof(hdmi->pwr_regs[0]),
+				      GFP_KERNEL);
 	if (!hdmi->pwr_regs) {
 		ret = -ENOMEM;
 		goto fail;
@@ -199,8 +203,10 @@ static struct hdmi *msm_hdmi_init(struct platform_device *pdev)
 		hdmi->pwr_regs[i] = reg;
 	}
 
-	hdmi->hpd_clks = devm_kzalloc(&pdev->dev, sizeof(hdmi->hpd_clks[0]) *
-			config->hpd_clk_cnt, GFP_KERNEL);
+	hdmi->hpd_clks = devm_kcalloc(&pdev->dev,
+				      config->hpd_clk_cnt,
+				      sizeof(hdmi->hpd_clks[0]),
+				      GFP_KERNEL);
 	if (!hdmi->hpd_clks) {
 		ret = -ENOMEM;
 		goto fail;
@@ -219,8 +225,10 @@ static struct hdmi *msm_hdmi_init(struct platform_device *pdev)
 		hdmi->hpd_clks[i] = clk;
 	}
 
-	hdmi->pwr_clks = devm_kzalloc(&pdev->dev, sizeof(hdmi->pwr_clks[0]) *
-			config->pwr_clk_cnt, GFP_KERNEL);
+	hdmi->pwr_clks = devm_kcalloc(&pdev->dev,
+				      config->pwr_clk_cnt,
+				      sizeof(hdmi->pwr_clks[0]),
+				      GFP_KERNEL);
 	if (!hdmi->pwr_clks) {
 		ret = -ENOMEM;
 		goto fail;
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi_connector.c b/drivers/gpu/drm/msm/hdmi/hdmi_connector.c
index c0848dfedd50..e9c9a0af508e 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi_connector.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi_connector.c
@@ -392,7 +392,7 @@ static int msm_hdmi_connector_get_modes(struct drm_connector *connector)
 	hdmi_write(hdmi, REG_HDMI_CTRL, hdmi_ctrl);
 
 	hdmi->hdmi_mode = drm_detect_hdmi_monitor(edid);
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 
 	if (edid) {
 		ret = drm_add_edid_modes(connector, edid);
@@ -477,7 +477,7 @@ struct drm_connector *msm_hdmi_connector_init(struct hdmi *hdmi)
 		return ERR_PTR(ret);
 	}
 
-	drm_mode_connector_attach_encoder(connector, hdmi->encoder);
+	drm_connector_attach_encoder(connector, hdmi->encoder);
 
 	return connector;
 }
diff --git a/drivers/gpu/drm/msm/hdmi/hdmi_phy.c b/drivers/gpu/drm/msm/hdmi/hdmi_phy.c
index 5e631392dc85..4157722d6b4d 100644
--- a/drivers/gpu/drm/msm/hdmi/hdmi_phy.c
+++ b/drivers/gpu/drm/msm/hdmi/hdmi_phy.c
@@ -21,12 +21,12 @@ static int msm_hdmi_phy_resource_init(struct hdmi_phy *phy)
 	struct device *dev = &phy->pdev->dev;
 	int i, ret;
 
-	phy->regs = devm_kzalloc(dev, sizeof(phy->regs[0]) * cfg->num_regs,
+	phy->regs = devm_kcalloc(dev, cfg->num_regs, sizeof(phy->regs[0]),
 				 GFP_KERNEL);
 	if (!phy->regs)
 		return -ENOMEM;
 
-	phy->clks = devm_kzalloc(dev, sizeof(phy->clks[0]) * cfg->num_clks,
+	phy->clks = devm_kcalloc(dev, cfg->num_clks, sizeof(phy->clks[0]),
 				 GFP_KERNEL);
 	if (!phy->clks)
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/msm/msm_atomic.c b/drivers/gpu/drm/msm/msm_atomic.c
index bf5f8c39f34d..f0635c3da7f4 100644
--- a/drivers/gpu/drm/msm/msm_atomic.c
+++ b/drivers/gpu/drm/msm/msm_atomic.c
@@ -16,69 +16,8 @@
  */
 
 #include "msm_drv.h"
-#include "msm_kms.h"
 #include "msm_gem.h"
-#include "msm_fence.h"
-
-struct msm_commit {
-	struct drm_device *dev;
-	struct drm_atomic_state *state;
-	struct work_struct work;
-	uint32_t crtc_mask;
-};
-
-static void commit_worker(struct work_struct *work);
-
-/* block until specified crtcs are no longer pending update, and
- * atomically mark them as pending update
- */
-static int start_atomic(struct msm_drm_private *priv, uint32_t crtc_mask)
-{
-	int ret;
-
-	spin_lock(&priv->pending_crtcs_event.lock);
-	ret = wait_event_interruptible_locked(priv->pending_crtcs_event,
-			!(priv->pending_crtcs & crtc_mask));
-	if (ret == 0) {
-		DBG("start: %08x", crtc_mask);
-		priv->pending_crtcs |= crtc_mask;
-	}
-	spin_unlock(&priv->pending_crtcs_event.lock);
-
-	return ret;
-}
-
-/* clear specified crtcs (no longer pending update)
- */
-static void end_atomic(struct msm_drm_private *priv, uint32_t crtc_mask)
-{
-	spin_lock(&priv->pending_crtcs_event.lock);
-	DBG("end: %08x", crtc_mask);
-	priv->pending_crtcs &= ~crtc_mask;
-	wake_up_all_locked(&priv->pending_crtcs_event);
-	spin_unlock(&priv->pending_crtcs_event.lock);
-}
-
-static struct msm_commit *commit_init(struct drm_atomic_state *state)
-{
-	struct msm_commit *c = kzalloc(sizeof(*c), GFP_KERNEL);
-
-	if (!c)
-		return NULL;
-
-	c->dev = state->dev;
-	c->state = state;
-
-	INIT_WORK(&c->work, commit_worker);
-
-	return c;
-}
-
-static void commit_destroy(struct msm_commit *c)
-{
-	end_atomic(c->dev->dev_private, c->crtc_mask);
-	kfree(c);
-}
+#include "msm_kms.h"
 
 static void msm_atomic_wait_for_commit_done(struct drm_device *dev,
 		struct drm_atomic_state *old_state)
@@ -97,195 +36,48 @@ static void msm_atomic_wait_for_commit_done(struct drm_device *dev,
 	}
 }
 
-/* The (potentially) asynchronous part of the commit.  At this point
- * nothing can fail short of armageddon.
- */
-static void complete_commit(struct msm_commit *c, bool async)
+int msm_atomic_prepare_fb(struct drm_plane *plane,
+			  struct drm_plane_state *new_state)
 {
-	struct drm_atomic_state *state = c->state;
-	struct drm_device *dev = state->dev;
-	struct msm_drm_private *priv = dev->dev_private;
+	struct msm_drm_private *priv = plane->dev->dev_private;
 	struct msm_kms *kms = priv->kms;
+	struct drm_gem_object *obj;
+	struct msm_gem_object *msm_obj;
+	struct dma_fence *fence;
 
-	drm_atomic_helper_wait_for_fences(dev, state, false);
-
-	kms->funcs->prepare_commit(kms, state);
-
-	drm_atomic_helper_commit_modeset_disables(dev, state);
-
-	drm_atomic_helper_commit_planes(dev, state, 0);
-
-	drm_atomic_helper_commit_modeset_enables(dev, state);
-
-	/* NOTE: _wait_for_vblanks() only waits for vblank on
-	 * enabled CRTCs.  So we end up faulting when disabling
-	 * due to (potentially) unref'ing the outgoing fb's
-	 * before the vblank when the disable has latched.
-	 *
-	 * But if it did wait on disabled (or newly disabled)
-	 * CRTCs, that would be racy (ie. we could have missed
-	 * the irq.  We need some way to poll for pipe shut
-	 * down.  Or just live with occasionally hitting the
-	 * timeout in the CRTC disable path (which really should
-	 * not be critical path)
-	 */
-
-	msm_atomic_wait_for_commit_done(dev, state);
-
-	drm_atomic_helper_cleanup_planes(dev, state);
+	if (!new_state->fb)
+		return 0;
 
-	kms->funcs->complete_commit(kms, state);
+	obj = msm_framebuffer_bo(new_state->fb, 0);
+	msm_obj = to_msm_bo(obj);
+	fence = reservation_object_get_excl_rcu(msm_obj->resv);
 
-	drm_atomic_state_put(state);
+	drm_atomic_set_fence_for_plane(new_state, fence);
 
-	commit_destroy(c);
-}
-
-static void commit_worker(struct work_struct *work)
-{
-	complete_commit(container_of(work, struct msm_commit, work), true);
+	return msm_framebuffer_prepare(new_state->fb, kms->aspace);
 }
 
-/**
- * drm_atomic_helper_commit - commit validated state object
- * @dev: DRM device
- * @state: the driver state object
- * @nonblock: nonblocking commit
- *
- * This function commits a with drm_atomic_helper_check() pre-validated state
- * object. This can still fail when e.g. the framebuffer reservation fails.
- *
- * RETURNS
- * Zero for success or -errno.
- */
-int msm_atomic_commit(struct drm_device *dev,
-		struct drm_atomic_state *state, bool nonblock)
+void msm_atomic_commit_tail(struct drm_atomic_state *state)
 {
+	struct drm_device *dev = state->dev;
 	struct msm_drm_private *priv = dev->dev_private;
-	struct msm_commit *c;
-	struct drm_crtc *crtc;
-	struct drm_crtc_state *crtc_state;
-	struct drm_plane *plane;
-	struct drm_plane_state *old_plane_state, *new_plane_state;
-	int i, ret;
-
-	ret = drm_atomic_helper_prepare_planes(dev, state);
-	if (ret)
-		return ret;
-
-	/*
-	 * Note that plane->atomic_async_check() should fail if we need
-	 * to re-assign hwpipe or anything that touches global atomic
-	 * state, so we'll never go down the async update path in those
-	 * cases.
-	 */
-	if (state->async_update) {
-		drm_atomic_helper_async_commit(dev, state);
-		drm_atomic_helper_cleanup_planes(dev, state);
-		return 0;
-	}
-
-	c = commit_init(state);
-	if (!c) {
-		ret = -ENOMEM;
-		goto error;
-	}
-
-	/*
-	 * Figure out what crtcs we have:
-	 */
-	for_each_new_crtc_in_state(state, crtc, crtc_state, i)
-		c->crtc_mask |= drm_crtc_mask(crtc);
-
-	/*
-	 * Figure out what fence to wait for:
-	 */
-	for_each_oldnew_plane_in_state(state, plane, old_plane_state, new_plane_state, i) {
-		if ((new_plane_state->fb != old_plane_state->fb) && new_plane_state->fb) {
-			struct drm_gem_object *obj = msm_framebuffer_bo(new_plane_state->fb, 0);
-			struct msm_gem_object *msm_obj = to_msm_bo(obj);
-			struct dma_fence *fence = reservation_object_get_excl_rcu(msm_obj->resv);
+	struct msm_kms *kms = priv->kms;
 
-			drm_atomic_set_fence_for_plane(new_plane_state, fence);
-		}
-	}
+	kms->funcs->prepare_commit(kms, state);
 
-	/*
-	 * Wait for pending updates on any of the same crtc's and then
-	 * mark our set of crtc's as busy:
-	 */
-	ret = start_atomic(dev->dev_private, c->crtc_mask);
-	if (ret)
-		goto err_free;
+	drm_atomic_helper_commit_modeset_disables(dev, state);
 
-	BUG_ON(drm_atomic_helper_swap_state(state, false) < 0);
+	drm_atomic_helper_commit_planes(dev, state, 0);
 
-	/*
-	 * This is the point of no return - everything below never fails except
-	 * when the hw goes bonghits. Which means we can commit the new state on
-	 * the software side now.
-	 *
-	 * swap driver private state while still holding state_lock
-	 */
-	if (to_kms_state(state)->state)
-		priv->kms->funcs->swap_state(priv->kms, state);
+	drm_atomic_helper_commit_modeset_enables(dev, state);
 
-	/*
-	 * Everything below can be run asynchronously without the need to grab
-	 * any modeset locks at all under one conditions: It must be guaranteed
-	 * that the asynchronous work has either been cancelled (if the driver
-	 * supports it, which at least requires that the framebuffers get
-	 * cleaned up with drm_atomic_helper_cleanup_planes()) or completed
-	 * before the new state gets committed on the software side with
-	 * drm_atomic_helper_swap_state().
-	 *
-	 * This scheme allows new atomic state updates to be prepared and
-	 * checked in parallel to the asynchronous completion of the previous
-	 * update. Which is important since compositors need to figure out the
-	 * composition of the next frame right after having submitted the
-	 * current layout.
-	 */
+	msm_atomic_wait_for_commit_done(dev, state);
 
-	drm_atomic_state_get(state);
-	if (nonblock) {
-		queue_work(priv->atomic_wq, &c->work);
-		return 0;
-	}
+	kms->funcs->complete_commit(kms, state);
 
-	complete_commit(c, false);
+	drm_atomic_helper_wait_for_vblanks(dev, state);
 
-	return 0;
+	drm_atomic_helper_commit_hw_done(state);
 
-err_free:
-	kfree(c);
-error:
 	drm_atomic_helper_cleanup_planes(dev, state);
-	return ret;
-}
-
-struct drm_atomic_state *msm_atomic_state_alloc(struct drm_device *dev)
-{
-	struct msm_kms_state *state = kzalloc(sizeof(*state), GFP_KERNEL);
-
-	if (!state || drm_atomic_state_init(dev, &state->base) < 0) {
-		kfree(state);
-		return NULL;
-	}
-
-	return &state->base;
-}
-
-void msm_atomic_state_clear(struct drm_atomic_state *s)
-{
-	struct msm_kms_state *state = to_kms_state(s);
-	drm_atomic_state_default_clear(&state->base);
-	kfree(state->state);
-	state->state = NULL;
-}
-
-void msm_atomic_state_free(struct drm_atomic_state *state)
-{
-	kfree(to_kms_state(state)->state);
-	drm_atomic_state_default_release(state);
-	kfree(state);
 }
diff --git a/drivers/gpu/drm/msm/msm_debugfs.c b/drivers/gpu/drm/msm/msm_debugfs.c
index ba74cb4f94df..1ff3fda245d1 100644
--- a/drivers/gpu/drm/msm/msm_debugfs.c
+++ b/drivers/gpu/drm/msm/msm_debugfs.c
@@ -140,9 +140,6 @@ int msm_debugfs_late_init(struct drm_device *dev)
 	if (ret)
 		return ret;
 	ret = late_init_minor(dev->render);
-	if (ret)
-		return ret;
-	ret = late_init_minor(dev->control);
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/msm/msm_drv.c b/drivers/gpu/drm/msm/msm_drv.c
index 30cd514d8f7c..021a0b6f9a59 100644
--- a/drivers/gpu/drm/msm/msm_drv.c
+++ b/drivers/gpu/drm/msm/msm_drv.c
@@ -41,10 +41,11 @@ static const struct drm_mode_config_funcs mode_config_funcs = {
 	.fb_create = msm_framebuffer_create,
 	.output_poll_changed = drm_fb_helper_output_poll_changed,
 	.atomic_check = drm_atomic_helper_check,
-	.atomic_commit = msm_atomic_commit,
-	.atomic_state_alloc = msm_atomic_state_alloc,
-	.atomic_state_clear = msm_atomic_state_clear,
-	.atomic_state_free = msm_atomic_state_free,
+	.atomic_commit = drm_atomic_helper_commit,
+};
+
+static const struct drm_mode_config_helper_funcs mode_config_helper_funcs = {
+	.atomic_commit_tail = msm_atomic_commit_tail,
 };
 
 #ifdef CONFIG_DRM_MSM_REGISTER_LOGGING
@@ -384,7 +385,6 @@ static int msm_drm_init(struct device *dev, struct drm_driver *drv)
 
 	priv->wq = alloc_ordered_workqueue("msm", 0);
 	priv->atomic_wq = alloc_ordered_workqueue("msm:atomic", 0);
-	init_waitqueue_head(&priv->pending_crtcs_event);
 
 	INIT_LIST_HEAD(&priv->inactive_list);
 	INIT_LIST_HEAD(&priv->vblank_ctrl.event_list);
@@ -442,6 +442,7 @@ static int msm_drm_init(struct device *dev, struct drm_driver *drv)
 	}
 
 	ddev->mode_config.funcs = &mode_config_funcs;
+	ddev->mode_config.helper_private = &mode_config_helper_funcs;
 
 	ret = drm_vblank_init(ddev, priv->num_crtcs);
 	if (ret < 0) {
diff --git a/drivers/gpu/drm/msm/msm_drv.h b/drivers/gpu/drm/msm/msm_drv.h
index 48ed5b9a8580..b2da1fbf81e0 100644
--- a/drivers/gpu/drm/msm/msm_drv.h
+++ b/drivers/gpu/drm/msm/msm_drv.h
@@ -117,10 +117,6 @@ struct msm_drm_private {
 	struct workqueue_struct *wq;
 	struct workqueue_struct *atomic_wq;
 
-	/* crtcs pending async atomic updates: */
-	uint32_t pending_crtcs;
-	wait_queue_head_t pending_crtcs_event;
-
 	unsigned int num_planes;
 	struct drm_plane *planes[16];
 
@@ -160,8 +156,9 @@ struct msm_format {
 	uint32_t pixel_format;
 };
 
-int msm_atomic_commit(struct drm_device *dev,
-		struct drm_atomic_state *state, bool nonblock);
+int msm_atomic_prepare_fb(struct drm_plane *plane,
+			  struct drm_plane_state *new_state);
+void msm_atomic_commit_tail(struct drm_atomic_state *state);
 struct drm_atomic_state *msm_atomic_state_alloc(struct drm_device *dev);
 void msm_atomic_state_clear(struct drm_atomic_state *state);
 void msm_atomic_state_free(struct drm_atomic_state *state);
diff --git a/drivers/gpu/drm/msm/msm_fb.c b/drivers/gpu/drm/msm/msm_fb.c
index 7a16242bf8bf..2a7348aeb38d 100644
--- a/drivers/gpu/drm/msm/msm_fb.c
+++ b/drivers/gpu/drm/msm/msm_fb.c
@@ -17,6 +17,7 @@
 
 #include <drm/drm_crtc.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 
 #include "msm_drv.h"
 #include "msm_kms.h"
@@ -25,49 +26,20 @@
 struct msm_framebuffer {
 	struct drm_framebuffer base;
 	const struct msm_format *format;
-	struct drm_gem_object *planes[MAX_PLANE];
 };
 #define to_msm_framebuffer(x) container_of(x, struct msm_framebuffer, base)
 
 static struct drm_framebuffer *msm_framebuffer_init(struct drm_device *dev,
 		const struct drm_mode_fb_cmd2 *mode_cmd, struct drm_gem_object **bos);
 
-static int msm_framebuffer_create_handle(struct drm_framebuffer *fb,
-		struct drm_file *file_priv,
-		unsigned int *handle)
-{
-	struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb);
-	return drm_gem_handle_create(file_priv,
-			msm_fb->planes[0], handle);
-}
-
-static void msm_framebuffer_destroy(struct drm_framebuffer *fb)
-{
-	struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb);
-	int i, n = fb->format->num_planes;
-
-	DBG("destroy: FB ID: %d (%p)", fb->base.id, fb);
-
-	drm_framebuffer_cleanup(fb);
-
-	for (i = 0; i < n; i++) {
-		struct drm_gem_object *bo = msm_fb->planes[i];
-
-		drm_gem_object_put_unlocked(bo);
-	}
-
-	kfree(msm_fb);
-}
-
 static const struct drm_framebuffer_funcs msm_framebuffer_funcs = {
-	.create_handle = msm_framebuffer_create_handle,
-	.destroy = msm_framebuffer_destroy,
+	.create_handle = drm_gem_fb_create_handle,
+	.destroy = drm_gem_fb_destroy,
 };
 
 #ifdef CONFIG_DEBUG_FS
 void msm_framebuffer_describe(struct drm_framebuffer *fb, struct seq_file *m)
 {
-	struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb);
 	int i, n = fb->format->num_planes;
 
 	seq_printf(m, "fb: %dx%d@%4.4s (%2d, ID:%d)\n",
@@ -77,7 +49,7 @@ void msm_framebuffer_describe(struct drm_framebuffer *fb, struct seq_file *m)
 	for (i = 0; i < n; i++) {
 		seq_printf(m, "   %d: offset=%d pitch=%d, obj: ",
 				i, fb->offsets[i], fb->pitches[i]);
-		msm_gem_describe(msm_fb->planes[i], m);
+		msm_gem_describe(fb->obj[i], m);
 	}
 }
 #endif
@@ -90,12 +62,11 @@ void msm_framebuffer_describe(struct drm_framebuffer *fb, struct seq_file *m)
 int msm_framebuffer_prepare(struct drm_framebuffer *fb,
 		struct msm_gem_address_space *aspace)
 {
-	struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb);
 	int ret, i, n = fb->format->num_planes;
 	uint64_t iova;
 
 	for (i = 0; i < n; i++) {
-		ret = msm_gem_get_iova(msm_fb->planes[i], aspace, &iova);
+		ret = msm_gem_get_iova(fb->obj[i], aspace, &iova);
 		DBG("FB[%u]: iova[%d]: %08llx (%d)", fb->base.id, i, iova, ret);
 		if (ret)
 			return ret;
@@ -107,26 +78,23 @@ int msm_framebuffer_prepare(struct drm_framebuffer *fb,
 void msm_framebuffer_cleanup(struct drm_framebuffer *fb,
 		struct msm_gem_address_space *aspace)
 {
-	struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb);
 	int i, n = fb->format->num_planes;
 
 	for (i = 0; i < n; i++)
-		msm_gem_put_iova(msm_fb->planes[i], aspace);
+		msm_gem_put_iova(fb->obj[i], aspace);
 }
 
 uint32_t msm_framebuffer_iova(struct drm_framebuffer *fb,
 		struct msm_gem_address_space *aspace, int plane)
 {
-	struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb);
-	if (!msm_fb->planes[plane])
+	if (!fb->obj[plane])
 		return 0;
-	return msm_gem_iova(msm_fb->planes[plane], aspace) + fb->offsets[plane];
+	return msm_gem_iova(fb->obj[plane], aspace) + fb->offsets[plane];
 }
 
 struct drm_gem_object *msm_framebuffer_bo(struct drm_framebuffer *fb, int plane)
 {
-	struct msm_framebuffer *msm_fb = to_msm_framebuffer(fb);
-	return msm_fb->planes[plane];
+	return drm_gem_fb_get_obj(fb, plane);
 }
 
 const struct msm_format *msm_framebuffer_format(struct drm_framebuffer *fb)
@@ -202,7 +170,7 @@ static struct drm_framebuffer *msm_framebuffer_init(struct drm_device *dev,
 
 	msm_fb->format = format;
 
-	if (n > ARRAY_SIZE(msm_fb->planes)) {
+	if (n > ARRAY_SIZE(fb->obj)) {
 		ret = -EINVAL;
 		goto fail;
 	}
@@ -221,7 +189,7 @@ static struct drm_framebuffer *msm_framebuffer_init(struct drm_device *dev,
 			goto fail;
 		}
 
-		msm_fb->planes[i] = bos[i];
+		msm_fb->base.obj[i] = bos[i];
 	}
 
 	drm_helper_mode_fill_fb_struct(dev, fb, mode_cmd);
diff --git a/drivers/gpu/drm/msm/msm_kms.h b/drivers/gpu/drm/msm/msm_kms.h
index aaa329dc020e..dfd92947de2c 100644
--- a/drivers/gpu/drm/msm/msm_kms.h
+++ b/drivers/gpu/drm/msm/msm_kms.h
@@ -40,8 +40,6 @@ struct msm_kms_funcs {
 	irqreturn_t (*irq)(struct msm_kms *kms);
 	int (*enable_vblank)(struct msm_kms *kms, struct drm_crtc *crtc);
 	void (*disable_vblank)(struct msm_kms *kms, struct drm_crtc *crtc);
-	/* swap global atomic state: */
-	void (*swap_state)(struct msm_kms *kms, struct drm_atomic_state *state);
 	/* modeset, bracketing atomic_commit(): */
 	void (*prepare_commit)(struct msm_kms *kms, struct drm_atomic_state *state);
 	void (*complete_commit)(struct msm_kms *kms, struct drm_atomic_state *state);
@@ -80,18 +78,6 @@ struct msm_kms {
 	struct msm_gem_address_space *aspace;
 };
 
-/**
- * Subclass of drm_atomic_state, to allow kms backend to have driver
- * private global state.  The kms backend can do whatever it wants
- * with the ->state ptr.  On ->atomic_state_clear() the ->state ptr
- * is kfree'd and set back to NULL.
- */
-struct msm_kms_state {
-	struct drm_atomic_state base;
-	void *state;
-};
-#define to_kms_state(x) container_of(x, struct msm_kms_state, base)
-
 static inline void msm_kms_init(struct msm_kms *kms,
 		const struct msm_kms_funcs *funcs)
 {
diff --git a/drivers/gpu/drm/nouveau/Kbuild b/drivers/gpu/drm/nouveau/Kbuild
index 9c0c650655e9..b17843dd050d 100644
--- a/drivers/gpu/drm/nouveau/Kbuild
+++ b/drivers/gpu/drm/nouveau/Kbuild
@@ -38,14 +38,16 @@ nouveau-y += nouveau_vmm.o
 
 # DRM - modesetting
 nouveau-$(CONFIG_DRM_NOUVEAU_BACKLIGHT) += nouveau_backlight.o
+nouveau-y += nouveau_bios.o
 nouveau-y += nouveau_connector.o
 nouveau-y += nouveau_display.o
-nouveau-y += nv50_display.o
 nouveau-y += nouveau_dp.o
 nouveau-y += nouveau_fbcon.o
 nouveau-y += nv04_fbcon.o
 nouveau-y += nv50_fbcon.o
 nouveau-y += nvc0_fbcon.o
+include $(src)/dispnv04/Kbuild
+include $(src)/dispnv50/Kbuild
 
 # DRM - command submission
 nouveau-y += nouveau_abi16.o
@@ -59,8 +61,4 @@ nouveau-y += nv50_fence.o
 nouveau-y += nv84_fence.o
 nouveau-y += nvc0_fence.o
 
-# DRM - prehistoric modesetting (NV04-G7x)
-nouveau-y += nouveau_bios.o
-include $(src)/dispnv04/Kbuild
-
 obj-$(CONFIG_DRM_NOUVEAU) += nouveau.o
diff --git a/drivers/gpu/drm/nouveau/dispnv04/crtc.c b/drivers/gpu/drm/nouveau/dispnv04/crtc.c
index 6aa6ee16dcbd..2c569e264df3 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/crtc.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/crtc.c
@@ -1017,7 +1017,7 @@ nv04_crtc_cursor_set(struct drm_crtc *crtc, struct drm_file *file_priv,
 	nv_crtc->cursor.set_offset(nv_crtc, nv_crtc->cursor.offset);
 	nv_crtc->cursor.show(nv_crtc, true);
 out:
-	drm_gem_object_unreference_unlocked(gem);
+	drm_gem_object_put_unlocked(gem);
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/nouveau/dispnv04/dac.c b/drivers/gpu/drm/nouveau/dispnv04/dac.c
index 4feab0a5419d..e7af95d37ddb 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/dac.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/dac.c
@@ -556,6 +556,6 @@ nv04_dac_create(struct drm_connector *connector, struct dcb_output *entry)
 	encoder->possible_crtcs = entry->heads;
 	encoder->possible_clones = 0;
 
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/dispnv04/dfp.c b/drivers/gpu/drm/nouveau/dispnv04/dfp.c
index 9805d2cdc1a1..73d41abbb510 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/dfp.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/dfp.c
@@ -716,6 +716,6 @@ nv04_dfp_create(struct drm_connector *connector, struct dcb_output *entry)
 	    entry->location != DCB_LOC_ON_CHIP)
 		nv04_tmds_slave_init(encoder);
 
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/dispnv04/tvnv04.c b/drivers/gpu/drm/nouveau/dispnv04/tvnv04.c
index 01664357d3e1..de4490b4ed30 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/tvnv04.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/tvnv04.c
@@ -244,7 +244,7 @@ nv04_tv_create(struct drm_connector *connector, struct dcb_output *entry)
 
 	/* Attach it to the specified connector. */
 	get_slave_funcs(encoder)->create_resources(encoder, connector);
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	return 0;
 
diff --git a/drivers/gpu/drm/nouveau/dispnv04/tvnv17.c b/drivers/gpu/drm/nouveau/dispnv04/tvnv17.c
index 6d99f11fee4e..6a4ca139cf5d 100644
--- a/drivers/gpu/drm/nouveau/dispnv04/tvnv17.c
+++ b/drivers/gpu/drm/nouveau/dispnv04/tvnv17.c
@@ -821,6 +821,6 @@ nv17_tv_create(struct drm_connector *connector, struct dcb_output *entry)
 	encoder->possible_clones = 0;
 
 	nv17_tv_create_resources(encoder, connector);
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/dispnv50/Kbuild b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
new file mode 100644
index 000000000000..849b0f45afb8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/Kbuild
@@ -0,0 +1,51 @@
+nouveau-y += dispnv50/disp.o
+nouveau-y += dispnv50/lut.o
+
+nouveau-y += dispnv50/core.o
+nouveau-y += dispnv50/core507d.o
+nouveau-y += dispnv50/core827d.o
+nouveau-y += dispnv50/core907d.o
+nouveau-y += dispnv50/core917d.o
+nouveau-y += dispnv50/corec37d.o
+
+nouveau-y += dispnv50/dac507d.o
+nouveau-y += dispnv50/dac907d.o
+
+nouveau-y += dispnv50/pior507d.o
+
+nouveau-y += dispnv50/sor507d.o
+nouveau-y += dispnv50/sor907d.o
+nouveau-y += dispnv50/sorc37d.o
+
+nouveau-y += dispnv50/head.o
+nouveau-y += dispnv50/head507d.o
+nouveau-y += dispnv50/head827d.o
+nouveau-y += dispnv50/head907d.o
+nouveau-y += dispnv50/head917d.o
+nouveau-y += dispnv50/headc37d.o
+
+nouveau-y += dispnv50/wimm.o
+nouveau-y += dispnv50/wimmc37b.o
+
+nouveau-y += dispnv50/wndw.o
+nouveau-y += dispnv50/wndwc37e.o
+
+nouveau-y += dispnv50/base.o
+nouveau-y += dispnv50/base507c.o
+nouveau-y += dispnv50/base827c.o
+nouveau-y += dispnv50/base907c.o
+nouveau-y += dispnv50/base917c.o
+
+nouveau-y += dispnv50/curs.o
+nouveau-y += dispnv50/curs507a.o
+nouveau-y += dispnv50/curs907a.o
+nouveau-y += dispnv50/cursc37a.o
+
+nouveau-y += dispnv50/oimm.o
+nouveau-y += dispnv50/oimm507b.o
+
+nouveau-y += dispnv50/ovly.o
+nouveau-y += dispnv50/ovly507e.o
+nouveau-y += dispnv50/ovly827e.o
+nouveau-y += dispnv50/ovly907e.o
+nouveau-y += dispnv50/ovly917e.o
diff --git a/drivers/gpu/drm/nouveau/dispnv50/atom.h b/drivers/gpu/drm/nouveau/dispnv50/atom.h
new file mode 100644
index 000000000000..908feb1fc60f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/atom.h
@@ -0,0 +1,222 @@
+#ifndef __NV50_KMS_ATOM_H__
+#define __NV50_KMS_ATOM_H__
+#define nv50_atom(p) container_of((p), struct nv50_atom, state)
+#include <drm/drm_atomic.h>
+
+struct nv50_atom {
+	struct drm_atomic_state state;
+
+	struct list_head outp;
+	bool lock_core;
+	bool flush_disable;
+};
+
+#define nv50_head_atom(p) container_of((p), struct nv50_head_atom, state)
+
+struct nv50_head_atom {
+	struct drm_crtc_state state;
+
+	struct {
+		u32 mask;
+		u32 olut;
+	} wndw;
+
+	struct {
+		u16 iW;
+		u16 iH;
+		u16 oW;
+		u16 oH;
+	} view;
+
+	struct nv50_head_mode {
+		bool interlace;
+		u32 clock;
+		struct {
+			u16 active;
+			u16 synce;
+			u16 blanke;
+			u16 blanks;
+		} h;
+		struct {
+			u32 active;
+			u16 synce;
+			u16 blanke;
+			u16 blanks;
+			u16 blank2s;
+			u16 blank2e;
+			u16 blankus;
+		} v;
+	} mode;
+
+	struct {
+		bool visible;
+		u32 handle;
+		u64 offset:40;
+		u8 buffer:1;
+		u8 mode:4;
+		u8 size:2;
+		u8 range:2;
+		u8 output_mode:2;
+	} olut;
+
+	struct {
+		bool visible;
+		u32 handle;
+		u64 offset:40;
+		u8  format;
+		u8  kind:7;
+		u8  layout:1;
+		u8  blockh:4;
+		u16 blocks:12;
+		u32 pitch:20;
+		u16 x;
+		u16 y;
+		u16 w;
+		u16 h;
+	} core;
+
+	struct {
+		bool visible;
+		u32 handle;
+		u64 offset:40;
+		u8  layout:2;
+		u8  format:8;
+	} curs;
+
+	struct {
+		u8  depth;
+		u8  cpp;
+		u16 x;
+		u16 y;
+		u16 w;
+		u16 h;
+	} base;
+
+	struct {
+		u8 cpp;
+	} ovly;
+
+	struct {
+		bool enable:1;
+		u8 bits:2;
+		u8 mode:4;
+	} dither;
+
+	struct {
+		struct {
+			u16 cos:12;
+			u16 sin:12;
+		} sat;
+	} procamp;
+
+	struct {
+		u8 nhsync:1;
+		u8 nvsync:1;
+		u8 depth:4;
+	} or;
+
+	union nv50_head_atom_mask {
+		struct {
+			bool olut:1;
+			bool core:1;
+			bool curs:1;
+			bool view:1;
+			bool mode:1;
+			bool base:1;
+			bool ovly:1;
+			bool dither:1;
+			bool procamp:1;
+			bool or:1;
+		};
+		u16 mask;
+	} set, clr;
+};
+
+static inline struct nv50_head_atom *
+nv50_head_atom_get(struct drm_atomic_state *state, struct drm_crtc *crtc)
+{
+	struct drm_crtc_state *statec = drm_atomic_get_crtc_state(state, crtc);
+	if (IS_ERR(statec))
+		return (void *)statec;
+	return nv50_head_atom(statec);
+}
+
+#define nv50_wndw_atom(p) container_of((p), struct nv50_wndw_atom, state)
+
+struct nv50_wndw_atom {
+	struct drm_plane_state state;
+
+	struct drm_property_blob *ilut;
+	bool visible;
+
+	struct {
+		u32  handle;
+		u16  offset:12;
+		bool awaken:1;
+	} ntfy;
+
+	struct {
+		u32 handle;
+		u16 offset:12;
+		u32 acquire;
+		u32 release;
+	} sema;
+
+	struct {
+		u32 handle;
+		struct {
+			u64 offset:40;
+			u8  buffer:1;
+			u8  enable:2;
+			u8  mode:4;
+			u8  size:2;
+			u8  range:2;
+			u8  output_mode:2;
+		} i;
+	} xlut;
+
+	struct {
+		u8  mode:2;
+		u8  interval:4;
+
+		u8  colorspace:2;
+		u8  format;
+		u8  kind:7;
+		u8  layout:1;
+		u8  blockh:4;
+		u16 blocks[3];
+		u32 pitch[3];
+		u16 w;
+		u16 h;
+
+		u32 handle[6];
+		u64 offset[6];
+	} image;
+
+	struct {
+		u16 sx;
+		u16 sy;
+		u16 sw;
+		u16 sh;
+		u16 dw;
+		u16 dh;
+	} scale;
+
+	struct {
+		u16 x;
+		u16 y;
+	} point;
+
+	union nv50_wndw_atom_mask {
+		struct {
+			bool ntfy:1;
+			bool sema:1;
+			bool xlut:1;
+			bool image:1;
+			bool scale:1;
+			bool point:1;
+		};
+		u8 mask;
+	} set, clr;
+};
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base.c b/drivers/gpu/drm/nouveau/dispnv50/base.c
new file mode 100644
index 000000000000..7c752acf2b48
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/base.c
@@ -0,0 +1,53 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "base.h"
+
+#include <nvif/class.h>
+
+int
+nv50_base_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
+{
+	struct {
+		s32 oclass;
+		int version;
+		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+	} bases[] = {
+		{ GK110_DISP_BASE_CHANNEL_DMA, 0, base917c_new },
+		{ GK104_DISP_BASE_CHANNEL_DMA, 0, base917c_new },
+		{ GF110_DISP_BASE_CHANNEL_DMA, 0, base907c_new },
+		{ GT214_DISP_BASE_CHANNEL_DMA, 0, base827c_new },
+		{ GT200_DISP_BASE_CHANNEL_DMA, 0, base827c_new },
+		{   G82_DISP_BASE_CHANNEL_DMA, 0, base827c_new },
+		{  NV50_DISP_BASE_CHANNEL_DMA, 0, base507c_new },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid;
+
+	cid = nvif_mclass(&disp->disp->object, bases);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported base class\n");
+		return cid;
+	}
+
+	return bases[cid].new(drm, head, bases[cid].oclass, pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base.h b/drivers/gpu/drm/nouveau/dispnv50/base.h
new file mode 100644
index 000000000000..e7f14f230f35
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/base.h
@@ -0,0 +1,31 @@
+#ifndef __NV50_KMS_BASE_H__
+#define __NV50_KMS_BASE_H__
+#include "wndw.h"
+
+int base507c_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+int base507c_new_(const struct nv50_wndw_func *, const u32 *format,
+		  struct nouveau_drm *, int head, s32 oclass,
+		  u32 interlock_data, struct nv50_wndw **);
+extern const u32 base507c_format[];
+int base507c_acquire(struct nv50_wndw *, struct nv50_wndw_atom *,
+		     struct nv50_head_atom *);
+void base507c_release(struct nv50_wndw *, struct nv50_wndw_atom *,
+		      struct nv50_head_atom *);
+void base507c_sema_set(struct nv50_wndw *, struct nv50_wndw_atom *);
+void base507c_sema_clr(struct nv50_wndw *);
+void base507c_ntfy_set(struct nv50_wndw *, struct nv50_wndw_atom *);
+void base507c_ntfy_clr(struct nv50_wndw *);
+void base507c_xlut_set(struct nv50_wndw *, struct nv50_wndw_atom *);
+void base507c_xlut_clr(struct nv50_wndw *);
+void base507c_image_clr(struct nv50_wndw *);
+void base507c_update(struct nv50_wndw *, u32 *);
+
+int base827c_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+
+int base907c_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+extern const struct nv50_wndw_func base907c;
+
+int base917c_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+
+int nv50_base_new(struct nouveau_drm *, int head, struct nv50_wndw **);
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base507c.c b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
new file mode 100644
index 000000000000..d5e295ca2caa
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/base507c.c
@@ -0,0 +1,286 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "base.h"
+
+#include <nvif/cl507c.h>
+#include <nvif/event.h>
+
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_plane_helper.h>
+#include "nouveau_bo.h"
+
+void
+base507c_update(struct nv50_wndw *wndw, u32 *interlock)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x0080, 1);
+		evo_data(push, interlock[NV50_DISP_INTERLOCK_CORE]);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+base507c_image_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 4))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+base507c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 10))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, asyw->image.mode << 8 |
+			       asyw->image.interval << 4);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, asyw->image.handle[0]);
+		evo_mthd(push, 0x0800, 5);
+		evo_data(push, asyw->image.offset[0] >> 8);
+		evo_data(push, 0x00000000);
+		evo_data(push, asyw->image.h << 16 | asyw->image.w);
+		evo_data(push, asyw->image.layout << 20 |
+			       (asyw->image.pitch[0] >> 8) << 8 |
+			       asyw->image.blocks[0] << 8 |
+			       asyw->image.blockh);
+		evo_data(push, asyw->image.kind << 16 |
+			       asyw->image.format << 8);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+base507c_xlut_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x00e0, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+base507c_xlut_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x00e0, 1);
+		evo_data(push, 0x40000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+int
+base507c_ntfy_wait_begun(struct nouveau_bo *bo, u32 offset,
+			 struct nvif_device *device)
+{
+	s64 time = nvif_msec(device, 2000ULL,
+		u32 data = nouveau_bo_rd32(bo, offset / 4);
+		if ((data & 0xc0000000) == 0x40000000)
+			break;
+		usleep_range(1, 2);
+	);
+	return time < 0 ? time : 0;
+}
+
+void
+base507c_ntfy_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x00a4, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+base507c_ntfy_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 3))) {
+		evo_mthd(push, 0x00a0, 2);
+		evo_data(push, asyw->ntfy.awaken << 30 | asyw->ntfy.offset);
+		evo_data(push, asyw->ntfy.handle);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+base507c_ntfy_reset(struct nouveau_bo *bo, u32 offset)
+{
+	nouveau_bo_wr32(bo, offset / 4, 0x00000000);
+}
+
+void
+base507c_sema_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x0094, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+base507c_sema_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 5))) {
+		evo_mthd(push, 0x0088, 4);
+		evo_data(push, asyw->sema.offset);
+		evo_data(push, asyw->sema.acquire);
+		evo_data(push, asyw->sema.release);
+		evo_data(push, asyw->sema.handle);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+base507c_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+	asyh->base.cpp = 0;
+}
+
+int
+base507c_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+	const struct drm_framebuffer *fb = asyw->state.fb;
+	int ret;
+
+	if (!fb->format->depth)
+		return -EINVAL;
+
+	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  false, true);
+	if (ret)
+		return ret;
+
+	if (!wndw->func->ilut) {
+		if ((asyh->base.cpp != 1) ^ (fb->format->cpp[0] != 1))
+			asyh->state.color_mgmt_changed = true;
+	}
+
+	asyh->base.depth = fb->format->depth;
+	asyh->base.cpp = fb->format->cpp[0];
+	asyh->base.x = asyw->state.src.x1 >> 16;
+	asyh->base.y = asyw->state.src.y1 >> 16;
+	asyh->base.w = asyw->state.fb->width;
+	asyh->base.h = asyw->state.fb->height;
+	return 0;
+}
+
+const u32
+base507c_format[] = {
+	DRM_FORMAT_C8,
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_ARGB1555,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ABGR2101010,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_ABGR8888,
+	0
+};
+
+static const struct nv50_wndw_func
+base507c = {
+	.acquire = base507c_acquire,
+	.release = base507c_release,
+	.sema_set = base507c_sema_set,
+	.sema_clr = base507c_sema_clr,
+	.ntfy_reset = base507c_ntfy_reset,
+	.ntfy_set = base507c_ntfy_set,
+	.ntfy_clr = base507c_ntfy_clr,
+	.ntfy_wait_begun = base507c_ntfy_wait_begun,
+	.olut_core = 1,
+	.xlut_set = base507c_xlut_set,
+	.xlut_clr = base507c_xlut_clr,
+	.image_set = base507c_image_set,
+	.image_clr = base507c_image_clr,
+	.update = base507c_update,
+};
+
+int
+base507c_new_(const struct nv50_wndw_func *func, const u32 *format,
+	      struct nouveau_drm *drm, int head, s32 oclass, u32 interlock_data,
+	      struct nv50_wndw **pwndw)
+{
+	struct nv50_disp_base_channel_dma_v0 args = {
+		.head = head,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_wndw *wndw;
+	int ret;
+
+	ret = nv50_wndw_new_(func, drm->dev, DRM_PLANE_TYPE_PRIMARY,
+			     "base", head, format, BIT(head),
+			     NV50_DISP_INTERLOCK_BASE, interlock_data, &wndw);
+	if (*pwndw = wndw, ret)
+		return ret;
+
+	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
+			       &oclass, head, &args, sizeof(args),
+			       disp->sync->bo.offset, &wndw->wndw);
+	if (ret) {
+		NV_ERROR(drm, "base%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	ret = nvif_notify_init(&wndw->wndw.base.user, wndw->notify.func,
+			       false, NV50_DISP_BASE_CHANNEL_DMA_V0_NTFY_UEVENT,
+			       &(struct nvif_notify_uevent_req) {},
+			       sizeof(struct nvif_notify_uevent_req),
+			       sizeof(struct nvif_notify_uevent_rep),
+			       &wndw->notify);
+	if (ret)
+		return ret;
+
+	wndw->ntfy = NV50_DISP_BASE_NTFY(wndw->id);
+	wndw->sema = NV50_DISP_BASE_SEM0(wndw->id);
+	wndw->data = 0x00000000;
+	return 0;
+}
+
+int
+base507c_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return base507c_new_(&base507c, base507c_format, drm, head, oclass,
+			     0x00000002 << (head * 8), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base827c.c b/drivers/gpu/drm/nouveau/dispnv50/base827c.c
new file mode 100644
index 000000000000..73646819a0d6
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/base827c.c
@@ -0,0 +1,71 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "base.h"
+
+static void
+base827c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 10))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, asyw->image.mode << 8 |
+			       asyw->image.interval << 4);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, asyw->image.handle[0]);
+		evo_mthd(push, 0x0800, 5);
+		evo_data(push, asyw->image.offset[0] >> 8);
+		evo_data(push, 0x00000000);
+		evo_data(push, asyw->image.h << 16 | asyw->image.w);
+		evo_data(push, asyw->image.layout << 20 |
+			       (asyw->image.pitch[0] >> 8) << 8 |
+			       asyw->image.blocks[0] << 8 |
+			       asyw->image.blockh);
+		evo_data(push, asyw->image.format << 8);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static const struct nv50_wndw_func
+base827c = {
+	.acquire = base507c_acquire,
+	.release = base507c_release,
+	.sema_set = base507c_sema_set,
+	.sema_clr = base507c_sema_clr,
+	.ntfy_reset = base507c_ntfy_reset,
+	.ntfy_set = base507c_ntfy_set,
+	.ntfy_clr = base507c_ntfy_clr,
+	.ntfy_wait_begun = base507c_ntfy_wait_begun,
+	.olut_core = 1,
+	.xlut_set = base507c_xlut_set,
+	.xlut_clr = base507c_xlut_clr,
+	.image_set = base827c_image_set,
+	.image_clr = base507c_image_clr,
+	.update = base507c_update,
+};
+
+int
+base827c_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return base507c_new_(&base827c, base507c_format, drm, head, oclass,
+			     0x00000002 << (head * 8), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base907c.c b/drivers/gpu/drm/nouveau/dispnv50/base907c.c
new file mode 100644
index 000000000000..a562fc94ce59
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/base907c.c
@@ -0,0 +1,110 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "base.h"
+
+static void
+base907c_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 10))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, asyw->image.mode << 8 |
+			       asyw->image.interval << 4);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, asyw->image.handle[0]);
+		evo_mthd(push, 0x0400, 5);
+		evo_data(push, asyw->image.offset[0] >> 8);
+		evo_data(push, 0x00000000);
+		evo_data(push, asyw->image.h << 16 | asyw->image.w);
+		evo_data(push, asyw->image.layout << 24 |
+			       (asyw->image.pitch[0] >> 8) << 8 |
+			       asyw->image.blocks[0] << 8 |
+			       asyw->image.blockh);
+		evo_data(push, asyw->image.format << 8);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+base907c_xlut_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 6))) {
+		evo_mthd(push, 0x00e0, 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x00e8, 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x00fc, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+base907c_xlut_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 6))) {
+		evo_mthd(push, 0x00e0, 3);
+		evo_data(push, asyw->xlut.i.enable << 30 |
+			       asyw->xlut.i.mode << 24);
+		evo_data(push, asyw->xlut.i.offset >> 8);
+		evo_data(push, 0x40000000);
+		evo_mthd(push, 0x00fc, 1);
+		evo_data(push, asyw->xlut.handle);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+base907c_ilut(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	asyw->xlut.i.mode = 7;
+	asyw->xlut.i.enable = 2;
+}
+
+const struct nv50_wndw_func
+base907c = {
+	.acquire = base507c_acquire,
+	.release = base507c_release,
+	.sema_set = base507c_sema_set,
+	.sema_clr = base507c_sema_clr,
+	.ntfy_reset = base507c_ntfy_reset,
+	.ntfy_set = base507c_ntfy_set,
+	.ntfy_clr = base507c_ntfy_clr,
+	.ntfy_wait_begun = base507c_ntfy_wait_begun,
+	.ilut = base907c_ilut,
+	.olut_core = true,
+	.xlut_set = base907c_xlut_set,
+	.xlut_clr = base907c_xlut_clr,
+	.image_set = base907c_image_set,
+	.image_clr = base507c_image_clr,
+	.update = base507c_update,
+};
+
+int
+base907c_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return base507c_new_(&base907c, base507c_format, drm, head, oclass,
+			     0x00000002 << (head * 4), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/base917c.c b/drivers/gpu/drm/nouveau/dispnv50/base917c.c
new file mode 100644
index 000000000000..54d705bb81a5
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/base917c.c
@@ -0,0 +1,48 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "base.h"
+#include "atom.h"
+
+const u32
+base917c_format[] = {
+	DRM_FORMAT_C8,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_ARGB1555,
+	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ABGR2101010,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_ABGR8888,
+	DRM_FORMAT_XRGB2101010,
+	DRM_FORMAT_ARGB2101010,
+	0
+};
+
+int
+base917c_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return base507c_new_(&base907c, base917c_format, drm, head, oclass,
+			     0x00000002 << (head * 4), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core.c b/drivers/gpu/drm/nouveau/dispnv50/core.c
new file mode 100644
index 000000000000..f3c49adb1bdb
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/core.c
@@ -0,0 +1,70 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+
+#include <nvif/class.h>
+
+void
+nv50_core_del(struct nv50_core **pcore)
+{
+	struct nv50_core *core = *pcore;
+	if (core) {
+		nv50_dmac_destroy(&core->chan);
+		kfree(*pcore);
+		*pcore = NULL;
+	}
+}
+
+int
+nv50_core_new(struct nouveau_drm *drm, struct nv50_core **pcore)
+{
+	struct {
+		s32 oclass;
+		int version;
+		int (*new)(struct nouveau_drm *, s32, struct nv50_core **);
+	} cores[] = {
+		{ GV100_DISP_CORE_CHANNEL_DMA, 0, corec37d_new },
+		{ GP102_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
+		{ GP100_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
+		{ GM200_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
+		{ GM107_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
+		{ GK110_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
+		{ GK104_DISP_CORE_CHANNEL_DMA, 0, core917d_new },
+		{ GF110_DISP_CORE_CHANNEL_DMA, 0, core907d_new },
+		{ GT214_DISP_CORE_CHANNEL_DMA, 0, core827d_new },
+		{ GT206_DISP_CORE_CHANNEL_DMA, 0, core827d_new },
+		{ GT200_DISP_CORE_CHANNEL_DMA, 0, core827d_new },
+		{   G82_DISP_CORE_CHANNEL_DMA, 0, core827d_new },
+		{  NV50_DISP_CORE_CHANNEL_DMA, 0, core507d_new },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid;
+
+	cid = nvif_mclass(&disp->disp->object, cores);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported core channel class\n");
+		return cid;
+	}
+
+	return cores[cid].new(drm, cores[cid].oclass, pcore);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core.h b/drivers/gpu/drm/nouveau/dispnv50/core.h
new file mode 100644
index 000000000000..8470df9dd13d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/core.h
@@ -0,0 +1,50 @@
+#ifndef __NV50_KMS_CORE_H__
+#define __NV50_KMS_CORE_H__
+#include "disp.h"
+#include "atom.h"
+
+struct nv50_core {
+	const struct nv50_core_func *func;
+	struct nv50_dmac chan;
+};
+
+int nv50_core_new(struct nouveau_drm *, struct nv50_core **);
+void nv50_core_del(struct nv50_core **);
+
+struct nv50_core_func {
+	void (*init)(struct nv50_core *);
+	void (*ntfy_init)(struct nouveau_bo *, u32 offset);
+	int (*ntfy_wait_done)(struct nouveau_bo *, u32 offset,
+			      struct nvif_device *);
+	void (*update)(struct nv50_core *, u32 *interlock, bool ntfy);
+
+	const struct nv50_head_func *head;
+	const struct nv50_outp_func {
+		void (*ctrl)(struct nv50_core *, int or, u32 ctrl,
+			     struct nv50_head_atom *);
+	} *dac, *pior, *sor;
+};
+
+int core507d_new(struct nouveau_drm *, s32, struct nv50_core **);
+int core507d_new_(const struct nv50_core_func *, struct nouveau_drm *, s32,
+		  struct nv50_core **);
+void core507d_init(struct nv50_core *);
+void core507d_ntfy_init(struct nouveau_bo *, u32);
+int core507d_ntfy_wait_done(struct nouveau_bo *, u32, struct nvif_device *);
+void core507d_update(struct nv50_core *, u32 *, bool);
+
+extern const struct nv50_outp_func dac507d;
+extern const struct nv50_outp_func sor507d;
+extern const struct nv50_outp_func pior507d;
+
+int core827d_new(struct nouveau_drm *, s32, struct nv50_core **);
+
+int core907d_new(struct nouveau_drm *, s32, struct nv50_core **);
+extern const struct nv50_outp_func dac907d;
+extern const struct nv50_outp_func sor907d;
+
+int core917d_new(struct nouveau_drm *, s32, struct nv50_core **);
+
+int corec37d_new(struct nouveau_drm *, s32, struct nv50_core **);
+extern const struct nv50_outp_func sorc37d;
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core507d.c b/drivers/gpu/drm/nouveau/dispnv50/core507d.c
new file mode 100644
index 000000000000..e7fcfa6e6467
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/core507d.c
@@ -0,0 +1,115 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+#include "head.h"
+
+#include <nvif/cl507d.h>
+
+#include "nouveau_bo.h"
+
+void
+core507d_update(struct nv50_core *core, u32 *interlock, bool ntfy)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 5))) {
+		if (ntfy) {
+			evo_mthd(push, 0x0084, 1);
+			evo_data(push, 0x80000000 | NV50_DISP_CORE_NTFY);
+		}
+		evo_mthd(push, 0x0080, 2);
+		evo_data(push, interlock[NV50_DISP_INTERLOCK_BASE] |
+			       interlock[NV50_DISP_INTERLOCK_OVLY]);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &core->chan);
+	}
+}
+
+int
+core507d_ntfy_wait_done(struct nouveau_bo *bo, u32 offset,
+			struct nvif_device *device)
+{
+	s64 time = nvif_msec(device, 2000ULL,
+		if (nouveau_bo_rd32(bo, offset / 4))
+			break;
+		usleep_range(1, 2);
+	);
+	return time < 0 ? time : 0;
+}
+
+void
+core507d_ntfy_init(struct nouveau_bo *bo, u32 offset)
+{
+	nouveau_bo_wr32(bo, offset / 4, 0x00000000);
+}
+
+void
+core507d_init(struct nv50_core *core)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 2))) {
+		evo_mthd(push, 0x0088, 1);
+		evo_data(push, core->chan.sync.handle);
+		evo_kick(push, &core->chan);
+	}
+}
+
+static const struct nv50_core_func
+core507d = {
+	.init = core507d_init,
+	.ntfy_init = core507d_ntfy_init,
+	.ntfy_wait_done = core507d_ntfy_wait_done,
+	.update = core507d_update,
+	.head = &head507d,
+	.dac = &dac507d,
+	.sor = &sor507d,
+	.pior = &pior507d,
+};
+
+int
+core507d_new_(const struct nv50_core_func *func, struct nouveau_drm *drm,
+	      s32 oclass, struct nv50_core **pcore)
+{
+	struct nv50_disp_core_channel_dma_v0 args = {};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_core *core;
+	int ret;
+
+	if (!(core = *pcore = kzalloc(sizeof(*core), GFP_KERNEL)))
+		return -ENOMEM;
+	core->func = func;
+
+	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
+			       &oclass, 0, &args, sizeof(args),
+			       disp->sync->bo.offset, &core->chan);
+	if (ret) {
+		NV_ERROR(drm, "core%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+int
+core507d_new(struct nouveau_drm *drm, s32 oclass, struct nv50_core **pcore)
+{
+	return core507d_new_(&core507d, drm, oclass, pcore);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core827d.c b/drivers/gpu/drm/nouveau/dispnv50/core827d.c
new file mode 100644
index 000000000000..6123a068f836
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/core827d.c
@@ -0,0 +1,41 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+#include "head.h"
+
+static const struct nv50_core_func
+core827d = {
+	.init = core507d_init,
+	.ntfy_init = core507d_ntfy_init,
+	.ntfy_wait_done = core507d_ntfy_wait_done,
+	.update = core507d_update,
+	.head = &head827d,
+	.dac = &dac507d,
+	.sor = &sor507d,
+	.pior = &pior507d,
+};
+
+int
+core827d_new(struct nouveau_drm *drm, s32 oclass, struct nv50_core **pcore)
+{
+	return core507d_new_(&core827d, drm, oclass, pcore);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegt200.c b/drivers/gpu/drm/nouveau/dispnv50/core907d.c
index 93451e46570c..ef822f813435 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegt200.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/core907d.c
@@ -1,5 +1,5 @@
 /*
- * Copyright 2015 Red Hat Inc.
+ * Copyright 2018 Red Hat Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -18,21 +18,23 @@
  * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
+#include "core.h"
+#include "head.h"
 
-const struct nv50_disp_dmac_oclass
-gt200_disp_base_oclass = {
-	.base.oclass = GT200_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &g84_disp_base_chan_mthd,
-	.chid = 1,
+static const struct nv50_core_func
+core907d = {
+	.init = core507d_init,
+	.ntfy_init = core507d_ntfy_init,
+	.ntfy_wait_done = core507d_ntfy_wait_done,
+	.update = core507d_update,
+	.head = &head907d,
+	.dac = &dac907d,
+	.sor = &sor907d,
 };
+
+int
+core907d_new(struct nouveau_drm *drm, s32 oclass, struct nv50_core **pcore)
+{
+	return core507d_new_(&core907d, drm, oclass, pcore);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/core917d.c b/drivers/gpu/drm/nouveau/dispnv50/core917d.c
new file mode 100644
index 000000000000..392338df5bfd
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/core917d.c
@@ -0,0 +1,40 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+#include "head.h"
+
+static const struct nv50_core_func
+core917d = {
+	.init = core507d_init,
+	.ntfy_init = core507d_ntfy_init,
+	.ntfy_wait_done = core507d_ntfy_wait_done,
+	.update = core507d_update,
+	.head = &head917d,
+	.dac = &dac907d,
+	.sor = &sor907d,
+};
+
+int
+core917d_new(struct nouveau_drm *drm, s32 oclass, struct nv50_core **pcore)
+{
+	return core507d_new_(&core917d, drm, oclass, pcore);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/corec37d.c b/drivers/gpu/drm/nouveau/dispnv50/corec37d.c
new file mode 100644
index 000000000000..b5c17c948918
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/corec37d.c
@@ -0,0 +1,110 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+#include "head.h"
+
+#include <nouveau_bo.h>
+
+static void
+corec37d_update(struct nv50_core *core, u32 *interlock, bool ntfy)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 9))) {
+		if (ntfy) {
+			evo_mthd(push, 0x020c, 1);
+			evo_data(push, 0x00001000 | NV50_DISP_CORE_NTFY);
+		}
+
+		evo_mthd(push, 0x0218, 2);
+		evo_data(push, interlock[NV50_DISP_INTERLOCK_CURS]);
+		evo_data(push, interlock[NV50_DISP_INTERLOCK_WNDW]);
+		evo_mthd(push, 0x0200, 1);
+		evo_data(push, 0x00000001);
+
+		if (ntfy) {
+			evo_mthd(push, 0x020c, 1);
+			evo_data(push, 0x00000000);
+		}
+		evo_kick(push, &core->chan);
+	}
+}
+
+int
+corec37d_ntfy_wait_done(struct nouveau_bo *bo, u32 offset,
+			struct nvif_device *device)
+{
+	u32 data;
+	s64 time = nvif_msec(device, 2000ULL,
+		data = nouveau_bo_rd32(bo, offset / 4 + 0);
+		if ((data & 0xc0000000) == 0x80000000)
+			break;
+		usleep_range(1, 2);
+	);
+	return time < 0 ? time : 0;
+}
+
+void
+corec37d_ntfy_init(struct nouveau_bo *bo, u32 offset)
+{
+	nouveau_bo_wr32(bo, offset / 4 + 0, 0x00000000);
+	nouveau_bo_wr32(bo, offset / 4 + 1, 0x00000000);
+	nouveau_bo_wr32(bo, offset / 4 + 2, 0x00000000);
+	nouveau_bo_wr32(bo, offset / 4 + 3, 0x00000000);
+}
+
+void
+corec37d_init(struct nv50_core *core)
+{
+	const u32 windows = 8; /*XXX*/
+	u32 *push, i;
+	if ((push = evo_wait(&core->chan, 2 + 6 * windows + 2))) {
+		evo_mthd(push, 0x0208, 1);
+		evo_data(push, core->chan.sync.handle);
+		for (i = 0; i < windows; i++) {
+			evo_mthd(push, 0x1000 + (i * 0x080), 3);
+			evo_data(push, i >> 1);
+			evo_data(push, 0x00000017);
+			evo_data(push, 0x00000000);
+			evo_mthd(push, 0x1010 + (i * 0x080), 1);
+			evo_data(push, 0x00127fff);
+		}
+		evo_mthd(push, 0x0200, 1);
+		evo_data(push, 0x00000001);
+		evo_kick(push, &core->chan);
+	}
+}
+
+static const struct nv50_core_func
+corec37d = {
+	.init = corec37d_init,
+	.ntfy_init = corec37d_ntfy_init,
+	.ntfy_wait_done = corec37d_ntfy_wait_done,
+	.update = corec37d_update,
+	.head = &headc37d,
+	.sor = &sorc37d,
+};
+
+int
+corec37d_new(struct nouveau_drm *drm, s32 oclass, struct nv50_core **pcore)
+{
+	return core507d_new_(&corec37d, drm, oclass, pcore);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs.c b/drivers/gpu/drm/nouveau/dispnv50/curs.c
new file mode 100644
index 000000000000..f592087338c4
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs.c
@@ -0,0 +1,52 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "curs.h"
+
+#include <nvif/class.h>
+
+int
+nv50_curs_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
+{
+	struct {
+		s32 oclass;
+		int version;
+		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+	} curses[] = {
+		{ GV100_DISP_CURSOR, 0, cursc37a_new },
+		{ GK104_DISP_CURSOR, 0, curs907a_new },
+		{ GF110_DISP_CURSOR, 0, curs907a_new },
+		{ GT214_DISP_CURSOR, 0, curs507a_new },
+		{   G82_DISP_CURSOR, 0, curs507a_new },
+		{  NV50_DISP_CURSOR, 0, curs507a_new },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid;
+
+	cid = nvif_mclass(&disp->disp->object, curses);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported cursor immediate class\n");
+		return cid;
+	}
+
+	return curses[cid].new(drm, head, curses[cid].oclass, pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs.h b/drivers/gpu/drm/nouveau/dispnv50/curs.h
new file mode 100644
index 000000000000..23aff5fd6747
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs.h
@@ -0,0 +1,14 @@
+#ifndef __NV50_KMS_CURS_H__
+#define __NV50_KMS_CURS_H__
+#include "wndw.h"
+
+int curs507a_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+int curs507a_new_(const struct nv50_wimm_func *, struct nouveau_drm *,
+		  int head, s32 oclass, u32 interlock_data,
+		  struct nv50_wndw **);
+
+int curs907a_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+int cursc37a_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+
+int nv50_curs_new(struct nouveau_drm *, int head, struct nv50_wndw **);
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/curs507a.c b/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
new file mode 100644
index 000000000000..397143b639c6
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs507a.c
@@ -0,0 +1,145 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "curs.h"
+#include "core.h"
+#include "head.h"
+
+#include <nvif/cl507a.h>
+
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_plane_helper.h>
+
+static void
+curs507a_update(struct nv50_wndw *wndw, u32 *interlock)
+{
+	nvif_wr32(&wndw->wimm.base.user, 0x0080, 0x00000000);
+}
+
+static void
+curs507a_point(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	nvif_wr32(&wndw->wimm.base.user, 0x0084, asyw->point.y << 16 |
+						 asyw->point.x);
+}
+
+const struct nv50_wimm_func
+curs507a = {
+	.point = curs507a_point,
+	.update = curs507a_update,
+};
+
+static void
+curs507a_prepare(struct nv50_wndw *wndw, struct nv50_head_atom *asyh,
+		 struct nv50_wndw_atom *asyw)
+{
+	u32 handle = nv50_disp(wndw->plane.dev)->core->chan.vram.handle;
+	u32 offset = asyw->image.offset[0];
+	if (asyh->curs.handle != handle || asyh->curs.offset != offset) {
+		asyh->curs.handle = handle;
+		asyh->curs.offset = offset;
+		asyh->set.curs = asyh->curs.visible;
+	}
+}
+
+static void
+curs507a_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+	asyh->curs.visible = false;
+}
+
+static int
+curs507a_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+	struct nv50_head *head = nv50_head(asyw->state.crtc);
+	int ret;
+
+	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  true, true);
+	asyh->curs.visible = asyw->state.visible;
+	if (ret || !asyh->curs.visible)
+		return ret;
+
+	if (asyw->image.w != asyw->image.h)
+		return -EINVAL;
+
+	ret = head->func->curs_layout(head, asyw, asyh);
+	if (ret)
+		return ret;
+
+	return head->func->curs_format(head, asyw, asyh);
+}
+
+static const u32
+curs507a_format[] = {
+	DRM_FORMAT_ARGB8888,
+	0
+};
+
+static const struct nv50_wndw_func
+curs507a_wndw = {
+	.acquire = curs507a_acquire,
+	.release = curs507a_release,
+	.prepare = curs507a_prepare,
+};
+
+int
+curs507a_new_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
+	      int head, s32 oclass, u32 interlock_data,
+	      struct nv50_wndw **pwndw)
+{
+	struct nv50_disp_cursor_v0 args = {
+		.head = head,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_wndw *wndw;
+	int ret;
+
+	ret = nv50_wndw_new_(&curs507a_wndw, drm->dev, DRM_PLANE_TYPE_CURSOR,
+			     "curs", head, curs507a_format, BIT(head),
+			     NV50_DISP_INTERLOCK_CURS, interlock_data, &wndw);
+	if (*pwndw = wndw, ret)
+		return ret;
+
+	ret = nvif_object_init(&disp->disp->object, 0, oclass, &args,
+			       sizeof(args), &wndw->wimm.base.user);
+	if (ret) {
+		NV_ERROR(drm, "curs%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	nvif_object_map(&wndw->wimm.base.user, NULL, 0);
+	wndw->immd = func;
+	wndw->ctxdma.parent = NULL;
+	return 0;
+}
+
+int
+curs507a_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return curs507a_new_(&curs507a, drm, head, oclass,
+			     0x00000001 << (head * 8), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogm200.c b/drivers/gpu/drm/nouveau/dispnv50/curs907a.c
index a13315147391..d742362de03e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogm200.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/curs907a.c
@@ -1,5 +1,5 @@
 /*
- * Copyright 2015 Red Hat Inc.
+ * Copyright 2018 Red Hat Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -18,17 +18,13 @@
  * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs
  */
-#include "changk104.h"
-
-#include <nvif/class.h>
+#include "curs.h"
 
-const struct nvkm_fifo_chan_oclass
-gm200_fifo_gpfifo_oclass = {
-	.base.oclass = MAXWELL_CHANNEL_GPFIFO_A,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = gk104_fifo_gpfifo_new,
-};
+int
+curs907a_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return curs507a_new_(&curs507a, drm, head, oclass,
+			     0x00000001 << (head * 4), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/cursc37a.c b/drivers/gpu/drm/nouveau/dispnv50/cursc37a.c
new file mode 100644
index 000000000000..23fb29d41efe
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/cursc37a.c
@@ -0,0 +1,50 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "curs.h"
+#include "atom.h"
+
+static void
+cursc37a_update(struct nv50_wndw *wndw, u32 *interlock)
+{
+	nvif_wr32(&wndw->wimm.base.user, 0x0200, 0x00000001);
+}
+
+static void
+cursc37a_point(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	nvif_wr32(&wndw->wimm.base.user, 0x0208, asyw->point.y << 16 |
+						 asyw->point.x);
+}
+
+static const struct nv50_wimm_func
+cursc37a = {
+	.point = cursc37a_point,
+	.update = cursc37a_update,
+};
+
+int
+cursc37a_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return curs507a_new_(&cursc37a, drm, head, oclass,
+			     0x00000001 << head, pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/dac507d.c b/drivers/gpu/drm/nouveau/dispnv50/dac507d.c
new file mode 100644
index 000000000000..2a10ef7d30a8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/dac507d.c
@@ -0,0 +1,44 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+
+static void
+dac507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	     struct nv50_head_atom *asyh)
+{
+	u32 *push, sync = 0;
+	if ((push = evo_wait(&core->chan, 3))) {
+		if (asyh) {
+			sync |= asyh->or.nvsync << 1;
+			sync |= asyh->or.nhsync;
+		}
+		evo_mthd(push, 0x0400 + (or * 0x080), 2);
+		evo_data(push, ctrl);
+		evo_data(push, sync);
+		evo_kick(push, &core->chan);
+	}
+}
+
+const struct nv50_outp_func
+dac507d = {
+	.ctrl = dac507d_ctrl,
+};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgt215.c b/drivers/gpu/drm/nouveau/dispnv50/dac907d.c
index 00a7f3564450..11e87fa53fac 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgt215.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/dac907d.c
@@ -1,5 +1,5 @@
 /*
- * Copyright 2015 Red Hat Inc.
+ * Copyright 2018 Red Hat Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -18,20 +18,22 @@
  * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
-#include "channv50.h"
-#include "rootnv50.h"
+#include "core.h"
 
-#include <nvif/class.h>
+static void
+dac907d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	     struct nv50_head_atom *asyh)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 2))) {
+		evo_mthd(push, 0x0180 + (or * 0x020), 1);
+		evo_data(push, ctrl);
+		evo_kick(push, &core->chan);
+	}
+}
 
-const struct nv50_disp_pioc_oclass
-gt215_disp_curs_oclass = {
-	.base.oclass = GT214_DISP_CURSOR,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_curs_new,
-	.func = &nv50_disp_pioc_func,
-	.chid = { 7, 7 },
+const struct nv50_outp_func
+dac907d = {
+	.ctrl = dac907d_ctrl,
 };
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.c b/drivers/gpu/drm/nouveau/dispnv50/disp.c
new file mode 100644
index 000000000000..0190377b02a6
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.c
@@ -0,0 +1,2266 @@
+/*
+ * Copyright 2011 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: Ben Skeggs
+ */
+#include "disp.h"
+#include "atom.h"
+#include "core.h"
+#include "head.h"
+#include "wndw.h"
+
+#include <linux/dma-mapping.h>
+#include <linux/hdmi.h>
+
+#include <drm/drmP.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_dp_helper.h>
+#include <drm/drm_fb_helper.h>
+#include <drm/drm_plane_helper.h>
+#include <drm/drm_edid.h>
+
+#include <nvif/class.h>
+#include <nvif/cl0002.h>
+#include <nvif/cl5070.h>
+#include <nvif/cl507d.h>
+#include <nvif/event.h>
+
+#include "nouveau_drv.h"
+#include "nouveau_dma.h"
+#include "nouveau_gem.h"
+#include "nouveau_connector.h"
+#include "nouveau_encoder.h"
+#include "nouveau_fence.h"
+#include "nouveau_fbcon.h"
+
+#include <subdev/bios/dp.h>
+
+/******************************************************************************
+ * Atomic state
+ *****************************************************************************/
+
+struct nv50_outp_atom {
+	struct list_head head;
+
+	struct drm_encoder *encoder;
+	bool flush_disable;
+
+	union nv50_outp_atom_mask {
+		struct {
+			bool ctrl:1;
+		};
+		u8 mask;
+	} set, clr;
+};
+
+/******************************************************************************
+ * EVO channel
+ *****************************************************************************/
+
+static int
+nv50_chan_create(struct nvif_device *device, struct nvif_object *disp,
+		 const s32 *oclass, u8 head, void *data, u32 size,
+		 struct nv50_chan *chan)
+{
+	struct nvif_sclass *sclass;
+	int ret, i, n;
+
+	chan->device = device;
+
+	ret = n = nvif_object_sclass_get(disp, &sclass);
+	if (ret < 0)
+		return ret;
+
+	while (oclass[0]) {
+		for (i = 0; i < n; i++) {
+			if (sclass[i].oclass == oclass[0]) {
+				ret = nvif_object_init(disp, 0, oclass[0],
+						       data, size, &chan->user);
+				if (ret == 0)
+					nvif_object_map(&chan->user, NULL, 0);
+				nvif_object_sclass_put(&sclass);
+				return ret;
+			}
+		}
+		oclass++;
+	}
+
+	nvif_object_sclass_put(&sclass);
+	return -ENOSYS;
+}
+
+static void
+nv50_chan_destroy(struct nv50_chan *chan)
+{
+	nvif_object_fini(&chan->user);
+}
+
+/******************************************************************************
+ * DMA EVO channel
+ *****************************************************************************/
+
+void
+nv50_dmac_destroy(struct nv50_dmac *dmac)
+{
+	nvif_object_fini(&dmac->vram);
+	nvif_object_fini(&dmac->sync);
+
+	nv50_chan_destroy(&dmac->base);
+
+	nvif_mem_fini(&dmac->push);
+}
+
+int
+nv50_dmac_create(struct nvif_device *device, struct nvif_object *disp,
+		 const s32 *oclass, u8 head, void *data, u32 size, u64 syncbuf,
+		 struct nv50_dmac *dmac)
+{
+	struct nouveau_cli *cli = (void *)device->object.client;
+	struct nv50_disp_core_channel_dma_v0 *args = data;
+	u8 type = NVIF_MEM_COHERENT;
+	int ret;
+
+	mutex_init(&dmac->lock);
+
+	/* Pascal added support for 47-bit physical addresses, but some
+	 * parts of EVO still only accept 40-bit PAs.
+	 *
+	 * To avoid issues on systems with large amounts of RAM, and on
+	 * systems where an IOMMU maps pages at a high address, we need
+	 * to allocate push buffers in VRAM instead.
+	 *
+	 * This appears to match NVIDIA's behaviour on Pascal.
+	 */
+	if (device->info.family == NV_DEVICE_INFO_V0_PASCAL)
+		type |= NVIF_MEM_VRAM;
+
+	ret = nvif_mem_init_map(&cli->mmu, type, 0x1000, &dmac->push);
+	if (ret)
+		return ret;
+
+	dmac->ptr = dmac->push.object.map.ptr;
+
+	args->pushbuf = nvif_handle(&dmac->push.object);
+
+	ret = nv50_chan_create(device, disp, oclass, head, data, size,
+			       &dmac->base);
+	if (ret)
+		return ret;
+
+	if (!syncbuf)
+		return 0;
+
+	ret = nvif_object_init(&dmac->base.user, 0xf0000000, NV_DMA_IN_MEMORY,
+			       &(struct nv_dma_v0) {
+					.target = NV_DMA_V0_TARGET_VRAM,
+					.access = NV_DMA_V0_ACCESS_RDWR,
+					.start = syncbuf + 0x0000,
+					.limit = syncbuf + 0x0fff,
+			       }, sizeof(struct nv_dma_v0),
+			       &dmac->sync);
+	if (ret)
+		return ret;
+
+	ret = nvif_object_init(&dmac->base.user, 0xf0000001, NV_DMA_IN_MEMORY,
+			       &(struct nv_dma_v0) {
+					.target = NV_DMA_V0_TARGET_VRAM,
+					.access = NV_DMA_V0_ACCESS_RDWR,
+					.start = 0,
+					.limit = device->info.ram_user - 1,
+			       }, sizeof(struct nv_dma_v0),
+			       &dmac->vram);
+	if (ret)
+		return ret;
+
+	return ret;
+}
+
+/******************************************************************************
+ * EVO channel helpers
+ *****************************************************************************/
+u32 *
+evo_wait(struct nv50_dmac *evoc, int nr)
+{
+	struct nv50_dmac *dmac = evoc;
+	struct nvif_device *device = dmac->base.device;
+	u32 put = nvif_rd32(&dmac->base.user, 0x0000) / 4;
+
+	mutex_lock(&dmac->lock);
+	if (put + nr >= (PAGE_SIZE / 4) - 8) {
+		dmac->ptr[put] = 0x20000000;
+
+		nvif_wr32(&dmac->base.user, 0x0000, 0x00000000);
+		if (nvif_msec(device, 2000,
+			if (!nvif_rd32(&dmac->base.user, 0x0004))
+				break;
+		) < 0) {
+			mutex_unlock(&dmac->lock);
+			pr_err("nouveau: evo channel stalled\n");
+			return NULL;
+		}
+
+		put = 0;
+	}
+
+	return dmac->ptr + put;
+}
+
+void
+evo_kick(u32 *push, struct nv50_dmac *evoc)
+{
+	struct nv50_dmac *dmac = evoc;
+
+	/* Push buffer fetches are not coherent with BAR1, we need to ensure
+	 * writes have been flushed right through to VRAM before writing PUT.
+	 */
+	if (dmac->push.type & NVIF_MEM_VRAM) {
+		struct nvif_device *device = dmac->base.device;
+		nvif_wr32(&device->object, 0x070000, 0x00000001);
+		nvif_msec(device, 2000,
+			if (!(nvif_rd32(&device->object, 0x070000) & 0x00000002))
+				break;
+		);
+	}
+
+	nvif_wr32(&dmac->base.user, 0x0000, (push - dmac->ptr) << 2);
+	mutex_unlock(&dmac->lock);
+}
+
+/******************************************************************************
+ * Output path helpers
+ *****************************************************************************/
+static void
+nv50_outp_release(struct nouveau_encoder *nv_encoder)
+{
+	struct nv50_disp *disp = nv50_disp(nv_encoder->base.base.dev);
+	struct {
+		struct nv50_disp_mthd_v1 base;
+	} args = {
+		.base.version = 1,
+		.base.method = NV50_DISP_MTHD_V1_RELEASE,
+		.base.hasht  = nv_encoder->dcb->hasht,
+		.base.hashm  = nv_encoder->dcb->hashm,
+	};
+
+	nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
+	nv_encoder->or = -1;
+	nv_encoder->link = 0;
+}
+
+static int
+nv50_outp_acquire(struct nouveau_encoder *nv_encoder)
+{
+	struct nouveau_drm *drm = nouveau_drm(nv_encoder->base.base.dev);
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_acquire_v0 info;
+	} args = {
+		.base.version = 1,
+		.base.method = NV50_DISP_MTHD_V1_ACQUIRE,
+		.base.hasht  = nv_encoder->dcb->hasht,
+		.base.hashm  = nv_encoder->dcb->hashm,
+	};
+	int ret;
+
+	ret = nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
+	if (ret) {
+		NV_ERROR(drm, "error acquiring output path: %d\n", ret);
+		return ret;
+	}
+
+	nv_encoder->or = args.info.or;
+	nv_encoder->link = args.info.link;
+	return 0;
+}
+
+static int
+nv50_outp_atomic_check_view(struct drm_encoder *encoder,
+			    struct drm_crtc_state *crtc_state,
+			    struct drm_connector_state *conn_state,
+			    struct drm_display_mode *native_mode)
+{
+	struct drm_display_mode *adjusted_mode = &crtc_state->adjusted_mode;
+	struct drm_display_mode *mode = &crtc_state->mode;
+	struct drm_connector *connector = conn_state->connector;
+	struct nouveau_conn_atom *asyc = nouveau_conn_atom(conn_state);
+	struct nouveau_drm *drm = nouveau_drm(encoder->dev);
+
+	NV_ATOMIC(drm, "%s atomic_check\n", encoder->name);
+	asyc->scaler.full = false;
+	if (!native_mode)
+		return 0;
+
+	if (asyc->scaler.mode == DRM_MODE_SCALE_NONE) {
+		switch (connector->connector_type) {
+		case DRM_MODE_CONNECTOR_LVDS:
+		case DRM_MODE_CONNECTOR_eDP:
+			/* Force use of scaler for non-EDID modes. */
+			if (adjusted_mode->type & DRM_MODE_TYPE_DRIVER)
+				break;
+			mode = native_mode;
+			asyc->scaler.full = true;
+			break;
+		default:
+			break;
+		}
+	} else {
+		mode = native_mode;
+	}
+
+	if (!drm_mode_equal(adjusted_mode, mode)) {
+		drm_mode_copy(adjusted_mode, mode);
+		crtc_state->mode_changed = true;
+	}
+
+	return 0;
+}
+
+static int
+nv50_outp_atomic_check(struct drm_encoder *encoder,
+		       struct drm_crtc_state *crtc_state,
+		       struct drm_connector_state *conn_state)
+{
+	struct nouveau_connector *nv_connector =
+		nouveau_connector(conn_state->connector);
+	return nv50_outp_atomic_check_view(encoder, crtc_state, conn_state,
+					   nv_connector->native_mode);
+}
+
+/******************************************************************************
+ * DAC
+ *****************************************************************************/
+static void
+nv50_dac_disable(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nv50_core *core = nv50_disp(encoder->dev)->core;
+	if (nv_encoder->crtc)
+		core->func->dac->ctrl(core, nv_encoder->or, 0x00000000, NULL);
+	nv_encoder->crtc = NULL;
+	nv50_outp_release(nv_encoder);
+}
+
+static void
+nv50_dac_enable(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
+	struct nv50_head_atom *asyh = nv50_head_atom(nv_crtc->base.state);
+	struct nv50_core *core = nv50_disp(encoder->dev)->core;
+
+	nv50_outp_acquire(nv_encoder);
+
+	core->func->dac->ctrl(core, nv_encoder->or, 1 << nv_crtc->index, asyh);
+	asyh->or.depth = 0;
+
+	nv_encoder->crtc = encoder->crtc;
+}
+
+static enum drm_connector_status
+nv50_dac_detect(struct drm_encoder *encoder, struct drm_connector *connector)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nv50_disp *disp = nv50_disp(encoder->dev);
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_dac_load_v0 load;
+	} args = {
+		.base.version = 1,
+		.base.method = NV50_DISP_MTHD_V1_DAC_LOAD,
+		.base.hasht  = nv_encoder->dcb->hasht,
+		.base.hashm  = nv_encoder->dcb->hashm,
+	};
+	int ret;
+
+	args.load.data = nouveau_drm(encoder->dev)->vbios.dactestval;
+	if (args.load.data == 0)
+		args.load.data = 340;
+
+	ret = nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
+	if (ret || !args.load.load)
+		return connector_status_disconnected;
+
+	return connector_status_connected;
+}
+
+static const struct drm_encoder_helper_funcs
+nv50_dac_help = {
+	.atomic_check = nv50_outp_atomic_check,
+	.enable = nv50_dac_enable,
+	.disable = nv50_dac_disable,
+	.detect = nv50_dac_detect
+};
+
+static void
+nv50_dac_destroy(struct drm_encoder *encoder)
+{
+	drm_encoder_cleanup(encoder);
+	kfree(encoder);
+}
+
+static const struct drm_encoder_funcs
+nv50_dac_func = {
+	.destroy = nv50_dac_destroy,
+};
+
+static int
+nv50_dac_create(struct drm_connector *connector, struct dcb_output *dcbe)
+{
+	struct nouveau_drm *drm = nouveau_drm(connector->dev);
+	struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device);
+	struct nvkm_i2c_bus *bus;
+	struct nouveau_encoder *nv_encoder;
+	struct drm_encoder *encoder;
+	int type = DRM_MODE_ENCODER_DAC;
+
+	nv_encoder = kzalloc(sizeof(*nv_encoder), GFP_KERNEL);
+	if (!nv_encoder)
+		return -ENOMEM;
+	nv_encoder->dcb = dcbe;
+
+	bus = nvkm_i2c_bus_find(i2c, dcbe->i2c_index);
+	if (bus)
+		nv_encoder->i2c = &bus->i2c;
+
+	encoder = to_drm_encoder(nv_encoder);
+	encoder->possible_crtcs = dcbe->heads;
+	encoder->possible_clones = 0;
+	drm_encoder_init(connector->dev, encoder, &nv50_dac_func, type,
+			 "dac-%04x-%04x", dcbe->hasht, dcbe->hashm);
+	drm_encoder_helper_add(encoder, &nv50_dac_help);
+
+	drm_connector_attach_encoder(connector, encoder);
+	return 0;
+}
+
+/******************************************************************************
+ * Audio
+ *****************************************************************************/
+static void
+nv50_audio_disable(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nv50_disp *disp = nv50_disp(encoder->dev);
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_sor_hda_eld_v0 eld;
+	} args = {
+		.base.version = 1,
+		.base.method  = NV50_DISP_MTHD_V1_SOR_HDA_ELD,
+		.base.hasht   = nv_encoder->dcb->hasht,
+		.base.hashm   = (0xf0ff & nv_encoder->dcb->hashm) |
+				(0x0100 << nv_crtc->index),
+	};
+
+	nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
+}
+
+static void
+nv50_audio_enable(struct drm_encoder *encoder, struct drm_display_mode *mode)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
+	struct nouveau_connector *nv_connector;
+	struct nv50_disp *disp = nv50_disp(encoder->dev);
+	struct __packed {
+		struct {
+			struct nv50_disp_mthd_v1 mthd;
+			struct nv50_disp_sor_hda_eld_v0 eld;
+		} base;
+		u8 data[sizeof(nv_connector->base.eld)];
+	} args = {
+		.base.mthd.version = 1,
+		.base.mthd.method  = NV50_DISP_MTHD_V1_SOR_HDA_ELD,
+		.base.mthd.hasht   = nv_encoder->dcb->hasht,
+		.base.mthd.hashm   = (0xf0ff & nv_encoder->dcb->hashm) |
+				     (0x0100 << nv_crtc->index),
+	};
+
+	nv_connector = nouveau_encoder_connector_get(nv_encoder);
+	if (!drm_detect_monitor_audio(nv_connector->edid))
+		return;
+
+	memcpy(args.data, nv_connector->base.eld, sizeof(args.data));
+
+	nvif_mthd(&disp->disp->object, 0, &args,
+		  sizeof(args.base) + drm_eld_size(args.data));
+}
+
+/******************************************************************************
+ * HDMI
+ *****************************************************************************/
+static void
+nv50_hdmi_disable(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nv50_disp *disp = nv50_disp(encoder->dev);
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_sor_hdmi_pwr_v0 pwr;
+	} args = {
+		.base.version = 1,
+		.base.method = NV50_DISP_MTHD_V1_SOR_HDMI_PWR,
+		.base.hasht  = nv_encoder->dcb->hasht,
+		.base.hashm  = (0xf0ff & nv_encoder->dcb->hashm) |
+			       (0x0100 << nv_crtc->index),
+	};
+
+	nvif_mthd(&disp->disp->object, 0, &args, sizeof(args));
+}
+
+static void
+nv50_hdmi_enable(struct drm_encoder *encoder, struct drm_display_mode *mode)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
+	struct nv50_disp *disp = nv50_disp(encoder->dev);
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_sor_hdmi_pwr_v0 pwr;
+		u8 infoframes[2 * 17]; /* two frames, up to 17 bytes each */
+	} args = {
+		.base.version = 1,
+		.base.method = NV50_DISP_MTHD_V1_SOR_HDMI_PWR,
+		.base.hasht  = nv_encoder->dcb->hasht,
+		.base.hashm  = (0xf0ff & nv_encoder->dcb->hashm) |
+			       (0x0100 << nv_crtc->index),
+		.pwr.state = 1,
+		.pwr.rekey = 56, /* binary driver, and tegra, constant */
+	};
+	struct nouveau_connector *nv_connector;
+	u32 max_ac_packet;
+	union hdmi_infoframe avi_frame;
+	union hdmi_infoframe vendor_frame;
+	int ret;
+	int size;
+
+	nv_connector = nouveau_encoder_connector_get(nv_encoder);
+	if (!drm_detect_hdmi_monitor(nv_connector->edid))
+		return;
+
+	ret = drm_hdmi_avi_infoframe_from_display_mode(&avi_frame.avi, mode,
+						       false);
+	if (!ret) {
+		/* We have an AVI InfoFrame, populate it to the display */
+		args.pwr.avi_infoframe_length
+			= hdmi_infoframe_pack(&avi_frame, args.infoframes, 17);
+	}
+
+	ret = drm_hdmi_vendor_infoframe_from_display_mode(&vendor_frame.vendor.hdmi,
+							  &nv_connector->base, mode);
+	if (!ret) {
+		/* We have a Vendor InfoFrame, populate it to the display */
+		args.pwr.vendor_infoframe_length
+			= hdmi_infoframe_pack(&vendor_frame,
+					      args.infoframes
+					      + args.pwr.avi_infoframe_length,
+					      17);
+	}
+
+	max_ac_packet  = mode->htotal - mode->hdisplay;
+	max_ac_packet -= args.pwr.rekey;
+	max_ac_packet -= 18; /* constant from tegra */
+	args.pwr.max_ac_packet = max_ac_packet / 32;
+
+	size = sizeof(args.base)
+		+ sizeof(args.pwr)
+		+ args.pwr.avi_infoframe_length
+		+ args.pwr.vendor_infoframe_length;
+	nvif_mthd(&disp->disp->object, 0, &args, size);
+	nv50_audio_enable(encoder, mode);
+}
+
+/******************************************************************************
+ * MST
+ *****************************************************************************/
+#define nv50_mstm(p) container_of((p), struct nv50_mstm, mgr)
+#define nv50_mstc(p) container_of((p), struct nv50_mstc, connector)
+#define nv50_msto(p) container_of((p), struct nv50_msto, encoder)
+
+struct nv50_mstm {
+	struct nouveau_encoder *outp;
+
+	struct drm_dp_mst_topology_mgr mgr;
+	struct nv50_msto *msto[4];
+
+	bool modified;
+	bool disabled;
+	int links;
+};
+
+struct nv50_mstc {
+	struct nv50_mstm *mstm;
+	struct drm_dp_mst_port *port;
+	struct drm_connector connector;
+
+	struct drm_display_mode *native;
+	struct edid *edid;
+
+	int pbn;
+};
+
+struct nv50_msto {
+	struct drm_encoder encoder;
+
+	struct nv50_head *head;
+	struct nv50_mstc *mstc;
+	bool disabled;
+};
+
+static struct drm_dp_payload *
+nv50_msto_payload(struct nv50_msto *msto)
+{
+	struct nouveau_drm *drm = nouveau_drm(msto->encoder.dev);
+	struct nv50_mstc *mstc = msto->mstc;
+	struct nv50_mstm *mstm = mstc->mstm;
+	int vcpi = mstc->port->vcpi.vcpi, i;
+
+	NV_ATOMIC(drm, "%s: vcpi %d\n", msto->encoder.name, vcpi);
+	for (i = 0; i < mstm->mgr.max_payloads; i++) {
+		struct drm_dp_payload *payload = &mstm->mgr.payloads[i];
+		NV_ATOMIC(drm, "%s: %d: vcpi %d start 0x%02x slots 0x%02x\n",
+			  mstm->outp->base.base.name, i, payload->vcpi,
+			  payload->start_slot, payload->num_slots);
+	}
+
+	for (i = 0; i < mstm->mgr.max_payloads; i++) {
+		struct drm_dp_payload *payload = &mstm->mgr.payloads[i];
+		if (payload->vcpi == vcpi)
+			return payload;
+	}
+
+	return NULL;
+}
+
+static void
+nv50_msto_cleanup(struct nv50_msto *msto)
+{
+	struct nouveau_drm *drm = nouveau_drm(msto->encoder.dev);
+	struct nv50_mstc *mstc = msto->mstc;
+	struct nv50_mstm *mstm = mstc->mstm;
+
+	NV_ATOMIC(drm, "%s: msto cleanup\n", msto->encoder.name);
+	if (mstc->port && mstc->port->vcpi.vcpi > 0 && !nv50_msto_payload(msto))
+		drm_dp_mst_deallocate_vcpi(&mstm->mgr, mstc->port);
+	if (msto->disabled) {
+		msto->mstc = NULL;
+		msto->head = NULL;
+		msto->disabled = false;
+	}
+}
+
+static void
+nv50_msto_prepare(struct nv50_msto *msto)
+{
+	struct nouveau_drm *drm = nouveau_drm(msto->encoder.dev);
+	struct nv50_mstc *mstc = msto->mstc;
+	struct nv50_mstm *mstm = mstc->mstm;
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_sor_dp_mst_vcpi_v0 vcpi;
+	} args = {
+		.base.version = 1,
+		.base.method = NV50_DISP_MTHD_V1_SOR_DP_MST_VCPI,
+		.base.hasht  = mstm->outp->dcb->hasht,
+		.base.hashm  = (0xf0ff & mstm->outp->dcb->hashm) |
+			       (0x0100 << msto->head->base.index),
+	};
+
+	NV_ATOMIC(drm, "%s: msto prepare\n", msto->encoder.name);
+	if (mstc->port && mstc->port->vcpi.vcpi > 0) {
+		struct drm_dp_payload *payload = nv50_msto_payload(msto);
+		if (payload) {
+			args.vcpi.start_slot = payload->start_slot;
+			args.vcpi.num_slots = payload->num_slots;
+			args.vcpi.pbn = mstc->port->vcpi.pbn;
+			args.vcpi.aligned_pbn = mstc->port->vcpi.aligned_pbn;
+		}
+	}
+
+	NV_ATOMIC(drm, "%s: %s: %02x %02x %04x %04x\n",
+		  msto->encoder.name, msto->head->base.base.name,
+		  args.vcpi.start_slot, args.vcpi.num_slots,
+		  args.vcpi.pbn, args.vcpi.aligned_pbn);
+	nvif_mthd(&drm->display->disp.object, 0, &args, sizeof(args));
+}
+
+static int
+nv50_msto_atomic_check(struct drm_encoder *encoder,
+		       struct drm_crtc_state *crtc_state,
+		       struct drm_connector_state *conn_state)
+{
+	struct nv50_mstc *mstc = nv50_mstc(conn_state->connector);
+	struct nv50_mstm *mstm = mstc->mstm;
+	int bpp = conn_state->connector->display_info.bpc * 3;
+	int slots;
+
+	mstc->pbn = drm_dp_calc_pbn_mode(crtc_state->adjusted_mode.clock, bpp);
+
+	slots = drm_dp_find_vcpi_slots(&mstm->mgr, mstc->pbn);
+	if (slots < 0)
+		return slots;
+
+	return nv50_outp_atomic_check_view(encoder, crtc_state, conn_state,
+					   mstc->native);
+}
+
+static void
+nv50_msto_enable(struct drm_encoder *encoder)
+{
+	struct nv50_head *head = nv50_head(encoder->crtc);
+	struct nv50_msto *msto = nv50_msto(encoder);
+	struct nv50_mstc *mstc = NULL;
+	struct nv50_mstm *mstm = NULL;
+	struct drm_connector *connector;
+	struct drm_connector_list_iter conn_iter;
+	u8 proto, depth;
+	int slots;
+	bool r;
+
+	drm_connector_list_iter_begin(encoder->dev, &conn_iter);
+	drm_for_each_connector_iter(connector, &conn_iter) {
+		if (connector->state->best_encoder == &msto->encoder) {
+			mstc = nv50_mstc(connector);
+			mstm = mstc->mstm;
+			break;
+		}
+	}
+	drm_connector_list_iter_end(&conn_iter);
+
+	if (WARN_ON(!mstc))
+		return;
+
+	slots = drm_dp_find_vcpi_slots(&mstm->mgr, mstc->pbn);
+	r = drm_dp_mst_allocate_vcpi(&mstm->mgr, mstc->port, mstc->pbn, slots);
+	WARN_ON(!r);
+
+	if (!mstm->links++)
+		nv50_outp_acquire(mstm->outp);
+
+	if (mstm->outp->link & 1)
+		proto = 0x8;
+	else
+		proto = 0x9;
+
+	switch (mstc->connector.display_info.bpc) {
+	case  6: depth = 0x2; break;
+	case  8: depth = 0x5; break;
+	case 10:
+	default: depth = 0x6; break;
+	}
+
+	mstm->outp->update(mstm->outp, head->base.index,
+			   nv50_head_atom(head->base.base.state), proto, depth);
+
+	msto->head = head;
+	msto->mstc = mstc;
+	mstm->modified = true;
+}
+
+static void
+nv50_msto_disable(struct drm_encoder *encoder)
+{
+	struct nv50_msto *msto = nv50_msto(encoder);
+	struct nv50_mstc *mstc = msto->mstc;
+	struct nv50_mstm *mstm = mstc->mstm;
+
+	if (mstc->port)
+		drm_dp_mst_reset_vcpi_slots(&mstm->mgr, mstc->port);
+
+	mstm->outp->update(mstm->outp, msto->head->base.index, NULL, 0, 0);
+	mstm->modified = true;
+	if (!--mstm->links)
+		mstm->disabled = true;
+	msto->disabled = true;
+}
+
+static const struct drm_encoder_helper_funcs
+nv50_msto_help = {
+	.disable = nv50_msto_disable,
+	.enable = nv50_msto_enable,
+	.atomic_check = nv50_msto_atomic_check,
+};
+
+static void
+nv50_msto_destroy(struct drm_encoder *encoder)
+{
+	struct nv50_msto *msto = nv50_msto(encoder);
+	drm_encoder_cleanup(&msto->encoder);
+	kfree(msto);
+}
+
+static const struct drm_encoder_funcs
+nv50_msto = {
+	.destroy = nv50_msto_destroy,
+};
+
+static int
+nv50_msto_new(struct drm_device *dev, u32 heads, const char *name, int id,
+	      struct nv50_msto **pmsto)
+{
+	struct nv50_msto *msto;
+	int ret;
+
+	if (!(msto = *pmsto = kzalloc(sizeof(*msto), GFP_KERNEL)))
+		return -ENOMEM;
+
+	ret = drm_encoder_init(dev, &msto->encoder, &nv50_msto,
+			       DRM_MODE_ENCODER_DPMST, "%s-mst-%d", name, id);
+	if (ret) {
+		kfree(*pmsto);
+		*pmsto = NULL;
+		return ret;
+	}
+
+	drm_encoder_helper_add(&msto->encoder, &nv50_msto_help);
+	msto->encoder.possible_crtcs = heads;
+	return 0;
+}
+
+static struct drm_encoder *
+nv50_mstc_atomic_best_encoder(struct drm_connector *connector,
+			      struct drm_connector_state *connector_state)
+{
+	struct nv50_head *head = nv50_head(connector_state->crtc);
+	struct nv50_mstc *mstc = nv50_mstc(connector);
+	if (mstc->port) {
+		struct nv50_mstm *mstm = mstc->mstm;
+		return &mstm->msto[head->base.index]->encoder;
+	}
+	return NULL;
+}
+
+static struct drm_encoder *
+nv50_mstc_best_encoder(struct drm_connector *connector)
+{
+	struct nv50_mstc *mstc = nv50_mstc(connector);
+	if (mstc->port) {
+		struct nv50_mstm *mstm = mstc->mstm;
+		return &mstm->msto[0]->encoder;
+	}
+	return NULL;
+}
+
+static enum drm_mode_status
+nv50_mstc_mode_valid(struct drm_connector *connector,
+		     struct drm_display_mode *mode)
+{
+	return MODE_OK;
+}
+
+static int
+nv50_mstc_get_modes(struct drm_connector *connector)
+{
+	struct nv50_mstc *mstc = nv50_mstc(connector);
+	int ret = 0;
+
+	mstc->edid = drm_dp_mst_get_edid(&mstc->connector, mstc->port->mgr, mstc->port);
+	drm_connector_update_edid_property(&mstc->connector, mstc->edid);
+	if (mstc->edid)
+		ret = drm_add_edid_modes(&mstc->connector, mstc->edid);
+
+	if (!mstc->connector.display_info.bpc)
+		mstc->connector.display_info.bpc = 8;
+
+	if (mstc->native)
+		drm_mode_destroy(mstc->connector.dev, mstc->native);
+	mstc->native = nouveau_conn_native_mode(&mstc->connector);
+	return ret;
+}
+
+static const struct drm_connector_helper_funcs
+nv50_mstc_help = {
+	.get_modes = nv50_mstc_get_modes,
+	.mode_valid = nv50_mstc_mode_valid,
+	.best_encoder = nv50_mstc_best_encoder,
+	.atomic_best_encoder = nv50_mstc_atomic_best_encoder,
+};
+
+static enum drm_connector_status
+nv50_mstc_detect(struct drm_connector *connector, bool force)
+{
+	struct nv50_mstc *mstc = nv50_mstc(connector);
+	if (!mstc->port)
+		return connector_status_disconnected;
+	return drm_dp_mst_detect_port(connector, mstc->port->mgr, mstc->port);
+}
+
+static void
+nv50_mstc_destroy(struct drm_connector *connector)
+{
+	struct nv50_mstc *mstc = nv50_mstc(connector);
+	drm_connector_cleanup(&mstc->connector);
+	kfree(mstc);
+}
+
+static const struct drm_connector_funcs
+nv50_mstc = {
+	.reset = nouveau_conn_reset,
+	.detect = nv50_mstc_detect,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = nv50_mstc_destroy,
+	.atomic_duplicate_state = nouveau_conn_atomic_duplicate_state,
+	.atomic_destroy_state = nouveau_conn_atomic_destroy_state,
+	.atomic_set_property = nouveau_conn_atomic_set_property,
+	.atomic_get_property = nouveau_conn_atomic_get_property,
+};
+
+static int
+nv50_mstc_new(struct nv50_mstm *mstm, struct drm_dp_mst_port *port,
+	      const char *path, struct nv50_mstc **pmstc)
+{
+	struct drm_device *dev = mstm->outp->base.base.dev;
+	struct nv50_mstc *mstc;
+	int ret, i;
+
+	if (!(mstc = *pmstc = kzalloc(sizeof(*mstc), GFP_KERNEL)))
+		return -ENOMEM;
+	mstc->mstm = mstm;
+	mstc->port = port;
+
+	ret = drm_connector_init(dev, &mstc->connector, &nv50_mstc,
+				 DRM_MODE_CONNECTOR_DisplayPort);
+	if (ret) {
+		kfree(*pmstc);
+		*pmstc = NULL;
+		return ret;
+	}
+
+	drm_connector_helper_add(&mstc->connector, &nv50_mstc_help);
+
+	mstc->connector.funcs->reset(&mstc->connector);
+	nouveau_conn_attach_properties(&mstc->connector);
+
+	for (i = 0; i < ARRAY_SIZE(mstm->msto) && mstm->msto[i]; i++)
+		drm_connector_attach_encoder(&mstc->connector, &mstm->msto[i]->encoder);
+
+	drm_object_attach_property(&mstc->connector.base, dev->mode_config.path_property, 0);
+	drm_object_attach_property(&mstc->connector.base, dev->mode_config.tile_property, 0);
+	drm_connector_set_path_property(&mstc->connector, path);
+	return 0;
+}
+
+static void
+nv50_mstm_cleanup(struct nv50_mstm *mstm)
+{
+	struct nouveau_drm *drm = nouveau_drm(mstm->outp->base.base.dev);
+	struct drm_encoder *encoder;
+	int ret;
+
+	NV_ATOMIC(drm, "%s: mstm cleanup\n", mstm->outp->base.base.name);
+	ret = drm_dp_check_act_status(&mstm->mgr);
+
+	ret = drm_dp_update_payload_part2(&mstm->mgr);
+
+	drm_for_each_encoder(encoder, mstm->outp->base.base.dev) {
+		if (encoder->encoder_type == DRM_MODE_ENCODER_DPMST) {
+			struct nv50_msto *msto = nv50_msto(encoder);
+			struct nv50_mstc *mstc = msto->mstc;
+			if (mstc && mstc->mstm == mstm)
+				nv50_msto_cleanup(msto);
+		}
+	}
+
+	mstm->modified = false;
+}
+
+static void
+nv50_mstm_prepare(struct nv50_mstm *mstm)
+{
+	struct nouveau_drm *drm = nouveau_drm(mstm->outp->base.base.dev);
+	struct drm_encoder *encoder;
+	int ret;
+
+	NV_ATOMIC(drm, "%s: mstm prepare\n", mstm->outp->base.base.name);
+	ret = drm_dp_update_payload_part1(&mstm->mgr);
+
+	drm_for_each_encoder(encoder, mstm->outp->base.base.dev) {
+		if (encoder->encoder_type == DRM_MODE_ENCODER_DPMST) {
+			struct nv50_msto *msto = nv50_msto(encoder);
+			struct nv50_mstc *mstc = msto->mstc;
+			if (mstc && mstc->mstm == mstm)
+				nv50_msto_prepare(msto);
+		}
+	}
+
+	if (mstm->disabled) {
+		if (!mstm->links)
+			nv50_outp_release(mstm->outp);
+		mstm->disabled = false;
+	}
+}
+
+static void
+nv50_mstm_hotplug(struct drm_dp_mst_topology_mgr *mgr)
+{
+	struct nv50_mstm *mstm = nv50_mstm(mgr);
+	drm_kms_helper_hotplug_event(mstm->outp->base.base.dev);
+}
+
+static void
+nv50_mstm_destroy_connector(struct drm_dp_mst_topology_mgr *mgr,
+			    struct drm_connector *connector)
+{
+	struct nouveau_drm *drm = nouveau_drm(connector->dev);
+	struct nv50_mstc *mstc = nv50_mstc(connector);
+
+	drm_connector_unregister(&mstc->connector);
+
+	drm_fb_helper_remove_one_connector(&drm->fbcon->helper, &mstc->connector);
+
+	drm_modeset_lock(&drm->dev->mode_config.connection_mutex, NULL);
+	mstc->port = NULL;
+	drm_modeset_unlock(&drm->dev->mode_config.connection_mutex);
+
+	drm_connector_put(&mstc->connector);
+}
+
+static void
+nv50_mstm_register_connector(struct drm_connector *connector)
+{
+	struct nouveau_drm *drm = nouveau_drm(connector->dev);
+
+	drm_fb_helper_add_one_connector(&drm->fbcon->helper, connector);
+
+	drm_connector_register(connector);
+}
+
+static struct drm_connector *
+nv50_mstm_add_connector(struct drm_dp_mst_topology_mgr *mgr,
+			struct drm_dp_mst_port *port, const char *path)
+{
+	struct nv50_mstm *mstm = nv50_mstm(mgr);
+	struct nv50_mstc *mstc;
+	int ret;
+
+	ret = nv50_mstc_new(mstm, port, path, &mstc);
+	if (ret) {
+		if (mstc)
+			mstc->connector.funcs->destroy(&mstc->connector);
+		return NULL;
+	}
+
+	return &mstc->connector;
+}
+
+static const struct drm_dp_mst_topology_cbs
+nv50_mstm = {
+	.add_connector = nv50_mstm_add_connector,
+	.register_connector = nv50_mstm_register_connector,
+	.destroy_connector = nv50_mstm_destroy_connector,
+	.hotplug = nv50_mstm_hotplug,
+};
+
+void
+nv50_mstm_service(struct nv50_mstm *mstm)
+{
+	struct drm_dp_aux *aux = mstm ? mstm->mgr.aux : NULL;
+	bool handled = true;
+	int ret;
+	u8 esi[8] = {};
+
+	if (!aux)
+		return;
+
+	while (handled) {
+		ret = drm_dp_dpcd_read(aux, DP_SINK_COUNT_ESI, esi, 8);
+		if (ret != 8) {
+			drm_dp_mst_topology_mgr_set_mst(&mstm->mgr, false);
+			return;
+		}
+
+		drm_dp_mst_hpd_irq(&mstm->mgr, esi, &handled);
+		if (!handled)
+			break;
+
+		drm_dp_dpcd_write(aux, DP_SINK_COUNT_ESI + 1, &esi[1], 3);
+	}
+}
+
+void
+nv50_mstm_remove(struct nv50_mstm *mstm)
+{
+	if (mstm)
+		drm_dp_mst_topology_mgr_set_mst(&mstm->mgr, false);
+}
+
+static int
+nv50_mstm_enable(struct nv50_mstm *mstm, u8 dpcd, int state)
+{
+	struct nouveau_encoder *outp = mstm->outp;
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_sor_dp_mst_link_v0 mst;
+	} args = {
+		.base.version = 1,
+		.base.method = NV50_DISP_MTHD_V1_SOR_DP_MST_LINK,
+		.base.hasht = outp->dcb->hasht,
+		.base.hashm = outp->dcb->hashm,
+		.mst.state = state,
+	};
+	struct nouveau_drm *drm = nouveau_drm(outp->base.base.dev);
+	struct nvif_object *disp = &drm->display->disp.object;
+	int ret;
+
+	if (dpcd >= 0x12) {
+		ret = drm_dp_dpcd_readb(mstm->mgr.aux, DP_MSTM_CTRL, &dpcd);
+		if (ret < 0)
+			return ret;
+
+		dpcd &= ~DP_MST_EN;
+		if (state)
+			dpcd |= DP_MST_EN;
+
+		ret = drm_dp_dpcd_writeb(mstm->mgr.aux, DP_MSTM_CTRL, dpcd);
+		if (ret < 0)
+			return ret;
+	}
+
+	return nvif_mthd(disp, 0, &args, sizeof(args));
+}
+
+int
+nv50_mstm_detect(struct nv50_mstm *mstm, u8 dpcd[8], int allow)
+{
+	int ret, state = 0;
+
+	if (!mstm)
+		return 0;
+
+	if (dpcd[0] >= 0x12) {
+		ret = drm_dp_dpcd_readb(mstm->mgr.aux, DP_MSTM_CAP, &dpcd[1]);
+		if (ret < 0)
+			return ret;
+
+		if (!(dpcd[1] & DP_MST_CAP))
+			dpcd[0] = 0x11;
+		else
+			state = allow;
+	}
+
+	ret = nv50_mstm_enable(mstm, dpcd[0], state);
+	if (ret)
+		return ret;
+
+	ret = drm_dp_mst_topology_mgr_set_mst(&mstm->mgr, state);
+	if (ret)
+		return nv50_mstm_enable(mstm, dpcd[0], 0);
+
+	return mstm->mgr.mst_state;
+}
+
+static void
+nv50_mstm_fini(struct nv50_mstm *mstm)
+{
+	if (mstm && mstm->mgr.mst_state)
+		drm_dp_mst_topology_mgr_suspend(&mstm->mgr);
+}
+
+static void
+nv50_mstm_init(struct nv50_mstm *mstm)
+{
+	if (mstm && mstm->mgr.mst_state)
+		drm_dp_mst_topology_mgr_resume(&mstm->mgr);
+}
+
+static void
+nv50_mstm_del(struct nv50_mstm **pmstm)
+{
+	struct nv50_mstm *mstm = *pmstm;
+	if (mstm) {
+		kfree(*pmstm);
+		*pmstm = NULL;
+	}
+}
+
+static int
+nv50_mstm_new(struct nouveau_encoder *outp, struct drm_dp_aux *aux, int aux_max,
+	      int conn_base_id, struct nv50_mstm **pmstm)
+{
+	const int max_payloads = hweight8(outp->dcb->heads);
+	struct drm_device *dev = outp->base.base.dev;
+	struct nv50_mstm *mstm;
+	int ret, i;
+	u8 dpcd;
+
+	/* This is a workaround for some monitors not functioning
+	 * correctly in MST mode on initial module load.  I think
+	 * some bad interaction with the VBIOS may be responsible.
+	 *
+	 * A good ol' off and on again seems to work here ;)
+	 */
+	ret = drm_dp_dpcd_readb(aux, DP_DPCD_REV, &dpcd);
+	if (ret >= 0 && dpcd >= 0x12)
+		drm_dp_dpcd_writeb(aux, DP_MSTM_CTRL, 0);
+
+	if (!(mstm = *pmstm = kzalloc(sizeof(*mstm), GFP_KERNEL)))
+		return -ENOMEM;
+	mstm->outp = outp;
+	mstm->mgr.cbs = &nv50_mstm;
+
+	ret = drm_dp_mst_topology_mgr_init(&mstm->mgr, dev, aux, aux_max,
+					   max_payloads, conn_base_id);
+	if (ret)
+		return ret;
+
+	for (i = 0; i < max_payloads; i++) {
+		ret = nv50_msto_new(dev, outp->dcb->heads, outp->base.base.name,
+				    i, &mstm->msto[i]);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+/******************************************************************************
+ * SOR
+ *****************************************************************************/
+static void
+nv50_sor_update(struct nouveau_encoder *nv_encoder, u8 head,
+		struct nv50_head_atom *asyh, u8 proto, u8 depth)
+{
+	struct nv50_disp *disp = nv50_disp(nv_encoder->base.base.dev);
+	struct nv50_core *core = disp->core;
+
+	if (!asyh) {
+		nv_encoder->ctrl &= ~BIT(head);
+		if (!(nv_encoder->ctrl & 0x0000000f))
+			nv_encoder->ctrl = 0;
+	} else {
+		nv_encoder->ctrl |= proto << 8;
+		nv_encoder->ctrl |= BIT(head);
+		asyh->or.depth = depth;
+	}
+
+	core->func->sor->ctrl(core, nv_encoder->or, nv_encoder->ctrl, asyh);
+}
+
+static void
+nv50_sor_disable(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(nv_encoder->crtc);
+
+	nv_encoder->crtc = NULL;
+
+	if (nv_crtc) {
+		struct nvkm_i2c_aux *aux = nv_encoder->aux;
+		u8 pwr;
+
+		if (aux) {
+			int ret = nvkm_rdaux(aux, DP_SET_POWER, &pwr, 1);
+			if (ret == 0) {
+				pwr &= ~DP_SET_POWER_MASK;
+				pwr |=  DP_SET_POWER_D3;
+				nvkm_wraux(aux, DP_SET_POWER, &pwr, 1);
+			}
+		}
+
+		nv_encoder->update(nv_encoder, nv_crtc->index, NULL, 0, 0);
+		nv50_audio_disable(encoder, nv_crtc);
+		nv50_hdmi_disable(&nv_encoder->base.base, nv_crtc);
+		nv50_outp_release(nv_encoder);
+	}
+}
+
+static void
+nv50_sor_enable(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
+	struct nv50_head_atom *asyh = nv50_head_atom(nv_crtc->base.state);
+	struct drm_display_mode *mode = &asyh->state.adjusted_mode;
+	struct {
+		struct nv50_disp_mthd_v1 base;
+		struct nv50_disp_sor_lvds_script_v0 lvds;
+	} lvds = {
+		.base.version = 1,
+		.base.method  = NV50_DISP_MTHD_V1_SOR_LVDS_SCRIPT,
+		.base.hasht   = nv_encoder->dcb->hasht,
+		.base.hashm   = nv_encoder->dcb->hashm,
+	};
+	struct nv50_disp *disp = nv50_disp(encoder->dev);
+	struct drm_device *dev = encoder->dev;
+	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nouveau_connector *nv_connector;
+	struct nvbios *bios = &drm->vbios;
+	u8 proto = 0xf;
+	u8 depth = 0x0;
+
+	nv_connector = nouveau_encoder_connector_get(nv_encoder);
+	nv_encoder->crtc = encoder->crtc;
+	nv50_outp_acquire(nv_encoder);
+
+	switch (nv_encoder->dcb->type) {
+	case DCB_OUTPUT_TMDS:
+		if (nv_encoder->link & 1) {
+			proto = 0x1;
+			/* Only enable dual-link if:
+			 *  - Need to (i.e. rate > 165MHz)
+			 *  - DCB says we can
+			 *  - Not an HDMI monitor, since there's no dual-link
+			 *    on HDMI.
+			 */
+			if (mode->clock >= 165000 &&
+			    nv_encoder->dcb->duallink_possible &&
+			    !drm_detect_hdmi_monitor(nv_connector->edid))
+				proto |= 0x4;
+		} else {
+			proto = 0x2;
+		}
+
+		nv50_hdmi_enable(&nv_encoder->base.base, mode);
+		break;
+	case DCB_OUTPUT_LVDS:
+		proto = 0x0;
+
+		if (bios->fp_no_ddc) {
+			if (bios->fp.dual_link)
+				lvds.lvds.script |= 0x0100;
+			if (bios->fp.if_is_24bit)
+				lvds.lvds.script |= 0x0200;
+		} else {
+			if (nv_connector->type == DCB_CONNECTOR_LVDS_SPWG) {
+				if (((u8 *)nv_connector->edid)[121] == 2)
+					lvds.lvds.script |= 0x0100;
+			} else
+			if (mode->clock >= bios->fp.duallink_transition_clk) {
+				lvds.lvds.script |= 0x0100;
+			}
+
+			if (lvds.lvds.script & 0x0100) {
+				if (bios->fp.strapless_is_24bit & 2)
+					lvds.lvds.script |= 0x0200;
+			} else {
+				if (bios->fp.strapless_is_24bit & 1)
+					lvds.lvds.script |= 0x0200;
+			}
+
+			if (nv_connector->base.display_info.bpc == 8)
+				lvds.lvds.script |= 0x0200;
+		}
+
+		nvif_mthd(&disp->disp->object, 0, &lvds, sizeof(lvds));
+		break;
+	case DCB_OUTPUT_DP:
+		if (nv_connector->base.display_info.bpc == 6)
+			depth = 0x2;
+		else
+		if (nv_connector->base.display_info.bpc == 8)
+			depth = 0x5;
+		else
+			depth = 0x6;
+
+		if (nv_encoder->link & 1)
+			proto = 0x8;
+		else
+			proto = 0x9;
+
+		nv50_audio_enable(encoder, mode);
+		break;
+	default:
+		BUG();
+		break;
+	}
+
+	nv_encoder->update(nv_encoder, nv_crtc->index, asyh, proto, depth);
+}
+
+static const struct drm_encoder_helper_funcs
+nv50_sor_help = {
+	.atomic_check = nv50_outp_atomic_check,
+	.enable = nv50_sor_enable,
+	.disable = nv50_sor_disable,
+};
+
+static void
+nv50_sor_destroy(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	nv50_mstm_del(&nv_encoder->dp.mstm);
+	drm_encoder_cleanup(encoder);
+	kfree(encoder);
+}
+
+static const struct drm_encoder_funcs
+nv50_sor_func = {
+	.destroy = nv50_sor_destroy,
+};
+
+static int
+nv50_sor_create(struct drm_connector *connector, struct dcb_output *dcbe)
+{
+	struct nouveau_connector *nv_connector = nouveau_connector(connector);
+	struct nouveau_drm *drm = nouveau_drm(connector->dev);
+	struct nvkm_bios *bios = nvxx_bios(&drm->client.device);
+	struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device);
+	struct nouveau_encoder *nv_encoder;
+	struct drm_encoder *encoder;
+	u8 ver, hdr, cnt, len;
+	u32 data;
+	int type, ret;
+
+	switch (dcbe->type) {
+	case DCB_OUTPUT_LVDS: type = DRM_MODE_ENCODER_LVDS; break;
+	case DCB_OUTPUT_TMDS:
+	case DCB_OUTPUT_DP:
+	default:
+		type = DRM_MODE_ENCODER_TMDS;
+		break;
+	}
+
+	nv_encoder = kzalloc(sizeof(*nv_encoder), GFP_KERNEL);
+	if (!nv_encoder)
+		return -ENOMEM;
+	nv_encoder->dcb = dcbe;
+	nv_encoder->update = nv50_sor_update;
+
+	encoder = to_drm_encoder(nv_encoder);
+	encoder->possible_crtcs = dcbe->heads;
+	encoder->possible_clones = 0;
+	drm_encoder_init(connector->dev, encoder, &nv50_sor_func, type,
+			 "sor-%04x-%04x", dcbe->hasht, dcbe->hashm);
+	drm_encoder_helper_add(encoder, &nv50_sor_help);
+
+	drm_connector_attach_encoder(connector, encoder);
+
+	if (dcbe->type == DCB_OUTPUT_DP) {
+		struct nv50_disp *disp = nv50_disp(encoder->dev);
+		struct nvkm_i2c_aux *aux =
+			nvkm_i2c_aux_find(i2c, dcbe->i2c_index);
+		if (aux) {
+			if (disp->disp->object.oclass < GF110_DISP) {
+				/* HW has no support for address-only
+				 * transactions, so we're required to
+				 * use custom I2C-over-AUX code.
+				 */
+				nv_encoder->i2c = &aux->i2c;
+			} else {
+				nv_encoder->i2c = &nv_connector->aux.ddc;
+			}
+			nv_encoder->aux = aux;
+		}
+
+		if ((data = nvbios_dp_table(bios, &ver, &hdr, &cnt, &len)) &&
+		    ver >= 0x40 && (nvbios_rd08(bios, data + 0x08) & 0x04)) {
+			ret = nv50_mstm_new(nv_encoder, &nv_connector->aux, 16,
+					    nv_connector->base.base.id,
+					    &nv_encoder->dp.mstm);
+			if (ret)
+				return ret;
+		}
+	} else {
+		struct nvkm_i2c_bus *bus =
+			nvkm_i2c_bus_find(i2c, dcbe->i2c_index);
+		if (bus)
+			nv_encoder->i2c = &bus->i2c;
+	}
+
+	return 0;
+}
+
+/******************************************************************************
+ * PIOR
+ *****************************************************************************/
+static int
+nv50_pior_atomic_check(struct drm_encoder *encoder,
+		       struct drm_crtc_state *crtc_state,
+		       struct drm_connector_state *conn_state)
+{
+	int ret = nv50_outp_atomic_check(encoder, crtc_state, conn_state);
+	if (ret)
+		return ret;
+	crtc_state->adjusted_mode.clock *= 2;
+	return 0;
+}
+
+static void
+nv50_pior_disable(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nv50_core *core = nv50_disp(encoder->dev)->core;
+	if (nv_encoder->crtc)
+		core->func->pior->ctrl(core, nv_encoder->or, 0x00000000, NULL);
+	nv_encoder->crtc = NULL;
+	nv50_outp_release(nv_encoder);
+}
+
+static void
+nv50_pior_enable(struct drm_encoder *encoder)
+{
+	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
+	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
+	struct nouveau_connector *nv_connector;
+	struct nv50_head_atom *asyh = nv50_head_atom(nv_crtc->base.state);
+	struct nv50_core *core = nv50_disp(encoder->dev)->core;
+	u8 owner = 1 << nv_crtc->index;
+	u8 proto;
+
+	nv50_outp_acquire(nv_encoder);
+
+	nv_connector = nouveau_encoder_connector_get(nv_encoder);
+	switch (nv_connector->base.display_info.bpc) {
+	case 10: asyh->or.depth = 0x6; break;
+	case  8: asyh->or.depth = 0x5; break;
+	case  6: asyh->or.depth = 0x2; break;
+	default: asyh->or.depth = 0x0; break;
+	}
+
+	switch (nv_encoder->dcb->type) {
+	case DCB_OUTPUT_TMDS:
+	case DCB_OUTPUT_DP:
+		proto = 0x0;
+		break;
+	default:
+		BUG();
+		break;
+	}
+
+	core->func->pior->ctrl(core, nv_encoder->or, (proto << 8) | owner, asyh);
+	nv_encoder->crtc = encoder->crtc;
+}
+
+static const struct drm_encoder_helper_funcs
+nv50_pior_help = {
+	.atomic_check = nv50_pior_atomic_check,
+	.enable = nv50_pior_enable,
+	.disable = nv50_pior_disable,
+};
+
+static void
+nv50_pior_destroy(struct drm_encoder *encoder)
+{
+	drm_encoder_cleanup(encoder);
+	kfree(encoder);
+}
+
+static const struct drm_encoder_funcs
+nv50_pior_func = {
+	.destroy = nv50_pior_destroy,
+};
+
+static int
+nv50_pior_create(struct drm_connector *connector, struct dcb_output *dcbe)
+{
+	struct nouveau_drm *drm = nouveau_drm(connector->dev);
+	struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device);
+	struct nvkm_i2c_bus *bus = NULL;
+	struct nvkm_i2c_aux *aux = NULL;
+	struct i2c_adapter *ddc;
+	struct nouveau_encoder *nv_encoder;
+	struct drm_encoder *encoder;
+	int type;
+
+	switch (dcbe->type) {
+	case DCB_OUTPUT_TMDS:
+		bus  = nvkm_i2c_bus_find(i2c, NVKM_I2C_BUS_EXT(dcbe->extdev));
+		ddc  = bus ? &bus->i2c : NULL;
+		type = DRM_MODE_ENCODER_TMDS;
+		break;
+	case DCB_OUTPUT_DP:
+		aux  = nvkm_i2c_aux_find(i2c, NVKM_I2C_AUX_EXT(dcbe->extdev));
+		ddc  = aux ? &aux->i2c : NULL;
+		type = DRM_MODE_ENCODER_TMDS;
+		break;
+	default:
+		return -ENODEV;
+	}
+
+	nv_encoder = kzalloc(sizeof(*nv_encoder), GFP_KERNEL);
+	if (!nv_encoder)
+		return -ENOMEM;
+	nv_encoder->dcb = dcbe;
+	nv_encoder->i2c = ddc;
+	nv_encoder->aux = aux;
+
+	encoder = to_drm_encoder(nv_encoder);
+	encoder->possible_crtcs = dcbe->heads;
+	encoder->possible_clones = 0;
+	drm_encoder_init(connector->dev, encoder, &nv50_pior_func, type,
+			 "pior-%04x-%04x", dcbe->hasht, dcbe->hashm);
+	drm_encoder_helper_add(encoder, &nv50_pior_help);
+
+	drm_connector_attach_encoder(connector, encoder);
+	return 0;
+}
+
+/******************************************************************************
+ * Atomic
+ *****************************************************************************/
+
+static void
+nv50_disp_atomic_commit_core(struct nouveau_drm *drm, u32 *interlock)
+{
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_core *core = disp->core;
+	struct nv50_mstm *mstm;
+	struct drm_encoder *encoder;
+
+	NV_ATOMIC(drm, "commit core %08x\n", interlock[NV50_DISP_INTERLOCK_BASE]);
+
+	drm_for_each_encoder(encoder, drm->dev) {
+		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
+			mstm = nouveau_encoder(encoder)->dp.mstm;
+			if (mstm && mstm->modified)
+				nv50_mstm_prepare(mstm);
+		}
+	}
+
+	core->func->ntfy_init(disp->sync, NV50_DISP_CORE_NTFY);
+	core->func->update(core, interlock, true);
+	if (core->func->ntfy_wait_done(disp->sync, NV50_DISP_CORE_NTFY,
+				       disp->core->chan.base.device))
+		NV_ERROR(drm, "core notifier timeout\n");
+
+	drm_for_each_encoder(encoder, drm->dev) {
+		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
+			mstm = nouveau_encoder(encoder)->dp.mstm;
+			if (mstm && mstm->modified)
+				nv50_mstm_cleanup(mstm);
+		}
+	}
+}
+
+static void
+nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
+{
+	struct drm_device *dev = state->dev;
+	struct drm_crtc_state *new_crtc_state, *old_crtc_state;
+	struct drm_crtc *crtc;
+	struct drm_plane_state *new_plane_state;
+	struct drm_plane *plane;
+	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nv50_disp *disp = nv50_disp(dev);
+	struct nv50_atom *atom = nv50_atom(state);
+	struct nv50_outp_atom *outp, *outt;
+	u32 interlock[NV50_DISP_INTERLOCK__SIZE] = {};
+	int i;
+
+	NV_ATOMIC(drm, "commit %d %d\n", atom->lock_core, atom->flush_disable);
+	drm_atomic_helper_wait_for_fences(dev, state, false);
+	drm_atomic_helper_wait_for_dependencies(state);
+	drm_atomic_helper_update_legacy_modeset_state(dev, state);
+
+	if (atom->lock_core)
+		mutex_lock(&disp->mutex);
+
+	/* Disable head(s). */
+	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
+		struct nv50_head_atom *asyh = nv50_head_atom(new_crtc_state);
+		struct nv50_head *head = nv50_head(crtc);
+
+		NV_ATOMIC(drm, "%s: clr %04x (set %04x)\n", crtc->name,
+			  asyh->clr.mask, asyh->set.mask);
+		if (old_crtc_state->active && !new_crtc_state->active)
+			drm_crtc_vblank_off(crtc);
+
+		if (asyh->clr.mask) {
+			nv50_head_flush_clr(head, asyh, atom->flush_disable);
+			interlock[NV50_DISP_INTERLOCK_CORE] |= 1;
+		}
+	}
+
+	/* Disable plane(s). */
+	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
+		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
+		struct nv50_wndw *wndw = nv50_wndw(plane);
+
+		NV_ATOMIC(drm, "%s: clr %02x (set %02x)\n", plane->name,
+			  asyw->clr.mask, asyw->set.mask);
+		if (!asyw->clr.mask)
+			continue;
+
+		nv50_wndw_flush_clr(wndw, interlock, atom->flush_disable, asyw);
+	}
+
+	/* Disable output path(s). */
+	list_for_each_entry(outp, &atom->outp, head) {
+		const struct drm_encoder_helper_funcs *help;
+		struct drm_encoder *encoder;
+
+		encoder = outp->encoder;
+		help = encoder->helper_private;
+
+		NV_ATOMIC(drm, "%s: clr %02x (set %02x)\n", encoder->name,
+			  outp->clr.mask, outp->set.mask);
+
+		if (outp->clr.mask) {
+			help->disable(encoder);
+			interlock[NV50_DISP_INTERLOCK_CORE] |= 1;
+			if (outp->flush_disable) {
+				nv50_disp_atomic_commit_core(drm, interlock);
+				memset(interlock, 0x00, sizeof(interlock));
+			}
+		}
+	}
+
+	/* Flush disable. */
+	if (interlock[NV50_DISP_INTERLOCK_CORE]) {
+		if (atom->flush_disable) {
+			for_each_new_plane_in_state(state, plane, new_plane_state, i) {
+				struct nv50_wndw *wndw = nv50_wndw(plane);
+				if (interlock[wndw->interlock.type] & wndw->interlock.data) {
+					if (wndw->func->update)
+						wndw->func->update(wndw, interlock);
+				}
+			}
+
+			nv50_disp_atomic_commit_core(drm, interlock);
+			memset(interlock, 0x00, sizeof(interlock));
+		}
+	}
+
+	/* Update output path(s). */
+	list_for_each_entry_safe(outp, outt, &atom->outp, head) {
+		const struct drm_encoder_helper_funcs *help;
+		struct drm_encoder *encoder;
+
+		encoder = outp->encoder;
+		help = encoder->helper_private;
+
+		NV_ATOMIC(drm, "%s: set %02x (clr %02x)\n", encoder->name,
+			  outp->set.mask, outp->clr.mask);
+
+		if (outp->set.mask) {
+			help->enable(encoder);
+			interlock[NV50_DISP_INTERLOCK_CORE] = 1;
+		}
+
+		list_del(&outp->head);
+		kfree(outp);
+	}
+
+	/* Update head(s). */
+	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
+		struct nv50_head_atom *asyh = nv50_head_atom(new_crtc_state);
+		struct nv50_head *head = nv50_head(crtc);
+
+		NV_ATOMIC(drm, "%s: set %04x (clr %04x)\n", crtc->name,
+			  asyh->set.mask, asyh->clr.mask);
+
+		if (asyh->set.mask) {
+			nv50_head_flush_set(head, asyh);
+			interlock[NV50_DISP_INTERLOCK_CORE] = 1;
+		}
+
+		if (new_crtc_state->active) {
+			if (!old_crtc_state->active)
+				drm_crtc_vblank_on(crtc);
+			if (new_crtc_state->event)
+				drm_crtc_vblank_get(crtc);
+		}
+	}
+
+	/* Update plane(s). */
+	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
+		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
+		struct nv50_wndw *wndw = nv50_wndw(plane);
+
+		NV_ATOMIC(drm, "%s: set %02x (clr %02x)\n", plane->name,
+			  asyw->set.mask, asyw->clr.mask);
+		if ( !asyw->set.mask &&
+		    (!asyw->clr.mask || atom->flush_disable))
+			continue;
+
+		nv50_wndw_flush_set(wndw, interlock, asyw);
+	}
+
+	/* Flush update. */
+	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
+		struct nv50_wndw *wndw = nv50_wndw(plane);
+		if (interlock[wndw->interlock.type] & wndw->interlock.data) {
+			if (wndw->func->update)
+				wndw->func->update(wndw, interlock);
+		}
+	}
+
+	if (interlock[NV50_DISP_INTERLOCK_CORE]) {
+		if (interlock[NV50_DISP_INTERLOCK_BASE] ||
+		    !atom->state.legacy_cursor_update)
+			nv50_disp_atomic_commit_core(drm, interlock);
+		else
+			disp->core->func->update(disp->core, interlock, false);
+	}
+
+	if (atom->lock_core)
+		mutex_unlock(&disp->mutex);
+
+	/* Wait for HW to signal completion. */
+	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
+		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
+		struct nv50_wndw *wndw = nv50_wndw(plane);
+		int ret = nv50_wndw_wait_armed(wndw, asyw);
+		if (ret)
+			NV_ERROR(drm, "%s: timeout\n", plane->name);
+	}
+
+	for_each_new_crtc_in_state(state, crtc, new_crtc_state, i) {
+		if (new_crtc_state->event) {
+			unsigned long flags;
+			/* Get correct count/ts if racing with vblank irq */
+			if (new_crtc_state->active)
+				drm_crtc_accurate_vblank_count(crtc);
+			spin_lock_irqsave(&crtc->dev->event_lock, flags);
+			drm_crtc_send_vblank_event(crtc, new_crtc_state->event);
+			spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
+
+			new_crtc_state->event = NULL;
+			if (new_crtc_state->active)
+				drm_crtc_vblank_put(crtc);
+		}
+	}
+
+	drm_atomic_helper_commit_hw_done(state);
+	drm_atomic_helper_cleanup_planes(dev, state);
+	drm_atomic_helper_commit_cleanup_done(state);
+	drm_atomic_state_put(state);
+}
+
+static void
+nv50_disp_atomic_commit_work(struct work_struct *work)
+{
+	struct drm_atomic_state *state =
+		container_of(work, typeof(*state), commit_work);
+	nv50_disp_atomic_commit_tail(state);
+}
+
+static int
+nv50_disp_atomic_commit(struct drm_device *dev,
+			struct drm_atomic_state *state, bool nonblock)
+{
+	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct drm_plane_state *new_plane_state;
+	struct drm_plane *plane;
+	struct drm_crtc *crtc;
+	bool active = false;
+	int ret, i;
+
+	ret = pm_runtime_get_sync(dev->dev);
+	if (ret < 0 && ret != -EACCES)
+		return ret;
+
+	ret = drm_atomic_helper_setup_commit(state, nonblock);
+	if (ret)
+		goto done;
+
+	INIT_WORK(&state->commit_work, nv50_disp_atomic_commit_work);
+
+	ret = drm_atomic_helper_prepare_planes(dev, state);
+	if (ret)
+		goto done;
+
+	if (!nonblock) {
+		ret = drm_atomic_helper_wait_for_fences(dev, state, true);
+		if (ret)
+			goto err_cleanup;
+	}
+
+	ret = drm_atomic_helper_swap_state(state, true);
+	if (ret)
+		goto err_cleanup;
+
+	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
+		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
+		struct nv50_wndw *wndw = nv50_wndw(plane);
+
+		if (asyw->set.image)
+			nv50_wndw_ntfy_enable(wndw, asyw);
+	}
+
+	drm_atomic_state_get(state);
+
+	if (nonblock)
+		queue_work(system_unbound_wq, &state->commit_work);
+	else
+		nv50_disp_atomic_commit_tail(state);
+
+	drm_for_each_crtc(crtc, dev) {
+		if (crtc->state->enable) {
+			if (!drm->have_disp_power_ref) {
+				drm->have_disp_power_ref = true;
+				return 0;
+			}
+			active = true;
+			break;
+		}
+	}
+
+	if (!active && drm->have_disp_power_ref) {
+		pm_runtime_put_autosuspend(dev->dev);
+		drm->have_disp_power_ref = false;
+	}
+
+err_cleanup:
+	if (ret)
+		drm_atomic_helper_cleanup_planes(dev, state);
+done:
+	pm_runtime_put_autosuspend(dev->dev);
+	return ret;
+}
+
+static struct nv50_outp_atom *
+nv50_disp_outp_atomic_add(struct nv50_atom *atom, struct drm_encoder *encoder)
+{
+	struct nv50_outp_atom *outp;
+
+	list_for_each_entry(outp, &atom->outp, head) {
+		if (outp->encoder == encoder)
+			return outp;
+	}
+
+	outp = kzalloc(sizeof(*outp), GFP_KERNEL);
+	if (!outp)
+		return ERR_PTR(-ENOMEM);
+
+	list_add(&outp->head, &atom->outp);
+	outp->encoder = encoder;
+	return outp;
+}
+
+static int
+nv50_disp_outp_atomic_check_clr(struct nv50_atom *atom,
+				struct drm_connector_state *old_connector_state)
+{
+	struct drm_encoder *encoder = old_connector_state->best_encoder;
+	struct drm_crtc_state *old_crtc_state, *new_crtc_state;
+	struct drm_crtc *crtc;
+	struct nv50_outp_atom *outp;
+
+	if (!(crtc = old_connector_state->crtc))
+		return 0;
+
+	old_crtc_state = drm_atomic_get_old_crtc_state(&atom->state, crtc);
+	new_crtc_state = drm_atomic_get_new_crtc_state(&atom->state, crtc);
+	if (old_crtc_state->active && drm_atomic_crtc_needs_modeset(new_crtc_state)) {
+		outp = nv50_disp_outp_atomic_add(atom, encoder);
+		if (IS_ERR(outp))
+			return PTR_ERR(outp);
+
+		if (outp->encoder->encoder_type == DRM_MODE_ENCODER_DPMST) {
+			outp->flush_disable = true;
+			atom->flush_disable = true;
+		}
+		outp->clr.ctrl = true;
+		atom->lock_core = true;
+	}
+
+	return 0;
+}
+
+static int
+nv50_disp_outp_atomic_check_set(struct nv50_atom *atom,
+				struct drm_connector_state *connector_state)
+{
+	struct drm_encoder *encoder = connector_state->best_encoder;
+	struct drm_crtc_state *new_crtc_state;
+	struct drm_crtc *crtc;
+	struct nv50_outp_atom *outp;
+
+	if (!(crtc = connector_state->crtc))
+		return 0;
+
+	new_crtc_state = drm_atomic_get_new_crtc_state(&atom->state, crtc);
+	if (new_crtc_state->active && drm_atomic_crtc_needs_modeset(new_crtc_state)) {
+		outp = nv50_disp_outp_atomic_add(atom, encoder);
+		if (IS_ERR(outp))
+			return PTR_ERR(outp);
+
+		outp->set.ctrl = true;
+		atom->lock_core = true;
+	}
+
+	return 0;
+}
+
+static int
+nv50_disp_atomic_check(struct drm_device *dev, struct drm_atomic_state *state)
+{
+	struct nv50_atom *atom = nv50_atom(state);
+	struct drm_connector_state *old_connector_state, *new_connector_state;
+	struct drm_connector *connector;
+	struct drm_crtc_state *new_crtc_state;
+	struct drm_crtc *crtc;
+	int ret, i;
+
+	/* We need to handle colour management on a per-plane basis. */
+	for_each_new_crtc_in_state(state, crtc, new_crtc_state, i) {
+		if (new_crtc_state->color_mgmt_changed) {
+			ret = drm_atomic_add_affected_planes(state, crtc);
+			if (ret)
+				return ret;
+		}
+	}
+
+	ret = drm_atomic_helper_check(dev, state);
+	if (ret)
+		return ret;
+
+	for_each_oldnew_connector_in_state(state, connector, old_connector_state, new_connector_state, i) {
+		ret = nv50_disp_outp_atomic_check_clr(atom, old_connector_state);
+		if (ret)
+			return ret;
+
+		ret = nv50_disp_outp_atomic_check_set(atom, new_connector_state);
+		if (ret)
+			return ret;
+	}
+
+	return 0;
+}
+
+static void
+nv50_disp_atomic_state_clear(struct drm_atomic_state *state)
+{
+	struct nv50_atom *atom = nv50_atom(state);
+	struct nv50_outp_atom *outp, *outt;
+
+	list_for_each_entry_safe(outp, outt, &atom->outp, head) {
+		list_del(&outp->head);
+		kfree(outp);
+	}
+
+	drm_atomic_state_default_clear(state);
+}
+
+static void
+nv50_disp_atomic_state_free(struct drm_atomic_state *state)
+{
+	struct nv50_atom *atom = nv50_atom(state);
+	drm_atomic_state_default_release(&atom->state);
+	kfree(atom);
+}
+
+static struct drm_atomic_state *
+nv50_disp_atomic_state_alloc(struct drm_device *dev)
+{
+	struct nv50_atom *atom;
+	if (!(atom = kzalloc(sizeof(*atom), GFP_KERNEL)) ||
+	    drm_atomic_state_init(dev, &atom->state) < 0) {
+		kfree(atom);
+		return NULL;
+	}
+	INIT_LIST_HEAD(&atom->outp);
+	return &atom->state;
+}
+
+static const struct drm_mode_config_funcs
+nv50_disp_func = {
+	.fb_create = nouveau_user_framebuffer_create,
+	.output_poll_changed = drm_fb_helper_output_poll_changed,
+	.atomic_check = nv50_disp_atomic_check,
+	.atomic_commit = nv50_disp_atomic_commit,
+	.atomic_state_alloc = nv50_disp_atomic_state_alloc,
+	.atomic_state_clear = nv50_disp_atomic_state_clear,
+	.atomic_state_free = nv50_disp_atomic_state_free,
+};
+
+/******************************************************************************
+ * Init
+ *****************************************************************************/
+
+void
+nv50_display_fini(struct drm_device *dev)
+{
+	struct nouveau_encoder *nv_encoder;
+	struct drm_encoder *encoder;
+	struct drm_plane *plane;
+
+	drm_for_each_plane(plane, dev) {
+		struct nv50_wndw *wndw = nv50_wndw(plane);
+		if (plane->funcs != &nv50_wndw)
+			continue;
+		nv50_wndw_fini(wndw);
+	}
+
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
+		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
+			nv_encoder = nouveau_encoder(encoder);
+			nv50_mstm_fini(nv_encoder->dp.mstm);
+		}
+	}
+}
+
+int
+nv50_display_init(struct drm_device *dev)
+{
+	struct nv50_core *core = nv50_disp(dev)->core;
+	struct drm_encoder *encoder;
+	struct drm_plane *plane;
+
+	core->func->init(core);
+
+	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
+		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
+			struct nouveau_encoder *nv_encoder =
+				nouveau_encoder(encoder);
+			nv50_mstm_init(nv_encoder->dp.mstm);
+		}
+	}
+
+	drm_for_each_plane(plane, dev) {
+		struct nv50_wndw *wndw = nv50_wndw(plane);
+		if (plane->funcs != &nv50_wndw)
+			continue;
+		nv50_wndw_init(wndw);
+	}
+
+	return 0;
+}
+
+void
+nv50_display_destroy(struct drm_device *dev)
+{
+	struct nv50_disp *disp = nv50_disp(dev);
+
+	nv50_core_del(&disp->core);
+
+	nouveau_bo_unmap(disp->sync);
+	if (disp->sync)
+		nouveau_bo_unpin(disp->sync);
+	nouveau_bo_ref(NULL, &disp->sync);
+
+	nouveau_display(dev)->priv = NULL;
+	kfree(disp);
+}
+
+MODULE_PARM_DESC(atomic, "Expose atomic ioctl (default: disabled)");
+static int nouveau_atomic = 0;
+module_param_named(atomic, nouveau_atomic, int, 0400);
+
+int
+nv50_display_create(struct drm_device *dev)
+{
+	struct nvif_device *device = &nouveau_drm(dev)->client.device;
+	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct dcb_table *dcb = &drm->vbios.dcb;
+	struct drm_connector *connector, *tmp;
+	struct nv50_disp *disp;
+	struct dcb_output *dcbe;
+	int crtcs, ret, i;
+
+	disp = kzalloc(sizeof(*disp), GFP_KERNEL);
+	if (!disp)
+		return -ENOMEM;
+
+	mutex_init(&disp->mutex);
+
+	nouveau_display(dev)->priv = disp;
+	nouveau_display(dev)->dtor = nv50_display_destroy;
+	nouveau_display(dev)->init = nv50_display_init;
+	nouveau_display(dev)->fini = nv50_display_fini;
+	disp->disp = &nouveau_display(dev)->disp;
+	dev->mode_config.funcs = &nv50_disp_func;
+	dev->driver->driver_features |= DRIVER_PREFER_XBGR_30BPP;
+	if (nouveau_atomic)
+		dev->driver->driver_features |= DRIVER_ATOMIC;
+
+	/* small shared memory area we use for notifiers and semaphores */
+	ret = nouveau_bo_new(&drm->client, 4096, 0x1000, TTM_PL_FLAG_VRAM,
+			     0, 0x0000, NULL, NULL, &disp->sync);
+	if (!ret) {
+		ret = nouveau_bo_pin(disp->sync, TTM_PL_FLAG_VRAM, true);
+		if (!ret) {
+			ret = nouveau_bo_map(disp->sync);
+			if (ret)
+				nouveau_bo_unpin(disp->sync);
+		}
+		if (ret)
+			nouveau_bo_ref(NULL, &disp->sync);
+	}
+
+	if (ret)
+		goto out;
+
+	/* allocate master evo channel */
+	ret = nv50_core_new(drm, &disp->core);
+	if (ret)
+		goto out;
+
+	/* create crtc objects to represent the hw heads */
+	if (disp->disp->object.oclass >= GV100_DISP)
+		crtcs = nvif_rd32(&device->object, 0x610060) & 0xff;
+	else
+	if (disp->disp->object.oclass >= GF110_DISP)
+		crtcs = nvif_rd32(&device->object, 0x612004) & 0xf;
+	else
+		crtcs = 0x3;
+
+	for (i = 0; i < fls(crtcs); i++) {
+		if (!(crtcs & (1 << i)))
+			continue;
+		ret = nv50_head_create(dev, i);
+		if (ret)
+			goto out;
+	}
+
+	/* create encoder/connector objects based on VBIOS DCB table */
+	for (i = 0, dcbe = &dcb->entry[0]; i < dcb->entries; i++, dcbe++) {
+		connector = nouveau_connector_create(dev, dcbe->connector);
+		if (IS_ERR(connector))
+			continue;
+
+		if (dcbe->location == DCB_LOC_ON_CHIP) {
+			switch (dcbe->type) {
+			case DCB_OUTPUT_TMDS:
+			case DCB_OUTPUT_LVDS:
+			case DCB_OUTPUT_DP:
+				ret = nv50_sor_create(connector, dcbe);
+				break;
+			case DCB_OUTPUT_ANALOG:
+				ret = nv50_dac_create(connector, dcbe);
+				break;
+			default:
+				ret = -ENODEV;
+				break;
+			}
+		} else {
+			ret = nv50_pior_create(connector, dcbe);
+		}
+
+		if (ret) {
+			NV_WARN(drm, "failed to create encoder %d/%d/%d: %d\n",
+				     dcbe->location, dcbe->type,
+				     ffs(dcbe->or) - 1, ret);
+			ret = 0;
+		}
+	}
+
+	/* cull any connectors we created that don't have an encoder */
+	list_for_each_entry_safe(connector, tmp, &dev->mode_config.connector_list, head) {
+		if (connector->encoder_ids[0])
+			continue;
+
+		NV_WARN(drm, "%s has no encoders, removing\n",
+			connector->name);
+		connector->funcs->destroy(connector);
+	}
+
+	/* Disable vblank irqs aggressively for power-saving, safe on nv50+ */
+	dev->vblank_disable_immediate = true;
+
+out:
+	if (ret)
+		nv50_display_destroy(dev);
+	return ret;
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/disp.h b/drivers/gpu/drm/nouveau/dispnv50/disp.h
new file mode 100644
index 000000000000..e48c5eb35b49
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/disp.h
@@ -0,0 +1,89 @@
+#ifndef __NV50_KMS_H__
+#define __NV50_KMS_H__
+#include <nvif/mem.h>
+
+#include "nouveau_display.h"
+
+struct nv50_disp {
+	struct nvif_disp *disp;
+	struct nv50_core *core;
+
+#define NV50_DISP_SYNC(c, o)                                ((c) * 0x040 + (o))
+#define NV50_DISP_CORE_NTFY                       NV50_DISP_SYNC(0      , 0x00)
+#define NV50_DISP_WNDW_SEM0(c)                    NV50_DISP_SYNC(1 + (c), 0x00)
+#define NV50_DISP_WNDW_SEM1(c)                    NV50_DISP_SYNC(1 + (c), 0x10)
+#define NV50_DISP_WNDW_NTFY(c)                    NV50_DISP_SYNC(1 + (c), 0x20)
+#define NV50_DISP_BASE_SEM0(c)                    NV50_DISP_WNDW_SEM0(0 + (c))
+#define NV50_DISP_BASE_SEM1(c)                    NV50_DISP_WNDW_SEM1(0 + (c))
+#define NV50_DISP_BASE_NTFY(c)                    NV50_DISP_WNDW_NTFY(0 + (c))
+#define NV50_DISP_OVLY_SEM0(c)                    NV50_DISP_WNDW_SEM0(4 + (c))
+#define NV50_DISP_OVLY_SEM1(c)                    NV50_DISP_WNDW_SEM1(4 + (c))
+#define NV50_DISP_OVLY_NTFY(c)                    NV50_DISP_WNDW_NTFY(4 + (c))
+	struct nouveau_bo *sync;
+
+	struct mutex mutex;
+};
+
+static inline struct nv50_disp *
+nv50_disp(struct drm_device *dev)
+{
+	return nouveau_display(dev)->priv;
+}
+
+struct nv50_disp_interlock {
+	enum nv50_disp_interlock_type {
+		NV50_DISP_INTERLOCK_CORE = 0,
+		NV50_DISP_INTERLOCK_CURS,
+		NV50_DISP_INTERLOCK_BASE,
+		NV50_DISP_INTERLOCK_OVLY,
+		NV50_DISP_INTERLOCK_WNDW,
+		NV50_DISP_INTERLOCK_WIMM,
+		NV50_DISP_INTERLOCK__SIZE
+	} type;
+	u32 data;
+};
+
+void corec37d_ntfy_init(struct nouveau_bo *, u32);
+
+struct nv50_chan {
+	struct nvif_object user;
+	struct nvif_device *device;
+};
+
+struct nv50_dmac {
+	struct nv50_chan base;
+
+	struct nvif_mem push;
+	u32 *ptr;
+
+	struct nvif_object sync;
+	struct nvif_object vram;
+
+	/* Protects against concurrent pushbuf access to this channel, lock is
+	 * grabbed by evo_wait (if the pushbuf reservation is successful) and
+	 * dropped again by evo_kick. */
+	struct mutex lock;
+};
+
+int nv50_dmac_create(struct nvif_device *device, struct nvif_object *disp,
+		     const s32 *oclass, u8 head, void *data, u32 size,
+		     u64 syncbuf, struct nv50_dmac *dmac);
+void nv50_dmac_destroy(struct nv50_dmac *);
+
+u32 *evo_wait(struct nv50_dmac *, int nr);
+void evo_kick(u32 *, struct nv50_dmac *);
+
+#define evo_mthd(p, m, s) do {						\
+	const u32 _m = (m), _s = (s);					\
+	if (drm_debug & DRM_UT_KMS)					\
+		pr_err("%04x %d %s\n", _m, _s, __func__);		\
+	*((p)++) = ((_s << 18) | _m);					\
+} while(0)
+
+#define evo_data(p, d) do {						\
+	const u32 _d = (d);						\
+	if (drm_debug & DRM_UT_KMS)					\
+		pr_err("\t%08x\n", _d);					\
+	*((p)++) = _d;							\
+} while(0)
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head.c b/drivers/gpu/drm/nouveau/dispnv50/head.c
new file mode 100644
index 000000000000..4f57e5379796
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/head.c
@@ -0,0 +1,511 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "head.h"
+#include "base.h"
+#include "core.h"
+#include "curs.h"
+#include "ovly.h"
+
+#include <nvif/class.h>
+
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include "nouveau_connector.h"
+void
+nv50_head_flush_clr(struct nv50_head *head,
+		    struct nv50_head_atom *asyh, bool flush)
+{
+	union nv50_head_atom_mask clr = {
+		.mask = asyh->clr.mask & ~(flush ? 0 : asyh->set.mask),
+	};
+	if (clr.olut) head->func->olut_clr(head);
+	if (clr.core) head->func->core_clr(head);
+	if (clr.curs) head->func->curs_clr(head);
+}
+
+void
+nv50_head_flush_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	if (asyh->set.view   ) head->func->view    (head, asyh);
+	if (asyh->set.mode   ) head->func->mode    (head, asyh);
+	if (asyh->set.core   ) head->func->core_set(head, asyh);
+	if (asyh->set.olut   ) {
+		asyh->olut.offset = nv50_lut_load(&head->olut,
+						  asyh->olut.mode <= 1,
+						  asyh->olut.buffer,
+						  asyh->state.gamma_lut);
+		head->func->olut_set(head, asyh);
+	}
+	if (asyh->set.curs   ) head->func->curs_set(head, asyh);
+	if (asyh->set.base   ) head->func->base    (head, asyh);
+	if (asyh->set.ovly   ) head->func->ovly    (head, asyh);
+	if (asyh->set.dither ) head->func->dither  (head, asyh);
+	if (asyh->set.procamp) head->func->procamp (head, asyh);
+	if (asyh->set.or     ) head->func->or      (head, asyh);
+}
+
+static void
+nv50_head_atomic_check_procamp(struct nv50_head_atom *armh,
+			       struct nv50_head_atom *asyh,
+			       struct nouveau_conn_atom *asyc)
+{
+	const int vib = asyc->procamp.color_vibrance - 100;
+	const int hue = asyc->procamp.vibrant_hue - 90;
+	const int adj = (vib > 0) ? 50 : 0;
+	asyh->procamp.sat.cos = ((vib * 2047 + adj) / 100) & 0xfff;
+	asyh->procamp.sat.sin = ((hue * 2047) / 100) & 0xfff;
+	asyh->set.procamp = true;
+}
+
+static void
+nv50_head_atomic_check_dither(struct nv50_head_atom *armh,
+			      struct nv50_head_atom *asyh,
+			      struct nouveau_conn_atom *asyc)
+{
+	struct drm_connector *connector = asyc->state.connector;
+	u32 mode = 0x00;
+
+	if (asyc->dither.mode == DITHERING_MODE_AUTO) {
+		if (asyh->base.depth > connector->display_info.bpc * 3)
+			mode = DITHERING_MODE_DYNAMIC2X2;
+	} else {
+		mode = asyc->dither.mode;
+	}
+
+	if (asyc->dither.depth == DITHERING_DEPTH_AUTO) {
+		if (connector->display_info.bpc >= 8)
+			mode |= DITHERING_DEPTH_8BPC;
+	} else {
+		mode |= asyc->dither.depth;
+	}
+
+	asyh->dither.enable = mode;
+	asyh->dither.bits = mode >> 1;
+	asyh->dither.mode = mode >> 3;
+	asyh->set.dither = true;
+}
+
+static void
+nv50_head_atomic_check_view(struct nv50_head_atom *armh,
+			    struct nv50_head_atom *asyh,
+			    struct nouveau_conn_atom *asyc)
+{
+	struct drm_connector *connector = asyc->state.connector;
+	struct drm_display_mode *omode = &asyh->state.adjusted_mode;
+	struct drm_display_mode *umode = &asyh->state.mode;
+	int mode = asyc->scaler.mode;
+	struct edid *edid;
+	int umode_vdisplay, omode_hdisplay, omode_vdisplay;
+
+	if (connector->edid_blob_ptr)
+		edid = (struct edid *)connector->edid_blob_ptr->data;
+	else
+		edid = NULL;
+
+	if (!asyc->scaler.full) {
+		if (mode == DRM_MODE_SCALE_NONE)
+			omode = umode;
+	} else {
+		/* Non-EDID LVDS/eDP mode. */
+		mode = DRM_MODE_SCALE_FULLSCREEN;
+	}
+
+	/* For the user-specified mode, we must ignore doublescan and
+	 * the like, but honor frame packing.
+	 */
+	umode_vdisplay = umode->vdisplay;
+	if ((umode->flags & DRM_MODE_FLAG_3D_MASK) == DRM_MODE_FLAG_3D_FRAME_PACKING)
+		umode_vdisplay += umode->vtotal;
+	asyh->view.iW = umode->hdisplay;
+	asyh->view.iH = umode_vdisplay;
+	/* For the output mode, we can just use the stock helper. */
+	drm_mode_get_hv_timing(omode, &omode_hdisplay, &omode_vdisplay);
+	asyh->view.oW = omode_hdisplay;
+	asyh->view.oH = omode_vdisplay;
+
+	/* Add overscan compensation if necessary, will keep the aspect
+	 * ratio the same as the backend mode unless overridden by the
+	 * user setting both hborder and vborder properties.
+	 */
+	if ((asyc->scaler.underscan.mode == UNDERSCAN_ON ||
+	    (asyc->scaler.underscan.mode == UNDERSCAN_AUTO &&
+	     drm_detect_hdmi_monitor(edid)))) {
+		u32 bX = asyc->scaler.underscan.hborder;
+		u32 bY = asyc->scaler.underscan.vborder;
+		u32 r = (asyh->view.oH << 19) / asyh->view.oW;
+
+		if (bX) {
+			asyh->view.oW -= (bX * 2);
+			if (bY) asyh->view.oH -= (bY * 2);
+			else    asyh->view.oH  = ((asyh->view.oW * r) + (r / 2)) >> 19;
+		} else {
+			asyh->view.oW -= (asyh->view.oW >> 4) + 32;
+			if (bY) asyh->view.oH -= (bY * 2);
+			else    asyh->view.oH  = ((asyh->view.oW * r) + (r / 2)) >> 19;
+		}
+	}
+
+	/* Handle CENTER/ASPECT scaling, taking into account the areas
+	 * removed already for overscan compensation.
+	 */
+	switch (mode) {
+	case DRM_MODE_SCALE_CENTER:
+		asyh->view.oW = min((u16)umode->hdisplay, asyh->view.oW);
+		asyh->view.oH = min((u16)umode_vdisplay, asyh->view.oH);
+		/* fall-through */
+	case DRM_MODE_SCALE_ASPECT:
+		if (asyh->view.oH < asyh->view.oW) {
+			u32 r = (asyh->view.iW << 19) / asyh->view.iH;
+			asyh->view.oW = ((asyh->view.oH * r) + (r / 2)) >> 19;
+		} else {
+			u32 r = (asyh->view.iH << 19) / asyh->view.iW;
+			asyh->view.oH = ((asyh->view.oW * r) + (r / 2)) >> 19;
+		}
+		break;
+	default:
+		break;
+	}
+
+	asyh->set.view = true;
+}
+
+static int
+nv50_head_atomic_check_lut(struct nv50_head *head,
+			   struct nv50_head_atom *asyh)
+{
+	struct nv50_disp *disp = nv50_disp(head->base.base.dev);
+	struct drm_property_blob *olut = asyh->state.gamma_lut;
+
+	/* Determine whether core output LUT should be enabled. */
+	if (olut) {
+		/* Check if any window(s) have stolen the core output LUT
+		 * to as an input LUT for legacy gamma + I8 colour format.
+		 */
+		if (asyh->wndw.olut) {
+			/* If any window has stolen the core output LUT,
+			 * all of them must.
+			 */
+			if (asyh->wndw.olut != asyh->wndw.mask)
+				return -EINVAL;
+			olut = NULL;
+		}
+	}
+
+	if (!olut) {
+		asyh->olut.handle = 0;
+		return 0;
+	}
+
+	asyh->olut.handle = disp->core->chan.vram.handle;
+	asyh->olut.buffer = !asyh->olut.buffer;
+	head->func->olut(head, asyh);
+	return 0;
+}
+
+static void
+nv50_head_atomic_check_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct drm_display_mode *mode = &asyh->state.adjusted_mode;
+	struct nv50_head_mode *m = &asyh->mode;
+	u32 blankus;
+
+	drm_mode_set_crtcinfo(mode, CRTC_INTERLACE_HALVE_V | CRTC_STEREO_DOUBLE);
+
+	/*
+	 * DRM modes are defined in terms of a repeating interval
+	 * starting with the active display area.  The hardware modes
+	 * are defined in terms of a repeating interval starting one
+	 * unit (pixel or line) into the sync pulse.  So, add bias.
+	 */
+
+	m->h.active = mode->crtc_htotal;
+	m->h.synce  = mode->crtc_hsync_end - mode->crtc_hsync_start - 1;
+	m->h.blanke = mode->crtc_hblank_end - mode->crtc_hsync_start - 1;
+	m->h.blanks = m->h.blanke + mode->crtc_hdisplay;
+
+	m->v.active = mode->crtc_vtotal;
+	m->v.synce  = mode->crtc_vsync_end - mode->crtc_vsync_start - 1;
+	m->v.blanke = mode->crtc_vblank_end - mode->crtc_vsync_start - 1;
+	m->v.blanks = m->v.blanke + mode->crtc_vdisplay;
+
+	/*XXX: Safe underestimate, even "0" works */
+	blankus = (m->v.active - mode->crtc_vdisplay - 2) * m->h.active;
+	blankus *= 1000;
+	blankus /= mode->crtc_clock;
+	m->v.blankus = blankus;
+
+	if (mode->flags & DRM_MODE_FLAG_INTERLACE) {
+		m->v.blank2e =  m->v.active + m->v.blanke;
+		m->v.blank2s =  m->v.blank2e + mode->crtc_vdisplay;
+		m->v.active  = (m->v.active * 2) + 1;
+		m->interlace = true;
+	} else {
+		m->v.blank2e = 0;
+		m->v.blank2s = 1;
+		m->interlace = false;
+	}
+	m->clock = mode->crtc_clock;
+
+	asyh->or.nhsync = !!(mode->flags & DRM_MODE_FLAG_NHSYNC);
+	asyh->or.nvsync = !!(mode->flags & DRM_MODE_FLAG_NVSYNC);
+	asyh->set.or = head->func->or != NULL;
+	asyh->set.mode = true;
+}
+
+static int
+nv50_head_atomic_check(struct drm_crtc *crtc, struct drm_crtc_state *state)
+{
+	struct nouveau_drm *drm = nouveau_drm(crtc->dev);
+	struct nv50_head *head = nv50_head(crtc);
+	struct nv50_head_atom *armh = nv50_head_atom(crtc->state);
+	struct nv50_head_atom *asyh = nv50_head_atom(state);
+	struct nouveau_conn_atom *asyc = NULL;
+	struct drm_connector_state *conns;
+	struct drm_connector *conn;
+	int i;
+
+	NV_ATOMIC(drm, "%s atomic_check %d\n", crtc->name, asyh->state.active);
+	if (asyh->state.active) {
+		for_each_new_connector_in_state(asyh->state.state, conn, conns, i) {
+			if (conns->crtc == crtc) {
+				asyc = nouveau_conn_atom(conns);
+				break;
+			}
+		}
+
+		if (armh->state.active) {
+			if (asyc) {
+				if (asyh->state.mode_changed)
+					asyc->set.scaler = true;
+				if (armh->base.depth != asyh->base.depth)
+					asyc->set.dither = true;
+			}
+		} else {
+			if (asyc)
+				asyc->set.mask = ~0;
+			asyh->set.mask = ~0;
+			asyh->set.or = head->func->or != NULL;
+		}
+
+		if (asyh->state.mode_changed)
+			nv50_head_atomic_check_mode(head, asyh);
+
+		if (asyh->state.color_mgmt_changed ||
+		    memcmp(&armh->wndw, &asyh->wndw, sizeof(asyh->wndw))) {
+			int ret = nv50_head_atomic_check_lut(head, asyh);
+			if (ret)
+				return ret;
+
+			asyh->olut.visible = asyh->olut.handle != 0;
+		}
+
+		if (asyc) {
+			if (asyc->set.scaler)
+				nv50_head_atomic_check_view(armh, asyh, asyc);
+			if (asyc->set.dither)
+				nv50_head_atomic_check_dither(armh, asyh, asyc);
+			if (asyc->set.procamp)
+				nv50_head_atomic_check_procamp(armh, asyh, asyc);
+		}
+
+		if (head->func->core_calc) {
+			head->func->core_calc(head, asyh);
+			if (!asyh->core.visible)
+				asyh->olut.visible = false;
+		}
+
+		asyh->set.base = armh->base.cpp != asyh->base.cpp;
+		asyh->set.ovly = armh->ovly.cpp != asyh->ovly.cpp;
+	} else {
+		asyh->olut.visible = false;
+		asyh->core.visible = false;
+		asyh->curs.visible = false;
+		asyh->base.cpp = 0;
+		asyh->ovly.cpp = 0;
+	}
+
+	if (!drm_atomic_crtc_needs_modeset(&asyh->state)) {
+		if (asyh->core.visible) {
+			if (memcmp(&armh->core, &asyh->core, sizeof(asyh->core)))
+				asyh->set.core = true;
+		} else
+		if (armh->core.visible) {
+			asyh->clr.core = true;
+		}
+
+		if (asyh->curs.visible) {
+			if (memcmp(&armh->curs, &asyh->curs, sizeof(asyh->curs)))
+				asyh->set.curs = true;
+		} else
+		if (armh->curs.visible) {
+			asyh->clr.curs = true;
+		}
+
+		if (asyh->olut.visible) {
+			if (memcmp(&armh->olut, &asyh->olut, sizeof(asyh->olut)))
+				asyh->set.olut = true;
+		} else
+		if (armh->olut.visible) {
+			asyh->clr.olut = true;
+		}
+	} else {
+		asyh->clr.olut = armh->olut.visible;
+		asyh->clr.core = armh->core.visible;
+		asyh->clr.curs = armh->curs.visible;
+		asyh->set.olut = asyh->olut.visible;
+		asyh->set.core = asyh->core.visible;
+		asyh->set.curs = asyh->curs.visible;
+	}
+
+	if (asyh->clr.mask || asyh->set.mask)
+		nv50_atom(asyh->state.state)->lock_core = true;
+	return 0;
+}
+
+static const struct drm_crtc_helper_funcs
+nv50_head_help = {
+	.atomic_check = nv50_head_atomic_check,
+};
+
+static void
+nv50_head_atomic_destroy_state(struct drm_crtc *crtc,
+			       struct drm_crtc_state *state)
+{
+	struct nv50_head_atom *asyh = nv50_head_atom(state);
+	__drm_atomic_helper_crtc_destroy_state(&asyh->state);
+	kfree(asyh);
+}
+
+static struct drm_crtc_state *
+nv50_head_atomic_duplicate_state(struct drm_crtc *crtc)
+{
+	struct nv50_head_atom *armh = nv50_head_atom(crtc->state);
+	struct nv50_head_atom *asyh;
+	if (!(asyh = kmalloc(sizeof(*asyh), GFP_KERNEL)))
+		return NULL;
+	__drm_atomic_helper_crtc_duplicate_state(crtc, &asyh->state);
+	asyh->wndw = armh->wndw;
+	asyh->view = armh->view;
+	asyh->mode = armh->mode;
+	asyh->olut = armh->olut;
+	asyh->core = armh->core;
+	asyh->curs = armh->curs;
+	asyh->base = armh->base;
+	asyh->ovly = armh->ovly;
+	asyh->dither = armh->dither;
+	asyh->procamp = armh->procamp;
+	asyh->clr.mask = 0;
+	asyh->set.mask = 0;
+	return &asyh->state;
+}
+
+static void
+__drm_atomic_helper_crtc_reset(struct drm_crtc *crtc,
+			       struct drm_crtc_state *state)
+{
+	if (crtc->state)
+		crtc->funcs->atomic_destroy_state(crtc, crtc->state);
+	crtc->state = state;
+	crtc->state->crtc = crtc;
+}
+
+static void
+nv50_head_reset(struct drm_crtc *crtc)
+{
+	struct nv50_head_atom *asyh;
+
+	if (WARN_ON(!(asyh = kzalloc(sizeof(*asyh), GFP_KERNEL))))
+		return;
+
+	__drm_atomic_helper_crtc_reset(crtc, &asyh->state);
+}
+
+static void
+nv50_head_destroy(struct drm_crtc *crtc)
+{
+	struct nv50_head *head = nv50_head(crtc);
+	nv50_lut_fini(&head->olut);
+	drm_crtc_cleanup(crtc);
+	kfree(head);
+}
+
+static const struct drm_crtc_funcs
+nv50_head_func = {
+	.reset = nv50_head_reset,
+	.gamma_set = drm_atomic_helper_legacy_gamma_set,
+	.destroy = nv50_head_destroy,
+	.set_config = drm_atomic_helper_set_config,
+	.page_flip = drm_atomic_helper_page_flip,
+	.atomic_duplicate_state = nv50_head_atomic_duplicate_state,
+	.atomic_destroy_state = nv50_head_atomic_destroy_state,
+};
+
+int
+nv50_head_create(struct drm_device *dev, int index)
+{
+	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nv50_disp *disp = nv50_disp(dev);
+	struct nv50_head *head;
+	struct nv50_wndw *curs, *wndw;
+	struct drm_crtc *crtc;
+	int ret;
+
+	head = kzalloc(sizeof(*head), GFP_KERNEL);
+	if (!head)
+		return -ENOMEM;
+
+	head->func = disp->core->func->head;
+	head->base.index = index;
+
+	if (disp->disp->object.oclass < GV100_DISP) {
+		ret = nv50_ovly_new(drm, head->base.index, &wndw);
+		ret = nv50_base_new(drm, head->base.index, &wndw);
+	} else {
+		ret = nv50_wndw_new(drm, DRM_PLANE_TYPE_OVERLAY,
+				    head->base.index * 2 + 1, &wndw);
+		ret = nv50_wndw_new(drm, DRM_PLANE_TYPE_PRIMARY,
+				    head->base.index * 2 + 0, &wndw);
+	}
+	if (ret == 0)
+		ret = nv50_curs_new(drm, head->base.index, &curs);
+	if (ret) {
+		kfree(head);
+		return ret;
+	}
+
+	crtc = &head->base.base;
+	drm_crtc_init_with_planes(dev, crtc, &wndw->plane, &curs->plane,
+				  &nv50_head_func, "head-%d", head->base.index);
+	drm_crtc_helper_add(crtc, &nv50_head_help);
+	drm_mode_crtc_set_gamma_size(crtc, 256);
+
+	if (head->func->olut_set) {
+		ret = nv50_lut_init(disp, &drm->client.mmu, &head->olut);
+		if (ret)
+			goto out;
+	}
+
+out:
+	if (ret)
+		nv50_head_destroy(crtc);
+	return ret;
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head.h b/drivers/gpu/drm/nouveau/dispnv50/head.h
new file mode 100644
index 000000000000..37b3248c6dae
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/head.h
@@ -0,0 +1,78 @@
+#ifndef __NV50_KMS_HEAD_H__
+#define __NV50_KMS_HEAD_H__
+#define nv50_head(c) container_of((c), struct nv50_head, base.base)
+#include "disp.h"
+#include "atom.h"
+#include "lut.h"
+
+#include "nouveau_crtc.h"
+
+struct nv50_head {
+	const struct nv50_head_func *func;
+	struct nouveau_crtc base;
+	struct nv50_lut olut;
+};
+
+int nv50_head_create(struct drm_device *, int index);
+void nv50_head_flush_set(struct nv50_head *, struct nv50_head_atom *);
+void nv50_head_flush_clr(struct nv50_head *, struct nv50_head_atom *, bool y);
+
+struct nv50_head_func {
+	void (*view)(struct nv50_head *, struct nv50_head_atom *);
+	void (*mode)(struct nv50_head *, struct nv50_head_atom *);
+	void (*olut)(struct nv50_head *, struct nv50_head_atom *);
+	void (*olut_set)(struct nv50_head *, struct nv50_head_atom *);
+	void (*olut_clr)(struct nv50_head *);
+	void (*core_calc)(struct nv50_head *, struct nv50_head_atom *);
+	void (*core_set)(struct nv50_head *, struct nv50_head_atom *);
+	void (*core_clr)(struct nv50_head *);
+	int (*curs_layout)(struct nv50_head *, struct nv50_wndw_atom *,
+			   struct nv50_head_atom *);
+	int (*curs_format)(struct nv50_head *, struct nv50_wndw_atom *,
+			   struct nv50_head_atom *);
+	void (*curs_set)(struct nv50_head *, struct nv50_head_atom *);
+	void (*curs_clr)(struct nv50_head *);
+	void (*base)(struct nv50_head *, struct nv50_head_atom *);
+	void (*ovly)(struct nv50_head *, struct nv50_head_atom *);
+	void (*dither)(struct nv50_head *, struct nv50_head_atom *);
+	void (*procamp)(struct nv50_head *, struct nv50_head_atom *);
+	void (*or)(struct nv50_head *, struct nv50_head_atom *);
+};
+
+extern const struct nv50_head_func head507d;
+void head507d_view(struct nv50_head *, struct nv50_head_atom *);
+void head507d_mode(struct nv50_head *, struct nv50_head_atom *);
+void head507d_olut(struct nv50_head *, struct nv50_head_atom *);
+void head507d_core_calc(struct nv50_head *, struct nv50_head_atom *);
+void head507d_core_clr(struct nv50_head *);
+int head507d_curs_layout(struct nv50_head *, struct nv50_wndw_atom *,
+			 struct nv50_head_atom *);
+int head507d_curs_format(struct nv50_head *, struct nv50_wndw_atom *,
+			 struct nv50_head_atom *);
+void head507d_base(struct nv50_head *, struct nv50_head_atom *);
+void head507d_ovly(struct nv50_head *, struct nv50_head_atom *);
+void head507d_dither(struct nv50_head *, struct nv50_head_atom *);
+void head507d_procamp(struct nv50_head *, struct nv50_head_atom *);
+
+extern const struct nv50_head_func head827d;
+
+extern const struct nv50_head_func head907d;
+void head907d_view(struct nv50_head *, struct nv50_head_atom *);
+void head907d_mode(struct nv50_head *, struct nv50_head_atom *);
+void head907d_olut(struct nv50_head *, struct nv50_head_atom *);
+void head907d_olut_set(struct nv50_head *, struct nv50_head_atom *);
+void head907d_olut_clr(struct nv50_head *);
+void head907d_core_set(struct nv50_head *, struct nv50_head_atom *);
+void head907d_core_clr(struct nv50_head *);
+void head907d_curs_set(struct nv50_head *, struct nv50_head_atom *);
+void head907d_curs_clr(struct nv50_head *);
+void head907d_ovly(struct nv50_head *, struct nv50_head_atom *);
+void head907d_procamp(struct nv50_head *, struct nv50_head_atom *);
+void head907d_or(struct nv50_head *, struct nv50_head_atom *);
+
+extern const struct nv50_head_func head917d;
+int head917d_curs_layout(struct nv50_head *, struct nv50_wndw_atom *,
+			 struct nv50_head_atom *);
+
+extern const struct nv50_head_func headc37d;
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head507d.c b/drivers/gpu/drm/nouveau/dispnv50/head507d.c
new file mode 100644
index 000000000000..51bc5996fd37
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/head507d.c
@@ -0,0 +1,325 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "head.h"
+#include "core.h"
+
+void
+head507d_procamp(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x08a8 + (head->base.index * 0x400), 1);
+		evo_data(push, asyh->procamp.sat.sin << 20 |
+			       asyh->procamp.sat.cos << 8);
+		evo_kick(push, core);
+	}
+}
+
+void
+head507d_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x08a0 + (head->base.index * 0x0400), 1);
+		evo_data(push, asyh->dither.mode << 3 |
+			       asyh->dither.bits << 1 |
+			       asyh->dither.enable);
+		evo_kick(push, core);
+	}
+}
+
+void
+head507d_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 bounds = 0;
+	u32 *push;
+
+	if (asyh->ovly.cpp) {
+		switch (asyh->ovly.cpp) {
+		case 4: bounds |= 0x00000300; break;
+		case 2: bounds |= 0x00000100; break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+		bounds |= 0x00000001;
+	} else {
+		bounds |= 0x00000100;
+	}
+
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x0904 + head->base.index * 0x400, 1);
+		evo_data(push, bounds);
+		evo_kick(push, core);
+	}
+}
+
+void
+head507d_base(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 bounds = 0;
+	u32 *push;
+
+	if (asyh->base.cpp) {
+		switch (asyh->base.cpp) {
+		case 8: bounds |= 0x00000500; break;
+		case 4: bounds |= 0x00000300; break;
+		case 2: bounds |= 0x00000100; break;
+		case 1: bounds |= 0x00000000; break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+		bounds |= 0x00000001;
+	}
+
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x0900 + head->base.index * 0x400, 1);
+		evo_data(push, bounds);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_curs_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
+		evo_data(push, 0x05000000);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 3))) {
+		evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
+		evo_data(push, 0x80000000 | asyh->curs.layout << 26 |
+					    asyh->curs.format << 24);
+		evo_data(push, asyh->curs.offset >> 8);
+		evo_kick(push, core);
+	}
+}
+
+int
+head507d_curs_format(struct nv50_head *head, struct nv50_wndw_atom *asyw,
+		     struct nv50_head_atom *asyh)
+{
+	switch (asyw->image.format) {
+	case 0xcf: asyh->curs.format = 1; break;
+	default:
+		WARN_ON(1);
+		return -EINVAL;
+	}
+	return 0;
+}
+
+int
+head507d_curs_layout(struct nv50_head *head, struct nv50_wndw_atom *asyw,
+		     struct nv50_head_atom *asyh)
+{
+	switch (asyw->image.w) {
+	case 32: asyh->curs.layout = 0; break;
+	case 64: asyh->curs.layout = 1; break;
+	default:
+		return -EINVAL;
+	}
+	return 0;
+}
+
+void
+head507d_core_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x0874 + head->base.index * 0x400, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 9))) {
+		evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
+		evo_data(push, asyh->core.offset >> 8);
+		evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
+		evo_data(push, asyh->core.h << 16 | asyh->core.w);
+		evo_data(push, asyh->core.layout << 20 |
+			       (asyh->core.pitch >> 8) << 8 |
+			       asyh->core.blocks << 8 |
+			       asyh->core.blockh);
+		evo_data(push, asyh->core.kind << 16 |
+			       asyh->core.format << 8);
+		evo_data(push, asyh->core.handle);
+		evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
+		evo_data(push, asyh->core.y << 16 | asyh->core.x);
+		evo_kick(push, core);
+
+		/* EVO will complain with INVALID_STATE if we have an
+		 * active cursor and (re)specify HeadSetContextDmaIso
+		 * without also updating HeadSetOffsetCursor.
+		 */
+		asyh->set.curs = asyh->curs.visible;
+		asyh->set.olut = asyh->olut.handle != 0;
+	}
+}
+
+void
+head507d_core_calc(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_disp *disp = nv50_disp(head->base.base.dev);
+	if ((asyh->core.visible = (asyh->base.cpp != 0))) {
+		asyh->core.x = asyh->base.x;
+		asyh->core.y = asyh->base.y;
+		asyh->core.w = asyh->base.w;
+		asyh->core.h = asyh->base.h;
+	} else
+	if ((asyh->core.visible = (asyh->ovly.cpp != 0)) ||
+	    (asyh->core.visible = asyh->curs.visible)) {
+		/*XXX: We need to either find some way of having the
+		 *     primary base layer appear black, while still
+		 *     being able to display the other layers, or we
+		 *     need to allocate a dummy black surface here.
+		 */
+		asyh->core.x = 0;
+		asyh->core.y = 0;
+		asyh->core.w = asyh->state.mode.hdisplay;
+		asyh->core.h = asyh->state.mode.vdisplay;
+	}
+	asyh->core.handle = disp->core->chan.vram.handle;
+	asyh->core.offset = 0;
+	asyh->core.format = 0xcf;
+	asyh->core.kind = 0;
+	asyh->core.layout = 1;
+	asyh->core.blockh = 0;
+	asyh->core.blocks = 0;
+	asyh->core.pitch = ALIGN(asyh->core.w, 64) * 4;
+}
+
+static void
+head507d_olut_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head507d_olut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 3))) {
+		evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
+		evo_data(push, 0x80000000 | asyh->olut.mode << 30);
+		evo_data(push, asyh->olut.offset >> 8);
+		evo_kick(push, core);
+	}
+}
+
+void
+head507d_olut(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	if (asyh->base.cpp == 1)
+		asyh->olut.mode = 0;
+	else
+		asyh->olut.mode = 1;
+}
+
+void
+head507d_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	struct nv50_head_mode *m = &asyh->mode;
+	u32 *push;
+	if ((push = evo_wait(core, 13))) {
+		evo_mthd(push, 0x0804 + (head->base.index * 0x400), 2);
+		evo_data(push, 0x00800000 | m->clock);
+		evo_data(push, m->interlace ? 0x00000002 : 0x00000000);
+		evo_mthd(push, 0x0810 + (head->base.index * 0x400), 7);
+		evo_data(push, 0x00000000);
+		evo_data(push, m->v.active  << 16 | m->h.active );
+		evo_data(push, m->v.synce   << 16 | m->h.synce  );
+		evo_data(push, m->v.blanke  << 16 | m->h.blanke );
+		evo_data(push, m->v.blanks  << 16 | m->h.blanks );
+		evo_data(push, m->v.blank2e << 16 | m->v.blank2s);
+		evo_data(push, asyh->mode.v.blankus);
+		evo_mthd(push, 0x082c + (head->base.index * 0x400), 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+void
+head507d_view(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 7))) {
+		evo_mthd(push, 0x08a4 + (head->base.index * 0x400), 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x08c8 + (head->base.index * 0x400), 1);
+		evo_data(push, asyh->view.iH << 16 | asyh->view.iW);
+		evo_mthd(push, 0x08d8 + (head->base.index * 0x400), 2);
+		evo_data(push, asyh->view.oH << 16 | asyh->view.oW);
+		evo_data(push, asyh->view.oH << 16 | asyh->view.oW);
+		evo_kick(push, core);
+	}
+}
+
+const struct nv50_head_func
+head507d = {
+	.view = head507d_view,
+	.mode = head507d_mode,
+	.olut = head507d_olut,
+	.olut_set = head507d_olut_set,
+	.olut_clr = head507d_olut_clr,
+	.core_calc = head507d_core_calc,
+	.core_set = head507d_core_set,
+	.core_clr = head507d_core_clr,
+	.curs_layout = head507d_curs_layout,
+	.curs_format = head507d_curs_format,
+	.curs_set = head507d_curs_set,
+	.curs_clr = head507d_curs_clr,
+	.base = head507d_base,
+	.ovly = head507d_ovly,
+	.dither = head507d_dither,
+	.procamp = head507d_procamp,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head827d.c b/drivers/gpu/drm/nouveau/dispnv50/head827d.c
new file mode 100644
index 000000000000..af5e7bd5978b
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/head827d.c
@@ -0,0 +1,124 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "head.h"
+#include "core.h"
+
+static void
+head827d_curs_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
+		evo_data(push, 0x05000000);
+		evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head827d_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 5))) {
+		evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
+		evo_data(push, 0x80000000 | asyh->curs.layout << 26 |
+					    asyh->curs.format << 24);
+		evo_data(push, asyh->curs.offset >> 8);
+		evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
+		evo_data(push, asyh->curs.handle);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head827d_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 9))) {
+		evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
+		evo_data(push, asyh->core.offset >> 8);
+		evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
+		evo_data(push, asyh->core.h << 16 | asyh->core.w);
+		evo_data(push, asyh->core.layout << 20 |
+			       (asyh->core.pitch >> 8) << 8 |
+			       asyh->core.blocks << 8 |
+			       asyh->core.blockh);
+		evo_data(push, asyh->core.format << 8);
+		evo_data(push, asyh->core.handle);
+		evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
+		evo_data(push, asyh->core.y << 16 | asyh->core.x);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head827d_olut_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head827d_olut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 5))) {
+		evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
+		evo_data(push, 0x80000000 | asyh->olut.mode << 30);
+		evo_data(push, asyh->olut.offset >> 8);
+		evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
+		evo_data(push, asyh->olut.handle);
+		evo_kick(push, core);
+	}
+}
+
+const struct nv50_head_func
+head827d = {
+	.view = head507d_view,
+	.mode = head507d_mode,
+	.olut = head507d_olut,
+	.olut_set = head827d_olut_set,
+	.olut_clr = head827d_olut_clr,
+	.core_calc = head507d_core_calc,
+	.core_set = head827d_core_set,
+	.core_clr = head507d_core_clr,
+	.curs_layout = head507d_curs_layout,
+	.curs_format = head507d_curs_format,
+	.curs_set = head827d_curs_set,
+	.curs_clr = head827d_curs_clr,
+	.base = head507d_base,
+	.ovly = head507d_ovly,
+	.dither = head507d_dither,
+	.procamp = head507d_procamp,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head907d.c b/drivers/gpu/drm/nouveau/dispnv50/head907d.c
new file mode 100644
index 000000000000..633907163eb1
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/head907d.c
@@ -0,0 +1,284 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "head.h"
+#include "core.h"
+
+void
+head907d_or(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 3))) {
+		evo_mthd(push, 0x0404 + (head->base.index * 0x300), 2);
+		evo_data(push, 0x00000001 | asyh->or.depth  << 6 |
+					    asyh->or.nvsync << 4 |
+					    asyh->or.nhsync << 3);
+		evo_data(push, 0x31ec6000 | head->base.index << 25 |
+					    asyh->mode.interlace);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_procamp(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x0498 + (head->base.index * 0x300), 1);
+		evo_data(push, asyh->procamp.sat.sin << 20 |
+			       asyh->procamp.sat.cos << 8);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head907d_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x0490 + (head->base.index * 0x0300), 1);
+		evo_data(push, asyh->dither.mode << 3 |
+			       asyh->dither.bits << 1 |
+			       asyh->dither.enable);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 bounds = 0;
+	u32 *push;
+
+	if (asyh->ovly.cpp) {
+		switch (asyh->ovly.cpp) {
+		case 8: bounds |= 0x00000500; break;
+		case 4: bounds |= 0x00000300; break;
+		case 2: bounds |= 0x00000100; break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+		bounds |= 0x00000001;
+	} else {
+		bounds |= 0x00000100;
+	}
+
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x04d4 + head->base.index * 0x300, 1);
+		evo_data(push, bounds);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head907d_base(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 bounds = 0;
+	u32 *push;
+
+	if (asyh->base.cpp) {
+		switch (asyh->base.cpp) {
+		case 8: bounds |= 0x00000500; break;
+		case 4: bounds |= 0x00000300; break;
+		case 2: bounds |= 0x00000100; break;
+		case 1: bounds |= 0x00000000; break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+		bounds |= 0x00000001;
+	}
+
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x04d0 + head->base.index * 0x300, 1);
+		evo_data(push, bounds);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_curs_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		evo_mthd(push, 0x0480 + head->base.index * 0x300, 1);
+		evo_data(push, 0x05000000);
+		evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 5))) {
+		evo_mthd(push, 0x0480 + head->base.index * 0x300, 2);
+		evo_data(push, 0x80000000 | asyh->curs.layout << 26 |
+					    asyh->curs.format << 24);
+		evo_data(push, asyh->curs.offset >> 8);
+		evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
+		evo_data(push, asyh->curs.handle);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_core_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x0474 + head->base.index * 0x300, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 9))) {
+		evo_mthd(push, 0x0460 + head->base.index * 0x300, 1);
+		evo_data(push, asyh->core.offset >> 8);
+		evo_mthd(push, 0x0468 + head->base.index * 0x300, 4);
+		evo_data(push, asyh->core.h << 16 | asyh->core.w);
+		evo_data(push, asyh->core.layout << 24 |
+			       (asyh->core.pitch >> 8) << 8 |
+			       asyh->core.blocks << 8 |
+			       asyh->core.blockh);
+		evo_data(push, asyh->core.format << 8);
+		evo_data(push, asyh->core.handle);
+		evo_mthd(push, 0x04b0 + head->base.index * 0x300, 1);
+		evo_data(push, asyh->core.y << 16 | asyh->core.x);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_olut_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		evo_mthd(push, 0x0448 + (head->base.index * 0x300), 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_olut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 5))) {
+		evo_mthd(push, 0x0448 + (head->base.index * 0x300), 2);
+		evo_data(push, 0x80000000 | asyh->olut.mode << 24);
+		evo_data(push, asyh->olut.offset >> 8);
+		evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
+		evo_data(push, asyh->olut.handle);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_olut(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	asyh->olut.mode = 7;
+}
+
+void
+head907d_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	struct nv50_head_mode *m = &asyh->mode;
+	u32 *push;
+	if ((push = evo_wait(core, 14))) {
+		evo_mthd(push, 0x0410 + (head->base.index * 0x300), 6);
+		evo_data(push, 0x00000000);
+		evo_data(push, m->v.active  << 16 | m->h.active );
+		evo_data(push, m->v.synce   << 16 | m->h.synce  );
+		evo_data(push, m->v.blanke  << 16 | m->h.blanke );
+		evo_data(push, m->v.blanks  << 16 | m->h.blanks );
+		evo_data(push, m->v.blank2e << 16 | m->v.blank2s);
+		evo_mthd(push, 0x042c + (head->base.index * 0x300), 2);
+		evo_data(push, 0x00000000); /* ??? */
+		evo_data(push, 0xffffff00);
+		evo_mthd(push, 0x0450 + (head->base.index * 0x300), 3);
+		evo_data(push, m->clock * 1000);
+		evo_data(push, 0x00200000); /* ??? */
+		evo_data(push, m->clock * 1000);
+		evo_kick(push, core);
+	}
+}
+
+void
+head907d_view(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 8))) {
+		evo_mthd(push, 0x0494 + (head->base.index * 0x300), 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x04b8 + (head->base.index * 0x300), 1);
+		evo_data(push, asyh->view.iH << 16 | asyh->view.iW);
+		evo_mthd(push, 0x04c0 + (head->base.index * 0x300), 3);
+		evo_data(push, asyh->view.oH << 16 | asyh->view.oW);
+		evo_data(push, asyh->view.oH << 16 | asyh->view.oW);
+		evo_data(push, asyh->view.oH << 16 | asyh->view.oW);
+		evo_kick(push, core);
+	}
+}
+
+const struct nv50_head_func
+head907d = {
+	.view = head907d_view,
+	.mode = head907d_mode,
+	.olut = head907d_olut,
+	.olut_set = head907d_olut_set,
+	.olut_clr = head907d_olut_clr,
+	.core_calc = head507d_core_calc,
+	.core_set = head907d_core_set,
+	.core_clr = head907d_core_clr,
+	.curs_layout = head507d_curs_layout,
+	.curs_format = head507d_curs_format,
+	.curs_set = head907d_curs_set,
+	.curs_clr = head907d_curs_clr,
+	.base = head907d_base,
+	.ovly = head907d_ovly,
+	.dither = head907d_dither,
+	.procamp = head907d_procamp,
+	.or = head907d_or,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/head917d.c b/drivers/gpu/drm/nouveau/dispnv50/head917d.c
new file mode 100644
index 000000000000..303df8459ca8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/head917d.c
@@ -0,0 +1,100 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "head.h"
+#include "core.h"
+
+static void
+head917d_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x04a0 + (head->base.index * 0x0300), 1);
+		evo_data(push, asyh->dither.mode << 3 |
+			       asyh->dither.bits << 1 |
+			       asyh->dither.enable);
+		evo_kick(push, core);
+	}
+}
+
+static void
+head917d_base(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 bounds = 0;
+	u32 *push;
+
+	if (asyh->base.cpp) {
+		switch (asyh->base.cpp) {
+		case 8: bounds |= 0x00000500; break;
+		case 4: bounds |= 0x00000300; break;
+		case 2: bounds |= 0x00000100; break;
+		case 1: bounds |= 0x00000000; break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+		bounds |= 0x00020001;
+	}
+
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x04d0 + head->base.index * 0x300, 1);
+		evo_data(push, bounds);
+		evo_kick(push, core);
+	}
+}
+
+int
+head917d_curs_layout(struct nv50_head *head, struct nv50_wndw_atom *asyw,
+		     struct nv50_head_atom *asyh)
+{
+	switch (asyw->state.fb->width) {
+	case  32: asyh->curs.layout = 0; break;
+	case  64: asyh->curs.layout = 1; break;
+	case 128: asyh->curs.layout = 2; break;
+	case 256: asyh->curs.layout = 3; break;
+	default:
+		return -EINVAL;
+	}
+	return 0;
+}
+
+const struct nv50_head_func
+head917d = {
+	.view = head907d_view,
+	.mode = head907d_mode,
+	.olut = head907d_olut,
+	.olut_set = head907d_olut_set,
+	.olut_clr = head907d_olut_clr,
+	.core_calc = head507d_core_calc,
+	.core_set = head907d_core_set,
+	.core_clr = head907d_core_clr,
+	.curs_layout = head917d_curs_layout,
+	.curs_format = head507d_curs_format,
+	.curs_set = head907d_curs_set,
+	.curs_clr = head907d_curs_clr,
+	.base = head917d_base,
+	.ovly = head907d_ovly,
+	.dither = head917d_dither,
+	.procamp = head907d_procamp,
+	.or = head907d_or,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/headc37d.c b/drivers/gpu/drm/nouveau/dispnv50/headc37d.c
new file mode 100644
index 000000000000..989c14083066
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/headc37d.c
@@ -0,0 +1,212 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "head.h"
+#include "atom.h"
+#include "core.h"
+
+static void
+headc37d_or(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		/*XXX: This is a dirty hack until OR depth handling is
+		 *     improved later for deep colour etc.
+		 */
+		switch (asyh->or.depth) {
+		case 6: asyh->or.depth = 5; break;
+		case 5: asyh->or.depth = 4; break;
+		case 2: asyh->or.depth = 1; break;
+		case 0:	asyh->or.depth = 4; break;
+		default:
+			WARN_ON(1);
+			break;
+		}
+
+		evo_mthd(push, 0x2004 + (head->base.index * 0x400), 1);
+		evo_data(push, 0x00000001 |
+			       asyh->or.depth << 4 |
+			       asyh->or.nvsync << 3 |
+			       asyh->or.nhsync << 2);
+		evo_kick(push, core);
+	}
+}
+
+static void
+headc37d_procamp(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x2000 + (head->base.index * 0x400), 1);
+		evo_data(push, 0x80000000 |
+			       asyh->procamp.sat.sin << 16 |
+			       asyh->procamp.sat.cos << 4);
+		evo_kick(push, core);
+	}
+}
+
+static void
+headc37d_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x2018 + (head->base.index * 0x0400), 1);
+		evo_data(push, asyh->dither.mode << 8 |
+			       asyh->dither.bits << 4 |
+			       asyh->dither.enable);
+		evo_kick(push, core);
+	}
+}
+
+static void
+headc37d_curs_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		evo_mthd(push, 0x209c + head->base.index * 0x400, 1);
+		evo_data(push, 0x000000cf);
+		evo_mthd(push, 0x2088 + head->base.index * 0x400, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+static void
+headc37d_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 7))) {
+		evo_mthd(push, 0x209c + head->base.index * 0x400, 2);
+		evo_data(push, 0x80000000 |
+			       asyh->curs.layout << 8 |
+			       asyh->curs.format << 0);
+		evo_data(push, 0x000072ff);
+		evo_mthd(push, 0x2088 + head->base.index * 0x400, 1);
+		evo_data(push, asyh->curs.handle);
+		evo_mthd(push, 0x2090 + head->base.index * 0x400, 1);
+		evo_data(push, asyh->curs.offset >> 8);
+		evo_kick(push, core);
+	}
+}
+
+static int
+headc37d_curs_format(struct nv50_head *head, struct nv50_wndw_atom *asyw,
+		     struct nv50_head_atom *asyh)
+{
+	asyh->curs.format = asyw->image.format;
+	return 0;
+}
+
+static void
+headc37d_olut_clr(struct nv50_head *head)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 2))) {
+		evo_mthd(push, 0x20ac + (head->base.index * 0x400), 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, core);
+	}
+}
+
+static void
+headc37d_olut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		evo_mthd(push, 0x20a4 + (head->base.index * 0x400), 3);
+		evo_data(push, asyh->olut.output_mode << 8 |
+			       asyh->olut.range << 4 |
+			       asyh->olut.size);
+		evo_data(push, asyh->olut.offset >> 8);
+		evo_data(push, asyh->olut.handle);
+		evo_kick(push, core);
+	}
+}
+
+static void
+headc37d_olut(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	asyh->olut.mode = 2;
+	asyh->olut.size = 0;
+	asyh->olut.range = 0;
+	asyh->olut.output_mode = 1;
+}
+
+static void
+headc37d_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	struct nv50_head_mode *m = &asyh->mode;
+	u32 *push;
+	if ((push = evo_wait(core, 12))) {
+		evo_mthd(push, 0x2064 + (head->base.index * 0x400), 5);
+		evo_data(push, (m->v.active  << 16) | m->h.active );
+		evo_data(push, (m->v.synce   << 16) | m->h.synce  );
+		evo_data(push, (m->v.blanke  << 16) | m->h.blanke );
+		evo_data(push, (m->v.blanks  << 16) | m->h.blanks );
+		evo_data(push, (m->v.blank2e << 16) | m->v.blank2s);
+		evo_mthd(push, 0x200c + (head->base.index * 0x400), 1);
+		evo_data(push, m->clock * 1000);
+		evo_mthd(push, 0x2028 + (head->base.index * 0x400), 1);
+		evo_data(push, m->clock * 1000);
+		/*XXX: HEAD_USAGE_BOUNDS, doesn't belong here. */
+		evo_mthd(push, 0x2030 + (head->base.index * 0x400), 1);
+		evo_data(push, 0x00000124);
+		evo_kick(push, core);
+	}
+}
+
+static void
+headc37d_view(struct nv50_head *head, struct nv50_head_atom *asyh)
+{
+	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->core->chan;
+	u32 *push;
+	if ((push = evo_wait(core, 4))) {
+		evo_mthd(push, 0x204c + (head->base.index * 0x400), 1);
+		evo_data(push, (asyh->view.iH << 16) | asyh->view.iW);
+		evo_mthd(push, 0x2058 + (head->base.index * 0x400), 1);
+		evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
+		evo_kick(push, core);
+	}
+}
+
+const struct nv50_head_func
+headc37d = {
+	.view = headc37d_view,
+	.mode = headc37d_mode,
+	.olut = headc37d_olut,
+	.olut_set = headc37d_olut_set,
+	.olut_clr = headc37d_olut_clr,
+	.curs_layout = head917d_curs_layout,
+	.curs_format = headc37d_curs_format,
+	.curs_set = headc37d_curs_set,
+	.curs_clr = headc37d_curs_clr,
+	.dither = headc37d_dither,
+	.procamp = headc37d_procamp,
+	.or = headc37d_or,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/lut.c b/drivers/gpu/drm/nouveau/dispnv50/lut.c
new file mode 100644
index 000000000000..a6b96ae2a22f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/lut.c
@@ -0,0 +1,95 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "lut.h"
+#include "disp.h"
+
+#include <drm/drm_color_mgmt.h>
+#include <drm/drm_mode.h>
+#include <drm/drm_property.h>
+
+#include <nvif/class.h>
+
+u32
+nv50_lut_load(struct nv50_lut *lut, bool legacy, int buffer,
+	      struct drm_property_blob *blob)
+{
+	struct drm_color_lut *in = (struct drm_color_lut *)blob->data;
+	void __iomem *mem = lut->mem[buffer].object.map.ptr;
+	const int size = blob->length / sizeof(*in);
+	int bits, shift, i;
+	u16 zero, r, g, b;
+	u32 addr = lut->mem[buffer].addr;
+
+	/* This can't happen.. But it shuts the compiler up. */
+	if (WARN_ON(size != 256))
+		return 0;
+
+	if (legacy) {
+		bits = 11;
+		shift = 3;
+		zero = 0x0000;
+	} else {
+		bits = 14;
+		shift = 0;
+		zero = 0x6000;
+	}
+
+	for (i = 0; i < size; i++) {
+		r = (drm_color_lut_extract(in[i].  red, bits) + zero) << shift;
+		g = (drm_color_lut_extract(in[i].green, bits) + zero) << shift;
+		b = (drm_color_lut_extract(in[i]. blue, bits) + zero) << shift;
+		writew(r, mem + (i * 0x08) + 0);
+		writew(g, mem + (i * 0x08) + 2);
+		writew(b, mem + (i * 0x08) + 4);
+	}
+
+	/* INTERPOLATE modes require a "next" entry to interpolate with,
+	 * so we replicate the last entry to deal with this for now.
+	 */
+	writew(r, mem + (i * 0x08) + 0);
+	writew(g, mem + (i * 0x08) + 2);
+	writew(b, mem + (i * 0x08) + 4);
+	return addr;
+}
+
+void
+nv50_lut_fini(struct nv50_lut *lut)
+{
+	int i;
+	for (i = 0; i < ARRAY_SIZE(lut->mem); i++)
+		nvif_mem_fini(&lut->mem[i]);
+}
+
+int
+nv50_lut_init(struct nv50_disp *disp, struct nvif_mmu *mmu,
+	      struct nv50_lut *lut)
+{
+	const u32 size = disp->disp->object.oclass < GF110_DISP ? 257 : 1025;
+	int i;
+	for (i = 0; i < ARRAY_SIZE(lut->mem); i++) {
+		int ret = nvif_mem_init_map(mmu, NVIF_MEM_VRAM, size * 8,
+					    &lut->mem[i]);
+		if (ret)
+			return ret;
+	}
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/lut.h b/drivers/gpu/drm/nouveau/dispnv50/lut.h
new file mode 100644
index 000000000000..6d7b8352e4cb
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/lut.h
@@ -0,0 +1,15 @@
+#ifndef __NV50_KMS_LUT_H__
+#define __NV50_KMS_LUT_H__
+#include <nvif/mem.h>
+struct drm_property_blob;
+struct nv50_disp;
+
+struct nv50_lut {
+	struct nvif_mem mem[2];
+};
+
+int nv50_lut_init(struct nv50_disp *, struct nvif_mmu *, struct nv50_lut *);
+void nv50_lut_fini(struct nv50_lut *);
+u32 nv50_lut_load(struct nv50_lut *, bool legacy, int buffer,
+		  struct drm_property_blob *);
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/oimm.c b/drivers/gpu/drm/nouveau/dispnv50/oimm.c
new file mode 100644
index 000000000000..2a2841d344c8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/oimm.c
@@ -0,0 +1,51 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "oimm.h"
+
+#include <nvif/class.h>
+
+int
+nv50_oimm_init(struct nouveau_drm *drm, struct nv50_wndw *wndw)
+{
+	static const struct {
+		s32 oclass;
+		int version;
+		int (*init)(struct nouveau_drm *, s32, struct nv50_wndw *);
+	} oimms[] = {
+		{ GK104_DISP_OVERLAY, 0, oimm507b_init },
+		{ GF110_DISP_OVERLAY, 0, oimm507b_init },
+		{ GT214_DISP_OVERLAY, 0, oimm507b_init },
+		{   G82_DISP_OVERLAY, 0, oimm507b_init },
+		{  NV50_DISP_OVERLAY, 0, oimm507b_init },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid;
+
+	cid = nvif_mclass(&disp->disp->object, oimms);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported overlay immediate class\n");
+		return cid;
+	}
+
+	return oimms[cid].init(drm, oimms[cid].oclass, wndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/oimm.h b/drivers/gpu/drm/nouveau/dispnv50/oimm.h
new file mode 100644
index 000000000000..6fa51f101e94
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/oimm.h
@@ -0,0 +1,8 @@
+#ifndef __NV50_KMS_OIMM_H__
+#define __NV50_KMS_OIMM_H__
+#include "wndw.h"
+
+int oimm507b_init(struct nouveau_drm *, s32, struct nv50_wndw *);
+
+int nv50_oimm_init(struct nouveau_drm *, struct nv50_wndw *);
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/oimm507b.c b/drivers/gpu/drm/nouveau/dispnv50/oimm507b.c
new file mode 100644
index 000000000000..2ee404b3e19f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/oimm507b.c
@@ -0,0 +1,52 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "oimm.h"
+
+#include <nvif/cl507b.h>
+
+static int
+oimm507b_init_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
+	       s32 oclass, struct nv50_wndw *wndw)
+{
+	struct nv50_disp_overlay_v0 args = {
+		.head = wndw->id,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int ret;
+
+	ret = nvif_object_init(&disp->disp->object, 0, oclass, &args,
+			       sizeof(args), &wndw->wimm.base.user);
+	if (ret) {
+		NV_ERROR(drm, "oimm%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	nvif_object_map(&wndw->wimm.base.user, NULL, 0);
+	wndw->immd = func;
+	return 0;
+}
+
+int
+oimm507b_init(struct nouveau_drm *drm, s32 oclass, struct nv50_wndw *wndw)
+{
+	return oimm507b_init_(&curs507a, drm, oclass, wndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly.c b/drivers/gpu/drm/nouveau/dispnv50/ovly.c
new file mode 100644
index 000000000000..90c246d47604
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly.c
@@ -0,0 +1,57 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ovly.h"
+#include "oimm.h"
+
+#include <nvif/class.h>
+
+int
+nv50_ovly_new(struct nouveau_drm *drm, int head, struct nv50_wndw **pwndw)
+{
+	static const struct {
+		s32 oclass;
+		int version;
+		int (*new)(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+	} ovlys[] = {
+		{ GK104_DISP_OVERLAY_CONTROL_DMA, 0, ovly917e_new },
+		{ GF110_DISP_OVERLAY_CONTROL_DMA, 0, ovly907e_new },
+		{ GT214_DISP_OVERLAY_CHANNEL_DMA, 0, ovly827e_new },
+		{ GT200_DISP_OVERLAY_CHANNEL_DMA, 0, ovly827e_new },
+		{   G82_DISP_OVERLAY_CHANNEL_DMA, 0, ovly827e_new },
+		{  NV50_DISP_OVERLAY_CHANNEL_DMA, 0, ovly507e_new },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid, ret;
+
+	cid = nvif_mclass(&disp->disp->object, ovlys);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported overlay class\n");
+		return cid;
+	}
+
+	ret = ovlys[cid].new(drm, head, ovlys[cid].oclass, pwndw);
+	if (ret)
+		return ret;
+
+	return nv50_oimm_init(drm, *pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly.h b/drivers/gpu/drm/nouveau/dispnv50/ovly.h
new file mode 100644
index 000000000000..4869d52d1786
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly.h
@@ -0,0 +1,30 @@
+#ifndef __NV50_KMS_OVLY_H__
+#define __NV50_KMS_OVLY_H__
+#include "wndw.h"
+
+int ovly507e_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+int ovly507e_new_(const struct nv50_wndw_func *, const u32 *format,
+		  struct nouveau_drm *, int head, s32 oclass,
+		  u32 interlock_data, struct nv50_wndw **);
+int ovly507e_acquire(struct nv50_wndw *, struct nv50_wndw_atom *,
+		     struct nv50_head_atom *);
+void ovly507e_release(struct nv50_wndw *, struct nv50_wndw_atom *,
+		      struct nv50_head_atom *);
+void ovly507e_ntfy_set(struct nv50_wndw *, struct nv50_wndw_atom *);
+void ovly507e_ntfy_clr(struct nv50_wndw *);
+void ovly507e_image_clr(struct nv50_wndw *);
+void ovly507e_scale_set(struct nv50_wndw *, struct nv50_wndw_atom *);
+void ovly507e_update(struct nv50_wndw *, u32 *);
+
+extern const u32 ovly827e_format[];
+void ovly827e_ntfy_reset(struct nouveau_bo *, u32);
+int ovly827e_ntfy_wait_begun(struct nouveau_bo *, u32, struct nvif_device *);
+
+extern const struct nv50_wndw_func ovly907e;
+
+int ovly827e_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+int ovly907e_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+int ovly917e_new(struct nouveau_drm *, int, s32, struct nv50_wndw **);
+
+int nv50_ovly_new(struct nouveau_drm *, int head, struct nv50_wndw **);
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c b/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c
new file mode 100644
index 000000000000..cc417664f823
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly507e.c
@@ -0,0 +1,217 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ovly.h"
+#include "atom.h"
+
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_plane_helper.h>
+
+#include <nvif/cl507e.h>
+#include <nvif/event.h>
+
+void
+ovly507e_update(struct nv50_wndw *wndw, u32 *interlock)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x0080, 1);
+		evo_data(push, interlock[NV50_DISP_INTERLOCK_CORE]);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+ovly507e_scale_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 4))) {
+		evo_mthd(push, 0x00e0, 3);
+		evo_data(push, asyw->scale.sy << 16 | asyw->scale.sx);
+		evo_data(push, asyw->scale.sh << 16 | asyw->scale.sw);
+		evo_data(push, asyw->scale.dw);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+ovly507e_image_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 4))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+ovly507e_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 12))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, asyw->image.interval << 4);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, asyw->image.handle[0]);
+		evo_mthd(push, 0x0100, 1);
+		evo_data(push, 0x00000002);
+		evo_mthd(push, 0x0800, 1);
+		evo_data(push, asyw->image.offset[0] >> 8);
+		evo_mthd(push, 0x0808, 3);
+		evo_data(push, asyw->image.h << 16 | asyw->image.w);
+		evo_data(push, asyw->image.layout << 20 |
+			       (asyw->image.pitch[0] >> 8) << 8 |
+			       asyw->image.blocks[0] << 8 |
+			       asyw->image.blockh);
+		evo_data(push, asyw->image.kind << 16 |
+			       asyw->image.format << 8 |
+			       asyw->image.colorspace);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+ovly507e_ntfy_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x00a4, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+ovly507e_ntfy_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 3))) {
+		evo_mthd(push, 0x00a0, 2);
+		evo_data(push, asyw->ntfy.awaken << 30 | asyw->ntfy.offset);
+		evo_data(push, asyw->ntfy.handle);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+void
+ovly507e_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+	asyh->ovly.cpp = 0;
+}
+
+int
+ovly507e_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+	const struct drm_framebuffer *fb = asyw->state.fb;
+	int ret;
+
+	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  true, true);
+	if (ret)
+		return ret;
+
+	asyh->ovly.cpp = fb->format->cpp[0];
+	return 0;
+}
+
+#include "nouveau_bo.h"
+
+static const struct nv50_wndw_func
+ovly507e = {
+	.acquire = ovly507e_acquire,
+	.release = ovly507e_release,
+	.ntfy_set = ovly507e_ntfy_set,
+	.ntfy_clr = ovly507e_ntfy_clr,
+	.ntfy_reset = base507c_ntfy_reset,
+	.ntfy_wait_begun = base507c_ntfy_wait_begun,
+	.image_set = ovly507e_image_set,
+	.image_clr = ovly507e_image_clr,
+	.scale_set = ovly507e_scale_set,
+	.update = ovly507e_update,
+};
+
+static const u32
+ovly507e_format[] = {
+	DRM_FORMAT_YUYV,
+	DRM_FORMAT_UYVY,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_ARGB1555,
+	0
+};
+
+int
+ovly507e_new_(const struct nv50_wndw_func *func, const u32 *format,
+	      struct nouveau_drm *drm, int head, s32 oclass, u32 interlock_data,
+	      struct nv50_wndw **pwndw)
+{
+	struct nv50_disp_overlay_channel_dma_v0 args = {
+		.head = head,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_wndw *wndw;
+	int ret;
+
+	ret = nv50_wndw_new_(func, drm->dev, DRM_PLANE_TYPE_OVERLAY,
+			     "ovly", head, format, BIT(head),
+			     NV50_DISP_INTERLOCK_OVLY, interlock_data,
+			     &wndw);
+	if (*pwndw = wndw, ret)
+		return ret;
+
+	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
+			       &oclass, 0, &args, sizeof(args),
+			       disp->sync->bo.offset, &wndw->wndw);
+	if (ret) {
+		NV_ERROR(drm, "ovly%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	ret = nvif_notify_init(&wndw->wndw.base.user, wndw->notify.func, false,
+			       NV50_DISP_OVERLAY_CHANNEL_DMA_V0_NTFY_UEVENT,
+			       &(struct nvif_notify_uevent_req) {},
+			       sizeof(struct nvif_notify_uevent_req),
+			       sizeof(struct nvif_notify_uevent_rep),
+			       &wndw->notify);
+	if (ret)
+		return ret;
+
+	wndw->ntfy = NV50_DISP_OVLY_NTFY(wndw->id);
+	wndw->sema = NV50_DISP_OVLY_SEM0(wndw->id);
+	wndw->data = 0x00000000;
+	return 0;
+}
+
+int
+ovly507e_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return ovly507e_new_(&ovly507e, ovly507e_format, drm, head, oclass,
+			     0x00000004 << (head * 8), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly827e.c b/drivers/gpu/drm/nouveau/dispnv50/ovly827e.c
new file mode 100644
index 000000000000..aaa9fe5a4fc8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly827e.c
@@ -0,0 +1,107 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ovly.h"
+#include "atom.h"
+
+#include <nouveau_bo.h>
+
+static void
+ovly827e_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 12))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, asyw->image.interval << 4);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, asyw->image.handle[0]);
+		evo_mthd(push, 0x0100, 1);
+		evo_data(push, 0x00000002);
+		evo_mthd(push, 0x0800, 1);
+		evo_data(push, asyw->image.offset[0] >> 8);
+		evo_mthd(push, 0x0808, 3);
+		evo_data(push, asyw->image.h << 16 | asyw->image.w);
+		evo_data(push, asyw->image.layout << 20 |
+			       (asyw->image.pitch[0] >> 8) << 8 |
+			       asyw->image.blocks[0] << 8 |
+			       asyw->image.blockh);
+		evo_data(push, asyw->image.format << 8 |
+			       asyw->image.colorspace);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+int
+ovly827e_ntfy_wait_begun(struct nouveau_bo *bo, u32 offset,
+			 struct nvif_device *device)
+{
+	s64 time = nvif_msec(device, 2000ULL,
+		u32 data = nouveau_bo_rd32(bo, offset / 4 + 3);
+		if ((data & 0xffff0000) == 0xffff0000)
+			break;
+		usleep_range(1, 2);
+	);
+	return time < 0 ? time : 0;
+}
+
+void
+ovly827e_ntfy_reset(struct nouveau_bo *bo, u32 offset)
+{
+	nouveau_bo_wr32(bo, offset / 4 + 0, 0x00000000);
+	nouveau_bo_wr32(bo, offset / 4 + 1, 0x00000000);
+	nouveau_bo_wr32(bo, offset / 4 + 2, 0x00000000);
+	nouveau_bo_wr32(bo, offset / 4 + 3, 0x80000000);
+}
+
+static const struct nv50_wndw_func
+ovly827e = {
+	.acquire = ovly507e_acquire,
+	.release = ovly507e_release,
+	.ntfy_set = ovly507e_ntfy_set,
+	.ntfy_clr = ovly507e_ntfy_clr,
+	.ntfy_reset = ovly827e_ntfy_reset,
+	.ntfy_wait_begun = ovly827e_ntfy_wait_begun,
+	.image_set = ovly827e_image_set,
+	.image_clr = ovly507e_image_clr,
+	.scale_set = ovly507e_scale_set,
+	.update = ovly507e_update,
+};
+
+const u32
+ovly827e_format[] = {
+	DRM_FORMAT_YUYV,
+	DRM_FORMAT_UYVY,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_ARGB1555,
+	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ABGR2101010,
+	0
+};
+
+int
+ovly827e_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return ovly507e_new_(&ovly827e, ovly827e_format, drm, head, oclass,
+			     0x00000004 << (head * 8), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly907e.c b/drivers/gpu/drm/nouveau/dispnv50/ovly907e.c
new file mode 100644
index 000000000000..a3ce53046015
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly907e.c
@@ -0,0 +1,70 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ovly.h"
+#include "atom.h"
+
+static void
+ovly907e_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 12))) {
+		evo_mthd(push, 0x0084, 1);
+		evo_data(push, asyw->image.interval << 4);
+		evo_mthd(push, 0x00c0, 1);
+		evo_data(push, asyw->image.handle[0]);
+		evo_mthd(push, 0x0100, 1);
+		evo_data(push, 0x00000002);
+		evo_mthd(push, 0x0400, 1);
+		evo_data(push, asyw->image.offset[0] >> 8);
+		evo_mthd(push, 0x0408, 3);
+		evo_data(push, asyw->image.h << 16 | asyw->image.w);
+		evo_data(push, asyw->image.layout << 24 |
+			       (asyw->image.pitch[0] >> 8) << 8 |
+			       asyw->image.blocks[0] << 8 |
+			       asyw->image.blockh);
+		evo_data(push, asyw->image.format << 8 |
+			       asyw->image.colorspace);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+const struct nv50_wndw_func
+ovly907e = {
+	.acquire = ovly507e_acquire,
+	.release = ovly507e_release,
+	.ntfy_set = ovly507e_ntfy_set,
+	.ntfy_clr = ovly507e_ntfy_clr,
+	.ntfy_reset = ovly827e_ntfy_reset,
+	.ntfy_wait_begun = ovly827e_ntfy_wait_begun,
+	.image_set = ovly907e_image_set,
+	.image_clr = ovly507e_image_clr,
+	.scale_set = ovly507e_scale_set,
+	.update = ovly507e_update,
+};
+
+int
+ovly907e_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return ovly507e_new_(&ovly907e, ovly827e_format, drm, head, oclass,
+			     0x00000004 << (head * 4), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/ovly917e.c b/drivers/gpu/drm/nouveau/dispnv50/ovly917e.c
new file mode 100644
index 000000000000..505fa7e78523
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/ovly917e.c
@@ -0,0 +1,45 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ovly.h"
+
+static const u32
+ovly917e_format[] = {
+	DRM_FORMAT_YUYV,
+	DRM_FORMAT_UYVY,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_ARGB1555,
+	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ABGR2101010,
+	DRM_FORMAT_XRGB2101010,
+	DRM_FORMAT_ARGB2101010,
+	0
+};
+
+int
+ovly917e_new(struct nouveau_drm *drm, int head, s32 oclass,
+	     struct nv50_wndw **pwndw)
+{
+	return ovly507e_new_(&ovly907e, ovly917e_format, drm, head, oclass,
+			     0x00000004 << (head * 4), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/pior507d.c b/drivers/gpu/drm/nouveau/dispnv50/pior507d.c
new file mode 100644
index 000000000000..d2bac6a341dc
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/pior507d.c
@@ -0,0 +1,44 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+
+static void
+pior507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	      struct nv50_head_atom *asyh)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 2))) {
+		if (asyh) {
+			ctrl |= asyh->or.depth  << 16;
+			ctrl |= asyh->or.nvsync << 13;
+			ctrl |= asyh->or.nhsync << 12;
+		}
+		evo_mthd(push, 0x0700 + (or * 0x040), 1);
+		evo_data(push, ctrl);
+		evo_kick(push, &core->chan);
+	}
+}
+
+const struct nv50_outp_func
+pior507d = {
+	.ctrl = pior507d_ctrl,
+};
diff --git a/drivers/gpu/drm/nouveau/dispnv50/sor507d.c b/drivers/gpu/drm/nouveau/dispnv50/sor507d.c
new file mode 100644
index 000000000000..5222fe6a9b21
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/sor507d.c
@@ -0,0 +1,44 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "core.h"
+
+static void
+sor507d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	     struct nv50_head_atom *asyh)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 2))) {
+		if (asyh) {
+			ctrl |= asyh->or.depth  << 16;
+			ctrl |= asyh->or.nvsync << 13;
+			ctrl |= asyh->or.nhsync << 12;
+		}
+		evo_mthd(push, 0x0600 + (or * 0x40), 1);
+		evo_data(push, ctrl);
+		evo_kick(push, &core->chan);
+	}
+}
+
+const struct nv50_outp_func
+sor507d = {
+	.ctrl = sor507d_ctrl,
+};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegt215.c b/drivers/gpu/drm/nouveau/dispnv50/sor907d.c
index 08e2b1fa3806..b0314ec11fb3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegt215.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/sor907d.c
@@ -1,5 +1,5 @@
 /*
- * Copyright 2015 Red Hat Inc.
+ * Copyright 2018 Red Hat Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -18,21 +18,24 @@
  * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
+#include "core.h"
 
 #include <nvif/class.h>
 
-const struct nv50_disp_dmac_oclass
-gt215_disp_base_oclass = {
-	.base.oclass = GT214_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &g84_disp_base_chan_mthd,
-	.chid = 1,
+static void
+sor907d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	     struct nv50_head_atom *asyh)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 2))) {
+		evo_mthd(push, 0x0200 + (or * 0x20), 1);
+		evo_data(push, ctrl);
+		evo_kick(push, &core->chan);
+	}
+}
+
+const struct nv50_outp_func
+sor907d = {
+	.ctrl = sor907d_ctrl,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgk104.c b/drivers/gpu/drm/nouveau/dispnv50/sorc37d.c
index 2a99db4bf8f8..dff059241c5d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgk104.c
+++ b/drivers/gpu/drm/nouveau/dispnv50/sorc37d.c
@@ -1,5 +1,5 @@
 /*
- * Copyright 2012 Red Hat Inc.
+ * Copyright 2018 Red Hat Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -18,20 +18,22 @@
  * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs
  */
-#include "channv50.h"
-#include "rootnv50.h"
+#include "core.h"
 
-#include <nvif/class.h>
+static void
+sorc37d_ctrl(struct nv50_core *core, int or, u32 ctrl,
+	     struct nv50_head_atom *asyh)
+{
+	u32 *push;
+	if ((push = evo_wait(&core->chan, 2))) {
+		evo_mthd(push, 0x0300 + (or * 0x20), 1);
+		evo_data(push, ctrl);
+		evo_kick(push, &core->chan);
+	}
+}
 
-const struct nv50_disp_pioc_oclass
-gk104_disp_curs_oclass = {
-	.base.oclass = GK104_DISP_CURSOR,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_curs_new,
-	.func = &gf119_disp_pioc_func,
-	.chid = { 13, 13 },
+const struct nv50_outp_func
+sorc37d = {
+	.ctrl = sorc37d_ctrl,
 };
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wimm.c b/drivers/gpu/drm/nouveau/dispnv50/wimm.c
new file mode 100644
index 000000000000..fc36e0696407
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/wimm.c
@@ -0,0 +1,47 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "wimm.h"
+
+#include <nvif/class.h>
+
+int
+nv50_wimm_init(struct nouveau_drm *drm, struct nv50_wndw *wndw)
+{
+	struct {
+		s32 oclass;
+		int version;
+		int (*init)(struct nouveau_drm *, s32, struct nv50_wndw *);
+	} wimms[] = {
+		{ GV100_DISP_WINDOW_IMM_CHANNEL_DMA, 0, wimmc37b_init },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid;
+
+	cid = nvif_mclass(&disp->disp->object, wimms);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported window immediate class\n");
+		return cid;
+	}
+
+	return wimms[cid].init(drm, wimms[cid].oclass, wndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wimm.h b/drivers/gpu/drm/nouveau/dispnv50/wimm.h
new file mode 100644
index 000000000000..363052309be9
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/wimm.h
@@ -0,0 +1,8 @@
+#ifndef __NV50_KMS_WIMM_H__
+#define __NV50_KMS_WIMM_H__
+#include "wndw.h"
+
+int nv50_wimm_init(struct nouveau_drm *drm, struct nv50_wndw *);
+
+int wimmc37b_init(struct nouveau_drm *, s32, struct nv50_wndw *);
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wimmc37b.c b/drivers/gpu/drm/nouveau/dispnv50/wimmc37b.c
new file mode 100644
index 000000000000..9103b8494279
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/wimmc37b.c
@@ -0,0 +1,86 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "wimm.h"
+#include "atom.h"
+#include "wndw.h"
+
+#include <nvif/clc37b.h>
+
+static void
+wimmc37b_update(struct nv50_wndw *wndw, u32 *interlock)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wimm, 2))) {
+		evo_mthd(push, 0x0200, 1);
+		if (interlock[NV50_DISP_INTERLOCK_WNDW] & wndw->interlock.data)
+			evo_data(push, 0x00000003);
+		else
+			evo_data(push, 0x00000001);
+		evo_kick(push, &wndw->wimm);
+	}
+}
+
+static void
+wimmc37b_point(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wimm, 2))) {
+		evo_mthd(push, 0x0208, 1);
+		evo_data(push, asyw->point.y << 16 | asyw->point.x);
+		evo_kick(push, &wndw->wimm);
+	}
+}
+
+static const struct nv50_wimm_func
+wimmc37b = {
+	.point = wimmc37b_point,
+	.update = wimmc37b_update,
+};
+
+static int
+wimmc37b_init_(const struct nv50_wimm_func *func, struct nouveau_drm *drm,
+	       s32 oclass, struct nv50_wndw *wndw)
+{
+	struct nvc37b_window_imm_channel_dma_v0 args = {
+		.pushbuf = 0xb0007b00 | wndw->id,
+		.index = wndw->id,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int ret;
+
+	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
+			       &oclass, 0, &args, sizeof(args), 0,
+			       &wndw->wimm);
+	if (ret) {
+		NV_ERROR(drm, "wimm%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	wndw->immd = func;
+	return 0;
+}
+
+int
+wimmc37b_init(struct nouveau_drm *drm, s32 oclass, struct nv50_wndw *wndw)
+{
+	return wimmc37b_init_(&wimmc37b, drm, oclass, wndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.c b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
new file mode 100644
index 000000000000..2187922e8dc2
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.c
@@ -0,0 +1,643 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "wndw.h"
+#include "wimm.h"
+
+#include <nvif/class.h>
+#include <nvif/cl0002.h>
+
+#include <drm/drm_atomic_helper.h>
+#include "nouveau_bo.h"
+
+static void
+nv50_wndw_ctxdma_del(struct nv50_wndw_ctxdma *ctxdma)
+{
+	nvif_object_fini(&ctxdma->object);
+	list_del(&ctxdma->head);
+	kfree(ctxdma);
+}
+
+static struct nv50_wndw_ctxdma *
+nv50_wndw_ctxdma_new(struct nv50_wndw *wndw, struct nouveau_framebuffer *fb)
+{
+	struct nouveau_drm *drm = nouveau_drm(fb->base.dev);
+	struct nv50_wndw_ctxdma *ctxdma;
+	const u8    kind = fb->nvbo->kind;
+	const u32 handle = 0xfb000000 | kind;
+	struct {
+		struct nv_dma_v0 base;
+		union {
+			struct nv50_dma_v0 nv50;
+			struct gf100_dma_v0 gf100;
+			struct gf119_dma_v0 gf119;
+		};
+	} args = {};
+	u32 argc = sizeof(args.base);
+	int ret;
+
+	list_for_each_entry(ctxdma, &wndw->ctxdma.list, head) {
+		if (ctxdma->object.handle == handle)
+			return ctxdma;
+	}
+
+	if (!(ctxdma = kzalloc(sizeof(*ctxdma), GFP_KERNEL)))
+		return ERR_PTR(-ENOMEM);
+	list_add(&ctxdma->head, &wndw->ctxdma.list);
+
+	args.base.target = NV_DMA_V0_TARGET_VRAM;
+	args.base.access = NV_DMA_V0_ACCESS_RDWR;
+	args.base.start  = 0;
+	args.base.limit  = drm->client.device.info.ram_user - 1;
+
+	if (drm->client.device.info.chipset < 0x80) {
+		args.nv50.part = NV50_DMA_V0_PART_256;
+		argc += sizeof(args.nv50);
+	} else
+	if (drm->client.device.info.chipset < 0xc0) {
+		args.nv50.part = NV50_DMA_V0_PART_256;
+		args.nv50.kind = kind;
+		argc += sizeof(args.nv50);
+	} else
+	if (drm->client.device.info.chipset < 0xd0) {
+		args.gf100.kind = kind;
+		argc += sizeof(args.gf100);
+	} else {
+		args.gf119.page = GF119_DMA_V0_PAGE_LP;
+		args.gf119.kind = kind;
+		argc += sizeof(args.gf119);
+	}
+
+	ret = nvif_object_init(wndw->ctxdma.parent, handle, NV_DMA_IN_MEMORY,
+			       &args, argc, &ctxdma->object);
+	if (ret) {
+		nv50_wndw_ctxdma_del(ctxdma);
+		return ERR_PTR(ret);
+	}
+
+	return ctxdma;
+}
+
+int
+nv50_wndw_wait_armed(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	struct nv50_disp *disp = nv50_disp(wndw->plane.dev);
+	if (asyw->set.ntfy) {
+		return wndw->func->ntfy_wait_begun(disp->sync,
+						   asyw->ntfy.offset,
+						   wndw->wndw.base.device);
+	}
+	return 0;
+}
+
+void
+nv50_wndw_flush_clr(struct nv50_wndw *wndw, u32 *interlock, bool flush,
+		    struct nv50_wndw_atom *asyw)
+{
+	union nv50_wndw_atom_mask clr = {
+		.mask = asyw->clr.mask & ~(flush ? 0 : asyw->set.mask),
+	};
+	if (clr.sema ) wndw->func-> sema_clr(wndw);
+	if (clr.ntfy ) wndw->func-> ntfy_clr(wndw);
+	if (clr.xlut ) wndw->func-> xlut_clr(wndw);
+	if (clr.image) wndw->func->image_clr(wndw);
+
+	interlock[wndw->interlock.type] |= wndw->interlock.data;
+}
+
+void
+nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 *interlock,
+		    struct nv50_wndw_atom *asyw)
+{
+	if (interlock) {
+		asyw->image.mode = 0;
+		asyw->image.interval = 1;
+	}
+
+	if (asyw->set.sema ) wndw->func->sema_set (wndw, asyw);
+	if (asyw->set.ntfy ) wndw->func->ntfy_set (wndw, asyw);
+	if (asyw->set.image) wndw->func->image_set(wndw, asyw);
+
+	if (asyw->set.xlut ) {
+		if (asyw->ilut) {
+			asyw->xlut.i.offset =
+				nv50_lut_load(&wndw->ilut,
+					      asyw->xlut.i.mode <= 1,
+					      asyw->xlut.i.buffer,
+					      asyw->ilut);
+		}
+		wndw->func->xlut_set(wndw, asyw);
+	}
+
+	if (asyw->set.scale) wndw->func->scale_set(wndw, asyw);
+	if (asyw->set.point) {
+		if (asyw->set.point = false, asyw->set.mask)
+			interlock[wndw->interlock.type] |= wndw->interlock.data;
+		interlock[NV50_DISP_INTERLOCK_WIMM] |= wndw->interlock.data;
+
+		wndw->immd->point(wndw, asyw);
+		wndw->immd->update(wndw, interlock);
+	} else {
+		interlock[wndw->interlock.type] |= wndw->interlock.data;
+	}
+}
+
+void
+nv50_wndw_ntfy_enable(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	struct nv50_disp *disp = nv50_disp(wndw->plane.dev);
+
+	asyw->ntfy.handle = wndw->wndw.sync.handle;
+	asyw->ntfy.offset = wndw->ntfy;
+	asyw->ntfy.awaken = false;
+	asyw->set.ntfy = true;
+
+	wndw->func->ntfy_reset(disp->sync, wndw->ntfy);
+	wndw->ntfy ^= 0x10;
+}
+
+static void
+nv50_wndw_atomic_check_release(struct nv50_wndw *wndw,
+			       struct nv50_wndw_atom *asyw,
+			       struct nv50_head_atom *asyh)
+{
+	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
+	NV_ATOMIC(drm, "%s release\n", wndw->plane.name);
+	wndw->func->release(wndw, asyw, asyh);
+	asyw->ntfy.handle = 0;
+	asyw->sema.handle = 0;
+}
+
+static int
+nv50_wndw_atomic_check_acquire_yuv(struct nv50_wndw_atom *asyw)
+{
+	switch (asyw->state.fb->format->format) {
+	case DRM_FORMAT_YUYV: asyw->image.format = 0x28; break;
+	case DRM_FORMAT_UYVY: asyw->image.format = 0x29; break;
+	default:
+		WARN_ON(1);
+		return -EINVAL;
+	}
+	asyw->image.colorspace = 1;
+	return 0;
+}
+
+static int
+nv50_wndw_atomic_check_acquire_rgb(struct nv50_wndw_atom *asyw)
+{
+	switch (asyw->state.fb->format->format) {
+	case DRM_FORMAT_C8         : asyw->image.format = 0x1e; break;
+	case DRM_FORMAT_XRGB8888   :
+	case DRM_FORMAT_ARGB8888   : asyw->image.format = 0xcf; break;
+	case DRM_FORMAT_RGB565     : asyw->image.format = 0xe8; break;
+	case DRM_FORMAT_XRGB1555   :
+	case DRM_FORMAT_ARGB1555   : asyw->image.format = 0xe9; break;
+	case DRM_FORMAT_XBGR2101010:
+	case DRM_FORMAT_ABGR2101010: asyw->image.format = 0xd1; break;
+	case DRM_FORMAT_XBGR8888   :
+	case DRM_FORMAT_ABGR8888   : asyw->image.format = 0xd5; break;
+	case DRM_FORMAT_XRGB2101010:
+	case DRM_FORMAT_ARGB2101010: asyw->image.format = 0xdf; break;
+	default:
+		return -EINVAL;
+	}
+	asyw->image.colorspace = 0;
+	return 0;
+}
+
+static int
+nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw, bool modeset,
+			       struct nv50_wndw_atom *armw,
+			       struct nv50_wndw_atom *asyw,
+			       struct nv50_head_atom *asyh)
+{
+	struct nouveau_framebuffer *fb = nouveau_framebuffer(asyw->state.fb);
+	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
+	int ret;
+
+	NV_ATOMIC(drm, "%s acquire\n", wndw->plane.name);
+
+	if (asyw->state.fb != armw->state.fb || !armw->visible || modeset) {
+		asyw->image.w = fb->base.width;
+		asyw->image.h = fb->base.height;
+		asyw->image.kind = fb->nvbo->kind;
+
+		ret = nv50_wndw_atomic_check_acquire_rgb(asyw);
+		if (ret) {
+			ret = nv50_wndw_atomic_check_acquire_yuv(asyw);
+			if (ret)
+				return ret;
+		}
+
+		if (asyw->image.kind) {
+			asyw->image.layout = 0;
+			if (drm->client.device.info.chipset >= 0xc0)
+				asyw->image.blockh = fb->nvbo->mode >> 4;
+			else
+				asyw->image.blockh = fb->nvbo->mode;
+			asyw->image.blocks[0] = fb->base.pitches[0] / 64;
+			asyw->image.pitch[0] = 0;
+		} else {
+			asyw->image.layout = 1;
+			asyw->image.blockh = 0;
+			asyw->image.blocks[0] = 0;
+			asyw->image.pitch[0] = fb->base.pitches[0];
+		}
+
+		if (!(asyh->state.pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC))
+			asyw->image.interval = 1;
+		else
+			asyw->image.interval = 0;
+		asyw->image.mode = asyw->image.interval ? 0 : 1;
+		asyw->set.image = wndw->func->image_set != NULL;
+	}
+
+	if (wndw->func->scale_set) {
+		asyw->scale.sx = asyw->state.src_x >> 16;
+		asyw->scale.sy = asyw->state.src_y >> 16;
+		asyw->scale.sw = asyw->state.src_w >> 16;
+		asyw->scale.sh = asyw->state.src_h >> 16;
+		asyw->scale.dw = asyw->state.crtc_w;
+		asyw->scale.dh = asyw->state.crtc_h;
+		if (memcmp(&armw->scale, &asyw->scale, sizeof(asyw->scale)))
+			asyw->set.scale = true;
+	}
+
+	if (wndw->immd) {
+		asyw->point.x = asyw->state.crtc_x;
+		asyw->point.y = asyw->state.crtc_y;
+		if (memcmp(&armw->point, &asyw->point, sizeof(asyw->point)))
+			asyw->set.point = true;
+	}
+
+	return wndw->func->acquire(wndw, asyw, asyh);
+}
+
+static void
+nv50_wndw_atomic_check_lut(struct nv50_wndw *wndw,
+			   struct nv50_wndw_atom *armw,
+			   struct nv50_wndw_atom *asyw,
+			   struct nv50_head_atom *asyh)
+{
+	struct drm_property_blob *ilut = asyh->state.degamma_lut;
+
+	/* I8 format without an input LUT makes no sense, and the
+	 * HW error-checks for this.
+	 *
+	 * In order to handle legacy gamma, when there's no input
+	 * LUT we need to steal the output LUT and use it instead.
+	 */
+	if (!ilut && asyw->state.fb->format->format == DRM_FORMAT_C8) {
+		/* This should be an error, but there's legacy clients
+		 * that do a modeset before providing a gamma table.
+		 *
+		 * We keep the window disabled to avoid angering HW.
+		 */
+		if (!(ilut = asyh->state.gamma_lut)) {
+			asyw->visible = false;
+			return;
+		}
+
+		if (wndw->func->ilut)
+			asyh->wndw.olut |= BIT(wndw->id);
+	} else {
+		asyh->wndw.olut &= ~BIT(wndw->id);
+	}
+
+	/* Recalculate LUT state. */
+	memset(&asyw->xlut, 0x00, sizeof(asyw->xlut));
+	if ((asyw->ilut = wndw->func->ilut ? ilut : NULL)) {
+		wndw->func->ilut(wndw, asyw);
+		asyw->xlut.handle = wndw->wndw.vram.handle;
+		asyw->xlut.i.buffer = !asyw->xlut.i.buffer;
+		asyw->set.xlut = true;
+	}
+
+	/* Handle setting base SET_OUTPUT_LUT_LO_ENABLE_USE_CORE_LUT. */
+	if (wndw->func->olut_core &&
+	    (!armw->visible || (armw->xlut.handle && !asyw->xlut.handle)))
+		asyw->set.xlut = true;
+
+	/* Can't do an immediate flip while changing the LUT. */
+	asyh->state.pageflip_flags &= ~DRM_MODE_PAGE_FLIP_ASYNC;
+}
+
+static int
+nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state)
+{
+	struct nouveau_drm *drm = nouveau_drm(plane->dev);
+	struct nv50_wndw *wndw = nv50_wndw(plane);
+	struct nv50_wndw_atom *armw = nv50_wndw_atom(wndw->plane.state);
+	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
+	struct nv50_head_atom *harm = NULL, *asyh = NULL;
+	bool modeset = false;
+	int ret;
+
+	NV_ATOMIC(drm, "%s atomic_check\n", plane->name);
+
+	/* Fetch the assembly state for the head the window will belong to,
+	 * and determine whether the window will be visible.
+	 */
+	if (asyw->state.crtc) {
+		asyh = nv50_head_atom_get(asyw->state.state, asyw->state.crtc);
+		if (IS_ERR(asyh))
+			return PTR_ERR(asyh);
+		modeset = drm_atomic_crtc_needs_modeset(&asyh->state);
+		asyw->visible = asyh->state.active;
+	} else {
+		asyw->visible = false;
+	}
+
+	/* Fetch assembly state for the head the window used to belong to. */
+	if (armw->state.crtc) {
+		harm = nv50_head_atom_get(asyw->state.state, armw->state.crtc);
+		if (IS_ERR(harm))
+			return PTR_ERR(harm);
+	}
+
+	/* LUT configuration can potentially cause the window to be disabled. */
+	if (asyw->visible && wndw->func->xlut_set &&
+	    (!armw->visible ||
+	     asyh->state.color_mgmt_changed ||
+	     asyw->state.fb->format->format !=
+	     armw->state.fb->format->format))
+		nv50_wndw_atomic_check_lut(wndw, armw, asyw, asyh);
+
+	/* Calculate new window state. */
+	if (asyw->visible) {
+		ret = nv50_wndw_atomic_check_acquire(wndw, modeset,
+						     armw, asyw, asyh);
+		if (ret)
+			return ret;
+
+		asyh->wndw.mask |= BIT(wndw->id);
+	} else
+	if (armw->visible) {
+		nv50_wndw_atomic_check_release(wndw, asyw, harm);
+		harm->wndw.mask &= ~BIT(wndw->id);
+	} else {
+		return 0;
+	}
+
+	/* Aside from the obvious case where the window is actively being
+	 * disabled, we might also need to temporarily disable the window
+	 * when performing certain modeset operations.
+	 */
+	if (!asyw->visible || modeset) {
+		asyw->clr.ntfy = armw->ntfy.handle != 0;
+		asyw->clr.sema = armw->sema.handle != 0;
+		asyw->clr.xlut = armw->xlut.handle != 0;
+		if (wndw->func->image_clr)
+			asyw->clr.image = armw->image.handle[0] != 0;
+	}
+
+	return 0;
+}
+
+static void
+nv50_wndw_cleanup_fb(struct drm_plane *plane, struct drm_plane_state *old_state)
+{
+	struct nouveau_framebuffer *fb = nouveau_framebuffer(old_state->fb);
+	struct nouveau_drm *drm = nouveau_drm(plane->dev);
+
+	NV_ATOMIC(drm, "%s cleanup: %p\n", plane->name, old_state->fb);
+	if (!old_state->fb)
+		return;
+
+	nouveau_bo_unpin(fb->nvbo);
+}
+
+static int
+nv50_wndw_prepare_fb(struct drm_plane *plane, struct drm_plane_state *state)
+{
+	struct nouveau_framebuffer *fb = nouveau_framebuffer(state->fb);
+	struct nouveau_drm *drm = nouveau_drm(plane->dev);
+	struct nv50_wndw *wndw = nv50_wndw(plane);
+	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
+	struct nv50_head_atom *asyh;
+	struct nv50_wndw_ctxdma *ctxdma;
+	int ret;
+
+	NV_ATOMIC(drm, "%s prepare: %p\n", plane->name, state->fb);
+	if (!asyw->state.fb)
+		return 0;
+
+	ret = nouveau_bo_pin(fb->nvbo, TTM_PL_FLAG_VRAM, true);
+	if (ret)
+		return ret;
+
+	if (wndw->ctxdma.parent) {
+		ctxdma = nv50_wndw_ctxdma_new(wndw, fb);
+		if (IS_ERR(ctxdma)) {
+			nouveau_bo_unpin(fb->nvbo);
+			return PTR_ERR(ctxdma);
+		}
+
+		asyw->image.handle[0] = ctxdma->object.handle;
+	}
+
+	asyw->state.fence = reservation_object_get_excl_rcu(fb->nvbo->bo.resv);
+	asyw->image.offset[0] = fb->nvbo->bo.offset;
+
+	if (wndw->func->prepare) {
+		asyh = nv50_head_atom_get(asyw->state.state, asyw->state.crtc);
+		if (IS_ERR(asyh))
+			return PTR_ERR(asyh);
+
+		wndw->func->prepare(wndw, asyh, asyw);
+	}
+
+	return 0;
+}
+
+static const struct drm_plane_helper_funcs
+nv50_wndw_helper = {
+	.prepare_fb = nv50_wndw_prepare_fb,
+	.cleanup_fb = nv50_wndw_cleanup_fb,
+	.atomic_check = nv50_wndw_atomic_check,
+};
+
+static void
+nv50_wndw_atomic_destroy_state(struct drm_plane *plane,
+			       struct drm_plane_state *state)
+{
+	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
+	__drm_atomic_helper_plane_destroy_state(&asyw->state);
+	kfree(asyw);
+}
+
+static struct drm_plane_state *
+nv50_wndw_atomic_duplicate_state(struct drm_plane *plane)
+{
+	struct nv50_wndw_atom *armw = nv50_wndw_atom(plane->state);
+	struct nv50_wndw_atom *asyw;
+	if (!(asyw = kmalloc(sizeof(*asyw), GFP_KERNEL)))
+		return NULL;
+	__drm_atomic_helper_plane_duplicate_state(plane, &asyw->state);
+	asyw->sema = armw->sema;
+	asyw->ntfy = armw->ntfy;
+	asyw->ilut = NULL;
+	asyw->xlut = armw->xlut;
+	asyw->image = armw->image;
+	asyw->point = armw->point;
+	asyw->clr.mask = 0;
+	asyw->set.mask = 0;
+	return &asyw->state;
+}
+
+static void
+nv50_wndw_reset(struct drm_plane *plane)
+{
+	struct nv50_wndw_atom *asyw;
+
+	if (WARN_ON(!(asyw = kzalloc(sizeof(*asyw), GFP_KERNEL))))
+		return;
+
+	if (plane->state)
+		plane->funcs->atomic_destroy_state(plane, plane->state);
+	plane->state = &asyw->state;
+	plane->state->plane = plane;
+	plane->state->rotation = DRM_MODE_ROTATE_0;
+}
+
+static void
+nv50_wndw_destroy(struct drm_plane *plane)
+{
+	struct nv50_wndw *wndw = nv50_wndw(plane);
+	struct nv50_wndw_ctxdma *ctxdma, *ctxtmp;
+
+	list_for_each_entry_safe(ctxdma, ctxtmp, &wndw->ctxdma.list, head) {
+		nv50_wndw_ctxdma_del(ctxdma);
+	}
+
+	nvif_notify_fini(&wndw->notify);
+	nv50_dmac_destroy(&wndw->wimm);
+	nv50_dmac_destroy(&wndw->wndw);
+
+	nv50_lut_fini(&wndw->ilut);
+
+	drm_plane_cleanup(&wndw->plane);
+	kfree(wndw);
+}
+
+const struct drm_plane_funcs
+nv50_wndw = {
+	.update_plane = drm_atomic_helper_update_plane,
+	.disable_plane = drm_atomic_helper_disable_plane,
+	.destroy = nv50_wndw_destroy,
+	.reset = nv50_wndw_reset,
+	.atomic_duplicate_state = nv50_wndw_atomic_duplicate_state,
+	.atomic_destroy_state = nv50_wndw_atomic_destroy_state,
+};
+
+static int
+nv50_wndw_notify(struct nvif_notify *notify)
+{
+	return NVIF_NOTIFY_KEEP;
+}
+
+void
+nv50_wndw_fini(struct nv50_wndw *wndw)
+{
+	nvif_notify_put(&wndw->notify);
+}
+
+void
+nv50_wndw_init(struct nv50_wndw *wndw)
+{
+	nvif_notify_get(&wndw->notify);
+}
+
+int
+nv50_wndw_new_(const struct nv50_wndw_func *func, struct drm_device *dev,
+	       enum drm_plane_type type, const char *name, int index,
+	       const u32 *format, u32 heads,
+	       enum nv50_disp_interlock_type interlock_type, u32 interlock_data,
+	       struct nv50_wndw **pwndw)
+{
+	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nvif_mmu *mmu = &drm->client.mmu;
+	struct nv50_disp *disp = nv50_disp(dev);
+	struct nv50_wndw *wndw;
+	int nformat;
+	int ret;
+
+	if (!(wndw = *pwndw = kzalloc(sizeof(*wndw), GFP_KERNEL)))
+		return -ENOMEM;
+	wndw->func = func;
+	wndw->id = index;
+	wndw->interlock.type = interlock_type;
+	wndw->interlock.data = interlock_data;
+
+	wndw->ctxdma.parent = &wndw->wndw.base.user;
+	INIT_LIST_HEAD(&wndw->ctxdma.list);
+
+	for (nformat = 0; format[nformat]; nformat++);
+
+	ret = drm_universal_plane_init(dev, &wndw->plane, heads, &nv50_wndw,
+				       format, nformat, NULL,
+				       type, "%s-%d", name, index);
+	if (ret) {
+		kfree(*pwndw);
+		*pwndw = NULL;
+		return ret;
+	}
+
+	drm_plane_helper_add(&wndw->plane, &nv50_wndw_helper);
+
+	if (wndw->func->ilut) {
+		ret = nv50_lut_init(disp, mmu, &wndw->ilut);
+		if (ret)
+			return ret;
+	}
+
+	wndw->notify.func = nv50_wndw_notify;
+	return 0;
+}
+
+int
+nv50_wndw_new(struct nouveau_drm *drm, enum drm_plane_type type, int index,
+	      struct nv50_wndw **pwndw)
+{
+	struct {
+		s32 oclass;
+		int version;
+		int (*new)(struct nouveau_drm *, enum drm_plane_type,
+			   int, s32, struct nv50_wndw **);
+	} wndws[] = {
+		{ GV100_DISP_WINDOW_CHANNEL_DMA, 0, wndwc37e_new },
+		{}
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	int cid, ret;
+
+	cid = nvif_mclass(&disp->disp->object, wndws);
+	if (cid < 0) {
+		NV_ERROR(drm, "No supported window class\n");
+		return cid;
+	}
+
+	ret = wndws[cid].new(drm, type, index, wndws[cid].oclass, pwndw);
+	if (ret)
+		return ret;
+
+	return nv50_wimm_init(drm, *pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndw.h b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
new file mode 100644
index 000000000000..b0b6428034b0
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndw.h
@@ -0,0 +1,96 @@
+#ifndef __NV50_KMS_WNDW_H__
+#define __NV50_KMS_WNDW_H__
+#define nv50_wndw(p) container_of((p), struct nv50_wndw, plane)
+#include "disp.h"
+#include "atom.h"
+#include "lut.h"
+
+#include <nvif/notify.h>
+
+struct nv50_wndw_ctxdma {
+	struct list_head head;
+	struct nvif_object object;
+};
+
+struct nv50_wndw {
+	const struct nv50_wndw_func *func;
+	const struct nv50_wimm_func *immd;
+	int id;
+	struct nv50_disp_interlock interlock;
+
+	struct {
+		struct nvif_object *parent;
+		struct list_head list;
+	} ctxdma;
+
+	struct drm_plane plane;
+
+	struct nv50_lut ilut;
+
+	struct nv50_dmac wndw;
+	struct nv50_dmac wimm;
+
+	struct nvif_notify notify;
+	u16 ntfy;
+	u16 sema;
+	u32 data;
+};
+
+int nv50_wndw_new_(const struct nv50_wndw_func *, struct drm_device *,
+		   enum drm_plane_type, const char *name, int index,
+		   const u32 *format, enum nv50_disp_interlock_type,
+		   u32 interlock_data, u32 heads, struct nv50_wndw **);
+void nv50_wndw_init(struct nv50_wndw *);
+void nv50_wndw_fini(struct nv50_wndw *);
+void nv50_wndw_flush_set(struct nv50_wndw *, u32 *interlock,
+			 struct nv50_wndw_atom *);
+void nv50_wndw_flush_clr(struct nv50_wndw *, u32 *interlock, bool flush,
+			 struct nv50_wndw_atom *);
+void nv50_wndw_ntfy_enable(struct nv50_wndw *, struct nv50_wndw_atom *);
+int nv50_wndw_wait_armed(struct nv50_wndw *, struct nv50_wndw_atom *);
+
+struct nv50_wndw_func {
+	int (*acquire)(struct nv50_wndw *, struct nv50_wndw_atom *asyw,
+		       struct nv50_head_atom *asyh);
+	void (*release)(struct nv50_wndw *, struct nv50_wndw_atom *asyw,
+			struct nv50_head_atom *asyh);
+	void (*prepare)(struct nv50_wndw *, struct nv50_head_atom *asyh,
+			struct nv50_wndw_atom *asyw);
+
+	void (*sema_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	void (*sema_clr)(struct nv50_wndw *);
+	void (*ntfy_reset)(struct nouveau_bo *, u32 offset);
+	void (*ntfy_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	void (*ntfy_clr)(struct nv50_wndw *);
+	int (*ntfy_wait_begun)(struct nouveau_bo *, u32 offset,
+			       struct nvif_device *);
+	void (*ilut)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	bool olut_core;
+	void (*xlut_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	void (*xlut_clr)(struct nv50_wndw *);
+	void (*image_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
+	void (*image_clr)(struct nv50_wndw *);
+	void (*scale_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
+
+	void (*update)(struct nv50_wndw *, u32 *interlock);
+};
+
+extern const struct drm_plane_funcs nv50_wndw;
+
+void base507c_ntfy_reset(struct nouveau_bo *, u32);
+int base507c_ntfy_wait_begun(struct nouveau_bo *, u32, struct nvif_device *);
+
+struct nv50_wimm_func {
+	void (*point)(struct nv50_wndw *, struct nv50_wndw_atom *);
+
+	void (*update)(struct nv50_wndw *, u32 *interlock);
+};
+
+extern const struct nv50_wimm_func curs507a;
+
+int wndwc37e_new(struct nouveau_drm *, enum drm_plane_type, int, s32,
+		 struct nv50_wndw **);
+
+int nv50_wndw_new(struct nouveau_drm *, enum drm_plane_type, int index,
+		  struct nv50_wndw **);
+#endif
diff --git a/drivers/gpu/drm/nouveau/dispnv50/wndwc37e.c b/drivers/gpu/drm/nouveau/dispnv50/wndwc37e.c
new file mode 100644
index 000000000000..44afb0f069a5
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/dispnv50/wndwc37e.c
@@ -0,0 +1,278 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "wndw.h"
+#include "atom.h"
+
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_plane_helper.h>
+#include <nouveau_bo.h>
+
+#include <nvif/clc37e.h>
+
+static void
+wndwc37e_ilut_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x02b8, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_ilut_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 4))) {
+		evo_mthd(push, 0x02b0, 3);
+		evo_data(push, asyw->xlut.i.output_mode << 8 |
+			       asyw->xlut.i.range << 4 |
+			       asyw->xlut.i.size);
+		evo_data(push, asyw->xlut.i.offset >> 8);
+		evo_data(push, asyw->xlut.handle);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_ilut(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	asyw->xlut.i.mode = 2;
+	asyw->xlut.i.size = 0;
+	asyw->xlut.i.range = 0;
+	asyw->xlut.i.output_mode = 1;
+}
+
+static void
+wndwc37e_image_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 4))) {
+		evo_mthd(push, 0x0308, 1);
+		evo_data(push, 0x00000000);
+		evo_mthd(push, 0x0240, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+
+	if (!(push = evo_wait(&wndw->wndw, 25)))
+		return;
+
+	evo_mthd(push, 0x0308, 1);
+	evo_data(push, asyw->image.mode << 4 | asyw->image.interval);
+	evo_mthd(push, 0x0224, 4);
+	evo_data(push, asyw->image.h << 16 | asyw->image.w);
+	evo_data(push, asyw->image.layout << 4 | asyw->image.blockh);
+	evo_data(push, asyw->image.colorspace << 8 | asyw->image.format);
+	evo_data(push, asyw->image.blocks[0] | (asyw->image.pitch[0] >> 6));
+	evo_mthd(push, 0x0240, 1);
+	evo_data(push, asyw->image.handle[0]);
+	evo_mthd(push, 0x0260, 1);
+	evo_data(push, asyw->image.offset[0] >> 8);
+	evo_mthd(push, 0x0290, 1);
+	evo_data(push, (asyw->state.src_y >> 16) << 16 |
+		       (asyw->state.src_x >> 16));
+	evo_mthd(push, 0x0298, 1);
+	evo_data(push, (asyw->state.src_h >> 16) << 16 |
+		       (asyw->state.src_w >> 16));
+	evo_mthd(push, 0x02a4, 1);
+	evo_data(push, asyw->state.crtc_h << 16 |
+		       asyw->state.crtc_w);
+
+	/*XXX: Composition-related stuff.  Need to implement properly. */
+	evo_mthd(push, 0x02ec, 1);
+	evo_data(push, (2 - (wndw->id & 1)) << 4);
+	evo_mthd(push, 0x02f4, 5);
+	evo_data(push, 0x00000011);
+	evo_data(push, 0xffff0000);
+	evo_data(push, 0xffff0000);
+	evo_data(push, 0xffff0000);
+	evo_data(push, 0xffff0000);
+	evo_kick(push, &wndw->wndw);
+}
+
+static void
+wndwc37e_ntfy_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x021c, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_ntfy_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 3))) {
+		evo_mthd(push, 0x021c, 2);
+		evo_data(push, asyw->ntfy.handle);
+		evo_data(push, asyw->ntfy.offset | asyw->ntfy.awaken);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_sema_clr(struct nv50_wndw *wndw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 2))) {
+		evo_mthd(push, 0x0218, 1);
+		evo_data(push, 0x00000000);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_sema_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 5))) {
+		evo_mthd(push, 0x020c, 4);
+		evo_data(push, asyw->sema.offset);
+		evo_data(push, asyw->sema.acquire);
+		evo_data(push, asyw->sema.release);
+		evo_data(push, asyw->sema.handle);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_update(struct nv50_wndw *wndw, u32 *interlock)
+{
+	u32 *push;
+	if ((push = evo_wait(&wndw->wndw, 5))) {
+		evo_mthd(push, 0x0370, 2);
+		evo_data(push, interlock[NV50_DISP_INTERLOCK_CURS] << 1 |
+			       interlock[NV50_DISP_INTERLOCK_CORE]);
+		evo_data(push, interlock[NV50_DISP_INTERLOCK_WNDW]);
+		evo_mthd(push, 0x0200, 1);
+		if (interlock[NV50_DISP_INTERLOCK_WIMM] & wndw->interlock.data)
+			evo_data(push, 0x00001001);
+		else
+			evo_data(push, 0x00000001);
+		evo_kick(push, &wndw->wndw);
+	}
+}
+
+static void
+wndwc37e_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+}
+
+static int
+wndwc37e_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
+		 struct nv50_head_atom *asyh)
+{
+	return drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
+						   DRM_PLANE_HELPER_NO_SCALING,
+						   DRM_PLANE_HELPER_NO_SCALING,
+						   true, true);
+}
+
+static const u32
+wndwc37e_format[] = {
+	DRM_FORMAT_C8,
+	DRM_FORMAT_YUYV,
+	DRM_FORMAT_UYVY,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_ARGB1555,
+	DRM_FORMAT_XBGR2101010,
+	DRM_FORMAT_ABGR2101010,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_ABGR8888,
+	DRM_FORMAT_XRGB2101010,
+	DRM_FORMAT_ARGB2101010,
+	0
+};
+
+static const struct nv50_wndw_func
+wndwc37e = {
+	.acquire = wndwc37e_acquire,
+	.release = wndwc37e_release,
+	.sema_set = wndwc37e_sema_set,
+	.sema_clr = wndwc37e_sema_clr,
+	.ntfy_set = wndwc37e_ntfy_set,
+	.ntfy_clr = wndwc37e_ntfy_clr,
+	.ntfy_reset = corec37d_ntfy_init,
+	.ntfy_wait_begun = base507c_ntfy_wait_begun,
+	.ilut = wndwc37e_ilut,
+	.xlut_set = wndwc37e_ilut_set,
+	.xlut_clr = wndwc37e_ilut_clr,
+	.image_set = wndwc37e_image_set,
+	.image_clr = wndwc37e_image_clr,
+	.update = wndwc37e_update,
+};
+
+static int
+wndwc37e_new_(const struct nv50_wndw_func *func, struct nouveau_drm *drm,
+	      enum drm_plane_type type, int index, s32 oclass, u32 heads,
+	      struct nv50_wndw **pwndw)
+{
+	struct nvc37e_window_channel_dma_v0 args = {
+		.pushbuf = 0xb0007e00 | index,
+		.index = index,
+	};
+	struct nv50_disp *disp = nv50_disp(drm->dev);
+	struct nv50_wndw *wndw;
+	int ret;
+
+	ret = nv50_wndw_new_(func, drm->dev, type, "wndw", index,
+			     wndwc37e_format, heads, NV50_DISP_INTERLOCK_WNDW,
+			     BIT(index), &wndw);
+	if (*pwndw = wndw, ret)
+		return ret;
+
+	ret = nv50_dmac_create(&drm->client.device, &disp->disp->object,
+			       &oclass, 0, &args, sizeof(args),
+			       disp->sync->bo.offset, &wndw->wndw);
+	if (ret) {
+		NV_ERROR(drm, "qndw%04x allocation failed: %d\n", oclass, ret);
+		return ret;
+	}
+
+	wndw->ntfy = NV50_DISP_WNDW_NTFY(wndw->id);
+	wndw->sema = NV50_DISP_WNDW_SEM0(wndw->id);
+	wndw->data = 0x00000000;
+	return 0;
+}
+
+int
+wndwc37e_new(struct nouveau_drm *drm, enum drm_plane_type type, int index,
+	     s32 oclass, struct nv50_wndw **pwndw)
+{
+	return wndwc37e_new_(&wndwc37e, drm, type, index, oclass,
+			     BIT(index >> 1), pwndw);
+}
diff --git a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
index 2740278d226b..4f5233107f5f 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/cl0080.h
@@ -31,6 +31,7 @@ struct nv_device_info_v0 {
 #define NV_DEVICE_INFO_V0_KEPLER                                           0x08
 #define NV_DEVICE_INFO_V0_MAXWELL                                          0x09
 #define NV_DEVICE_INFO_V0_PASCAL                                           0x0a
+#define NV_DEVICE_INFO_V0_VOLTA                                            0x0b
 	__u8  family;
 	__u8  pad06[2];
 	__u64 ram_size;
@@ -39,9 +40,55 @@ struct nv_device_info_v0 {
 	char  name[64];
 };
 
+struct nv_device_info_v1 {
+	__u8  version;
+	__u8  count;
+	__u8  pad02[6];
+	struct nv_device_info_v1_data {
+		__u64 mthd; /* NV_DEVICE_INFO_* (see below). */
+		__u64 data;
+	} data[];
+};
+
 struct nv_device_time_v0 {
 	__u8  version;
 	__u8  pad01[7];
 	__u64 time;
 };
+
+#define NV_DEVICE_INFO_UNIT                               (0xffffffffULL << 32)
+#define NV_DEVICE_INFO(n)                          ((n) | (0x00000000ULL << 32))
+#define NV_DEVICE_FIFO(n)                          ((n) | (0x00000001ULL << 32))
+
+/* This will be returned for unsupported queries. */
+#define NV_DEVICE_INFO_INVALID                                           ~0ULL
+
+/* These return a mask of available engines of particular type. */
+#define NV_DEVICE_INFO_ENGINE_SW                     NV_DEVICE_INFO(0x00000000)
+#define NV_DEVICE_INFO_ENGINE_GR                     NV_DEVICE_INFO(0x00000001)
+#define NV_DEVICE_INFO_ENGINE_MPEG                   NV_DEVICE_INFO(0x00000002)
+#define NV_DEVICE_INFO_ENGINE_ME                     NV_DEVICE_INFO(0x00000003)
+#define NV_DEVICE_INFO_ENGINE_CIPHER                 NV_DEVICE_INFO(0x00000004)
+#define NV_DEVICE_INFO_ENGINE_BSP                    NV_DEVICE_INFO(0x00000005)
+#define NV_DEVICE_INFO_ENGINE_VP                     NV_DEVICE_INFO(0x00000006)
+#define NV_DEVICE_INFO_ENGINE_CE                     NV_DEVICE_INFO(0x00000007)
+#define NV_DEVICE_INFO_ENGINE_SEC                    NV_DEVICE_INFO(0x00000008)
+#define NV_DEVICE_INFO_ENGINE_MSVLD                  NV_DEVICE_INFO(0x00000009)
+#define NV_DEVICE_INFO_ENGINE_MSPDEC                 NV_DEVICE_INFO(0x0000000a)
+#define NV_DEVICE_INFO_ENGINE_MSPPP                  NV_DEVICE_INFO(0x0000000b)
+#define NV_DEVICE_INFO_ENGINE_MSENC                  NV_DEVICE_INFO(0x0000000c)
+#define NV_DEVICE_INFO_ENGINE_VIC                    NV_DEVICE_INFO(0x0000000d)
+#define NV_DEVICE_INFO_ENGINE_SEC2                   NV_DEVICE_INFO(0x0000000e)
+#define NV_DEVICE_INFO_ENGINE_NVDEC                  NV_DEVICE_INFO(0x0000000f)
+#define NV_DEVICE_INFO_ENGINE_NVENC                  NV_DEVICE_INFO(0x00000010)
+
+/* Returns the number of available channels. */
+#define NV_DEVICE_FIFO_CHANNELS                      NV_DEVICE_FIFO(0x00000000)
+
+/* Returns a mask of available runlists. */
+#define NV_DEVICE_FIFO_RUNLISTS                      NV_DEVICE_FIFO(0x00000001)
+
+/* These return a mask of engines available on a particular runlist. */
+#define NV_DEVICE_FIFO_RUNLIST_ENGINES(n)     ((n) + NV_DEVICE_FIFO(0x00000010))
+#define NV_DEVICE_FIFO_RUNLIST_ENGINES__SIZE                                64
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvif/cla06f.h b/drivers/gpu/drm/nouveau/include/nvif/cla06f.h
index 56f5bd81e480..fbfcffc5feb2 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/cla06f.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/cla06f.h
@@ -4,25 +4,11 @@
 
 struct kepler_channel_gpfifo_a_v0 {
 	__u8  version;
-	__u8  pad01[5];
+	__u8  pad01[1];
 	__u16 chid;
-#define NVA06F_V0_ENGINE_SW                                          0x00000001
-#define NVA06F_V0_ENGINE_GR                                          0x00000002
-#define NVA06F_V0_ENGINE_SEC                                         0x00000004
-#define NVA06F_V0_ENGINE_MSVLD                                       0x00000010
-#define NVA06F_V0_ENGINE_MSPDEC                                      0x00000020
-#define NVA06F_V0_ENGINE_MSPPP                                       0x00000040
-#define NVA06F_V0_ENGINE_MSENC                                       0x00000080
-#define NVA06F_V0_ENGINE_VIC                                         0x00000100
-#define NVA06F_V0_ENGINE_NVDEC                                       0x00000200
-#define NVA06F_V0_ENGINE_NVENC0                                      0x00000400
-#define NVA06F_V0_ENGINE_NVENC1                                      0x00000800
-#define NVA06F_V0_ENGINE_CE0                                         0x00010000
-#define NVA06F_V0_ENGINE_CE1                                         0x00020000
-#define NVA06F_V0_ENGINE_CE2                                         0x00040000
-	__u32 engines;
 	__u32 ilength;
 	__u64 ioffset;
+	__u64 runlist;
 	__u64 vmm;
 };
 
diff --git a/drivers/gpu/drm/nouveau/include/nvif/class.h b/drivers/gpu/drm/nouveau/include/nvif/class.h
index a7c5bf572788..6db56bd7d67e 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/class.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/class.h
@@ -52,6 +52,8 @@
 
 #define NV04_DISP                                     /* cl0046.h */ 0x00000046
 
+#define VOLTA_USERMODE_A                                             0x0000c361
+
 #define NV03_CHANNEL_DMA                              /* cl506b.h */ 0x0000006b
 #define NV10_CHANNEL_DMA                              /* cl506b.h */ 0x0000006e
 #define NV17_CHANNEL_DMA                              /* cl506b.h */ 0x0000176e
@@ -66,6 +68,7 @@
 #define KEPLER_CHANNEL_GPFIFO_B                       /* cla06f.h */ 0x0000a16f
 #define MAXWELL_CHANNEL_GPFIFO_A                      /* cla06f.h */ 0x0000b06f
 #define PASCAL_CHANNEL_GPFIFO_A                       /* cla06f.h */ 0x0000c06f
+#define VOLTA_CHANNEL_GPFIFO_A                        /* cla06f.h */ 0x0000c36f
 
 #define NV50_DISP                                     /* cl5070.h */ 0x00005070
 #define G82_DISP                                      /* cl5070.h */ 0x00008270
@@ -79,6 +82,7 @@
 #define GM200_DISP                                    /* cl5070.h */ 0x00009570
 #define GP100_DISP                                    /* cl5070.h */ 0x00009770
 #define GP102_DISP                                    /* cl5070.h */ 0x00009870
+#define GV100_DISP                                    /* cl5070.h */ 0x0000c370
 
 #define NV31_MPEG                                                    0x00003174
 #define G82_MPEG                                                     0x00008274
@@ -90,6 +94,7 @@
 #define GT214_DISP_CURSOR                             /* cl507a.h */ 0x0000857a
 #define GF110_DISP_CURSOR                             /* cl507a.h */ 0x0000907a
 #define GK104_DISP_CURSOR                             /* cl507a.h */ 0x0000917a
+#define GV100_DISP_CURSOR                             /* cl507a.h */ 0x0000c37a
 
 #define NV50_DISP_OVERLAY                             /* cl507b.h */ 0x0000507b
 #define G82_DISP_OVERLAY                              /* cl507b.h */ 0x0000827b
@@ -97,6 +102,8 @@
 #define GF110_DISP_OVERLAY                            /* cl507b.h */ 0x0000907b
 #define GK104_DISP_OVERLAY                            /* cl507b.h */ 0x0000917b
 
+#define GV100_DISP_WINDOW_IMM_CHANNEL_DMA             /* clc37b.h */ 0x0000c37b
+
 #define NV50_DISP_BASE_CHANNEL_DMA                    /* cl507c.h */ 0x0000507c
 #define G82_DISP_BASE_CHANNEL_DMA                     /* cl507c.h */ 0x0000827c
 #define GT200_DISP_BASE_CHANNEL_DMA                   /* cl507c.h */ 0x0000837c
@@ -117,6 +124,7 @@
 #define GM200_DISP_CORE_CHANNEL_DMA                   /* cl507d.h */ 0x0000957d
 #define GP100_DISP_CORE_CHANNEL_DMA                   /* cl507d.h */ 0x0000977d
 #define GP102_DISP_CORE_CHANNEL_DMA                   /* cl507d.h */ 0x0000987d
+#define GV100_DISP_CORE_CHANNEL_DMA                   /* cl507d.h */ 0x0000c37d
 
 #define NV50_DISP_OVERLAY_CHANNEL_DMA                 /* cl507e.h */ 0x0000507e
 #define G82_DISP_OVERLAY_CHANNEL_DMA                  /* cl507e.h */ 0x0000827e
@@ -125,6 +133,8 @@
 #define GF110_DISP_OVERLAY_CONTROL_DMA                /* cl507e.h */ 0x0000907e
 #define GK104_DISP_OVERLAY_CONTROL_DMA                /* cl507e.h */ 0x0000917e
 
+#define GV100_DISP_WINDOW_CHANNEL_DMA                 /* clc37e.h */ 0x0000c37e
+
 #define NV50_TESLA                                                   0x00005097
 #define G82_TESLA                                                    0x00008297
 #define GT200_TESLA                                                  0x00008397
@@ -145,6 +155,8 @@
 #define PASCAL_A                                      /* cl9097.h */ 0x0000c097
 #define PASCAL_B                                      /* cl9097.h */ 0x0000c197
 
+#define VOLTA_A                                       /* cl9097.h */ 0x0000c397
+
 #define NV74_BSP                                                     0x000074b0
 
 #define GT212_MSVLD                                                  0x000085b1
@@ -170,6 +182,7 @@
 #define MAXWELL_DMA_COPY_A                                           0x0000b0b5
 #define PASCAL_DMA_COPY_A                                            0x0000c0b5
 #define PASCAL_DMA_COPY_B                                            0x0000c1b5
+#define VOLTA_DMA_COPY_A                                             0x0000c3b5
 
 #define FERMI_DECOMPRESS                                             0x000090b8
 
@@ -183,6 +196,7 @@
 #define MAXWELL_COMPUTE_B                                            0x0000b1c0
 #define PASCAL_COMPUTE_A                                             0x0000c0c0
 #define PASCAL_COMPUTE_B                                             0x0000c1c0
+#define VOLTA_COMPUTE_A                                              0x0000c3c0
 
 #define NV74_CIPHER                                                  0x000074c1
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvif/clc37b.h b/drivers/gpu/drm/nouveau/include/nvif/clc37b.h
new file mode 100644
index 000000000000..89b18189d43b
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/include/nvif/clc37b.h
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __NVIF_CLC37B_H__
+#define __NVIF_CLC37B_H__
+
+struct nvc37b_window_imm_channel_dma_v0 {
+	__u8  version;
+	__u8  index;
+	__u8  pad02[6];
+	__u64 pushbuf;
+};
+#endif
diff --git a/drivers/gpu/drm/nouveau/include/nvif/clc37e.h b/drivers/gpu/drm/nouveau/include/nvif/clc37e.h
new file mode 100644
index 000000000000..899db9e915ef
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/include/nvif/clc37e.h
@@ -0,0 +1,13 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __NVIF_CLC37E_H__
+#define __NVIF_CLC37E_H__
+
+struct nvc37e_window_channel_dma_v0 {
+	__u8  version;
+	__u8  index;
+	__u8  pad02[6];
+	__u64 pushbuf;
+};
+
+#define NVC37E_WINDOW_CHANNEL_DMA_V0_NTFY_UEVENT                           0x00
+#endif
diff --git a/drivers/gpu/drm/nouveau/include/nvif/device.h b/drivers/gpu/drm/nouveau/include/nvif/device.h
index 6edb6266857e..ef839bd1d37e 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/device.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/device.h
@@ -4,10 +4,18 @@
 
 #include <nvif/object.h>
 #include <nvif/cl0080.h>
+#include <nvif/user.h>
 
 struct nvif_device {
 	struct nvif_object object;
 	struct nv_device_info_v0 info;
+
+	struct nvif_fifo_runlist {
+		u64 engines;
+	} *runlist;
+	int runlists;
+
+	struct nvif_user user;
 };
 
 int  nvif_device_init(struct nvif_object *, u32 handle, s32 oclass, void *, u32,
@@ -67,6 +75,5 @@ u64  nvif_device_time(struct nvif_device *);
 #include <engine/fifo.h>
 #include <engine/gr.h>
 
-#define nvxx_fifo(a) nvxx_device(a)->fifo
 #define nvxx_gr(a) nvxx_device(a)->gr
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvif/disp.h b/drivers/gpu/drm/nouveau/include/nvif/disp.h
new file mode 100644
index 000000000000..7c0eda375c01
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/include/nvif/disp.h
@@ -0,0 +1,12 @@
+#ifndef __NVIF_DISP_H__
+#define __NVIF_DISP_H__
+#include <nvif/object.h>
+struct nvif_device;
+
+struct nvif_disp {
+	struct nvif_object object;
+};
+
+int nvif_disp_ctor(struct nvif_device *, s32 oclass, struct nvif_disp *);
+void nvif_disp_dtor(struct nvif_disp *);
+#endif
diff --git a/drivers/gpu/drm/nouveau/include/nvif/fifo.h b/drivers/gpu/drm/nouveau/include/nvif/fifo.h
new file mode 100644
index 000000000000..e9468c9f9abf
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/include/nvif/fifo.h
@@ -0,0 +1,18 @@
+#ifndef __NVIF_FIFO_H__
+#define __NVIF_FIFO_H__
+#include <nvif/device.h>
+
+/* Returns mask of runlists that support a NV_DEVICE_INFO_ENGINE_* type. */
+u64 nvif_fifo_runlist(struct nvif_device *, u64 engine);
+
+/* CE-supporting runlists (excluding GRCE, if others exist). */
+static inline u64
+nvif_fifo_runlist_ce(struct nvif_device *device)
+{
+	u64 runmgr = nvif_fifo_runlist(device, NV_DEVICE_INFO_ENGINE_GR);
+	u64 runmce = nvif_fifo_runlist(device, NV_DEVICE_INFO_ENGINE_CE);
+	if (runmce && !(runmce &= ~runmgr))
+		runmce = runmgr;
+	return runmce;
+}
+#endif
diff --git a/drivers/gpu/drm/nouveau/include/nvif/mem.h b/drivers/gpu/drm/nouveau/include/nvif/mem.h
index b542fe38398e..80ee4ab0f016 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/mem.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/mem.h
@@ -15,4 +15,6 @@ int nvif_mem_init_type(struct nvif_mmu *mmu, s32 oclass, int type, u8 page,
 int nvif_mem_init(struct nvif_mmu *mmu, s32 oclass, u8 type, u8 page,
 		  u64 size, void *argv, u32 argc, struct nvif_mem *);
 void nvif_mem_fini(struct nvif_mem *);
+
+int nvif_mem_init_map(struct nvif_mmu *, u8 type, u64 size, struct nvif_mem *);
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvif/mmu.h b/drivers/gpu/drm/nouveau/include/nvif/mmu.h
index c8cd5b5b0688..747ecf67e403 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/mmu.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/mmu.h
@@ -8,6 +8,7 @@ struct nvif_mmu {
 	u8  heap_nr;
 	u8  type_nr;
 	u16 kind_nr;
+	s32 mem;
 
 	struct {
 		u64 size;
diff --git a/drivers/gpu/drm/nouveau/include/nvif/object.h b/drivers/gpu/drm/nouveau/include/nvif/object.h
index a2d5244ff2b7..8407651f6ac6 100644
--- a/drivers/gpu/drm/nouveau/include/nvif/object.h
+++ b/drivers/gpu/drm/nouveau/include/nvif/object.h
@@ -78,7 +78,7 @@ struct nvif_mclass {
 #define nvif_mclass(o,m) ({                                                    \
 	struct nvif_object *object = (o);                                      \
 	struct nvif_sclass *sclass;                                            \
-	const typeof(m[0]) *mclass = (m);                                      \
+	typeof(m[0]) *mclass = (m);                                            \
 	int ret = -ENODEV;                                                     \
 	int cnt, i, j;                                                         \
                                                                                \
@@ -99,6 +99,22 @@ struct nvif_mclass {
 	ret;                                                                   \
 })
 
+#define nvif_sclass(o,m,u) ({                                                  \
+	const typeof(m[0]) *_mclass = (m);                                     \
+	s32 _oclass = (u);                                                     \
+	int _cid;                                                              \
+	if (_oclass) {                                                         \
+		for (_cid = 0; _mclass[_cid].oclass; _cid++) {                 \
+			if (_mclass[_cid].oclass == _oclass)                   \
+				break;                                         \
+		}                                                              \
+		_cid = _mclass[_cid].oclass ? _cid : -ENOSYS;                  \
+	} else {                                                               \
+		_cid = nvif_mclass((o), _mclass);                              \
+	}                                                                      \
+	_cid;                                                                  \
+})
+
 /*XXX*/
 #include <core/object.h>
 #define nvxx_object(a) ({                                                      \
diff --git a/drivers/gpu/drm/nouveau/include/nvif/user.h b/drivers/gpu/drm/nouveau/include/nvif/user.h
new file mode 100644
index 000000000000..03c11826b693
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/include/nvif/user.h
@@ -0,0 +1,19 @@
+#ifndef __NVIF_USER_H__
+#define __NVIF_USER_H__
+#include <nvif/object.h>
+struct nvif_device;
+
+struct nvif_user {
+	const struct nvif_user_func *func;
+	struct nvif_object object;
+};
+
+struct nvif_user_func {
+	void (*doorbell)(struct nvif_user *, u32 token);
+};
+
+int nvif_user_init(struct nvif_device *);
+void nvif_user_fini(struct nvif_device *);
+
+extern const struct nvif_user_func nvif_userc361;
+#endif
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/device.h b/drivers/gpu/drm/nouveau/include/nvkm/core/device.h
index 560265b15ec2..d83d834b7452 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/core/device.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/core/device.h
@@ -22,6 +22,7 @@ enum nvkm_devidx {
 	NVKM_SUBDEV_LTC,
 	NVKM_SUBDEV_MMU,
 	NVKM_SUBDEV_BAR,
+	NVKM_SUBDEV_FAULT,
 	NVKM_SUBDEV_PMU,
 	NVKM_SUBDEV_VOLT,
 	NVKM_SUBDEV_ICCSENSE,
@@ -37,7 +38,10 @@ enum nvkm_devidx {
 	NVKM_ENGINE_CE3,
 	NVKM_ENGINE_CE4,
 	NVKM_ENGINE_CE5,
-	NVKM_ENGINE_CE_LAST = NVKM_ENGINE_CE5,
+	NVKM_ENGINE_CE6,
+	NVKM_ENGINE_CE7,
+	NVKM_ENGINE_CE8,
+	NVKM_ENGINE_CE_LAST = NVKM_ENGINE_CE8,
 
 	NVKM_ENGINE_CIPHER,
 	NVKM_ENGINE_DISP,
@@ -109,6 +113,7 @@ struct nvkm_device {
 		NV_E0    = 0xe0,
 		GM100    = 0x110,
 		GP100    = 0x130,
+		GV100    = 0x140,
 	} card_type;
 	u32 chipset;
 	u8  chiprev;
@@ -123,6 +128,7 @@ struct nvkm_device {
 	struct nvkm_bus *bus;
 	struct nvkm_clk *clk;
 	struct nvkm_devinit *devinit;
+	struct nvkm_fault *fault;
 	struct nvkm_fb *fb;
 	struct nvkm_fuse *fuse;
 	struct nvkm_gpio *gpio;
@@ -143,7 +149,7 @@ struct nvkm_device {
 	struct nvkm_volt *volt;
 
 	struct nvkm_engine *bsp;
-	struct nvkm_engine *ce[6];
+	struct nvkm_engine *ce[9];
 	struct nvkm_engine *cipher;
 	struct nvkm_disp *disp;
 	struct nvkm_dma *dma;
@@ -194,6 +200,7 @@ struct nvkm_device_chip {
 	int (*bus     )(struct nvkm_device *, int idx, struct nvkm_bus **);
 	int (*clk     )(struct nvkm_device *, int idx, struct nvkm_clk **);
 	int (*devinit )(struct nvkm_device *, int idx, struct nvkm_devinit **);
+	int (*fault   )(struct nvkm_device *, int idx, struct nvkm_fault **);
 	int (*fb      )(struct nvkm_device *, int idx, struct nvkm_fb **);
 	int (*fuse    )(struct nvkm_device *, int idx, struct nvkm_fuse **);
 	int (*gpio    )(struct nvkm_device *, int idx, struct nvkm_gpio **);
@@ -214,7 +221,7 @@ struct nvkm_device_chip {
 	int (*volt    )(struct nvkm_device *, int idx, struct nvkm_volt **);
 
 	int (*bsp     )(struct nvkm_device *, int idx, struct nvkm_engine **);
-	int (*ce[6]   )(struct nvkm_device *, int idx, struct nvkm_engine **);
+	int (*ce[9]   )(struct nvkm_device *, int idx, struct nvkm_engine **);
 	int (*cipher  )(struct nvkm_device *, int idx, struct nvkm_engine **);
 	int (*disp    )(struct nvkm_device *, int idx, struct nvkm_disp **);
 	int (*dma     )(struct nvkm_device *, int idx, struct nvkm_dma **);
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/engine.h b/drivers/gpu/drm/nouveau/include/nvkm/core/engine.h
index ebf8473a39fe..8a2be5b635e2 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/core/engine.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/core/engine.h
@@ -18,6 +18,7 @@ struct nvkm_engine_func {
 	void *(*dtor)(struct nvkm_engine *);
 	void (*preinit)(struct nvkm_engine *);
 	int (*oneinit)(struct nvkm_engine *);
+	int (*info)(struct nvkm_engine *, u64 mthd, u64 *data);
 	int (*init)(struct nvkm_engine *);
 	int (*fini)(struct nvkm_engine *, bool suspend);
 	void (*intr)(struct nvkm_engine *);
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/core/subdev.h b/drivers/gpu/drm/nouveau/include/nvkm/core/subdev.h
index 63df2290177f..85a0777c2ce4 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/core/subdev.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/core/subdev.h
@@ -17,6 +17,7 @@ struct nvkm_subdev_func {
 	void *(*dtor)(struct nvkm_subdev *);
 	int (*preinit)(struct nvkm_subdev *);
 	int (*oneinit)(struct nvkm_subdev *);
+	int (*info)(struct nvkm_subdev *, u64 mthd, u64 *data);
 	int (*init)(struct nvkm_subdev *);
 	int (*fini)(struct nvkm_subdev *, bool suspend);
 	void (*intr)(struct nvkm_subdev *);
@@ -29,6 +30,7 @@ void nvkm_subdev_del(struct nvkm_subdev **);
 int  nvkm_subdev_preinit(struct nvkm_subdev *);
 int  nvkm_subdev_init(struct nvkm_subdev *);
 int  nvkm_subdev_fini(struct nvkm_subdev *, bool suspend);
+int  nvkm_subdev_info(struct nvkm_subdev *, u64, u64 *);
 void nvkm_subdev_intr(struct nvkm_subdev *);
 
 /* subdev logging */
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/ce.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/ce.h
index 553245994450..fc295e1faa19 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/engine/ce.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/ce.h
@@ -10,4 +10,5 @@ int gm107_ce_new(struct nvkm_device *, int, struct nvkm_engine **);
 int gm200_ce_new(struct nvkm_device *, int, struct nvkm_engine **);
 int gp100_ce_new(struct nvkm_device *, int, struct nvkm_engine **);
 int gp102_ce_new(struct nvkm_device *, int, struct nvkm_engine **);
+int gv100_ce_new(struct nvkm_device *, int, struct nvkm_engine **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/disp.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/disp.h
index e83193d3ccab..ef7dc0844d26 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/engine/disp.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/disp.h
@@ -35,4 +35,5 @@ int gm107_disp_new(struct nvkm_device *, int, struct nvkm_disp **);
 int gm200_disp_new(struct nvkm_device *, int, struct nvkm_disp **);
 int gp100_disp_new(struct nvkm_device *, int, struct nvkm_disp **);
 int gp102_disp_new(struct nvkm_device *, int, struct nvkm_disp **);
+int gv100_disp_new(struct nvkm_device *, int, struct nvkm_disp **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/dma.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/dma.h
index 0f9c1c702ed6..f0c1b2c8c78c 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/engine/dma.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/dma.h
@@ -27,4 +27,5 @@ int nv04_dma_new(struct nvkm_device *, int, struct nvkm_dma **);
 int nv50_dma_new(struct nvkm_device *, int, struct nvkm_dma **);
 int gf100_dma_new(struct nvkm_device *, int, struct nvkm_dma **);
 int gf119_dma_new(struct nvkm_device *, int, struct nvkm_dma **);
+int gv100_dma_new(struct nvkm_device *, int, struct nvkm_dma **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h
index c17b3a9bf8fb..7e39fbed2519 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/fifo.h
@@ -4,6 +4,7 @@
 #include <core/engine.h>
 #include <core/object.h>
 #include <core/event.h>
+struct nvkm_fault_data;
 
 #define NVKM_FIFO_CHID_NR 4096
 
@@ -45,6 +46,7 @@ struct nvkm_fifo {
 	struct nvkm_event kevent; /* channel killed */
 };
 
+void nvkm_fifo_fault(struct nvkm_fifo *, struct nvkm_fault_data *);
 void nvkm_fifo_pause(struct nvkm_fifo *, unsigned long *);
 void nvkm_fifo_start(struct nvkm_fifo *, unsigned long *);
 
@@ -71,4 +73,5 @@ int gm200_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **);
 int gm20b_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **);
 int gp100_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **);
 int gp10b_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **);
+int gv100_fifo_new(struct nvkm_device *, int, struct nvkm_fifo **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h b/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h
index fb18f105fc43..ba1518ff8b66 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/engine/gr.h
@@ -45,6 +45,8 @@ int gm200_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 int gm20b_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 int gp100_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 int gp102_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
+int gp104_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 int gp107_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 int gp10b_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
+int gv100_gr_new(struct nvkm_device *, int, struct nvkm_gr **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/dp.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/dp.h
index df34b41838d6..512e25a41803 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/dp.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/bios/dp.h
@@ -1,6 +1,10 @@
 /* SPDX-License-Identifier: GPL-2.0 */
 #ifndef __NVBIOS_DP_H__
 #define __NVBIOS_DP_H__
+
+u16
+nvbios_dp_table(struct nvkm_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len);
+
 struct nvbios_dpout {
 	u16 type;
 	u16 mask;
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/devinit.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/devinit.h
index 40558064d589..486e7635c29d 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/devinit.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/devinit.h
@@ -30,4 +30,5 @@ int mcp89_devinit_new(struct nvkm_device *, int, struct nvkm_devinit **);
 int gf100_devinit_new(struct nvkm_device *, int, struct nvkm_devinit **);
 int gm107_devinit_new(struct nvkm_device *, int, struct nvkm_devinit **);
 int gm200_devinit_new(struct nvkm_device *, int, struct nvkm_devinit **);
+int gv100_devinit_new(struct nvkm_device *, int, struct nvkm_devinit **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
new file mode 100644
index 000000000000..5a77498fe6a0
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fault.h
@@ -0,0 +1,33 @@
+#ifndef __NVKM_FAULT_H__
+#define __NVKM_FAULT_H__
+#include <core/subdev.h>
+#include <core/notify.h>
+
+struct nvkm_fault {
+	const struct nvkm_fault_func *func;
+	struct nvkm_subdev subdev;
+
+	struct nvkm_fault_buffer *buffer[2];
+	int buffer_nr;
+
+	struct nvkm_event event;
+
+	struct nvkm_notify nrpfb;
+};
+
+struct nvkm_fault_data {
+	u64  addr;
+	u64  inst;
+	u64  time;
+	u8 engine;
+	u8  valid;
+	u8    gpc;
+	u8    hub;
+	u8 access;
+	u8 client;
+	u8 reason;
+};
+
+int gp100_fault_new(struct nvkm_device *, int, struct nvkm_fault **);
+int gv100_fault_new(struct nvkm_device *, int, struct nvkm_fault **);
+#endif
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h
index 92be0e5269c6..96ccc624ee81 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/fb.h
@@ -83,6 +83,7 @@ int gm20b_fb_new(struct nvkm_device *, int, struct nvkm_fb **);
 int gp100_fb_new(struct nvkm_device *, int, struct nvkm_fb **);
 int gp102_fb_new(struct nvkm_device *, int, struct nvkm_fb **);
 int gp10b_fb_new(struct nvkm_device *, int, struct nvkm_fb **);
+int gv100_fb_new(struct nvkm_device *, int, struct nvkm_fb **);
 
 #include <subdev/bios.h>
 #include <subdev/bios/ramcfg.h>
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/ltc.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/ltc.h
index 95b611554d53..9db5f8293198 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/ltc.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/ltc.h
@@ -21,12 +21,14 @@ struct nvkm_ltc {
 	int zbc_max;
 	u32 zbc_color[NVKM_LTC_MAX_ZBC_CNT][4];
 	u32 zbc_depth[NVKM_LTC_MAX_ZBC_CNT];
+	u32 zbc_stencil[NVKM_LTC_MAX_ZBC_CNT];
 };
 
 void nvkm_ltc_tags_clear(struct nvkm_device *, u32 first, u32 count);
 
 int nvkm_ltc_zbc_color_get(struct nvkm_ltc *, int index, const u32[4]);
 int nvkm_ltc_zbc_depth_get(struct nvkm_ltc *, int index, const u32);
+int nvkm_ltc_zbc_stencil_get(struct nvkm_ltc *, int index, const u32);
 
 void nvkm_ltc_invalidate(struct nvkm_ltc *);
 void nvkm_ltc_flush(struct nvkm_ltc *);
@@ -37,4 +39,5 @@ int gk20a_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **);
 int gm107_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **);
 int gm200_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **);
 int gp100_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **);
+int gp102_ltc_new(struct nvkm_device *, int, struct nvkm_ltc **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/include/nvkm/subdev/mmu.h b/drivers/gpu/drm/nouveau/include/nvkm/subdev/mmu.h
index baab93398e54..688595545e21 100644
--- a/drivers/gpu/drm/nouveau/include/nvkm/subdev/mmu.h
+++ b/drivers/gpu/drm/nouveau/include/nvkm/subdev/mmu.h
@@ -129,4 +129,5 @@ int gm200_mmu_new(struct nvkm_device *, int, struct nvkm_mmu **);
 int gm20b_mmu_new(struct nvkm_device *, int, struct nvkm_mmu **);
 int gp100_mmu_new(struct nvkm_device *, int, struct nvkm_mmu **);
 int gp10b_mmu_new(struct nvkm_device *, int, struct nvkm_mmu **);
+int gv100_mmu_new(struct nvkm_device *, int, struct nvkm_mmu **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nouveau_abi16.c b/drivers/gpu/drm/nouveau/nouveau_abi16.c
index ece650a0c5f9..e67a471331b5 100644
--- a/drivers/gpu/drm/nouveau/nouveau_abi16.c
+++ b/drivers/gpu/drm/nouveau/nouveau_abi16.c
@@ -23,6 +23,7 @@
 
 #include <nvif/client.h>
 #include <nvif/driver.h>
+#include <nvif/fifo.h>
 #include <nvif/ioctl.h>
 #include <nvif/class.h>
 #include <nvif/cl0002.h>
@@ -102,6 +103,7 @@ nouveau_abi16_swclass(struct nouveau_drm *drm)
 	case NV_DEVICE_INFO_V0_KEPLER:
 	case NV_DEVICE_INFO_V0_MAXWELL:
 	case NV_DEVICE_INFO_V0_PASCAL:
+	case NV_DEVICE_INFO_V0_VOLTA:
 		return NVIF_CLASS_SW_GF100;
 	}
 
@@ -137,7 +139,7 @@ nouveau_abi16_chan_fini(struct nouveau_abi16 *abi16,
 	if (chan->ntfy) {
 		nouveau_vma_del(&chan->ntfy_vma);
 		nouveau_bo_unpin(chan->ntfy);
-		drm_gem_object_unreference_unlocked(&chan->ntfy->gem);
+		drm_gem_object_put_unlocked(&chan->ntfy->gem);
 	}
 
 	if (chan->heap.block_size)
@@ -256,6 +258,7 @@ nouveau_abi16_ioctl_channel_alloc(ABI16_IOCTL_ARGS)
 	struct nouveau_abi16 *abi16 = nouveau_abi16_get(file_priv);
 	struct nouveau_abi16_chan *chan;
 	struct nvif_device *device;
+	u64 engine;
 	int ret;
 
 	if (unlikely(!abi16))
@@ -268,25 +271,26 @@ nouveau_abi16_ioctl_channel_alloc(ABI16_IOCTL_ARGS)
 
 	/* hack to allow channel engine type specification on kepler */
 	if (device->info.family >= NV_DEVICE_INFO_V0_KEPLER) {
-		if (init->fb_ctxdma_handle != ~0)
-			init->fb_ctxdma_handle = NVA06F_V0_ENGINE_GR;
-		else {
-			init->fb_ctxdma_handle = 0;
-#define _(A,B) if (init->tt_ctxdma_handle & (A)) init->fb_ctxdma_handle |= (B)
-			_(0x01, NVA06F_V0_ENGINE_GR);
-			_(0x02, NVA06F_V0_ENGINE_MSPDEC);
-			_(0x04, NVA06F_V0_ENGINE_MSPPP);
-			_(0x08, NVA06F_V0_ENGINE_MSVLD);
-			_(0x10, NVA06F_V0_ENGINE_CE0);
-			_(0x20, NVA06F_V0_ENGINE_CE1);
-			_(0x40, NVA06F_V0_ENGINE_MSENC);
-#undef _
+		if (init->fb_ctxdma_handle == ~0) {
+			switch (init->tt_ctxdma_handle) {
+			case 0x01: engine = NV_DEVICE_INFO_ENGINE_GR    ; break;
+			case 0x02: engine = NV_DEVICE_INFO_ENGINE_MSPDEC; break;
+			case 0x04: engine = NV_DEVICE_INFO_ENGINE_MSPPP ; break;
+			case 0x08: engine = NV_DEVICE_INFO_ENGINE_MSVLD ; break;
+			case 0x30: engine = NV_DEVICE_INFO_ENGINE_CE    ; break;
+			default:
+				return nouveau_abi16_put(abi16, -ENOSYS);
+			}
+		} else {
+			engine = NV_DEVICE_INFO_ENGINE_GR;
 		}
 
-		/* allow flips to be executed if this is a graphics channel */
+		if (engine != NV_DEVICE_INFO_ENGINE_CE)
+			engine = nvif_fifo_runlist(device, engine);
+		else
+			engine = nvif_fifo_runlist_ce(device);
+		init->fb_ctxdma_handle = engine;
 		init->tt_ctxdma_handle = 0;
-		if (init->fb_ctxdma_handle == NVA06F_V0_ENGINE_GR)
-			init->tt_ctxdma_handle = 1;
 	}
 
 	if (init->fb_ctxdma_handle == ~0 || init->tt_ctxdma_handle == ~0)
diff --git a/drivers/gpu/drm/nouveau/nouveau_acpi.c b/drivers/gpu/drm/nouveau/nouveau_acpi.c
index 5ffcb6683776..ffb195850314 100644
--- a/drivers/gpu/drm/nouveau/nouveau_acpi.c
+++ b/drivers/gpu/drm/nouveau/nouveau_acpi.c
@@ -193,7 +193,7 @@ static int nouveau_dsm_power_state(enum vga_switcheroo_client_id id,
 	return nouveau_dsm_set_discrete_state(nouveau_dsm_priv.dhandle, state);
 }
 
-static int nouveau_dsm_get_client_id(struct pci_dev *pdev)
+static enum vga_switcheroo_client_id nouveau_dsm_get_client_id(struct pci_dev *pdev)
 {
 	/* easy option one - intel vendor ID means Integrated */
 	if (pdev->vendor == PCI_VENDOR_ID_INTEL)
diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c
index 6f402c4f2bdd..7214022dfb91 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.c
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.c
@@ -214,7 +214,6 @@ nouveau_bo_new(struct nouveau_cli *cli, u64 size, int align,
 	INIT_LIST_HEAD(&nvbo->entry);
 	INIT_LIST_HEAD(&nvbo->vma_list);
 	nvbo->bo.bdev = &drm->ttm.bdev;
-	nvbo->cli = cli;
 
 	/* This is confusing, and doesn't actually mean we want an uncached
 	 * mapping, but is what NOUVEAU_GEM_DOMAIN_COHERENT gets translated
@@ -1142,6 +1141,8 @@ nouveau_bo_move_init(struct nouveau_drm *drm)
 			    struct ttm_mem_reg *, struct ttm_mem_reg *);
 		int (*init)(struct nouveau_channel *, u32 handle);
 	} _methods[] = {
+		{  "COPY", 4, 0xc3b5, nve0_bo_move_copy, nve0_bo_move_init },
+		{  "GRCE", 0, 0xc3b5, nve0_bo_move_copy, nvc0_bo_move_init },
 		{  "COPY", 4, 0xc1b5, nve0_bo_move_copy, nve0_bo_move_init },
 		{  "GRCE", 0, 0xc1b5, nve0_bo_move_copy, nvc0_bo_move_init },
 		{  "COPY", 4, 0xc0b5, nve0_bo_move_copy, nve0_bo_move_init },
diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.h b/drivers/gpu/drm/nouveau/nouveau_bo.h
index be8e00b49cde..73c48440d4d7 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.h
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.h
@@ -26,8 +26,6 @@ struct nouveau_bo {
 
 	struct list_head vma_list;
 
-	struct nouveau_cli *cli;
-
 	unsigned contig:1;
 	unsigned page:5;
 	unsigned kind:8;
diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.c b/drivers/gpu/drm/nouveau/nouveau_chan.c
index af1116655910..92d3115f96b5 100644
--- a/drivers/gpu/drm/nouveau/nouveau_chan.c
+++ b/drivers/gpu/drm/nouveau/nouveau_chan.c
@@ -163,12 +163,15 @@ nouveau_channel_prep(struct nouveau_drm *drm, struct nvif_device *device,
 			return ret;
 		}
 
+		chan->push.addr = chan->push.vma->addr;
+
+		if (device->info.family >= NV_DEVICE_INFO_V0_FERMI)
+			return 0;
+
 		args.target = NV_DMA_V0_TARGET_VM;
 		args.access = NV_DMA_V0_ACCESS_VM;
 		args.start = 0;
 		args.limit = cli->vmm.vmm.limit - 1;
-
-		chan->push.addr = chan->push.vma->addr;
 	} else
 	if (chan->push.buffer->bo.mem.mem_type == TTM_PL_VRAM) {
 		if (device->info.family == NV_DEVICE_INFO_V0_TNT) {
@@ -214,10 +217,11 @@ nouveau_channel_prep(struct nouveau_drm *drm, struct nvif_device *device,
 
 static int
 nouveau_channel_ind(struct nouveau_drm *drm, struct nvif_device *device,
-		    u32 engine, struct nouveau_channel **pchan)
+		    u64 runlist, struct nouveau_channel **pchan)
 {
 	struct nouveau_cli *cli = (void *)device->object.client;
-	static const u16 oclasses[] = { PASCAL_CHANNEL_GPFIFO_A,
+	static const u16 oclasses[] = { VOLTA_CHANNEL_GPFIFO_A,
+					PASCAL_CHANNEL_GPFIFO_A,
 					MAXWELL_CHANNEL_GPFIFO_A,
 					KEPLER_CHANNEL_GPFIFO_B,
 					KEPLER_CHANNEL_GPFIFO_A,
@@ -245,9 +249,9 @@ nouveau_channel_ind(struct nouveau_drm *drm, struct nvif_device *device,
 	do {
 		if (oclass[0] >= KEPLER_CHANNEL_GPFIFO_A) {
 			args.kepler.version = 0;
-			args.kepler.engines = engine;
 			args.kepler.ilength = 0x02000;
 			args.kepler.ioffset = 0x10000 + chan->push.addr;
+			args.kepler.runlist = runlist;
 			args.kepler.vmm = nvif_handle(&cli->vmm.vmm.object);
 			size = sizeof(args.kepler);
 		} else
@@ -474,3 +478,28 @@ done:
 	cli->base.super = super;
 	return ret;
 }
+
+int
+nouveau_channels_init(struct nouveau_drm *drm)
+{
+	struct {
+		struct nv_device_info_v1 m;
+		struct {
+			struct nv_device_info_v1_data channels;
+		} v;
+	} args = {
+		.m.version = 1,
+		.m.count = sizeof(args.v) / sizeof(args.v.channels),
+		.v.channels.mthd = NV_DEVICE_FIFO_CHANNELS,
+	};
+	struct nvif_object *device = &drm->client.device.object;
+	int ret;
+
+	ret = nvif_object_mthd(device, NV_DEVICE_V0_INFO, &args, sizeof(args));
+	if (ret || args.v.channels.mthd == NV_DEVICE_INFO_INVALID)
+		return -ENODEV;
+
+	drm->chan.nr = args.v.channels.data;
+	drm->chan.context_base = dma_fence_context_alloc(drm->chan.nr);
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/nouveau_chan.h b/drivers/gpu/drm/nouveau/nouveau_chan.h
index 14607c16a2bd..64454c2ebd90 100644
--- a/drivers/gpu/drm/nouveau/nouveau_chan.h
+++ b/drivers/gpu/drm/nouveau/nouveau_chan.h
@@ -45,6 +45,7 @@ struct nouveau_channel {
 	atomic_t killed;
 };
 
+int nouveau_channels_init(struct nouveau_drm *);
 
 int  nouveau_channel_new(struct nouveau_drm *, struct nvif_device *,
 			 u32 arg0, u32 arg1, struct nouveau_channel **);
diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c
index 6ed9cb053dfa..22a15478d23d 100644
--- a/drivers/gpu/drm/nouveau/nouveau_connector.c
+++ b/drivers/gpu/drm/nouveau/nouveau_connector.c
@@ -151,7 +151,7 @@ nouveau_conn_atomic_set_property(struct drm_connector *connector,
 				/* ... except prior to G80, where the code
 				 * doesn't support such things.
 				 */
-				if (disp->disp.oclass < NV50_DISP)
+				if (disp->disp.object.oclass < NV50_DISP)
 					return -EINVAL;
 				break;
 			default:
@@ -260,7 +260,7 @@ nouveau_conn_reset(struct drm_connector *connector)
 	asyc->procamp.color_vibrance = 150;
 	asyc->procamp.vibrant_hue = 90;
 
-	if (nouveau_display(connector->dev)->disp.oclass < NV50_DISP) {
+	if (nouveau_display(connector->dev)->disp.object.oclass < NV50_DISP) {
 		switch (connector->connector_type) {
 		case DRM_MODE_CONNECTOR_LVDS:
 			/* See note in nouveau_conn_atomic_set_property(). */
@@ -314,7 +314,7 @@ nouveau_conn_attach_properties(struct drm_connector *connector)
 	case DRM_MODE_CONNECTOR_TV:
 		break;
 	case DRM_MODE_CONNECTOR_VGA:
-		if (disp->disp.oclass < NV50_DISP)
+		if (disp->disp.object.oclass < NV50_DISP)
 			break; /* Can only scale on DFPs. */
 		/* Fall-through. */
 	default:
@@ -363,19 +363,11 @@ module_param_named(hdmimhz, nouveau_hdmimhz, int, 0400);
 struct nouveau_encoder *
 find_encoder(struct drm_connector *connector, int type)
 {
-	struct drm_device *dev = connector->dev;
 	struct nouveau_encoder *nv_encoder;
 	struct drm_encoder *enc;
-	int i, id;
-
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		id = connector->encoder_ids[i];
-		if (!id)
-			break;
+	int i;
 
-		enc = drm_encoder_find(dev, NULL, id);
-		if (!enc)
-			continue;
+	drm_connector_for_each_possible_encoder(connector, enc, i) {
 		nv_encoder = nouveau_encoder(enc);
 
 		if (type == DCB_OUTPUT_ANY ||
@@ -420,7 +412,7 @@ nouveau_connector_ddc_detect(struct drm_connector *connector)
 	struct nouveau_connector *nv_connector = nouveau_connector(connector);
 	struct nouveau_drm *drm = nouveau_drm(dev);
 	struct nvkm_gpio *gpio = nvxx_gpio(&drm->client.device);
-	struct nouveau_encoder *nv_encoder;
+	struct nouveau_encoder *nv_encoder = NULL;
 	struct drm_encoder *encoder;
 	int i, panel = -ENODEV;
 
@@ -436,14 +428,7 @@ nouveau_connector_ddc_detect(struct drm_connector *connector)
 		}
 	}
 
-	for (i = 0; nv_encoder = NULL, i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		int id = connector->encoder_ids[i];
-		if (id == 0)
-			break;
-
-		encoder = drm_encoder_find(dev, NULL, id);
-		if (!encoder)
-			continue;
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		nv_encoder = nouveau_encoder(encoder);
 
 		if (nv_encoder->dcb->type == DCB_OUTPUT_DP) {
@@ -565,7 +550,7 @@ nouveau_connector_detect(struct drm_connector *connector, bool force)
 
 	/* Cleanup the previous EDID block. */
 	if (nv_connector->edid) {
-		drm_mode_connector_update_edid_property(connector, NULL);
+		drm_connector_update_edid_property(connector, NULL);
 		kfree(nv_connector->edid);
 		nv_connector->edid = NULL;
 	}
@@ -590,7 +575,7 @@ nouveau_connector_detect(struct drm_connector *connector, bool force)
 		else
 			nv_connector->edid = drm_get_edid(connector, i2c);
 
-		drm_mode_connector_update_edid_property(connector,
+		drm_connector_update_edid_property(connector,
 							nv_connector->edid);
 		if (!nv_connector->edid) {
 			NV_ERROR(drm, "DDC responded, but no EDID for %s\n",
@@ -672,7 +657,7 @@ nouveau_connector_detect_lvds(struct drm_connector *connector, bool force)
 
 	/* Cleanup the previous EDID block. */
 	if (nv_connector->edid) {
-		drm_mode_connector_update_edid_property(connector, NULL);
+		drm_connector_update_edid_property(connector, NULL);
 		kfree(nv_connector->edid);
 		nv_connector->edid = NULL;
 	}
@@ -736,7 +721,7 @@ out:
 		status = connector_status_unknown;
 #endif
 
-	drm_mode_connector_update_edid_property(connector, nv_connector->edid);
+	drm_connector_update_edid_property(connector, nv_connector->edid);
 	nouveau_connector_set_encoder(connector, nv_encoder);
 	return status;
 }
@@ -1005,7 +990,7 @@ get_tmds_link_bandwidth(struct drm_connector *connector, bool hdmi)
 		return 112000;
 }
 
-static int
+static enum drm_mode_status
 nouveau_connector_mode_valid(struct drm_connector *connector,
 			     struct drm_display_mode *mode)
 {
@@ -1321,7 +1306,7 @@ nouveau_connector_create(struct drm_device *dev, int index)
 	}
 
 	/* HDMI 3D support */
-	if ((disp->disp.oclass >= G82_DISP)
+	if ((disp->disp.object.oclass >= G82_DISP)
 	    && ((type == DRM_MODE_CONNECTOR_DisplayPort)
 		|| (type == DRM_MODE_CONNECTOR_eDP)
 		|| (type == DRM_MODE_CONNECTOR_HDMIA)))
@@ -1343,7 +1328,7 @@ nouveau_connector_create(struct drm_device *dev, int index)
 	case DCB_CONNECTOR_LVDS_SPWG:
 	case DCB_CONNECTOR_eDP:
 		/* see note in nouveau_connector_set_property() */
-		if (disp->disp.oclass < NV50_DISP) {
+		if (disp->disp.object.oclass < NV50_DISP) {
 			nv_connector->scaling_mode = DRM_MODE_SCALE_FULLSCREEN;
 			break;
 		}
@@ -1366,8 +1351,8 @@ nouveau_connector_create(struct drm_device *dev, int index)
 		break;
 	}
 
-	ret = nvif_notify_init(&disp->disp, nouveau_connector_hotplug, true,
-			       NV04_DISP_NTFY_CONN,
+	ret = nvif_notify_init(&disp->disp.object, nouveau_connector_hotplug,
+			       true, NV04_DISP_NTFY_CONN,
 			       &(struct nvif_notify_conn_req_v0) {
 				.mask = NVIF_NOTIFY_CONN_V0_ANY,
 				.conn = index,
diff --git a/drivers/gpu/drm/nouveau/nouveau_debugfs.c b/drivers/gpu/drm/nouveau/nouveau_debugfs.c
index 963a4dba8213..9109b69cd052 100644
--- a/drivers/gpu/drm/nouveau/nouveau_debugfs.c
+++ b/drivers/gpu/drm/nouveau/nouveau_debugfs.c
@@ -160,7 +160,11 @@ nouveau_debugfs_pstate_set(struct file *file, const char __user *ubuf,
 		args.ustate = value;
 	}
 
+	ret = pm_runtime_get_sync(drm->dev);
+	if (IS_ERR_VALUE(ret) && ret != -EACCES)
+		return ret;
 	ret = nvif_mthd(ctrl, NVIF_CONTROL_PSTATE_USER, &args, sizeof(args));
+	pm_runtime_put_autosuspend(drm->dev);
 	if (ret < 0)
 		return ret;
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.c b/drivers/gpu/drm/nouveau/nouveau_display.c
index 009713404cc4..dfa236370726 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.c
+++ b/drivers/gpu/drm/nouveau/nouveau_display.c
@@ -116,7 +116,7 @@ nouveau_display_scanoutpos_head(struct drm_crtc *crtc, int *vpos, int *hpos,
 	bool ret = false;
 
 	do {
-		ret = nvif_mthd(&disp->disp, 0, &args, sizeof(args));
+		ret = nvif_mthd(&disp->disp.object, 0, &args, sizeof(args));
 		if (ret != 0)
 			return false;
 
@@ -175,7 +175,7 @@ nouveau_display_vblank_init(struct drm_device *dev)
 
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
 		struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
-		ret = nvif_notify_init(&disp->disp,
+		ret = nvif_notify_init(&disp->disp.object,
 				       nouveau_display_vblank_handler, false,
 				       NV04_DISP_NTFY_VBLANK,
 				       &(struct nvif_notify_head_req_v0) {
@@ -205,7 +205,7 @@ nouveau_user_framebuffer_destroy(struct drm_framebuffer *drm_fb)
 	struct nouveau_framebuffer *fb = nouveau_framebuffer(drm_fb);
 
 	if (fb->nvbo)
-		drm_gem_object_unreference_unlocked(&fb->nvbo->gem);
+		drm_gem_object_put_unlocked(&fb->nvbo->gem);
 
 	drm_framebuffer_cleanup(drm_fb);
 	kfree(fb);
@@ -287,7 +287,7 @@ nouveau_user_framebuffer_create(struct drm_device *dev,
 	if (ret == 0)
 		return &fb->base;
 
-	drm_gem_object_unreference_unlocked(gem);
+	drm_gem_object_put_unlocked(gem);
 	return ERR_PTR(ret);
 }
 
@@ -338,11 +338,9 @@ static struct nouveau_drm_prop_enum_list dither_depth[] = {
 	if (c) {                                                               \
 		p = drm_property_create(dev, DRM_MODE_PROP_ENUM, n, c);        \
 		l = (list);                                                    \
-		c = 0;                                                         \
 		while (p && l->gen_mask) {                                     \
 			if (l->gen_mask & (1 << (gen))) {                      \
-				drm_property_add_enum(p, c, l->type, l->name); \
-				c++;                                           \
+				drm_property_add_enum(p, l->type, l->name);    \
 			}                                                      \
 			l++;                                                   \
 		}                                                              \
@@ -456,10 +454,10 @@ nouveau_display_create_properties(struct drm_device *dev)
 	struct nouveau_display *disp = nouveau_display(dev);
 	int gen;
 
-	if (disp->disp.oclass < NV50_DISP)
+	if (disp->disp.object.oclass < NV50_DISP)
 		gen = 0;
 	else
-	if (disp->disp.oclass < GF110_DISP)
+	if (disp->disp.object.oclass < GF110_DISP)
 		gen = 1;
 	else
 		gen = 2;
@@ -535,31 +533,10 @@ nouveau_display_create(struct drm_device *dev)
 	drm_kms_helper_poll_disable(dev);
 
 	if (nouveau_modeset != 2 && drm->vbios.dcb.entries) {
-		static const u16 oclass[] = {
-			GP102_DISP,
-			GP100_DISP,
-			GM200_DISP,
-			GM107_DISP,
-			GK110_DISP,
-			GK104_DISP,
-			GF110_DISP,
-			GT214_DISP,
-			GT206_DISP,
-			GT200_DISP,
-			G82_DISP,
-			NV50_DISP,
-			NV04_DISP,
-		};
-		int i;
-
-		for (i = 0, ret = -ENODEV; ret && i < ARRAY_SIZE(oclass); i++) {
-			ret = nvif_object_init(&drm->client.device.object, 0,
-					       oclass[i], NULL, 0, &disp->disp);
-		}
-
+		ret = nvif_disp_ctor(&drm->client.device, 0, &disp->disp);
 		if (ret == 0) {
 			nouveau_display_create_properties(dev);
-			if (disp->disp.oclass < NV50_DISP)
+			if (disp->disp.object.oclass < NV50_DISP)
 				ret = nv04_display_create(dev);
 			else
 				ret = nv50_display_create(dev);
@@ -613,7 +590,7 @@ nouveau_display_destroy(struct drm_device *dev)
 	if (disp->dtor)
 		disp->dtor(dev);
 
-	nvif_object_fini(&disp->disp);
+	nvif_disp_dtor(&disp->disp);
 
 	nouveau_drm(dev)->display = NULL;
 	kfree(disp);
@@ -962,7 +939,7 @@ nouveau_display_dumb_create(struct drm_file *file_priv, struct drm_device *dev,
 		return ret;
 
 	ret = drm_gem_handle_create(file_priv, &bo->gem, &args->handle);
-	drm_gem_object_unreference_unlocked(&bo->gem);
+	drm_gem_object_put_unlocked(&bo->gem);
 	return ret;
 }
 
@@ -977,7 +954,7 @@ nouveau_display_dumb_map_offset(struct drm_file *file_priv,
 	if (gem) {
 		struct nouveau_bo *bo = nouveau_gem_object(gem);
 		*poffset = drm_vma_node_offset_addr(&bo->bo.vma_node);
-		drm_gem_object_unreference_unlocked(gem);
+		drm_gem_object_put_unlocked(gem);
 		return 0;
 	}
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_display.h b/drivers/gpu/drm/nouveau/nouveau_display.h
index 270ba56f2756..54aa7c3fa42d 100644
--- a/drivers/gpu/drm/nouveau/nouveau_display.h
+++ b/drivers/gpu/drm/nouveau/nouveau_display.h
@@ -2,6 +2,7 @@
 #ifndef __NOUVEAU_DISPLAY_H__
 #define __NOUVEAU_DISPLAY_H__
 #include "nouveau_drv.h"
+#include <nvif/disp.h>
 
 struct nouveau_framebuffer {
 	struct drm_framebuffer base;
@@ -38,7 +39,7 @@ struct nouveau_display {
 	int  (*init)(struct drm_device *);
 	void (*fini)(struct drm_device *);
 
-	struct nvif_object disp;
+	struct nvif_disp disp;
 
 	struct drm_property *dithering_mode;
 	struct drm_property *dithering_depth;
diff --git a/drivers/gpu/drm/nouveau/nouveau_dma.c b/drivers/gpu/drm/nouveau/nouveau_dma.c
index 10e84f6ca2b7..945afd34138e 100644
--- a/drivers/gpu/drm/nouveau/nouveau_dma.c
+++ b/drivers/gpu/drm/nouveau/nouveau_dma.c
@@ -28,6 +28,8 @@
 #include "nouveau_dma.h"
 #include "nouveau_vmm.h"
 
+#include <nvif/user.h>
+
 void
 OUT_RINGp(struct nouveau_channel *chan, const void *data, unsigned nr_dwords)
 {
@@ -80,18 +82,11 @@ READ_GET(struct nouveau_channel *chan, uint64_t *prev_get, int *timeout)
 }
 
 void
-nv50_dma_push(struct nouveau_channel *chan, struct nouveau_bo *bo,
-	      int delta, int length)
+nv50_dma_push(struct nouveau_channel *chan, u64 offset, int length)
 {
-	struct nouveau_cli *cli = (void *)chan->user.client;
+	struct nvif_user *user = &chan->drm->client.device.user;
 	struct nouveau_bo *pb = chan->push.buffer;
-	struct nouveau_vma *vma;
 	int ip = (chan->dma.ib_put * 2) + chan->dma.ib_base;
-	u64 offset;
-
-	vma = nouveau_vma_find(bo, &cli->vmm);
-	BUG_ON(!vma);
-	offset = vma->addr + delta;
 
 	BUG_ON(chan->dma.ib_free < 1);
 
@@ -105,6 +100,8 @@ nv50_dma_push(struct nouveau_channel *chan, struct nouveau_bo *bo,
 	nouveau_bo_rd32(pb, 0);
 
 	nvif_wr32(&chan->user, 0x8c, chan->dma.ib_put);
+	if (user->func && user->func->doorbell)
+		user->func->doorbell(user, chan->chid);
 	chan->dma.ib_free--;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_dma.h b/drivers/gpu/drm/nouveau/nouveau_dma.h
index 74e10b14a7da..fc5e3f41282d 100644
--- a/drivers/gpu/drm/nouveau/nouveau_dma.h
+++ b/drivers/gpu/drm/nouveau/nouveau_dma.h
@@ -31,8 +31,7 @@
 #include "nouveau_chan.h"
 
 int nouveau_dma_wait(struct nouveau_channel *, int slots, int size);
-void nv50_dma_push(struct nouveau_channel *, struct nouveau_bo *,
-		   int delta, int length);
+void nv50_dma_push(struct nouveau_channel *, u64 addr, int length);
 
 /*
  * There's a hw race condition where you can't jump to your PUT offset,
@@ -55,7 +54,6 @@ enum {
 
 	NvSub2D		= 3, /* DO NOT CHANGE - hardcoded for kepler gr fifo */
 	NvSubCopy	= 4, /* DO NOT CHANGE - hardcoded for kepler gr fifo */
-	FermiSw		= 5, /* DO NOT CHANGE (well.. 6/7 will work...) */
 };
 
 /* Object handles - for stuff that's doesn't use handle == oclass. */
@@ -151,7 +149,7 @@ FIRE_RING(struct nouveau_channel *chan)
 	chan->accel_done = true;
 
 	if (chan->dma.ib_max) {
-		nv50_dma_push(chan, chan->push.buffer, chan->dma.put << 2,
+		nv50_dma_push(chan, chan->push.addr + (chan->dma.put << 2),
 			      (chan->dma.cur - chan->dma.put) << 2);
 	} else {
 		WRITE_PUT(chan->dma.cur);
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index bbbf353682e1..c779ee3c665b 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -38,6 +38,8 @@
 #include <core/tegra.h>
 
 #include <nvif/driver.h>
+#include <nvif/fifo.h>
+#include <nvif/user.h>
 
 #include <nvif/class.h>
 #include <nvif/cl0002.h>
@@ -112,24 +114,22 @@ nouveau_name(struct drm_device *dev)
 }
 
 static inline bool
-nouveau_cli_work_ready(struct dma_fence *fence, bool wait)
+nouveau_cli_work_ready(struct dma_fence *fence)
 {
-	if (!dma_fence_is_signaled(fence)) {
-		if (!wait)
-			return false;
-		WARN_ON(dma_fence_wait_timeout(fence, false, 2 * HZ) <= 0);
-	}
+	if (!dma_fence_is_signaled(fence))
+		return false;
 	dma_fence_put(fence);
 	return true;
 }
 
 static void
-nouveau_cli_work_flush(struct nouveau_cli *cli, bool wait)
+nouveau_cli_work(struct work_struct *w)
 {
+	struct nouveau_cli *cli = container_of(w, typeof(*cli), work);
 	struct nouveau_cli_work *work, *wtmp;
 	mutex_lock(&cli->lock);
 	list_for_each_entry_safe(work, wtmp, &cli->worker, head) {
-		if (!work->fence || nouveau_cli_work_ready(work->fence, wait)) {
+		if (!work->fence || nouveau_cli_work_ready(work->fence)) {
 			list_del(&work->head);
 			work->func(work);
 		}
@@ -158,16 +158,16 @@ nouveau_cli_work_queue(struct nouveau_cli *cli, struct dma_fence *fence,
 }
 
 static void
-nouveau_cli_work(struct work_struct *w)
-{
-	struct nouveau_cli *cli = container_of(w, typeof(*cli), work);
-	nouveau_cli_work_flush(cli, false);
-}
-
-static void
 nouveau_cli_fini(struct nouveau_cli *cli)
 {
-	nouveau_cli_work_flush(cli, true);
+	/* All our channels are dead now, which means all the fences they
+	 * own are signalled, and all callback functions have been called.
+	 *
+	 * So, after flushing the workqueue, there should be nothing left.
+	 */
+	flush_work(&cli->work);
+	WARN_ON(!list_empty(&cli->worker));
+
 	usif_client_fini(cli);
 	nouveau_vmm_fini(&cli->vmm);
 	nvif_mmu_fini(&cli->mmu);
@@ -307,6 +307,16 @@ nouveau_accel_init(struct nouveau_drm *drm)
 	if (nouveau_noaccel)
 		return;
 
+	ret = nouveau_channels_init(drm);
+	if (ret)
+		return;
+
+	if (drm->client.device.info.family >= NV_DEVICE_INFO_V0_VOLTA) {
+		ret = nvif_user_init(device);
+		if (ret)
+			return;
+	}
+
 	/* initialise synchronisation routines */
 	/*XXX: this is crap, but the fence/channel stuff is a little
 	 *     backwards in some places.  this will be fixed.
@@ -338,6 +348,7 @@ nouveau_accel_init(struct nouveau_drm *drm)
 		case KEPLER_CHANNEL_GPFIFO_B:
 		case MAXWELL_CHANNEL_GPFIFO_A:
 		case PASCAL_CHANNEL_GPFIFO_A:
+		case VOLTA_CHANNEL_GPFIFO_A:
 			ret = nvc0_fence_create(drm);
 			break;
 		default:
@@ -354,13 +365,12 @@ nouveau_accel_init(struct nouveau_drm *drm)
 
 	if (device->info.family >= NV_DEVICE_INFO_V0_KEPLER) {
 		ret = nouveau_channel_new(drm, &drm->client.device,
-					  NVA06F_V0_ENGINE_CE0 |
-					  NVA06F_V0_ENGINE_CE1,
-					  0, &drm->cechan);
+					  nvif_fifo_runlist_ce(device), 0,
+					  &drm->cechan);
 		if (ret)
 			NV_ERROR(drm, "failed to create ce channel, %d\n", ret);
 
-		arg0 = NVA06F_V0_ENGINE_GR;
+		arg0 = nvif_fifo_runlist(device, NV_DEVICE_INFO_ENGINE_GR);
 		arg1 = 1;
 	} else
 	if (device->info.chipset >= 0xa3 &&
@@ -386,38 +396,36 @@ nouveau_accel_init(struct nouveau_drm *drm)
 		return;
 	}
 
-	ret = nvif_object_init(&drm->channel->user, NVDRM_NVSW,
-			       nouveau_abi16_swclass(drm), NULL, 0, &drm->nvsw);
-	if (ret == 0) {
-		ret = RING_SPACE(drm->channel, 2);
+	if (device->info.family < NV_DEVICE_INFO_V0_TESLA) {
+		ret = nvif_object_init(&drm->channel->user, NVDRM_NVSW,
+				       nouveau_abi16_swclass(drm), NULL, 0,
+				       &drm->nvsw);
 		if (ret == 0) {
-			if (device->info.family < NV_DEVICE_INFO_V0_FERMI) {
+			ret = RING_SPACE(drm->channel, 2);
+			if (ret == 0) {
 				BEGIN_NV04(drm->channel, NvSubSw, 0, 1);
-				OUT_RING  (drm->channel, NVDRM_NVSW);
-			} else
-			if (device->info.family < NV_DEVICE_INFO_V0_KEPLER) {
-				BEGIN_NVC0(drm->channel, FermiSw, 0, 1);
-				OUT_RING  (drm->channel, 0x001f0000);
+				OUT_RING  (drm->channel, drm->nvsw.handle);
+			}
+
+			ret = nvif_notify_init(&drm->nvsw,
+					       nouveau_flip_complete,
+					       false, NV04_NVSW_NTFY_UEVENT,
+					       NULL, 0, 0, &drm->flip);
+			if (ret == 0)
+				ret = nvif_notify_get(&drm->flip);
+			if (ret) {
+				nouveau_accel_fini(drm);
+				return;
 			}
 		}
 
-		ret = nvif_notify_init(&drm->nvsw, nouveau_flip_complete,
-				       false, NV04_NVSW_NTFY_UEVENT,
-				       NULL, 0, 0, &drm->flip);
-		if (ret == 0)
-			ret = nvif_notify_get(&drm->flip);
 		if (ret) {
+			NV_ERROR(drm, "failed to allocate sw class, %d\n", ret);
 			nouveau_accel_fini(drm);
 			return;
 		}
 	}
 
-	if (ret) {
-		NV_ERROR(drm, "failed to allocate software object, %d\n", ret);
-		nouveau_accel_fini(drm);
-		return;
-	}
-
 	if (device->info.family < NV_DEVICE_INFO_V0_FERMI) {
 		ret = nvkm_gpuobj_new(nvxx_device(&drm->client.device), 32, 0,
 				      false, NULL, &drm->notify);
@@ -904,8 +912,10 @@ nouveau_drm_open(struct drm_device *dev, struct drm_file *fpriv)
 	get_task_comm(tmpname, current);
 	snprintf(name, sizeof(name), "%s[%d]", tmpname, pid_nr(fpriv->pid));
 
-	if (!(cli = kzalloc(sizeof(*cli), GFP_KERNEL)))
-		return ret;
+	if (!(cli = kzalloc(sizeof(*cli), GFP_KERNEL))) {
+		ret = -ENOMEM;
+		goto done;
+	}
 
 	ret = nouveau_cli_init(drm, name, cli);
 	if (ret)
diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h
index 881b44b89a01..6e1acaec3400 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drv.h
+++ b/drivers/gpu/drm/nouveau/nouveau_drv.h
@@ -170,6 +170,12 @@ struct nouveau_drm {
 	/* synchronisation */
 	void *fence;
 
+	/* Global channel management. */
+	struct {
+		int nr;
+		u64 context_base;
+	} chan;
+
 	/* context for accelerated drm-internal operations */
 	struct nouveau_channel *cechan;
 	struct nouveau_channel *channel;
diff --git a/drivers/gpu/drm/nouveau/nouveau_encoder.h b/drivers/gpu/drm/nouveau/nouveau_encoder.h
index e28d966946a1..3517f920bf89 100644
--- a/drivers/gpu/drm/nouveau/nouveau_encoder.h
+++ b/drivers/gpu/drm/nouveau/nouveau_encoder.h
@@ -32,6 +32,7 @@
 #include <drm/drm_encoder_slave.h>
 #include <drm/drm_dp_mst_helper.h>
 #include "dispnv04/disp.h"
+struct nv50_head_atom;
 
 #define NV_DPMS_CLEARED 0x80
 
@@ -68,7 +69,7 @@ struct nouveau_encoder {
 	void (*enc_save)(struct drm_encoder *encoder);
 	void (*enc_restore)(struct drm_encoder *encoder);
 	void (*update)(struct nouveau_encoder *, u8 head,
-		       struct drm_display_mode *, u8 proto, u8 depth);
+		       struct nv50_head_atom *, u8 proto, u8 depth);
 };
 
 struct nouveau_encoder *
diff --git a/drivers/gpu/drm/nouveau/nouveau_fbcon.c b/drivers/gpu/drm/nouveau/nouveau_fbcon.c
index 85c1f10bc2b6..844498c4267c 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fbcon.c
+++ b/drivers/gpu/drm/nouveau/nouveau_fbcon.c
@@ -429,7 +429,7 @@ nouveau_fbcon_destroy(struct drm_device *dev, struct nouveau_fbdev *fbcon)
 		nouveau_vma_del(&nouveau_fb->vma);
 		nouveau_bo_unmap(nouveau_fb->nvbo);
 		nouveau_bo_unpin(nouveau_fb->nvbo);
-		drm_framebuffer_unreference(&nouveau_fb->base);
+		drm_framebuffer_put(&nouveau_fb->base);
 	}
 
 	return 0;
diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.c b/drivers/gpu/drm/nouveau/nouveau_fence.c
index 503fa94dc06d..412d49bc6e56 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fence.c
+++ b/drivers/gpu/drm/nouveau/nouveau_fence.c
@@ -74,15 +74,14 @@ nouveau_fence_signal(struct nouveau_fence *fence)
 }
 
 static struct nouveau_fence *
-nouveau_local_fence(struct dma_fence *fence, struct nouveau_drm *drm) {
-	struct nouveau_fence_priv *priv = (void*)drm->fence;
-
+nouveau_local_fence(struct dma_fence *fence, struct nouveau_drm *drm)
+{
 	if (fence->ops != &nouveau_fence_ops_legacy &&
 	    fence->ops != &nouveau_fence_ops_uevent)
 		return NULL;
 
-	if (fence->context < priv->context_base ||
-	    fence->context >= priv->context_base + priv->contexts)
+	if (fence->context < drm->chan.context_base ||
+	    fence->context >= drm->chan.context_base + drm->chan.nr)
 		return NULL;
 
 	return from_fence(fence);
@@ -176,7 +175,7 @@ nouveau_fence_context_new(struct nouveau_channel *chan, struct nouveau_fence_cha
 	INIT_LIST_HEAD(&fctx->flip);
 	INIT_LIST_HEAD(&fctx->pending);
 	spin_lock_init(&fctx->lock);
-	fctx->context = priv->context_base + chan->chid;
+	fctx->context = chan->drm->chan.context_base + chan->chid;
 
 	if (chan == chan->drm->cechan)
 		strcpy(fctx->name, "copy engine channel");
diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.h b/drivers/gpu/drm/nouveau/nouveau_fence.h
index 5bd8d30d1657..b999e6058046 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fence.h
+++ b/drivers/gpu/drm/nouveau/nouveau_fence.h
@@ -55,8 +55,6 @@ struct nouveau_fence_priv {
 	int  (*context_new)(struct nouveau_channel *);
 	void (*context_del)(struct nouveau_channel *);
 
-	u32 contexts;
-	u64 context_base;
 	bool uevent;
 };
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c
index e72a7e37eb0a..df73bec354e8 100644
--- a/drivers/gpu/drm/nouveau/nouveau_gem.c
+++ b/drivers/gpu/drm/nouveau/nouveau_gem.c
@@ -99,6 +99,7 @@ struct nouveau_gem_object_unmap {
 static void
 nouveau_gem_object_delete(struct nouveau_vma *vma)
 {
+	nouveau_fence_unref(&vma->fence);
 	nouveau_vma_del(&vma);
 }
 
@@ -114,25 +115,12 @@ nouveau_gem_object_delete_work(struct nouveau_cli_work *w)
 static void
 nouveau_gem_object_unmap(struct nouveau_bo *nvbo, struct nouveau_vma *vma)
 {
-	const bool mapped = nvbo->bo.mem.mem_type != TTM_PL_SYSTEM;
-	struct reservation_object *resv = nvbo->bo.resv;
-	struct reservation_object_list *fobj;
+	struct dma_fence *fence = vma->fence ? &vma->fence->base : NULL;
 	struct nouveau_gem_object_unmap *work;
-	struct dma_fence *fence = NULL;
-
-	fobj = reservation_object_get_list(resv);
 
 	list_del_init(&vma->head);
 
-	if (fobj && fobj->shared_count > 1)
-		ttm_bo_wait(&nvbo->bo, false, false);
-	else if (fobj && fobj->shared_count == 1)
-		fence = rcu_dereference_protected(fobj->shared[0],
-						reservation_object_held(resv));
-	else
-		fence = reservation_object_get_excl(nvbo->bo.resv);
-
-	if (!fence || !mapped) {
+	if (!fence) {
 		nouveau_gem_object_delete(vma);
 		return;
 	}
@@ -286,7 +274,7 @@ nouveau_gem_ioctl_new(struct drm_device *dev, void *data,
 	}
 
 	/* drop reference from allocate - handle holds it now */
-	drm_gem_object_unreference_unlocked(&nvbo->gem);
+	drm_gem_object_put_unlocked(&nvbo->gem);
 	return ret;
 }
 
@@ -344,9 +332,20 @@ validate_fini_no_ticket(struct validate_op *op, struct nouveau_fence *fence,
 		nvbo = list_entry(op->list.next, struct nouveau_bo, entry);
 		b = &pbbo[nvbo->pbbo_index];
 
-		if (likely(fence))
+		if (likely(fence)) {
+			struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev);
+			struct nouveau_vma *vma;
+
 			nouveau_bo_fence(nvbo, fence, !!b->write_domains);
 
+			if (drm->client.vmm.vmm.object.oclass >= NVIF_CLASS_VMM_NV50) {
+				vma = (void *)(unsigned long)b->user_priv;
+				nouveau_fence_unref(&vma->fence);
+				dma_fence_get(&fence->base);
+				vma->fence = fence;
+			}
+		}
+
 		if (unlikely(nvbo->validate_mapped)) {
 			ttm_bo_kunmap(&nvbo->kmap);
 			nvbo->validate_mapped = false;
@@ -355,7 +354,7 @@ validate_fini_no_ticket(struct validate_op *op, struct nouveau_fence *fence,
 		list_del(&nvbo->entry);
 		nvbo->reserved_by = NULL;
 		ttm_bo_unreserve(&nvbo->bo);
-		drm_gem_object_unreference_unlocked(&nvbo->gem);
+		drm_gem_object_put_unlocked(&nvbo->gem);
 	}
 }
 
@@ -401,14 +400,14 @@ retry:
 		nvbo = nouveau_gem_object(gem);
 		if (nvbo == res_bo) {
 			res_bo = NULL;
-			drm_gem_object_unreference_unlocked(gem);
+			drm_gem_object_put_unlocked(gem);
 			continue;
 		}
 
 		if (nvbo->reserved_by && nvbo->reserved_by == file_priv) {
 			NV_PRINTK(err, cli, "multiple instances of buffer %d on "
 				      "validation list\n", b->handle);
-			drm_gem_object_unreference_unlocked(gem);
+			drm_gem_object_put_unlocked(gem);
 			ret = -EINVAL;
 			break;
 		}
@@ -432,7 +431,20 @@ retry:
 			}
 		}
 
-		b->user_priv = (uint64_t)(unsigned long)nvbo;
+		if (cli->vmm.vmm.object.oclass >= NVIF_CLASS_VMM_NV50) {
+			struct nouveau_vmm *vmm = &cli->vmm;
+			struct nouveau_vma *vma = nouveau_vma_find(nvbo, vmm);
+			if (!vma) {
+				NV_PRINTK(err, cli, "vma not found!\n");
+				ret = -EINVAL;
+				break;
+			}
+
+			b->user_priv = (uint64_t)(unsigned long)vma;
+		} else {
+			b->user_priv = (uint64_t)(unsigned long)nvbo;
+		}
+
 		nvbo->reserved_by = file_priv;
 		nvbo->pbbo_index = i;
 		if ((b->valid_domains & NOUVEAU_GEM_DOMAIN_VRAM) &&
@@ -763,10 +775,10 @@ nouveau_gem_ioctl_pushbuf(struct drm_device *dev, void *data,
 		}
 
 		for (i = 0; i < req->nr_push; i++) {
-			struct nouveau_bo *nvbo = (void *)(unsigned long)
+			struct nouveau_vma *vma = (void *)(unsigned long)
 				bo[push[i].bo_index].user_priv;
 
-			nv50_dma_push(chan, nvbo, push[i].offset,
+			nv50_dma_push(chan, vma->addr + push[i].offset,
 				      push[i].length);
 		}
 	} else
@@ -882,7 +894,7 @@ nouveau_gem_ioctl_cpu_prep(struct drm_device *dev, void *data,
 		ret = lret;
 
 	nouveau_bo_sync_for_cpu(nvbo);
-	drm_gem_object_unreference_unlocked(gem);
+	drm_gem_object_put_unlocked(gem);
 
 	return ret;
 }
@@ -901,7 +913,7 @@ nouveau_gem_ioctl_cpu_fini(struct drm_device *dev, void *data,
 	nvbo = nouveau_gem_object(gem);
 
 	nouveau_bo_sync_for_device(nvbo);
-	drm_gem_object_unreference_unlocked(gem);
+	drm_gem_object_put_unlocked(gem);
 	return 0;
 }
 
@@ -918,7 +930,7 @@ nouveau_gem_ioctl_info(struct drm_device *dev, void *data,
 		return -ENOENT;
 
 	ret = nouveau_gem_info(file_priv, gem, req);
-	drm_gem_object_unreference_unlocked(gem);
+	drm_gem_object_put_unlocked(gem);
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_hwmon.c b/drivers/gpu/drm/nouveau/nouveau_hwmon.c
index 7c965648df80..08a1ab6b150d 100644
--- a/drivers/gpu/drm/nouveau/nouveau_hwmon.c
+++ b/drivers/gpu/drm/nouveau/nouveau_hwmon.c
@@ -69,8 +69,8 @@ nouveau_hwmon_set_temp1_auto_point1_temp(struct device *d,
 	struct nvkm_therm *therm = nvxx_therm(&drm->client.device);
 	long value;
 
-	if (kstrtol(buf, 10, &value) == -EINVAL)
-		return count;
+	if (kstrtol(buf, 10, &value))
+		return -EINVAL;
 
 	therm->attr_set(therm, NVKM_THERM_ATTR_THRS_FAN_BOOST,
 			value / 1000);
@@ -102,8 +102,8 @@ nouveau_hwmon_set_temp1_auto_point1_temp_hyst(struct device *d,
 	struct nvkm_therm *therm = nvxx_therm(&drm->client.device);
 	long value;
 
-	if (kstrtol(buf, 10, &value) == -EINVAL)
-		return count;
+	if (kstrtol(buf, 10, &value))
+		return -EINVAL;
 
 	therm->attr_set(therm, NVKM_THERM_ATTR_THRS_FAN_BOOST_HYST,
 			value / 1000);
@@ -156,7 +156,7 @@ nouveau_hwmon_set_pwm1_min(struct device *d, struct device_attribute *a,
 	long value;
 	int ret;
 
-	if (kstrtol(buf, 10, &value) == -EINVAL)
+	if (kstrtol(buf, 10, &value))
 		return -EINVAL;
 
 	ret = therm->attr_set(therm, NVKM_THERM_ATTR_FAN_MIN_DUTY, value);
@@ -179,7 +179,7 @@ nouveau_hwmon_set_pwm1_max(struct device *d, struct device_attribute *a,
 	long value;
 	int ret;
 
-	if (kstrtol(buf, 10, &value) == -EINVAL)
+	if (kstrtol(buf, 10, &value))
 		return -EINVAL;
 
 	ret = therm->attr_set(therm, NVKM_THERM_ATTR_FAN_MAX_DUTY, value);
@@ -327,7 +327,7 @@ nouveau_temp_is_visible(const void *data, u32 attr, int channel)
 	struct nouveau_drm *drm = nouveau_drm((struct drm_device *)data);
 	struct nvkm_therm *therm = nvxx_therm(&drm->client.device);
 
-	if (therm && therm->attr_get && nvkm_therm_temp_get(therm) < 0)
+	if (!therm || !therm->attr_get || nvkm_therm_temp_get(therm) < 0)
 		return 0;
 
 	switch (attr) {
@@ -351,8 +351,8 @@ nouveau_pwm_is_visible(const void *data, u32 attr, int channel)
 	struct nouveau_drm *drm = nouveau_drm((struct drm_device *)data);
 	struct nvkm_therm *therm = nvxx_therm(&drm->client.device);
 
-	if (therm && therm->attr_get && therm->fan_get &&
-				therm->fan_get(therm) < 0)
+	if (!therm || !therm->attr_get || !therm->fan_get ||
+	    therm->fan_get(therm) < 0)
 		return 0;
 
 	switch (attr) {
@@ -707,13 +707,20 @@ nouveau_hwmon_init(struct drm_device *dev)
 {
 #if defined(CONFIG_HWMON) || (defined(MODULE) && defined(CONFIG_HWMON_MODULE))
 	struct nouveau_drm *drm = nouveau_drm(dev);
+	struct nvkm_iccsense *iccsense = nvxx_iccsense(&drm->client.device);
 	struct nvkm_therm *therm = nvxx_therm(&drm->client.device);
+	struct nvkm_volt *volt = nvxx_volt(&drm->client.device);
 	const struct attribute_group *special_groups[N_ATTR_GROUPS];
 	struct nouveau_hwmon *hwmon;
 	struct device *hwmon_dev;
 	int ret = 0;
 	int i = 0;
 
+	if (!iccsense && !therm && !volt) {
+		NV_DEBUG(drm, "Skipping hwmon registration\n");
+		return 0;
+	}
+
 	hwmon = drm->hwmon = kzalloc(sizeof(*hwmon), GFP_KERNEL);
 	if (!hwmon)
 		return -ENOMEM;
@@ -749,6 +756,9 @@ nouveau_hwmon_fini(struct drm_device *dev)
 #if defined(CONFIG_HWMON) || (defined(MODULE) && defined(CONFIG_HWMON_MODULE))
 	struct nouveau_hwmon *hwmon = nouveau_hwmon(dev);
 
+	if (!hwmon)
+		return;
+
 	if (hwmon->hwmon)
 		hwmon_device_unregister(hwmon->hwmon);
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_platform.c b/drivers/gpu/drm/nouveau/nouveau_platform.c
index 1ada186fab77..039e23548e08 100644
--- a/drivers/gpu/drm/nouveau/nouveau_platform.c
+++ b/drivers/gpu/drm/nouveau/nouveau_platform.c
@@ -36,7 +36,7 @@ static int nouveau_platform_probe(struct platform_device *pdev)
 
 	ret = drm_dev_register(drm, 0);
 	if (ret < 0) {
-		drm_dev_unref(drm);
+		drm_dev_put(drm);
 		return ret;
 	}
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_ttm.c b/drivers/gpu/drm/nouveau/nouveau_ttm.c
index dff51a0ee028..8edb9f2a4269 100644
--- a/drivers/gpu/drm/nouveau/nouveau_ttm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_ttm.c
@@ -1,8 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /*
  * Copyright (c) 2007-2008 Tungsten Graphics, Inc., Cedar Park, TX., USA,
- * All Rights Reserved.
  * Copyright (c) 2009 VMware, Inc., Palo Alto, CA., USA,
- * All Rights Reserved.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -63,7 +62,7 @@ nouveau_vram_manager_new(struct ttm_mem_type_manager *man,
 			 struct ttm_mem_reg *reg)
 {
 	struct nouveau_bo *nvbo = nouveau_bo(bo);
-	struct nouveau_drm *drm = nvbo->cli->drm;
+	struct nouveau_drm *drm = nouveau_bdev(bo->bdev);
 	struct nouveau_mem *mem;
 	int ret;
 
@@ -103,7 +102,7 @@ nouveau_gart_manager_new(struct ttm_mem_type_manager *man,
 			 struct ttm_mem_reg *reg)
 {
 	struct nouveau_bo *nvbo = nouveau_bo(bo);
-	struct nouveau_drm *drm = nvbo->cli->drm;
+	struct nouveau_drm *drm = nouveau_bdev(bo->bdev);
 	struct nouveau_mem *mem;
 	int ret;
 
@@ -131,7 +130,7 @@ nv04_gart_manager_new(struct ttm_mem_type_manager *man,
 		      struct ttm_mem_reg *reg)
 {
 	struct nouveau_bo *nvbo = nouveau_bo(bo);
-	struct nouveau_drm *drm = nvbo->cli->drm;
+	struct nouveau_drm *drm = nouveau_bdev(bo->bdev);
 	struct nouveau_mem *mem;
 	int ret;
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_vmm.c b/drivers/gpu/drm/nouveau/nouveau_vmm.c
index f5371d96b003..2032c3e4f6e5 100644
--- a/drivers/gpu/drm/nouveau/nouveau_vmm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_vmm.c
@@ -92,6 +92,7 @@ nouveau_vma_new(struct nouveau_bo *nvbo, struct nouveau_vmm *vmm,
 	vma->refs = 1;
 	vma->addr = ~0ULL;
 	vma->mem = NULL;
+	vma->fence = NULL;
 	list_add_tail(&vma->head, &nvbo->vma_list);
 
 	if (nvbo->bo.mem.mem_type != TTM_PL_SYSTEM &&
diff --git a/drivers/gpu/drm/nouveau/nouveau_vmm.h b/drivers/gpu/drm/nouveau/nouveau_vmm.h
index 5c31f43678d3..7e3b118cf7c4 100644
--- a/drivers/gpu/drm/nouveau/nouveau_vmm.h
+++ b/drivers/gpu/drm/nouveau/nouveau_vmm.h
@@ -11,6 +11,8 @@ struct nouveau_vma {
 	u64 addr;
 
 	struct nouveau_mem *mem;
+
+	struct nouveau_fence *fence;
 };
 
 struct nouveau_vma *nouveau_vma_find(struct nouveau_bo *, struct nouveau_vmm *);
diff --git a/drivers/gpu/drm/nouveau/nv04_fence.c b/drivers/gpu/drm/nouveau/nv04_fence.c
index fa8f2375c398..c41e82be4893 100644
--- a/drivers/gpu/drm/nouveau/nv04_fence.c
+++ b/drivers/gpu/drm/nouveau/nv04_fence.c
@@ -109,7 +109,5 @@ nv04_fence_create(struct nouveau_drm *drm)
 	priv->base.dtor = nv04_fence_destroy;
 	priv->base.context_new = nv04_fence_context_new;
 	priv->base.context_del = nv04_fence_context_del;
-	priv->base.contexts = 15;
-	priv->base.context_base = dma_fence_context_alloc(priv->base.contexts);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/nv10_fence.c b/drivers/gpu/drm/nouveau/nv10_fence.c
index 2998bde29211..4476b712dc84 100644
--- a/drivers/gpu/drm/nouveau/nv10_fence.c
+++ b/drivers/gpu/drm/nouveau/nv10_fence.c
@@ -103,8 +103,6 @@ nv10_fence_create(struct nouveau_drm *drm)
 	priv->base.dtor = nv10_fence_destroy;
 	priv->base.context_new = nv10_fence_context_new;
 	priv->base.context_del = nv10_fence_context_del;
-	priv->base.contexts = 31;
-	priv->base.context_base = dma_fence_context_alloc(priv->base.contexts);
 	spin_lock_init(&priv->lock);
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/nv17_fence.c b/drivers/gpu/drm/nouveau/nv17_fence.c
index 6477b7069e14..5d613d43b84d 100644
--- a/drivers/gpu/drm/nouveau/nv17_fence.c
+++ b/drivers/gpu/drm/nouveau/nv17_fence.c
@@ -125,8 +125,6 @@ nv17_fence_create(struct nouveau_drm *drm)
 	priv->base.resume = nv17_fence_resume;
 	priv->base.context_new = nv17_fence_context_new;
 	priv->base.context_del = nv10_fence_context_del;
-	priv->base.contexts = 31;
-	priv->base.context_base = dma_fence_context_alloc(priv->base.contexts);
 	spin_lock_init(&priv->lock);
 
 	ret = nouveau_bo_new(&drm->client, 4096, 0x1000, TTM_PL_FLAG_VRAM,
diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c
deleted file mode 100644
index 8bd739cfd00d..000000000000
--- a/drivers/gpu/drm/nouveau/nv50_display.c
+++ /dev/null
@@ -1,4559 +0,0 @@
-/*
- * Copyright 2011 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs
- */
-
-#include <linux/dma-mapping.h>
-#include <linux/hdmi.h>
-
-#include <drm/drmP.h>
-#include <drm/drm_atomic.h>
-#include <drm/drm_atomic_helper.h>
-#include <drm/drm_crtc_helper.h>
-#include <drm/drm_dp_helper.h>
-#include <drm/drm_fb_helper.h>
-#include <drm/drm_plane_helper.h>
-#include <drm/drm_edid.h>
-
-#include <nvif/class.h>
-#include <nvif/cl0002.h>
-#include <nvif/cl5070.h>
-#include <nvif/cl507a.h>
-#include <nvif/cl507b.h>
-#include <nvif/cl507c.h>
-#include <nvif/cl507d.h>
-#include <nvif/cl507e.h>
-#include <nvif/event.h>
-
-#include "nouveau_drv.h"
-#include "nouveau_dma.h"
-#include "nouveau_gem.h"
-#include "nouveau_connector.h"
-#include "nouveau_encoder.h"
-#include "nouveau_crtc.h"
-#include "nouveau_fence.h"
-#include "nouveau_fbcon.h"
-#include "nv50_display.h"
-
-#define EVO_DMA_NR 9
-
-#define EVO_MASTER  (0x00)
-#define EVO_FLIP(c) (0x01 + (c))
-#define EVO_OVLY(c) (0x05 + (c))
-#define EVO_OIMM(c) (0x09 + (c))
-#define EVO_CURS(c) (0x0d + (c))
-
-/* offsets in shared sync bo of various structures */
-#define EVO_SYNC(c, o) ((c) * 0x0100 + (o))
-#define EVO_MAST_NTFY     EVO_SYNC(      0, 0x00)
-#define EVO_FLIP_SEM0(c)  EVO_SYNC((c) + 1, 0x00)
-#define EVO_FLIP_SEM1(c)  EVO_SYNC((c) + 1, 0x10)
-#define EVO_FLIP_NTFY0(c) EVO_SYNC((c) + 1, 0x20)
-#define EVO_FLIP_NTFY1(c) EVO_SYNC((c) + 1, 0x30)
-
-/******************************************************************************
- * Atomic state
- *****************************************************************************/
-#define nv50_atom(p) container_of((p), struct nv50_atom, state)
-
-struct nv50_atom {
-	struct drm_atomic_state state;
-
-	struct list_head outp;
-	bool lock_core;
-	bool flush_disable;
-};
-
-struct nv50_outp_atom {
-	struct list_head head;
-
-	struct drm_encoder *encoder;
-	bool flush_disable;
-
-	union {
-		struct {
-			bool ctrl:1;
-		};
-		u8 mask;
-	} clr;
-
-	union {
-		struct {
-			bool ctrl:1;
-		};
-		u8 mask;
-	} set;
-};
-
-#define nv50_head_atom(p) container_of((p), struct nv50_head_atom, state)
-
-struct nv50_head_atom {
-	struct drm_crtc_state state;
-
-	struct {
-		u16 iW;
-		u16 iH;
-		u16 oW;
-		u16 oH;
-	} view;
-
-	struct nv50_head_mode {
-		bool interlace;
-		u32 clock;
-		struct {
-			u16 active;
-			u16 synce;
-			u16 blanke;
-			u16 blanks;
-		} h;
-		struct {
-			u32 active;
-			u16 synce;
-			u16 blanke;
-			u16 blanks;
-			u16 blank2s;
-			u16 blank2e;
-			u16 blankus;
-		} v;
-	} mode;
-
-	struct {
-		bool visible;
-		u32 handle;
-		u64 offset:40;
-		u8  mode:4;
-	} lut;
-
-	struct {
-		bool visible;
-		u32 handle;
-		u64 offset:40;
-		u8  format;
-		u8  kind:7;
-		u8  layout:1;
-		u8  block:4;
-		u32 pitch:20;
-		u16 x;
-		u16 y;
-		u16 w;
-		u16 h;
-	} core;
-
-	struct {
-		bool visible;
-		u32 handle;
-		u64 offset:40;
-		u8  layout:1;
-		u8  format:1;
-	} curs;
-
-	struct {
-		u8  depth;
-		u8  cpp;
-		u16 x;
-		u16 y;
-		u16 w;
-		u16 h;
-	} base;
-
-	struct {
-		u8 cpp;
-	} ovly;
-
-	struct {
-		bool enable:1;
-		u8 bits:2;
-		u8 mode:4;
-	} dither;
-
-	struct {
-		struct {
-			u16 cos:12;
-			u16 sin:12;
-		} sat;
-	} procamp;
-
-	union {
-		struct {
-			bool ilut:1;
-			bool core:1;
-			bool curs:1;
-		};
-		u8 mask;
-	} clr;
-
-	union {
-		struct {
-			bool ilut:1;
-			bool core:1;
-			bool curs:1;
-			bool view:1;
-			bool mode:1;
-			bool base:1;
-			bool ovly:1;
-			bool dither:1;
-			bool procamp:1;
-		};
-		u16 mask;
-	} set;
-};
-
-static inline struct nv50_head_atom *
-nv50_head_atom_get(struct drm_atomic_state *state, struct drm_crtc *crtc)
-{
-	struct drm_crtc_state *statec = drm_atomic_get_crtc_state(state, crtc);
-	if (IS_ERR(statec))
-		return (void *)statec;
-	return nv50_head_atom(statec);
-}
-
-#define nv50_wndw_atom(p) container_of((p), struct nv50_wndw_atom, state)
-
-struct nv50_wndw_atom {
-	struct drm_plane_state state;
-	u8 interval;
-
-	struct {
-		u32  handle;
-		u16  offset:12;
-		bool awaken:1;
-	} ntfy;
-
-	struct {
-		u32 handle;
-		u16 offset:12;
-		u32 acquire;
-		u32 release;
-	} sema;
-
-	struct {
-		u8 enable:2;
-	} lut;
-
-	struct {
-		u8  mode:2;
-		u8  interval:4;
-
-		u8  format;
-		u8  kind:7;
-		u8  layout:1;
-		u8  block:4;
-		u32 pitch:20;
-		u16 w;
-		u16 h;
-
-		u32 handle;
-		u64 offset;
-	} image;
-
-	struct {
-		u16 x;
-		u16 y;
-	} point;
-
-	union {
-		struct {
-			bool ntfy:1;
-			bool sema:1;
-			bool image:1;
-		};
-		u8 mask;
-	} clr;
-
-	union {
-		struct {
-			bool ntfy:1;
-			bool sema:1;
-			bool image:1;
-			bool lut:1;
-			bool point:1;
-		};
-		u8 mask;
-	} set;
-};
-
-/******************************************************************************
- * EVO channel
- *****************************************************************************/
-
-struct nv50_chan {
-	struct nvif_object user;
-	struct nvif_device *device;
-};
-
-static int
-nv50_chan_create(struct nvif_device *device, struct nvif_object *disp,
-		 const s32 *oclass, u8 head, void *data, u32 size,
-		 struct nv50_chan *chan)
-{
-	struct nvif_sclass *sclass;
-	int ret, i, n;
-
-	chan->device = device;
-
-	ret = n = nvif_object_sclass_get(disp, &sclass);
-	if (ret < 0)
-		return ret;
-
-	while (oclass[0]) {
-		for (i = 0; i < n; i++) {
-			if (sclass[i].oclass == oclass[0]) {
-				ret = nvif_object_init(disp, 0, oclass[0],
-						       data, size, &chan->user);
-				if (ret == 0)
-					nvif_object_map(&chan->user, NULL, 0);
-				nvif_object_sclass_put(&sclass);
-				return ret;
-			}
-		}
-		oclass++;
-	}
-
-	nvif_object_sclass_put(&sclass);
-	return -ENOSYS;
-}
-
-static void
-nv50_chan_destroy(struct nv50_chan *chan)
-{
-	nvif_object_fini(&chan->user);
-}
-
-/******************************************************************************
- * PIO EVO channel
- *****************************************************************************/
-
-struct nv50_pioc {
-	struct nv50_chan base;
-};
-
-static void
-nv50_pioc_destroy(struct nv50_pioc *pioc)
-{
-	nv50_chan_destroy(&pioc->base);
-}
-
-static int
-nv50_pioc_create(struct nvif_device *device, struct nvif_object *disp,
-		 const s32 *oclass, u8 head, void *data, u32 size,
-		 struct nv50_pioc *pioc)
-{
-	return nv50_chan_create(device, disp, oclass, head, data, size,
-				&pioc->base);
-}
-
-/******************************************************************************
- * Overlay Immediate
- *****************************************************************************/
-
-struct nv50_oimm {
-	struct nv50_pioc base;
-};
-
-static int
-nv50_oimm_create(struct nvif_device *device, struct nvif_object *disp,
-		 int head, struct nv50_oimm *oimm)
-{
-	struct nv50_disp_cursor_v0 args = {
-		.head = head,
-	};
-	static const s32 oclass[] = {
-		GK104_DISP_OVERLAY,
-		GF110_DISP_OVERLAY,
-		GT214_DISP_OVERLAY,
-		G82_DISP_OVERLAY,
-		NV50_DISP_OVERLAY,
-		0
-	};
-
-	return nv50_pioc_create(device, disp, oclass, head, &args, sizeof(args),
-				&oimm->base);
-}
-
-/******************************************************************************
- * DMA EVO channel
- *****************************************************************************/
-
-struct nv50_dmac_ctxdma {
-	struct list_head head;
-	struct nvif_object object;
-};
-
-struct nv50_dmac {
-	struct nv50_chan base;
-	dma_addr_t handle;
-	u32 *ptr;
-
-	struct nvif_object sync;
-	struct nvif_object vram;
-	struct list_head ctxdma;
-
-	/* Protects against concurrent pushbuf access to this channel, lock is
-	 * grabbed by evo_wait (if the pushbuf reservation is successful) and
-	 * dropped again by evo_kick. */
-	struct mutex lock;
-};
-
-static void
-nv50_dmac_ctxdma_del(struct nv50_dmac_ctxdma *ctxdma)
-{
-	nvif_object_fini(&ctxdma->object);
-	list_del(&ctxdma->head);
-	kfree(ctxdma);
-}
-
-static struct nv50_dmac_ctxdma *
-nv50_dmac_ctxdma_new(struct nv50_dmac *dmac, struct nouveau_framebuffer *fb)
-{
-	struct nouveau_drm *drm = nouveau_drm(fb->base.dev);
-	struct nv50_dmac_ctxdma *ctxdma;
-	const u8    kind = fb->nvbo->kind;
-	const u32 handle = 0xfb000000 | kind;
-	struct {
-		struct nv_dma_v0 base;
-		union {
-			struct nv50_dma_v0 nv50;
-			struct gf100_dma_v0 gf100;
-			struct gf119_dma_v0 gf119;
-		};
-	} args = {};
-	u32 argc = sizeof(args.base);
-	int ret;
-
-	list_for_each_entry(ctxdma, &dmac->ctxdma, head) {
-		if (ctxdma->object.handle == handle)
-			return ctxdma;
-	}
-
-	if (!(ctxdma = kzalloc(sizeof(*ctxdma), GFP_KERNEL)))
-		return ERR_PTR(-ENOMEM);
-	list_add(&ctxdma->head, &dmac->ctxdma);
-
-	args.base.target = NV_DMA_V0_TARGET_VRAM;
-	args.base.access = NV_DMA_V0_ACCESS_RDWR;
-	args.base.start  = 0;
-	args.base.limit  = drm->client.device.info.ram_user - 1;
-
-	if (drm->client.device.info.chipset < 0x80) {
-		args.nv50.part = NV50_DMA_V0_PART_256;
-		argc += sizeof(args.nv50);
-	} else
-	if (drm->client.device.info.chipset < 0xc0) {
-		args.nv50.part = NV50_DMA_V0_PART_256;
-		args.nv50.kind = kind;
-		argc += sizeof(args.nv50);
-	} else
-	if (drm->client.device.info.chipset < 0xd0) {
-		args.gf100.kind = kind;
-		argc += sizeof(args.gf100);
-	} else {
-		args.gf119.page = GF119_DMA_V0_PAGE_LP;
-		args.gf119.kind = kind;
-		argc += sizeof(args.gf119);
-	}
-
-	ret = nvif_object_init(&dmac->base.user, handle, NV_DMA_IN_MEMORY,
-			       &args, argc, &ctxdma->object);
-	if (ret) {
-		nv50_dmac_ctxdma_del(ctxdma);
-		return ERR_PTR(ret);
-	}
-
-	return ctxdma;
-}
-
-static void
-nv50_dmac_destroy(struct nv50_dmac *dmac, struct nvif_object *disp)
-{
-	struct nvif_device *device = dmac->base.device;
-	struct nv50_dmac_ctxdma *ctxdma, *ctxtmp;
-
-	list_for_each_entry_safe(ctxdma, ctxtmp, &dmac->ctxdma, head) {
-		nv50_dmac_ctxdma_del(ctxdma);
-	}
-
-	nvif_object_fini(&dmac->vram);
-	nvif_object_fini(&dmac->sync);
-
-	nv50_chan_destroy(&dmac->base);
-
-	if (dmac->ptr) {
-		struct device *dev = nvxx_device(device)->dev;
-		dma_free_coherent(dev, PAGE_SIZE, dmac->ptr, dmac->handle);
-	}
-}
-
-static int
-nv50_dmac_create(struct nvif_device *device, struct nvif_object *disp,
-		 const s32 *oclass, u8 head, void *data, u32 size, u64 syncbuf,
-		 struct nv50_dmac *dmac)
-{
-	struct nv50_disp_core_channel_dma_v0 *args = data;
-	struct nvif_object pushbuf;
-	int ret;
-
-	mutex_init(&dmac->lock);
-	INIT_LIST_HEAD(&dmac->ctxdma);
-
-	dmac->ptr = dma_alloc_coherent(nvxx_device(device)->dev, PAGE_SIZE,
-				       &dmac->handle, GFP_KERNEL);
-	if (!dmac->ptr)
-		return -ENOMEM;
-
-	ret = nvif_object_init(&device->object, 0, NV_DMA_FROM_MEMORY,
-			       &(struct nv_dma_v0) {
-					.target = NV_DMA_V0_TARGET_PCI_US,
-					.access = NV_DMA_V0_ACCESS_RD,
-					.start = dmac->handle + 0x0000,
-					.limit = dmac->handle + 0x0fff,
-			       }, sizeof(struct nv_dma_v0), &pushbuf);
-	if (ret)
-		return ret;
-
-	args->pushbuf = nvif_handle(&pushbuf);
-
-	ret = nv50_chan_create(device, disp, oclass, head, data, size,
-			       &dmac->base);
-	nvif_object_fini(&pushbuf);
-	if (ret)
-		return ret;
-
-	ret = nvif_object_init(&dmac->base.user, 0xf0000000, NV_DMA_IN_MEMORY,
-			       &(struct nv_dma_v0) {
-					.target = NV_DMA_V0_TARGET_VRAM,
-					.access = NV_DMA_V0_ACCESS_RDWR,
-					.start = syncbuf + 0x0000,
-					.limit = syncbuf + 0x0fff,
-			       }, sizeof(struct nv_dma_v0),
-			       &dmac->sync);
-	if (ret)
-		return ret;
-
-	ret = nvif_object_init(&dmac->base.user, 0xf0000001, NV_DMA_IN_MEMORY,
-			       &(struct nv_dma_v0) {
-					.target = NV_DMA_V0_TARGET_VRAM,
-					.access = NV_DMA_V0_ACCESS_RDWR,
-					.start = 0,
-					.limit = device->info.ram_user - 1,
-			       }, sizeof(struct nv_dma_v0),
-			       &dmac->vram);
-	if (ret)
-		return ret;
-
-	return ret;
-}
-
-/******************************************************************************
- * Core
- *****************************************************************************/
-
-struct nv50_mast {
-	struct nv50_dmac base;
-};
-
-static int
-nv50_core_create(struct nvif_device *device, struct nvif_object *disp,
-		 u64 syncbuf, struct nv50_mast *core)
-{
-	struct nv50_disp_core_channel_dma_v0 args = {
-		.pushbuf = 0xb0007d00,
-	};
-	static const s32 oclass[] = {
-		GP102_DISP_CORE_CHANNEL_DMA,
-		GP100_DISP_CORE_CHANNEL_DMA,
-		GM200_DISP_CORE_CHANNEL_DMA,
-		GM107_DISP_CORE_CHANNEL_DMA,
-		GK110_DISP_CORE_CHANNEL_DMA,
-		GK104_DISP_CORE_CHANNEL_DMA,
-		GF110_DISP_CORE_CHANNEL_DMA,
-		GT214_DISP_CORE_CHANNEL_DMA,
-		GT206_DISP_CORE_CHANNEL_DMA,
-		GT200_DISP_CORE_CHANNEL_DMA,
-		G82_DISP_CORE_CHANNEL_DMA,
-		NV50_DISP_CORE_CHANNEL_DMA,
-		0
-	};
-
-	return nv50_dmac_create(device, disp, oclass, 0, &args, sizeof(args),
-				syncbuf, &core->base);
-}
-
-/******************************************************************************
- * Base
- *****************************************************************************/
-
-struct nv50_sync {
-	struct nv50_dmac base;
-	u32 addr;
-	u32 data;
-};
-
-static int
-nv50_base_create(struct nvif_device *device, struct nvif_object *disp,
-		 int head, u64 syncbuf, struct nv50_sync *base)
-{
-	struct nv50_disp_base_channel_dma_v0 args = {
-		.pushbuf = 0xb0007c00 | head,
-		.head = head,
-	};
-	static const s32 oclass[] = {
-		GK110_DISP_BASE_CHANNEL_DMA,
-		GK104_DISP_BASE_CHANNEL_DMA,
-		GF110_DISP_BASE_CHANNEL_DMA,
-		GT214_DISP_BASE_CHANNEL_DMA,
-		GT200_DISP_BASE_CHANNEL_DMA,
-		G82_DISP_BASE_CHANNEL_DMA,
-		NV50_DISP_BASE_CHANNEL_DMA,
-		0
-	};
-
-	return nv50_dmac_create(device, disp, oclass, head, &args, sizeof(args),
-				syncbuf, &base->base);
-}
-
-/******************************************************************************
- * Overlay
- *****************************************************************************/
-
-struct nv50_ovly {
-	struct nv50_dmac base;
-};
-
-static int
-nv50_ovly_create(struct nvif_device *device, struct nvif_object *disp,
-		 int head, u64 syncbuf, struct nv50_ovly *ovly)
-{
-	struct nv50_disp_overlay_channel_dma_v0 args = {
-		.pushbuf = 0xb0007e00 | head,
-		.head = head,
-	};
-	static const s32 oclass[] = {
-		GK104_DISP_OVERLAY_CONTROL_DMA,
-		GF110_DISP_OVERLAY_CONTROL_DMA,
-		GT214_DISP_OVERLAY_CHANNEL_DMA,
-		GT200_DISP_OVERLAY_CHANNEL_DMA,
-		G82_DISP_OVERLAY_CHANNEL_DMA,
-		NV50_DISP_OVERLAY_CHANNEL_DMA,
-		0
-	};
-
-	return nv50_dmac_create(device, disp, oclass, head, &args, sizeof(args),
-				syncbuf, &ovly->base);
-}
-
-struct nv50_head {
-	struct nouveau_crtc base;
-	struct {
-		struct nouveau_bo *nvbo[2];
-		int next;
-	} lut;
-	struct nv50_ovly ovly;
-	struct nv50_oimm oimm;
-};
-
-#define nv50_head(c) ((struct nv50_head *)nouveau_crtc(c))
-#define nv50_ovly(c) (&nv50_head(c)->ovly)
-#define nv50_oimm(c) (&nv50_head(c)->oimm)
-#define nv50_chan(c) (&(c)->base.base)
-#define nv50_vers(c) nv50_chan(c)->user.oclass
-
-struct nv50_disp {
-	struct nvif_object *disp;
-	struct nv50_mast mast;
-
-	struct nouveau_bo *sync;
-
-	struct mutex mutex;
-};
-
-static struct nv50_disp *
-nv50_disp(struct drm_device *dev)
-{
-	return nouveau_display(dev)->priv;
-}
-
-#define nv50_mast(d) (&nv50_disp(d)->mast)
-
-/******************************************************************************
- * EVO channel helpers
- *****************************************************************************/
-static u32 *
-evo_wait(void *evoc, int nr)
-{
-	struct nv50_dmac *dmac = evoc;
-	struct nvif_device *device = dmac->base.device;
-	u32 put = nvif_rd32(&dmac->base.user, 0x0000) / 4;
-
-	mutex_lock(&dmac->lock);
-	if (put + nr >= (PAGE_SIZE / 4) - 8) {
-		dmac->ptr[put] = 0x20000000;
-
-		nvif_wr32(&dmac->base.user, 0x0000, 0x00000000);
-		if (nvif_msec(device, 2000,
-			if (!nvif_rd32(&dmac->base.user, 0x0004))
-				break;
-		) < 0) {
-			mutex_unlock(&dmac->lock);
-			pr_err("nouveau: evo channel stalled\n");
-			return NULL;
-		}
-
-		put = 0;
-	}
-
-	return dmac->ptr + put;
-}
-
-static void
-evo_kick(u32 *push, void *evoc)
-{
-	struct nv50_dmac *dmac = evoc;
-	nvif_wr32(&dmac->base.user, 0x0000, (push - dmac->ptr) << 2);
-	mutex_unlock(&dmac->lock);
-}
-
-#define evo_mthd(p, m, s) do {						\
-	const u32 _m = (m), _s = (s);					\
-	if (drm_debug & DRM_UT_KMS)					\
-		pr_err("%04x %d %s\n", _m, _s, __func__);		\
-	*((p)++) = ((_s << 18) | _m);					\
-} while(0)
-
-#define evo_data(p, d) do {						\
-	const u32 _d = (d);						\
-	if (drm_debug & DRM_UT_KMS)					\
-		pr_err("\t%08x\n", _d);					\
-	*((p)++) = _d;							\
-} while(0)
-
-/******************************************************************************
- * Plane
- *****************************************************************************/
-#define nv50_wndw(p) container_of((p), struct nv50_wndw, plane)
-
-struct nv50_wndw {
-	const struct nv50_wndw_func *func;
-	struct nv50_dmac *dmac;
-
-	struct drm_plane plane;
-
-	struct nvif_notify notify;
-	u16 ntfy;
-	u16 sema;
-	u32 data;
-};
-
-struct nv50_wndw_func {
-	void *(*dtor)(struct nv50_wndw *);
-	int (*acquire)(struct nv50_wndw *, struct nv50_wndw_atom *asyw,
-		       struct nv50_head_atom *asyh);
-	void (*release)(struct nv50_wndw *, struct nv50_wndw_atom *asyw,
-			struct nv50_head_atom *asyh);
-	void (*prepare)(struct nv50_wndw *, struct nv50_head_atom *asyh,
-			struct nv50_wndw_atom *asyw);
-
-	void (*sema_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*sema_clr)(struct nv50_wndw *);
-	void (*ntfy_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*ntfy_clr)(struct nv50_wndw *);
-	int (*ntfy_wait_begun)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*image_set)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*image_clr)(struct nv50_wndw *);
-	void (*lut)(struct nv50_wndw *, struct nv50_wndw_atom *);
-	void (*point)(struct nv50_wndw *, struct nv50_wndw_atom *);
-
-	u32 (*update)(struct nv50_wndw *, u32 interlock);
-};
-
-static int
-nv50_wndw_wait_armed(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	if (asyw->set.ntfy)
-		return wndw->func->ntfy_wait_begun(wndw, asyw);
-	return 0;
-}
-
-static u32
-nv50_wndw_flush_clr(struct nv50_wndw *wndw, u32 interlock, bool flush,
-		    struct nv50_wndw_atom *asyw)
-{
-	if (asyw->clr.sema && (!asyw->set.sema || flush))
-		wndw->func->sema_clr(wndw);
-	if (asyw->clr.ntfy && (!asyw->set.ntfy || flush))
-		wndw->func->ntfy_clr(wndw);
-	if (asyw->clr.image && (!asyw->set.image || flush))
-		wndw->func->image_clr(wndw);
-
-	return flush ? wndw->func->update(wndw, interlock) : 0;
-}
-
-static u32
-nv50_wndw_flush_set(struct nv50_wndw *wndw, u32 interlock,
-		    struct nv50_wndw_atom *asyw)
-{
-	if (interlock) {
-		asyw->image.mode = 0;
-		asyw->image.interval = 1;
-	}
-
-	if (asyw->set.sema ) wndw->func->sema_set (wndw, asyw);
-	if (asyw->set.ntfy ) wndw->func->ntfy_set (wndw, asyw);
-	if (asyw->set.image) wndw->func->image_set(wndw, asyw);
-	if (asyw->set.lut  ) wndw->func->lut      (wndw, asyw);
-	if (asyw->set.point) wndw->func->point    (wndw, asyw);
-
-	return wndw->func->update(wndw, interlock);
-}
-
-static void
-nv50_wndw_atomic_check_release(struct nv50_wndw *wndw,
-			       struct nv50_wndw_atom *asyw,
-			       struct nv50_head_atom *asyh)
-{
-	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
-	NV_ATOMIC(drm, "%s release\n", wndw->plane.name);
-	wndw->func->release(wndw, asyw, asyh);
-	asyw->ntfy.handle = 0;
-	asyw->sema.handle = 0;
-}
-
-static int
-nv50_wndw_atomic_check_acquire(struct nv50_wndw *wndw,
-			       struct nv50_wndw_atom *asyw,
-			       struct nv50_head_atom *asyh)
-{
-	struct nouveau_framebuffer *fb = nouveau_framebuffer(asyw->state.fb);
-	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
-	int ret;
-
-	NV_ATOMIC(drm, "%s acquire\n", wndw->plane.name);
-
-	asyw->image.w = fb->base.width;
-	asyw->image.h = fb->base.height;
-	asyw->image.kind = fb->nvbo->kind;
-
-	if (asyh->state.pageflip_flags & DRM_MODE_PAGE_FLIP_ASYNC)
-		asyw->interval = 0;
-	else
-		asyw->interval = 1;
-
-	if (asyw->image.kind) {
-		asyw->image.layout = 0;
-		if (drm->client.device.info.chipset >= 0xc0)
-			asyw->image.block = fb->nvbo->mode >> 4;
-		else
-			asyw->image.block = fb->nvbo->mode;
-		asyw->image.pitch = (fb->base.pitches[0] / 4) << 4;
-	} else {
-		asyw->image.layout = 1;
-		asyw->image.block  = 0;
-		asyw->image.pitch  = fb->base.pitches[0];
-	}
-
-	ret = wndw->func->acquire(wndw, asyw, asyh);
-	if (ret)
-		return ret;
-
-	if (asyw->set.image) {
-		if (!(asyw->image.mode = asyw->interval ? 0 : 1))
-			asyw->image.interval = asyw->interval;
-		else
-			asyw->image.interval = 0;
-	}
-
-	return 0;
-}
-
-static int
-nv50_wndw_atomic_check(struct drm_plane *plane, struct drm_plane_state *state)
-{
-	struct nouveau_drm *drm = nouveau_drm(plane->dev);
-	struct nv50_wndw *wndw = nv50_wndw(plane);
-	struct nv50_wndw_atom *armw = nv50_wndw_atom(wndw->plane.state);
-	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
-	struct nv50_head_atom *harm = NULL, *asyh = NULL;
-	bool varm = false, asyv = false, asym = false;
-	int ret;
-
-	NV_ATOMIC(drm, "%s atomic_check\n", plane->name);
-	if (asyw->state.crtc) {
-		asyh = nv50_head_atom_get(asyw->state.state, asyw->state.crtc);
-		if (IS_ERR(asyh))
-			return PTR_ERR(asyh);
-		asym = drm_atomic_crtc_needs_modeset(&asyh->state);
-		asyv = asyh->state.active;
-	}
-
-	if (armw->state.crtc) {
-		harm = nv50_head_atom_get(asyw->state.state, armw->state.crtc);
-		if (IS_ERR(harm))
-			return PTR_ERR(harm);
-		varm = harm->state.crtc->state->active;
-	}
-
-	if (asyv) {
-		asyw->point.x = asyw->state.crtc_x;
-		asyw->point.y = asyw->state.crtc_y;
-		if (memcmp(&armw->point, &asyw->point, sizeof(asyw->point)))
-			asyw->set.point = true;
-
-		ret = nv50_wndw_atomic_check_acquire(wndw, asyw, asyh);
-		if (ret)
-			return ret;
-	} else
-	if (varm) {
-		nv50_wndw_atomic_check_release(wndw, asyw, harm);
-	} else {
-		return 0;
-	}
-
-	if (!asyv || asym) {
-		asyw->clr.ntfy = armw->ntfy.handle != 0;
-		asyw->clr.sema = armw->sema.handle != 0;
-		if (wndw->func->image_clr)
-			asyw->clr.image = armw->image.handle != 0;
-		asyw->set.lut = wndw->func->lut && asyv;
-	}
-
-	return 0;
-}
-
-static void
-nv50_wndw_cleanup_fb(struct drm_plane *plane, struct drm_plane_state *old_state)
-{
-	struct nouveau_framebuffer *fb = nouveau_framebuffer(old_state->fb);
-	struct nouveau_drm *drm = nouveau_drm(plane->dev);
-
-	NV_ATOMIC(drm, "%s cleanup: %p\n", plane->name, old_state->fb);
-	if (!old_state->fb)
-		return;
-
-	nouveau_bo_unpin(fb->nvbo);
-}
-
-static int
-nv50_wndw_prepare_fb(struct drm_plane *plane, struct drm_plane_state *state)
-{
-	struct nouveau_framebuffer *fb = nouveau_framebuffer(state->fb);
-	struct nouveau_drm *drm = nouveau_drm(plane->dev);
-	struct nv50_wndw *wndw = nv50_wndw(plane);
-	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
-	struct nv50_head_atom *asyh;
-	struct nv50_dmac_ctxdma *ctxdma;
-	int ret;
-
-	NV_ATOMIC(drm, "%s prepare: %p\n", plane->name, state->fb);
-	if (!asyw->state.fb)
-		return 0;
-
-	ret = nouveau_bo_pin(fb->nvbo, TTM_PL_FLAG_VRAM, true);
-	if (ret)
-		return ret;
-
-	ctxdma = nv50_dmac_ctxdma_new(wndw->dmac, fb);
-	if (IS_ERR(ctxdma)) {
-		nouveau_bo_unpin(fb->nvbo);
-		return PTR_ERR(ctxdma);
-	}
-
-	asyw->state.fence = reservation_object_get_excl_rcu(fb->nvbo->bo.resv);
-	asyw->image.handle = ctxdma->object.handle;
-	asyw->image.offset = fb->nvbo->bo.offset;
-
-	if (wndw->func->prepare) {
-		asyh = nv50_head_atom_get(asyw->state.state, asyw->state.crtc);
-		if (IS_ERR(asyh))
-			return PTR_ERR(asyh);
-
-		wndw->func->prepare(wndw, asyh, asyw);
-	}
-
-	return 0;
-}
-
-static const struct drm_plane_helper_funcs
-nv50_wndw_helper = {
-	.prepare_fb = nv50_wndw_prepare_fb,
-	.cleanup_fb = nv50_wndw_cleanup_fb,
-	.atomic_check = nv50_wndw_atomic_check,
-};
-
-static void
-nv50_wndw_atomic_destroy_state(struct drm_plane *plane,
-			       struct drm_plane_state *state)
-{
-	struct nv50_wndw_atom *asyw = nv50_wndw_atom(state);
-	__drm_atomic_helper_plane_destroy_state(&asyw->state);
-	kfree(asyw);
-}
-
-static struct drm_plane_state *
-nv50_wndw_atomic_duplicate_state(struct drm_plane *plane)
-{
-	struct nv50_wndw_atom *armw = nv50_wndw_atom(plane->state);
-	struct nv50_wndw_atom *asyw;
-	if (!(asyw = kmalloc(sizeof(*asyw), GFP_KERNEL)))
-		return NULL;
-	__drm_atomic_helper_plane_duplicate_state(plane, &asyw->state);
-	asyw->interval = 1;
-	asyw->sema = armw->sema;
-	asyw->ntfy = armw->ntfy;
-	asyw->image = armw->image;
-	asyw->point = armw->point;
-	asyw->lut = armw->lut;
-	asyw->clr.mask = 0;
-	asyw->set.mask = 0;
-	return &asyw->state;
-}
-
-static void
-nv50_wndw_reset(struct drm_plane *plane)
-{
-	struct nv50_wndw_atom *asyw;
-
-	if (WARN_ON(!(asyw = kzalloc(sizeof(*asyw), GFP_KERNEL))))
-		return;
-
-	if (plane->state)
-		plane->funcs->atomic_destroy_state(plane, plane->state);
-	plane->state = &asyw->state;
-	plane->state->plane = plane;
-	plane->state->rotation = DRM_MODE_ROTATE_0;
-}
-
-static void
-nv50_wndw_destroy(struct drm_plane *plane)
-{
-	struct nv50_wndw *wndw = nv50_wndw(plane);
-	void *data;
-	nvif_notify_fini(&wndw->notify);
-	data = wndw->func->dtor(wndw);
-	drm_plane_cleanup(&wndw->plane);
-	kfree(data);
-}
-
-static const struct drm_plane_funcs
-nv50_wndw = {
-	.update_plane = drm_atomic_helper_update_plane,
-	.disable_plane = drm_atomic_helper_disable_plane,
-	.destroy = nv50_wndw_destroy,
-	.reset = nv50_wndw_reset,
-	.atomic_duplicate_state = nv50_wndw_atomic_duplicate_state,
-	.atomic_destroy_state = nv50_wndw_atomic_destroy_state,
-};
-
-static void
-nv50_wndw_fini(struct nv50_wndw *wndw)
-{
-	nvif_notify_put(&wndw->notify);
-}
-
-static void
-nv50_wndw_init(struct nv50_wndw *wndw)
-{
-	nvif_notify_get(&wndw->notify);
-}
-
-static int
-nv50_wndw_ctor(const struct nv50_wndw_func *func, struct drm_device *dev,
-	       enum drm_plane_type type, const char *name, int index,
-	       struct nv50_dmac *dmac, const u32 *format, int nformat,
-	       struct nv50_wndw *wndw)
-{
-	int ret;
-
-	wndw->func = func;
-	wndw->dmac = dmac;
-
-	ret = drm_universal_plane_init(dev, &wndw->plane, 0, &nv50_wndw,
-				       format, nformat, NULL,
-				       type, "%s-%d", name, index);
-	if (ret)
-		return ret;
-
-	drm_plane_helper_add(&wndw->plane, &nv50_wndw_helper);
-	return 0;
-}
-
-/******************************************************************************
- * Cursor plane
- *****************************************************************************/
-#define nv50_curs(p) container_of((p), struct nv50_curs, wndw)
-
-struct nv50_curs {
-	struct nv50_wndw wndw;
-	struct nvif_object chan;
-};
-
-static u32
-nv50_curs_update(struct nv50_wndw *wndw, u32 interlock)
-{
-	struct nv50_curs *curs = nv50_curs(wndw);
-	nvif_wr32(&curs->chan, 0x0080, 0x00000000);
-	return 0;
-}
-
-static void
-nv50_curs_point(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	struct nv50_curs *curs = nv50_curs(wndw);
-	nvif_wr32(&curs->chan, 0x0084, (asyw->point.y << 16) | asyw->point.x);
-}
-
-static void
-nv50_curs_prepare(struct nv50_wndw *wndw, struct nv50_head_atom *asyh,
-		  struct nv50_wndw_atom *asyw)
-{
-	u32 handle = nv50_disp(wndw->plane.dev)->mast.base.vram.handle;
-	u32 offset = asyw->image.offset;
-	if (asyh->curs.handle != handle || asyh->curs.offset != offset) {
-		asyh->curs.handle = handle;
-		asyh->curs.offset = offset;
-		asyh->set.curs = asyh->curs.visible;
-	}
-}
-
-static void
-nv50_curs_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
-		  struct nv50_head_atom *asyh)
-{
-	asyh->curs.visible = false;
-}
-
-static int
-nv50_curs_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
-		  struct nv50_head_atom *asyh)
-{
-	int ret;
-
-	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
-						  DRM_PLANE_HELPER_NO_SCALING,
-						  DRM_PLANE_HELPER_NO_SCALING,
-						  true, true);
-	asyh->curs.visible = asyw->state.visible;
-	if (ret || !asyh->curs.visible)
-		return ret;
-
-	switch (asyw->state.fb->width) {
-	case 32: asyh->curs.layout = 0; break;
-	case 64: asyh->curs.layout = 1; break;
-	default:
-		return -EINVAL;
-	}
-
-	if (asyw->state.fb->width != asyw->state.fb->height)
-		return -EINVAL;
-
-	switch (asyw->state.fb->format->format) {
-	case DRM_FORMAT_ARGB8888: asyh->curs.format = 1; break;
-	default:
-		WARN_ON(1);
-		return -EINVAL;
-	}
-
-	return 0;
-}
-
-static void *
-nv50_curs_dtor(struct nv50_wndw *wndw)
-{
-	struct nv50_curs *curs = nv50_curs(wndw);
-	nvif_object_fini(&curs->chan);
-	return curs;
-}
-
-static const u32
-nv50_curs_format[] = {
-	DRM_FORMAT_ARGB8888,
-};
-
-static const struct nv50_wndw_func
-nv50_curs = {
-	.dtor = nv50_curs_dtor,
-	.acquire = nv50_curs_acquire,
-	.release = nv50_curs_release,
-	.prepare = nv50_curs_prepare,
-	.point = nv50_curs_point,
-	.update = nv50_curs_update,
-};
-
-static int
-nv50_curs_new(struct nouveau_drm *drm, struct nv50_head *head,
-	      struct nv50_curs **pcurs)
-{
-	static const struct nvif_mclass curses[] = {
-		{ GK104_DISP_CURSOR, 0 },
-		{ GF110_DISP_CURSOR, 0 },
-		{ GT214_DISP_CURSOR, 0 },
-		{   G82_DISP_CURSOR, 0 },
-		{  NV50_DISP_CURSOR, 0 },
-		{}
-	};
-	struct nv50_disp_cursor_v0 args = {
-		.head = head->base.index,
-	};
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct nv50_curs *curs;
-	int cid, ret;
-
-	cid = nvif_mclass(disp->disp, curses);
-	if (cid < 0) {
-		NV_ERROR(drm, "No supported cursor immediate class\n");
-		return cid;
-	}
-
-	if (!(curs = *pcurs = kzalloc(sizeof(*curs), GFP_KERNEL)))
-		return -ENOMEM;
-
-	ret = nv50_wndw_ctor(&nv50_curs, drm->dev, DRM_PLANE_TYPE_CURSOR,
-			     "curs", head->base.index, &disp->mast.base,
-			     nv50_curs_format, ARRAY_SIZE(nv50_curs_format),
-			     &curs->wndw);
-	if (ret) {
-		kfree(curs);
-		return ret;
-	}
-
-	ret = nvif_object_init(disp->disp, 0, curses[cid].oclass, &args,
-			       sizeof(args), &curs->chan);
-	if (ret) {
-		NV_ERROR(drm, "curs%04x allocation failed: %d\n",
-			 curses[cid].oclass, ret);
-		return ret;
-	}
-
-	return 0;
-}
-
-/******************************************************************************
- * Primary plane
- *****************************************************************************/
-#define nv50_base(p) container_of((p), struct nv50_base, wndw)
-
-struct nv50_base {
-	struct nv50_wndw wndw;
-	struct nv50_sync chan;
-	int id;
-};
-
-static int
-nv50_base_notify(struct nvif_notify *notify)
-{
-	return NVIF_NOTIFY_KEEP;
-}
-
-static void
-nv50_base_lut(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	u32 *push;
-	if ((push = evo_wait(&base->chan, 2))) {
-		evo_mthd(push, 0x00e0, 1);
-		evo_data(push, asyw->lut.enable << 30);
-		evo_kick(push, &base->chan);
-	}
-}
-
-static void
-nv50_base_image_clr(struct nv50_wndw *wndw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	u32 *push;
-	if ((push = evo_wait(&base->chan, 4))) {
-		evo_mthd(push, 0x0084, 1);
-		evo_data(push, 0x00000000);
-		evo_mthd(push, 0x00c0, 1);
-		evo_data(push, 0x00000000);
-		evo_kick(push, &base->chan);
-	}
-}
-
-static void
-nv50_base_image_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	const s32 oclass = base->chan.base.base.user.oclass;
-	u32 *push;
-	if ((push = evo_wait(&base->chan, 10))) {
-		evo_mthd(push, 0x0084, 1);
-		evo_data(push, (asyw->image.mode << 8) |
-			       (asyw->image.interval << 4));
-		evo_mthd(push, 0x00c0, 1);
-		evo_data(push, asyw->image.handle);
-		if (oclass < G82_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0800, 5);
-			evo_data(push, asyw->image.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
-			evo_data(push, (asyw->image.layout << 20) |
-					asyw->image.pitch |
-					asyw->image.block);
-			evo_data(push, (asyw->image.kind << 16) |
-				       (asyw->image.format << 8));
-		} else
-		if (oclass < GF110_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0800, 5);
-			evo_data(push, asyw->image.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
-			evo_data(push, (asyw->image.layout << 20) |
-					asyw->image.pitch |
-					asyw->image.block);
-			evo_data(push, asyw->image.format << 8);
-		} else {
-			evo_mthd(push, 0x0400, 5);
-			evo_data(push, asyw->image.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, (asyw->image.h << 16) | asyw->image.w);
-			evo_data(push, (asyw->image.layout << 24) |
-					asyw->image.pitch |
-					asyw->image.block);
-			evo_data(push, asyw->image.format << 8);
-		}
-		evo_kick(push, &base->chan);
-	}
-}
-
-static void
-nv50_base_ntfy_clr(struct nv50_wndw *wndw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	u32 *push;
-	if ((push = evo_wait(&base->chan, 2))) {
-		evo_mthd(push, 0x00a4, 1);
-		evo_data(push, 0x00000000);
-		evo_kick(push, &base->chan);
-	}
-}
-
-static void
-nv50_base_ntfy_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	u32 *push;
-	if ((push = evo_wait(&base->chan, 3))) {
-		evo_mthd(push, 0x00a0, 2);
-		evo_data(push, (asyw->ntfy.awaken << 30) | asyw->ntfy.offset);
-		evo_data(push, asyw->ntfy.handle);
-		evo_kick(push, &base->chan);
-	}
-}
-
-static void
-nv50_base_sema_clr(struct nv50_wndw *wndw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	u32 *push;
-	if ((push = evo_wait(&base->chan, 2))) {
-		evo_mthd(push, 0x0094, 1);
-		evo_data(push, 0x00000000);
-		evo_kick(push, &base->chan);
-	}
-}
-
-static void
-nv50_base_sema_set(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	u32 *push;
-	if ((push = evo_wait(&base->chan, 5))) {
-		evo_mthd(push, 0x0088, 4);
-		evo_data(push, asyw->sema.offset);
-		evo_data(push, asyw->sema.acquire);
-		evo_data(push, asyw->sema.release);
-		evo_data(push, asyw->sema.handle);
-		evo_kick(push, &base->chan);
-	}
-}
-
-static u32
-nv50_base_update(struct nv50_wndw *wndw, u32 interlock)
-{
-	struct nv50_base *base = nv50_base(wndw);
-	u32 *push;
-
-	if (!(push = evo_wait(&base->chan, 2)))
-		return 0;
-	evo_mthd(push, 0x0080, 1);
-	evo_data(push, interlock);
-	evo_kick(push, &base->chan);
-
-	if (base->chan.base.base.user.oclass < GF110_DISP_BASE_CHANNEL_DMA)
-		return interlock ? 2 << (base->id * 8) : 0;
-	return interlock ? 2 << (base->id * 4) : 0;
-}
-
-static int
-nv50_base_ntfy_wait_begun(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw)
-{
-	struct nouveau_drm *drm = nouveau_drm(wndw->plane.dev);
-	struct nv50_disp *disp = nv50_disp(wndw->plane.dev);
-	if (nvif_msec(&drm->client.device, 2000ULL,
-		u32 data = nouveau_bo_rd32(disp->sync, asyw->ntfy.offset / 4);
-		if ((data & 0xc0000000) == 0x40000000)
-			break;
-		usleep_range(1, 2);
-	) < 0)
-		return -ETIMEDOUT;
-	return 0;
-}
-
-static void
-nv50_base_release(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
-		  struct nv50_head_atom *asyh)
-{
-	asyh->base.cpp = 0;
-}
-
-static int
-nv50_base_acquire(struct nv50_wndw *wndw, struct nv50_wndw_atom *asyw,
-		  struct nv50_head_atom *asyh)
-{
-	const struct drm_framebuffer *fb = asyw->state.fb;
-	int ret;
-
-	if (!fb->format->depth)
-		return -EINVAL;
-
-	ret = drm_atomic_helper_check_plane_state(&asyw->state, &asyh->state,
-						  DRM_PLANE_HELPER_NO_SCALING,
-						  DRM_PLANE_HELPER_NO_SCALING,
-						  false, true);
-	if (ret)
-		return ret;
-
-	asyh->base.depth = fb->format->depth;
-	asyh->base.cpp = fb->format->cpp[0];
-	asyh->base.x = asyw->state.src.x1 >> 16;
-	asyh->base.y = asyw->state.src.y1 >> 16;
-	asyh->base.w = asyw->state.fb->width;
-	asyh->base.h = asyw->state.fb->height;
-
-	switch (fb->format->format) {
-	case DRM_FORMAT_C8         : asyw->image.format = 0x1e; break;
-	case DRM_FORMAT_RGB565     : asyw->image.format = 0xe8; break;
-	case DRM_FORMAT_XRGB1555   :
-	case DRM_FORMAT_ARGB1555   : asyw->image.format = 0xe9; break;
-	case DRM_FORMAT_XRGB8888   :
-	case DRM_FORMAT_ARGB8888   : asyw->image.format = 0xcf; break;
-	case DRM_FORMAT_XBGR2101010:
-	case DRM_FORMAT_ABGR2101010: asyw->image.format = 0xd1; break;
-	case DRM_FORMAT_XBGR8888   :
-	case DRM_FORMAT_ABGR8888   : asyw->image.format = 0xd5; break;
-	default:
-		WARN_ON(1);
-		return -EINVAL;
-	}
-
-	asyw->lut.enable = 1;
-	asyw->set.image = true;
-	return 0;
-}
-
-static void *
-nv50_base_dtor(struct nv50_wndw *wndw)
-{
-	struct nv50_disp *disp = nv50_disp(wndw->plane.dev);
-	struct nv50_base *base = nv50_base(wndw);
-	nv50_dmac_destroy(&base->chan.base, disp->disp);
-	return base;
-}
-
-static const u32
-nv50_base_format[] = {
-	DRM_FORMAT_C8,
-	DRM_FORMAT_RGB565,
-	DRM_FORMAT_XRGB1555,
-	DRM_FORMAT_ARGB1555,
-	DRM_FORMAT_XRGB8888,
-	DRM_FORMAT_ARGB8888,
-	DRM_FORMAT_XBGR2101010,
-	DRM_FORMAT_ABGR2101010,
-	DRM_FORMAT_XBGR8888,
-	DRM_FORMAT_ABGR8888,
-};
-
-static const struct nv50_wndw_func
-nv50_base = {
-	.dtor = nv50_base_dtor,
-	.acquire = nv50_base_acquire,
-	.release = nv50_base_release,
-	.sema_set = nv50_base_sema_set,
-	.sema_clr = nv50_base_sema_clr,
-	.ntfy_set = nv50_base_ntfy_set,
-	.ntfy_clr = nv50_base_ntfy_clr,
-	.ntfy_wait_begun = nv50_base_ntfy_wait_begun,
-	.image_set = nv50_base_image_set,
-	.image_clr = nv50_base_image_clr,
-	.lut = nv50_base_lut,
-	.update = nv50_base_update,
-};
-
-static int
-nv50_base_new(struct nouveau_drm *drm, struct nv50_head *head,
-	      struct nv50_base **pbase)
-{
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct nv50_base *base;
-	int ret;
-
-	if (!(base = *pbase = kzalloc(sizeof(*base), GFP_KERNEL)))
-		return -ENOMEM;
-	base->id = head->base.index;
-	base->wndw.ntfy = EVO_FLIP_NTFY0(base->id);
-	base->wndw.sema = EVO_FLIP_SEM0(base->id);
-	base->wndw.data = 0x00000000;
-
-	ret = nv50_wndw_ctor(&nv50_base, drm->dev, DRM_PLANE_TYPE_PRIMARY,
-			     "base", base->id, &base->chan.base,
-			     nv50_base_format, ARRAY_SIZE(nv50_base_format),
-			     &base->wndw);
-	if (ret) {
-		kfree(base);
-		return ret;
-	}
-
-	ret = nv50_base_create(&drm->client.device, disp->disp, base->id,
-			       disp->sync->bo.offset, &base->chan);
-	if (ret)
-		return ret;
-
-	return nvif_notify_init(&base->chan.base.base.user, nv50_base_notify,
-				false,
-				NV50_DISP_BASE_CHANNEL_DMA_V0_NTFY_UEVENT,
-				&(struct nvif_notify_uevent_req) {},
-				sizeof(struct nvif_notify_uevent_req),
-				sizeof(struct nvif_notify_uevent_rep),
-				&base->wndw.notify);
-}
-
-/******************************************************************************
- * Head
- *****************************************************************************/
-static void
-nv50_head_procamp(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x08a8 + (head->base.index * 0x400), 1);
-		else
-			evo_mthd(push, 0x0498 + (head->base.index * 0x300), 1);
-		evo_data(push, (asyh->procamp.sat.sin << 20) |
-			       (asyh->procamp.sat.cos << 8));
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_dither(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x08a0 + (head->base.index * 0x0400), 1);
-		else
-		if (core->base.user.oclass < GK104_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0490 + (head->base.index * 0x0300), 1);
-		else
-			evo_mthd(push, 0x04a0 + (head->base.index * 0x0300), 1);
-		evo_data(push, (asyh->dither.mode << 3) |
-			       (asyh->dither.bits << 1) |
-			        asyh->dither.enable);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_ovly(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 bounds = 0;
-	u32 *push;
-
-	if (asyh->base.cpp) {
-		switch (asyh->base.cpp) {
-		case 8: bounds |= 0x00000500; break;
-		case 4: bounds |= 0x00000300; break;
-		case 2: bounds |= 0x00000100; break;
-		default:
-			WARN_ON(1);
-			break;
-		}
-		bounds |= 0x00000001;
-	}
-
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0904 + head->base.index * 0x400, 1);
-		else
-			evo_mthd(push, 0x04d4 + head->base.index * 0x300, 1);
-		evo_data(push, bounds);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_base(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 bounds = 0;
-	u32 *push;
-
-	if (asyh->base.cpp) {
-		switch (asyh->base.cpp) {
-		case 8: bounds |= 0x00000500; break;
-		case 4: bounds |= 0x00000300; break;
-		case 2: bounds |= 0x00000100; break;
-		case 1: bounds |= 0x00000000; break;
-		default:
-			WARN_ON(1);
-			break;
-		}
-		bounds |= 0x00000001;
-	}
-
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0900 + head->base.index * 0x400, 1);
-		else
-			evo_mthd(push, 0x04d0 + head->base.index * 0x300, 1);
-		evo_data(push, bounds);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_curs_clr(struct nv50_head *head)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 4))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
-			evo_data(push, 0x05000000);
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 1);
-			evo_data(push, 0x05000000);
-			evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
-			evo_data(push, 0x00000000);
-		} else {
-			evo_mthd(push, 0x0480 + head->base.index * 0x300, 1);
-			evo_data(push, 0x05000000);
-			evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
-			evo_data(push, 0x00000000);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_curs_set(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 5))) {
-		if (core->base.user.oclass < G82_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
-			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
-						    (asyh->curs.format << 24));
-			evo_data(push, asyh->curs.offset >> 8);
-		} else
-		if (core->base.user.oclass < GF110_DISP_BASE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0880 + head->base.index * 0x400, 2);
-			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
-						    (asyh->curs.format << 24));
-			evo_data(push, asyh->curs.offset >> 8);
-			evo_mthd(push, 0x089c + head->base.index * 0x400, 1);
-			evo_data(push, asyh->curs.handle);
-		} else {
-			evo_mthd(push, 0x0480 + head->base.index * 0x300, 2);
-			evo_data(push, 0x80000000 | (asyh->curs.layout << 26) |
-						    (asyh->curs.format << 24));
-			evo_data(push, asyh->curs.offset >> 8);
-			evo_mthd(push, 0x048c + head->base.index * 0x300, 1);
-			evo_data(push, asyh->curs.handle);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_core_clr(struct nv50_head *head)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 2))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA)
-			evo_mthd(push, 0x0874 + head->base.index * 0x400, 1);
-		else
-			evo_mthd(push, 0x0474 + head->base.index * 0x300, 1);
-		evo_data(push, 0x00000000);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_core_set(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 9))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
-			evo_data(push, asyh->core.offset >> 8);
-			evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
-			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
-			evo_data(push, asyh->core.layout << 20 |
-				       (asyh->core.pitch >> 8) << 8 |
-				       asyh->core.block);
-			evo_data(push, asyh->core.kind << 16 |
-				       asyh->core.format << 8);
-			evo_data(push, asyh->core.handle);
-			evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
-			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
-			/* EVO will complain with INVALID_STATE if we have an
-			 * active cursor and (re)specify HeadSetContextDmaIso
-			 * without also updating HeadSetOffsetCursor.
-			 */
-			asyh->set.curs = asyh->curs.visible;
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0860 + head->base.index * 0x400, 1);
-			evo_data(push, asyh->core.offset >> 8);
-			evo_mthd(push, 0x0868 + head->base.index * 0x400, 4);
-			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
-			evo_data(push, asyh->core.layout << 20 |
-				       (asyh->core.pitch >> 8) << 8 |
-				       asyh->core.block);
-			evo_data(push, asyh->core.format << 8);
-			evo_data(push, asyh->core.handle);
-			evo_mthd(push, 0x08c0 + head->base.index * 0x400, 1);
-			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
-		} else {
-			evo_mthd(push, 0x0460 + head->base.index * 0x300, 1);
-			evo_data(push, asyh->core.offset >> 8);
-			evo_mthd(push, 0x0468 + head->base.index * 0x300, 4);
-			evo_data(push, (asyh->core.h << 16) | asyh->core.w);
-			evo_data(push, asyh->core.layout << 24 |
-				       (asyh->core.pitch >> 8) << 8 |
-				       asyh->core.block);
-			evo_data(push, asyh->core.format << 8);
-			evo_data(push, asyh->core.handle);
-			evo_mthd(push, 0x04b0 + head->base.index * 0x300, 1);
-			evo_data(push, (asyh->core.y << 16) | asyh->core.x);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_lut_clr(struct nv50_head *head)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 4))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
-			evo_data(push, 0x40000000);
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 1);
-			evo_data(push, 0x40000000);
-			evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
-			evo_data(push, 0x00000000);
-		} else {
-			evo_mthd(push, 0x0440 + (head->base.index * 0x300), 1);
-			evo_data(push, 0x03000000);
-			evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
-			evo_data(push, 0x00000000);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_lut_load(struct drm_property_blob *blob, int mode,
-		   struct nouveau_bo *nvbo)
-{
-	struct drm_color_lut *in = (struct drm_color_lut *)blob->data;
-	void __iomem *lut = (u8 *)nvbo_kmap_obj_iovirtual(nvbo);
-	const int size = blob->length / sizeof(*in);
-	int bits, shift, i;
-	u16 zero, r, g, b;
-
-	/* This can't happen.. But it shuts the compiler up. */
-	if (WARN_ON(size != 256))
-		return;
-
-	switch (mode) {
-	case 0: /* LORES. */
-	case 1: /* HIRES. */
-		bits = 11;
-		shift = 3;
-		zero = 0x0000;
-		break;
-	case 7: /* INTERPOLATE_257_UNITY_RANGE. */
-		bits = 14;
-		shift = 0;
-		zero = 0x6000;
-		break;
-	default:
-		WARN_ON(1);
-		return;
-	}
-
-	for (i = 0; i < size; i++) {
-		r = (drm_color_lut_extract(in[i].  red, bits) + zero) << shift;
-		g = (drm_color_lut_extract(in[i].green, bits) + zero) << shift;
-		b = (drm_color_lut_extract(in[i]. blue, bits) + zero) << shift;
-		writew(r, lut + (i * 0x08) + 0);
-		writew(g, lut + (i * 0x08) + 2);
-		writew(b, lut + (i * 0x08) + 4);
-	}
-
-	/* INTERPOLATE modes require a "next" entry to interpolate with,
-	 * so we replicate the last entry to deal with this for now.
-	 */
-	writew(r, lut + (i * 0x08) + 0);
-	writew(g, lut + (i * 0x08) + 2);
-	writew(b, lut + (i * 0x08) + 4);
-}
-
-static void
-nv50_head_lut_set(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 7))) {
-		if (core->base.user.oclass < G82_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
-			evo_data(push, 0x80000000 | asyh->lut.mode << 30);
-			evo_data(push, asyh->lut.offset >> 8);
-		} else
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0840 + (head->base.index * 0x400), 2);
-			evo_data(push, 0x80000000 | asyh->lut.mode << 30);
-			evo_data(push, asyh->lut.offset >> 8);
-			evo_mthd(push, 0x085c + (head->base.index * 0x400), 1);
-			evo_data(push, asyh->lut.handle);
-		} else {
-			evo_mthd(push, 0x0440 + (head->base.index * 0x300), 4);
-			evo_data(push, 0x80000000 | asyh->lut.mode << 24);
-			evo_data(push, asyh->lut.offset >> 8);
-			evo_data(push, 0x00000000);
-			evo_data(push, 0x00000000);
-			evo_mthd(push, 0x045c + (head->base.index * 0x300), 1);
-			evo_data(push, asyh->lut.handle);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	struct nv50_head_mode *m = &asyh->mode;
-	u32 *push;
-	if ((push = evo_wait(core, 14))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x0804 + (head->base.index * 0x400), 2);
-			evo_data(push, 0x00800000 | m->clock);
-			evo_data(push, m->interlace ? 0x00000002 : 0x00000000);
-			evo_mthd(push, 0x0810 + (head->base.index * 0x400), 7);
-			evo_data(push, 0x00000000);
-			evo_data(push, (m->v.active  << 16) | m->h.active );
-			evo_data(push, (m->v.synce   << 16) | m->h.synce  );
-			evo_data(push, (m->v.blanke  << 16) | m->h.blanke );
-			evo_data(push, (m->v.blanks  << 16) | m->h.blanks );
-			evo_data(push, (m->v.blank2e << 16) | m->v.blank2s);
-			evo_data(push, asyh->mode.v.blankus);
-			evo_mthd(push, 0x082c + (head->base.index * 0x400), 1);
-			evo_data(push, 0x00000000);
-		} else {
-			evo_mthd(push, 0x0410 + (head->base.index * 0x300), 6);
-			evo_data(push, 0x00000000);
-			evo_data(push, (m->v.active  << 16) | m->h.active );
-			evo_data(push, (m->v.synce   << 16) | m->h.synce  );
-			evo_data(push, (m->v.blanke  << 16) | m->h.blanke );
-			evo_data(push, (m->v.blanks  << 16) | m->h.blanks );
-			evo_data(push, (m->v.blank2e << 16) | m->v.blank2s);
-			evo_mthd(push, 0x042c + (head->base.index * 0x300), 2);
-			evo_data(push, 0x00000000); /* ??? */
-			evo_data(push, 0xffffff00);
-			evo_mthd(push, 0x0450 + (head->base.index * 0x300), 3);
-			evo_data(push, m->clock * 1000);
-			evo_data(push, 0x00200000); /* ??? */
-			evo_data(push, m->clock * 1000);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_view(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct nv50_dmac *core = &nv50_disp(head->base.base.dev)->mast.base;
-	u32 *push;
-	if ((push = evo_wait(core, 10))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			evo_mthd(push, 0x08a4 + (head->base.index * 0x400), 1);
-			evo_data(push, 0x00000000);
-			evo_mthd(push, 0x08c8 + (head->base.index * 0x400), 1);
-			evo_data(push, (asyh->view.iH << 16) | asyh->view.iW);
-			evo_mthd(push, 0x08d8 + (head->base.index * 0x400), 2);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-		} else {
-			evo_mthd(push, 0x0494 + (head->base.index * 0x300), 1);
-			evo_data(push, 0x00000000);
-			evo_mthd(push, 0x04b8 + (head->base.index * 0x300), 1);
-			evo_data(push, (asyh->view.iH << 16) | asyh->view.iW);
-			evo_mthd(push, 0x04c0 + (head->base.index * 0x300), 3);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-			evo_data(push, (asyh->view.oH << 16) | asyh->view.oW);
-		}
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_head_flush_clr(struct nv50_head *head, struct nv50_head_atom *asyh, bool y)
-{
-	if (asyh->clr.ilut && (!asyh->set.ilut || y))
-		nv50_head_lut_clr(head);
-	if (asyh->clr.core && (!asyh->set.core || y))
-		nv50_head_core_clr(head);
-	if (asyh->clr.curs && (!asyh->set.curs || y))
-		nv50_head_curs_clr(head);
-}
-
-static void
-nv50_head_flush_set(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	if (asyh->set.view   ) nv50_head_view    (head, asyh);
-	if (asyh->set.mode   ) nv50_head_mode    (head, asyh);
-	if (asyh->set.ilut   ) {
-		struct nouveau_bo *nvbo = head->lut.nvbo[head->lut.next];
-		struct drm_property_blob *blob = asyh->state.gamma_lut;
-		if (blob)
-			nv50_head_lut_load(blob, asyh->lut.mode, nvbo);
-		asyh->lut.offset = nvbo->bo.offset;
-		head->lut.next ^= 1;
-		nv50_head_lut_set(head, asyh);
-	}
-	if (asyh->set.core   ) nv50_head_core_set(head, asyh);
-	if (asyh->set.curs   ) nv50_head_curs_set(head, asyh);
-	if (asyh->set.base   ) nv50_head_base    (head, asyh);
-	if (asyh->set.ovly   ) nv50_head_ovly    (head, asyh);
-	if (asyh->set.dither ) nv50_head_dither  (head, asyh);
-	if (asyh->set.procamp) nv50_head_procamp (head, asyh);
-}
-
-static void
-nv50_head_atomic_check_procamp(struct nv50_head_atom *armh,
-			       struct nv50_head_atom *asyh,
-			       struct nouveau_conn_atom *asyc)
-{
-	const int vib = asyc->procamp.color_vibrance - 100;
-	const int hue = asyc->procamp.vibrant_hue - 90;
-	const int adj = (vib > 0) ? 50 : 0;
-	asyh->procamp.sat.cos = ((vib * 2047 + adj) / 100) & 0xfff;
-	asyh->procamp.sat.sin = ((hue * 2047) / 100) & 0xfff;
-	asyh->set.procamp = true;
-}
-
-static void
-nv50_head_atomic_check_dither(struct nv50_head_atom *armh,
-			      struct nv50_head_atom *asyh,
-			      struct nouveau_conn_atom *asyc)
-{
-	struct drm_connector *connector = asyc->state.connector;
-	u32 mode = 0x00;
-
-	if (asyc->dither.mode == DITHERING_MODE_AUTO) {
-		if (asyh->base.depth > connector->display_info.bpc * 3)
-			mode = DITHERING_MODE_DYNAMIC2X2;
-	} else {
-		mode = asyc->dither.mode;
-	}
-
-	if (asyc->dither.depth == DITHERING_DEPTH_AUTO) {
-		if (connector->display_info.bpc >= 8)
-			mode |= DITHERING_DEPTH_8BPC;
-	} else {
-		mode |= asyc->dither.depth;
-	}
-
-	asyh->dither.enable = mode;
-	asyh->dither.bits = mode >> 1;
-	asyh->dither.mode = mode >> 3;
-	asyh->set.dither = true;
-}
-
-static void
-nv50_head_atomic_check_view(struct nv50_head_atom *armh,
-			    struct nv50_head_atom *asyh,
-			    struct nouveau_conn_atom *asyc)
-{
-	struct drm_connector *connector = asyc->state.connector;
-	struct drm_display_mode *omode = &asyh->state.adjusted_mode;
-	struct drm_display_mode *umode = &asyh->state.mode;
-	int mode = asyc->scaler.mode;
-	struct edid *edid;
-	int umode_vdisplay, omode_hdisplay, omode_vdisplay;
-
-	if (connector->edid_blob_ptr)
-		edid = (struct edid *)connector->edid_blob_ptr->data;
-	else
-		edid = NULL;
-
-	if (!asyc->scaler.full) {
-		if (mode == DRM_MODE_SCALE_NONE)
-			omode = umode;
-	} else {
-		/* Non-EDID LVDS/eDP mode. */
-		mode = DRM_MODE_SCALE_FULLSCREEN;
-	}
-
-	/* For the user-specified mode, we must ignore doublescan and
-	 * the like, but honor frame packing.
-	 */
-	umode_vdisplay = umode->vdisplay;
-	if ((umode->flags & DRM_MODE_FLAG_3D_MASK) == DRM_MODE_FLAG_3D_FRAME_PACKING)
-		umode_vdisplay += umode->vtotal;
-	asyh->view.iW = umode->hdisplay;
-	asyh->view.iH = umode_vdisplay;
-	/* For the output mode, we can just use the stock helper. */
-	drm_mode_get_hv_timing(omode, &omode_hdisplay, &omode_vdisplay);
-	asyh->view.oW = omode_hdisplay;
-	asyh->view.oH = omode_vdisplay;
-
-	/* Add overscan compensation if necessary, will keep the aspect
-	 * ratio the same as the backend mode unless overridden by the
-	 * user setting both hborder and vborder properties.
-	 */
-	if ((asyc->scaler.underscan.mode == UNDERSCAN_ON ||
-	    (asyc->scaler.underscan.mode == UNDERSCAN_AUTO &&
-	     drm_detect_hdmi_monitor(edid)))) {
-		u32 bX = asyc->scaler.underscan.hborder;
-		u32 bY = asyc->scaler.underscan.vborder;
-		u32 r = (asyh->view.oH << 19) / asyh->view.oW;
-
-		if (bX) {
-			asyh->view.oW -= (bX * 2);
-			if (bY) asyh->view.oH -= (bY * 2);
-			else    asyh->view.oH  = ((asyh->view.oW * r) + (r / 2)) >> 19;
-		} else {
-			asyh->view.oW -= (asyh->view.oW >> 4) + 32;
-			if (bY) asyh->view.oH -= (bY * 2);
-			else    asyh->view.oH  = ((asyh->view.oW * r) + (r / 2)) >> 19;
-		}
-	}
-
-	/* Handle CENTER/ASPECT scaling, taking into account the areas
-	 * removed already for overscan compensation.
-	 */
-	switch (mode) {
-	case DRM_MODE_SCALE_CENTER:
-		asyh->view.oW = min((u16)umode->hdisplay, asyh->view.oW);
-		asyh->view.oH = min((u16)umode_vdisplay, asyh->view.oH);
-		/* fall-through */
-	case DRM_MODE_SCALE_ASPECT:
-		if (asyh->view.oH < asyh->view.oW) {
-			u32 r = (asyh->view.iW << 19) / asyh->view.iH;
-			asyh->view.oW = ((asyh->view.oH * r) + (r / 2)) >> 19;
-		} else {
-			u32 r = (asyh->view.iH << 19) / asyh->view.iW;
-			asyh->view.oH = ((asyh->view.oW * r) + (r / 2)) >> 19;
-		}
-		break;
-	default:
-		break;
-	}
-
-	asyh->set.view = true;
-}
-
-static void
-nv50_head_atomic_check_lut(struct nv50_head *head,
-			   struct nv50_head_atom *armh,
-			   struct nv50_head_atom *asyh)
-{
-	struct nv50_disp *disp = nv50_disp(head->base.base.dev);
-
-	/* An I8 surface without an input LUT makes no sense, and
-	 * EVO will throw an error if you try.
-	 *
-	 * Legacy clients actually cause this due to the order in
-	 * which they call ioctls, so we will enable the LUT with
-	 * whatever contents the buffer already contains to avoid
-	 * triggering the error check.
-	 */
-	if (!asyh->state.gamma_lut && asyh->base.cpp != 1) {
-		asyh->lut.handle = 0;
-		asyh->clr.ilut = armh->lut.visible;
-		return;
-	}
-
-	if (disp->disp->oclass < GF110_DISP) {
-		asyh->lut.mode = (asyh->base.cpp == 1) ? 0 : 1;
-		asyh->set.ilut = true;
-	} else {
-		asyh->lut.mode = 7;
-		asyh->set.ilut = asyh->state.color_mgmt_changed;
-	}
-	asyh->lut.handle = disp->mast.base.vram.handle;
-}
-
-static void
-nv50_head_atomic_check_mode(struct nv50_head *head, struct nv50_head_atom *asyh)
-{
-	struct drm_display_mode *mode = &asyh->state.adjusted_mode;
-	struct nv50_head_mode *m = &asyh->mode;
-	u32 blankus;
-
-	drm_mode_set_crtcinfo(mode, CRTC_INTERLACE_HALVE_V | CRTC_STEREO_DOUBLE);
-
-	/*
-	 * DRM modes are defined in terms of a repeating interval
-	 * starting with the active display area.  The hardware modes
-	 * are defined in terms of a repeating interval starting one
-	 * unit (pixel or line) into the sync pulse.  So, add bias.
-	 */
-
-	m->h.active = mode->crtc_htotal;
-	m->h.synce  = mode->crtc_hsync_end - mode->crtc_hsync_start - 1;
-	m->h.blanke = mode->crtc_hblank_end - mode->crtc_hsync_start - 1;
-	m->h.blanks = m->h.blanke + mode->crtc_hdisplay;
-
-	m->v.active = mode->crtc_vtotal;
-	m->v.synce  = mode->crtc_vsync_end - mode->crtc_vsync_start - 1;
-	m->v.blanke = mode->crtc_vblank_end - mode->crtc_vsync_start - 1;
-	m->v.blanks = m->v.blanke + mode->crtc_vdisplay;
-
-	/*XXX: Safe underestimate, even "0" works */
-	blankus = (m->v.active - mode->crtc_vdisplay - 2) * m->h.active;
-	blankus *= 1000;
-	blankus /= mode->crtc_clock;
-	m->v.blankus = blankus;
-
-	if (mode->flags & DRM_MODE_FLAG_INTERLACE) {
-		m->v.blank2e =  m->v.active + m->v.blanke;
-		m->v.blank2s =  m->v.blank2e + mode->crtc_vdisplay;
-		m->v.active  = (m->v.active * 2) + 1;
-		m->interlace = true;
-	} else {
-		m->v.blank2e = 0;
-		m->v.blank2s = 1;
-		m->interlace = false;
-	}
-	m->clock = mode->crtc_clock;
-
-	asyh->set.mode = true;
-}
-
-static int
-nv50_head_atomic_check(struct drm_crtc *crtc, struct drm_crtc_state *state)
-{
-	struct nouveau_drm *drm = nouveau_drm(crtc->dev);
-	struct nv50_disp *disp = nv50_disp(crtc->dev);
-	struct nv50_head *head = nv50_head(crtc);
-	struct nv50_head_atom *armh = nv50_head_atom(crtc->state);
-	struct nv50_head_atom *asyh = nv50_head_atom(state);
-	struct nouveau_conn_atom *asyc = NULL;
-	struct drm_connector_state *conns;
-	struct drm_connector *conn;
-	int i;
-
-	NV_ATOMIC(drm, "%s atomic_check %d\n", crtc->name, asyh->state.active);
-	if (asyh->state.active) {
-		for_each_new_connector_in_state(asyh->state.state, conn, conns, i) {
-			if (conns->crtc == crtc) {
-				asyc = nouveau_conn_atom(conns);
-				break;
-			}
-		}
-
-		if (armh->state.active) {
-			if (asyc) {
-				if (asyh->state.mode_changed)
-					asyc->set.scaler = true;
-				if (armh->base.depth != asyh->base.depth)
-					asyc->set.dither = true;
-			}
-		} else {
-			if (asyc)
-				asyc->set.mask = ~0;
-			asyh->set.mask = ~0;
-		}
-
-		if (asyh->state.mode_changed)
-			nv50_head_atomic_check_mode(head, asyh);
-
-		if (asyh->state.color_mgmt_changed ||
-		    asyh->base.cpp != armh->base.cpp)
-			nv50_head_atomic_check_lut(head, armh, asyh);
-		asyh->lut.visible = asyh->lut.handle != 0;
-
-		if (asyc) {
-			if (asyc->set.scaler)
-				nv50_head_atomic_check_view(armh, asyh, asyc);
-			if (asyc->set.dither)
-				nv50_head_atomic_check_dither(armh, asyh, asyc);
-			if (asyc->set.procamp)
-				nv50_head_atomic_check_procamp(armh, asyh, asyc);
-		}
-
-		if ((asyh->core.visible = (asyh->base.cpp != 0))) {
-			asyh->core.x = asyh->base.x;
-			asyh->core.y = asyh->base.y;
-			asyh->core.w = asyh->base.w;
-			asyh->core.h = asyh->base.h;
-		} else
-		if ((asyh->core.visible = asyh->curs.visible) ||
-		    (asyh->core.visible = asyh->lut.visible)) {
-			/*XXX: We need to either find some way of having the
-			 *     primary base layer appear black, while still
-			 *     being able to display the other layers, or we
-			 *     need to allocate a dummy black surface here.
-			 */
-			asyh->core.x = 0;
-			asyh->core.y = 0;
-			asyh->core.w = asyh->state.mode.hdisplay;
-			asyh->core.h = asyh->state.mode.vdisplay;
-		}
-		asyh->core.handle = disp->mast.base.vram.handle;
-		asyh->core.offset = 0;
-		asyh->core.format = 0xcf;
-		asyh->core.kind = 0;
-		asyh->core.layout = 1;
-		asyh->core.block = 0;
-		asyh->core.pitch = ALIGN(asyh->core.w, 64) * 4;
-		asyh->set.base = armh->base.cpp != asyh->base.cpp;
-		asyh->set.ovly = armh->ovly.cpp != asyh->ovly.cpp;
-	} else {
-		asyh->lut.visible = false;
-		asyh->core.visible = false;
-		asyh->curs.visible = false;
-		asyh->base.cpp = 0;
-		asyh->ovly.cpp = 0;
-	}
-
-	if (!drm_atomic_crtc_needs_modeset(&asyh->state)) {
-		if (asyh->core.visible) {
-			if (memcmp(&armh->core, &asyh->core, sizeof(asyh->core)))
-				asyh->set.core = true;
-		} else
-		if (armh->core.visible) {
-			asyh->clr.core = true;
-		}
-
-		if (asyh->curs.visible) {
-			if (memcmp(&armh->curs, &asyh->curs, sizeof(asyh->curs)))
-				asyh->set.curs = true;
-		} else
-		if (armh->curs.visible) {
-			asyh->clr.curs = true;
-		}
-	} else {
-		asyh->clr.ilut = armh->lut.visible;
-		asyh->clr.core = armh->core.visible;
-		asyh->clr.curs = armh->curs.visible;
-		asyh->set.ilut = asyh->lut.visible;
-		asyh->set.core = asyh->core.visible;
-		asyh->set.curs = asyh->curs.visible;
-	}
-
-	if (asyh->clr.mask || asyh->set.mask)
-		nv50_atom(asyh->state.state)->lock_core = true;
-	return 0;
-}
-
-static const struct drm_crtc_helper_funcs
-nv50_head_help = {
-	.atomic_check = nv50_head_atomic_check,
-};
-
-static void
-nv50_head_atomic_destroy_state(struct drm_crtc *crtc,
-			       struct drm_crtc_state *state)
-{
-	struct nv50_head_atom *asyh = nv50_head_atom(state);
-	__drm_atomic_helper_crtc_destroy_state(&asyh->state);
-	kfree(asyh);
-}
-
-static struct drm_crtc_state *
-nv50_head_atomic_duplicate_state(struct drm_crtc *crtc)
-{
-	struct nv50_head_atom *armh = nv50_head_atom(crtc->state);
-	struct nv50_head_atom *asyh;
-	if (!(asyh = kmalloc(sizeof(*asyh), GFP_KERNEL)))
-		return NULL;
-	__drm_atomic_helper_crtc_duplicate_state(crtc, &asyh->state);
-	asyh->view = armh->view;
-	asyh->mode = armh->mode;
-	asyh->lut  = armh->lut;
-	asyh->core = armh->core;
-	asyh->curs = armh->curs;
-	asyh->base = armh->base;
-	asyh->ovly = armh->ovly;
-	asyh->dither = armh->dither;
-	asyh->procamp = armh->procamp;
-	asyh->clr.mask = 0;
-	asyh->set.mask = 0;
-	return &asyh->state;
-}
-
-static void
-__drm_atomic_helper_crtc_reset(struct drm_crtc *crtc,
-			       struct drm_crtc_state *state)
-{
-	if (crtc->state)
-		crtc->funcs->atomic_destroy_state(crtc, crtc->state);
-	crtc->state = state;
-	crtc->state->crtc = crtc;
-}
-
-static void
-nv50_head_reset(struct drm_crtc *crtc)
-{
-	struct nv50_head_atom *asyh;
-
-	if (WARN_ON(!(asyh = kzalloc(sizeof(*asyh), GFP_KERNEL))))
-		return;
-
-	__drm_atomic_helper_crtc_reset(crtc, &asyh->state);
-}
-
-static void
-nv50_head_destroy(struct drm_crtc *crtc)
-{
-	struct nv50_disp *disp = nv50_disp(crtc->dev);
-	struct nv50_head *head = nv50_head(crtc);
-	int i;
-
-	nv50_dmac_destroy(&head->ovly.base, disp->disp);
-	nv50_pioc_destroy(&head->oimm.base);
-
-	for (i = 0; i < ARRAY_SIZE(head->lut.nvbo); i++)
-		nouveau_bo_unmap_unpin_unref(&head->lut.nvbo[i]);
-
-	drm_crtc_cleanup(crtc);
-	kfree(crtc);
-}
-
-static const struct drm_crtc_funcs
-nv50_head_func = {
-	.reset = nv50_head_reset,
-	.gamma_set = drm_atomic_helper_legacy_gamma_set,
-	.destroy = nv50_head_destroy,
-	.set_config = drm_atomic_helper_set_config,
-	.page_flip = drm_atomic_helper_page_flip,
-	.atomic_duplicate_state = nv50_head_atomic_duplicate_state,
-	.atomic_destroy_state = nv50_head_atomic_destroy_state,
-};
-
-static int
-nv50_head_create(struct drm_device *dev, int index)
-{
-	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct nvif_device *device = &drm->client.device;
-	struct nv50_disp *disp = nv50_disp(dev);
-	struct nv50_head *head;
-	struct nv50_base *base;
-	struct nv50_curs *curs;
-	struct drm_crtc *crtc;
-	int ret, i;
-
-	head = kzalloc(sizeof(*head), GFP_KERNEL);
-	if (!head)
-		return -ENOMEM;
-
-	head->base.index = index;
-	ret = nv50_base_new(drm, head, &base);
-	if (ret == 0)
-		ret = nv50_curs_new(drm, head, &curs);
-	if (ret) {
-		kfree(head);
-		return ret;
-	}
-
-	crtc = &head->base.base;
-	drm_crtc_init_with_planes(dev, crtc, &base->wndw.plane,
-				  &curs->wndw.plane, &nv50_head_func,
-				  "head-%d", head->base.index);
-	drm_crtc_helper_add(crtc, &nv50_head_help);
-	drm_mode_crtc_set_gamma_size(crtc, 256);
-
-	for (i = 0; i < ARRAY_SIZE(head->lut.nvbo); i++) {
-		ret = nouveau_bo_new_pin_map(&drm->client, 1025 * 8, 0x100,
-					     TTM_PL_FLAG_VRAM,
-					     &head->lut.nvbo[i]);
-		if (ret)
-			goto out;
-	}
-
-	/* allocate overlay resources */
-	ret = nv50_oimm_create(device, disp->disp, index, &head->oimm);
-	if (ret)
-		goto out;
-
-	ret = nv50_ovly_create(device, disp->disp, index, disp->sync->bo.offset,
-			       &head->ovly);
-	if (ret)
-		goto out;
-
-out:
-	if (ret)
-		nv50_head_destroy(crtc);
-	return ret;
-}
-
-/******************************************************************************
- * Output path helpers
- *****************************************************************************/
-static void
-nv50_outp_release(struct nouveau_encoder *nv_encoder)
-{
-	struct nv50_disp *disp = nv50_disp(nv_encoder->base.base.dev);
-	struct {
-		struct nv50_disp_mthd_v1 base;
-	} args = {
-		.base.version = 1,
-		.base.method = NV50_DISP_MTHD_V1_RELEASE,
-		.base.hasht  = nv_encoder->dcb->hasht,
-		.base.hashm  = nv_encoder->dcb->hashm,
-	};
-
-	nvif_mthd(disp->disp, 0, &args, sizeof(args));
-	nv_encoder->or = -1;
-	nv_encoder->link = 0;
-}
-
-static int
-nv50_outp_acquire(struct nouveau_encoder *nv_encoder)
-{
-	struct nouveau_drm *drm = nouveau_drm(nv_encoder->base.base.dev);
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_acquire_v0 info;
-	} args = {
-		.base.version = 1,
-		.base.method = NV50_DISP_MTHD_V1_ACQUIRE,
-		.base.hasht  = nv_encoder->dcb->hasht,
-		.base.hashm  = nv_encoder->dcb->hashm,
-	};
-	int ret;
-
-	ret = nvif_mthd(disp->disp, 0, &args, sizeof(args));
-	if (ret) {
-		NV_ERROR(drm, "error acquiring output path: %d\n", ret);
-		return ret;
-	}
-
-	nv_encoder->or = args.info.or;
-	nv_encoder->link = args.info.link;
-	return 0;
-}
-
-static int
-nv50_outp_atomic_check_view(struct drm_encoder *encoder,
-			    struct drm_crtc_state *crtc_state,
-			    struct drm_connector_state *conn_state,
-			    struct drm_display_mode *native_mode)
-{
-	struct drm_display_mode *adjusted_mode = &crtc_state->adjusted_mode;
-	struct drm_display_mode *mode = &crtc_state->mode;
-	struct drm_connector *connector = conn_state->connector;
-	struct nouveau_conn_atom *asyc = nouveau_conn_atom(conn_state);
-	struct nouveau_drm *drm = nouveau_drm(encoder->dev);
-
-	NV_ATOMIC(drm, "%s atomic_check\n", encoder->name);
-	asyc->scaler.full = false;
-	if (!native_mode)
-		return 0;
-
-	if (asyc->scaler.mode == DRM_MODE_SCALE_NONE) {
-		switch (connector->connector_type) {
-		case DRM_MODE_CONNECTOR_LVDS:
-		case DRM_MODE_CONNECTOR_eDP:
-			/* Force use of scaler for non-EDID modes. */
-			if (adjusted_mode->type & DRM_MODE_TYPE_DRIVER)
-				break;
-			mode = native_mode;
-			asyc->scaler.full = true;
-			break;
-		default:
-			break;
-		}
-	} else {
-		mode = native_mode;
-	}
-
-	if (!drm_mode_equal(adjusted_mode, mode)) {
-		drm_mode_copy(adjusted_mode, mode);
-		crtc_state->mode_changed = true;
-	}
-
-	return 0;
-}
-
-static int
-nv50_outp_atomic_check(struct drm_encoder *encoder,
-		       struct drm_crtc_state *crtc_state,
-		       struct drm_connector_state *conn_state)
-{
-	struct nouveau_connector *nv_connector =
-		nouveau_connector(conn_state->connector);
-	return nv50_outp_atomic_check_view(encoder, crtc_state, conn_state,
-					   nv_connector->native_mode);
-}
-
-/******************************************************************************
- * DAC
- *****************************************************************************/
-static void
-nv50_dac_disable(struct drm_encoder *encoder)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_mast *mast = nv50_mast(encoder->dev);
-	const int or = nv_encoder->or;
-	u32 *push;
-
-	if (nv_encoder->crtc) {
-		push = evo_wait(mast, 4);
-		if (push) {
-			if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
-				evo_mthd(push, 0x0400 + (or * 0x080), 1);
-				evo_data(push, 0x00000000);
-			} else {
-				evo_mthd(push, 0x0180 + (or * 0x020), 1);
-				evo_data(push, 0x00000000);
-			}
-			evo_kick(push, mast);
-		}
-	}
-
-	nv_encoder->crtc = NULL;
-	nv50_outp_release(nv_encoder);
-}
-
-static void
-nv50_dac_enable(struct drm_encoder *encoder)
-{
-	struct nv50_mast *mast = nv50_mast(encoder->dev);
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
-	struct drm_display_mode *mode = &nv_crtc->base.state->adjusted_mode;
-	u32 *push;
-
-	nv50_outp_acquire(nv_encoder);
-
-	push = evo_wait(mast, 8);
-	if (push) {
-		if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
-			u32 syncs = 0x00000000;
-
-			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-				syncs |= 0x00000001;
-			if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-				syncs |= 0x00000002;
-
-			evo_mthd(push, 0x0400 + (nv_encoder->or * 0x080), 2);
-			evo_data(push, 1 << nv_crtc->index);
-			evo_data(push, syncs);
-		} else {
-			u32 magic = 0x31ec6000 | (nv_crtc->index << 25);
-			u32 syncs = 0x00000001;
-
-			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-				syncs |= 0x00000008;
-			if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-				syncs |= 0x00000010;
-
-			if (mode->flags & DRM_MODE_FLAG_INTERLACE)
-				magic |= 0x00000001;
-
-			evo_mthd(push, 0x0404 + (nv_crtc->index * 0x300), 2);
-			evo_data(push, syncs);
-			evo_data(push, magic);
-			evo_mthd(push, 0x0180 + (nv_encoder->or * 0x020), 1);
-			evo_data(push, 1 << nv_crtc->index);
-		}
-
-		evo_kick(push, mast);
-	}
-
-	nv_encoder->crtc = encoder->crtc;
-}
-
-static enum drm_connector_status
-nv50_dac_detect(struct drm_encoder *encoder, struct drm_connector *connector)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_disp *disp = nv50_disp(encoder->dev);
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_dac_load_v0 load;
-	} args = {
-		.base.version = 1,
-		.base.method = NV50_DISP_MTHD_V1_DAC_LOAD,
-		.base.hasht  = nv_encoder->dcb->hasht,
-		.base.hashm  = nv_encoder->dcb->hashm,
-	};
-	int ret;
-
-	args.load.data = nouveau_drm(encoder->dev)->vbios.dactestval;
-	if (args.load.data == 0)
-		args.load.data = 340;
-
-	ret = nvif_mthd(disp->disp, 0, &args, sizeof(args));
-	if (ret || !args.load.load)
-		return connector_status_disconnected;
-
-	return connector_status_connected;
-}
-
-static const struct drm_encoder_helper_funcs
-nv50_dac_help = {
-	.atomic_check = nv50_outp_atomic_check,
-	.enable = nv50_dac_enable,
-	.disable = nv50_dac_disable,
-	.detect = nv50_dac_detect
-};
-
-static void
-nv50_dac_destroy(struct drm_encoder *encoder)
-{
-	drm_encoder_cleanup(encoder);
-	kfree(encoder);
-}
-
-static const struct drm_encoder_funcs
-nv50_dac_func = {
-	.destroy = nv50_dac_destroy,
-};
-
-static int
-nv50_dac_create(struct drm_connector *connector, struct dcb_output *dcbe)
-{
-	struct nouveau_drm *drm = nouveau_drm(connector->dev);
-	struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device);
-	struct nvkm_i2c_bus *bus;
-	struct nouveau_encoder *nv_encoder;
-	struct drm_encoder *encoder;
-	int type = DRM_MODE_ENCODER_DAC;
-
-	nv_encoder = kzalloc(sizeof(*nv_encoder), GFP_KERNEL);
-	if (!nv_encoder)
-		return -ENOMEM;
-	nv_encoder->dcb = dcbe;
-
-	bus = nvkm_i2c_bus_find(i2c, dcbe->i2c_index);
-	if (bus)
-		nv_encoder->i2c = &bus->i2c;
-
-	encoder = to_drm_encoder(nv_encoder);
-	encoder->possible_crtcs = dcbe->heads;
-	encoder->possible_clones = 0;
-	drm_encoder_init(connector->dev, encoder, &nv50_dac_func, type,
-			 "dac-%04x-%04x", dcbe->hasht, dcbe->hashm);
-	drm_encoder_helper_add(encoder, &nv50_dac_help);
-
-	drm_mode_connector_attach_encoder(connector, encoder);
-	return 0;
-}
-
-/******************************************************************************
- * Audio
- *****************************************************************************/
-static void
-nv50_audio_disable(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_disp *disp = nv50_disp(encoder->dev);
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_sor_hda_eld_v0 eld;
-	} args = {
-		.base.version = 1,
-		.base.method  = NV50_DISP_MTHD_V1_SOR_HDA_ELD,
-		.base.hasht   = nv_encoder->dcb->hasht,
-		.base.hashm   = (0xf0ff & nv_encoder->dcb->hashm) |
-				(0x0100 << nv_crtc->index),
-	};
-
-	nvif_mthd(disp->disp, 0, &args, sizeof(args));
-}
-
-static void
-nv50_audio_enable(struct drm_encoder *encoder, struct drm_display_mode *mode)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
-	struct nouveau_connector *nv_connector;
-	struct nv50_disp *disp = nv50_disp(encoder->dev);
-	struct __packed {
-		struct {
-			struct nv50_disp_mthd_v1 mthd;
-			struct nv50_disp_sor_hda_eld_v0 eld;
-		} base;
-		u8 data[sizeof(nv_connector->base.eld)];
-	} args = {
-		.base.mthd.version = 1,
-		.base.mthd.method  = NV50_DISP_MTHD_V1_SOR_HDA_ELD,
-		.base.mthd.hasht   = nv_encoder->dcb->hasht,
-		.base.mthd.hashm   = (0xf0ff & nv_encoder->dcb->hashm) |
-				     (0x0100 << nv_crtc->index),
-	};
-
-	nv_connector = nouveau_encoder_connector_get(nv_encoder);
-	if (!drm_detect_monitor_audio(nv_connector->edid))
-		return;
-
-	memcpy(args.data, nv_connector->base.eld, sizeof(args.data));
-
-	nvif_mthd(disp->disp, 0, &args,
-		  sizeof(args.base) + drm_eld_size(args.data));
-}
-
-/******************************************************************************
- * HDMI
- *****************************************************************************/
-static void
-nv50_hdmi_disable(struct drm_encoder *encoder, struct nouveau_crtc *nv_crtc)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_disp *disp = nv50_disp(encoder->dev);
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_sor_hdmi_pwr_v0 pwr;
-	} args = {
-		.base.version = 1,
-		.base.method = NV50_DISP_MTHD_V1_SOR_HDMI_PWR,
-		.base.hasht  = nv_encoder->dcb->hasht,
-		.base.hashm  = (0xf0ff & nv_encoder->dcb->hashm) |
-			       (0x0100 << nv_crtc->index),
-	};
-
-	nvif_mthd(disp->disp, 0, &args, sizeof(args));
-}
-
-static void
-nv50_hdmi_enable(struct drm_encoder *encoder, struct drm_display_mode *mode)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
-	struct nv50_disp *disp = nv50_disp(encoder->dev);
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_sor_hdmi_pwr_v0 pwr;
-		u8 infoframes[2 * 17]; /* two frames, up to 17 bytes each */
-	} args = {
-		.base.version = 1,
-		.base.method = NV50_DISP_MTHD_V1_SOR_HDMI_PWR,
-		.base.hasht  = nv_encoder->dcb->hasht,
-		.base.hashm  = (0xf0ff & nv_encoder->dcb->hashm) |
-			       (0x0100 << nv_crtc->index),
-		.pwr.state = 1,
-		.pwr.rekey = 56, /* binary driver, and tegra, constant */
-	};
-	struct nouveau_connector *nv_connector;
-	u32 max_ac_packet;
-	union hdmi_infoframe avi_frame;
-	union hdmi_infoframe vendor_frame;
-	int ret;
-	int size;
-
-	nv_connector = nouveau_encoder_connector_get(nv_encoder);
-	if (!drm_detect_hdmi_monitor(nv_connector->edid))
-		return;
-
-	ret = drm_hdmi_avi_infoframe_from_display_mode(&avi_frame.avi, mode,
-						       false);
-	if (!ret) {
-		/* We have an AVI InfoFrame, populate it to the display */
-		args.pwr.avi_infoframe_length
-			= hdmi_infoframe_pack(&avi_frame, args.infoframes, 17);
-	}
-
-	ret = drm_hdmi_vendor_infoframe_from_display_mode(&vendor_frame.vendor.hdmi,
-							  &nv_connector->base, mode);
-	if (!ret) {
-		/* We have a Vendor InfoFrame, populate it to the display */
-		args.pwr.vendor_infoframe_length
-			= hdmi_infoframe_pack(&vendor_frame,
-					      args.infoframes
-					      + args.pwr.avi_infoframe_length,
-					      17);
-	}
-
-	max_ac_packet  = mode->htotal - mode->hdisplay;
-	max_ac_packet -= args.pwr.rekey;
-	max_ac_packet -= 18; /* constant from tegra */
-	args.pwr.max_ac_packet = max_ac_packet / 32;
-
-	size = sizeof(args.base)
-		+ sizeof(args.pwr)
-		+ args.pwr.avi_infoframe_length
-		+ args.pwr.vendor_infoframe_length;
-	nvif_mthd(disp->disp, 0, &args, size);
-	nv50_audio_enable(encoder, mode);
-}
-
-/******************************************************************************
- * MST
- *****************************************************************************/
-#define nv50_mstm(p) container_of((p), struct nv50_mstm, mgr)
-#define nv50_mstc(p) container_of((p), struct nv50_mstc, connector)
-#define nv50_msto(p) container_of((p), struct nv50_msto, encoder)
-
-struct nv50_mstm {
-	struct nouveau_encoder *outp;
-
-	struct drm_dp_mst_topology_mgr mgr;
-	struct nv50_msto *msto[4];
-
-	bool modified;
-	bool disabled;
-	int links;
-};
-
-struct nv50_mstc {
-	struct nv50_mstm *mstm;
-	struct drm_dp_mst_port *port;
-	struct drm_connector connector;
-
-	struct drm_display_mode *native;
-	struct edid *edid;
-
-	int pbn;
-};
-
-struct nv50_msto {
-	struct drm_encoder encoder;
-
-	struct nv50_head *head;
-	struct nv50_mstc *mstc;
-	bool disabled;
-};
-
-static struct drm_dp_payload *
-nv50_msto_payload(struct nv50_msto *msto)
-{
-	struct nouveau_drm *drm = nouveau_drm(msto->encoder.dev);
-	struct nv50_mstc *mstc = msto->mstc;
-	struct nv50_mstm *mstm = mstc->mstm;
-	int vcpi = mstc->port->vcpi.vcpi, i;
-
-	NV_ATOMIC(drm, "%s: vcpi %d\n", msto->encoder.name, vcpi);
-	for (i = 0; i < mstm->mgr.max_payloads; i++) {
-		struct drm_dp_payload *payload = &mstm->mgr.payloads[i];
-		NV_ATOMIC(drm, "%s: %d: vcpi %d start 0x%02x slots 0x%02x\n",
-			  mstm->outp->base.base.name, i, payload->vcpi,
-			  payload->start_slot, payload->num_slots);
-	}
-
-	for (i = 0; i < mstm->mgr.max_payloads; i++) {
-		struct drm_dp_payload *payload = &mstm->mgr.payloads[i];
-		if (payload->vcpi == vcpi)
-			return payload;
-	}
-
-	return NULL;
-}
-
-static void
-nv50_msto_cleanup(struct nv50_msto *msto)
-{
-	struct nouveau_drm *drm = nouveau_drm(msto->encoder.dev);
-	struct nv50_mstc *mstc = msto->mstc;
-	struct nv50_mstm *mstm = mstc->mstm;
-
-	NV_ATOMIC(drm, "%s: msto cleanup\n", msto->encoder.name);
-	if (mstc->port && mstc->port->vcpi.vcpi > 0 && !nv50_msto_payload(msto))
-		drm_dp_mst_deallocate_vcpi(&mstm->mgr, mstc->port);
-	if (msto->disabled) {
-		msto->mstc = NULL;
-		msto->head = NULL;
-		msto->disabled = false;
-	}
-}
-
-static void
-nv50_msto_prepare(struct nv50_msto *msto)
-{
-	struct nouveau_drm *drm = nouveau_drm(msto->encoder.dev);
-	struct nv50_mstc *mstc = msto->mstc;
-	struct nv50_mstm *mstm = mstc->mstm;
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_sor_dp_mst_vcpi_v0 vcpi;
-	} args = {
-		.base.version = 1,
-		.base.method = NV50_DISP_MTHD_V1_SOR_DP_MST_VCPI,
-		.base.hasht  = mstm->outp->dcb->hasht,
-		.base.hashm  = (0xf0ff & mstm->outp->dcb->hashm) |
-			       (0x0100 << msto->head->base.index),
-	};
-
-	NV_ATOMIC(drm, "%s: msto prepare\n", msto->encoder.name);
-	if (mstc->port && mstc->port->vcpi.vcpi > 0) {
-		struct drm_dp_payload *payload = nv50_msto_payload(msto);
-		if (payload) {
-			args.vcpi.start_slot = payload->start_slot;
-			args.vcpi.num_slots = payload->num_slots;
-			args.vcpi.pbn = mstc->port->vcpi.pbn;
-			args.vcpi.aligned_pbn = mstc->port->vcpi.aligned_pbn;
-		}
-	}
-
-	NV_ATOMIC(drm, "%s: %s: %02x %02x %04x %04x\n",
-		  msto->encoder.name, msto->head->base.base.name,
-		  args.vcpi.start_slot, args.vcpi.num_slots,
-		  args.vcpi.pbn, args.vcpi.aligned_pbn);
-	nvif_mthd(&drm->display->disp, 0, &args, sizeof(args));
-}
-
-static int
-nv50_msto_atomic_check(struct drm_encoder *encoder,
-		       struct drm_crtc_state *crtc_state,
-		       struct drm_connector_state *conn_state)
-{
-	struct nv50_mstc *mstc = nv50_mstc(conn_state->connector);
-	struct nv50_mstm *mstm = mstc->mstm;
-	int bpp = conn_state->connector->display_info.bpc * 3;
-	int slots;
-
-	mstc->pbn = drm_dp_calc_pbn_mode(crtc_state->adjusted_mode.clock, bpp);
-
-	slots = drm_dp_find_vcpi_slots(&mstm->mgr, mstc->pbn);
-	if (slots < 0)
-		return slots;
-
-	return nv50_outp_atomic_check_view(encoder, crtc_state, conn_state,
-					   mstc->native);
-}
-
-static void
-nv50_msto_enable(struct drm_encoder *encoder)
-{
-	struct nv50_head *head = nv50_head(encoder->crtc);
-	struct nv50_msto *msto = nv50_msto(encoder);
-	struct nv50_mstc *mstc = NULL;
-	struct nv50_mstm *mstm = NULL;
-	struct drm_connector *connector;
-	struct drm_connector_list_iter conn_iter;
-	u8 proto, depth;
-	int slots;
-	bool r;
-
-	drm_connector_list_iter_begin(encoder->dev, &conn_iter);
-	drm_for_each_connector_iter(connector, &conn_iter) {
-		if (connector->state->best_encoder == &msto->encoder) {
-			mstc = nv50_mstc(connector);
-			mstm = mstc->mstm;
-			break;
-		}
-	}
-	drm_connector_list_iter_end(&conn_iter);
-
-	if (WARN_ON(!mstc))
-		return;
-
-	slots = drm_dp_find_vcpi_slots(&mstm->mgr, mstc->pbn);
-	r = drm_dp_mst_allocate_vcpi(&mstm->mgr, mstc->port, mstc->pbn, slots);
-	WARN_ON(!r);
-
-	if (!mstm->links++)
-		nv50_outp_acquire(mstm->outp);
-
-	if (mstm->outp->link & 1)
-		proto = 0x8;
-	else
-		proto = 0x9;
-
-	switch (mstc->connector.display_info.bpc) {
-	case  6: depth = 0x2; break;
-	case  8: depth = 0x5; break;
-	case 10:
-	default: depth = 0x6; break;
-	}
-
-	mstm->outp->update(mstm->outp, head->base.index,
-			   &head->base.base.state->adjusted_mode, proto, depth);
-
-	msto->head = head;
-	msto->mstc = mstc;
-	mstm->modified = true;
-}
-
-static void
-nv50_msto_disable(struct drm_encoder *encoder)
-{
-	struct nv50_msto *msto = nv50_msto(encoder);
-	struct nv50_mstc *mstc = msto->mstc;
-	struct nv50_mstm *mstm = mstc->mstm;
-
-	if (mstc->port)
-		drm_dp_mst_reset_vcpi_slots(&mstm->mgr, mstc->port);
-
-	mstm->outp->update(mstm->outp, msto->head->base.index, NULL, 0, 0);
-	mstm->modified = true;
-	if (!--mstm->links)
-		mstm->disabled = true;
-	msto->disabled = true;
-}
-
-static const struct drm_encoder_helper_funcs
-nv50_msto_help = {
-	.disable = nv50_msto_disable,
-	.enable = nv50_msto_enable,
-	.atomic_check = nv50_msto_atomic_check,
-};
-
-static void
-nv50_msto_destroy(struct drm_encoder *encoder)
-{
-	struct nv50_msto *msto = nv50_msto(encoder);
-	drm_encoder_cleanup(&msto->encoder);
-	kfree(msto);
-}
-
-static const struct drm_encoder_funcs
-nv50_msto = {
-	.destroy = nv50_msto_destroy,
-};
-
-static int
-nv50_msto_new(struct drm_device *dev, u32 heads, const char *name, int id,
-	      struct nv50_msto **pmsto)
-{
-	struct nv50_msto *msto;
-	int ret;
-
-	if (!(msto = *pmsto = kzalloc(sizeof(*msto), GFP_KERNEL)))
-		return -ENOMEM;
-
-	ret = drm_encoder_init(dev, &msto->encoder, &nv50_msto,
-			       DRM_MODE_ENCODER_DPMST, "%s-mst-%d", name, id);
-	if (ret) {
-		kfree(*pmsto);
-		*pmsto = NULL;
-		return ret;
-	}
-
-	drm_encoder_helper_add(&msto->encoder, &nv50_msto_help);
-	msto->encoder.possible_crtcs = heads;
-	return 0;
-}
-
-static struct drm_encoder *
-nv50_mstc_atomic_best_encoder(struct drm_connector *connector,
-			      struct drm_connector_state *connector_state)
-{
-	struct nv50_head *head = nv50_head(connector_state->crtc);
-	struct nv50_mstc *mstc = nv50_mstc(connector);
-	if (mstc->port) {
-		struct nv50_mstm *mstm = mstc->mstm;
-		return &mstm->msto[head->base.index]->encoder;
-	}
-	return NULL;
-}
-
-static struct drm_encoder *
-nv50_mstc_best_encoder(struct drm_connector *connector)
-{
-	struct nv50_mstc *mstc = nv50_mstc(connector);
-	if (mstc->port) {
-		struct nv50_mstm *mstm = mstc->mstm;
-		return &mstm->msto[0]->encoder;
-	}
-	return NULL;
-}
-
-static enum drm_mode_status
-nv50_mstc_mode_valid(struct drm_connector *connector,
-		     struct drm_display_mode *mode)
-{
-	return MODE_OK;
-}
-
-static int
-nv50_mstc_get_modes(struct drm_connector *connector)
-{
-	struct nv50_mstc *mstc = nv50_mstc(connector);
-	int ret = 0;
-
-	mstc->edid = drm_dp_mst_get_edid(&mstc->connector, mstc->port->mgr, mstc->port);
-	drm_mode_connector_update_edid_property(&mstc->connector, mstc->edid);
-	if (mstc->edid)
-		ret = drm_add_edid_modes(&mstc->connector, mstc->edid);
-
-	if (!mstc->connector.display_info.bpc)
-		mstc->connector.display_info.bpc = 8;
-
-	if (mstc->native)
-		drm_mode_destroy(mstc->connector.dev, mstc->native);
-	mstc->native = nouveau_conn_native_mode(&mstc->connector);
-	return ret;
-}
-
-static const struct drm_connector_helper_funcs
-nv50_mstc_help = {
-	.get_modes = nv50_mstc_get_modes,
-	.mode_valid = nv50_mstc_mode_valid,
-	.best_encoder = nv50_mstc_best_encoder,
-	.atomic_best_encoder = nv50_mstc_atomic_best_encoder,
-};
-
-static enum drm_connector_status
-nv50_mstc_detect(struct drm_connector *connector, bool force)
-{
-	struct nv50_mstc *mstc = nv50_mstc(connector);
-	if (!mstc->port)
-		return connector_status_disconnected;
-	return drm_dp_mst_detect_port(connector, mstc->port->mgr, mstc->port);
-}
-
-static void
-nv50_mstc_destroy(struct drm_connector *connector)
-{
-	struct nv50_mstc *mstc = nv50_mstc(connector);
-	drm_connector_cleanup(&mstc->connector);
-	kfree(mstc);
-}
-
-static const struct drm_connector_funcs
-nv50_mstc = {
-	.reset = nouveau_conn_reset,
-	.detect = nv50_mstc_detect,
-	.fill_modes = drm_helper_probe_single_connector_modes,
-	.destroy = nv50_mstc_destroy,
-	.atomic_duplicate_state = nouveau_conn_atomic_duplicate_state,
-	.atomic_destroy_state = nouveau_conn_atomic_destroy_state,
-	.atomic_set_property = nouveau_conn_atomic_set_property,
-	.atomic_get_property = nouveau_conn_atomic_get_property,
-};
-
-static int
-nv50_mstc_new(struct nv50_mstm *mstm, struct drm_dp_mst_port *port,
-	      const char *path, struct nv50_mstc **pmstc)
-{
-	struct drm_device *dev = mstm->outp->base.base.dev;
-	struct nv50_mstc *mstc;
-	int ret, i;
-
-	if (!(mstc = *pmstc = kzalloc(sizeof(*mstc), GFP_KERNEL)))
-		return -ENOMEM;
-	mstc->mstm = mstm;
-	mstc->port = port;
-
-	ret = drm_connector_init(dev, &mstc->connector, &nv50_mstc,
-				 DRM_MODE_CONNECTOR_DisplayPort);
-	if (ret) {
-		kfree(*pmstc);
-		*pmstc = NULL;
-		return ret;
-	}
-
-	drm_connector_helper_add(&mstc->connector, &nv50_mstc_help);
-
-	mstc->connector.funcs->reset(&mstc->connector);
-	nouveau_conn_attach_properties(&mstc->connector);
-
-	for (i = 0; i < ARRAY_SIZE(mstm->msto) && mstm->msto[i]; i++)
-		drm_mode_connector_attach_encoder(&mstc->connector, &mstm->msto[i]->encoder);
-
-	drm_object_attach_property(&mstc->connector.base, dev->mode_config.path_property, 0);
-	drm_object_attach_property(&mstc->connector.base, dev->mode_config.tile_property, 0);
-	drm_mode_connector_set_path_property(&mstc->connector, path);
-	return 0;
-}
-
-static void
-nv50_mstm_cleanup(struct nv50_mstm *mstm)
-{
-	struct nouveau_drm *drm = nouveau_drm(mstm->outp->base.base.dev);
-	struct drm_encoder *encoder;
-	int ret;
-
-	NV_ATOMIC(drm, "%s: mstm cleanup\n", mstm->outp->base.base.name);
-	ret = drm_dp_check_act_status(&mstm->mgr);
-
-	ret = drm_dp_update_payload_part2(&mstm->mgr);
-
-	drm_for_each_encoder(encoder, mstm->outp->base.base.dev) {
-		if (encoder->encoder_type == DRM_MODE_ENCODER_DPMST) {
-			struct nv50_msto *msto = nv50_msto(encoder);
-			struct nv50_mstc *mstc = msto->mstc;
-			if (mstc && mstc->mstm == mstm)
-				nv50_msto_cleanup(msto);
-		}
-	}
-
-	mstm->modified = false;
-}
-
-static void
-nv50_mstm_prepare(struct nv50_mstm *mstm)
-{
-	struct nouveau_drm *drm = nouveau_drm(mstm->outp->base.base.dev);
-	struct drm_encoder *encoder;
-	int ret;
-
-	NV_ATOMIC(drm, "%s: mstm prepare\n", mstm->outp->base.base.name);
-	ret = drm_dp_update_payload_part1(&mstm->mgr);
-
-	drm_for_each_encoder(encoder, mstm->outp->base.base.dev) {
-		if (encoder->encoder_type == DRM_MODE_ENCODER_DPMST) {
-			struct nv50_msto *msto = nv50_msto(encoder);
-			struct nv50_mstc *mstc = msto->mstc;
-			if (mstc && mstc->mstm == mstm)
-				nv50_msto_prepare(msto);
-		}
-	}
-
-	if (mstm->disabled) {
-		if (!mstm->links)
-			nv50_outp_release(mstm->outp);
-		mstm->disabled = false;
-	}
-}
-
-static void
-nv50_mstm_hotplug(struct drm_dp_mst_topology_mgr *mgr)
-{
-	struct nv50_mstm *mstm = nv50_mstm(mgr);
-	drm_kms_helper_hotplug_event(mstm->outp->base.base.dev);
-}
-
-static void
-nv50_mstm_destroy_connector(struct drm_dp_mst_topology_mgr *mgr,
-			    struct drm_connector *connector)
-{
-	struct nouveau_drm *drm = nouveau_drm(connector->dev);
-	struct nv50_mstc *mstc = nv50_mstc(connector);
-
-	drm_connector_unregister(&mstc->connector);
-
-	drm_modeset_lock_all(drm->dev);
-	drm_fb_helper_remove_one_connector(&drm->fbcon->helper, &mstc->connector);
-	mstc->port = NULL;
-	drm_modeset_unlock_all(drm->dev);
-
-	drm_connector_unreference(&mstc->connector);
-}
-
-static void
-nv50_mstm_register_connector(struct drm_connector *connector)
-{
-	struct nouveau_drm *drm = nouveau_drm(connector->dev);
-
-	drm_modeset_lock_all(drm->dev);
-	drm_fb_helper_add_one_connector(&drm->fbcon->helper, connector);
-	drm_modeset_unlock_all(drm->dev);
-
-	drm_connector_register(connector);
-}
-
-static struct drm_connector *
-nv50_mstm_add_connector(struct drm_dp_mst_topology_mgr *mgr,
-			struct drm_dp_mst_port *port, const char *path)
-{
-	struct nv50_mstm *mstm = nv50_mstm(mgr);
-	struct nv50_mstc *mstc;
-	int ret;
-
-	ret = nv50_mstc_new(mstm, port, path, &mstc);
-	if (ret) {
-		if (mstc)
-			mstc->connector.funcs->destroy(&mstc->connector);
-		return NULL;
-	}
-
-	return &mstc->connector;
-}
-
-static const struct drm_dp_mst_topology_cbs
-nv50_mstm = {
-	.add_connector = nv50_mstm_add_connector,
-	.register_connector = nv50_mstm_register_connector,
-	.destroy_connector = nv50_mstm_destroy_connector,
-	.hotplug = nv50_mstm_hotplug,
-};
-
-void
-nv50_mstm_service(struct nv50_mstm *mstm)
-{
-	struct drm_dp_aux *aux = mstm ? mstm->mgr.aux : NULL;
-	bool handled = true;
-	int ret;
-	u8 esi[8] = {};
-
-	if (!aux)
-		return;
-
-	while (handled) {
-		ret = drm_dp_dpcd_read(aux, DP_SINK_COUNT_ESI, esi, 8);
-		if (ret != 8) {
-			drm_dp_mst_topology_mgr_set_mst(&mstm->mgr, false);
-			return;
-		}
-
-		drm_dp_mst_hpd_irq(&mstm->mgr, esi, &handled);
-		if (!handled)
-			break;
-
-		drm_dp_dpcd_write(aux, DP_SINK_COUNT_ESI + 1, &esi[1], 3);
-	}
-}
-
-void
-nv50_mstm_remove(struct nv50_mstm *mstm)
-{
-	if (mstm)
-		drm_dp_mst_topology_mgr_set_mst(&mstm->mgr, false);
-}
-
-static int
-nv50_mstm_enable(struct nv50_mstm *mstm, u8 dpcd, int state)
-{
-	struct nouveau_encoder *outp = mstm->outp;
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_sor_dp_mst_link_v0 mst;
-	} args = {
-		.base.version = 1,
-		.base.method = NV50_DISP_MTHD_V1_SOR_DP_MST_LINK,
-		.base.hasht = outp->dcb->hasht,
-		.base.hashm = outp->dcb->hashm,
-		.mst.state = state,
-	};
-	struct nouveau_drm *drm = nouveau_drm(outp->base.base.dev);
-	struct nvif_object *disp = &drm->display->disp;
-	int ret;
-
-	if (dpcd >= 0x12) {
-		ret = drm_dp_dpcd_readb(mstm->mgr.aux, DP_MSTM_CTRL, &dpcd);
-		if (ret < 0)
-			return ret;
-
-		dpcd &= ~DP_MST_EN;
-		if (state)
-			dpcd |= DP_MST_EN;
-
-		ret = drm_dp_dpcd_writeb(mstm->mgr.aux, DP_MSTM_CTRL, dpcd);
-		if (ret < 0)
-			return ret;
-	}
-
-	return nvif_mthd(disp, 0, &args, sizeof(args));
-}
-
-int
-nv50_mstm_detect(struct nv50_mstm *mstm, u8 dpcd[8], int allow)
-{
-	int ret, state = 0;
-
-	if (!mstm)
-		return 0;
-
-	if (dpcd[0] >= 0x12) {
-		ret = drm_dp_dpcd_readb(mstm->mgr.aux, DP_MSTM_CAP, &dpcd[1]);
-		if (ret < 0)
-			return ret;
-
-		if (!(dpcd[1] & DP_MST_CAP))
-			dpcd[0] = 0x11;
-		else
-			state = allow;
-	}
-
-	ret = nv50_mstm_enable(mstm, dpcd[0], state);
-	if (ret)
-		return ret;
-
-	ret = drm_dp_mst_topology_mgr_set_mst(&mstm->mgr, state);
-	if (ret)
-		return nv50_mstm_enable(mstm, dpcd[0], 0);
-
-	return mstm->mgr.mst_state;
-}
-
-static void
-nv50_mstm_fini(struct nv50_mstm *mstm)
-{
-	if (mstm && mstm->mgr.mst_state)
-		drm_dp_mst_topology_mgr_suspend(&mstm->mgr);
-}
-
-static void
-nv50_mstm_init(struct nv50_mstm *mstm)
-{
-	if (mstm && mstm->mgr.mst_state)
-		drm_dp_mst_topology_mgr_resume(&mstm->mgr);
-}
-
-static void
-nv50_mstm_del(struct nv50_mstm **pmstm)
-{
-	struct nv50_mstm *mstm = *pmstm;
-	if (mstm) {
-		kfree(*pmstm);
-		*pmstm = NULL;
-	}
-}
-
-static int
-nv50_mstm_new(struct nouveau_encoder *outp, struct drm_dp_aux *aux, int aux_max,
-	      int conn_base_id, struct nv50_mstm **pmstm)
-{
-	const int max_payloads = hweight8(outp->dcb->heads);
-	struct drm_device *dev = outp->base.base.dev;
-	struct nv50_mstm *mstm;
-	int ret, i;
-	u8 dpcd;
-
-	/* This is a workaround for some monitors not functioning
-	 * correctly in MST mode on initial module load.  I think
-	 * some bad interaction with the VBIOS may be responsible.
-	 *
-	 * A good ol' off and on again seems to work here ;)
-	 */
-	ret = drm_dp_dpcd_readb(aux, DP_DPCD_REV, &dpcd);
-	if (ret >= 0 && dpcd >= 0x12)
-		drm_dp_dpcd_writeb(aux, DP_MSTM_CTRL, 0);
-
-	if (!(mstm = *pmstm = kzalloc(sizeof(*mstm), GFP_KERNEL)))
-		return -ENOMEM;
-	mstm->outp = outp;
-	mstm->mgr.cbs = &nv50_mstm;
-
-	ret = drm_dp_mst_topology_mgr_init(&mstm->mgr, dev, aux, aux_max,
-					   max_payloads, conn_base_id);
-	if (ret)
-		return ret;
-
-	for (i = 0; i < max_payloads; i++) {
-		ret = nv50_msto_new(dev, outp->dcb->heads, outp->base.base.name,
-				    i, &mstm->msto[i]);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
-}
-
-/******************************************************************************
- * SOR
- *****************************************************************************/
-static void
-nv50_sor_update(struct nouveau_encoder *nv_encoder, u8 head,
-		struct drm_display_mode *mode, u8 proto, u8 depth)
-{
-	struct nv50_dmac *core = &nv50_mast(nv_encoder->base.base.dev)->base;
-	u32 *push;
-
-	if (!mode) {
-		nv_encoder->ctrl &= ~BIT(head);
-		if (!(nv_encoder->ctrl & 0x0000000f))
-			nv_encoder->ctrl = 0;
-	} else {
-		nv_encoder->ctrl |= proto << 8;
-		nv_encoder->ctrl |= BIT(head);
-	}
-
-	if ((push = evo_wait(core, 6))) {
-		if (core->base.user.oclass < GF110_DISP_CORE_CHANNEL_DMA) {
-			if (mode) {
-				if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-					nv_encoder->ctrl |= 0x00001000;
-				if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-					nv_encoder->ctrl |= 0x00002000;
-				nv_encoder->ctrl |= depth << 16;
-			}
-			evo_mthd(push, 0x0600 + (nv_encoder->or * 0x40), 1);
-		} else {
-			if (mode) {
-				u32 magic = 0x31ec6000 | (head << 25);
-				u32 syncs = 0x00000001;
-				if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-					syncs |= 0x00000008;
-				if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-					syncs |= 0x00000010;
-				if (mode->flags & DRM_MODE_FLAG_INTERLACE)
-					magic |= 0x00000001;
-
-				evo_mthd(push, 0x0404 + (head * 0x300), 2);
-				evo_data(push, syncs | (depth << 6));
-				evo_data(push, magic);
-			}
-			evo_mthd(push, 0x0200 + (nv_encoder->or * 0x20), 1);
-		}
-		evo_data(push, nv_encoder->ctrl);
-		evo_kick(push, core);
-	}
-}
-
-static void
-nv50_sor_disable(struct drm_encoder *encoder)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_crtc *nv_crtc = nouveau_crtc(nv_encoder->crtc);
-
-	nv_encoder->crtc = NULL;
-
-	if (nv_crtc) {
-		struct nvkm_i2c_aux *aux = nv_encoder->aux;
-		u8 pwr;
-
-		if (aux) {
-			int ret = nvkm_rdaux(aux, DP_SET_POWER, &pwr, 1);
-			if (ret == 0) {
-				pwr &= ~DP_SET_POWER_MASK;
-				pwr |=  DP_SET_POWER_D3;
-				nvkm_wraux(aux, DP_SET_POWER, &pwr, 1);
-			}
-		}
-
-		nv_encoder->update(nv_encoder, nv_crtc->index, NULL, 0, 0);
-		nv50_audio_disable(encoder, nv_crtc);
-		nv50_hdmi_disable(&nv_encoder->base.base, nv_crtc);
-		nv50_outp_release(nv_encoder);
-	}
-}
-
-static void
-nv50_sor_enable(struct drm_encoder *encoder)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
-	struct drm_display_mode *mode = &nv_crtc->base.state->adjusted_mode;
-	struct {
-		struct nv50_disp_mthd_v1 base;
-		struct nv50_disp_sor_lvds_script_v0 lvds;
-	} lvds = {
-		.base.version = 1,
-		.base.method  = NV50_DISP_MTHD_V1_SOR_LVDS_SCRIPT,
-		.base.hasht   = nv_encoder->dcb->hasht,
-		.base.hashm   = nv_encoder->dcb->hashm,
-	};
-	struct nv50_disp *disp = nv50_disp(encoder->dev);
-	struct drm_device *dev = encoder->dev;
-	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct nouveau_connector *nv_connector;
-	struct nvbios *bios = &drm->vbios;
-	u8 proto = 0xf;
-	u8 depth = 0x0;
-
-	nv_connector = nouveau_encoder_connector_get(nv_encoder);
-	nv_encoder->crtc = encoder->crtc;
-	nv50_outp_acquire(nv_encoder);
-
-	switch (nv_encoder->dcb->type) {
-	case DCB_OUTPUT_TMDS:
-		if (nv_encoder->link & 1) {
-			proto = 0x1;
-			/* Only enable dual-link if:
-			 *  - Need to (i.e. rate > 165MHz)
-			 *  - DCB says we can
-			 *  - Not an HDMI monitor, since there's no dual-link
-			 *    on HDMI.
-			 */
-			if (mode->clock >= 165000 &&
-			    nv_encoder->dcb->duallink_possible &&
-			    !drm_detect_hdmi_monitor(nv_connector->edid))
-				proto |= 0x4;
-		} else {
-			proto = 0x2;
-		}
-
-		nv50_hdmi_enable(&nv_encoder->base.base, mode);
-		break;
-	case DCB_OUTPUT_LVDS:
-		proto = 0x0;
-
-		if (bios->fp_no_ddc) {
-			if (bios->fp.dual_link)
-				lvds.lvds.script |= 0x0100;
-			if (bios->fp.if_is_24bit)
-				lvds.lvds.script |= 0x0200;
-		} else {
-			if (nv_connector->type == DCB_CONNECTOR_LVDS_SPWG) {
-				if (((u8 *)nv_connector->edid)[121] == 2)
-					lvds.lvds.script |= 0x0100;
-			} else
-			if (mode->clock >= bios->fp.duallink_transition_clk) {
-				lvds.lvds.script |= 0x0100;
-			}
-
-			if (lvds.lvds.script & 0x0100) {
-				if (bios->fp.strapless_is_24bit & 2)
-					lvds.lvds.script |= 0x0200;
-			} else {
-				if (bios->fp.strapless_is_24bit & 1)
-					lvds.lvds.script |= 0x0200;
-			}
-
-			if (nv_connector->base.display_info.bpc == 8)
-				lvds.lvds.script |= 0x0200;
-		}
-
-		nvif_mthd(disp->disp, 0, &lvds, sizeof(lvds));
-		break;
-	case DCB_OUTPUT_DP:
-		if (nv_connector->base.display_info.bpc == 6)
-			depth = 0x2;
-		else
-		if (nv_connector->base.display_info.bpc == 8)
-			depth = 0x5;
-		else
-			depth = 0x6;
-
-		if (nv_encoder->link & 1)
-			proto = 0x8;
-		else
-			proto = 0x9;
-
-		nv50_audio_enable(encoder, mode);
-		break;
-	default:
-		BUG();
-		break;
-	}
-
-	nv_encoder->update(nv_encoder, nv_crtc->index, mode, proto, depth);
-}
-
-static const struct drm_encoder_helper_funcs
-nv50_sor_help = {
-	.atomic_check = nv50_outp_atomic_check,
-	.enable = nv50_sor_enable,
-	.disable = nv50_sor_disable,
-};
-
-static void
-nv50_sor_destroy(struct drm_encoder *encoder)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	nv50_mstm_del(&nv_encoder->dp.mstm);
-	drm_encoder_cleanup(encoder);
-	kfree(encoder);
-}
-
-static const struct drm_encoder_funcs
-nv50_sor_func = {
-	.destroy = nv50_sor_destroy,
-};
-
-static int
-nv50_sor_create(struct drm_connector *connector, struct dcb_output *dcbe)
-{
-	struct nouveau_connector *nv_connector = nouveau_connector(connector);
-	struct nouveau_drm *drm = nouveau_drm(connector->dev);
-	struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device);
-	struct nouveau_encoder *nv_encoder;
-	struct drm_encoder *encoder;
-	int type, ret;
-
-	switch (dcbe->type) {
-	case DCB_OUTPUT_LVDS: type = DRM_MODE_ENCODER_LVDS; break;
-	case DCB_OUTPUT_TMDS:
-	case DCB_OUTPUT_DP:
-	default:
-		type = DRM_MODE_ENCODER_TMDS;
-		break;
-	}
-
-	nv_encoder = kzalloc(sizeof(*nv_encoder), GFP_KERNEL);
-	if (!nv_encoder)
-		return -ENOMEM;
-	nv_encoder->dcb = dcbe;
-	nv_encoder->update = nv50_sor_update;
-
-	encoder = to_drm_encoder(nv_encoder);
-	encoder->possible_crtcs = dcbe->heads;
-	encoder->possible_clones = 0;
-	drm_encoder_init(connector->dev, encoder, &nv50_sor_func, type,
-			 "sor-%04x-%04x", dcbe->hasht, dcbe->hashm);
-	drm_encoder_helper_add(encoder, &nv50_sor_help);
-
-	drm_mode_connector_attach_encoder(connector, encoder);
-
-	if (dcbe->type == DCB_OUTPUT_DP) {
-		struct nv50_disp *disp = nv50_disp(encoder->dev);
-		struct nvkm_i2c_aux *aux =
-			nvkm_i2c_aux_find(i2c, dcbe->i2c_index);
-		if (aux) {
-			if (disp->disp->oclass < GF110_DISP) {
-				/* HW has no support for address-only
-				 * transactions, so we're required to
-				 * use custom I2C-over-AUX code.
-				 */
-				nv_encoder->i2c = &aux->i2c;
-			} else {
-				nv_encoder->i2c = &nv_connector->aux.ddc;
-			}
-			nv_encoder->aux = aux;
-		}
-
-		/*TODO: Use DP Info Table to check for support. */
-		if (disp->disp->oclass >= GF110_DISP) {
-			ret = nv50_mstm_new(nv_encoder, &nv_connector->aux, 16,
-					    nv_connector->base.base.id,
-					    &nv_encoder->dp.mstm);
-			if (ret)
-				return ret;
-		}
-	} else {
-		struct nvkm_i2c_bus *bus =
-			nvkm_i2c_bus_find(i2c, dcbe->i2c_index);
-		if (bus)
-			nv_encoder->i2c = &bus->i2c;
-	}
-
-	return 0;
-}
-
-/******************************************************************************
- * PIOR
- *****************************************************************************/
-static int
-nv50_pior_atomic_check(struct drm_encoder *encoder,
-		       struct drm_crtc_state *crtc_state,
-		       struct drm_connector_state *conn_state)
-{
-	int ret = nv50_outp_atomic_check(encoder, crtc_state, conn_state);
-	if (ret)
-		return ret;
-	crtc_state->adjusted_mode.clock *= 2;
-	return 0;
-}
-
-static void
-nv50_pior_disable(struct drm_encoder *encoder)
-{
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nv50_mast *mast = nv50_mast(encoder->dev);
-	const int or = nv_encoder->or;
-	u32 *push;
-
-	if (nv_encoder->crtc) {
-		push = evo_wait(mast, 4);
-		if (push) {
-			if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
-				evo_mthd(push, 0x0700 + (or * 0x040), 1);
-				evo_data(push, 0x00000000);
-			}
-			evo_kick(push, mast);
-		}
-	}
-
-	nv_encoder->crtc = NULL;
-	nv50_outp_release(nv_encoder);
-}
-
-static void
-nv50_pior_enable(struct drm_encoder *encoder)
-{
-	struct nv50_mast *mast = nv50_mast(encoder->dev);
-	struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
-	struct nouveau_crtc *nv_crtc = nouveau_crtc(encoder->crtc);
-	struct nouveau_connector *nv_connector;
-	struct drm_display_mode *mode = &nv_crtc->base.state->adjusted_mode;
-	u8 owner = 1 << nv_crtc->index;
-	u8 proto, depth;
-	u32 *push;
-
-	nv50_outp_acquire(nv_encoder);
-
-	nv_connector = nouveau_encoder_connector_get(nv_encoder);
-	switch (nv_connector->base.display_info.bpc) {
-	case 10: depth = 0x6; break;
-	case  8: depth = 0x5; break;
-	case  6: depth = 0x2; break;
-	default: depth = 0x0; break;
-	}
-
-	switch (nv_encoder->dcb->type) {
-	case DCB_OUTPUT_TMDS:
-	case DCB_OUTPUT_DP:
-		proto = 0x0;
-		break;
-	default:
-		BUG();
-		break;
-	}
-
-	push = evo_wait(mast, 8);
-	if (push) {
-		if (nv50_vers(mast) < GF110_DISP_CORE_CHANNEL_DMA) {
-			u32 ctrl = (depth << 16) | (proto << 8) | owner;
-			if (mode->flags & DRM_MODE_FLAG_NHSYNC)
-				ctrl |= 0x00001000;
-			if (mode->flags & DRM_MODE_FLAG_NVSYNC)
-				ctrl |= 0x00002000;
-			evo_mthd(push, 0x0700 + (nv_encoder->or * 0x040), 1);
-			evo_data(push, ctrl);
-		}
-
-		evo_kick(push, mast);
-	}
-
-	nv_encoder->crtc = encoder->crtc;
-}
-
-static const struct drm_encoder_helper_funcs
-nv50_pior_help = {
-	.atomic_check = nv50_pior_atomic_check,
-	.enable = nv50_pior_enable,
-	.disable = nv50_pior_disable,
-};
-
-static void
-nv50_pior_destroy(struct drm_encoder *encoder)
-{
-	drm_encoder_cleanup(encoder);
-	kfree(encoder);
-}
-
-static const struct drm_encoder_funcs
-nv50_pior_func = {
-	.destroy = nv50_pior_destroy,
-};
-
-static int
-nv50_pior_create(struct drm_connector *connector, struct dcb_output *dcbe)
-{
-	struct nouveau_connector *nv_connector = nouveau_connector(connector);
-	struct nouveau_drm *drm = nouveau_drm(connector->dev);
-	struct nvkm_i2c *i2c = nvxx_i2c(&drm->client.device);
-	struct nvkm_i2c_bus *bus = NULL;
-	struct nvkm_i2c_aux *aux = NULL;
-	struct i2c_adapter *ddc;
-	struct nouveau_encoder *nv_encoder;
-	struct drm_encoder *encoder;
-	int type;
-
-	switch (dcbe->type) {
-	case DCB_OUTPUT_TMDS:
-		bus  = nvkm_i2c_bus_find(i2c, NVKM_I2C_BUS_EXT(dcbe->extdev));
-		ddc  = bus ? &bus->i2c : NULL;
-		type = DRM_MODE_ENCODER_TMDS;
-		break;
-	case DCB_OUTPUT_DP:
-		aux  = nvkm_i2c_aux_find(i2c, NVKM_I2C_AUX_EXT(dcbe->extdev));
-		ddc  = aux ? &nv_connector->aux.ddc : NULL;
-		type = DRM_MODE_ENCODER_TMDS;
-		break;
-	default:
-		return -ENODEV;
-	}
-
-	nv_encoder = kzalloc(sizeof(*nv_encoder), GFP_KERNEL);
-	if (!nv_encoder)
-		return -ENOMEM;
-	nv_encoder->dcb = dcbe;
-	nv_encoder->i2c = ddc;
-	nv_encoder->aux = aux;
-
-	encoder = to_drm_encoder(nv_encoder);
-	encoder->possible_crtcs = dcbe->heads;
-	encoder->possible_clones = 0;
-	drm_encoder_init(connector->dev, encoder, &nv50_pior_func, type,
-			 "pior-%04x-%04x", dcbe->hasht, dcbe->hashm);
-	drm_encoder_helper_add(encoder, &nv50_pior_help);
-
-	drm_mode_connector_attach_encoder(connector, encoder);
-	return 0;
-}
-
-/******************************************************************************
- * Atomic
- *****************************************************************************/
-
-static void
-nv50_disp_atomic_commit_core(struct nouveau_drm *drm, u32 interlock)
-{
-	struct nv50_disp *disp = nv50_disp(drm->dev);
-	struct nv50_dmac *core = &disp->mast.base;
-	struct nv50_mstm *mstm;
-	struct drm_encoder *encoder;
-	u32 *push;
-
-	NV_ATOMIC(drm, "commit core %08x\n", interlock);
-
-	drm_for_each_encoder(encoder, drm->dev) {
-		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
-			mstm = nouveau_encoder(encoder)->dp.mstm;
-			if (mstm && mstm->modified)
-				nv50_mstm_prepare(mstm);
-		}
-	}
-
-	if ((push = evo_wait(core, 5))) {
-		evo_mthd(push, 0x0084, 1);
-		evo_data(push, 0x80000000);
-		evo_mthd(push, 0x0080, 2);
-		evo_data(push, interlock);
-		evo_data(push, 0x00000000);
-		nouveau_bo_wr32(disp->sync, 0, 0x00000000);
-		evo_kick(push, core);
-		if (nvif_msec(&drm->client.device, 2000ULL,
-			if (nouveau_bo_rd32(disp->sync, 0))
-				break;
-			usleep_range(1, 2);
-		) < 0)
-			NV_ERROR(drm, "EVO timeout\n");
-	}
-
-	drm_for_each_encoder(encoder, drm->dev) {
-		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
-			mstm = nouveau_encoder(encoder)->dp.mstm;
-			if (mstm && mstm->modified)
-				nv50_mstm_cleanup(mstm);
-		}
-	}
-}
-
-static void
-nv50_disp_atomic_commit_tail(struct drm_atomic_state *state)
-{
-	struct drm_device *dev = state->dev;
-	struct drm_crtc_state *new_crtc_state, *old_crtc_state;
-	struct drm_crtc *crtc;
-	struct drm_plane_state *new_plane_state;
-	struct drm_plane *plane;
-	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct nv50_disp *disp = nv50_disp(dev);
-	struct nv50_atom *atom = nv50_atom(state);
-	struct nv50_outp_atom *outp, *outt;
-	u32 interlock_core = 0;
-	u32 interlock_chan = 0;
-	int i;
-
-	NV_ATOMIC(drm, "commit %d %d\n", atom->lock_core, atom->flush_disable);
-	drm_atomic_helper_wait_for_fences(dev, state, false);
-	drm_atomic_helper_wait_for_dependencies(state);
-	drm_atomic_helper_update_legacy_modeset_state(dev, state);
-
-	if (atom->lock_core)
-		mutex_lock(&disp->mutex);
-
-	/* Disable head(s). */
-	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
-		struct nv50_head_atom *asyh = nv50_head_atom(new_crtc_state);
-		struct nv50_head *head = nv50_head(crtc);
-
-		NV_ATOMIC(drm, "%s: clr %04x (set %04x)\n", crtc->name,
-			  asyh->clr.mask, asyh->set.mask);
-		if (old_crtc_state->active && !new_crtc_state->active)
-			drm_crtc_vblank_off(crtc);
-
-		if (asyh->clr.mask) {
-			nv50_head_flush_clr(head, asyh, atom->flush_disable);
-			interlock_core |= 1;
-		}
-	}
-
-	/* Disable plane(s). */
-	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
-		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
-		struct nv50_wndw *wndw = nv50_wndw(plane);
-
-		NV_ATOMIC(drm, "%s: clr %02x (set %02x)\n", plane->name,
-			  asyw->clr.mask, asyw->set.mask);
-		if (!asyw->clr.mask)
-			continue;
-
-		interlock_chan |= nv50_wndw_flush_clr(wndw, interlock_core,
-						      atom->flush_disable,
-						      asyw);
-	}
-
-	/* Disable output path(s). */
-	list_for_each_entry(outp, &atom->outp, head) {
-		const struct drm_encoder_helper_funcs *help;
-		struct drm_encoder *encoder;
-
-		encoder = outp->encoder;
-		help = encoder->helper_private;
-
-		NV_ATOMIC(drm, "%s: clr %02x (set %02x)\n", encoder->name,
-			  outp->clr.mask, outp->set.mask);
-
-		if (outp->clr.mask) {
-			help->disable(encoder);
-			interlock_core |= 1;
-			if (outp->flush_disable) {
-				nv50_disp_atomic_commit_core(drm, interlock_chan);
-				interlock_core = 0;
-				interlock_chan = 0;
-			}
-		}
-	}
-
-	/* Flush disable. */
-	if (interlock_core) {
-		if (atom->flush_disable) {
-			nv50_disp_atomic_commit_core(drm, interlock_chan);
-			interlock_core = 0;
-			interlock_chan = 0;
-		}
-	}
-
-	/* Update output path(s). */
-	list_for_each_entry_safe(outp, outt, &atom->outp, head) {
-		const struct drm_encoder_helper_funcs *help;
-		struct drm_encoder *encoder;
-
-		encoder = outp->encoder;
-		help = encoder->helper_private;
-
-		NV_ATOMIC(drm, "%s: set %02x (clr %02x)\n", encoder->name,
-			  outp->set.mask, outp->clr.mask);
-
-		if (outp->set.mask) {
-			help->enable(encoder);
-			interlock_core = 1;
-		}
-
-		list_del(&outp->head);
-		kfree(outp);
-	}
-
-	/* Update head(s). */
-	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
-		struct nv50_head_atom *asyh = nv50_head_atom(new_crtc_state);
-		struct nv50_head *head = nv50_head(crtc);
-
-		NV_ATOMIC(drm, "%s: set %04x (clr %04x)\n", crtc->name,
-			  asyh->set.mask, asyh->clr.mask);
-
-		if (asyh->set.mask) {
-			nv50_head_flush_set(head, asyh);
-			interlock_core = 1;
-		}
-
-		if (new_crtc_state->active) {
-			if (!old_crtc_state->active)
-				drm_crtc_vblank_on(crtc);
-			if (new_crtc_state->event)
-				drm_crtc_vblank_get(crtc);
-		}
-	}
-
-	/* Update plane(s). */
-	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
-		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
-		struct nv50_wndw *wndw = nv50_wndw(plane);
-
-		NV_ATOMIC(drm, "%s: set %02x (clr %02x)\n", plane->name,
-			  asyw->set.mask, asyw->clr.mask);
-		if ( !asyw->set.mask &&
-		    (!asyw->clr.mask || atom->flush_disable))
-			continue;
-
-		interlock_chan |= nv50_wndw_flush_set(wndw, interlock_core, asyw);
-	}
-
-	/* Flush update. */
-	if (interlock_core) {
-		if (!interlock_chan && atom->state.legacy_cursor_update) {
-			u32 *push = evo_wait(&disp->mast, 2);
-			if (push) {
-				evo_mthd(push, 0x0080, 1);
-				evo_data(push, 0x00000000);
-				evo_kick(push, &disp->mast);
-			}
-		} else {
-			nv50_disp_atomic_commit_core(drm, interlock_chan);
-		}
-	}
-
-	if (atom->lock_core)
-		mutex_unlock(&disp->mutex);
-
-	/* Wait for HW to signal completion. */
-	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
-		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
-		struct nv50_wndw *wndw = nv50_wndw(plane);
-		int ret = nv50_wndw_wait_armed(wndw, asyw);
-		if (ret)
-			NV_ERROR(drm, "%s: timeout\n", plane->name);
-	}
-
-	for_each_new_crtc_in_state(state, crtc, new_crtc_state, i) {
-		if (new_crtc_state->event) {
-			unsigned long flags;
-			/* Get correct count/ts if racing with vblank irq */
-			if (new_crtc_state->active)
-				drm_crtc_accurate_vblank_count(crtc);
-			spin_lock_irqsave(&crtc->dev->event_lock, flags);
-			drm_crtc_send_vblank_event(crtc, new_crtc_state->event);
-			spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
-
-			new_crtc_state->event = NULL;
-			if (new_crtc_state->active)
-				drm_crtc_vblank_put(crtc);
-		}
-	}
-
-	drm_atomic_helper_commit_hw_done(state);
-	drm_atomic_helper_cleanup_planes(dev, state);
-	drm_atomic_helper_commit_cleanup_done(state);
-	drm_atomic_state_put(state);
-}
-
-static void
-nv50_disp_atomic_commit_work(struct work_struct *work)
-{
-	struct drm_atomic_state *state =
-		container_of(work, typeof(*state), commit_work);
-	nv50_disp_atomic_commit_tail(state);
-}
-
-static int
-nv50_disp_atomic_commit(struct drm_device *dev,
-			struct drm_atomic_state *state, bool nonblock)
-{
-	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct nv50_disp *disp = nv50_disp(dev);
-	struct drm_plane_state *new_plane_state;
-	struct drm_plane *plane;
-	struct drm_crtc *crtc;
-	bool active = false;
-	int ret, i;
-
-	ret = pm_runtime_get_sync(dev->dev);
-	if (ret < 0 && ret != -EACCES)
-		return ret;
-
-	ret = drm_atomic_helper_setup_commit(state, nonblock);
-	if (ret)
-		goto done;
-
-	INIT_WORK(&state->commit_work, nv50_disp_atomic_commit_work);
-
-	ret = drm_atomic_helper_prepare_planes(dev, state);
-	if (ret)
-		goto done;
-
-	if (!nonblock) {
-		ret = drm_atomic_helper_wait_for_fences(dev, state, true);
-		if (ret)
-			goto err_cleanup;
-	}
-
-	ret = drm_atomic_helper_swap_state(state, true);
-	if (ret)
-		goto err_cleanup;
-
-	for_each_new_plane_in_state(state, plane, new_plane_state, i) {
-		struct nv50_wndw_atom *asyw = nv50_wndw_atom(new_plane_state);
-		struct nv50_wndw *wndw = nv50_wndw(plane);
-
-		if (asyw->set.image) {
-			asyw->ntfy.handle = wndw->dmac->sync.handle;
-			asyw->ntfy.offset = wndw->ntfy;
-			asyw->ntfy.awaken = false;
-			asyw->set.ntfy = true;
-			nouveau_bo_wr32(disp->sync, wndw->ntfy / 4, 0x00000000);
-			wndw->ntfy ^= 0x10;
-		}
-	}
-
-	drm_atomic_state_get(state);
-
-	if (nonblock)
-		queue_work(system_unbound_wq, &state->commit_work);
-	else
-		nv50_disp_atomic_commit_tail(state);
-
-	drm_for_each_crtc(crtc, dev) {
-		if (crtc->state->enable) {
-			if (!drm->have_disp_power_ref) {
-				drm->have_disp_power_ref = true;
-				return 0;
-			}
-			active = true;
-			break;
-		}
-	}
-
-	if (!active && drm->have_disp_power_ref) {
-		pm_runtime_put_autosuspend(dev->dev);
-		drm->have_disp_power_ref = false;
-	}
-
-err_cleanup:
-	if (ret)
-		drm_atomic_helper_cleanup_planes(dev, state);
-done:
-	pm_runtime_put_autosuspend(dev->dev);
-	return ret;
-}
-
-static struct nv50_outp_atom *
-nv50_disp_outp_atomic_add(struct nv50_atom *atom, struct drm_encoder *encoder)
-{
-	struct nv50_outp_atom *outp;
-
-	list_for_each_entry(outp, &atom->outp, head) {
-		if (outp->encoder == encoder)
-			return outp;
-	}
-
-	outp = kzalloc(sizeof(*outp), GFP_KERNEL);
-	if (!outp)
-		return ERR_PTR(-ENOMEM);
-
-	list_add(&outp->head, &atom->outp);
-	outp->encoder = encoder;
-	return outp;
-}
-
-static int
-nv50_disp_outp_atomic_check_clr(struct nv50_atom *atom,
-				struct drm_connector_state *old_connector_state)
-{
-	struct drm_encoder *encoder = old_connector_state->best_encoder;
-	struct drm_crtc_state *old_crtc_state, *new_crtc_state;
-	struct drm_crtc *crtc;
-	struct nv50_outp_atom *outp;
-
-	if (!(crtc = old_connector_state->crtc))
-		return 0;
-
-	old_crtc_state = drm_atomic_get_old_crtc_state(&atom->state, crtc);
-	new_crtc_state = drm_atomic_get_new_crtc_state(&atom->state, crtc);
-	if (old_crtc_state->active && drm_atomic_crtc_needs_modeset(new_crtc_state)) {
-		outp = nv50_disp_outp_atomic_add(atom, encoder);
-		if (IS_ERR(outp))
-			return PTR_ERR(outp);
-
-		if (outp->encoder->encoder_type == DRM_MODE_ENCODER_DPMST) {
-			outp->flush_disable = true;
-			atom->flush_disable = true;
-		}
-		outp->clr.ctrl = true;
-		atom->lock_core = true;
-	}
-
-	return 0;
-}
-
-static int
-nv50_disp_outp_atomic_check_set(struct nv50_atom *atom,
-				struct drm_connector_state *connector_state)
-{
-	struct drm_encoder *encoder = connector_state->best_encoder;
-	struct drm_crtc_state *new_crtc_state;
-	struct drm_crtc *crtc;
-	struct nv50_outp_atom *outp;
-
-	if (!(crtc = connector_state->crtc))
-		return 0;
-
-	new_crtc_state = drm_atomic_get_new_crtc_state(&atom->state, crtc);
-	if (new_crtc_state->active && drm_atomic_crtc_needs_modeset(new_crtc_state)) {
-		outp = nv50_disp_outp_atomic_add(atom, encoder);
-		if (IS_ERR(outp))
-			return PTR_ERR(outp);
-
-		outp->set.ctrl = true;
-		atom->lock_core = true;
-	}
-
-	return 0;
-}
-
-static int
-nv50_disp_atomic_check(struct drm_device *dev, struct drm_atomic_state *state)
-{
-	struct nv50_atom *atom = nv50_atom(state);
-	struct drm_connector_state *old_connector_state, *new_connector_state;
-	struct drm_connector *connector;
-	int ret, i;
-
-	ret = drm_atomic_helper_check(dev, state);
-	if (ret)
-		return ret;
-
-	for_each_oldnew_connector_in_state(state, connector, old_connector_state, new_connector_state, i) {
-		ret = nv50_disp_outp_atomic_check_clr(atom, old_connector_state);
-		if (ret)
-			return ret;
-
-		ret = nv50_disp_outp_atomic_check_set(atom, new_connector_state);
-		if (ret)
-			return ret;
-	}
-
-	return 0;
-}
-
-static void
-nv50_disp_atomic_state_clear(struct drm_atomic_state *state)
-{
-	struct nv50_atom *atom = nv50_atom(state);
-	struct nv50_outp_atom *outp, *outt;
-
-	list_for_each_entry_safe(outp, outt, &atom->outp, head) {
-		list_del(&outp->head);
-		kfree(outp);
-	}
-
-	drm_atomic_state_default_clear(state);
-}
-
-static void
-nv50_disp_atomic_state_free(struct drm_atomic_state *state)
-{
-	struct nv50_atom *atom = nv50_atom(state);
-	drm_atomic_state_default_release(&atom->state);
-	kfree(atom);
-}
-
-static struct drm_atomic_state *
-nv50_disp_atomic_state_alloc(struct drm_device *dev)
-{
-	struct nv50_atom *atom;
-	if (!(atom = kzalloc(sizeof(*atom), GFP_KERNEL)) ||
-	    drm_atomic_state_init(dev, &atom->state) < 0) {
-		kfree(atom);
-		return NULL;
-	}
-	INIT_LIST_HEAD(&atom->outp);
-	return &atom->state;
-}
-
-static const struct drm_mode_config_funcs
-nv50_disp_func = {
-	.fb_create = nouveau_user_framebuffer_create,
-	.output_poll_changed = drm_fb_helper_output_poll_changed,
-	.atomic_check = nv50_disp_atomic_check,
-	.atomic_commit = nv50_disp_atomic_commit,
-	.atomic_state_alloc = nv50_disp_atomic_state_alloc,
-	.atomic_state_clear = nv50_disp_atomic_state_clear,
-	.atomic_state_free = nv50_disp_atomic_state_free,
-};
-
-/******************************************************************************
- * Init
- *****************************************************************************/
-
-void
-nv50_display_fini(struct drm_device *dev)
-{
-	struct nouveau_encoder *nv_encoder;
-	struct drm_encoder *encoder;
-	struct drm_plane *plane;
-
-	drm_for_each_plane(plane, dev) {
-		struct nv50_wndw *wndw = nv50_wndw(plane);
-		if (plane->funcs != &nv50_wndw)
-			continue;
-		nv50_wndw_fini(wndw);
-	}
-
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
-		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
-			nv_encoder = nouveau_encoder(encoder);
-			nv50_mstm_fini(nv_encoder->dp.mstm);
-		}
-	}
-}
-
-int
-nv50_display_init(struct drm_device *dev)
-{
-	struct drm_encoder *encoder;
-	struct drm_plane *plane;
-	u32 *push;
-
-	push = evo_wait(nv50_mast(dev), 32);
-	if (!push)
-		return -EBUSY;
-
-	evo_mthd(push, 0x0088, 1);
-	evo_data(push, nv50_mast(dev)->base.sync.handle);
-	evo_kick(push, nv50_mast(dev));
-
-	list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
-		if (encoder->encoder_type != DRM_MODE_ENCODER_DPMST) {
-			struct nouveau_encoder *nv_encoder =
-				nouveau_encoder(encoder);
-			nv50_mstm_init(nv_encoder->dp.mstm);
-		}
-	}
-
-	drm_for_each_plane(plane, dev) {
-		struct nv50_wndw *wndw = nv50_wndw(plane);
-		if (plane->funcs != &nv50_wndw)
-			continue;
-		nv50_wndw_init(wndw);
-	}
-
-	return 0;
-}
-
-void
-nv50_display_destroy(struct drm_device *dev)
-{
-	struct nv50_disp *disp = nv50_disp(dev);
-
-	nv50_dmac_destroy(&disp->mast.base, disp->disp);
-
-	nouveau_bo_unmap(disp->sync);
-	if (disp->sync)
-		nouveau_bo_unpin(disp->sync);
-	nouveau_bo_ref(NULL, &disp->sync);
-
-	nouveau_display(dev)->priv = NULL;
-	kfree(disp);
-}
-
-MODULE_PARM_DESC(atomic, "Expose atomic ioctl (default: disabled)");
-static int nouveau_atomic = 0;
-module_param_named(atomic, nouveau_atomic, int, 0400);
-
-int
-nv50_display_create(struct drm_device *dev)
-{
-	struct nvif_device *device = &nouveau_drm(dev)->client.device;
-	struct nouveau_drm *drm = nouveau_drm(dev);
-	struct dcb_table *dcb = &drm->vbios.dcb;
-	struct drm_connector *connector, *tmp;
-	struct nv50_disp *disp;
-	struct dcb_output *dcbe;
-	int crtcs, ret, i;
-
-	disp = kzalloc(sizeof(*disp), GFP_KERNEL);
-	if (!disp)
-		return -ENOMEM;
-
-	mutex_init(&disp->mutex);
-
-	nouveau_display(dev)->priv = disp;
-	nouveau_display(dev)->dtor = nv50_display_destroy;
-	nouveau_display(dev)->init = nv50_display_init;
-	nouveau_display(dev)->fini = nv50_display_fini;
-	disp->disp = &nouveau_display(dev)->disp;
-	dev->mode_config.funcs = &nv50_disp_func;
-	dev->driver->driver_features |= DRIVER_PREFER_XBGR_30BPP;
-	if (nouveau_atomic)
-		dev->driver->driver_features |= DRIVER_ATOMIC;
-
-	/* small shared memory area we use for notifiers and semaphores */
-	ret = nouveau_bo_new(&drm->client, 4096, 0x1000, TTM_PL_FLAG_VRAM,
-			     0, 0x0000, NULL, NULL, &disp->sync);
-	if (!ret) {
-		ret = nouveau_bo_pin(disp->sync, TTM_PL_FLAG_VRAM, true);
-		if (!ret) {
-			ret = nouveau_bo_map(disp->sync);
-			if (ret)
-				nouveau_bo_unpin(disp->sync);
-		}
-		if (ret)
-			nouveau_bo_ref(NULL, &disp->sync);
-	}
-
-	if (ret)
-		goto out;
-
-	/* allocate master evo channel */
-	ret = nv50_core_create(device, disp->disp, disp->sync->bo.offset,
-			      &disp->mast);
-	if (ret)
-		goto out;
-
-	/* create crtc objects to represent the hw heads */
-	if (disp->disp->oclass >= GF110_DISP)
-		crtcs = nvif_rd32(&device->object, 0x612004) & 0xf;
-	else
-		crtcs = 0x3;
-
-	for (i = 0; i < fls(crtcs); i++) {
-		if (!(crtcs & (1 << i)))
-			continue;
-		ret = nv50_head_create(dev, i);
-		if (ret)
-			goto out;
-	}
-
-	/* create encoder/connector objects based on VBIOS DCB table */
-	for (i = 0, dcbe = &dcb->entry[0]; i < dcb->entries; i++, dcbe++) {
-		connector = nouveau_connector_create(dev, dcbe->connector);
-		if (IS_ERR(connector))
-			continue;
-
-		if (dcbe->location == DCB_LOC_ON_CHIP) {
-			switch (dcbe->type) {
-			case DCB_OUTPUT_TMDS:
-			case DCB_OUTPUT_LVDS:
-			case DCB_OUTPUT_DP:
-				ret = nv50_sor_create(connector, dcbe);
-				break;
-			case DCB_OUTPUT_ANALOG:
-				ret = nv50_dac_create(connector, dcbe);
-				break;
-			default:
-				ret = -ENODEV;
-				break;
-			}
-		} else {
-			ret = nv50_pior_create(connector, dcbe);
-		}
-
-		if (ret) {
-			NV_WARN(drm, "failed to create encoder %d/%d/%d: %d\n",
-				     dcbe->location, dcbe->type,
-				     ffs(dcbe->or) - 1, ret);
-			ret = 0;
-		}
-	}
-
-	/* cull any connectors we created that don't have an encoder */
-	list_for_each_entry_safe(connector, tmp, &dev->mode_config.connector_list, head) {
-		if (connector->encoder_ids[0])
-			continue;
-
-		NV_WARN(drm, "%s has no encoders, removing\n",
-			connector->name);
-		connector->funcs->destroy(connector);
-	}
-
-out:
-	if (ret)
-		nv50_display_destroy(dev);
-	return ret;
-}
diff --git a/drivers/gpu/drm/nouveau/nv50_display.h b/drivers/gpu/drm/nouveau/nv50_display.h
index 918187cee84b..fbd3b15583bc 100644
--- a/drivers/gpu/drm/nouveau/nv50_display.h
+++ b/drivers/gpu/drm/nouveau/nv50_display.h
@@ -28,7 +28,6 @@
 #define __NV50_DISPLAY_H__
 
 #include "nouveau_display.h"
-#include "nouveau_crtc.h"
 #include "nouveau_reg.h"
 
 int  nv50_display_create(struct drm_device *);
diff --git a/drivers/gpu/drm/nouveau/nv50_fence.c b/drivers/gpu/drm/nouveau/nv50_fence.c
index a369d978e267..a00ecc3de053 100644
--- a/drivers/gpu/drm/nouveau/nv50_fence.c
+++ b/drivers/gpu/drm/nouveau/nv50_fence.c
@@ -78,8 +78,6 @@ nv50_fence_create(struct nouveau_drm *drm)
 	priv->base.resume = nv17_fence_resume;
 	priv->base.context_new = nv50_fence_context_new;
 	priv->base.context_del = nv10_fence_context_del;
-	priv->base.contexts = 127;
-	priv->base.context_base = dma_fence_context_alloc(priv->base.contexts);
 	spin_lock_init(&priv->lock);
 
 	ret = nouveau_bo_new(&drm->client, 4096, 0x1000, TTM_PL_FLAG_VRAM,
diff --git a/drivers/gpu/drm/nouveau/nv84_fence.c b/drivers/gpu/drm/nouveau/nv84_fence.c
index 5f0c0c27d5dc..e721bb2163a0 100644
--- a/drivers/gpu/drm/nouveau/nv84_fence.c
+++ b/drivers/gpu/drm/nouveau/nv84_fence.c
@@ -141,9 +141,9 @@ nv84_fence_suspend(struct nouveau_drm *drm)
 	struct nv84_fence_priv *priv = drm->fence;
 	int i;
 
-	priv->suspend = vmalloc(priv->base.contexts * sizeof(u32));
+	priv->suspend = vmalloc(array_size(sizeof(u32), drm->chan.nr));
 	if (priv->suspend) {
-		for (i = 0; i < priv->base.contexts; i++)
+		for (i = 0; i < drm->chan.nr; i++)
 			priv->suspend[i] = nouveau_bo_rd32(priv->bo, i*4);
 	}
 
@@ -157,7 +157,7 @@ nv84_fence_resume(struct nouveau_drm *drm)
 	int i;
 
 	if (priv->suspend) {
-		for (i = 0; i < priv->base.contexts; i++)
+		for (i = 0; i < drm->chan.nr; i++)
 			nouveau_bo_wr32(priv->bo, i*4, priv->suspend[i]);
 		vfree(priv->suspend);
 		priv->suspend = NULL;
@@ -179,7 +179,6 @@ nv84_fence_destroy(struct nouveau_drm *drm)
 int
 nv84_fence_create(struct nouveau_drm *drm)
 {
-	struct nvkm_fifo *fifo = nvxx_fifo(&drm->client.device);
 	struct nv84_fence_priv *priv;
 	u32 domain;
 	int ret;
@@ -194,8 +193,6 @@ nv84_fence_create(struct nouveau_drm *drm)
 	priv->base.context_new = nv84_fence_context_new;
 	priv->base.context_del = nv84_fence_context_del;
 
-	priv->base.contexts = fifo->nr;
-	priv->base.context_base = dma_fence_context_alloc(priv->base.contexts);
 	priv->base.uevent = true;
 
 	mutex_init(&priv->mutex);
@@ -207,7 +204,7 @@ nv84_fence_create(struct nouveau_drm *drm)
 			  * will lose CPU/GPU coherency!
 			  */
 			 TTM_PL_FLAG_TT | TTM_PL_FLAG_UNCACHED;
-	ret = nouveau_bo_new(&drm->client, 16 * priv->base.contexts, 0,
+	ret = nouveau_bo_new(&drm->client, 16 * drm->chan.nr, 0,
 			     domain, 0, 0, NULL, NULL, &priv->bo);
 	if (ret == 0) {
 		ret = nouveau_bo_pin(priv->bo, domain, false);
diff --git a/drivers/gpu/drm/nouveau/nvif/Kbuild b/drivers/gpu/drm/nouveau/nvif/Kbuild
index f1675a4ab6fa..42e8c85caa33 100644
--- a/drivers/gpu/drm/nouveau/nvif/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvif/Kbuild
@@ -1,8 +1,14 @@
 nvif-y := nvif/object.o
 nvif-y += nvif/client.o
 nvif-y += nvif/device.o
+nvif-y += nvif/disp.o
 nvif-y += nvif/driver.o
+nvif-y += nvif/fifo.o
 nvif-y += nvif/mem.o
 nvif-y += nvif/mmu.o
 nvif-y += nvif/notify.o
 nvif-y += nvif/vmm.o
+
+# Usermode classes
+nvif-y += nvif/user.o
+nvif-y += nvif/userc361.o
diff --git a/drivers/gpu/drm/nouveau/nvif/device.c b/drivers/gpu/drm/nouveau/nvif/device.c
index 252d8c33215b..1ec101ba3b42 100644
--- a/drivers/gpu/drm/nouveau/nvif/device.c
+++ b/drivers/gpu/drm/nouveau/nvif/device.c
@@ -37,6 +37,9 @@ nvif_device_time(struct nvif_device *device)
 void
 nvif_device_fini(struct nvif_device *device)
 {
+	nvif_user_fini(device);
+	kfree(device->runlist);
+	device->runlist = NULL;
 	nvif_object_fini(&device->object);
 }
 
@@ -46,6 +49,8 @@ nvif_device_init(struct nvif_object *parent, u32 handle, s32 oclass,
 {
 	int ret = nvif_object_init(parent, handle, oclass, data, size,
 				   &device->object);
+	device->runlist = NULL;
+	device->user.func = NULL;
 	if (ret == 0) {
 		device->info.version = 0;
 		ret = nvif_object_mthd(&device->object, NV_DEVICE_V0_INFO,
diff --git a/drivers/gpu/drm/amd/powerplay/inc/pp_soc15.h b/drivers/gpu/drm/nouveau/nvif/disp.c
index 214f370c5efd..18c7d064f75c 100644
--- a/drivers/gpu/drm/amd/powerplay/inc/pp_soc15.h
+++ b/drivers/gpu/drm/nouveau/nvif/disp.c
@@ -1,5 +1,5 @@
 /*
- * Copyright 2016 Advanced Micro Devices, Inc.
+ * Copyright 2018 Red Hat Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -18,35 +18,43 @@
  * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
- *
  */
-#ifndef PP_SOC15_H
-#define PP_SOC15_H
+#include <nvif/disp.h>
+#include <nvif/device.h>
 
-#include "soc15_hw_ip.h"
-#include "vega10_ip_offset.h"
+#include <nvif/class.h>
 
-inline static uint32_t soc15_get_register_offset(
-		uint32_t hw_id,
-		uint32_t inst,
-		uint32_t segment,
-		uint32_t offset)
+void
+nvif_disp_dtor(struct nvif_disp *disp)
 {
-	uint32_t reg = 0;
-
-	if (hw_id == THM_HWID)
-		reg = THM_BASE.instance[inst].segment[segment] + offset;
-	else if (hw_id == NBIF_HWID)
-		reg = NBIF_BASE.instance[inst].segment[segment] + offset;
-	else if (hw_id == MP1_HWID)
-		reg = MP1_BASE.instance[inst].segment[segment] + offset;
-	else if (hw_id == DF_HWID)
-		reg = DF_BASE.instance[inst].segment[segment] + offset;
-	else if (hw_id == GC_HWID)
-		reg = GC_BASE.instance[inst].segment[segment] + offset;
-	else if (hw_id == SMUIO_HWID)
-		reg = SMUIO_BASE.instance[inst].segment[segment] + offset;
-	return reg;
+	nvif_object_fini(&disp->object);
 }
 
-#endif
+int
+nvif_disp_ctor(struct nvif_device *device, s32 oclass, struct nvif_disp *disp)
+{
+	static const struct nvif_mclass disps[] = {
+		{ GV100_DISP, -1 },
+		{ GP102_DISP, -1 },
+		{ GP100_DISP, -1 },
+		{ GM200_DISP, -1 },
+		{ GM107_DISP, -1 },
+		{ GK110_DISP, -1 },
+		{ GK104_DISP, -1 },
+		{ GF110_DISP, -1 },
+		{ GT214_DISP, -1 },
+		{ GT206_DISP, -1 },
+		{ GT200_DISP, -1 },
+		{   G82_DISP, -1 },
+		{  NV50_DISP, -1 },
+		{  NV04_DISP, -1 },
+		{}
+	};
+	int cid = nvif_sclass(&device->object, disps, oclass);
+	disp->object.client = NULL;
+	if (cid < 0)
+		return cid;
+
+	return nvif_object_init(&device->object, 0, disps[cid].oclass,
+				NULL, 0, &disp->object);
+}
diff --git a/drivers/gpu/drm/nouveau/nvif/fifo.c b/drivers/gpu/drm/nouveau/nvif/fifo.c
new file mode 100644
index 000000000000..e84a2e2ff043
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvif/fifo.c
@@ -0,0 +1,99 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include <nvif/fifo.h>
+
+static int
+nvif_fifo_runlists(struct nvif_device *device)
+{
+	struct nvif_object *object = &device->object;
+	struct {
+		struct nv_device_info_v1 m;
+		struct {
+			struct nv_device_info_v1_data runlists;
+			struct nv_device_info_v1_data runlist[64];
+		} v;
+	} *a;
+	int ret, i;
+
+	if (device->runlist)
+		return 0;
+
+	if (!(a = kmalloc(sizeof(*a), GFP_KERNEL)))
+		return -ENOMEM;
+	a->m.version = 1;
+	a->m.count = sizeof(a->v) / sizeof(a->v.runlists);
+	a->v.runlists.mthd = NV_DEVICE_FIFO_RUNLISTS;
+	for (i = 0; i < ARRAY_SIZE(a->v.runlist); i++)
+		a->v.runlist[i].mthd = NV_DEVICE_FIFO_RUNLIST_ENGINES(i);
+
+	ret = nvif_object_mthd(object, NV_DEVICE_V0_INFO, a, sizeof(*a));
+	if (ret)
+		goto done;
+
+	device->runlists = fls64(a->v.runlists.data);
+	device->runlist = kcalloc(device->runlists, sizeof(*device->runlist),
+				  GFP_KERNEL);
+	if (!device->runlist) {
+		ret = -ENOMEM;
+		goto done;
+	}
+
+	for (i = 0; i < device->runlists; i++) {
+		if (a->v.runlists.data & BIT_ULL(i))
+			device->runlist[i].engines = a->v.runlist[i].data;
+	}
+
+done:
+	kfree(a);
+	return ret;
+}
+
+u64
+nvif_fifo_runlist(struct nvif_device *device, u64 engine)
+{
+	struct nvif_object *object = &device->object;
+	struct {
+		struct nv_device_info_v1 m;
+		struct {
+			struct nv_device_info_v1_data engine;
+		} v;
+	} a = {
+		.m.version = 1,
+		.m.count = sizeof(a.v) / sizeof(a.v.engine),
+		.v.engine.mthd = engine,
+	};
+	u64 runm = 0;
+	int ret, i;
+
+	if ((ret = nvif_fifo_runlists(device)))
+		return runm;
+
+	ret = nvif_object_mthd(object, NV_DEVICE_V0_INFO, &a, sizeof(a));
+	if (ret == 0) {
+		for (i = 0; i < device->runlists; i++) {
+			if (device->runlist[i].engines & a.v.engine.data)
+				runm |= BIT_ULL(i);
+		}
+	}
+
+	return runm;
+}
diff --git a/drivers/gpu/drm/nouveau/nvif/mem.c b/drivers/gpu/drm/nouveau/nvif/mem.c
index 0f9382c60145..b6ebb3b58673 100644
--- a/drivers/gpu/drm/nouveau/nvif/mem.c
+++ b/drivers/gpu/drm/nouveau/nvif/mem.c
@@ -24,6 +24,19 @@
 
 #include <nvif/if000a.h>
 
+int
+nvif_mem_init_map(struct nvif_mmu *mmu, u8 type, u64 size, struct nvif_mem *mem)
+{
+	int ret = nvif_mem_init(mmu, mmu->mem, NVIF_MEM_MAPPABLE | type, 0,
+				size, NULL, 0, mem);
+	if (ret == 0) {
+		ret = nvif_object_map(&mem->object, NULL, 0);
+		if (ret)
+			nvif_mem_fini(mem);
+	}
+	return ret;
+}
+
 void
 nvif_mem_fini(struct nvif_mem *mem)
 {
diff --git a/drivers/gpu/drm/nouveau/nvif/mmu.c b/drivers/gpu/drm/nouveau/nvif/mmu.c
index 15d0dcbf7ab4..ae08a1ca8044 100644
--- a/drivers/gpu/drm/nouveau/nvif/mmu.c
+++ b/drivers/gpu/drm/nouveau/nvif/mmu.c
@@ -36,6 +36,12 @@ nvif_mmu_fini(struct nvif_mmu *mmu)
 int
 nvif_mmu_init(struct nvif_object *parent, s32 oclass, struct nvif_mmu *mmu)
 {
+	static const struct nvif_mclass mems[] = {
+		{ NVIF_CLASS_MEM_GF100, -1 },
+		{ NVIF_CLASS_MEM_NV50 , -1 },
+		{ NVIF_CLASS_MEM_NV04 , -1 },
+		{}
+	};
 	struct nvif_mmu_v0 args;
 	int ret, i;
 
@@ -54,12 +60,20 @@ nvif_mmu_init(struct nvif_object *parent, s32 oclass, struct nvif_mmu *mmu)
 	mmu->type_nr = args.type_nr;
 	mmu->kind_nr = args.kind_nr;
 
-	mmu->heap = kmalloc(sizeof(*mmu->heap) * mmu->heap_nr, GFP_KERNEL);
-	mmu->type = kmalloc(sizeof(*mmu->type) * mmu->type_nr, GFP_KERNEL);
+	ret = nvif_mclass(&mmu->object, mems);
+	if (ret < 0)
+		goto done;
+	mmu->mem = mems[ret].oclass;
+
+	mmu->heap = kmalloc_array(mmu->heap_nr, sizeof(*mmu->heap),
+				  GFP_KERNEL);
+	mmu->type = kmalloc_array(mmu->type_nr, sizeof(*mmu->type),
+				  GFP_KERNEL);
 	if (ret = -ENOMEM, !mmu->heap || !mmu->type)
 		goto done;
 
-	mmu->kind = kmalloc(sizeof(*mmu->kind) * mmu->kind_nr, GFP_KERNEL);
+	mmu->kind = kmalloc_array(mmu->kind_nr, sizeof(*mmu->kind),
+				  GFP_KERNEL);
 	if (!mmu->kind && mmu->kind_nr)
 		goto done;
 
diff --git a/drivers/gpu/drm/nouveau/nvif/object.c b/drivers/gpu/drm/nouveau/nvif/object.c
index 40adfe9b334b..ef3f62840e83 100644
--- a/drivers/gpu/drm/nouveau/nvif/object.c
+++ b/drivers/gpu/drm/nouveau/nvif/object.c
@@ -83,7 +83,7 @@ nvif_object_sclass_get(struct nvif_object *object, struct nvif_sclass **psclass)
 			return ret;
 	}
 
-	*psclass = kzalloc(sizeof(**psclass) * args->sclass.count, GFP_KERNEL);
+	*psclass = kcalloc(args->sclass.count, sizeof(**psclass), GFP_KERNEL);
 	if (*psclass) {
 		for (i = 0; i < args->sclass.count; i++) {
 			(*psclass)[i].oclass = args->sclass.oclass[i].oclass;
diff --git a/drivers/gpu/drm/nouveau/nvif/user.c b/drivers/gpu/drm/nouveau/nvif/user.c
new file mode 100644
index 000000000000..10da3cdca647
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvif/user.c
@@ -0,0 +1,64 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include <nvif/user.h>
+#include <nvif/device.h>
+
+#include <nvif/class.h>
+
+void
+nvif_user_fini(struct nvif_device *device)
+{
+	if (device->user.func) {
+		nvif_object_fini(&device->user.object);
+		device->user.func = NULL;
+	}
+}
+
+int
+nvif_user_init(struct nvif_device *device)
+{
+	struct {
+		s32 oclass;
+		int version;
+		const struct nvif_user_func *func;
+	} users[] = {
+		{ VOLTA_USERMODE_A, -1, &nvif_userc361 },
+		{}
+	};
+	int cid, ret;
+
+	if (device->user.func)
+		return 0;
+
+	cid = nvif_mclass(&device->object, users);
+	if (cid < 0)
+		return cid;
+
+	ret = nvif_object_init(&device->object, 0, users[cid].oclass, NULL, 0,
+			       &device->user.object);
+	if (ret)
+		return ret;
+
+	nvif_object_map(&device->user.object, NULL, 0);
+	device->user.func = users[cid].func;
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogp100.c b/drivers/gpu/drm/nouveau/nvif/userc361.c
index 1530a9217aea..19f9958e7e01 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogp100.c
+++ b/drivers/gpu/drm/nouveau/nvif/userc361.c
@@ -1,5 +1,5 @@
 /*
- * Copyright 2016 Red Hat Inc.
+ * Copyright 2018 Red Hat Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -18,17 +18,16 @@
  * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs
  */
-#include "changk104.h"
+#include <nvif/user.h>
 
-#include <nvif/class.h>
+static void
+nvif_userc361_doorbell(struct nvif_user *user, u32 token)
+{
+	nvif_wr32(&user->object, 0x90, token);
+}
 
-const struct nvkm_fifo_chan_oclass
-gp100_fifo_gpfifo_oclass = {
-	.base.oclass = PASCAL_CHANNEL_GPFIFO_A,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = gk104_fifo_gpfifo_new,
+const struct nvif_user_func
+nvif_userc361 = {
+	.doorbell = nvif_userc361_doorbell,
 };
diff --git a/drivers/gpu/drm/nouveau/nvif/vmm.c b/drivers/gpu/drm/nouveau/nvif/vmm.c
index 31cdb2d2e1ff..6b9c5776547f 100644
--- a/drivers/gpu/drm/nouveau/nvif/vmm.c
+++ b/drivers/gpu/drm/nouveau/nvif/vmm.c
@@ -37,7 +37,7 @@ nvif_vmm_map(struct nvif_vmm *vmm, u64 addr, u64 size, void *argv, u32 argc,
 	     struct nvif_mem *mem, u64 offset)
 {
 	struct nvif_vmm_map_v0 *args;
-	u8 stack[16];
+	u8 stack[48];
 	int ret;
 
 	if (sizeof(*args) + argc > sizeof(stack)) {
@@ -138,7 +138,8 @@ nvif_vmm_init(struct nvif_mmu *mmu, s32 oclass, u64 addr, u64 size,
 	vmm->limit = args->size;
 
 	vmm->page_nr = args->page_nr;
-	vmm->page = kmalloc(sizeof(*vmm->page) * vmm->page_nr, GFP_KERNEL);
+	vmm->page = kmalloc_array(vmm->page_nr, sizeof(*vmm->page),
+				  GFP_KERNEL);
 	if (!vmm->page) {
 		ret = -ENOMEM;
 		goto done;
diff --git a/drivers/gpu/drm/nouveau/nvkm/core/engine.c b/drivers/gpu/drm/nouveau/nvkm/core/engine.c
index 657231c3c098..1a47c40e171b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/core/engine.c
+++ b/drivers/gpu/drm/nouveau/nvkm/core/engine.c
@@ -83,6 +83,21 @@ nvkm_engine_intr(struct nvkm_subdev *subdev)
 }
 
 static int
+nvkm_engine_info(struct nvkm_subdev *subdev, u64 mthd, u64 *data)
+{
+	struct nvkm_engine *engine = nvkm_engine(subdev);
+	if (engine->func->info) {
+		if (!IS_ERR((engine = nvkm_engine_ref(engine)))) {
+			int ret = engine->func->info(engine, mthd, data);
+			nvkm_engine_unref(&engine);
+			return ret;
+		}
+		return PTR_ERR(engine);
+	}
+	return -ENOSYS;
+}
+
+static int
 nvkm_engine_fini(struct nvkm_subdev *subdev, bool suspend)
 {
 	struct nvkm_engine *engine = nvkm_engine(subdev);
@@ -150,6 +165,7 @@ nvkm_engine_func = {
 	.preinit = nvkm_engine_preinit,
 	.init = nvkm_engine_init,
 	.fini = nvkm_engine_fini,
+	.info = nvkm_engine_info,
 	.intr = nvkm_engine_intr,
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/core/event.c b/drivers/gpu/drm/nouveau/nvkm/core/event.c
index 4e8d3fa042df..006618d77aa4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/core/event.c
+++ b/drivers/gpu/drm/nouveau/nvkm/core/event.c
@@ -84,7 +84,8 @@ int
 nvkm_event_init(const struct nvkm_event_func *func, int types_nr, int index_nr,
 		struct nvkm_event *event)
 {
-	event->refs = kzalloc(sizeof(*event->refs) * index_nr * types_nr,
+	event->refs = kzalloc(array3_size(index_nr, types_nr,
+					  sizeof(*event->refs)),
 			      GFP_KERNEL);
 	if (!event->refs)
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/nouveau/nvkm/core/ramht.c b/drivers/gpu/drm/nouveau/nvkm/core/ramht.c
index ccba4ae73cc5..8162e3d2359c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/core/ramht.c
+++ b/drivers/gpu/drm/nouveau/nvkm/core/ramht.c
@@ -144,8 +144,7 @@ nvkm_ramht_new(struct nvkm_device *device, u32 size, u32 align,
 	struct nvkm_ramht *ramht;
 	int ret, i;
 
-	if (!(ramht = *pramht = vzalloc(sizeof(*ramht) +
-					(size >> 3) * sizeof(*ramht->data))))
+	if (!(ramht = *pramht = vzalloc(struct_size(ramht, data, (size >> 3)))))
 		return -ENOMEM;
 
 	ramht->device = device;
diff --git a/drivers/gpu/drm/nouveau/nvkm/core/subdev.c b/drivers/gpu/drm/nouveau/nvkm/core/subdev.c
index a134d225f958..03f676c18aad 100644
--- a/drivers/gpu/drm/nouveau/nvkm/core/subdev.c
+++ b/drivers/gpu/drm/nouveau/nvkm/core/subdev.c
@@ -35,6 +35,7 @@ nvkm_subdev_name[NVKM_SUBDEV_NR] = {
 	[NVKM_SUBDEV_BUS     ] = "bus",
 	[NVKM_SUBDEV_CLK     ] = "clk",
 	[NVKM_SUBDEV_DEVINIT ] = "devinit",
+	[NVKM_SUBDEV_FAULT   ] = "fault",
 	[NVKM_SUBDEV_FB      ] = "fb",
 	[NVKM_SUBDEV_FUSE    ] = "fuse",
 	[NVKM_SUBDEV_GPIO    ] = "gpio",
@@ -60,6 +61,9 @@ nvkm_subdev_name[NVKM_SUBDEV_NR] = {
 	[NVKM_ENGINE_CE3     ] = "ce3",
 	[NVKM_ENGINE_CE4     ] = "ce4",
 	[NVKM_ENGINE_CE5     ] = "ce5",
+	[NVKM_ENGINE_CE6     ] = "ce6",
+	[NVKM_ENGINE_CE7     ] = "ce7",
+	[NVKM_ENGINE_CE8     ] = "ce8",
 	[NVKM_ENGINE_CIPHER  ] = "cipher",
 	[NVKM_ENGINE_DISP    ] = "disp",
 	[NVKM_ENGINE_DMAOBJ  ] = "dma",
@@ -92,6 +96,14 @@ nvkm_subdev_intr(struct nvkm_subdev *subdev)
 }
 
 int
+nvkm_subdev_info(struct nvkm_subdev *subdev, u64 mthd, u64 *data)
+{
+	if (subdev->func->info)
+		return subdev->func->info(subdev, mthd, data);
+	return -ENOSYS;
+}
+
+int
 nvkm_subdev_fini(struct nvkm_subdev *subdev, bool suspend)
 {
 	struct nvkm_device *device = subdev->device;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/ce/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/ce/Kbuild
index 255d81ccf916..80d784441904 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/ce/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/ce/Kbuild
@@ -5,3 +5,4 @@ nvkm-y += nvkm/engine/ce/gm107.o
 nvkm-y += nvkm/engine/ce/gm200.o
 nvkm-y += nvkm/engine/ce/gp100.o
 nvkm-y += nvkm/engine/ce/gp102.o
+nvkm-y += nvkm/engine/ce/gv100.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/ce/gv100.c
index 5ad5d0f5db05..fcda3de45857 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/ce/gv100.c
@@ -1,5 +1,5 @@
 /*
- * Copyright 2012 Red Hat Inc.
+ * Copyright 2018 Red Hat Inc.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -18,20 +18,23 @@
  * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
  * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs
  */
-#include "channv50.h"
-#include "rootnv50.h"
+#include "priv.h"
 
 #include <nvif/class.h>
 
-const struct nv50_disp_pioc_oclass
-g84_disp_oimm_oclass = {
-	.base.oclass = G82_DISP_OVERLAY,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_oimm_new,
-	.func = &nv50_disp_pioc_func,
-	.chid = { 5, 5 },
+static const struct nvkm_engine_func
+gv100_ce = {
+	.intr = gp100_ce_intr,
+	.sclass = {
+		{ -1, -1, VOLTA_DMA_COPY_A },
+		{}
+	}
 };
+
+int
+gv100_ce_new(struct nvkm_device *device, int index,
+	     struct nvkm_engine **pengine)
+{
+	return nvkm_engine_new_(&gv100_ce, device, index, true, pengine);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
index 05cd674326a6..e294013426ce 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c
@@ -2161,6 +2161,7 @@ nv130_chipset = {
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gm200_devinit_new,
+	.fault = gp100_fault_new,
 	.fb = gp100_fb_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
@@ -2196,13 +2197,14 @@ nv132_chipset = {
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gm200_devinit_new,
+	.fault = gp100_fault_new,
 	.fb = gp102_fb_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
 	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
 	.imem = nv50_instmem_new,
-	.ltc = gp100_ltc_new,
+	.ltc = gp102_ltc_new,
 	.mc = gp100_mc_new,
 	.mmu = gp100_mmu_new,
 	.therm = gp100_therm_new,
@@ -2231,13 +2233,14 @@ nv134_chipset = {
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gm200_devinit_new,
+	.fault = gp100_fault_new,
 	.fb = gp102_fb_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
 	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
 	.imem = nv50_instmem_new,
-	.ltc = gp100_ltc_new,
+	.ltc = gp102_ltc_new,
 	.mc = gp100_mc_new,
 	.mmu = gp100_mmu_new,
 	.therm = gp100_therm_new,
@@ -2253,7 +2256,7 @@ nv134_chipset = {
 	.disp = gp102_disp_new,
 	.dma = gf119_dma_new,
 	.fifo = gp100_fifo_new,
-	.gr = gp102_gr_new,
+	.gr = gp104_gr_new,
 	.nvdec = gp102_nvdec_new,
 	.sec2 = gp102_sec2_new,
 	.sw = gf100_sw_new,
@@ -2266,13 +2269,14 @@ nv136_chipset = {
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gm200_devinit_new,
+	.fault = gp100_fault_new,
 	.fb = gp102_fb_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
 	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
 	.imem = nv50_instmem_new,
-	.ltc = gp100_ltc_new,
+	.ltc = gp102_ltc_new,
 	.mc = gp100_mc_new,
 	.mmu = gp100_mmu_new,
 	.therm = gp100_therm_new,
@@ -2288,7 +2292,7 @@ nv136_chipset = {
 	.disp = gp102_disp_new,
 	.dma = gf119_dma_new,
 	.fifo = gp100_fifo_new,
-	.gr = gp102_gr_new,
+	.gr = gp104_gr_new,
 	.nvdec = gp102_nvdec_new,
 	.sec2 = gp102_sec2_new,
 	.sw = gf100_sw_new,
@@ -2301,13 +2305,14 @@ nv137_chipset = {
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gm200_devinit_new,
+	.fault = gp100_fault_new,
 	.fb = gp102_fb_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
 	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
 	.imem = nv50_instmem_new,
-	.ltc = gp100_ltc_new,
+	.ltc = gp102_ltc_new,
 	.mc = gp100_mc_new,
 	.mmu = gp100_mmu_new,
 	.therm = gp100_therm_new,
@@ -2336,13 +2341,14 @@ nv138_chipset = {
 	.bios = nvkm_bios_new,
 	.bus = gf100_bus_new,
 	.devinit = gm200_devinit_new,
+	.fault = gp100_fault_new,
 	.fb = gp102_fb_new,
 	.fuse = gm107_fuse_new,
 	.gpio = gk104_gpio_new,
 	.i2c = gm200_i2c_new,
 	.ibus = gm200_ibus_new,
 	.imem = nv50_instmem_new,
-	.ltc = gp100_ltc_new,
+	.ltc = gp102_ltc_new,
 	.mc = gp100_mc_new,
 	.mmu = gp100_mmu_new,
 	.therm = gp100_therm_new,
@@ -2369,11 +2375,12 @@ nv13b_chipset = {
 	.name = "GP10B",
 	.bar = gm20b_bar_new,
 	.bus = gf100_bus_new,
+	.fault = gp100_fault_new,
 	.fb = gp10b_fb_new,
 	.fuse = gm107_fuse_new,
 	.ibus = gp10b_ibus_new,
 	.imem = gk20a_instmem_new,
-	.ltc = gp100_ltc_new,
+	.ltc = gp102_ltc_new,
 	.mc = gp10b_mc_new,
 	.mmu = gp10b_mmu_new,
 	.secboot = gp10b_secboot_new,
@@ -2387,6 +2394,46 @@ nv13b_chipset = {
 	.sw = gf100_sw_new,
 };
 
+static const struct nvkm_device_chip
+nv140_chipset = {
+	.name = "GV100",
+	.bar = gm107_bar_new,
+	.bios = nvkm_bios_new,
+	.bus = gf100_bus_new,
+	.devinit = gv100_devinit_new,
+	.fault = gv100_fault_new,
+	.fb = gv100_fb_new,
+	.fuse = gm107_fuse_new,
+	.gpio = gk104_gpio_new,
+	.i2c = gm200_i2c_new,
+	.ibus = gm200_ibus_new,
+	.imem = nv50_instmem_new,
+	.ltc = gp102_ltc_new,
+	.mc = gp100_mc_new,
+	.mmu = gv100_mmu_new,
+	.pci = gp100_pci_new,
+	.pmu = gp102_pmu_new,
+	.secboot = gp108_secboot_new,
+	.therm = gp100_therm_new,
+	.timer = gk20a_timer_new,
+	.top = gk104_top_new,
+	.disp = gv100_disp_new,
+	.ce[0] = gv100_ce_new,
+	.ce[1] = gv100_ce_new,
+	.ce[2] = gv100_ce_new,
+	.ce[3] = gv100_ce_new,
+	.ce[4] = gv100_ce_new,
+	.ce[5] = gv100_ce_new,
+	.ce[6] = gv100_ce_new,
+	.ce[7] = gv100_ce_new,
+	.ce[8] = gv100_ce_new,
+	.dma = gv100_dma_new,
+	.fifo = gv100_fifo_new,
+	.gr = gv100_gr_new,
+	.nvdec = gp102_nvdec_new,
+	.sec2 = gp102_sec2_new,
+};
+
 static int
 nvkm_device_event_ctor(struct nvkm_object *object, void *data, u32 size,
 		       struct nvkm_notify *notify)
@@ -2420,6 +2467,7 @@ nvkm_device_subdev(struct nvkm_device *device, int index)
 	_(BUS     , device->bus     , &device->bus->subdev);
 	_(CLK     , device->clk     , &device->clk->subdev);
 	_(DEVINIT , device->devinit , &device->devinit->subdev);
+	_(FAULT   , device->fault   , &device->fault->subdev);
 	_(FB      , device->fb      , &device->fb->subdev);
 	_(FUSE    , device->fuse    , &device->fuse->subdev);
 	_(GPIO    , device->gpio    , &device->gpio->subdev);
@@ -2463,6 +2511,9 @@ nvkm_device_engine(struct nvkm_device *device, int index)
 	_(CE3    , device->ce[3]   ,  device->ce[3]);
 	_(CE4    , device->ce[4]   ,  device->ce[4]);
 	_(CE5    , device->ce[5]   ,  device->ce[5]);
+	_(CE6    , device->ce[6]   ,  device->ce[6]);
+	_(CE7    , device->ce[7]   ,  device->ce[7]);
+	_(CE8    , device->ce[8]   ,  device->ce[8]);
 	_(CIPHER , device->cipher  ,  device->cipher);
 	_(DISP   , device->disp    , &device->disp->engine);
 	_(DMAOBJ , device->dma     , &device->dma->engine);
@@ -2739,6 +2790,7 @@ nvkm_device_ctor(const struct nvkm_device_func *func,
 			case 0x110:
 			case 0x120: device->card_type = GM100; break;
 			case 0x130: device->card_type = GP100; break;
+			case 0x140: device->card_type = GV100; break;
 			default:
 				break;
 			}
@@ -2830,6 +2882,7 @@ nvkm_device_ctor(const struct nvkm_device_func *func,
 		case 0x137: device->chip = &nv137_chipset; break;
 		case 0x138: device->chip = &nv138_chipset; break;
 		case 0x13b: device->chip = &nv13b_chipset; break;
+		case 0x140: device->chip = &nv140_chipset; break;
 		default:
 			nvdev_error(device, "unknown chipset (%08x)\n", boot0);
 			goto done;
@@ -2891,6 +2944,7 @@ nvkm_device_ctor(const struct nvkm_device_func *func,
 		_(NVKM_SUBDEV_BUS     ,      bus);
 		_(NVKM_SUBDEV_CLK     ,      clk);
 		_(NVKM_SUBDEV_DEVINIT ,  devinit);
+		_(NVKM_SUBDEV_FAULT   ,    fault);
 		_(NVKM_SUBDEV_FB      ,       fb);
 		_(NVKM_SUBDEV_FUSE    ,     fuse);
 		_(NVKM_SUBDEV_GPIO    ,     gpio);
@@ -2916,6 +2970,9 @@ nvkm_device_ctor(const struct nvkm_device_func *func,
 		_(NVKM_ENGINE_CE3     ,    ce[3]);
 		_(NVKM_ENGINE_CE4     ,    ce[4]);
 		_(NVKM_ENGINE_CE5     ,    ce[5]);
+		_(NVKM_ENGINE_CE6     ,    ce[6]);
+		_(NVKM_ENGINE_CE7     ,    ce[7]);
+		_(NVKM_ENGINE_CE8     ,    ce[8]);
 		_(NVKM_ENGINE_CIPHER  ,   cipher);
 		_(NVKM_ENGINE_DISP    ,     disp);
 		_(NVKM_ENGINE_DMAOBJ  ,      dma);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/priv.h b/drivers/gpu/drm/nouveau/nvkm/engine/device/priv.h
index 08d0bf605722..253ab914a8ef 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/priv.h
@@ -8,6 +8,7 @@
 #include <subdev/bus.h>
 #include <subdev/clk.h>
 #include <subdev/devinit.h>
+#include <subdev/fault.h>
 #include <subdev/fb.h>
 #include <subdev/fuse.h>
 #include <subdev/gpio.h>
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/tegra.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/tegra.c
index 78597da6313a..0e372a190d3f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/tegra.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/tegra.c
@@ -23,6 +23,10 @@
 #ifdef CONFIG_NOUVEAU_PLATFORM_DRIVER
 #include "priv.h"
 
+#if IS_ENABLED(CONFIG_ARM_DMA_USE_IOMMU)
+#include <asm/dma-iommu.h>
+#endif
+
 static int
 nvkm_device_tegra_power_up(struct nvkm_device_tegra *tdev)
 {
@@ -105,6 +109,15 @@ nvkm_device_tegra_probe_iommu(struct nvkm_device_tegra *tdev)
 	unsigned long pgsize_bitmap;
 	int ret;
 
+#if IS_ENABLED(CONFIG_ARM_DMA_USE_IOMMU)
+	if (dev->archdata.mapping) {
+		struct dma_iommu_mapping *mapping = to_dma_iommu_mapping(dev);
+
+		arm_iommu_detach_device(dev);
+		arm_iommu_release_mapping(mapping);
+	}
+#endif
+
 	if (!tdev->func->iommu_bit)
 		return;
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
index 17adcb4e8854..dde6bbafa709 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/user.c
@@ -40,6 +40,66 @@ struct nvkm_udevice {
 };
 
 static int
+nvkm_udevice_info_subdev(struct nvkm_device *device, u64 mthd, u64 *data)
+{
+	struct nvkm_subdev *subdev;
+	enum nvkm_devidx subidx;
+
+	switch (mthd & NV_DEVICE_INFO_UNIT) {
+	case NV_DEVICE_FIFO(0): subidx = NVKM_ENGINE_FIFO; break;
+	default:
+		return -EINVAL;
+	}
+
+	subdev = nvkm_device_subdev(device, subidx);
+	if (subdev)
+		return nvkm_subdev_info(subdev, mthd, data);
+	return -ENODEV;
+}
+
+static void
+nvkm_udevice_info_v1(struct nvkm_device *device,
+		     struct nv_device_info_v1_data *args)
+{
+	if (args->mthd & NV_DEVICE_INFO_UNIT) {
+		if (nvkm_udevice_info_subdev(device, args->mthd, &args->data))
+			args->mthd = NV_DEVICE_INFO_INVALID;
+		return;
+	}
+
+	switch (args->mthd) {
+#define ENGINE__(A,B,C) NV_DEVICE_INFO_ENGINE_##A: { int _i;                   \
+	for (_i = (B), args->data = 0ULL; _i <= (C); _i++) {                   \
+		if (nvkm_device_engine(device, _i))                            \
+			args->data |= BIT_ULL(_i);                             \
+	}                                                                      \
+}
+#define ENGINE_A(A) ENGINE__(A, NVKM_ENGINE_##A   , NVKM_ENGINE_##A)
+#define ENGINE_B(A) ENGINE__(A, NVKM_ENGINE_##A##0, NVKM_ENGINE_##A##_LAST)
+	case ENGINE_A(SW    ); break;
+	case ENGINE_A(GR    ); break;
+	case ENGINE_A(MPEG  ); break;
+	case ENGINE_A(ME    ); break;
+	case ENGINE_A(CIPHER); break;
+	case ENGINE_A(BSP   ); break;
+	case ENGINE_A(VP    ); break;
+	case ENGINE_B(CE    ); break;
+	case ENGINE_A(SEC   ); break;
+	case ENGINE_A(MSVLD ); break;
+	case ENGINE_A(MSPDEC); break;
+	case ENGINE_A(MSPPP ); break;
+	case ENGINE_A(MSENC ); break;
+	case ENGINE_A(VIC   ); break;
+	case ENGINE_A(SEC2  ); break;
+	case ENGINE_A(NVDEC ); break;
+	case ENGINE_B(NVENC ); break;
+	default:
+		args->mthd = NV_DEVICE_INFO_INVALID;
+		break;
+	}
+}
+
+static int
 nvkm_udevice_info(struct nvkm_udevice *udev, void *data, u32 size)
 {
 	struct nvkm_object *object = &udev->object;
@@ -48,10 +108,21 @@ nvkm_udevice_info(struct nvkm_udevice *udev, void *data, u32 size)
 	struct nvkm_instmem *imem = device->imem;
 	union {
 		struct nv_device_info_v0 v0;
+		struct nv_device_info_v1 v1;
 	} *args = data;
-	int ret = -ENOSYS;
+	int ret = -ENOSYS, i;
 
 	nvif_ioctl(object, "device info size %d\n", size);
+	if (!(ret = nvif_unpack(ret, &data, &size, args->v1, 1, 1, true))) {
+		nvif_ioctl(object, "device info vers %d count %d\n",
+			   args->v1.version, args->v1.count);
+		if (args->v1.count * sizeof(args->v1.data[0]) == size) {
+			for (i = 0; i < args->v1.count; i++)
+				nvkm_udevice_info_v1(device, &args->v1.data[i]);
+			return 0;
+		}
+		return -EINVAL;
+	} else
 	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
 		nvif_ioctl(object, "device info vers %d\n", args->v0.version);
 	} else
@@ -103,6 +174,7 @@ nvkm_udevice_info(struct nvkm_udevice *udev, void *data, u32 size)
 	case NV_E0: args->v0.family = NV_DEVICE_INFO_V0_KEPLER; break;
 	case GM100: args->v0.family = NV_DEVICE_INFO_V0_MAXWELL; break;
 	case GP100: args->v0.family = NV_DEVICE_INFO_V0_PASCAL; break;
+	case GV100: args->v0.family = NV_DEVICE_INFO_V0_VOLTA; break;
 	default:
 		args->v0.family = 0;
 		break;
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
index 48ce6699183e..3d485dbf310a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/Kbuild
@@ -14,12 +14,14 @@ nvkm-y += nvkm/engine/disp/gm107.o
 nvkm-y += nvkm/engine/disp/gm200.o
 nvkm-y += nvkm/engine/disp/gp100.o
 nvkm-y += nvkm/engine/disp/gp102.o
+nvkm-y += nvkm/engine/disp/gv100.o
 nvkm-y += nvkm/engine/disp/vga.o
 
 nvkm-y += nvkm/engine/disp/head.o
 nvkm-y += nvkm/engine/disp/headnv04.o
 nvkm-y += nvkm/engine/disp/headnv50.o
 nvkm-y += nvkm/engine/disp/headgf119.o
+nvkm-y += nvkm/engine/disp/headgv100.o
 
 nvkm-y += nvkm/engine/disp/ior.o
 nvkm-y += nvkm/engine/disp/dacnv50.o
@@ -35,6 +37,7 @@ nvkm-y += nvkm/engine/disp/sorgf119.o
 nvkm-y += nvkm/engine/disp/sorgk104.o
 nvkm-y += nvkm/engine/disp/sorgm107.o
 nvkm-y += nvkm/engine/disp/sorgm200.o
+nvkm-y += nvkm/engine/disp/sorgv100.o
 
 nvkm-y += nvkm/engine/disp/outp.o
 nvkm-y += nvkm/engine/disp/dp.o
@@ -47,6 +50,7 @@ nvkm-y += nvkm/engine/disp/hdmig84.o
 nvkm-y += nvkm/engine/disp/hdmigt215.o
 nvkm-y += nvkm/engine/disp/hdmigf119.o
 nvkm-y += nvkm/engine/disp/hdmigk104.o
+nvkm-y += nvkm/engine/disp/hdmigv100.o
 
 nvkm-y += nvkm/engine/disp/conn.o
 
@@ -63,57 +67,49 @@ nvkm-y += nvkm/engine/disp/rootgm107.o
 nvkm-y += nvkm/engine/disp/rootgm200.o
 nvkm-y += nvkm/engine/disp/rootgp100.o
 nvkm-y += nvkm/engine/disp/rootgp102.o
+nvkm-y += nvkm/engine/disp/rootgv100.o
 
 nvkm-y += nvkm/engine/disp/channv50.o
 nvkm-y += nvkm/engine/disp/changf119.o
+nvkm-y += nvkm/engine/disp/changv100.o
 
 nvkm-y += nvkm/engine/disp/dmacnv50.o
 nvkm-y += nvkm/engine/disp/dmacgf119.o
 nvkm-y += nvkm/engine/disp/dmacgp102.o
+nvkm-y += nvkm/engine/disp/dmacgv100.o
 
 nvkm-y += nvkm/engine/disp/basenv50.o
 nvkm-y += nvkm/engine/disp/baseg84.o
-nvkm-y += nvkm/engine/disp/basegt200.o
-nvkm-y += nvkm/engine/disp/basegt215.o
 nvkm-y += nvkm/engine/disp/basegf119.o
-nvkm-y += nvkm/engine/disp/basegk104.o
-nvkm-y += nvkm/engine/disp/basegk110.o
 nvkm-y += nvkm/engine/disp/basegp102.o
 
 nvkm-y += nvkm/engine/disp/corenv50.o
 nvkm-y += nvkm/engine/disp/coreg84.o
 nvkm-y += nvkm/engine/disp/coreg94.o
-nvkm-y += nvkm/engine/disp/coregt200.o
-nvkm-y += nvkm/engine/disp/coregt215.o
 nvkm-y += nvkm/engine/disp/coregf119.o
 nvkm-y += nvkm/engine/disp/coregk104.o
-nvkm-y += nvkm/engine/disp/coregk110.o
-nvkm-y += nvkm/engine/disp/coregm107.o
-nvkm-y += nvkm/engine/disp/coregm200.o
-nvkm-y += nvkm/engine/disp/coregp100.o
 nvkm-y += nvkm/engine/disp/coregp102.o
+nvkm-y += nvkm/engine/disp/coregv100.o
 
 nvkm-y += nvkm/engine/disp/ovlynv50.o
 nvkm-y += nvkm/engine/disp/ovlyg84.o
 nvkm-y += nvkm/engine/disp/ovlygt200.o
-nvkm-y += nvkm/engine/disp/ovlygt215.o
 nvkm-y += nvkm/engine/disp/ovlygf119.o
 nvkm-y += nvkm/engine/disp/ovlygk104.o
 nvkm-y += nvkm/engine/disp/ovlygp102.o
 
+nvkm-y += nvkm/engine/disp/wimmgv100.o
+
+nvkm-y += nvkm/engine/disp/wndwgv100.o
+
 nvkm-y += nvkm/engine/disp/piocnv50.o
 nvkm-y += nvkm/engine/disp/piocgf119.o
 
 nvkm-y += nvkm/engine/disp/cursnv50.o
-nvkm-y += nvkm/engine/disp/cursg84.o
-nvkm-y += nvkm/engine/disp/cursgt215.o
 nvkm-y += nvkm/engine/disp/cursgf119.o
-nvkm-y += nvkm/engine/disp/cursgk104.o
 nvkm-y += nvkm/engine/disp/cursgp102.o
+nvkm-y += nvkm/engine/disp/cursgv100.o
 
 nvkm-y += nvkm/engine/disp/oimmnv50.o
-nvkm-y += nvkm/engine/disp/oimmg84.o
-nvkm-y += nvkm/engine/disp/oimmgt215.o
 nvkm-y += nvkm/engine/disp/oimmgf119.o
-nvkm-y += nvkm/engine/disp/oimmgk104.o
 nvkm-y += nvkm/engine/disp/oimmgp102.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/base.c
index 93a75e5b2791..32fa94a9773f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/base.c
@@ -220,6 +220,9 @@ nvkm_disp_fini(struct nvkm_engine *engine, bool suspend)
 	struct nvkm_conn *conn;
 	struct nvkm_outp *outp;
 
+	if (disp->func->fini)
+		disp->func->fini(disp);
+
 	list_for_each_entry(outp, &disp->outp, head) {
 		nvkm_outp_fini(outp);
 	}
@@ -237,6 +240,7 @@ nvkm_disp_init(struct nvkm_engine *engine)
 	struct nvkm_disp *disp = nvkm_disp(engine);
 	struct nvkm_conn *conn;
 	struct nvkm_outp *outp;
+	struct nvkm_ior *ior;
 
 	list_for_each_entry(conn, &disp->conn, head) {
 		nvkm_conn_init(conn);
@@ -246,6 +250,19 @@ nvkm_disp_init(struct nvkm_engine *engine)
 		nvkm_outp_init(outp);
 	}
 
+	if (disp->func->init) {
+		int ret = disp->func->init(disp);
+		if (ret)
+			return ret;
+	}
+
+	/* Set 'normal' (ie. when it's attached to a head) state for
+	 * each output resource to 'fully enabled'.
+	 */
+	list_for_each_entry(ior, &disp->ior, head) {
+		ior->func->power(ior, true, true, true, true, true);
+	}
+
 	return 0;
 }
 
@@ -376,6 +393,12 @@ nvkm_disp_oneinit(struct nvkm_engine *engine)
 	if (ret)
 		return ret;
 
+	if (disp->func->oneinit) {
+		ret = disp->func->oneinit(disp);
+		if (ret)
+			return ret;
+	}
+
 	i = 0;
 	list_for_each_entry(head, &disp->head, head)
 		i = max(i, head->id + 1);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/baseg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/baseg84.c
index 6d17630a3dee..01253f4a9946 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/baseg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/baseg84.c
@@ -21,10 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 g84_disp_base_mthd_base = {
@@ -56,8 +53,8 @@ g84_disp_base_mthd_base = {
 	}
 };
 
-const struct nv50_disp_chan_mthd
-g84_disp_base_chan_mthd = {
+static const struct nv50_disp_chan_mthd
+g84_disp_base_mthd = {
 	.name = "Base",
 	.addr = 0x000540,
 	.prev = 0x000004,
@@ -68,13 +65,10 @@ g84_disp_base_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-g84_disp_base_oclass = {
-	.base.oclass = G82_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &g84_disp_base_chan_mthd,
-	.chid = 1,
-};
+int
+g84_disp_base_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		  struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_base_new_(&nv50_disp_dmac_func, &g84_disp_base_mthd,
+				   disp, 1, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegf119.c
index ebcb925e9d90..389e19dfc514 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegf119.c
@@ -21,10 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 gf119_disp_base_mthd_base = {
@@ -91,7 +88,7 @@ gf119_disp_base_mthd_image = {
 };
 
 const struct nv50_disp_chan_mthd
-gf119_disp_base_chan_mthd = {
+gf119_disp_base_mthd = {
 	.name = "Base",
 	.addr = 0x001000,
 	.prev = -0x020000,
@@ -102,13 +99,10 @@ gf119_disp_base_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-gf119_disp_base_oclass = {
-	.base.oclass = GF110_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &gf119_disp_dmac_func,
-	.mthd = &gf119_disp_base_chan_mthd,
-	.chid = 1,
-};
+int
+gf119_disp_base_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_base_new_(&gf119_disp_dmac_func, &gf119_disp_base_mthd,
+				   disp, 1, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp102.c
index 8a3cdeef8d2c..0cb23d673aa0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basegp102.c
@@ -21,18 +21,12 @@
  *
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
+#include "channv50.h"
 
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gp102_disp_base_oclass = {
-	.base.oclass = GK110_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &gp102_disp_dmac_func,
-	.mthd = &gf119_disp_base_chan_mthd,
-	.chid = 1,
-};
+int
+gp102_disp_base_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_base_new_(&gp102_disp_dmac_func, &gf119_disp_base_mthd,
+				   disp, 1, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c
index f1d6b820d482..19eb7dde01f2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/basenv50.c
@@ -21,33 +21,30 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 #include "head.h"
-#include "rootnv50.h"
 
 #include <core/client.h>
 
-#include <nvif/class.h>
 #include <nvif/cl507c.h>
 #include <nvif/unpack.h>
 
 int
-nv50_disp_base_new(const struct nv50_disp_dmac_func *func,
-		   const struct nv50_disp_chan_mthd *mthd,
-		   struct nv50_disp_root *root, int chid,
-		   const struct nvkm_oclass *oclass, void *data, u32 size,
-		   struct nvkm_object **pobject)
+nv50_disp_base_new_(const struct nv50_disp_chan_func *func,
+		    const struct nv50_disp_chan_mthd *mthd,
+		    struct nv50_disp *disp, int chid,
+		    const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nvkm_object **pobject)
 {
 	union {
 		struct nv50_disp_base_channel_dma_v0 v0;
-	} *args = data;
+	} *args = argv;
 	struct nvkm_object *parent = oclass->parent;
-	struct nv50_disp *disp = root->disp;
 	int head, ret = -ENOSYS;
 	u64 push;
 
-	nvif_ioctl(parent, "create disp base channel dma size %d\n", size);
-	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+	nvif_ioctl(parent, "create disp base channel dma size %d\n", argc);
+	if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))) {
 		nvif_ioctl(parent, "create disp base channel dma vers %d "
 				   "pushbuf %016llx head %d\n",
 			   args->v0.version, args->v0.pushbuf, args->v0.head);
@@ -58,7 +55,7 @@ nv50_disp_base_new(const struct nv50_disp_dmac_func *func,
 	} else
 		return ret;
 
-	return nv50_disp_dmac_new_(func, mthd, root, chid + head,
+	return nv50_disp_dmac_new_(func, mthd, disp, chid + head,
 				   head, push, oclass, pobject);
 }
 
@@ -102,7 +99,7 @@ nv50_disp_base_mthd_image = {
 };
 
 static const struct nv50_disp_chan_mthd
-nv50_disp_base_chan_mthd = {
+nv50_disp_base_mthd = {
 	.name = "Base",
 	.addr = 0x000540,
 	.prev = 0x000004,
@@ -113,13 +110,10 @@ nv50_disp_base_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-nv50_disp_base_oclass = {
-	.base.oclass = NV50_DISP_BASE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_base_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &nv50_disp_base_chan_mthd,
-	.chid = 1,
-};
+int
+nv50_disp_base_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		   struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_base_new_(&nv50_disp_dmac_func, &nv50_disp_base_mthd,
+				   disp, 1, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/changf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/changf119.c
index 17a3d835cb42..525f95d06429 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/changf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/changf119.c
@@ -47,3 +47,16 @@ gf119_disp_chan_uevent = {
 	.init = gf119_disp_chan_uevent_init,
 	.fini = gf119_disp_chan_uevent_fini,
 };
+
+void
+gf119_disp_chan_intr(struct nv50_disp_chan *chan, bool en)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 mask = 0x00000001 << chan->chid.user;
+	if (!en) {
+		nvkm_mask(device, 0x610090, mask, 0x00000000);
+		nvkm_mask(device, 0x6100a0, mask, 0x00000000);
+	} else {
+		nvkm_mask(device, 0x6100a0, mask, mask);
+	}
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/changv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/changv100.c
new file mode 100644
index 000000000000..75247c9c7e10
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/changv100.c
@@ -0,0 +1,34 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "channv50.h"
+
+const struct nvkm_event_func
+gv100_disp_chan_uevent = {
+	.ctor = nv50_disp_chan_uevent_ctor,
+};
+
+u64
+gv100_disp_chan_user(struct nv50_disp_chan *chan, u64 *psize)
+{
+	*psize = 0x1000;
+	return 0x690000 + ((chan->chid.user - 1) * 0x1000);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
index 723dcbde2ac2..bcf32d92ee5a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.c
@@ -26,6 +26,7 @@
 
 #include <core/client.h>
 #include <core/notify.h>
+#include <core/oproxy.h>
 #include <core/ramht.h>
 #include <engine/dma.h>
 
@@ -65,7 +66,7 @@ nv50_disp_mthd_list(struct nv50_disp *disp, int debug, u32 base, int c,
 void
 nv50_disp_chan_mthd(struct nv50_disp_chan *chan, int debug)
 {
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	const struct nv50_disp_chan_mthd *mthd = chan->mthd;
 	const struct nv50_disp_mthd_list *list;
@@ -154,13 +155,29 @@ nv50_disp_chan_uevent = {
 	.fini = nv50_disp_chan_uevent_fini,
 };
 
+u64
+nv50_disp_chan_user(struct nv50_disp_chan *chan, u64 *psize)
+{
+	*psize = 0x1000;
+	return 0x640000 + (chan->chid.user * 0x1000);
+}
+
+void
+nv50_disp_chan_intr(struct nv50_disp_chan *chan, bool en)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 mask = 0x00010001 << chan->chid.user;
+	const u32 data = en ? 0x00010000 << chan->chid.user : 0x00000000;
+	nvkm_mask(device, 0x610028, mask, data);
+}
+
 static int
 nv50_disp_chan_rd32(struct nvkm_object *object, u64 addr, u32 *data)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	struct nv50_disp *disp = chan->root->disp;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
-	*data = nvkm_rd32(device, 0x640000 + (chan->chid.user * 0x1000) + addr);
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	u64 size, base = chan->func->user(chan, &size);
+	*data = nvkm_rd32(device, base + addr);
 	return 0;
 }
 
@@ -168,9 +185,9 @@ static int
 nv50_disp_chan_wr32(struct nvkm_object *object, u64 addr, u32 data)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	struct nv50_disp *disp = chan->root->disp;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
-	nvkm_wr32(device, 0x640000 + (chan->chid.user * 0x1000) + addr, data);
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	u64 size, base = chan->func->user(chan, &size);
+	nvkm_wr32(device, base + addr, data);
 	return 0;
 }
 
@@ -179,7 +196,7 @@ nv50_disp_chan_ntfy(struct nvkm_object *object, u32 type,
 		    struct nvkm_event **pevent)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	switch (type) {
 	case NV50_DISP_CORE_CHANNEL_DMA_V0_NTFY_UEVENT:
 		*pevent = &disp->uevent;
@@ -195,34 +212,83 @@ nv50_disp_chan_map(struct nvkm_object *object, void *argv, u32 argc,
 		   enum nvkm_object_map *type, u64 *addr, u64 *size)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	struct nv50_disp *disp = chan->root->disp;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u64 base = device->func->resource_addr(device, 0);
 	*type = NVKM_OBJECT_MAP_IO;
-	*addr = device->func->resource_addr(device, 0) +
-		0x640000 + (chan->chid.user * 0x1000);
-	*size = 0x001000;
+	*addr = base + chan->func->user(chan, size);
 	return 0;
 }
 
+struct nv50_disp_chan_object {
+	struct nvkm_oproxy oproxy;
+	struct nv50_disp *disp;
+	int hash;
+};
+
+static void
+nv50_disp_chan_child_del_(struct nvkm_oproxy *base)
+{
+	struct nv50_disp_chan_object *object =
+		container_of(base, typeof(*object), oproxy);
+	nvkm_ramht_remove(object->disp->ramht, object->hash);
+}
+
+static const struct nvkm_oproxy_func
+nv50_disp_chan_child_func_ = {
+	.dtor[0] = nv50_disp_chan_child_del_,
+};
+
 static int
 nv50_disp_chan_child_new(const struct nvkm_oclass *oclass,
-			 void *data, u32 size, struct nvkm_object **pobject)
+			 void *argv, u32 argc, struct nvkm_object **pobject)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(oclass->parent);
-	return chan->func->child_new(chan, oclass, data, size, pobject);
+	struct nv50_disp *disp = chan->disp;
+	struct nvkm_device *device = disp->base.engine.subdev.device;
+	const struct nvkm_device_oclass *sclass = oclass->priv;
+	struct nv50_disp_chan_object *object;
+	int ret;
+
+	if (!(object = kzalloc(sizeof(*object), GFP_KERNEL)))
+		return -ENOMEM;
+	nvkm_oproxy_ctor(&nv50_disp_chan_child_func_, oclass, &object->oproxy);
+	object->disp = disp;
+	*pobject = &object->oproxy.base;
+
+	ret = sclass->ctor(device, oclass, argv, argc, &object->oproxy.object);
+	if (ret)
+		return ret;
+
+	object->hash = chan->func->bind(chan, object->oproxy.object,
+					      oclass->handle);
+	if (object->hash < 0)
+		return object->hash;
+
+	return 0;
 }
 
 static int
 nv50_disp_chan_child_get(struct nvkm_object *object, int index,
-			 struct nvkm_oclass *oclass)
+			 struct nvkm_oclass *sclass)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	if (chan->func->child_get) {
-		int ret = chan->func->child_get(chan, index, oclass);
-		if (ret == 0)
-			oclass->ctor = nv50_disp_chan_child_new;
-		return ret;
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const struct nvkm_device_oclass *oclass = NULL;
+
+	if (chan->func->bind)
+		sclass->engine = nvkm_device_engine(device, NVKM_ENGINE_DMAOBJ);
+	else
+		sclass->engine = NULL;
+
+	if (sclass->engine && sclass->engine->func->base.sclass) {
+		sclass->engine->func->base.sclass(sclass, index, &oclass);
+		if (oclass) {
+			sclass->ctor = nv50_disp_chan_child_new,
+			sclass->priv = oclass;
+			return 0;
+		}
 	}
+
 	return -EINVAL;
 }
 
@@ -231,6 +297,7 @@ nv50_disp_chan_fini(struct nvkm_object *object, bool suspend)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
 	chan->func->fini(chan);
+	chan->func->intr(chan, false);
 	return 0;
 }
 
@@ -238,6 +305,7 @@ static int
 nv50_disp_chan_init(struct nvkm_object *object)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
+	chan->func->intr(chan, true);
 	return chan->func->init(chan);
 }
 
@@ -245,10 +313,11 @@ static void *
 nv50_disp_chan_dtor(struct nvkm_object *object)
 {
 	struct nv50_disp_chan *chan = nv50_disp_chan(object);
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	if (chan->chid.user >= 0)
 		disp->chan[chan->chid.user] = NULL;
-	return chan->func->dtor ? chan->func->dtor(chan) : chan;
+	nvkm_memory_unref(&chan->memory);
+	return chan;
 }
 
 static const struct nvkm_object_func
@@ -264,18 +333,22 @@ nv50_disp_chan = {
 };
 
 int
-nv50_disp_chan_ctor(const struct nv50_disp_chan_func *func,
+nv50_disp_chan_new_(const struct nv50_disp_chan_func *func,
 		    const struct nv50_disp_chan_mthd *mthd,
-		    struct nv50_disp_root *root, int ctrl, int user, int head,
+		    struct nv50_disp *disp, int ctrl, int user, int head,
 		    const struct nvkm_oclass *oclass,
-		    struct nv50_disp_chan *chan)
+		    struct nvkm_object **pobject)
 {
-	struct nv50_disp *disp = root->disp;
+	struct nv50_disp_chan *chan;
+
+	if (!(chan = kzalloc(sizeof(*chan), GFP_KERNEL)))
+		return -ENOMEM;
+	*pobject = &chan->object;
 
 	nvkm_object_ctor(&nv50_disp_chan, oclass, &chan->object);
 	chan->func = func;
 	chan->mthd = mthd;
-	chan->root = root;
+	chan->disp = disp;
 	chan->chid.ctrl = ctrl;
 	chan->chid.user = user;
 	chan->head = head;
@@ -287,20 +360,3 @@ nv50_disp_chan_ctor(const struct nv50_disp_chan_func *func,
 	disp->chan[chan->chid.user] = chan;
 	return 0;
 }
-
-int
-nv50_disp_chan_new_(const struct nv50_disp_chan_func *func,
-		    const struct nv50_disp_chan_mthd *mthd,
-		    struct nv50_disp_root *root, int ctrl, int user, int head,
-		    const struct nvkm_oclass *oclass,
-		    struct nvkm_object **pobject)
-{
-	struct nv50_disp_chan *chan;
-
-	if (!(chan = kzalloc(sizeof(*chan), GFP_KERNEL)))
-		return -ENOMEM;
-	*pobject = &chan->object;
-
-	return nv50_disp_chan_ctor(func, mthd, root, ctrl, user,
-				   head, oclass, chan);
-}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
index 40681db91a02..adc9d76d09cc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/channv50.h
@@ -4,11 +4,12 @@
 #define nv50_disp_chan(p) container_of((p), struct nv50_disp_chan, object)
 #include <core/object.h>
 #include "nv50.h"
+struct nv50_disp_root;
 
 struct nv50_disp_chan {
 	const struct nv50_disp_chan_func *func;
 	const struct nv50_disp_chan_mthd *mthd;
-	struct nv50_disp_root *root;
+	struct nv50_disp *disp;
 
 	struct {
 		int ctrl;
@@ -17,36 +18,133 @@ struct nv50_disp_chan {
 	int head;
 
 	struct nvkm_object object;
+
+	struct nvkm_memory *memory;
+	u64 push;
 };
 
 struct nv50_disp_chan_func {
-	void *(*dtor)(struct nv50_disp_chan *);
 	int (*init)(struct nv50_disp_chan *);
 	void (*fini)(struct nv50_disp_chan *);
-	int (*child_get)(struct nv50_disp_chan *, int index,
-			 struct nvkm_oclass *);
-	int (*child_new)(struct nv50_disp_chan *, const struct nvkm_oclass *,
-			 void *data, u32 size, struct nvkm_object **);
+	void (*intr)(struct nv50_disp_chan *, bool en);
+	u64 (*user)(struct nv50_disp_chan *, u64 *size);
+	int (*bind)(struct nv50_disp_chan *, struct nvkm_object *, u32 handle);
 };
 
-int nv50_disp_chan_ctor(const struct nv50_disp_chan_func *,
-			const struct nv50_disp_chan_mthd *,
-			struct nv50_disp_root *, int ctrl, int user, int head,
-			const struct nvkm_oclass *, struct nv50_disp_chan *);
 int nv50_disp_chan_new_(const struct nv50_disp_chan_func *,
 			const struct nv50_disp_chan_mthd *,
-			struct nv50_disp_root *, int ctrl, int user, int head,
+			struct nv50_disp *, int ctrl, int user, int head,
+			const struct nvkm_oclass *, struct nvkm_object **);
+int nv50_disp_dmac_new_(const struct nv50_disp_chan_func *,
+			const struct nv50_disp_chan_mthd *,
+			struct nv50_disp *, int chid, int head, u64 push,
 			const struct nvkm_oclass *, struct nvkm_object **);
 
+void nv50_disp_chan_intr(struct nv50_disp_chan *, bool);
+u64 nv50_disp_chan_user(struct nv50_disp_chan *, u64 *);
 extern const struct nv50_disp_chan_func nv50_disp_pioc_func;
-extern const struct nv50_disp_chan_func gf119_disp_pioc_func;
-
-extern const struct nvkm_event_func nv50_disp_chan_uevent;
-int  nv50_disp_chan_uevent_ctor(struct nvkm_object *, void *, u32,
-				struct nvkm_notify *);
-void nv50_disp_chan_uevent_send(struct nv50_disp *, int);
+extern const struct nv50_disp_chan_func nv50_disp_dmac_func;
+int nv50_disp_dmac_bind(struct nv50_disp_chan *, struct nvkm_object *, u32);
+extern const struct nv50_disp_chan_func nv50_disp_core_func;
 
-extern const struct nvkm_event_func gf119_disp_chan_uevent;
+void gf119_disp_chan_intr(struct nv50_disp_chan *, bool);
+extern const struct nv50_disp_chan_func gf119_disp_pioc_func;
+extern const struct nv50_disp_chan_func gf119_disp_dmac_func;
+void gf119_disp_dmac_fini(struct nv50_disp_chan *);
+int gf119_disp_dmac_bind(struct nv50_disp_chan *, struct nvkm_object *, u32);
+extern const struct nv50_disp_chan_func gf119_disp_core_func;
+void gf119_disp_core_fini(struct nv50_disp_chan *);
+
+extern const struct nv50_disp_chan_func gp102_disp_dmac_func;
+
+u64 gv100_disp_chan_user(struct nv50_disp_chan *, u64 *);
+int gv100_disp_dmac_init(struct nv50_disp_chan *);
+void gv100_disp_dmac_fini(struct nv50_disp_chan *);
+int gv100_disp_dmac_bind(struct nv50_disp_chan *, struct nvkm_object *, u32);
+
+int nv50_disp_curs_new_(const struct nv50_disp_chan_func *,
+			struct nv50_disp *, int ctrl, int user,
+			const struct nvkm_oclass *, void *argv, u32 argc,
+			struct nvkm_object **);
+int nv50_disp_oimm_new_(const struct nv50_disp_chan_func *,
+			struct nv50_disp *, int ctrl, int user,
+			const struct nvkm_oclass *, void *argv, u32 argc,
+			struct nvkm_object **);
+int nv50_disp_base_new_(const struct nv50_disp_chan_func *,
+			const struct nv50_disp_chan_mthd *,
+			struct nv50_disp *, int chid,
+			const struct nvkm_oclass *, void *argv, u32 argc,
+			struct nvkm_object **);
+int nv50_disp_core_new_(const struct nv50_disp_chan_func *,
+			const struct nv50_disp_chan_mthd *,
+			struct nv50_disp *, int chid,
+			const struct nvkm_oclass *oclass, void *argv, u32 argc,
+			struct nvkm_object **);
+int nv50_disp_ovly_new_(const struct nv50_disp_chan_func *,
+			const struct nv50_disp_chan_mthd *,
+			struct nv50_disp *, int chid,
+			const struct nvkm_oclass *, void *argv, u32 argc,
+			struct nvkm_object **);
+
+int nv50_disp_curs_new(const struct nvkm_oclass *, void *, u32,
+		       struct nv50_disp *, struct nvkm_object **);
+int nv50_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
+		       struct nv50_disp *, struct nvkm_object **);
+int nv50_disp_base_new(const struct nvkm_oclass *, void *, u32,
+		       struct nv50_disp *, struct nvkm_object **);
+int nv50_disp_core_new(const struct nvkm_oclass *, void *, u32,
+		       struct nv50_disp *, struct nvkm_object **);
+int nv50_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
+		       struct nv50_disp *, struct nvkm_object **);
+
+int g84_disp_base_new(const struct nvkm_oclass *, void *, u32,
+		      struct nv50_disp *, struct nvkm_object **);
+int g84_disp_core_new(const struct nvkm_oclass *, void *, u32,
+		      struct nv50_disp *, struct nvkm_object **);
+int g84_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
+		      struct nv50_disp *, struct nvkm_object **);
+
+int g94_disp_core_new(const struct nvkm_oclass *, void *, u32,
+		      struct nv50_disp *, struct nvkm_object **);
+
+int gt200_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+
+int gf119_disp_curs_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gf119_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gf119_disp_base_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gf119_disp_core_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gf119_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+
+int gk104_disp_core_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gk104_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+
+int gp102_disp_curs_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gp102_disp_oimm_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gp102_disp_base_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gp102_disp_core_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gp102_disp_ovly_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+
+int gv100_disp_curs_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gv100_disp_wimm_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gv100_disp_core_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
+int gv100_disp_wndw_new(const struct nvkm_oclass *, void *, u32,
+			struct nv50_disp *, struct nvkm_object **);
 
 struct nv50_disp_mthd_list {
 	u32 mthd;
@@ -76,64 +174,18 @@ extern const struct nv50_disp_mthd_list nv50_disp_core_mthd_sor;
 extern const struct nv50_disp_mthd_list nv50_disp_core_mthd_pior;
 extern const struct nv50_disp_mthd_list nv50_disp_base_mthd_image;
 
-extern const struct nv50_disp_chan_mthd g84_disp_core_chan_mthd;
+extern const struct nv50_disp_chan_mthd g84_disp_core_mthd;
 extern const struct nv50_disp_mthd_list g84_disp_core_mthd_dac;
 extern const struct nv50_disp_mthd_list g84_disp_core_mthd_head;
-extern const struct nv50_disp_chan_mthd g84_disp_base_chan_mthd;
-extern const struct nv50_disp_chan_mthd g84_disp_ovly_chan_mthd;
 
-extern const struct nv50_disp_chan_mthd g94_disp_core_chan_mthd;
+extern const struct nv50_disp_chan_mthd g94_disp_core_mthd;
 
 extern const struct nv50_disp_mthd_list gf119_disp_core_mthd_base;
 extern const struct nv50_disp_mthd_list gf119_disp_core_mthd_dac;
 extern const struct nv50_disp_mthd_list gf119_disp_core_mthd_sor;
 extern const struct nv50_disp_mthd_list gf119_disp_core_mthd_pior;
-extern const struct nv50_disp_chan_mthd gf119_disp_base_chan_mthd;
-
-extern const struct nv50_disp_chan_mthd gk104_disp_core_chan_mthd;
-extern const struct nv50_disp_chan_mthd gk104_disp_ovly_chan_mthd;
-
-struct nv50_disp_pioc_oclass {
-	int (*ctor)(const struct nv50_disp_chan_func *,
-		    const struct nv50_disp_chan_mthd *,
-		    struct nv50_disp_root *, int ctrl, int user,
-		    const struct nvkm_oclass *, void *data, u32 size,
-		    struct nvkm_object **);
-	struct nvkm_sclass base;
-	const struct nv50_disp_chan_func *func;
-	const struct nv50_disp_chan_mthd *mthd;
-	struct {
-		int ctrl;
-		int user;
-	} chid;
-};
-
-extern const struct nv50_disp_pioc_oclass nv50_disp_oimm_oclass;
-extern const struct nv50_disp_pioc_oclass nv50_disp_curs_oclass;
-
-extern const struct nv50_disp_pioc_oclass g84_disp_oimm_oclass;
-extern const struct nv50_disp_pioc_oclass g84_disp_curs_oclass;
-
-extern const struct nv50_disp_pioc_oclass gt215_disp_oimm_oclass;
-extern const struct nv50_disp_pioc_oclass gt215_disp_curs_oclass;
-
-extern const struct nv50_disp_pioc_oclass gf119_disp_oimm_oclass;
-extern const struct nv50_disp_pioc_oclass gf119_disp_curs_oclass;
-
-extern const struct nv50_disp_pioc_oclass gk104_disp_oimm_oclass;
-extern const struct nv50_disp_pioc_oclass gk104_disp_curs_oclass;
-
-extern const struct nv50_disp_pioc_oclass gp102_disp_oimm_oclass;
-extern const struct nv50_disp_pioc_oclass gp102_disp_curs_oclass;
+extern const struct nv50_disp_chan_mthd gf119_disp_base_mthd;
 
-int nv50_disp_curs_new(const struct nv50_disp_chan_func *,
-		       const struct nv50_disp_chan_mthd *,
-		       struct nv50_disp_root *, int ctrl, int user,
-		       const struct nvkm_oclass *, void *data, u32 size,
-		       struct nvkm_object **);
-int nv50_disp_oimm_new(const struct nv50_disp_chan_func *,
-		       const struct nv50_disp_chan_mthd *,
-		       struct nv50_disp_root *, int ctrl, int user,
-		       const struct nvkm_oclass *, void *data, u32 size,
-		       struct nvkm_object **);
+extern const struct nv50_disp_chan_mthd gk104_disp_core_mthd;
+extern const struct nv50_disp_chan_mthd gk104_disp_ovly_mthd;
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg84.c
index 1baa5c34b327..cfc54aad3e7c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg84.c
@@ -21,10 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
+#include "channv50.h"
 
 const struct nv50_disp_mthd_list
 g84_disp_core_mthd_dac = {
@@ -91,7 +88,7 @@ g84_disp_core_mthd_head = {
 };
 
 const struct nv50_disp_chan_mthd
-g84_disp_core_chan_mthd = {
+g84_disp_core_mthd = {
 	.name = "Core",
 	.addr = 0x000000,
 	.prev = 0x000004,
@@ -105,13 +102,10 @@ g84_disp_core_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-g84_disp_core_oclass = {
-	.base.oclass = G82_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &nv50_disp_core_func,
-	.mthd = &g84_disp_core_chan_mthd,
-	.chid = 0,
-};
+int
+g84_disp_core_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		  struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_core_new_(&nv50_disp_core_func, &g84_disp_core_mthd,
+				   disp, 0, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg94.c
index c65c9f3ff69f..e911925f1182 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coreg94.c
@@ -21,10 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 g94_disp_core_mthd_sor = {
@@ -37,7 +34,7 @@ g94_disp_core_mthd_sor = {
 };
 
 const struct nv50_disp_chan_mthd
-g94_disp_core_chan_mthd = {
+g94_disp_core_mthd = {
 	.name = "Core",
 	.addr = 0x000000,
 	.prev = 0x000004,
@@ -51,13 +48,10 @@ g94_disp_core_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-g94_disp_core_oclass = {
-	.base.oclass = GT206_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &nv50_disp_core_func,
-	.mthd = &g94_disp_core_chan_mthd,
-	.chid = 0,
-};
+int
+g94_disp_core_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		  struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_core_new_(&nv50_disp_core_func, &g94_disp_core_mthd,
+				   disp, 0, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
index 21fbf89b6319..d162b9cf4eac 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregf119.c
@@ -21,15 +21,10 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
+#include "channv50.h"
 
-#include <core/client.h>
 #include <subdev/timer.h>
 
-#include <nvif/class.h>
-#include <nvif/unpack.h>
-
 const struct nv50_disp_mthd_list
 gf119_disp_core_mthd_base = {
 	.mthd = 0x0000,
@@ -157,7 +152,7 @@ gf119_disp_core_mthd_head = {
 };
 
 static const struct nv50_disp_chan_mthd
-gf119_disp_core_chan_mthd = {
+gf119_disp_core_mthd = {
 	.name = "Core",
 	.addr = 0x000000,
 	.prev = -0x020000,
@@ -172,10 +167,9 @@ gf119_disp_core_chan_mthd = {
 };
 
 void
-gf119_disp_core_fini(struct nv50_disp_dmac *chan)
+gf119_disp_core_fini(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
 	/* deactivate channel */
@@ -188,22 +182,14 @@ gf119_disp_core_fini(struct nv50_disp_dmac *chan)
 		nvkm_error(subdev, "core fini: %08x\n",
 			   nvkm_rd32(device, 0x610490));
 	}
-
-	/* disable error reporting and completion notification */
-	nvkm_mask(device, 0x610090, 0x00000001, 0x00000000);
-	nvkm_mask(device, 0x6100a0, 0x00000001, 0x00000000);
 }
 
 static int
-gf119_disp_core_init(struct nv50_disp_dmac *chan)
+gf119_disp_core_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
-	/* enable error reporting */
-	nvkm_mask(device, 0x6100a0, 0x00000001, 0x00000001);
-
 	/* initialise channel for dma command submission */
 	nvkm_wr32(device, 0x610494, chan->push);
 	nvkm_wr32(device, 0x610498, 0x00010000);
@@ -225,20 +211,19 @@ gf119_disp_core_init(struct nv50_disp_dmac *chan)
 	return 0;
 }
 
-const struct nv50_disp_dmac_func
+const struct nv50_disp_chan_func
 gf119_disp_core_func = {
 	.init = gf119_disp_core_init,
 	.fini = gf119_disp_core_fini,
+	.intr = gf119_disp_chan_intr,
+	.user = nv50_disp_chan_user,
 	.bind = gf119_disp_dmac_bind,
 };
 
-const struct nv50_disp_dmac_oclass
-gf119_disp_core_oclass = {
-	.base.oclass = GF110_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &gf119_disp_core_func,
-	.mthd = &gf119_disp_core_chan_mthd,
-	.chid = 0,
-};
+int
+gf119_disp_core_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_core_new_(&gf119_disp_core_func, &gf119_disp_core_mthd,
+				   disp, 0, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk104.c
index 088ab222e823..5c800174e079 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk104.c
@@ -21,10 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 gk104_disp_core_mthd_head = {
@@ -106,7 +103,7 @@ gk104_disp_core_mthd_head = {
 };
 
 const struct nv50_disp_chan_mthd
-gk104_disp_core_chan_mthd = {
+gk104_disp_core_mthd = {
 	.name = "Core",
 	.addr = 0x000000,
 	.prev = -0x020000,
@@ -120,13 +117,10 @@ gk104_disp_core_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-gk104_disp_core_oclass = {
-	.base.oclass = GK104_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &gf119_disp_core_func,
-	.mthd = &gk104_disp_core_chan_mthd,
-	.chid = 0,
-};
+int
+gk104_disp_core_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_core_new_(&gf119_disp_core_func, &gk104_disp_core_mthd,
+				   disp, 0, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk110.c
deleted file mode 100644
index df0f45c20108..000000000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregk110.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gk110_disp_core_oclass = {
-	.base.oclass = GK110_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &gf119_disp_core_func,
-	.mthd = &gk104_disp_core_chan_mthd,
-	.chid = 0,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregm107.c
deleted file mode 100644
index 9e27f8fd98b6..000000000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregm107.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gm107_disp_core_oclass = {
-	.base.oclass = GM107_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &gf119_disp_core_func,
-	.mthd = &gk104_disp_core_chan_mthd,
-	.chid = 0,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregm200.c
deleted file mode 100644
index bb23a8658ac0..000000000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregm200.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gm200_disp_core_oclass = {
-	.base.oclass = GM200_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &gf119_disp_core_func,
-	.mthd = &gk104_disp_core_chan_mthd,
-	.chid = 0,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp100.c
deleted file mode 100644
index d5dff6619d4d..000000000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp100.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gp100_disp_core_oclass = {
-	.base.oclass = GP100_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &gf119_disp_core_func,
-	.mthd = &gk104_disp_core_chan_mthd,
-	.chid = 0,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
index b0df4b752b8c..5b7f993c73c7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregp102.c
@@ -21,23 +21,16 @@
  *
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
+#include "channv50.h"
 
 #include <subdev/timer.h>
 
-#include <nvif/class.h>
-
 static int
-gp102_disp_core_init(struct nv50_disp_dmac *chan)
+gp102_disp_core_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
-	/* enable error reporting */
-	nvkm_mask(device, 0x6100a0, 0x00000001, 0x00000001);
-
 	/* initialise channel for dma command submission */
 	nvkm_wr32(device, 0x611494, chan->push);
 	nvkm_wr32(device, 0x611498, 0x00010000);
@@ -59,20 +52,19 @@ gp102_disp_core_init(struct nv50_disp_dmac *chan)
 	return 0;
 }
 
-static const struct nv50_disp_dmac_func
+static const struct nv50_disp_chan_func
 gp102_disp_core_func = {
 	.init = gp102_disp_core_init,
 	.fini = gf119_disp_core_fini,
+	.intr = gf119_disp_chan_intr,
+	.user = nv50_disp_chan_user,
 	.bind = gf119_disp_dmac_bind,
 };
 
-const struct nv50_disp_dmac_oclass
-gp102_disp_core_oclass = {
-	.base.oclass = GP102_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &gp102_disp_core_func,
-	.mthd = &gk104_disp_core_chan_mthd,
-	.chid = 0,
-};
+int
+gp102_disp_core_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_core_new_(&gp102_disp_core_func, &gk104_disp_core_mthd,
+				   disp, 0, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregt200.c
deleted file mode 100644
index b234547708fc..000000000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregt200.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gt200_disp_core_oclass = {
-	.base.oclass = GT200_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &nv50_disp_core_func,
-	.mthd = &g84_disp_core_chan_mthd,
-	.chid = 0,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregt215.c
deleted file mode 100644
index 8f5ba2018975..000000000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregt215.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gt215_disp_core_oclass = {
-	.base.oclass = GT214_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &nv50_disp_core_func,
-	.mthd = &g94_disp_core_chan_mthd,
-	.chid = 0,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregv100.c
new file mode 100644
index 000000000000..4592d0e69fec
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/coregv100.c
@@ -0,0 +1,204 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "channv50.h"
+
+#include <subdev/timer.h>
+
+const struct nv50_disp_mthd_list
+gv100_disp_core_mthd_base = {
+	.mthd = 0x0000,
+	.addr = 0x000000,
+	.data = {
+		{ 0x0200, 0x680200 },
+		{ 0x0208, 0x680208 },
+		{ 0x020c, 0x68020c },
+		{ 0x0210, 0x680210 },
+		{ 0x0214, 0x680214 },
+		{ 0x0218, 0x680218 },
+		{ 0x021c, 0x68021c },
+		{}
+	}
+};
+
+const struct nv50_disp_mthd_list
+gv100_disp_core_mthd_sor = {
+	.mthd = 0x0020,
+	.addr = 0x000020,
+	.data = {
+		{ 0x0300, 0x680300 },
+		{ 0x0304, 0x680304 },
+		{ 0x0308, 0x680308 },
+		{ 0x030c, 0x68030c },
+		{}
+	}
+};
+
+static const struct nv50_disp_mthd_list
+gv100_disp_core_mthd_wndw = {
+	.mthd = 0x0080,
+	.addr = 0x000080,
+	.data = {
+		{ 0x1000, 0x681000 },
+		{ 0x1004, 0x681004 },
+		{ 0x1008, 0x681008 },
+		{ 0x100c, 0x68100c },
+		{ 0x1010, 0x681010 },
+		{}
+	}
+};
+
+static const struct nv50_disp_mthd_list
+gv100_disp_core_mthd_head = {
+	.mthd = 0x0400,
+	.addr = 0x000400,
+	.data = {
+		{ 0x2000, 0x682000 },
+		{ 0x2004, 0x682004 },
+		{ 0x2008, 0x682008 },
+		{ 0x200c, 0x68200c },
+		{ 0x2014, 0x682014 },
+		{ 0x2018, 0x682018 },
+		{ 0x201c, 0x68201c },
+		{ 0x2020, 0x682020 },
+		{ 0x2028, 0x682028 },
+		{ 0x202c, 0x68202c },
+		{ 0x2030, 0x682030 },
+		{ 0x2038, 0x682038 },
+		{ 0x203c, 0x68203c },
+		{ 0x2048, 0x682048 },
+		{ 0x204c, 0x68204c },
+		{ 0x2050, 0x682050 },
+		{ 0x2054, 0x682054 },
+		{ 0x2058, 0x682058 },
+		{ 0x205c, 0x68205c },
+		{ 0x2060, 0x682060 },
+		{ 0x2064, 0x682064 },
+		{ 0x2068, 0x682068 },
+		{ 0x206c, 0x68206c },
+		{ 0x2070, 0x682070 },
+		{ 0x2074, 0x682074 },
+		{ 0x2078, 0x682078 },
+		{ 0x207c, 0x68207c },
+		{ 0x2080, 0x682080 },
+		{ 0x2088, 0x682088 },
+		{ 0x2090, 0x682090 },
+		{ 0x209c, 0x68209c },
+		{ 0x20a0, 0x6820a0 },
+		{ 0x20a4, 0x6820a4 },
+		{ 0x20a8, 0x6820a8 },
+		{ 0x20ac, 0x6820ac },
+		{ 0x218c, 0x68218c },
+		{ 0x2194, 0x682194 },
+		{ 0x2198, 0x682198 },
+		{ 0x219c, 0x68219c },
+		{ 0x21a0, 0x6821a0 },
+		{ 0x21a4, 0x6821a4 },
+		{ 0x2214, 0x682214 },
+		{ 0x2218, 0x682218 },
+		{}
+	}
+};
+
+static const struct nv50_disp_chan_mthd
+gv100_disp_core_mthd = {
+	.name = "Core",
+	.addr = 0x000000,
+	.prev = 0x008000,
+	.data = {
+		{ "Global", 1, &gv100_disp_core_mthd_base },
+		{    "SOR", 4, &gv100_disp_core_mthd_sor  },
+		{ "WINDOW", 8, &gv100_disp_core_mthd_wndw },
+		{   "HEAD", 4, &gv100_disp_core_mthd_head },
+		{}
+	}
+};
+
+static int
+gv100_disp_core_idle(struct nv50_disp_chan *chan)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	nvkm_msec(device, 2000,
+		u32 stat = nvkm_rd32(device, 0x610630);
+		if ((stat & 0x001f0000) == 0x000b0000)
+			return 0;
+	);
+	return -EBUSY;
+}
+
+static u64
+gv100_disp_core_user(struct nv50_disp_chan *chan, u64 *psize)
+{
+	*psize = 0x10000;
+	return 0x680000;
+}
+
+static void
+gv100_disp_core_intr(struct nv50_disp_chan *chan, bool en)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 mask = 0x00000001;
+	const u32 data = en ? mask : 0;
+	nvkm_mask(device, 0x611dac, mask, data);
+}
+
+static void
+gv100_disp_core_fini(struct nv50_disp_chan *chan)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	nvkm_mask(device, 0x6104e0, 0x00000010, 0x00000000);
+	gv100_disp_core_idle(chan);
+	nvkm_mask(device, 0x6104e0, 0x00000002, 0x00000000);
+}
+
+static int
+gv100_disp_core_init(struct nv50_disp_chan *chan)
+{
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+
+	nvkm_wr32(device, 0x610b24, lower_32_bits(chan->push));
+	nvkm_wr32(device, 0x610b20, upper_32_bits(chan->push));
+	nvkm_wr32(device, 0x610b28, 0x00000001);
+	nvkm_wr32(device, 0x610b2c, 0x00000040);
+
+	nvkm_mask(device, 0x6104e0, 0x00000010, 0x00000010);
+	nvkm_wr32(device, 0x680000, 0x00000000);
+	nvkm_wr32(device, 0x6104e0, 0x00000013);
+	return gv100_disp_core_idle(chan);
+}
+
+static const struct nv50_disp_chan_func
+gv100_disp_core = {
+	.init = gv100_disp_core_init,
+	.fini = gv100_disp_core_fini,
+	.intr = gv100_disp_core_intr,
+	.user = gv100_disp_core_user,
+	.bind = gv100_disp_dmac_bind,
+};
+
+int
+gv100_disp_core_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_core_new_(&gv100_disp_core, &gv100_disp_core_mthd,
+				   disp, 0, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
index b547c8b833ca..55db9a22b4be 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/corenv50.c
@@ -21,32 +21,30 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
+#include "channv50.h"
 
 #include <core/client.h>
 #include <subdev/timer.h>
 
-#include <nvif/class.h>
 #include <nvif/cl507d.h>
 #include <nvif/unpack.h>
 
 int
-nv50_disp_core_new(const struct nv50_disp_dmac_func *func,
-		   const struct nv50_disp_chan_mthd *mthd,
-		   struct nv50_disp_root *root, int chid,
-		   const struct nvkm_oclass *oclass, void *data, u32 size,
-		   struct nvkm_object **pobject)
+nv50_disp_core_new_(const struct nv50_disp_chan_func *func,
+		    const struct nv50_disp_chan_mthd *mthd,
+		    struct nv50_disp *disp, int chid,
+		    const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nvkm_object **pobject)
 {
 	union {
 		struct nv50_disp_core_channel_dma_v0 v0;
-	} *args = data;
+	} *args = argv;
 	struct nvkm_object *parent = oclass->parent;
 	u64 push;
 	int ret = -ENOSYS;
 
-	nvif_ioctl(parent, "create disp core channel dma size %d\n", size);
-	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+	nvif_ioctl(parent, "create disp core channel dma size %d\n", argc);
+	if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))) {
 		nvif_ioctl(parent, "create disp core channel dma vers %d "
 				   "pushbuf %016llx\n",
 			   args->v0.version, args->v0.pushbuf);
@@ -54,7 +52,7 @@ nv50_disp_core_new(const struct nv50_disp_dmac_func *func,
 	} else
 		return ret;
 
-	return nv50_disp_dmac_new_(func, mthd, root, chid, 0,
+	return nv50_disp_dmac_new_(func, mthd, disp, chid, 0,
 				   push, oclass, pobject);
 }
 
@@ -151,7 +149,7 @@ nv50_disp_core_mthd_head = {
 };
 
 static const struct nv50_disp_chan_mthd
-nv50_disp_core_chan_mthd = {
+nv50_disp_core_mthd = {
 	.name = "Core",
 	.addr = 0x000000,
 	.prev = 0x000004,
@@ -166,10 +164,9 @@ nv50_disp_core_chan_mthd = {
 };
 
 static void
-nv50_disp_core_fini(struct nv50_disp_dmac *chan)
+nv50_disp_core_fini(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
 	/* deactivate channel */
@@ -182,21 +179,14 @@ nv50_disp_core_fini(struct nv50_disp_dmac *chan)
 		nvkm_error(subdev, "core fini: %08x\n",
 			   nvkm_rd32(device, 0x610200));
 	}
-
-	/* disable error reporting and completion notifications */
-	nvkm_mask(device, 0x610028, 0x00010001, 0x00000000);
 }
 
 static int
-nv50_disp_core_init(struct nv50_disp_dmac *chan)
+nv50_disp_core_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 
-	/* enable error reporting */
-	nvkm_mask(device, 0x610028, 0x00010000, 0x00010000);
-
 	/* attempt to unstick channel from some unknown state */
 	if ((nvkm_rd32(device, 0x610200) & 0x009f0000) == 0x00020000)
 		nvkm_mask(device, 0x610200, 0x00800000, 0x00800000);
@@ -224,20 +214,19 @@ nv50_disp_core_init(struct nv50_disp_dmac *chan)
 	return 0;
 }
 
-const struct nv50_disp_dmac_func
+const struct nv50_disp_chan_func
 nv50_disp_core_func = {
 	.init = nv50_disp_core_init,
 	.fini = nv50_disp_core_fini,
+	.intr = nv50_disp_chan_intr,
+	.user = nv50_disp_chan_user,
 	.bind = nv50_disp_dmac_bind,
 };
 
-const struct nv50_disp_dmac_oclass
-nv50_disp_core_oclass = {
-	.base.oclass = NV50_DISP_CORE_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_core_new,
-	.func = &nv50_disp_core_func,
-	.mthd = &nv50_disp_core_chan_mthd,
-	.chid = 0,
-};
+int
+nv50_disp_core_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		   struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_core_new_(&nv50_disp_core_func, &nv50_disp_core_mthd,
+				   disp, 0, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgf119.c
index 2be6fb052c65..cdda3658dcb3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgf119.c
@@ -22,16 +22,11 @@
  * Authors: Ben Skeggs
  */
 #include "channv50.h"
-#include "rootnv50.h"
 
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-gf119_disp_curs_oclass = {
-	.base.oclass = GF110_DISP_CURSOR,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_curs_new,
-	.func = &gf119_disp_pioc_func,
-	.chid = { 13, 13 },
-};
+int
+gf119_disp_curs_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_curs_new_(&gf119_disp_pioc_func, disp, 13, 13,
+				   oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgp102.c
index e958210d8105..1a4601f975e6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgp102.c
@@ -22,16 +22,11 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 #include "channv50.h"
-#include "rootnv50.h"
 
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-gp102_disp_curs_oclass = {
-	.base.oclass = GK104_DISP_CURSOR,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_curs_new,
-	.func = &gf119_disp_pioc_func,
-	.chid = { 13, 17 },
-};
+int
+gp102_disp_curs_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_curs_new_(&gf119_disp_pioc_func, disp, 13, 17,
+				   oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgv100.c
new file mode 100644
index 000000000000..a3e4f6900245
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursgv100.c
@@ -0,0 +1,81 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "channv50.h"
+
+#include <subdev/timer.h>
+
+static int
+gv100_disp_curs_idle(struct nv50_disp_chan *chan)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 soff = (chan->chid.ctrl - 1) * 0x04;
+	nvkm_msec(device, 2000,
+		u32 stat = nvkm_rd32(device, 0x610664 + soff);
+		if ((stat & 0x00070000) == 0x00040000)
+			return 0;
+	);
+	return -EBUSY;
+}
+
+static void
+gv100_disp_curs_intr(struct nv50_disp_chan *chan, bool en)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 mask = 0x00010000 << chan->head;
+	const u32 data = en ? mask : 0;
+	nvkm_mask(device, 0x611dac, mask, data);
+}
+
+static void
+gv100_disp_curs_fini(struct nv50_disp_chan *chan)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 hoff = chan->chid.ctrl * 4;
+	nvkm_mask(device, 0x6104e0 + hoff, 0x00000010, 0x00000010);
+	gv100_disp_curs_idle(chan);
+	nvkm_mask(device, 0x6104e0 + hoff, 0x00000001, 0x00000000);
+}
+
+static int
+gv100_disp_curs_init(struct nv50_disp_chan *chan)
+{
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	nvkm_wr32(device, 0x6104e0 + chan->chid.ctrl * 4, 0x00000001);
+	return gv100_disp_curs_idle(chan);
+}
+
+static const struct nv50_disp_chan_func
+gv100_disp_curs = {
+	.init = gv100_disp_curs_init,
+	.fini = gv100_disp_curs_fini,
+	.intr = gv100_disp_curs_intr,
+	.user = gv100_disp_chan_user,
+};
+
+int
+gv100_disp_curs_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_curs_new_(&gv100_disp_curs, disp, 73, 73,
+				   oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursnv50.c
index ab51121b7982..d29758504a5f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/cursnv50.c
@@ -23,30 +23,26 @@
  */
 #include "channv50.h"
 #include "head.h"
-#include "rootnv50.h"
 
 #include <core/client.h>
 
-#include <nvif/class.h>
 #include <nvif/cl507a.h>
 #include <nvif/unpack.h>
 
 int
-nv50_disp_curs_new(const struct nv50_disp_chan_func *func,
-		   const struct nv50_disp_chan_mthd *mthd,
-		   struct nv50_disp_root *root, int ctrl, int user,
-		   const struct nvkm_oclass *oclass, void *data, u32 size,
-		   struct nvkm_object **pobject)
+nv50_disp_curs_new_(const struct nv50_disp_chan_func *func,
+		    struct nv50_disp *disp, int ctrl, int user,
+		    const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nvkm_object **pobject)
 {
 	union {
 		struct nv50_disp_cursor_v0 v0;
-	} *args = data;
+	} *args = argv;
 	struct nvkm_object *parent = oclass->parent;
-	struct nv50_disp *disp = root->disp;
 	int head, ret = -ENOSYS;
 
-	nvif_ioctl(parent, "create disp cursor size %d\n", size);
-	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+	nvif_ioctl(parent, "create disp cursor size %d\n", argc);
+	if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))) {
 		nvif_ioctl(parent, "create disp cursor vers %d head %d\n",
 			   args->v0.version, args->v0.head);
 		if (!nvkm_head_find(&disp->base, args->v0.head))
@@ -55,16 +51,14 @@ nv50_disp_curs_new(const struct nv50_disp_chan_func *func,
 	} else
 		return ret;
 
-	return nv50_disp_chan_new_(func, mthd, root, ctrl + head, user + head,
+	return nv50_disp_chan_new_(func, NULL, disp, ctrl + head, user + head,
 				   head, oclass, pobject);
 }
 
-const struct nv50_disp_pioc_oclass
-nv50_disp_curs_oclass = {
-	.base.oclass = NV50_DISP_CURSOR,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_curs_new,
-	.func = &nv50_disp_pioc_func,
-	.chid = { 7, 7 },
-};
+int
+nv50_disp_curs_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		   struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_curs_new_(&nv50_disp_pioc_func, disp, 7, 7,
+				   oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacgf119.c
index dbd032ef352a..71a94777ea2e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacgf119.c
@@ -58,8 +58,13 @@ gf119_dac = {
 int
 gf119_dac_new(struct nvkm_disp *disp, int id)
 {
-	struct nvkm_device *device = disp->engine.subdev.device;
-	if (!(nvkm_rd32(device, 0x612004) & (0x00000010 << id)))
-		return 0;
 	return nvkm_ior_new_(&gf119_dac, disp, DAC, id);
 }
+
+int
+gf119_dac_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = (nvkm_rd32(device, 0x612004) & 0x000000f0) >> 4;
+	return 4;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacnv50.c
index 85e692b12260..558012db35f8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dacnv50.c
@@ -109,8 +109,13 @@ nv50_dac = {
 int
 nv50_dac_new(struct nvkm_disp *disp, int id)
 {
-	struct nvkm_device *device = disp->engine.subdev.device;
-	if (!(nvkm_rd32(device, 0x610184) & (0x00100000 << id)))
-		return 0;
 	return nvkm_ior_new_(&nv50_dac, disp, DAC, id);
 }
+
+int
+nv50_dac_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = (nvkm_rd32(device, 0x610184) & 0x00700000) >> 20;
+	return 3;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
index ce7cd74fbd5d..edf7dd0d931d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgf119.c
@@ -21,29 +21,27 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
+#include "channv50.h"
 
 #include <core/ramht.h>
 #include <subdev/timer.h>
 
 int
-gf119_disp_dmac_bind(struct nv50_disp_dmac *chan,
+gf119_disp_dmac_bind(struct nv50_disp_chan *chan,
 		     struct nvkm_object *object, u32 handle)
 {
-	return nvkm_ramht_insert(chan->base.root->ramht, object,
-				 chan->base.chid.user, -9, handle,
-				 chan->base.chid.user << 27 | 0x00000001);
+	return nvkm_ramht_insert(chan->disp->ramht, object,
+				 chan->chid.user, -9, handle,
+				 chan->chid.user << 27 | 0x00000001);
 }
 
 void
-gf119_disp_dmac_fini(struct nv50_disp_dmac *chan)
+gf119_disp_dmac_fini(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
-	int ctrl = chan->base.chid.ctrl;
-	int user = chan->base.chid.user;
+	int ctrl = chan->chid.ctrl;
+	int user = chan->chid.user;
 
 	/* deactivate channel */
 	nvkm_mask(device, 0x610490 + (ctrl * 0x0010), 0x00001010, 0x00001000);
@@ -55,23 +53,15 @@ gf119_disp_dmac_fini(struct nv50_disp_dmac *chan)
 		nvkm_error(subdev, "ch %d fini: %08x\n", user,
 			   nvkm_rd32(device, 0x610490 + (ctrl * 0x10)));
 	}
-
-	/* disable error reporting and completion notification */
-	nvkm_mask(device, 0x610090, 0x00000001 << user, 0x00000000);
-	nvkm_mask(device, 0x6100a0, 0x00000001 << user, 0x00000000);
 }
 
 static int
-gf119_disp_dmac_init(struct nv50_disp_dmac *chan)
+gf119_disp_dmac_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
-	int ctrl = chan->base.chid.ctrl;
-	int user = chan->base.chid.user;
-
-	/* enable error reporting */
-	nvkm_mask(device, 0x6100a0, 0x00000001 << user, 0x00000001 << user);
+	int ctrl = chan->chid.ctrl;
+	int user = chan->chid.user;
 
 	/* initialise channel for dma command submission */
 	nvkm_wr32(device, 0x610494 + (ctrl * 0x0010), chan->push);
@@ -94,9 +84,11 @@ gf119_disp_dmac_init(struct nv50_disp_dmac *chan)
 	return 0;
 }
 
-const struct nv50_disp_dmac_func
+const struct nv50_disp_chan_func
 gf119_disp_dmac_func = {
 	.init = gf119_disp_dmac_init,
 	.fini = gf119_disp_dmac_fini,
+	.intr = gf119_disp_chan_intr,
+	.user = nv50_disp_chan_user,
 	.bind = gf119_disp_dmac_bind,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
index cdead9500343..f21a433199aa 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgp102.c
@@ -21,22 +21,17 @@
  *
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
+#include "channv50.h"
 
 #include <subdev/timer.h>
 
 static int
-gp102_disp_dmac_init(struct nv50_disp_dmac *chan)
+gp102_disp_dmac_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
-	int ctrl = chan->base.chid.ctrl;
-	int user = chan->base.chid.user;
-
-	/* enable error reporting */
-	nvkm_mask(device, 0x6100a0, 0x00000001 << user, 0x00000001 << user);
+	int ctrl = chan->chid.ctrl;
+	int user = chan->chid.user;
 
 	/* initialise channel for dma command submission */
 	nvkm_wr32(device, 0x611494 + (ctrl * 0x0010), chan->push);
@@ -59,9 +54,11 @@ gp102_disp_dmac_init(struct nv50_disp_dmac *chan)
 	return 0;
 }
 
-const struct nv50_disp_dmac_func
+const struct nv50_disp_chan_func
 gp102_disp_dmac_func = {
 	.init = gp102_disp_dmac_init,
 	.fini = gf119_disp_dmac_fini,
+	.intr = gf119_disp_chan_intr,
+	.user = nv50_disp_chan_user,
 	.bind = gf119_disp_dmac_bind,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgv100.c
new file mode 100644
index 000000000000..eac0e42da354
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacgv100.c
@@ -0,0 +1,77 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "channv50.h"
+
+#include <core/ramht.h>
+#include <subdev/timer.h>
+
+static int
+gv100_disp_dmac_idle(struct nv50_disp_chan *chan)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 soff = (chan->chid.ctrl - 1) * 0x04;
+	nvkm_msec(device, 2000,
+		u32 stat = nvkm_rd32(device, 0x610664 + soff);
+		if ((stat & 0x000f0000) == 0x00040000)
+			return 0;
+	);
+	return -EBUSY;
+}
+
+int
+gv100_disp_dmac_bind(struct nv50_disp_chan *chan,
+		     struct nvkm_object *object, u32 handle)
+{
+	return nvkm_ramht_insert(chan->disp->ramht, object,
+				 chan->chid.user, -9, handle,
+				 chan->chid.user << 25 | 0x00000040);
+}
+
+void
+gv100_disp_dmac_fini(struct nv50_disp_chan *chan)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 coff = chan->chid.ctrl * 0x04;
+	nvkm_mask(device, 0x6104e0 + coff, 0x00000010, 0x00000000);
+	gv100_disp_dmac_idle(chan);
+	nvkm_mask(device, 0x6104e0 + coff, 0x00000002, 0x00000000);
+}
+
+int
+gv100_disp_dmac_init(struct nv50_disp_chan *chan)
+{
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	const u32 uoff = (chan->chid.ctrl - 1) * 0x1000;
+	const u32 poff = chan->chid.ctrl * 0x10;
+	const u32 coff = chan->chid.ctrl * 0x04;
+
+	nvkm_wr32(device, 0x610b24 + poff, lower_32_bits(chan->push));
+	nvkm_wr32(device, 0x610b20 + poff, upper_32_bits(chan->push));
+	nvkm_wr32(device, 0x610b28 + poff, 0x00000001);
+	nvkm_wr32(device, 0x610b2c + poff, 0x00000040);
+
+	nvkm_mask(device, 0x6104e0 + coff, 0x00000010, 0x00000010);
+	nvkm_wr32(device, 0x690000 + uoff, 0x00000000);
+	nvkm_wr32(device, 0x6104e0 + coff, 0x00000013);
+	return gv100_disp_dmac_idle(chan);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
index 070ec5e18fdb..9e8a9d7a9b68 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.c
@@ -21,176 +21,68 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
+#include "channv50.h"
 
 #include <core/client.h>
-#include <core/oproxy.h>
 #include <core/ramht.h>
 #include <subdev/fb.h>
+#include <subdev/mmu.h>
 #include <subdev/timer.h>
 #include <engine/dma.h>
 
-struct nv50_disp_dmac_object {
-	struct nvkm_oproxy oproxy;
-	struct nv50_disp_root *root;
-	int hash;
-};
-
-static void
-nv50_disp_dmac_child_del_(struct nvkm_oproxy *base)
-{
-	struct nv50_disp_dmac_object *object =
-		container_of(base, typeof(*object), oproxy);
-	nvkm_ramht_remove(object->root->ramht, object->hash);
-}
-
-static const struct nvkm_oproxy_func
-nv50_disp_dmac_child_func_ = {
-	.dtor[0] = nv50_disp_dmac_child_del_,
-};
-
-static int
-nv50_disp_dmac_child_new_(struct nv50_disp_chan *base,
-			  const struct nvkm_oclass *oclass,
-			  void *data, u32 size, struct nvkm_object **pobject)
-{
-	struct nv50_disp_dmac *chan = nv50_disp_dmac(base);
-	struct nv50_disp_root *root = chan->base.root;
-	struct nvkm_device *device = root->disp->base.engine.subdev.device;
-	const struct nvkm_device_oclass *sclass = oclass->priv;
-	struct nv50_disp_dmac_object *object;
-	int ret;
-
-	if (!(object = kzalloc(sizeof(*object), GFP_KERNEL)))
-		return -ENOMEM;
-	nvkm_oproxy_ctor(&nv50_disp_dmac_child_func_, oclass, &object->oproxy);
-	object->root = root;
-	*pobject = &object->oproxy.base;
-
-	ret = sclass->ctor(device, oclass, data, size, &object->oproxy.object);
-	if (ret)
-		return ret;
-
-	object->hash = chan->func->bind(chan, object->oproxy.object,
-					      oclass->handle);
-	if (object->hash < 0)
-		return object->hash;
-
-	return 0;
-}
-
-static int
-nv50_disp_dmac_child_get_(struct nv50_disp_chan *base, int index,
-			  struct nvkm_oclass *sclass)
-{
-	struct nv50_disp_dmac *chan = nv50_disp_dmac(base);
-	struct nv50_disp *disp = chan->base.root->disp;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
-	const struct nvkm_device_oclass *oclass = NULL;
-
-	sclass->engine = nvkm_device_engine(device, NVKM_ENGINE_DMAOBJ);
-	if (sclass->engine && sclass->engine->func->base.sclass) {
-		sclass->engine->func->base.sclass(sclass, index, &oclass);
-		if (oclass) {
-			sclass->priv = oclass;
-			return 0;
-		}
-	}
-
-	return -EINVAL;
-}
-
-static void
-nv50_disp_dmac_fini_(struct nv50_disp_chan *base)
-{
-	struct nv50_disp_dmac *chan = nv50_disp_dmac(base);
-	chan->func->fini(chan);
-}
-
-static int
-nv50_disp_dmac_init_(struct nv50_disp_chan *base)
-{
-	struct nv50_disp_dmac *chan = nv50_disp_dmac(base);
-	return chan->func->init(chan);
-}
-
-static void *
-nv50_disp_dmac_dtor_(struct nv50_disp_chan *base)
-{
-	return nv50_disp_dmac(base);
-}
-
-static const struct nv50_disp_chan_func
-nv50_disp_dmac_func_ = {
-	.dtor = nv50_disp_dmac_dtor_,
-	.init = nv50_disp_dmac_init_,
-	.fini = nv50_disp_dmac_fini_,
-	.child_get = nv50_disp_dmac_child_get_,
-	.child_new = nv50_disp_dmac_child_new_,
-};
-
 int
-nv50_disp_dmac_new_(const struct nv50_disp_dmac_func *func,
+nv50_disp_dmac_new_(const struct nv50_disp_chan_func *func,
 		    const struct nv50_disp_chan_mthd *mthd,
-		    struct nv50_disp_root *root, int chid, int head, u64 push,
+		    struct nv50_disp *disp, int chid, int head, u64 push,
 		    const struct nvkm_oclass *oclass,
 		    struct nvkm_object **pobject)
 {
 	struct nvkm_client *client = oclass->client;
-	struct nvkm_dmaobj *dmaobj;
-	struct nv50_disp_dmac *chan;
+	struct nv50_disp_chan *chan;
 	int ret;
 
-	if (!(chan = kzalloc(sizeof(*chan), GFP_KERNEL)))
-		return -ENOMEM;
-	*pobject = &chan->base.object;
-	chan->func = func;
-
-	ret = nv50_disp_chan_ctor(&nv50_disp_dmac_func_, mthd, root,
-				  chid, chid, head, oclass, &chan->base);
+	ret = nv50_disp_chan_new_(func, mthd, disp, chid, chid, head, oclass,
+				  pobject);
+	chan = nv50_disp_chan(*pobject);
 	if (ret)
 		return ret;
 
-	dmaobj = nvkm_dmaobj_search(client, push);
-	if (IS_ERR(dmaobj))
-		return PTR_ERR(dmaobj);
+	chan->memory = nvkm_umem_search(client, push);
+	if (IS_ERR(chan->memory))
+		return PTR_ERR(chan->memory);
 
-	if (dmaobj->limit - dmaobj->start != 0xfff)
+	if (nvkm_memory_size(chan->memory) < 0x1000)
 		return -EINVAL;
 
-	switch (dmaobj->target) {
-	case NV_MEM_TARGET_VRAM:
-		chan->push = 0x00000001 | dmaobj->start >> 8;
-		break;
-	case NV_MEM_TARGET_PCI_NOSNOOP:
-		chan->push = 0x00000003 | dmaobj->start >> 8;
-		break;
+	switch (nvkm_memory_target(chan->memory)) {
+	case NVKM_MEM_TARGET_VRAM: chan->push = 0x00000001; break;
+	case NVKM_MEM_TARGET_NCOH: chan->push = 0x00000002; break;
+	case NVKM_MEM_TARGET_HOST: chan->push = 0x00000003; break;
 	default:
 		return -EINVAL;
 	}
 
+	chan->push |= nvkm_memory_addr(chan->memory) >> 8;
 	return 0;
 }
 
 int
-nv50_disp_dmac_bind(struct nv50_disp_dmac *chan,
+nv50_disp_dmac_bind(struct nv50_disp_chan *chan,
 		    struct nvkm_object *object, u32 handle)
 {
-	return nvkm_ramht_insert(chan->base.root->ramht, object,
-				 chan->base.chid.user, -10, handle,
-				 chan->base.chid.user << 28 |
-				 chan->base.chid.user);
+	return nvkm_ramht_insert(chan->disp->ramht, object,
+				 chan->chid.user, -10, handle,
+				 chan->chid.user << 28 |
+				 chan->chid.user);
 }
 
 static void
-nv50_disp_dmac_fini(struct nv50_disp_dmac *chan)
+nv50_disp_dmac_fini(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
-	int ctrl = chan->base.chid.ctrl;
-	int user = chan->base.chid.user;
+	int ctrl = chan->chid.ctrl;
+	int user = chan->chid.user;
 
 	/* deactivate channel */
 	nvkm_mask(device, 0x610200 + (ctrl * 0x0010), 0x00001010, 0x00001000);
@@ -202,22 +94,15 @@ nv50_disp_dmac_fini(struct nv50_disp_dmac *chan)
 		nvkm_error(subdev, "ch %d fini timeout, %08x\n", user,
 			   nvkm_rd32(device, 0x610200 + (ctrl * 0x10)));
 	}
-
-	/* disable error reporting and completion notifications */
-	nvkm_mask(device, 0x610028, 0x00010001 << user, 0x00000000 << user);
 }
 
 static int
-nv50_disp_dmac_init(struct nv50_disp_dmac *chan)
+nv50_disp_dmac_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->base.root->disp;
-	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_subdev *subdev = &chan->disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
-	int ctrl = chan->base.chid.ctrl;
-	int user = chan->base.chid.user;
-
-	/* enable error reporting */
-	nvkm_mask(device, 0x610028, 0x00010000 << user, 0x00010000 << user);
+	int ctrl = chan->chid.ctrl;
+	int user = chan->chid.user;
 
 	/* initialise channel for dma command submission */
 	nvkm_wr32(device, 0x610204 + (ctrl * 0x0010), chan->push);
@@ -240,9 +125,11 @@ nv50_disp_dmac_init(struct nv50_disp_dmac *chan)
 	return 0;
 }
 
-const struct nv50_disp_dmac_func
+const struct nv50_disp_chan_func
 nv50_disp_dmac_func = {
 	.init = nv50_disp_dmac_init,
 	.fini = nv50_disp_dmac_fini,
+	.intr = nv50_disp_chan_intr,
+	.user = nv50_disp_chan_user,
 	.bind = nv50_disp_dmac_bind,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
deleted file mode 100644
index f9b98211da6a..000000000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/dmacnv50.h
+++ /dev/null
@@ -1,102 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#ifndef __NV50_DISP_DMAC_H__
-#define __NV50_DISP_DMAC_H__
-#define nv50_disp_dmac(p) container_of((p), struct nv50_disp_dmac, base)
-#include "channv50.h"
-
-struct nv50_disp_dmac {
-	const struct nv50_disp_dmac_func *func;
-	struct nv50_disp_chan base;
-	u32 push;
-};
-
-struct nv50_disp_dmac_func {
-	int  (*init)(struct nv50_disp_dmac *);
-	void (*fini)(struct nv50_disp_dmac *);
-	int  (*bind)(struct nv50_disp_dmac *, struct nvkm_object *, u32 handle);
-};
-
-int nv50_disp_dmac_new_(const struct nv50_disp_dmac_func *,
-			const struct nv50_disp_chan_mthd *,
-			struct nv50_disp_root *, int chid, int head, u64 push,
-			const struct nvkm_oclass *, struct nvkm_object **);
-
-extern const struct nv50_disp_dmac_func nv50_disp_dmac_func;
-int nv50_disp_dmac_bind(struct nv50_disp_dmac *, struct nvkm_object *, u32);
-extern const struct nv50_disp_dmac_func nv50_disp_core_func;
-
-extern const struct nv50_disp_dmac_func gf119_disp_dmac_func;
-void gf119_disp_dmac_fini(struct nv50_disp_dmac *);
-int gf119_disp_dmac_bind(struct nv50_disp_dmac *, struct nvkm_object *, u32);
-extern const struct nv50_disp_dmac_func gf119_disp_core_func;
-void gf119_disp_core_fini(struct nv50_disp_dmac *);
-
-extern const struct nv50_disp_dmac_func gp102_disp_dmac_func;
-
-struct nv50_disp_dmac_oclass {
-	int (*ctor)(const struct nv50_disp_dmac_func *,
-		    const struct nv50_disp_chan_mthd *,
-		    struct nv50_disp_root *, int chid,
-		    const struct nvkm_oclass *, void *data, u32 size,
-		    struct nvkm_object **);
-	struct nvkm_sclass base;
-	const struct nv50_disp_dmac_func *func;
-	const struct nv50_disp_chan_mthd *mthd;
-	int chid;
-};
-
-int nv50_disp_core_new(const struct nv50_disp_dmac_func *,
-		       const struct nv50_disp_chan_mthd *,
-		       struct nv50_disp_root *, int chid,
-		       const struct nvkm_oclass *oclass, void *data, u32 size,
-		       struct nvkm_object **);
-int nv50_disp_base_new(const struct nv50_disp_dmac_func *,
-		       const struct nv50_disp_chan_mthd *,
-		       struct nv50_disp_root *, int chid,
-		       const struct nvkm_oclass *oclass, void *data, u32 size,
-		       struct nvkm_object **);
-int nv50_disp_ovly_new(const struct nv50_disp_dmac_func *,
-		       const struct nv50_disp_chan_mthd *,
-		       struct nv50_disp_root *, int chid,
-		       const struct nvkm_oclass *oclass, void *data, u32 size,
-		       struct nvkm_object **);
-
-extern const struct nv50_disp_dmac_oclass nv50_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass nv50_disp_base_oclass;
-extern const struct nv50_disp_dmac_oclass nv50_disp_ovly_oclass;
-
-extern const struct nv50_disp_dmac_oclass g84_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass g84_disp_base_oclass;
-extern const struct nv50_disp_dmac_oclass g84_disp_ovly_oclass;
-
-extern const struct nv50_disp_dmac_oclass g94_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass gt200_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass gt200_disp_base_oclass;
-extern const struct nv50_disp_dmac_oclass gt200_disp_ovly_oclass;
-
-extern const struct nv50_disp_dmac_oclass gt215_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass gt215_disp_base_oclass;
-extern const struct nv50_disp_dmac_oclass gt215_disp_ovly_oclass;
-
-extern const struct nv50_disp_dmac_oclass gf119_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass gf119_disp_base_oclass;
-extern const struct nv50_disp_dmac_oclass gf119_disp_ovly_oclass;
-
-extern const struct nv50_disp_dmac_oclass gk104_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass gk104_disp_base_oclass;
-extern const struct nv50_disp_dmac_oclass gk104_disp_ovly_oclass;
-
-extern const struct nv50_disp_dmac_oclass gk110_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass gk110_disp_base_oclass;
-
-extern const struct nv50_disp_dmac_oclass gm107_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass gm200_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass gp100_disp_core_oclass;
-
-extern const struct nv50_disp_dmac_oclass gp102_disp_core_oclass;
-extern const struct nv50_disp_dmac_oclass gp102_disp_base_oclass;
-extern const struct nv50_disp_dmac_oclass gp102_disp_ovly_oclass;
-#endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
index 842e1b72ee42..731f188fc1ee 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g84.c
@@ -28,18 +28,20 @@
 
 static const struct nv50_disp_func
 g84_disp = {
+	.init = nv50_disp_init,
+	.fini = nv50_disp_fini,
 	.intr = nv50_disp_intr,
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
 	.root = &g84_disp_root_oclass,
-	.head.new = nv50_head_new,
-	.dac = { .nr = 3, .new = nv50_dac_new },
-	.sor = { .nr = 2, .new = g84_sor_new },
-	.pior = { .nr = 3, .new = nv50_pior_new },
+	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
+	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
+	.sor = { .cnt = nv50_sor_cnt, .new = g84_sor_new },
+	.pior = { .cnt = nv50_pior_cnt, .new = nv50_pior_new },
 };
 
 int
 g84_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return nv50_disp_new_(&g84_disp, device, index, 2, pdisp);
+	return nv50_disp_new_(&g84_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
index d184e6ab8918..def54fe1951e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/g94.c
@@ -28,18 +28,20 @@
 
 static const struct nv50_disp_func
 g94_disp = {
+	.init = nv50_disp_init,
+	.fini = nv50_disp_fini,
 	.intr = nv50_disp_intr,
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
 	.root = &g94_disp_root_oclass,
-	.head.new = nv50_head_new,
-	.dac = { .nr = 3, .new = nv50_dac_new },
-	.sor = { .nr = 4, .new = g94_sor_new },
-	.pior = { .nr = 3, .new = nv50_pior_new },
+	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
+	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
+	.sor = { .cnt = g94_sor_cnt, .new = g94_sor_new },
+	.pior = { .cnt = nv50_pior_cnt, .new = nv50_pior_new },
 };
 
 int
 g94_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return nv50_disp_new_(&g94_disp, device, index, 2, pdisp);
+	return nv50_disp_new_(&g94_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
index d8765b57180b..794e90982641 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gf119.c
@@ -24,8 +24,12 @@
 #include "nv50.h"
 #include "head.h"
 #include "ior.h"
+#include "channv50.h"
 #include "rootnv50.h"
 
+#include <core/ramht.h>
+#include <subdev/timer.h>
+
 void
 gf119_disp_super(struct work_struct *work)
 {
@@ -164,28 +168,99 @@ gf119_disp_intr(struct nv50_disp *disp)
 	}
 }
 
+void
+gf119_disp_fini(struct nv50_disp *disp)
+{
+	struct nvkm_device *device = disp->base.engine.subdev.device;
+	/* disable all interrupts */
+	nvkm_wr32(device, 0x6100b0, 0x00000000);
+}
+
 int
-gf119_disp_new_(const struct nv50_disp_func *func, struct nvkm_device *device,
-		int index, struct nvkm_disp **pdisp)
+gf119_disp_init(struct nv50_disp *disp)
 {
-	u32 heads = nvkm_rd32(device, 0x022448);
-	return nv50_disp_new_(func, device, index, heads, pdisp);
+	struct nvkm_device *device = disp->base.engine.subdev.device;
+	struct nvkm_head *head;
+	u32 tmp;
+	int i;
+
+	/* The below segments of code copying values from one register to
+	 * another appear to inform EVO of the display capabilities or
+	 * something similar.
+	 */
+
+	/* ... CRTC caps */
+	list_for_each_entry(head, &disp->base.head, head) {
+		const u32 hoff = head->id * 0x800;
+		tmp = nvkm_rd32(device, 0x616104 + hoff);
+		nvkm_wr32(device, 0x6101b4 + hoff, tmp);
+		tmp = nvkm_rd32(device, 0x616108 + hoff);
+		nvkm_wr32(device, 0x6101b8 + hoff, tmp);
+		tmp = nvkm_rd32(device, 0x61610c + hoff);
+		nvkm_wr32(device, 0x6101bc + hoff, tmp);
+	}
+
+	/* ... DAC caps */
+	for (i = 0; i < disp->dac.nr; i++) {
+		tmp = nvkm_rd32(device, 0x61a000 + (i * 0x800));
+		nvkm_wr32(device, 0x6101c0 + (i * 0x800), tmp);
+	}
+
+	/* ... SOR caps */
+	for (i = 0; i < disp->sor.nr; i++) {
+		tmp = nvkm_rd32(device, 0x61c000 + (i * 0x800));
+		nvkm_wr32(device, 0x6301c4 + (i * 0x800), tmp);
+	}
+
+	/* steal display away from vbios, or something like that */
+	if (nvkm_rd32(device, 0x6100ac) & 0x00000100) {
+		nvkm_wr32(device, 0x6100ac, 0x00000100);
+		nvkm_mask(device, 0x6194e8, 0x00000001, 0x00000000);
+		if (nvkm_msec(device, 2000,
+			if (!(nvkm_rd32(device, 0x6194e8) & 0x00000002))
+				break;
+		) < 0)
+			return -EBUSY;
+	}
+
+	/* point at display engine memory area (hash table, objects) */
+	nvkm_wr32(device, 0x610010, (disp->inst->addr >> 8) | 9);
+
+	/* enable supervisor interrupts, disable everything else */
+	nvkm_wr32(device, 0x610090, 0x00000000);
+	nvkm_wr32(device, 0x6100a0, 0x00000000);
+	nvkm_wr32(device, 0x6100b0, 0x00000307);
+
+	/* disable underflow reporting, preventing an intermittent issue
+	 * on some gk104 boards where the production vbios left this
+	 * setting enabled by default.
+	 *
+	 * ftp://download.nvidia.com/open-gpu-doc/gk104-disable-underflow-reporting/1/gk104-disable-underflow-reporting.txt
+	 */
+	list_for_each_entry(head, &disp->base.head, head) {
+		const u32 hoff = head->id * 0x800;
+		nvkm_mask(device, 0x616308 + hoff, 0x00000111, 0x00000010);
+	}
+
+	return 0;
 }
 
 static const struct nv50_disp_func
 gf119_disp = {
+	.init = gf119_disp_init,
+	.fini = gf119_disp_fini,
 	.intr = gf119_disp_intr,
 	.intr_error = gf119_disp_intr_error,
 	.uevent = &gf119_disp_chan_uevent,
 	.super = gf119_disp_super,
 	.root = &gf119_disp_root_oclass,
-	.head.new = gf119_head_new,
-	.dac = { .nr = 3, .new = gf119_dac_new },
-	.sor = { .nr = 4, .new = gf119_sor_new },
+	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
+	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
+	.sor = { .cnt = gf119_sor_cnt, .new = gf119_sor_new },
 };
 
 int
 gf119_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return gf119_disp_new_(&gf119_disp, device, index, pdisp);
+	return nv50_disp_new_(&gf119_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
index e8fe9f315d64..4c3439b1a62d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk104.c
@@ -28,18 +28,20 @@
 
 static const struct nv50_disp_func
 gk104_disp = {
+	.init = gf119_disp_init,
+	.fini = gf119_disp_fini,
 	.intr = gf119_disp_intr,
 	.intr_error = gf119_disp_intr_error,
 	.uevent = &gf119_disp_chan_uevent,
 	.super = gf119_disp_super,
 	.root = &gk104_disp_root_oclass,
-	.head.new = gf119_head_new,
-	.dac = { .nr = 3, .new = gf119_dac_new },
-	.sor = { .nr = 4, .new = gk104_sor_new },
+	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
+	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
+	.sor = { .cnt = gf119_sor_cnt, .new = gk104_sor_new },
 };
 
 int
 gk104_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return gf119_disp_new_(&gk104_disp, device, index, pdisp);
+	return nv50_disp_new_(&gk104_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
index 769687502e7a..bc6f4750c942 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gk110.c
@@ -28,18 +28,20 @@
 
 static const struct nv50_disp_func
 gk110_disp = {
+	.init = gf119_disp_init,
+	.fini = gf119_disp_fini,
 	.intr = gf119_disp_intr,
 	.intr_error = gf119_disp_intr_error,
 	.uevent = &gf119_disp_chan_uevent,
 	.super = gf119_disp_super,
 	.root = &gk110_disp_root_oclass,
-	.head.new = gf119_head_new,
-	.dac = { .nr = 3, .new = gf119_dac_new },
-	.sor = { .nr = 4, .new = gk104_sor_new },
+	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
+	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
+	.sor = { .cnt = gf119_sor_cnt, .new = gk104_sor_new },
 };
 
 int
 gk110_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return gf119_disp_new_(&gk110_disp, device, index, pdisp);
+	return nv50_disp_new_(&gk110_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
index ede70e5d188e..031cf6b03a76 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm107.c
@@ -28,18 +28,20 @@
 
 static const struct nv50_disp_func
 gm107_disp = {
+	.init = gf119_disp_init,
+	.fini = gf119_disp_fini,
 	.intr = gf119_disp_intr,
 	.intr_error = gf119_disp_intr_error,
 	.uevent = &gf119_disp_chan_uevent,
 	.super = gf119_disp_super,
 	.root = &gm107_disp_root_oclass,
-	.head.new = gf119_head_new,
-	.dac = { .nr = 3, .new = gf119_dac_new },
-	.sor = { .nr = 4, .new = gm107_sor_new },
+	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
+	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
+	.sor = { .cnt = gf119_sor_cnt, .new = gm107_sor_new },
 };
 
 int
 gm107_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return gf119_disp_new_(&gm107_disp, device, index, pdisp);
+	return nv50_disp_new_(&gm107_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
index 292d3b5f9704..ec9c33a5162d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gm200.c
@@ -28,18 +28,20 @@
 
 static const struct nv50_disp_func
 gm200_disp = {
+	.init = gf119_disp_init,
+	.fini = gf119_disp_fini,
 	.intr = gf119_disp_intr,
 	.intr_error = gf119_disp_intr_error,
 	.uevent = &gf119_disp_chan_uevent,
 	.super = gf119_disp_super,
 	.root = &gm200_disp_root_oclass,
-	.head.new = gf119_head_new,
-	.dac = { .nr = 3, .new = gf119_dac_new },
-	.sor = { .nr = 4, .new = gm200_sor_new },
+	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
+	.dac = { .cnt = gf119_dac_cnt, .new = gf119_dac_new },
+	.sor = { .cnt = gf119_sor_cnt, .new = gm200_sor_new },
 };
 
 int
 gm200_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return gf119_disp_new_(&gm200_disp, device, index, pdisp);
+	return nv50_disp_new_(&gm200_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c
index 39eb98b2c3a2..fd6216684f6d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp100.c
@@ -28,17 +28,19 @@
 
 static const struct nv50_disp_func
 gp100_disp = {
+	.init = gf119_disp_init,
+	.fini = gf119_disp_fini,
 	.intr = gf119_disp_intr,
 	.intr_error = gf119_disp_intr_error,
 	.uevent = &gf119_disp_chan_uevent,
 	.super = gf119_disp_super,
 	.root = &gp100_disp_root_oclass,
-	.head.new = gf119_head_new,
-	.sor = { .nr = 4, .new = gm200_sor_new },
+	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
+	.sor = { .cnt = gf119_sor_cnt, .new = gm200_sor_new },
 };
 
 int
 gp100_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return gf119_disp_new_(&gp100_disp, device, index, pdisp);
+	return nv50_disp_new_(&gp100_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
index 91d70fe18275..3468ddec1270 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gp102.c
@@ -24,6 +24,7 @@
 #include "nv50.h"
 #include "head.h"
 #include "ior.h"
+#include "channv50.h"
 #include "rootnv50.h"
 
 static void
@@ -54,17 +55,19 @@ gp102_disp_intr_error(struct nv50_disp *disp, int chid)
 
 static const struct nv50_disp_func
 gp102_disp = {
+	.init = gf119_disp_init,
+	.fini = gf119_disp_fini,
 	.intr = gf119_disp_intr,
 	.intr_error = gp102_disp_intr_error,
 	.uevent = &gf119_disp_chan_uevent,
 	.super = gf119_disp_super,
 	.root = &gp102_disp_root_oclass,
-	.head.new = gf119_head_new,
-	.sor = { .nr = 4, .new = gm200_sor_new },
+	.head = { .cnt = gf119_head_cnt, .new = gf119_head_new },
+	.sor = { .cnt = gf119_sor_cnt, .new = gm200_sor_new },
 };
 
 int
 gp102_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return gf119_disp_new_(&gp102_disp, device, index, pdisp);
+	return nv50_disp_new_(&gp102_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
index bf00c4e3be3a..f80183701f44 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt200.c
@@ -28,18 +28,20 @@
 
 static const struct nv50_disp_func
 gt200_disp = {
+	.init = nv50_disp_init,
+	.fini = nv50_disp_fini,
 	.intr = nv50_disp_intr,
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
 	.root = &gt200_disp_root_oclass,
-	.head.new = nv50_head_new,
-	.dac = { .nr = 3, .new = nv50_dac_new },
-	.sor = { .nr = 2, .new = g84_sor_new },
-	.pior = { .nr = 3, .new = nv50_pior_new },
+	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
+	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
+	.sor = { .cnt = nv50_sor_cnt, .new = g84_sor_new },
+	.pior = { .cnt = nv50_pior_cnt, .new = nv50_pior_new },
 };
 
 int
 gt200_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return nv50_disp_new_(&gt200_disp, device, index, 2, pdisp);
+	return nv50_disp_new_(&gt200_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
index 2cdd4d7a98d3..7581efc1357e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gt215.c
@@ -28,18 +28,20 @@
 
 static const struct nv50_disp_func
 gt215_disp = {
+	.init = nv50_disp_init,
+	.fini = nv50_disp_fini,
 	.intr = nv50_disp_intr,
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
 	.root = &gt215_disp_root_oclass,
-	.head.new = nv50_head_new,
-	.dac = { .nr = 3, .new = nv50_dac_new },
-	.sor = { .nr = 4, .new = gt215_sor_new },
-	.pior = { .nr = 3, .new = nv50_pior_new },
+	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
+	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
+	.sor = { .cnt = g94_sor_cnt, .new = gt215_sor_new },
+	.pior = { .cnt = nv50_pior_cnt, .new = nv50_pior_new },
 };
 
 int
 gt215_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return nv50_disp_new_(&gt215_disp, device, index, 2, pdisp);
+	return nv50_disp_new_(&gt215_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/gv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gv100.c
new file mode 100644
index 000000000000..d0a7e3456da1
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/gv100.c
@@ -0,0 +1,427 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "nv50.h"
+#include "head.h"
+#include "ior.h"
+#include "channv50.h"
+#include "rootnv50.h"
+
+#include <core/gpuobj.h>
+#include <subdev/timer.h>
+
+static int
+gv100_disp_wndw_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = nvkm_rd32(device, 0x610064);
+	return (nvkm_rd32(device, 0x610074) & 0x03f00000) >> 20;
+}
+
+static void
+gv100_disp_super(struct work_struct *work)
+{
+	struct nv50_disp *disp =
+		container_of(work, struct nv50_disp, supervisor);
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	struct nvkm_head *head;
+	u32 stat = nvkm_rd32(device, 0x6107a8);
+	u32 mask[4];
+
+	nvkm_debug(subdev, "supervisor %d: %08x\n", ffs(disp->super), stat);
+	list_for_each_entry(head, &disp->base.head, head) {
+		mask[head->id] = nvkm_rd32(device, 0x6107ac + (head->id * 4));
+		HEAD_DBG(head, "%08x", mask[head->id]);
+	}
+
+	if (disp->super & 0x00000001) {
+		nv50_disp_chan_mthd(disp->chan[0], NV_DBG_DEBUG);
+		nv50_disp_super_1(disp);
+		list_for_each_entry(head, &disp->base.head, head) {
+			if (!(mask[head->id] & 0x00001000))
+				continue;
+			nv50_disp_super_1_0(disp, head);
+		}
+	} else
+	if (disp->super & 0x00000002) {
+		list_for_each_entry(head, &disp->base.head, head) {
+			if (!(mask[head->id] & 0x00001000))
+				continue;
+			nv50_disp_super_2_0(disp, head);
+		}
+		nvkm_outp_route(&disp->base);
+		list_for_each_entry(head, &disp->base.head, head) {
+			if (!(mask[head->id] & 0x00010000))
+				continue;
+			nv50_disp_super_2_1(disp, head);
+		}
+		list_for_each_entry(head, &disp->base.head, head) {
+			if (!(mask[head->id] & 0x00001000))
+				continue;
+			nv50_disp_super_2_2(disp, head);
+		}
+	} else
+	if (disp->super & 0x00000004) {
+		list_for_each_entry(head, &disp->base.head, head) {
+			if (!(mask[head->id] & 0x00001000))
+				continue;
+			nv50_disp_super_3_0(disp, head);
+		}
+	}
+
+	list_for_each_entry(head, &disp->base.head, head)
+		nvkm_wr32(device, 0x6107ac + (head->id * 4), 0x00000000);
+	nvkm_wr32(device, 0x6107a8, 0x80000000);
+}
+
+static void
+gv100_disp_exception(struct nv50_disp *disp, int chid)
+{
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 stat = nvkm_rd32(device, 0x611020 + (chid * 12));
+	u32 type = (stat & 0x00007000) >> 12;
+	u32 mthd = (stat & 0x00000fff) << 2;
+	u32 data = nvkm_rd32(device, 0x611024 + (chid * 12));
+	u32 code = nvkm_rd32(device, 0x611028 + (chid * 12));
+
+	nvkm_error(subdev, "chid %d %08x [type %d mthd %04x] "
+			   "data %08x code %08x\n",
+		   chid, stat, type, mthd, data, code);
+
+	if (chid < ARRAY_SIZE(disp->chan) && disp->chan[chid]) {
+		switch (mthd) {
+		case 0x0200:
+			nv50_disp_chan_mthd(disp->chan[chid], NV_DBG_ERROR);
+			break;
+		default:
+			break;
+		}
+	}
+
+	nvkm_wr32(device, 0x611020 + (chid * 12), 0x90000000);
+}
+
+static void
+gv100_disp_intr_ctrl_disp(struct nv50_disp *disp)
+{
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 stat = nvkm_rd32(device, 0x611c30);
+
+	if (stat & 0x00000007) {
+		disp->super = (stat & 0x00000007);
+		queue_work(disp->wq, &disp->supervisor);
+		nvkm_wr32(device, 0x611860, disp->super);
+		stat &= ~0x00000007;
+	}
+
+	/*TODO: I would guess this is VBIOS_RELEASE, however, NFI how to
+	 *      ACK it, nor does RM appear to bother.
+	 */
+	if (stat & 0x00000008)
+		stat &= ~0x00000008;
+
+	if (stat & 0x00000100) {
+		unsigned long wndws = nvkm_rd32(device, 0x611858);
+		unsigned long other = nvkm_rd32(device, 0x61185c);
+		int wndw;
+
+		nvkm_wr32(device, 0x611858, wndws);
+		nvkm_wr32(device, 0x61185c, other);
+
+		/* AWAKEN_OTHER_CORE. */
+		if (other & 0x00000001)
+			nv50_disp_chan_uevent_send(disp, 0);
+
+		/* AWAKEN_WIN_CH(n). */
+		for_each_set_bit(wndw, &wndws, disp->wndw.nr) {
+			nv50_disp_chan_uevent_send(disp, 1 + wndw);
+		}
+	}
+
+	if (stat)
+		nvkm_warn(subdev, "ctrl %08x\n", stat);
+}
+
+static void
+gv100_disp_intr_exc_other(struct nv50_disp *disp)
+{
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 stat = nvkm_rd32(device, 0x611854);
+	unsigned long mask;
+	int head;
+
+	if (stat & 0x00000001) {
+		nvkm_wr32(device, 0x611854, 0x00000001);
+		gv100_disp_exception(disp, 0);
+		stat &= ~0x00000001;
+	}
+
+	if ((mask = (stat & 0x00ff0000) >> 16)) {
+		for_each_set_bit(head, &mask, disp->wndw.nr) {
+			nvkm_wr32(device, 0x611854, 0x00010000 << head);
+			gv100_disp_exception(disp, 73 + head);
+			stat &= ~(0x00010000 << head);
+		}
+	}
+
+	if (stat) {
+		nvkm_warn(subdev, "exception %08x\n", stat);
+		nvkm_wr32(device, 0x611854, stat);
+	}
+}
+
+static void
+gv100_disp_intr_exc_winim(struct nv50_disp *disp)
+{
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	unsigned long stat = nvkm_rd32(device, 0x611850);
+	int wndw;
+
+	for_each_set_bit(wndw, &stat, disp->wndw.nr) {
+		nvkm_wr32(device, 0x611850, BIT(wndw));
+		gv100_disp_exception(disp, 33 + wndw);
+		stat &= ~BIT(wndw);
+	}
+
+	if (stat) {
+		nvkm_warn(subdev, "wimm %08x\n", (u32)stat);
+		nvkm_wr32(device, 0x611850, stat);
+	}
+}
+
+static void
+gv100_disp_intr_exc_win(struct nv50_disp *disp)
+{
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	unsigned long stat = nvkm_rd32(device, 0x61184c);
+	int wndw;
+
+	for_each_set_bit(wndw, &stat, disp->wndw.nr) {
+		nvkm_wr32(device, 0x61184c, BIT(wndw));
+		gv100_disp_exception(disp, 1 + wndw);
+		stat &= ~BIT(wndw);
+	}
+
+	if (stat) {
+		nvkm_warn(subdev, "wndw %08x\n", (u32)stat);
+		nvkm_wr32(device, 0x61184c, stat);
+	}
+}
+
+static void
+gv100_disp_intr_head_timing(struct nv50_disp *disp, int head)
+{
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 stat = nvkm_rd32(device, 0x611800 + (head * 0x04));
+
+	/* LAST_DATA, LOADV. */
+	if (stat & 0x00000003) {
+		nvkm_wr32(device, 0x611800 + (head * 0x04), stat & 0x00000003);
+		stat &= ~0x00000003;
+	}
+
+	if (stat & 0x00000004) {
+		nvkm_disp_vblank(&disp->base, head);
+		nvkm_wr32(device, 0x611800 + (head * 0x04), 0x00000004);
+		stat &= ~0x00000004;
+	}
+
+	if (stat) {
+		nvkm_warn(subdev, "head %08x\n", stat);
+		nvkm_wr32(device, 0x611800 + (head * 0x04), stat);
+	}
+}
+
+static void
+gv100_disp_intr(struct nv50_disp *disp)
+{
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 stat = nvkm_rd32(device, 0x611ec0);
+	unsigned long mask;
+	int head;
+
+	if ((mask = (stat & 0x000000ff))) {
+		for_each_set_bit(head, &mask, 8) {
+			gv100_disp_intr_head_timing(disp, head);
+			stat &= ~BIT(head);
+		}
+	}
+
+	if (stat & 0x00000200) {
+		gv100_disp_intr_exc_win(disp);
+		stat &= ~0x00000200;
+	}
+
+	if (stat & 0x00000400) {
+		gv100_disp_intr_exc_winim(disp);
+		stat &= ~0x00000400;
+	}
+
+	if (stat & 0x00000800) {
+		gv100_disp_intr_exc_other(disp);
+		stat &= ~0x00000800;
+	}
+
+	if (stat & 0x00001000) {
+		gv100_disp_intr_ctrl_disp(disp);
+		stat &= ~0x00001000;
+	}
+
+	if (stat)
+		nvkm_warn(subdev, "intr %08x\n", stat);
+}
+
+static void
+gv100_disp_fini(struct nv50_disp *disp)
+{
+	struct nvkm_device *device = disp->base.engine.subdev.device;
+	nvkm_wr32(device, 0x611db0, 0x00000000);
+}
+
+static int
+gv100_disp_init(struct nv50_disp *disp)
+{
+	struct nvkm_device *device = disp->base.engine.subdev.device;
+	struct nvkm_head *head;
+	int i, j;
+	u32 tmp;
+
+	/* Claim ownership of display. */
+	if (nvkm_rd32(device, 0x6254e8) & 0x00000002) {
+		nvkm_mask(device, 0x6254e8, 0x00000001, 0x00000000);
+		if (nvkm_msec(device, 2000,
+			if (!(nvkm_rd32(device, 0x6254e8) & 0x00000002))
+				break;
+		) < 0)
+			return -EBUSY;
+	}
+
+	/* Lock pin capabilities. */
+	tmp = nvkm_rd32(device, 0x610068);
+	nvkm_wr32(device, 0x640008, tmp);
+
+	/* SOR capabilities. */
+	for (i = 0; i < disp->sor.nr; i++) {
+		tmp = nvkm_rd32(device, 0x61c000 + (i * 0x800));
+		nvkm_mask(device, 0x640000, 0x00000100 << i, 0x00000100 << i);
+		nvkm_wr32(device, 0x640144 + (i * 0x08), tmp);
+	}
+
+	/* Head capabilities. */
+	list_for_each_entry(head, &disp->base.head, head) {
+		const int id = head->id;
+
+		/* RG. */
+		tmp = nvkm_rd32(device, 0x616300 + (id * 0x800));
+		nvkm_wr32(device, 0x640048 + (id * 0x020), tmp);
+
+		/* POSTCOMP. */
+		for (j = 0; j < 6 * 4; j += 4) {
+			tmp = nvkm_rd32(device, 0x616100 + (id * 0x800) + j);
+			nvkm_wr32(device, 0x640030 + (id * 0x20) + j, tmp);
+		}
+	}
+
+	/* Window capabilities. */
+	for (i = 0; i < disp->wndw.nr; i++) {
+		nvkm_mask(device, 0x640004, 1 << i, 1 << i);
+		for (j = 0; j < 6 * 4; j += 4) {
+			tmp = nvkm_rd32(device, 0x630050 + (i * 0x800) + j);
+			nvkm_wr32(device, 0x6401e4 + (i * 0x20) + j, tmp);
+		}
+	}
+
+	/* IHUB capabilities. */
+	for (i = 0; i < 4; i++) {
+		tmp = nvkm_rd32(device, 0x62e000 + (i * 0x04));
+		nvkm_wr32(device, 0x640010 + (i * 0x04), tmp);
+	}
+
+	nvkm_mask(device, 0x610078, 0x00000001, 0x00000001);
+
+	/* Setup instance memory. */
+	switch (nvkm_memory_target(disp->inst->memory)) {
+	case NVKM_MEM_TARGET_VRAM: tmp = 0x00000001; break;
+	case NVKM_MEM_TARGET_NCOH: tmp = 0x00000002; break;
+	case NVKM_MEM_TARGET_HOST: tmp = 0x00000003; break;
+	default:
+		break;
+	}
+	nvkm_wr32(device, 0x610010, 0x00000008 | tmp);
+	nvkm_wr32(device, 0x610014, disp->inst->addr >> 16);
+
+	/* CTRL_DISP: AWAKEN, ERROR, SUPERVISOR[1-3]. */
+	nvkm_wr32(device, 0x611cf0, 0x00000187); /* MSK. */
+	nvkm_wr32(device, 0x611db0, 0x00000187); /* EN. */
+
+	/* EXC_OTHER: CURSn, CORE. */
+	nvkm_wr32(device, 0x611cec, disp->head.mask << 16 |
+				    0x00000001); /* MSK. */
+	nvkm_wr32(device, 0x611dac, 0x00000000); /* EN. */
+
+	/* EXC_WINIM. */
+	nvkm_wr32(device, 0x611ce8, disp->wndw.mask); /* MSK. */
+	nvkm_wr32(device, 0x611da8, 0x00000000); /* EN. */
+
+	/* EXC_WIN. */
+	nvkm_wr32(device, 0x611ce4, disp->wndw.mask); /* MSK. */
+	nvkm_wr32(device, 0x611da4, 0x00000000); /* EN. */
+
+	/* HEAD_TIMING(n): VBLANK. */
+	list_for_each_entry(head, &disp->base.head, head) {
+		const u32 hoff = head->id * 4;
+		nvkm_wr32(device, 0x611cc0 + hoff, 0x00000004); /* MSK. */
+		nvkm_wr32(device, 0x611d80 + hoff, 0x00000000); /* EN. */
+	}
+
+	/* OR. */
+	nvkm_wr32(device, 0x611cf4, 0x00000000); /* MSK. */
+	nvkm_wr32(device, 0x611db4, 0x00000000); /* EN. */
+	return 0;
+}
+
+static const struct nv50_disp_func
+gv100_disp = {
+	.init = gv100_disp_init,
+	.fini = gv100_disp_fini,
+	.intr = gv100_disp_intr,
+	.uevent = &gv100_disp_chan_uevent,
+	.super = gv100_disp_super,
+	.root = &gv100_disp_root_oclass,
+	.wndw = { .cnt = gv100_disp_wndw_cnt },
+	.head = { .cnt = gv100_head_cnt, .new = gv100_head_new },
+	.sor = { .cnt = gv100_sor_cnt, .new = gv100_sor_new },
+	.ramht_size = 0x2000,
+};
+
+int
+gv100_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
+{
+	return nv50_disp_new_(&gv100_disp, device, index, pdisp);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/hdmigv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/hdmigv100.c
new file mode 100644
index 000000000000..6e3c450eaace
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/hdmigv100.c
@@ -0,0 +1,85 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "hdmi.h"
+
+void
+gv100_hdmi_ctrl(struct nvkm_ior *ior, int head, bool enable, u8 max_ac_packet,
+		u8 rekey, u8 *avi, u8 avi_size, u8 *vendor, u8 vendor_size)
+{
+	struct nvkm_device *device = ior->disp->engine.subdev.device;
+	const u32 ctrl = 0x40000000 * enable |
+			 max_ac_packet << 16 |
+			 rekey;
+	const u32 hoff = head * 0x800;
+	const u32 hdmi = head * 0x400;
+	struct packed_hdmi_infoframe avi_infoframe;
+	struct packed_hdmi_infoframe vendor_infoframe;
+
+	pack_hdmi_infoframe(&avi_infoframe, avi, avi_size);
+	pack_hdmi_infoframe(&vendor_infoframe, vendor, vendor_size);
+
+	if (!(ctrl & 0x40000000)) {
+		nvkm_mask(device, 0x6165c0 + hoff, 0x40000000, 0x00000000);
+		nvkm_mask(device, 0x6f0100 + hdmi, 0x00000001, 0x00000000);
+		nvkm_mask(device, 0x6f00c0 + hdmi, 0x00000001, 0x00000000);
+		nvkm_mask(device, 0x6f0000 + hdmi, 0x00000001, 0x00000000);
+		return;
+	}
+
+	/* AVI InfoFrame (AVI). */
+	nvkm_mask(device, 0x6f0000 + hdmi, 0x00000001, 0x00000000);
+	if (avi_size) {
+		nvkm_wr32(device, 0x6f0008 + hdmi, avi_infoframe.header);
+		nvkm_wr32(device, 0x6f000c + hdmi, avi_infoframe.subpack0_low);
+		nvkm_wr32(device, 0x6f0010 + hdmi, avi_infoframe.subpack0_high);
+		nvkm_wr32(device, 0x6f0014 + hdmi, avi_infoframe.subpack1_low);
+		nvkm_wr32(device, 0x6f0018 + hdmi, avi_infoframe.subpack1_high);
+		nvkm_mask(device, 0x6f0000 + hdmi, 0x00000001, 0x00000001);
+	}
+
+	/* Vendor-specific InfoFrame (VSI). */
+	nvkm_mask(device, 0x6f0100 + hdmi, 0x00010001, 0x00000000);
+	if (vendor_size) {
+		nvkm_wr32(device, 0x6f0108 + hdmi, vendor_infoframe.header);
+		nvkm_wr32(device, 0x6f010c + hdmi, vendor_infoframe.subpack0_low);
+		nvkm_wr32(device, 0x6f0110 + hdmi, vendor_infoframe.subpack0_high);
+		nvkm_wr32(device, 0x6f0110 + hdmi, 0x00000000);
+		nvkm_wr32(device, 0x6f0114 + hdmi, 0x00000000);
+		nvkm_wr32(device, 0x6f0118 + hdmi, 0x00000000);
+		nvkm_wr32(device, 0x6f011c + hdmi, 0x00000000);
+		nvkm_wr32(device, 0x6f0120 + hdmi, 0x00000000);
+		nvkm_wr32(device, 0x6f0124 + hdmi, 0x00000000);
+		nvkm_mask(device, 0x6f0100 + hdmi, 0x00000001, 0x00000001);
+	}
+
+
+	/* General Control (GCP). */
+	nvkm_mask(device, 0x6f00c0 + hdmi, 0x00000001, 0x00000000);
+	nvkm_wr32(device, 0x6f00cc + hdmi, 0x00000010);
+	nvkm_mask(device, 0x6f00c0 + hdmi, 0x00000001, 0x00000001);
+
+	/* Audio Clock Regeneration (ACR). */
+	nvkm_wr32(device, 0x6f0080 + hdmi, 0x82000000);
+
+	/* NV_PDISP_SF_HDMI_CTRL. */
+	nvkm_mask(device, 0x6165c0 + hoff, 0x401f007f, ctrl);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/head.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/head.h
index 57030b3a4a75..7d55faf52fcb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/head.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/head.h
@@ -52,6 +52,14 @@ void nv50_head_rgpos(struct nvkm_head *, u16 *, u16 *);
 #define HEAD_DBG(h,f,a...) HEAD_MSG((h), debug, f, ##a)
 
 int nv04_head_new(struct nvkm_disp *, int id);
+
+int nv50_head_cnt(struct nvkm_disp *, unsigned long *);
 int nv50_head_new(struct nvkm_disp *, int id);
+
+int gf119_head_cnt(struct nvkm_disp *, unsigned long *);
 int gf119_head_new(struct nvkm_disp *, int id);
+void gf119_head_rgclk(struct nvkm_head *, int);
+
+int gv100_head_cnt(struct nvkm_disp *, unsigned long *);
+int gv100_head_new(struct nvkm_disp *, int id);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgf119.c
index 9fd7ae331308..e86298b35902 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgf119.c
@@ -39,7 +39,7 @@ gf119_head_vblank_get(struct nvkm_head *head)
 	nvkm_mask(device, 0x6100c0 + hoff, 0x00000001, 0x00000001);
 }
 
-static void
+void
 gf119_head_rgclk(struct nvkm_head *head, int div)
 {
 	struct nvkm_device *device = head->disp->engine.subdev.device;
@@ -92,8 +92,13 @@ gf119_head = {
 int
 gf119_head_new(struct nvkm_disp *disp, int id)
 {
-	struct nvkm_device *device = disp->engine.subdev.device;
-	if (!(nvkm_rd32(device, 0x612004) & (0x00000001 << id)))
-		return 0;
 	return nvkm_head_new_(&gf119_head, disp, id);
 }
+
+int
+gf119_head_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = nvkm_rd32(device, 0x612004) & 0x0000000f;
+	return nvkm_rd32(device, 0x022448);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgv100.c
new file mode 100644
index 000000000000..1a061b42ae5c
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/headgv100.c
@@ -0,0 +1,105 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "head.h"
+
+static void
+gv100_head_vblank_put(struct nvkm_head *head)
+{
+	struct nvkm_device *device = head->disp->engine.subdev.device;
+	nvkm_mask(device, 0x611d80 + (head->id * 4), 0x00000004, 0x00000000);
+}
+
+static void
+gv100_head_vblank_get(struct nvkm_head *head)
+{
+	struct nvkm_device *device = head->disp->engine.subdev.device;
+	nvkm_mask(device, 0x611d80 + (head->id * 4), 0x00000004, 0x00000004);
+}
+
+static void
+gv100_head_rgpos(struct nvkm_head *head, u16 *hline, u16 *vline)
+{
+	struct nvkm_device *device = head->disp->engine.subdev.device;
+	const u32 hoff = head->id * 0x800;
+	/* vline read locks hline. */
+	*vline = nvkm_rd32(device, 0x616330 + hoff) & 0x0000ffff;
+	*hline = nvkm_rd32(device, 0x616334 + hoff) & 0x0000ffff;
+}
+
+static void
+gv100_head_state(struct nvkm_head *head, struct nvkm_head_state *state)
+{
+	struct nvkm_device *device = head->disp->engine.subdev.device;
+	const u32 hoff = (state == &head->arm) * 0x8000 + head->id * 0x400;
+	u32 data;
+
+	data = nvkm_rd32(device, 0x682064 + hoff);
+	state->vtotal = (data & 0xffff0000) >> 16;
+	state->htotal = (data & 0x0000ffff);
+	data = nvkm_rd32(device, 0x682068 + hoff);
+	state->vsynce = (data & 0xffff0000) >> 16;
+	state->hsynce = (data & 0x0000ffff);
+	data = nvkm_rd32(device, 0x68206c + hoff);
+	state->vblanke = (data & 0xffff0000) >> 16;
+	state->hblanke = (data & 0x0000ffff);
+	data = nvkm_rd32(device, 0x682070 + hoff);
+	state->vblanks = (data & 0xffff0000) >> 16;
+	state->hblanks = (data & 0x0000ffff);
+	state->hz = nvkm_rd32(device, 0x68200c + hoff);
+
+	data = nvkm_rd32(device, 0x682004 + hoff);
+	switch ((data & 0x000000f0) >> 4) {
+	case 5: state->or.depth = 30; break;
+	case 4: state->or.depth = 24; break;
+	case 1: state->or.depth = 18; break;
+	default:
+		state->or.depth = 18;
+		WARN_ON(1);
+		break;
+	}
+}
+
+static const struct nvkm_head_func
+gv100_head = {
+	.state = gv100_head_state,
+	.rgpos = gv100_head_rgpos,
+	.rgclk = gf119_head_rgclk,
+	.vblank_get = gv100_head_vblank_get,
+	.vblank_put = gv100_head_vblank_put,
+};
+
+int
+gv100_head_new(struct nvkm_disp *disp, int id)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	if (!(nvkm_rd32(device, 0x610060) & (0x00000001 << id)))
+		return 0;
+	return nvkm_head_new_(&gv100_head, disp, id);
+}
+
+int
+gv100_head_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = nvkm_rd32(device, 0x610060) & 0x000000ff;
+	return nvkm_rd32(device, 0x610074) & 0x0000000f;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/headnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/headnv50.c
index c80d06d5168f..e7d5c397cd29 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/headnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/headnv50.c
@@ -90,3 +90,10 @@ nv50_head_new(struct nvkm_disp *disp, int id)
 {
 	return nvkm_head_new_(&nv50_head, disp, id);
 }
+
+int
+nv50_head_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	*pmask = 3;
+	return 2;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
index 4548c031b937..e0b4e0c5704e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ior.h
@@ -30,7 +30,7 @@ struct nvkm_ior {
 			UNKNOWN
 		} proto:3;
 		unsigned link:2;
-		unsigned head:4;
+		unsigned head:8;
 	} arm, asy;
 
 	/* Armed DP state. */
@@ -106,7 +106,6 @@ nv50_sor_link(struct nvkm_ior *ior)
 	return nv50_ior_base(ior) + ((ior->asy.link == 2) * 0x80);
 }
 
-int nv50_sor_new_(const struct nvkm_ior_func *, struct nvkm_disp *, int id);
 void nv50_sor_state(struct nvkm_ior *, struct nvkm_ior_state *);
 void nv50_sor_power(struct nvkm_ior *, bool, bool, bool, bool, bool);
 void nv50_sor_clock(struct nvkm_ior *);
@@ -122,7 +121,6 @@ void g94_sor_dp_watermark(struct nvkm_ior *, int, u8);
 
 void gt215_sor_dp_audio(struct nvkm_ior *, int, bool);
 
-int gf119_sor_new_(const struct nvkm_ior_func *, struct nvkm_disp *, int id);
 void gf119_sor_state(struct nvkm_ior *, struct nvkm_ior_state *);
 void gf119_sor_clock(struct nvkm_ior *);
 int gf119_sor_dp_links(struct nvkm_ior *, struct nvkm_i2c_aux *);
@@ -135,10 +133,15 @@ void gf119_sor_dp_watermark(struct nvkm_ior *, int, u8);
 
 void gm107_sor_dp_pattern(struct nvkm_ior *, int);
 
+void gm200_sor_route_set(struct nvkm_outp *, struct nvkm_ior *);
+int gm200_sor_route_get(struct nvkm_outp *, int *);
+void gm200_sor_dp_drive(struct nvkm_ior *, int, int, int, int, int);
+
 void g84_hdmi_ctrl(struct nvkm_ior *, int, bool, u8, u8, u8 *, u8 , u8 *, u8);
 void gt215_hdmi_ctrl(struct nvkm_ior *, int, bool, u8, u8, u8 *, u8 , u8 *, u8);
 void gf119_hdmi_ctrl(struct nvkm_ior *, int, bool, u8, u8, u8 *, u8 , u8 *, u8);
 void gk104_hdmi_ctrl(struct nvkm_ior *, int, bool, u8, u8, u8 *, u8 , u8 *, u8);
+void gv100_hdmi_ctrl(struct nvkm_ior *, int, bool, u8, u8, u8 *, u8 , u8 *, u8);
 
 void gt215_hda_hpd(struct nvkm_ior *, int, bool);
 void gt215_hda_eld(struct nvkm_ior *, u8 *, u8);
@@ -153,19 +156,34 @@ void gf119_hda_eld(struct nvkm_ior *, u8 *, u8);
 #define IOR_WARN(i,f,a...) IOR_MSG((i), warn, f, ##a)
 #define IOR_DBG(i,f,a...) IOR_MSG((i), debug, f, ##a)
 
+int nv50_dac_cnt(struct nvkm_disp *, unsigned long *);
 int nv50_dac_new(struct nvkm_disp *, int);
+
+int gf119_dac_cnt(struct nvkm_disp *, unsigned long *);
 int gf119_dac_new(struct nvkm_disp *, int);
 
+int nv50_pior_cnt(struct nvkm_disp *, unsigned long *);
 int nv50_pior_new(struct nvkm_disp *, int);
 
+int nv50_sor_cnt(struct nvkm_disp *, unsigned long *);
 int nv50_sor_new(struct nvkm_disp *, int);
+
 int g84_sor_new(struct nvkm_disp *, int);
+
+int g94_sor_cnt(struct nvkm_disp *, unsigned long *);
 int g94_sor_new(struct nvkm_disp *, int);
+
 int mcp77_sor_new(struct nvkm_disp *, int);
 int gt215_sor_new(struct nvkm_disp *, int);
 int mcp89_sor_new(struct nvkm_disp *, int);
+
+int gf119_sor_cnt(struct nvkm_disp *, unsigned long *);
 int gf119_sor_new(struct nvkm_disp *, int);
+
 int gk104_sor_new(struct nvkm_disp *, int);
 int gm107_sor_new(struct nvkm_disp *, int);
 int gm200_sor_new(struct nvkm_disp *, int);
+
+int gv100_sor_cnt(struct nvkm_disp *, unsigned long *);
+int gv100_sor_new(struct nvkm_disp *, int);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
index d7e0fbb12bf1..cfdce23ab83a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp77.c
@@ -26,18 +26,20 @@
 
 static const struct nv50_disp_func
 mcp77_disp = {
+	.init = nv50_disp_init,
+	.fini = nv50_disp_fini,
 	.intr = nv50_disp_intr,
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
 	.root = &g94_disp_root_oclass,
-	.head.new = nv50_head_new,
-	.dac = { .nr = 3, .new = nv50_dac_new },
-	.sor = { .nr = 4, .new = mcp77_sor_new },
-	.pior = { .nr = 3, .new = nv50_pior_new },
+	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
+	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
+	.sor = { .cnt = g94_sor_cnt, .new = mcp77_sor_new },
+	.pior = { .cnt = nv50_pior_cnt, .new = nv50_pior_new },
 };
 
 int
 mcp77_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return nv50_disp_new_(&mcp77_disp, device, index, 2, pdisp);
+	return nv50_disp_new_(&mcp77_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
index 7b75c57c12ed..85d9329cfa0e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/mcp89.c
@@ -26,18 +26,20 @@
 
 static const struct nv50_disp_func
 mcp89_disp = {
+	.init = nv50_disp_init,
+	.fini = nv50_disp_fini,
 	.intr = nv50_disp_intr,
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
 	.root = &gt215_disp_root_oclass,
-	.head.new = nv50_head_new,
-	.dac = { .nr = 3, .new = nv50_dac_new },
-	.sor = { .nr = 4, .new = mcp89_sor_new },
-	.pior = { .nr = 3, .new = nv50_pior_new },
+	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
+	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
+	.sor = { .cnt = g94_sor_cnt, .new = mcp89_sor_new },
+	.pior = { .cnt = nv50_pior_cnt, .new = nv50_pior_new },
 };
 
 int
 mcp89_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return nv50_disp_new_(&mcp89_disp, device, index, 2, pdisp);
+	return nv50_disp_new_(&mcp89_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
index 0c570dbd3021..f89c7b977aa5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.c
@@ -24,11 +24,12 @@
 #include "nv50.h"
 #include "head.h"
 #include "ior.h"
+#include "channv50.h"
 #include "rootnv50.h"
 
 #include <core/client.h>
 #include <core/enum.h>
-#include <core/gpuobj.h>
+#include <core/ramht.h>
 #include <subdev/bios.h>
 #include <subdev/bios/disp.h>
 #include <subdev/bios/init.h>
@@ -49,29 +50,115 @@ nv50_disp_intr_(struct nvkm_disp *base)
 	disp->func->intr(disp);
 }
 
+static void
+nv50_disp_fini_(struct nvkm_disp *base)
+{
+	struct nv50_disp *disp = nv50_disp(base);
+	disp->func->fini(disp);
+}
+
+static int
+nv50_disp_init_(struct nvkm_disp *base)
+{
+	struct nv50_disp *disp = nv50_disp(base);
+	return disp->func->init(disp);
+}
+
 static void *
 nv50_disp_dtor_(struct nvkm_disp *base)
 {
 	struct nv50_disp *disp = nv50_disp(base);
+
+	nvkm_ramht_del(&disp->ramht);
+	nvkm_gpuobj_del(&disp->inst);
+
 	nvkm_event_fini(&disp->uevent);
 	if (disp->wq)
 		destroy_workqueue(disp->wq);
+
 	return disp;
 }
 
+static int
+nv50_disp_oneinit_(struct nvkm_disp *base)
+{
+	struct nv50_disp *disp = nv50_disp(base);
+	const struct nv50_disp_func *func = disp->func;
+	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	int ret, i;
+
+	if (func->wndw.cnt) {
+		disp->wndw.nr = func->wndw.cnt(&disp->base, &disp->wndw.mask);
+		nvkm_debug(subdev, "Window(s): %d (%08lx)\n",
+			   disp->wndw.nr, disp->wndw.mask);
+	}
+
+	disp->head.nr = func->head.cnt(&disp->base, &disp->head.mask);
+	nvkm_debug(subdev, "  Head(s): %d (%02lx)\n",
+		   disp->head.nr, disp->head.mask);
+	for_each_set_bit(i, &disp->head.mask, disp->head.nr) {
+		ret = func->head.new(&disp->base, i);
+		if (ret)
+			return ret;
+	}
+
+	if (func->dac.cnt) {
+		disp->dac.nr = func->dac.cnt(&disp->base, &disp->dac.mask);
+		nvkm_debug(subdev, "   DAC(s): %d (%02lx)\n",
+			   disp->dac.nr, disp->dac.mask);
+		for_each_set_bit(i, &disp->dac.mask, disp->dac.nr) {
+			ret = func->dac.new(&disp->base, i);
+			if (ret)
+				return ret;
+		}
+	}
+
+	if (func->pior.cnt) {
+		disp->pior.nr = func->pior.cnt(&disp->base, &disp->pior.mask);
+		nvkm_debug(subdev, "  PIOR(s): %d (%02lx)\n",
+			   disp->pior.nr, disp->pior.mask);
+		for_each_set_bit(i, &disp->pior.mask, disp->pior.nr) {
+			ret = func->pior.new(&disp->base, i);
+			if (ret)
+				return ret;
+		}
+	}
+
+	disp->sor.nr = func->sor.cnt(&disp->base, &disp->sor.mask);
+	nvkm_debug(subdev, "   SOR(s): %d (%02lx)\n",
+		   disp->sor.nr, disp->sor.mask);
+	for_each_set_bit(i, &disp->sor.mask, disp->sor.nr) {
+		ret = func->sor.new(&disp->base, i);
+		if (ret)
+			return ret;
+	}
+
+	ret = nvkm_gpuobj_new(device, 0x10000, 0x10000, false, NULL,
+			      &disp->inst);
+	if (ret)
+		return ret;
+
+	return nvkm_ramht_new(device, func->ramht_size ? func->ramht_size :
+			      0x1000, 0, disp->inst, &disp->ramht);
+}
+
 static const struct nvkm_disp_func
 nv50_disp_ = {
 	.dtor = nv50_disp_dtor_,
+	.oneinit = nv50_disp_oneinit_,
+	.init = nv50_disp_init_,
+	.fini = nv50_disp_fini_,
 	.intr = nv50_disp_intr_,
 	.root = nv50_disp_root_,
 };
 
 int
 nv50_disp_new_(const struct nv50_disp_func *func, struct nvkm_device *device,
-	       int index, int heads, struct nvkm_disp **pdisp)
+	       int index, struct nvkm_disp **pdisp)
 {
 	struct nv50_disp *disp;
-	int ret, i;
+	int ret;
 
 	if (!(disp = kzalloc(sizeof(*disp), GFP_KERNEL)))
 		return -ENOMEM;
@@ -85,33 +172,11 @@ nv50_disp_new_(const struct nv50_disp_func *func, struct nvkm_device *device,
 	disp->wq = create_singlethread_workqueue("nvkm-disp");
 	if (!disp->wq)
 		return -ENOMEM;
-	INIT_WORK(&disp->supervisor, func->super);
-
-	for (i = 0; func->head.new && i < heads; i++) {
-		ret = func->head.new(&disp->base, i);
-		if (ret)
-			return ret;
-	}
-
-	for (i = 0; func->dac.new && i < func->dac.nr; i++) {
-		ret = func->dac.new(&disp->base, i);
-		if (ret)
-			return ret;
-	}
-
-	for (i = 0; func->pior.new && i < func->pior.nr; i++) {
-		ret = func->pior.new(&disp->base, i);
-		if (ret)
-			return ret;
-	}
 
-	for (i = 0; func->sor.new && i < func->sor.nr; i++) {
-		ret = func->sor.new(&disp->base, i);
-		if (ret)
-			return ret;
-	}
+	INIT_WORK(&disp->supervisor, func->super);
 
-	return nvkm_event_init(func->uevent, 1, 1 + (heads * 4), &disp->uevent);
+	return nvkm_event_init(func->uevent, 1, ARRAY_SIZE(disp->chan),
+			       &disp->uevent);
 }
 
 static u32
@@ -613,20 +678,96 @@ nv50_disp_intr(struct nv50_disp *disp)
 	}
 }
 
+void
+nv50_disp_fini(struct nv50_disp *disp)
+{
+	struct nvkm_device *device = disp->base.engine.subdev.device;
+	/* disable all interrupts */
+	nvkm_wr32(device, 0x610024, 0x00000000);
+	nvkm_wr32(device, 0x610020, 0x00000000);
+}
+
+int
+nv50_disp_init(struct nv50_disp *disp)
+{
+	struct nvkm_device *device = disp->base.engine.subdev.device;
+	struct nvkm_head *head;
+	u32 tmp;
+	int i;
+
+	/* The below segments of code copying values from one register to
+	 * another appear to inform EVO of the display capabilities or
+	 * something similar.  NFI what the 0x614004 caps are for..
+	 */
+	tmp = nvkm_rd32(device, 0x614004);
+	nvkm_wr32(device, 0x610184, tmp);
+
+	/* ... CRTC caps */
+	list_for_each_entry(head, &disp->base.head, head) {
+		tmp = nvkm_rd32(device, 0x616100 + (head->id * 0x800));
+		nvkm_wr32(device, 0x610190 + (head->id * 0x10), tmp);
+		tmp = nvkm_rd32(device, 0x616104 + (head->id * 0x800));
+		nvkm_wr32(device, 0x610194 + (head->id * 0x10), tmp);
+		tmp = nvkm_rd32(device, 0x616108 + (head->id * 0x800));
+		nvkm_wr32(device, 0x610198 + (head->id * 0x10), tmp);
+		tmp = nvkm_rd32(device, 0x61610c + (head->id * 0x800));
+		nvkm_wr32(device, 0x61019c + (head->id * 0x10), tmp);
+	}
+
+	/* ... DAC caps */
+	for (i = 0; i < disp->dac.nr; i++) {
+		tmp = nvkm_rd32(device, 0x61a000 + (i * 0x800));
+		nvkm_wr32(device, 0x6101d0 + (i * 0x04), tmp);
+	}
+
+	/* ... SOR caps */
+	for (i = 0; i < disp->sor.nr; i++) {
+		tmp = nvkm_rd32(device, 0x61c000 + (i * 0x800));
+		nvkm_wr32(device, 0x6101e0 + (i * 0x04), tmp);
+	}
+
+	/* ... PIOR caps */
+	for (i = 0; i < disp->pior.nr; i++) {
+		tmp = nvkm_rd32(device, 0x61e000 + (i * 0x800));
+		nvkm_wr32(device, 0x6101f0 + (i * 0x04), tmp);
+	}
+
+	/* steal display away from vbios, or something like that */
+	if (nvkm_rd32(device, 0x610024) & 0x00000100) {
+		nvkm_wr32(device, 0x610024, 0x00000100);
+		nvkm_mask(device, 0x6194e8, 0x00000001, 0x00000000);
+		if (nvkm_msec(device, 2000,
+			if (!(nvkm_rd32(device, 0x6194e8) & 0x00000002))
+				break;
+		) < 0)
+			return -EBUSY;
+	}
+
+	/* point at display engine memory area (hash table, objects) */
+	nvkm_wr32(device, 0x610010, (disp->inst->addr >> 8) | 9);
+
+	/* enable supervisor interrupts, disable everything else */
+	nvkm_wr32(device, 0x61002c, 0x00000370);
+	nvkm_wr32(device, 0x610028, 0x00000000);
+	return 0;
+}
+
 static const struct nv50_disp_func
 nv50_disp = {
+	.init = nv50_disp_init,
+	.fini = nv50_disp_fini,
 	.intr = nv50_disp_intr,
 	.uevent = &nv50_disp_chan_uevent,
 	.super = nv50_disp_super,
 	.root = &nv50_disp_root_oclass,
-	.head.new = nv50_head_new,
-	.dac = { .nr = 3, .new = nv50_dac_new },
-	.sor = { .nr = 2, .new = nv50_sor_new },
-	.pior = { .nr = 3, .new = nv50_pior_new },
+	.head = { .cnt = nv50_head_cnt, .new = nv50_head_new },
+	.dac = { .cnt = nv50_dac_cnt, .new = nv50_dac_new },
+	.sor = { .cnt = nv50_sor_cnt, .new = nv50_sor_new },
+	.pior = { .cnt = nv50_pior_cnt, .new = nv50_pior_new },
 };
 
 int
 nv50_disp_new(struct nvkm_device *device, int index, struct nvkm_disp **pdisp)
 {
-	return nv50_disp_new_(&nv50_disp, device, index, 2, pdisp);
+	return nv50_disp_new_(&nv50_disp, device, index, pdisp);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
index eb0b8acb1c5b..8580382ab248 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/nv50.h
@@ -16,14 +16,26 @@ struct nv50_disp {
 	struct nvkm_event uevent;
 
 	struct {
+		unsigned long mask;
+		int nr;
+	} wndw, head, dac;
+
+	struct {
+		unsigned long mask;
+		int nr;
 		u32 lvdsconf;
 	} sor;
 
 	struct {
+		unsigned long mask;
+		int nr;
 		u8 type[3];
 	} pior;
 
-	struct nv50_disp_chan *chan[21];
+	struct nvkm_gpuobj *inst;
+	struct nvkm_ramht *ramht;
+
+	struct nv50_disp_chan *chan[81];
 };
 
 void nv50_disp_super_1(struct nv50_disp *);
@@ -34,11 +46,11 @@ void nv50_disp_super_2_2(struct nv50_disp *, struct nvkm_head *);
 void nv50_disp_super_3_0(struct nv50_disp *, struct nvkm_head *);
 
 int nv50_disp_new_(const struct nv50_disp_func *, struct nvkm_device *,
-		   int index, int heads, struct nvkm_disp **);
-int gf119_disp_new_(const struct nv50_disp_func *, struct nvkm_device *,
-		    int index, struct nvkm_disp **);
+		   int index, struct nvkm_disp **);
 
 struct nv50_disp_func {
+	int (*init)(struct nv50_disp *);
+	void (*fini)(struct nv50_disp *);
 	void (*intr)(struct nv50_disp *);
 	void (*intr_error)(struct nv50_disp *, int chid);
 
@@ -48,28 +60,20 @@ struct nv50_disp_func {
 	const struct nvkm_disp_oclass *root;
 
 	struct {
+		int (*cnt)(struct nvkm_disp *, unsigned long *mask);
 		int (*new)(struct nvkm_disp *, int id);
-	} head;
+	} wndw, head, dac, sor, pior;
 
-	struct {
-		int nr;
-		int (*new)(struct nvkm_disp *, int id);
-	} dac;
-
-	struct {
-		int nr;
-		int (*new)(struct nvkm_disp *, int id);
-	} sor;
-
-	struct {
-		int nr;
-		int (*new)(struct nvkm_disp *, int id);
-	} pior;
+	u16 ramht_size;
 };
 
+int nv50_disp_init(struct nv50_disp *);
+void nv50_disp_fini(struct nv50_disp *);
 void nv50_disp_intr(struct nv50_disp *);
 void nv50_disp_super(struct work_struct *);
 
+int gf119_disp_init(struct nv50_disp *);
+void gf119_disp_fini(struct nv50_disp *);
 void gf119_disp_intr(struct nv50_disp *);
 void gf119_disp_super(struct work_struct *);
 void gf119_disp_intr_error(struct nv50_disp *, int);
@@ -77,4 +81,12 @@ void gf119_disp_intr_error(struct nv50_disp *, int);
 void nv50_disp_dptmds_war_2(struct nv50_disp *, struct dcb_output *);
 void nv50_disp_dptmds_war_3(struct nv50_disp *, struct dcb_output *);
 void nv50_disp_update_sppll1(struct nv50_disp *);
+
+extern const struct nvkm_event_func nv50_disp_chan_uevent;
+int  nv50_disp_chan_uevent_ctor(struct nvkm_object *, void *, u32,
+				struct nvkm_notify *);
+void nv50_disp_chan_uevent_send(struct nv50_disp *, int);
+
+extern const struct nvkm_event_func gf119_disp_chan_uevent;
+extern const struct nvkm_event_func gv100_disp_chan_uevent;
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgf119.c
index 1f9fd3403f07..1ae0bcfc89b9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgf119.c
@@ -22,16 +22,11 @@
  * Authors: Ben Skeggs
  */
 #include "channv50.h"
-#include "rootnv50.h"
 
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-gf119_disp_oimm_oclass = {
-	.base.oclass = GF110_DISP_OVERLAY,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_oimm_new,
-	.func = &gf119_disp_pioc_func,
-	.chid = { 9, 9 },
-};
+int
+gf119_disp_oimm_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_oimm_new_(&gf119_disp_pioc_func, disp, 9, 9,
+				   oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgp102.c
index abf82365c671..30ffb1008505 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgp102.c
@@ -22,16 +22,11 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 #include "channv50.h"
-#include "rootnv50.h"
 
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-gp102_disp_oimm_oclass = {
-	.base.oclass = GK104_DISP_OVERLAY,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_oimm_new,
-	.func = &gf119_disp_pioc_func,
-	.chid = { 9, 13 },
-};
+int
+gp102_disp_oimm_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_oimm_new_(&gf119_disp_pioc_func, disp, 9, 13,
+				   oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgt215.c
deleted file mode 100644
index 1281db28aebd..000000000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmgt215.c
+++ /dev/null
@@ -1,37 +0,0 @@
-/*
- * Copyright 2012 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs
- */
-#include "channv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_pioc_oclass
-gt215_disp_oimm_oclass = {
-	.base.oclass = GT214_DISP_OVERLAY,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_oimm_new,
-	.func = &nv50_disp_pioc_func,
-	.chid = { 5, 5 },
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmnv50.c
index f3b0fa2c5924..0db99bfe9db9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/oimmnv50.c
@@ -23,30 +23,26 @@
  */
 #include "channv50.h"
 #include "head.h"
-#include "rootnv50.h"
 
 #include <core/client.h>
 
-#include <nvif/class.h>
 #include <nvif/cl507b.h>
 #include <nvif/unpack.h>
 
 int
-nv50_disp_oimm_new(const struct nv50_disp_chan_func *func,
-		   const struct nv50_disp_chan_mthd *mthd,
-		   struct nv50_disp_root *root, int ctrl, int user,
-		   const struct nvkm_oclass *oclass, void *data, u32 size,
-		   struct nvkm_object **pobject)
+nv50_disp_oimm_new_(const struct nv50_disp_chan_func *func,
+		    struct nv50_disp *disp, int ctrl, int user,
+		    const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nvkm_object **pobject)
 {
 	union {
 		struct nv50_disp_overlay_v0 v0;
-	} *args = data;
+	} *args = argv;
 	struct nvkm_object *parent = oclass->parent;
-	struct nv50_disp *disp = root->disp;
 	int head, ret = -ENOSYS;
 
-	nvif_ioctl(parent, "create disp overlay size %d\n", size);
-	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+	nvif_ioctl(parent, "create disp overlay size %d\n", argc);
+	if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))) {
 		nvif_ioctl(parent, "create disp overlay vers %d head %d\n",
 			   args->v0.version, args->v0.head);
 		if (!nvkm_head_find(&disp->base, args->v0.head))
@@ -55,16 +51,14 @@ nv50_disp_oimm_new(const struct nv50_disp_chan_func *func,
 	} else
 		return ret;
 
-	return nv50_disp_chan_new_(func, mthd, root, ctrl + head, user + head,
+	return nv50_disp_chan_new_(func, NULL, disp, ctrl + head, user + head,
 				   head, oclass, pobject);
 }
 
-const struct nv50_disp_pioc_oclass
-nv50_disp_oimm_oclass = {
-	.base.oclass = NV50_DISP_OVERLAY,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_oimm_new,
-	.func = &nv50_disp_pioc_func,
-	.chid = { 5, 5 },
-};
+int
+nv50_disp_oimm_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		   struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_oimm_new_(&nv50_disp_pioc_func, disp, 5, 5,
+				   oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlyg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlyg84.c
index db6234eebc61..31b915d48699 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlyg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlyg84.c
@@ -21,10 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 g84_disp_ovly_mthd_base = {
@@ -54,8 +51,8 @@ g84_disp_ovly_mthd_base = {
 	}
 };
 
-const struct nv50_disp_chan_mthd
-g84_disp_ovly_chan_mthd = {
+static const struct nv50_disp_chan_mthd
+g84_disp_ovly_mthd = {
 	.name = "Overlay",
 	.addr = 0x000540,
 	.prev = 0x000004,
@@ -65,13 +62,10 @@ g84_disp_ovly_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-g84_disp_ovly_oclass = {
-	.base.oclass = G82_DISP_OVERLAY_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_ovly_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &g84_disp_ovly_chan_mthd,
-	.chid = 3,
-};
+int
+g84_disp_ovly_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		  struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_ovly_new_(&nv50_disp_dmac_func, &g84_disp_ovly_mthd,
+				   disp, 3, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygf119.c
index 5985879abd23..83fd534c44da 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygf119.c
@@ -21,10 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 gf119_disp_ovly_mthd_base = {
@@ -79,7 +76,7 @@ gf119_disp_ovly_mthd_base = {
 };
 
 static const struct nv50_disp_chan_mthd
-gf119_disp_ovly_chan_mthd = {
+gf119_disp_ovly_mthd = {
 	.name = "Overlay",
 	.addr = 0x001000,
 	.prev = -0x020000,
@@ -89,13 +86,10 @@ gf119_disp_ovly_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-gf119_disp_ovly_oclass = {
-	.base.oclass = GF110_DISP_OVERLAY_CONTROL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_ovly_new,
-	.func = &gf119_disp_dmac_func,
-	.mthd = &gf119_disp_ovly_chan_mthd,
-	.chid = 5,
-};
+int
+gf119_disp_ovly_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_ovly_new_(&gf119_disp_dmac_func, &gf119_disp_ovly_mthd,
+				   disp, 5, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c
index 2f0220b39f34..a7acacbc92c1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygk104.c
@@ -21,10 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 gk104_disp_ovly_mthd_base = {
@@ -81,7 +78,7 @@ gk104_disp_ovly_mthd_base = {
 };
 
 const struct nv50_disp_chan_mthd
-gk104_disp_ovly_chan_mthd = {
+gk104_disp_ovly_mthd = {
 	.name = "Overlay",
 	.addr = 0x001000,
 	.prev = -0x020000,
@@ -91,13 +88,10 @@ gk104_disp_ovly_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-gk104_disp_ovly_oclass = {
-	.base.oclass = GK104_DISP_OVERLAY_CONTROL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_ovly_new,
-	.func = &gf119_disp_dmac_func,
-	.mthd = &gk104_disp_ovly_chan_mthd,
-	.chid = 5,
-};
+int
+gk104_disp_ovly_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_ovly_new_(&gf119_disp_dmac_func, &gk104_disp_ovly_mthd,
+				   disp, 5, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp102.c
index 589bd2f12b41..e0eca6ea914c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygp102.c
@@ -21,18 +21,12 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
+#include "channv50.h"
 
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gp102_disp_ovly_oclass = {
-	.base.oclass = GK104_DISP_OVERLAY_CONTROL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_ovly_new,
-	.func = &gp102_disp_dmac_func,
-	.mthd = &gk104_disp_ovly_chan_mthd,
-	.chid = 5,
-};
+int
+gp102_disp_ovly_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_ovly_new_(&gp102_disp_dmac_func, &gk104_disp_ovly_mthd,
+				   disp, 5, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt200.c
index f858053db83d..dc60cd00dc16 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt200.c
@@ -21,10 +21,7 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
+#include "channv50.h"
 
 static const struct nv50_disp_mthd_list
 gt200_disp_ovly_mthd_base = {
@@ -58,7 +55,7 @@ gt200_disp_ovly_mthd_base = {
 };
 
 static const struct nv50_disp_chan_mthd
-gt200_disp_ovly_chan_mthd = {
+gt200_disp_ovly_mthd = {
 	.name = "Overlay",
 	.addr = 0x000540,
 	.prev = 0x000004,
@@ -68,13 +65,10 @@ gt200_disp_ovly_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-gt200_disp_ovly_oclass = {
-	.base.oclass = GT200_DISP_OVERLAY_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_ovly_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &gt200_disp_ovly_chan_mthd,
-	.chid = 3,
-};
+int
+gt200_disp_ovly_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_ovly_new_(&nv50_disp_dmac_func, &gt200_disp_ovly_mthd,
+				   disp, 3, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt215.c
deleted file mode 100644
index c947e1e16a37..000000000000
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlygt215.c
+++ /dev/null
@@ -1,38 +0,0 @@
-/*
- * Copyright 2015 Red Hat Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: Ben Skeggs <bskeggs@redhat.com>
- */
-#include "dmacnv50.h"
-#include "rootnv50.h"
-
-#include <nvif/class.h>
-
-const struct nv50_disp_dmac_oclass
-gt215_disp_ovly_oclass = {
-	.base.oclass = GT214_DISP_OVERLAY_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_ovly_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &g84_disp_ovly_chan_mthd,
-	.chid = 3,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c
index 9ebaaa6e9e33..6974c12c4518 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/ovlynv50.c
@@ -21,33 +21,30 @@
  *
  * Authors: Ben Skeggs
  */
-#include "dmacnv50.h"
+#include "channv50.h"
 #include "head.h"
-#include "rootnv50.h"
 
 #include <core/client.h>
 
-#include <nvif/class.h>
 #include <nvif/cl507e.h>
 #include <nvif/unpack.h>
 
 int
-nv50_disp_ovly_new(const struct nv50_disp_dmac_func *func,
-		   const struct nv50_disp_chan_mthd *mthd,
-		   struct nv50_disp_root *root, int chid,
-		   const struct nvkm_oclass *oclass, void *data, u32 size,
-		   struct nvkm_object **pobject)
+nv50_disp_ovly_new_(const struct nv50_disp_chan_func *func,
+		    const struct nv50_disp_chan_mthd *mthd,
+		    struct nv50_disp *disp, int chid,
+		    const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nvkm_object **pobject)
 {
 	union {
 		struct nv50_disp_overlay_channel_dma_v0 v0;
-	} *args = data;
+	} *args = argv;
 	struct nvkm_object *parent = oclass->parent;
-	struct nv50_disp *disp = root->disp;
 	int head, ret = -ENOSYS;
 	u64 push;
 
-	nvif_ioctl(parent, "create disp overlay channel dma size %d\n", size);
-	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+	nvif_ioctl(parent, "create disp overlay channel dma size %d\n", argc);
+	if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))) {
 		nvif_ioctl(parent, "create disp overlay channel dma vers %d "
 				   "pushbuf %016llx head %d\n",
 			   args->v0.version, args->v0.pushbuf, args->v0.head);
@@ -58,7 +55,7 @@ nv50_disp_ovly_new(const struct nv50_disp_dmac_func *func,
 	} else
 		return ret;
 
-	return nv50_disp_dmac_new_(func, mthd, root, chid + head,
+	return nv50_disp_dmac_new_(func, mthd, disp, chid + head,
 				   head, push, oclass, pobject);
 }
 
@@ -91,7 +88,7 @@ nv50_disp_ovly_mthd_base = {
 };
 
 static const struct nv50_disp_chan_mthd
-nv50_disp_ovly_chan_mthd = {
+nv50_disp_ovly_mthd = {
 	.name = "Overlay",
 	.addr = 0x000540,
 	.prev = 0x000004,
@@ -101,13 +98,10 @@ nv50_disp_ovly_chan_mthd = {
 	}
 };
 
-const struct nv50_disp_dmac_oclass
-nv50_disp_ovly_oclass = {
-	.base.oclass = NV50_DISP_OVERLAY_CHANNEL_DMA,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = nv50_disp_ovly_new,
-	.func = &nv50_disp_dmac_func,
-	.mthd = &nv50_disp_ovly_chan_mthd,
-	.chid = 3,
-};
+int
+nv50_disp_ovly_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		   struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return nv50_disp_ovly_new_(&nv50_disp_dmac_func, &nv50_disp_ovly_mthd,
+				   disp, 3, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c
index 0abaa6431943..5296e7bee813 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocgf119.c
@@ -29,7 +29,7 @@
 static void
 gf119_disp_pioc_fini(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	int ctrl = chan->chid.ctrl;
@@ -43,24 +43,17 @@ gf119_disp_pioc_fini(struct nv50_disp_chan *chan)
 		nvkm_error(subdev, "ch %d fini: %08x\n", user,
 			   nvkm_rd32(device, 0x610490 + (ctrl * 0x10)));
 	}
-
-	/* disable error reporting and completion notification */
-	nvkm_mask(device, 0x610090, 0x00000001 << user, 0x00000000);
-	nvkm_mask(device, 0x6100a0, 0x00000001 << user, 0x00000000);
 }
 
 static int
 gf119_disp_pioc_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	int ctrl = chan->chid.ctrl;
 	int user = chan->chid.user;
 
-	/* enable error reporting */
-	nvkm_mask(device, 0x6100a0, 0x00000001 << user, 0x00000001 << user);
-
 	/* activate channel */
 	nvkm_wr32(device, 0x610490 + (ctrl * 0x10), 0x00000001);
 	if (nvkm_msec(device, 2000,
@@ -80,4 +73,6 @@ const struct nv50_disp_chan_func
 gf119_disp_pioc_func = {
 	.init = gf119_disp_pioc_init,
 	.fini = gf119_disp_pioc_fini,
+	.intr = gf119_disp_chan_intr,
+	.user = nv50_disp_chan_user,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c
index 0211e0e8a35f..4faed6fce682 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piocnv50.c
@@ -29,7 +29,7 @@
 static void
 nv50_disp_pioc_fini(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	int ctrl = chan->chid.ctrl;
@@ -48,7 +48,7 @@ nv50_disp_pioc_fini(struct nv50_disp_chan *chan)
 static int
 nv50_disp_pioc_init(struct nv50_disp_chan *chan)
 {
-	struct nv50_disp *disp = chan->root->disp;
+	struct nv50_disp *disp = chan->disp;
 	struct nvkm_subdev *subdev = &disp->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	int ctrl = chan->chid.ctrl;
@@ -82,4 +82,6 @@ const struct nv50_disp_chan_func
 nv50_disp_pioc_func = {
 	.init = nv50_disp_pioc_init,
 	.fini = nv50_disp_pioc_fini,
+	.intr = nv50_disp_chan_intr,
+	.user = nv50_disp_chan_user,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piornv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piornv50.c
index 99b3b9050635..e997a207f546 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/piornv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/piornv50.c
@@ -127,8 +127,13 @@ nv50_pior = {
 int
 nv50_pior_new(struct nvkm_disp *disp, int id)
 {
-	struct nvkm_device *device = disp->engine.subdev.device;
-	if (!(nvkm_rd32(device, 0x610184) & (0x10000000 << id)))
-		return 0;
 	return nvkm_ior_new_(&nv50_pior, disp, PIOR, id);
 }
+
+int
+nv50_pior_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = (nvkm_rd32(device, 0x610184) & 0x70000000) >> 28;
+	return 3;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/priv.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/priv.h
index 6c9bfff6d043..ef66c5f38ad5 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/priv.h
@@ -12,6 +12,9 @@ void nvkm_disp_vblank(struct nvkm_disp *, int head);
 
 struct nvkm_disp_func {
 	void *(*dtor)(struct nvkm_disp *);
+	int (*oneinit)(struct nvkm_disp *);
+	int (*init)(struct nvkm_disp *);
+	void (*fini)(struct nvkm_disp *);
 	void (*intr)(struct nvkm_disp *);
 
 	const struct nvkm_disp_oclass *(*root)(struct nvkm_disp *);
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
index 721e4f74d1fc..1ed371fd7ddf 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg84.c
@@ -22,22 +22,19 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
 static const struct nv50_disp_root_func
 g84_disp_root = {
-	.init = nv50_disp_root_init,
-	.fini = nv50_disp_root_fini,
-	.dmac = {
-		&g84_disp_core_oclass,
-		&g84_disp_base_oclass,
-		&g84_disp_ovly_oclass,
-	},
-	.pioc = {
-		&g84_disp_oimm_oclass,
-		&g84_disp_curs_oclass,
+	.user = {
+		{{0,0,G82_DISP_CURSOR             }, nv50_disp_curs_new },
+		{{0,0,G82_DISP_OVERLAY            }, nv50_disp_oimm_new },
+		{{0,0,G82_DISP_BASE_CHANNEL_DMA   },  g84_disp_base_new },
+		{{0,0,G82_DISP_CORE_CHANNEL_DMA   },  g84_disp_core_new },
+		{{0,0,G82_DISP_OVERLAY_CHANNEL_DMA},  g84_disp_ovly_new },
+		{}
 	},
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
index 9493f6edf62b..ef579eb00238 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootg94.c
@@ -22,22 +22,19 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
 static const struct nv50_disp_root_func
 g94_disp_root = {
-	.init = nv50_disp_root_init,
-	.fini = nv50_disp_root_fini,
-	.dmac = {
-		&g94_disp_core_oclass,
-		&gt200_disp_base_oclass,
-		&gt200_disp_ovly_oclass,
-	},
-	.pioc = {
-		&g84_disp_oimm_oclass,
-		&g84_disp_curs_oclass,
+	.user = {
+		{{0,0,  G82_DISP_CURSOR             },  nv50_disp_curs_new },
+		{{0,0,  G82_DISP_OVERLAY            },  nv50_disp_oimm_new },
+		{{0,0,GT200_DISP_BASE_CHANNEL_DMA   },   g84_disp_base_new },
+		{{0,0,GT206_DISP_CORE_CHANNEL_DMA   },   g94_disp_core_new },
+		{{0,0,GT200_DISP_OVERLAY_CHANNEL_DMA}, gt200_disp_ovly_new },
+		{}
 	},
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
index 333c8424b413..fe011165dc02 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgf119.c
@@ -22,104 +22,19 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "head.h"
-#include "dmacnv50.h"
-
-#include <core/ramht.h>
-#include <subdev/timer.h>
+#include "channv50.h"
 
 #include <nvif/class.h>
 
-void
-gf119_disp_root_fini(struct nv50_disp_root *root)
-{
-	struct nvkm_device *device = root->disp->base.engine.subdev.device;
-	/* disable all interrupts */
-	nvkm_wr32(device, 0x6100b0, 0x00000000);
-}
-
-int
-gf119_disp_root_init(struct nv50_disp_root *root)
-{
-	struct nv50_disp *disp = root->disp;
-	struct nvkm_head *head;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
-	u32 tmp;
-	int i;
-
-	/* The below segments of code copying values from one register to
-	 * another appear to inform EVO of the display capabilities or
-	 * something similar.
-	 */
-
-	/* ... CRTC caps */
-	list_for_each_entry(head, &disp->base.head, head) {
-		const u32 hoff = head->id * 0x800;
-		tmp = nvkm_rd32(device, 0x616104 + hoff);
-		nvkm_wr32(device, 0x6101b4 + hoff, tmp);
-		tmp = nvkm_rd32(device, 0x616108 + hoff);
-		nvkm_wr32(device, 0x6101b8 + hoff, tmp);
-		tmp = nvkm_rd32(device, 0x61610c + hoff);
-		nvkm_wr32(device, 0x6101bc + hoff, tmp);
-	}
-
-	/* ... DAC caps */
-	for (i = 0; i < disp->func->dac.nr; i++) {
-		tmp = nvkm_rd32(device, 0x61a000 + (i * 0x800));
-		nvkm_wr32(device, 0x6101c0 + (i * 0x800), tmp);
-	}
-
-	/* ... SOR caps */
-	for (i = 0; i < disp->func->sor.nr; i++) {
-		tmp = nvkm_rd32(device, 0x61c000 + (i * 0x800));
-		nvkm_wr32(device, 0x6301c4 + (i * 0x800), tmp);
-	}
-
-	/* steal display away from vbios, or something like that */
-	if (nvkm_rd32(device, 0x6100ac) & 0x00000100) {
-		nvkm_wr32(device, 0x6100ac, 0x00000100);
-		nvkm_mask(device, 0x6194e8, 0x00000001, 0x00000000);
-		if (nvkm_msec(device, 2000,
-			if (!(nvkm_rd32(device, 0x6194e8) & 0x00000002))
-				break;
-		) < 0)
-			return -EBUSY;
-	}
-
-	/* point at display engine memory area (hash table, objects) */
-	nvkm_wr32(device, 0x610010, (root->instmem->addr >> 8) | 9);
-
-	/* enable supervisor interrupts, disable everything else */
-	nvkm_wr32(device, 0x610090, 0x00000000);
-	nvkm_wr32(device, 0x6100a0, 0x00000000);
-	nvkm_wr32(device, 0x6100b0, 0x00000307);
-
-	/* disable underflow reporting, preventing an intermittent issue
-	 * on some gk104 boards where the production vbios left this
-	 * setting enabled by default.
-	 *
-	 * ftp://download.nvidia.com/open-gpu-doc/gk104-disable-underflow-reporting/1/gk104-disable-underflow-reporting.txt
-	 */
-	list_for_each_entry(head, &disp->base.head, head) {
-		const u32 hoff = head->id * 0x800;
-		nvkm_mask(device, 0x616308 + hoff, 0x00000111, 0x00000010);
-	}
-
-	return 0;
-}
-
 static const struct nv50_disp_root_func
 gf119_disp_root = {
-	.init = gf119_disp_root_init,
-	.fini = gf119_disp_root_fini,
-	.dmac = {
-		&gf119_disp_core_oclass,
-		&gf119_disp_base_oclass,
-		&gf119_disp_ovly_oclass,
-	},
-	.pioc = {
-		&gf119_disp_oimm_oclass,
-		&gf119_disp_curs_oclass,
+	.user = {
+		{{0,0,GF110_DISP_CURSOR             }, gf119_disp_curs_new },
+		{{0,0,GF110_DISP_OVERLAY            }, gf119_disp_oimm_new },
+		{{0,0,GF110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
+		{{0,0,GF110_DISP_CORE_CHANNEL_DMA   }, gf119_disp_core_new },
+		{{0,0,GF110_DISP_OVERLAY_CONTROL_DMA}, gf119_disp_ovly_new },
+		{}
 	},
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
index 0bfdb1d1c6ab..9e8ffd348b50 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk104.c
@@ -22,22 +22,19 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
 static const struct nv50_disp_root_func
 gk104_disp_root = {
-	.init = gf119_disp_root_init,
-	.fini = gf119_disp_root_fini,
-	.dmac = {
-		&gk104_disp_core_oclass,
-		&gk104_disp_base_oclass,
-		&gk104_disp_ovly_oclass,
-	},
-	.pioc = {
-		&gk104_disp_oimm_oclass,
-		&gk104_disp_curs_oclass,
+	.user = {
+		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
+		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
+		{{0,0,GK104_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
+		{{0,0,GK104_DISP_CORE_CHANNEL_DMA   }, gk104_disp_core_new },
+		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
+		{}
 	},
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
index 1e8dbed8a67c..dc85cc1c9490 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgk110.c
@@ -22,22 +22,19 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
 static const struct nv50_disp_root_func
 gk110_disp_root = {
-	.init = gf119_disp_root_init,
-	.fini = gf119_disp_root_fini,
-	.dmac = {
-		&gk110_disp_core_oclass,
-		&gk110_disp_base_oclass,
-		&gk104_disp_ovly_oclass,
-	},
-	.pioc = {
-		&gk104_disp_oimm_oclass,
-		&gk104_disp_curs_oclass,
+	.user = {
+		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
+		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
+		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
+		{{0,0,GK110_DISP_CORE_CHANNEL_DMA   }, gk104_disp_core_new },
+		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
+		{}
 	},
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
index 44c55be69e99..e0181ca08840 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm107.c
@@ -22,22 +22,19 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
 static const struct nv50_disp_root_func
 gm107_disp_root = {
-	.init = gf119_disp_root_init,
-	.fini = gf119_disp_root_fini,
-	.dmac = {
-		&gm107_disp_core_oclass,
-		&gk110_disp_base_oclass,
-		&gk104_disp_ovly_oclass,
-	},
-	.pioc = {
-		&gk104_disp_oimm_oclass,
-		&gk104_disp_curs_oclass,
+	.user = {
+		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
+		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
+		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
+		{{0,0,GM107_DISP_CORE_CHANNEL_DMA   }, gk104_disp_core_new },
+		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
+		{}
 	},
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
index 38f5ee1dfc58..e5e590e19f62 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgm200.c
@@ -22,22 +22,19 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
 static const struct nv50_disp_root_func
 gm200_disp_root = {
-	.init = gf119_disp_root_init,
-	.fini = gf119_disp_root_fini,
-	.dmac = {
-		&gm200_disp_core_oclass,
-		&gk110_disp_base_oclass,
-		&gk104_disp_ovly_oclass,
-	},
-	.pioc = {
-		&gk104_disp_oimm_oclass,
-		&gk104_disp_curs_oclass,
+	.user = {
+		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
+		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
+		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
+		{{0,0,GM200_DISP_CORE_CHANNEL_DMA   }, gk104_disp_core_new },
+		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
+		{}
 	},
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
index ac8fdd728ec6..762a1a922e05 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp100.c
@@ -22,22 +22,19 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
 static const struct nv50_disp_root_func
 gp100_disp_root = {
-	.init = gf119_disp_root_init,
-	.fini = gf119_disp_root_fini,
-	.dmac = {
-		&gp100_disp_core_oclass,
-		&gk110_disp_base_oclass,
-		&gk104_disp_ovly_oclass,
-	},
-	.pioc = {
-		&gk104_disp_oimm_oclass,
-		&gk104_disp_curs_oclass,
+	.user = {
+		{{0,0,GK104_DISP_CURSOR             }, gf119_disp_curs_new },
+		{{0,0,GK104_DISP_OVERLAY            }, gf119_disp_oimm_new },
+		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gf119_disp_base_new },
+		{{0,0,GP100_DISP_CORE_CHANNEL_DMA   }, gk104_disp_core_new },
+		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gk104_disp_ovly_new },
+		{}
 	},
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
index 37122ca579ad..c7f00946c9af 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgp102.c
@@ -22,22 +22,19 @@
  * Authors: Ben Skeggs <bskeggs@redhat.com>
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
 static const struct nv50_disp_root_func
 gp102_disp_root = {
-	.init = gf119_disp_root_init,
-	.fini = gf119_disp_root_fini,
-	.dmac = {
-		&gp102_disp_core_oclass,
-		&gp102_disp_base_oclass,
-		&gp102_disp_ovly_oclass,
-	},
-	.pioc = {
-		&gp102_disp_oimm_oclass,
-		&gp102_disp_curs_oclass,
+	.user = {
+		{{0,0,GK104_DISP_CURSOR             }, gp102_disp_curs_new },
+		{{0,0,GK104_DISP_OVERLAY            }, gp102_disp_oimm_new },
+		{{0,0,GK110_DISP_BASE_CHANNEL_DMA   }, gp102_disp_base_new },
+		{{0,0,GP102_DISP_CORE_CHANNEL_DMA   }, gp102_disp_core_new },
+		{{0,0,GK104_DISP_OVERLAY_CONTROL_DMA}, gp102_disp_ovly_new },
+		{}
 	},
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
index 124a0c24f92c..a6963654087c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt200.c
@@ -22,22 +22,19 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
 static const struct nv50_disp_root_func
 gt200_disp_root = {
-	.init = nv50_disp_root_init,
-	.fini = nv50_disp_root_fini,
-	.dmac = {
-		&gt200_disp_core_oclass,
-		&gt200_disp_base_oclass,
-		&gt200_disp_ovly_oclass,
-	},
-	.pioc = {
-		&g84_disp_oimm_oclass,
-		&g84_disp_curs_oclass,
+	.user = {
+		{{0,0,  G82_DISP_CURSOR             },  nv50_disp_curs_new },
+		{{0,0,  G82_DISP_OVERLAY            },  nv50_disp_oimm_new },
+		{{0,0,GT200_DISP_BASE_CHANNEL_DMA   },   g84_disp_base_new },
+		{{0,0,GT200_DISP_CORE_CHANNEL_DMA   },   g84_disp_core_new },
+		{{0,0,GT200_DISP_OVERLAY_CHANNEL_DMA}, gt200_disp_ovly_new },
+		{}
 	},
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
index dff52f30668b..4fe0a3ae8891 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgt215.c
@@ -22,22 +22,19 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 
 #include <nvif/class.h>
 
 static const struct nv50_disp_root_func
 gt215_disp_root = {
-	.init = nv50_disp_root_init,
-	.fini = nv50_disp_root_fini,
-	.dmac = {
-		&gt215_disp_core_oclass,
-		&gt215_disp_base_oclass,
-		&gt215_disp_ovly_oclass,
-	},
-	.pioc = {
-		&gt215_disp_oimm_oclass,
-		&gt215_disp_curs_oclass,
+	.user = {
+		{{0,0,GT214_DISP_CURSOR             },  nv50_disp_curs_new },
+		{{0,0,GT214_DISP_OVERLAY            },  nv50_disp_oimm_new },
+		{{0,0,GT214_DISP_BASE_CHANNEL_DMA   },   g84_disp_base_new },
+		{{0,0,GT214_DISP_CORE_CHANNEL_DMA   },   g94_disp_core_new },
+		{{0,0,GT214_DISP_OVERLAY_CHANNEL_DMA},   g84_disp_ovly_new },
+		{}
 	},
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgv100.c
new file mode 100644
index 000000000000..9c658d632d37
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootgv100.c
@@ -0,0 +1,52 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "rootnv50.h"
+#include "channv50.h"
+
+#include <nvif/class.h>
+
+static const struct nv50_disp_root_func
+gv100_disp_root = {
+	.user = {
+		{{0,0,GV100_DISP_CURSOR                }, gv100_disp_curs_new },
+		{{0,0,GV100_DISP_WINDOW_IMM_CHANNEL_DMA}, gv100_disp_wimm_new },
+		{{0,0,GV100_DISP_CORE_CHANNEL_DMA      }, gv100_disp_core_new },
+		{{0,0,GV100_DISP_WINDOW_CHANNEL_DMA    }, gv100_disp_wndw_new },
+		{}
+	},
+};
+
+static int
+gv100_disp_root_new(struct nvkm_disp *disp, const struct nvkm_oclass *oclass,
+		    void *data, u32 size, struct nvkm_object **pobject)
+{
+	return nv50_disp_root_new_(&gv100_disp_root, disp, oclass,
+				   data, size, pobject);
+}
+
+const struct nvkm_disp_oclass
+gv100_disp_root_oclass = {
+	.base.oclass = GV100_DISP,
+	.base.minver = -1,
+	.base.maxver = -1,
+	.ctor = gv100_disp_root_new,
+};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
index 1208524aae14..3aa5a2879239 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.c
@@ -22,14 +22,12 @@
  * Authors: Ben Skeggs
  */
 #include "rootnv50.h"
-#include "dmacnv50.h"
+#include "channv50.h"
 #include "dp.h"
 #include "head.h"
 #include "ior.h"
 
 #include <core/client.h>
-#include <core/ramht.h>
-#include <subdev/timer.h>
 
 #include <nvif/class.h>
 #include <nvif/cl5070.h>
@@ -271,23 +269,12 @@ nv50_disp_root_mthd_(struct nvkm_object *object, u32 mthd, void *data, u32 size)
 }
 
 static int
-nv50_disp_root_dmac_new_(const struct nvkm_oclass *oclass,
-			 void *data, u32 size, struct nvkm_object **pobject)
+nv50_disp_root_child_new_(const struct nvkm_oclass *oclass,
+			  void *argv, u32 argc, struct nvkm_object **pobject)
 {
-	const struct nv50_disp_dmac_oclass *sclass = oclass->priv;
-	struct nv50_disp_root *root = nv50_disp_root(oclass->parent);
-	return sclass->ctor(sclass->func, sclass->mthd, root, sclass->chid,
-			    oclass, data, size, pobject);
-}
-
-static int
-nv50_disp_root_pioc_new_(const struct nvkm_oclass *oclass,
-			 void *data, u32 size, struct nvkm_object **pobject)
-{
-	const struct nv50_disp_pioc_oclass *sclass = oclass->priv;
-	struct nv50_disp_root *root = nv50_disp_root(oclass->parent);
-	return sclass->ctor(sclass->func, sclass->mthd, root, sclass->chid.ctrl,
-			    sclass->chid.user, oclass, data, size, pobject);
+	struct nv50_disp *disp = nv50_disp_root(oclass->parent)->disp;
+	const struct nv50_disp_user *user = oclass->priv;
+	return user->ctor(oclass, argv, argc, disp, pobject);
 }
 
 static int
@@ -296,68 +283,26 @@ nv50_disp_root_child_get_(struct nvkm_object *object, int index,
 {
 	struct nv50_disp_root *root = nv50_disp_root(object);
 
-	if (index < ARRAY_SIZE(root->func->dmac)) {
-		sclass->base = root->func->dmac[index]->base;
-		sclass->priv = root->func->dmac[index];
-		sclass->ctor = nv50_disp_root_dmac_new_;
-		return 0;
-	}
-
-	index -= ARRAY_SIZE(root->func->dmac);
-
-	if (index < ARRAY_SIZE(root->func->pioc)) {
-		sclass->base = root->func->pioc[index]->base;
-		sclass->priv = root->func->pioc[index];
-		sclass->ctor = nv50_disp_root_pioc_new_;
+	if (root->func->user[index].ctor) {
+		sclass->base = root->func->user[index].base;
+		sclass->priv = root->func->user + index;
+		sclass->ctor = nv50_disp_root_child_new_;
 		return 0;
 	}
 
 	return -EINVAL;
 }
 
-static int
-nv50_disp_root_fini_(struct nvkm_object *object, bool suspend)
-{
-	struct nv50_disp_root *root = nv50_disp_root(object);
-	root->func->fini(root);
-	return 0;
-}
-
-static int
-nv50_disp_root_init_(struct nvkm_object *object)
-{
-	struct nv50_disp_root *root = nv50_disp_root(object);
-	struct nvkm_ior *ior;
-	int ret;
-
-	ret = root->func->init(root);
-	if (ret)
-		return ret;
-
-	/* Set 'normal' (ie. when it's attached to a head) state for
-	 * each output resource to 'fully enabled'.
-	 */
-	list_for_each_entry(ior, &root->disp->base.ior, head) {
-		ior->func->power(ior, true, true, true, true, true);
-	}
-
-	return 0;
-}
-
 static void *
 nv50_disp_root_dtor_(struct nvkm_object *object)
 {
 	struct nv50_disp_root *root = nv50_disp_root(object);
-	nvkm_ramht_del(&root->ramht);
-	nvkm_gpuobj_del(&root->instmem);
 	return root;
 }
 
 static const struct nvkm_object_func
 nv50_disp_root_ = {
 	.dtor = nv50_disp_root_dtor_,
-	.init = nv50_disp_root_init_,
-	.fini = nv50_disp_root_fini_,
 	.mthd = nv50_disp_root_mthd_,
 	.ntfy = nvkm_disp_ntfy,
 	.sclass = nv50_disp_root_child_get_,
@@ -370,8 +315,6 @@ nv50_disp_root_new_(const struct nv50_disp_root_func *func,
 {
 	struct nv50_disp *disp = nv50_disp(base);
 	struct nv50_disp_root *root;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
-	int ret;
 
 	if (!(root = kzalloc(sizeof(*root), GFP_KERNEL)))
 		return -ENOMEM;
@@ -380,102 +323,18 @@ nv50_disp_root_new_(const struct nv50_disp_root_func *func,
 	nvkm_object_ctor(&nv50_disp_root_, oclass, &root->object);
 	root->func = func;
 	root->disp = disp;
-
-	ret = nvkm_gpuobj_new(disp->base.engine.subdev.device, 0x10000, 0x10000,
-			      false, NULL, &root->instmem);
-	if (ret)
-		return ret;
-
-	return nvkm_ramht_new(device, 0x1000, 0, root->instmem, &root->ramht);
-}
-
-void
-nv50_disp_root_fini(struct nv50_disp_root *root)
-{
-	struct nvkm_device *device = root->disp->base.engine.subdev.device;
-	/* disable all interrupts */
-	nvkm_wr32(device, 0x610024, 0x00000000);
-	nvkm_wr32(device, 0x610020, 0x00000000);
-}
-
-int
-nv50_disp_root_init(struct nv50_disp_root *root)
-{
-	struct nv50_disp *disp = root->disp;
-	struct nvkm_head *head;
-	struct nvkm_device *device = disp->base.engine.subdev.device;
-	u32 tmp;
-	int i;
-
-	/* The below segments of code copying values from one register to
-	 * another appear to inform EVO of the display capabilities or
-	 * something similar.  NFI what the 0x614004 caps are for..
-	 */
-	tmp = nvkm_rd32(device, 0x614004);
-	nvkm_wr32(device, 0x610184, tmp);
-
-	/* ... CRTC caps */
-	list_for_each_entry(head, &disp->base.head, head) {
-		tmp = nvkm_rd32(device, 0x616100 + (head->id * 0x800));
-		nvkm_wr32(device, 0x610190 + (head->id * 0x10), tmp);
-		tmp = nvkm_rd32(device, 0x616104 + (head->id * 0x800));
-		nvkm_wr32(device, 0x610194 + (head->id * 0x10), tmp);
-		tmp = nvkm_rd32(device, 0x616108 + (head->id * 0x800));
-		nvkm_wr32(device, 0x610198 + (head->id * 0x10), tmp);
-		tmp = nvkm_rd32(device, 0x61610c + (head->id * 0x800));
-		nvkm_wr32(device, 0x61019c + (head->id * 0x10), tmp);
-	}
-
-	/* ... DAC caps */
-	for (i = 0; i < disp->func->dac.nr; i++) {
-		tmp = nvkm_rd32(device, 0x61a000 + (i * 0x800));
-		nvkm_wr32(device, 0x6101d0 + (i * 0x04), tmp);
-	}
-
-	/* ... SOR caps */
-	for (i = 0; i < disp->func->sor.nr; i++) {
-		tmp = nvkm_rd32(device, 0x61c000 + (i * 0x800));
-		nvkm_wr32(device, 0x6101e0 + (i * 0x04), tmp);
-	}
-
-	/* ... PIOR caps */
-	for (i = 0; i < disp->func->pior.nr; i++) {
-		tmp = nvkm_rd32(device, 0x61e000 + (i * 0x800));
-		nvkm_wr32(device, 0x6101f0 + (i * 0x04), tmp);
-	}
-
-	/* steal display away from vbios, or something like that */
-	if (nvkm_rd32(device, 0x610024) & 0x00000100) {
-		nvkm_wr32(device, 0x610024, 0x00000100);
-		nvkm_mask(device, 0x6194e8, 0x00000001, 0x00000000);
-		if (nvkm_msec(device, 2000,
-			if (!(nvkm_rd32(device, 0x6194e8) & 0x00000002))
-				break;
-		) < 0)
-			return -EBUSY;
-	}
-
-	/* point at display engine memory area (hash table, objects) */
-	nvkm_wr32(device, 0x610010, (root->instmem->addr >> 8) | 9);
-
-	/* enable supervisor interrupts, disable everything else */
-	nvkm_wr32(device, 0x61002c, 0x00000370);
-	nvkm_wr32(device, 0x610028, 0x00000000);
 	return 0;
 }
 
 static const struct nv50_disp_root_func
 nv50_disp_root = {
-	.init = nv50_disp_root_init,
-	.fini = nv50_disp_root_fini,
-	.dmac = {
-		&nv50_disp_core_oclass,
-		&nv50_disp_base_oclass,
-		&nv50_disp_ovly_oclass,
-	},
-	.pioc = {
-		&nv50_disp_oimm_oclass,
-		&nv50_disp_curs_oclass,
+	.user = {
+		{{0,0,NV50_DISP_CURSOR             }, nv50_disp_curs_new },
+		{{0,0,NV50_DISP_OVERLAY            }, nv50_disp_oimm_new },
+		{{0,0,NV50_DISP_BASE_CHANNEL_DMA   }, nv50_disp_base_new },
+		{{0,0,NV50_DISP_CORE_CHANNEL_DMA   }, nv50_disp_core_new },
+		{{0,0,NV50_DISP_OVERLAY_CHANNEL_DMA}, nv50_disp_ovly_new },
+		{}
 	},
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
index 4818fa69ae6c..6ca4f9184b51 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/rootnv50.h
@@ -2,34 +2,27 @@
 #ifndef __NV50_DISP_ROOT_H__
 #define __NV50_DISP_ROOT_H__
 #define nv50_disp_root(p) container_of((p), struct nv50_disp_root, object)
+#include <core/object.h>
 #include "nv50.h"
-#include "channv50.h"
-#include "dmacnv50.h"
 
 struct nv50_disp_root {
 	const struct nv50_disp_root_func *func;
 	struct nv50_disp *disp;
 	struct nvkm_object object;
-
-	struct nvkm_gpuobj *instmem;
-	struct nvkm_ramht *ramht;
 };
 
 struct nv50_disp_root_func {
-	int (*init)(struct nv50_disp_root *);
-	void (*fini)(struct nv50_disp_root *);
-	const struct nv50_disp_dmac_oclass *dmac[3];
-	const struct nv50_disp_pioc_oclass *pioc[2];
+	int blah;
+	struct nv50_disp_user {
+		struct nvkm_sclass base;
+		int (*ctor)(const struct nvkm_oclass *, void *argv, u32 argc,
+			    struct nv50_disp *, struct nvkm_object **);
+	} user[];
 };
 
 int  nv50_disp_root_new_(const struct nv50_disp_root_func *, struct nvkm_disp *,
 			 const struct nvkm_oclass *, void *data, u32 size,
 			 struct nvkm_object **);
-int  nv50_disp_root_init(struct nv50_disp_root *);
-void nv50_disp_root_fini(struct nv50_disp_root *);
-
-int  gf119_disp_root_init(struct nv50_disp_root *);
-void gf119_disp_root_fini(struct nv50_disp_root *);
 
 extern const struct nvkm_disp_oclass nv50_disp_root_oclass;
 extern const struct nvkm_disp_oclass g84_disp_root_oclass;
@@ -43,4 +36,5 @@ extern const struct nvkm_disp_oclass gm107_disp_root_oclass;
 extern const struct nvkm_disp_oclass gm200_disp_root_oclass;
 extern const struct nvkm_disp_oclass gp100_disp_root_oclass;
 extern const struct nvkm_disp_oclass gp102_disp_root_oclass;
+extern const struct nvkm_disp_oclass gv100_disp_root_oclass;
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg84.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg84.c
index f40b909b4ca2..ec3a7db08118 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg84.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg84.c
@@ -34,5 +34,5 @@ g84_sor = {
 int
 g84_sor_new(struct nvkm_disp *disp, int id)
 {
-	return nv50_sor_new_(&g84_sor, disp, id);
+	return nvkm_ior_new_(&g84_sor, disp, SOR, id);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg94.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg94.c
index 49aeafde0031..4d59d02525d9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg94.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorg94.c
@@ -279,5 +279,13 @@ g94_sor = {
 int
 g94_sor_new(struct nvkm_disp *disp, int id)
 {
-	return nv50_sor_new_(&g94_sor, disp, id);
+	return nvkm_ior_new_(&g94_sor, disp, SOR, id);
+}
+
+int
+g94_sor_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = (nvkm_rd32(device, 0x610184) & 0x0f000000) >> 24;
+	return 4;
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c
index 700fc754f28a..e6e6dfbb1283 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgf119.c
@@ -152,15 +152,6 @@ gf119_sor_state(struct nvkm_ior *sor, struct nvkm_ior_state *state)
 	state->head = ctrl & 0x0000000f;
 }
 
-int
-gf119_sor_new_(const struct nvkm_ior_func *func, struct nvkm_disp *disp, int id)
-{
-	struct nvkm_device *device = disp->engine.subdev.device;
-	if (!(nvkm_rd32(device, 0x612004) & (0x00000100 << id)))
-		return 0;
-	return nvkm_ior_new_(func, disp, SOR, id);
-}
-
 static const struct nvkm_ior_func
 gf119_sor = {
 	.state = gf119_sor_state,
@@ -189,5 +180,13 @@ gf119_sor = {
 int
 gf119_sor_new(struct nvkm_disp *disp, int id)
 {
-	return gf119_sor_new_(&gf119_sor, disp, id);
+	return nvkm_ior_new_(&gf119_sor, disp, SOR, id);
+}
+
+int
+gf119_sor_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = (nvkm_rd32(device, 0x612004) & 0x0000ff00) >> 8;
+	return 8;
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgk104.c
index a1547bdf490b..b94090edaebf 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgk104.c
@@ -49,5 +49,5 @@ gk104_sor = {
 int
 gk104_sor_new(struct nvkm_disp *disp, int id)
 {
-	return gf119_sor_new_(&gk104_sor, disp, id);
+	return nvkm_ior_new_(&gk104_sor, disp, SOR, id);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm107.c
index 60230957d82b..e6965dec09c9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm107.c
@@ -63,5 +63,5 @@ gm107_sor = {
 int
 gm107_sor_new(struct nvkm_disp *disp, int id)
 {
-	return gf119_sor_new_(&gm107_sor, disp, id);
+	return nvkm_ior_new_(&gm107_sor, disp, SOR, id);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c
index f9b8107aa2a2..d892bdf04034 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgm200.c
@@ -23,7 +23,7 @@
  */
 #include "ior.h"
 
-static void
+void
 gm200_sor_dp_drive(struct nvkm_ior *sor, int ln, int pc, int dc, int pe, int pu)
 {
 	struct nvkm_device *device = sor->disp->engine.subdev.device;
@@ -45,7 +45,7 @@ gm200_sor_dp_drive(struct nvkm_ior *sor, int ln, int pc, int dc, int pe, int pu)
 	nvkm_wr32(device, 0x61c13c + loff, data[3] | (pc << shift));
 }
 
-static void
+void
 gm200_sor_route_set(struct nvkm_outp *outp, struct nvkm_ior *ior)
 {
 	struct nvkm_device *device = outp->disp->engine.subdev.device;
@@ -62,7 +62,7 @@ gm200_sor_route_set(struct nvkm_outp *outp, struct nvkm_ior *ior)
 		nvkm_mask(device, 0x612388 + moff, 0x0000001f, link << 4 | sor);
 }
 
-static int
+int
 gm200_sor_route_get(struct nvkm_outp *outp, int *link)
 {
 	struct nvkm_device *device = outp->disp->engine.subdev.device;
@@ -120,5 +120,5 @@ gm200_sor = {
 int
 gm200_sor_new(struct nvkm_disp *disp, int id)
 {
-	return gf119_sor_new_(&gm200_sor, disp, id);
+	return nvkm_ior_new_(&gm200_sor, disp, SOR, id);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgt215.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgt215.c
index da228b54b43e..54d134d4ca1d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgt215.c
@@ -65,5 +65,5 @@ gt215_sor = {
 int
 gt215_sor_new(struct nvkm_disp *disp, int id)
 {
-	return nv50_sor_new_(&gt215_sor, disp, id);
+	return nvkm_ior_new_(&gt215_sor, disp, SOR, id);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgv100.c
new file mode 100644
index 000000000000..040db8a338de
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sorgv100.c
@@ -0,0 +1,120 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ior.h"
+
+#include <subdev/timer.h>
+
+static void
+gv100_sor_dp_watermark(struct nvkm_ior *sor, int head, u8 watermark)
+{
+	struct nvkm_device *device = sor->disp->engine.subdev.device;
+	const u32 hoff = head * 0x800;
+	nvkm_mask(device, 0x616550 + hoff, 0x0c00003f, 0x08000000 | watermark);
+}
+
+static void
+gv100_sor_dp_audio_sym(struct nvkm_ior *sor, int head, u16 h, u32 v)
+{
+	struct nvkm_device *device = sor->disp->engine.subdev.device;
+	const u32 hoff = head * 0x800;
+	nvkm_mask(device, 0x616568 + hoff, 0x0000ffff, h);
+	nvkm_mask(device, 0x61656c + hoff, 0x00ffffff, v);
+}
+
+static void
+gv100_sor_dp_audio(struct nvkm_ior *sor, int head, bool enable)
+{
+	struct nvkm_device *device = sor->disp->engine.subdev.device;
+	const u32 hoff = 0x800 * head;
+	const u32 data = 0x80000000 | (0x00000001 * enable);
+	const u32 mask = 0x8000000d;
+	nvkm_mask(device, 0x616560 + hoff, mask, data);
+	nvkm_msec(device, 2000,
+		if (!(nvkm_rd32(device, 0x616560 + hoff) & 0x80000000))
+			break;
+	);
+}
+
+static void
+gv100_sor_state(struct nvkm_ior *sor, struct nvkm_ior_state *state)
+{
+	struct nvkm_device *device = sor->disp->engine.subdev.device;
+	const u32 coff = (state == &sor->arm) * 0x8000 + sor->id * 0x20;
+	u32 ctrl = nvkm_rd32(device, 0x680300 + coff);
+
+	state->proto_evo = (ctrl & 0x00000f00) >> 8;
+	switch (state->proto_evo) {
+	case 0: state->proto = LVDS; state->link = 1; break;
+	case 1: state->proto = TMDS; state->link = 1; break;
+	case 2: state->proto = TMDS; state->link = 2; break;
+	case 5: state->proto = TMDS; state->link = 3; break;
+	case 8: state->proto =   DP; state->link = 1; break;
+	case 9: state->proto =   DP; state->link = 2; break;
+	default:
+		state->proto = UNKNOWN;
+		break;
+	}
+
+	state->head = ctrl & 0x000000ff;
+}
+
+static const struct nvkm_ior_func
+gv100_sor = {
+	.route = {
+		.get = gm200_sor_route_get,
+		.set = gm200_sor_route_set,
+	},
+	.state = gv100_sor_state,
+	.power = nv50_sor_power,
+	.clock = gf119_sor_clock,
+	.hdmi = {
+		.ctrl = gv100_hdmi_ctrl,
+	},
+	.dp = {
+		.lanes = { 0, 1, 2, 3 },
+		.links = gf119_sor_dp_links,
+		.power = g94_sor_dp_power,
+		.pattern = gm107_sor_dp_pattern,
+		.drive = gm200_sor_dp_drive,
+		.audio = gv100_sor_dp_audio,
+		.audio_sym = gv100_sor_dp_audio_sym,
+		.watermark = gv100_sor_dp_watermark,
+	},
+	.hda = {
+		.hpd = gf119_hda_hpd,
+		.eld = gf119_hda_eld,
+	},
+};
+
+int
+gv100_sor_new(struct nvkm_disp *disp, int id)
+{
+	return nvkm_ior_new_(&gv100_sor, disp, SOR, id);
+}
+
+int
+gv100_sor_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = (nvkm_rd32(device, 0x610060) & 0x0000ff00) >> 8;
+	return (nvkm_rd32(device, 0x610074) & 0x00000f00) >> 8;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp77.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp77.c
index c0179ccb956d..8a70dd25b13a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp77.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp77.c
@@ -44,5 +44,5 @@ mcp77_sor = {
 int
 mcp77_sor_new(struct nvkm_disp *disp, int id)
 {
-	return nv50_sor_new_(&mcp77_sor, disp, id);
+	return nvkm_ior_new_(&mcp77_sor, disp, SOR, id);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp89.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp89.c
index 9bb01cd96697..eac9c5be9166 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp89.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sormcp89.c
@@ -49,5 +49,5 @@ mcp89_sor = {
 int
 mcp89_sor_new(struct nvkm_disp *disp, int id)
 {
-	return nv50_sor_new_(&mcp89_sor, disp, id);
+	return nvkm_ior_new_(&mcp89_sor, disp, SOR, id);
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sornv50.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sornv50.c
index f3ebd0c22e7d..b4729f8798af 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/disp/sornv50.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/sornv50.c
@@ -84,15 +84,6 @@ nv50_sor_state(struct nvkm_ior *sor, struct nvkm_ior_state *state)
 	state->head = ctrl & 0x00000003;
 }
 
-int
-nv50_sor_new_(const struct nvkm_ior_func *func, struct nvkm_disp *disp, int id)
-{
-	struct nvkm_device *device = disp->engine.subdev.device;
-	if (!(nvkm_rd32(device, 0x610184) & (0x01000000 << id)))
-		return 0;
-	return nvkm_ior_new_(func, disp, SOR, id);
-}
-
 static const struct nvkm_ior_func
 nv50_sor = {
 	.state = nv50_sor_state,
@@ -103,5 +94,13 @@ nv50_sor = {
 int
 nv50_sor_new(struct nvkm_disp *disp, int id)
 {
-	return nv50_sor_new_(&nv50_sor, disp, id);
+	return nvkm_ior_new_(&nv50_sor, disp, SOR, id);
+}
+
+int
+nv50_sor_cnt(struct nvkm_disp *disp, unsigned long *pmask)
+{
+	struct nvkm_device *device = disp->engine.subdev.device;
+	*pmask = (nvkm_rd32(device, 0x610184) & 0x03000000) >> 24;
+	return 2;
 }
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/wimmgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/wimmgv100.c
new file mode 100644
index 000000000000..89d783368b4f
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/wimmgv100.c
@@ -0,0 +1,82 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "channv50.h"
+
+#include <core/client.h>
+
+#include <nvif/clc37b.h>
+#include <nvif/unpack.h>
+
+static void
+gv100_disp_wimm_intr(struct nv50_disp_chan *chan, bool en)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 mask = 0x00000001 << chan->head;
+	const u32 data = en ? mask : 0;
+	nvkm_mask(device, 0x611da8, mask, data);
+}
+
+const struct nv50_disp_chan_func
+gv100_disp_wimm = {
+	.init = gv100_disp_dmac_init,
+	.fini = gv100_disp_dmac_fini,
+	.intr = gv100_disp_wimm_intr,
+	.user = gv100_disp_chan_user,
+};
+
+static int
+gv100_disp_wimm_new_(const struct nv50_disp_chan_func *func,
+		     const struct nv50_disp_chan_mthd *mthd,
+		     struct nv50_disp *disp, int chid,
+		     const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		     struct nvkm_object **pobject)
+{
+	union {
+		struct nvc37b_window_imm_channel_dma_v0 v0;
+	} *args = argv;
+	struct nvkm_object *parent = oclass->parent;
+	int wndw, ret = -ENOSYS;
+	u64 push;
+
+	nvif_ioctl(parent, "create window imm channel dma size %d\n", argc);
+	if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))) {
+		nvif_ioctl(parent, "create window imm channel dma vers %d "
+				   "pushbuf %016llx index %d\n",
+			   args->v0.version, args->v0.pushbuf, args->v0.index);
+		if (!(disp->wndw.mask & BIT(args->v0.index)))
+			return -EINVAL;
+		push = args->v0.pushbuf;
+		wndw = args->v0.index;
+	} else
+		return ret;
+
+	return nv50_disp_dmac_new_(func, mthd, disp, chid + wndw,
+				   wndw, push, oclass, pobject);
+}
+
+int
+gv100_disp_wimm_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return gv100_disp_wimm_new_(&gv100_disp_wimm, NULL, disp, 33,
+				    oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/disp/wndwgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/disp/wndwgv100.c
new file mode 100644
index 000000000000..98911805aabf
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/disp/wndwgv100.c
@@ -0,0 +1,184 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "channv50.h"
+
+#include <core/client.h>
+
+#include <nvif/clc37e.h>
+#include <nvif/unpack.h>
+
+static const struct nv50_disp_mthd_list
+gv100_disp_wndw_mthd_base = {
+	.mthd = 0x0000,
+	.addr = 0x000000,
+	.data = {
+		{ 0x0200, 0x690200 },
+		{ 0x020c, 0x69020c },
+		{ 0x0210, 0x690210 },
+		{ 0x0214, 0x690214 },
+		{ 0x0218, 0x690218 },
+		{ 0x021c, 0x69021c },
+		{ 0x0220, 0x690220 },
+		{ 0x0224, 0x690224 },
+		{ 0x0228, 0x690228 },
+		{ 0x022c, 0x69022c },
+		{ 0x0230, 0x690230 },
+		{ 0x0234, 0x690234 },
+		{ 0x0238, 0x690238 },
+		{ 0x0240, 0x690240 },
+		{ 0x0244, 0x690244 },
+		{ 0x0248, 0x690248 },
+		{ 0x024c, 0x69024c },
+		{ 0x0250, 0x690250 },
+		{ 0x0254, 0x690254 },
+		{ 0x0260, 0x690260 },
+		{ 0x0264, 0x690264 },
+		{ 0x0268, 0x690268 },
+		{ 0x026c, 0x69026c },
+		{ 0x0270, 0x690270 },
+		{ 0x0274, 0x690274 },
+		{ 0x0280, 0x690280 },
+		{ 0x0284, 0x690284 },
+		{ 0x0288, 0x690288 },
+		{ 0x028c, 0x69028c },
+		{ 0x0290, 0x690290 },
+		{ 0x0298, 0x690298 },
+		{ 0x029c, 0x69029c },
+		{ 0x02a0, 0x6902a0 },
+		{ 0x02a4, 0x6902a4 },
+		{ 0x02a8, 0x6902a8 },
+		{ 0x02ac, 0x6902ac },
+		{ 0x02b0, 0x6902b0 },
+		{ 0x02b4, 0x6902b4 },
+		{ 0x02b8, 0x6902b8 },
+		{ 0x02bc, 0x6902bc },
+		{ 0x02c0, 0x6902c0 },
+		{ 0x02c4, 0x6902c4 },
+		{ 0x02c8, 0x6902c8 },
+		{ 0x02cc, 0x6902cc },
+		{ 0x02d0, 0x6902d0 },
+		{ 0x02d4, 0x6902d4 },
+		{ 0x02d8, 0x6902d8 },
+		{ 0x02dc, 0x6902dc },
+		{ 0x02e0, 0x6902e0 },
+		{ 0x02e4, 0x6902e4 },
+		{ 0x02e8, 0x6902e8 },
+		{ 0x02ec, 0x6902ec },
+		{ 0x02f0, 0x6902f0 },
+		{ 0x02f4, 0x6902f4 },
+		{ 0x02f8, 0x6902f8 },
+		{ 0x02fc, 0x6902fc },
+		{ 0x0300, 0x690300 },
+		{ 0x0304, 0x690304 },
+		{ 0x0308, 0x690308 },
+		{ 0x0310, 0x690310 },
+		{ 0x0314, 0x690314 },
+		{ 0x0318, 0x690318 },
+		{ 0x031c, 0x69031c },
+		{ 0x0320, 0x690320 },
+		{ 0x0324, 0x690324 },
+		{ 0x0328, 0x690328 },
+		{ 0x032c, 0x69032c },
+		{ 0x033c, 0x69033c },
+		{ 0x0340, 0x690340 },
+		{ 0x0344, 0x690344 },
+		{ 0x0348, 0x690348 },
+		{ 0x034c, 0x69034c },
+		{ 0x0350, 0x690350 },
+		{ 0x0354, 0x690354 },
+		{ 0x0358, 0x690358 },
+		{ 0x0364, 0x690364 },
+		{ 0x0368, 0x690368 },
+		{ 0x036c, 0x69036c },
+		{ 0x0370, 0x690370 },
+		{ 0x0374, 0x690374 },
+		{ 0x0380, 0x690380 },
+		{}
+	}
+};
+
+const struct nv50_disp_chan_mthd
+gv100_disp_wndw_mthd = {
+	.name = "Base",
+	.addr = 0x001000,
+	.prev = 0x000800,
+	.data = {
+		{ "Global", 1, &gv100_disp_wndw_mthd_base },
+		{}
+	}
+};
+
+static void
+gv100_disp_wndw_intr(struct nv50_disp_chan *chan, bool en)
+{
+	struct nvkm_device *device = chan->disp->base.engine.subdev.device;
+	const u32 mask = 0x00000001 << chan->head;
+	const u32 data = en ? mask : 0;
+	nvkm_mask(device, 0x611da4, mask, data);
+}
+
+const struct nv50_disp_chan_func
+gv100_disp_wndw = {
+	.init = gv100_disp_dmac_init,
+	.fini = gv100_disp_dmac_fini,
+	.intr = gv100_disp_wndw_intr,
+	.user = gv100_disp_chan_user,
+	.bind = gv100_disp_dmac_bind,
+};
+
+static int
+gv100_disp_wndw_new_(const struct nv50_disp_chan_func *func,
+		     const struct nv50_disp_chan_mthd *mthd,
+		     struct nv50_disp *disp, int chid,
+		     const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		     struct nvkm_object **pobject)
+{
+	union {
+		struct nvc37e_window_channel_dma_v0 v0;
+	} *args = argv;
+	struct nvkm_object *parent = oclass->parent;
+	int wndw, ret = -ENOSYS;
+	u64 push;
+
+	nvif_ioctl(parent, "create window channel dma size %d\n", argc);
+	if (!(ret = nvif_unpack(ret, &argv, &argc, args->v0, 0, 0, false))) {
+		nvif_ioctl(parent, "create window channel dma vers %d "
+				   "pushbuf %016llx index %d\n",
+			   args->v0.version, args->v0.pushbuf, args->v0.index);
+		if (!(disp->wndw.mask & BIT(args->v0.index)))
+			return -EINVAL;
+		push = args->v0.pushbuf;
+		wndw = args->v0.index;
+	} else
+		return ret;
+
+	return nv50_disp_dmac_new_(func, mthd, disp, chid + wndw,
+				   wndw, push, oclass, pobject);
+}
+
+int
+gv100_disp_wndw_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nv50_disp *disp, struct nvkm_object **pobject)
+{
+	return gv100_disp_wndw_new_(&gv100_disp_wndw, &gv100_disp_wndw_mthd,
+				    disp, 1, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/dma/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/dma/Kbuild
index c4a2ce9b0d71..e96d1f57f9f9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/dma/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/dma/Kbuild
@@ -3,9 +3,11 @@ nvkm-y += nvkm/engine/dma/nv04.o
 nvkm-y += nvkm/engine/dma/nv50.o
 nvkm-y += nvkm/engine/dma/gf100.o
 nvkm-y += nvkm/engine/dma/gf119.o
+nvkm-y += nvkm/engine/dma/gv100.o
 
 nvkm-y += nvkm/engine/dma/user.o
 nvkm-y += nvkm/engine/dma/usernv04.o
 nvkm-y += nvkm/engine/dma/usernv50.o
 nvkm-y += nvkm/engine/dma/usergf100.o
 nvkm-y += nvkm/engine/dma/usergf119.o
+nvkm-y += nvkm/engine/dma/usergv100.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/dma/gv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/dma/gv100.c
new file mode 100644
index 000000000000..c65a4c2ea93d
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/dma/gv100.c
@@ -0,0 +1,34 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "priv.h"
+#include "user.h"
+
+static const struct nvkm_dma_func
+gv100_dma = {
+	.class_new = gv100_dmaobj_new,
+};
+
+int
+gv100_dma_new(struct nvkm_device *device, int index, struct nvkm_dma **pdma)
+{
+	return nvkm_dma_new_(&gv100_dma, device, index, pdma);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/dma/user.h b/drivers/gpu/drm/nouveau/nvkm/engine/dma/user.h
index 4bbac8a21c71..9fe01fd75474 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/dma/user.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/dma/user.h
@@ -16,4 +16,6 @@ int gf100_dmaobj_new(struct nvkm_dma *, const struct nvkm_oclass *, void *, u32,
 		     struct nvkm_dmaobj **);
 int gf119_dmaobj_new(struct nvkm_dma *, const struct nvkm_oclass *, void *, u32,
 		     struct nvkm_dmaobj **);
+int gv100_dmaobj_new(struct nvkm_dma *, const struct nvkm_oclass *, void *, u32,
+		     struct nvkm_dmaobj **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/dma/usergv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/dma/usergv100.c
new file mode 100644
index 000000000000..39eba9fc82be
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/dma/usergv100.c
@@ -0,0 +1,119 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#define gv100_dmaobj(p) container_of((p), struct gv100_dmaobj, base)
+#include "user.h"
+
+#include <core/client.h>
+#include <core/gpuobj.h>
+#include <subdev/fb.h>
+
+#include <nvif/cl0002.h>
+#include <nvif/unpack.h>
+
+struct gv100_dmaobj {
+	struct nvkm_dmaobj base;
+	u32 flags0;
+};
+
+static int
+gv100_dmaobj_bind(struct nvkm_dmaobj *base, struct nvkm_gpuobj *parent,
+		  int align, struct nvkm_gpuobj **pgpuobj)
+{
+	struct gv100_dmaobj *dmaobj = gv100_dmaobj(base);
+	struct nvkm_device *device = dmaobj->base.dma->engine.subdev.device;
+	u64 start = dmaobj->base.start >> 8;
+	u64 limit = dmaobj->base.limit >> 8;
+	int ret;
+
+	ret = nvkm_gpuobj_new(device, 24, align, false, parent, pgpuobj);
+	if (ret == 0) {
+		nvkm_kmap(*pgpuobj);
+		nvkm_wo32(*pgpuobj, 0x00, dmaobj->flags0);
+		nvkm_wo32(*pgpuobj, 0x04, lower_32_bits(start));
+		nvkm_wo32(*pgpuobj, 0x08, upper_32_bits(start));
+		nvkm_wo32(*pgpuobj, 0x0c, lower_32_bits(limit));
+		nvkm_wo32(*pgpuobj, 0x10, upper_32_bits(limit));
+		nvkm_done(*pgpuobj);
+	}
+
+	return ret;
+}
+
+static const struct nvkm_dmaobj_func
+gv100_dmaobj_func = {
+	.bind = gv100_dmaobj_bind,
+};
+
+int
+gv100_dmaobj_new(struct nvkm_dma *dma, const struct nvkm_oclass *oclass,
+		 void *data, u32 size, struct nvkm_dmaobj **pdmaobj)
+{
+	union {
+		struct gf119_dma_v0 v0;
+	} *args;
+	struct nvkm_object *parent = oclass->parent;
+	struct gv100_dmaobj *dmaobj;
+	u32 kind, page;
+	int ret;
+
+	if (!(dmaobj = kzalloc(sizeof(*dmaobj), GFP_KERNEL)))
+		return -ENOMEM;
+	*pdmaobj = &dmaobj->base;
+
+	ret = nvkm_dmaobj_ctor(&gv100_dmaobj_func, dma, oclass,
+			       &data, &size, &dmaobj->base);
+	if (ret)
+		return ret;
+
+	ret  = -ENOSYS;
+	args = data;
+
+	nvif_ioctl(parent, "create gv100 dma size %d\n", size);
+	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+		nvif_ioctl(parent,
+			   "create gv100 dma vers %d page %d kind %02x\n",
+			   args->v0.version, args->v0.page, args->v0.kind);
+		kind = args->v0.kind != 0;
+		page = args->v0.page != 0;
+	} else
+	if (size == 0) {
+		kind = 0;
+		page = GF119_DMA_V0_PAGE_SP;
+	} else
+		return ret;
+
+	if (kind)
+		dmaobj->flags0 |= 0x00100000;
+	if (page)
+		dmaobj->flags0 |= 0x00000040;
+	dmaobj->flags0 |= 0x00000004; /* rw */
+
+	switch (dmaobj->base.target) {
+	case NV_MEM_TARGET_VRAM       : dmaobj->flags0 |= 0x00000001; break;
+	case NV_MEM_TARGET_PCI        : dmaobj->flags0 |= 0x00000002; break;
+	case NV_MEM_TARGET_PCI_NOSNOOP: dmaobj->flags0 |= 0x00000003; break;
+	default:
+		return -EINVAL;
+	}
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild
index 64e51838edf8..f00408577a6a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/Kbuild
@@ -15,6 +15,7 @@ nvkm-y += nvkm/engine/fifo/gm200.o
 nvkm-y += nvkm/engine/fifo/gm20b.o
 nvkm-y += nvkm/engine/fifo/gp100.o
 nvkm-y += nvkm/engine/fifo/gp10b.o
+nvkm-y += nvkm/engine/fifo/gv100.o
 
 nvkm-y += nvkm/engine/fifo/chan.o
 nvkm-y += nvkm/engine/fifo/channv50.o
@@ -31,6 +32,6 @@ nvkm-y += nvkm/engine/fifo/gpfifonv50.o
 nvkm-y += nvkm/engine/fifo/gpfifog84.o
 nvkm-y += nvkm/engine/fifo/gpfifogf100.o
 nvkm-y += nvkm/engine/fifo/gpfifogk104.o
-nvkm-y += nvkm/engine/fifo/gpfifogk110.o
-nvkm-y += nvkm/engine/fifo/gpfifogm200.o
-nvkm-y += nvkm/engine/fifo/gpfifogp100.o
+nvkm-y += nvkm/engine/fifo/gpfifogv100.o
+
+nvkm-y += nvkm/engine/fifo/usergv100.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
index 64f6b7654a08..c773caf21f6b 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/base.c
@@ -30,6 +30,7 @@
 #include <subdev/mc.h>
 
 #include <nvif/event.h>
+#include <nvif/cl0080.h>
 #include <nvif/unpack.h>
 
 void
@@ -56,6 +57,12 @@ nvkm_fifo_start(struct nvkm_fifo *fifo, unsigned long *flags)
 }
 
 void
+nvkm_fifo_fault(struct nvkm_fifo *fifo, struct nvkm_fault_data *info)
+{
+	return fifo->func->fault(fifo, info);
+}
+
+void
 nvkm_fifo_chan_put(struct nvkm_fifo *fifo, unsigned long flags,
 		   struct nvkm_fifo_chan **pchan)
 {
@@ -209,6 +216,20 @@ nvkm_fifo_uevent(struct nvkm_fifo *fifo)
 }
 
 static int
+nvkm_fifo_class_new_(struct nvkm_device *device,
+		     const struct nvkm_oclass *oclass, void *data, u32 size,
+		     struct nvkm_object **pobject)
+{
+	struct nvkm_fifo *fifo = nvkm_fifo(oclass->engine);
+	return fifo->func->class_new(fifo, oclass, data, size, pobject);
+}
+
+static const struct nvkm_device_oclass
+nvkm_fifo_class_ = {
+	.ctor = nvkm_fifo_class_new_,
+};
+
+static int
 nvkm_fifo_class_new(struct nvkm_device *device,
 		    const struct nvkm_oclass *oclass, void *data, u32 size,
 		    struct nvkm_object **pobject)
@@ -232,13 +253,9 @@ nvkm_fifo_class_get(struct nvkm_oclass *oclass, int index,
 	int c = 0;
 
 	if (fifo->func->class_get) {
-		int ret = fifo->func->class_get(fifo, index, &sclass);
-		if (ret == 0) {
-			oclass->base = sclass->base;
-			oclass->engn = sclass;
-			*class = &nvkm_fifo_class;
-			return 0;
-		}
+		int ret = fifo->func->class_get(fifo, index, oclass);
+		if (ret == 0)
+			*class = &nvkm_fifo_class_;
 		return ret;
 	}
 
@@ -271,6 +288,20 @@ nvkm_fifo_fini(struct nvkm_engine *engine, bool suspend)
 }
 
 static int
+nvkm_fifo_info(struct nvkm_engine *engine, u64 mthd, u64 *data)
+{
+	struct nvkm_fifo *fifo = nvkm_fifo(engine);
+	switch (mthd) {
+	case NV_DEVICE_FIFO_CHANNELS: *data = fifo->nr; return 0;
+	default:
+		if (fifo->func->info)
+			return fifo->func->info(fifo, mthd, data);
+		break;
+	}
+	return -ENOSYS;
+}
+
+static int
 nvkm_fifo_oneinit(struct nvkm_engine *engine)
 {
 	struct nvkm_fifo *fifo = nvkm_fifo(engine);
@@ -311,6 +342,7 @@ nvkm_fifo = {
 	.dtor = nvkm_fifo_dtor,
 	.preinit = nvkm_fifo_preinit,
 	.oneinit = nvkm_fifo_oneinit,
+	.info = nvkm_fifo_info,
 	.init = nvkm_fifo_init,
 	.fini = nvkm_fifo_fini,
 	.intr = nvkm_fifo_intr,
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/cgrp.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/cgrp.h
new file mode 100644
index 000000000000..d0ac60b06720
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/cgrp.h
@@ -0,0 +1,11 @@
+#ifndef __NVKM_FIFO_CGRP_H__
+#define __NVKM_FIFO_CGRP_H__
+#include "priv.h"
+
+struct nvkm_fifo_cgrp {
+	int id;
+	struct list_head head;
+	struct list_head chan;
+	int chan_nr;
+};
+#endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h
index 1208e3d9dbe2..8e28ba6b2307 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/changk104.h
@@ -10,6 +10,7 @@ struct gk104_fifo_chan {
 	struct gk104_fifo *fifo;
 	int runl;
 
+	struct nvkm_fifo_cgrp *cgrp;
 	struct list_head head;
 	bool killed;
 
@@ -19,11 +20,20 @@ struct gk104_fifo_chan {
 	} engn[NVKM_SUBDEV_NR];
 };
 
-int gk104_fifo_gpfifo_new(struct nvkm_fifo *, const struct nvkm_oclass *,
+extern const struct nvkm_fifo_chan_func gk104_fifo_gpfifo_func;
+
+int gk104_fifo_gpfifo_new(struct gk104_fifo *, const struct nvkm_oclass *,
 			  void *data, u32 size, struct nvkm_object **);
+void *gk104_fifo_gpfifo_dtor(struct nvkm_fifo_chan *);
+void gk104_fifo_gpfifo_init(struct nvkm_fifo_chan *);
+void gk104_fifo_gpfifo_fini(struct nvkm_fifo_chan *);
+int gk104_fifo_gpfifo_engine_ctor(struct nvkm_fifo_chan *, struct nvkm_engine *,
+				  struct nvkm_object *);
+void gk104_fifo_gpfifo_engine_dtor(struct nvkm_fifo_chan *,
+				   struct nvkm_engine *);
+int gk104_fifo_gpfifo_kick(struct gk104_fifo_chan *);
+int gk104_fifo_gpfifo_kick_locked(struct gk104_fifo_chan *);
 
-extern const struct nvkm_fifo_chan_oclass gk104_fifo_gpfifo_oclass;
-extern const struct nvkm_fifo_chan_oclass gk110_fifo_gpfifo_oclass;
-extern const struct nvkm_fifo_chan_oclass gm200_fifo_gpfifo_oclass;
-extern const struct nvkm_fifo_chan_oclass gp100_fifo_gpfifo_oclass;
+int gv100_fifo_gpfifo_new(struct gk104_fifo *, const struct nvkm_oclass *,
+			  void *data, u32 size, struct nvkm_object **);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
index 84bd703dd897..afccf9721cf0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.c
@@ -22,16 +22,19 @@
  * Authors: Ben Skeggs
  */
 #include "gk104.h"
+#include "cgrp.h"
 #include "changk104.h"
 
 #include <core/client.h>
 #include <core/gpuobj.h>
 #include <subdev/bar.h>
+#include <subdev/fault.h>
 #include <subdev/timer.h>
 #include <subdev/top.h>
 #include <engine/sw.h>
 
 #include <nvif/class.h>
+#include <nvif/cl0080.h>
 
 struct gk104_fifo_engine_status {
 	bool busy;
@@ -93,15 +96,39 @@ gk104_fifo_engine_status(struct gk104_fifo *fifo, int engn,
 }
 
 static int
+gk104_fifo_class_new(struct nvkm_fifo *base, const struct nvkm_oclass *oclass,
+		     void *argv, u32 argc, struct nvkm_object **pobject)
+{
+	struct gk104_fifo *fifo = gk104_fifo(base);
+	if (oclass->engn == &fifo->func->chan) {
+		const struct gk104_fifo_chan_user *user = oclass->engn;
+		return user->ctor(fifo, oclass, argv, argc, pobject);
+	} else
+	if (oclass->engn == &fifo->func->user) {
+		const struct gk104_fifo_user_user *user = oclass->engn;
+		return user->ctor(oclass, argv, argc, pobject);
+	}
+	WARN_ON(1);
+	return -EINVAL;
+}
+
+static int
 gk104_fifo_class_get(struct nvkm_fifo *base, int index,
-		     const struct nvkm_fifo_chan_oclass **psclass)
+		     struct nvkm_oclass *oclass)
 {
 	struct gk104_fifo *fifo = gk104_fifo(base);
 	int c = 0;
 
-	while ((*psclass = fifo->func->chan[c])) {
-		if (c++ == index)
-			return 0;
+	if (fifo->func->user.ctor && c++ == index) {
+		oclass->base =  fifo->func->user.user;
+		oclass->engn = &fifo->func->user;
+		return 0;
+	}
+
+	if (fifo->func->chan.ctor && c++ == index) {
+		oclass->base =  fifo->func->chan.user;
+		oclass->engn = &fifo->func->chan;
+		return 0;
 	}
 
 	return c;
@@ -124,10 +151,12 @@ gk104_fifo_uevent_init(struct nvkm_fifo *fifo)
 void
 gk104_fifo_runlist_commit(struct gk104_fifo *fifo, int runl)
 {
+	const struct gk104_fifo_runlist_func *func = fifo->func->runlist;
 	struct gk104_fifo_chan *chan;
 	struct nvkm_subdev *subdev = &fifo->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	struct nvkm_memory *mem;
+	struct nvkm_fifo_cgrp *cgrp;
 	int nr = 0;
 	int target;
 
@@ -137,9 +166,14 @@ gk104_fifo_runlist_commit(struct gk104_fifo *fifo, int runl)
 
 	nvkm_kmap(mem);
 	list_for_each_entry(chan, &fifo->runlist[runl].chan, head) {
-		nvkm_wo32(mem, (nr * 8) + 0, chan->base.chid);
-		nvkm_wo32(mem, (nr * 8) + 4, 0x00000000);
-		nr++;
+		func->chan(chan, mem, nr++ * func->size);
+	}
+
+	list_for_each_entry(cgrp, &fifo->runlist[runl].cgrp, head) {
+		func->cgrp(cgrp, mem, nr++ * func->size);
+		list_for_each_entry(chan, &cgrp->chan, head) {
+			func->chan(chan, mem, nr++ * func->size);
+		}
 	}
 	nvkm_done(mem);
 
@@ -155,10 +189,10 @@ gk104_fifo_runlist_commit(struct gk104_fifo *fifo, int runl)
 				    (target << 28));
 	nvkm_wr32(device, 0x002274, (runl << 20) | nr);
 
-	if (wait_event_timeout(fifo->runlist[runl].wait,
-			       !(nvkm_rd32(device, 0x002284 + (runl * 0x08))
-				       & 0x00100000),
-			       msecs_to_jiffies(2000)) == 0)
+	if (nvkm_msec(device, 2000,
+		if (!(nvkm_rd32(device, 0x002284 + (runl * 0x08)) & 0x00100000))
+			break;
+	) < 0)
 		nvkm_error(subdev, "runlist %d update timeout\n", runl);
 unlock:
 	mutex_unlock(&subdev->mutex);
@@ -167,19 +201,45 @@ unlock:
 void
 gk104_fifo_runlist_remove(struct gk104_fifo *fifo, struct gk104_fifo_chan *chan)
 {
+	struct nvkm_fifo_cgrp *cgrp = chan->cgrp;
 	mutex_lock(&fifo->base.engine.subdev.mutex);
-	list_del_init(&chan->head);
+	if (!list_empty(&chan->head)) {
+		list_del_init(&chan->head);
+		if (cgrp && !--cgrp->chan_nr)
+			list_del_init(&cgrp->head);
+	}
 	mutex_unlock(&fifo->base.engine.subdev.mutex);
 }
 
 void
 gk104_fifo_runlist_insert(struct gk104_fifo *fifo, struct gk104_fifo_chan *chan)
 {
+	struct nvkm_fifo_cgrp *cgrp = chan->cgrp;
 	mutex_lock(&fifo->base.engine.subdev.mutex);
-	list_add_tail(&chan->head, &fifo->runlist[chan->runl].chan);
+	if (cgrp) {
+		if (!cgrp->chan_nr++)
+			list_add_tail(&cgrp->head, &fifo->runlist[chan->runl].cgrp);
+		list_add_tail(&chan->head, &cgrp->chan);
+	} else {
+		list_add_tail(&chan->head, &fifo->runlist[chan->runl].chan);
+	}
 	mutex_unlock(&fifo->base.engine.subdev.mutex);
 }
 
+void
+gk104_fifo_runlist_chan(struct gk104_fifo_chan *chan,
+			struct nvkm_memory *memory, u32 offset)
+{
+	nvkm_wo32(memory, offset + 0, chan->base.chid);
+	nvkm_wo32(memory, offset + 4, 0x00000000);
+}
+
+const struct gk104_fifo_runlist_func
+gk104_fifo_runlist = {
+	.size = 8,
+	.chan = gk104_fifo_runlist_chan,
+};
+
 static void
 gk104_fifo_recover_work(struct work_struct *w)
 {
@@ -235,6 +295,32 @@ gk104_fifo_recover_runl(struct gk104_fifo *fifo, int runl)
 	schedule_work(&fifo->recover.work);
 }
 
+static struct gk104_fifo_chan *
+gk104_fifo_recover_chid(struct gk104_fifo *fifo, int runl, int chid)
+{
+	struct gk104_fifo_chan *chan;
+	struct nvkm_fifo_cgrp *cgrp;
+
+	list_for_each_entry(chan, &fifo->runlist[runl].chan, head) {
+		if (chan->base.chid == chid) {
+			list_del_init(&chan->head);
+			return chan;
+		}
+	}
+
+	list_for_each_entry(cgrp, &fifo->runlist[runl].cgrp, head) {
+		if (cgrp->id == chid) {
+			chan = list_first_entry(&cgrp->chan, typeof(*chan), head);
+			list_del_init(&chan->head);
+			if (!--cgrp->chan_nr)
+				list_del_init(&cgrp->head);
+			return chan;
+		}
+	}
+
+	return NULL;
+}
+
 static void
 gk104_fifo_recover_chan(struct nvkm_fifo *base, int chid)
 {
@@ -252,13 +338,10 @@ gk104_fifo_recover_chan(struct nvkm_fifo *base, int chid)
 		return;
 
 	/* Lookup SW state for channel, and mark it as dead. */
-	list_for_each_entry(chan, &fifo->runlist[runl].chan, head) {
-		if (chan->base.chid == chid) {
-			list_del_init(&chan->head);
-			chan->killed = true;
-			nvkm_fifo_kevent(&fifo->base, chid);
-			break;
-		}
+	chan = gk104_fifo_recover_chid(fifo, runl, chid);
+	if (chan) {
+		chan->killed = true;
+		nvkm_fifo_kevent(&fifo->base, chid);
 	}
 
 	/* Disable channel. */
@@ -347,6 +430,90 @@ gk104_fifo_recover_engn(struct gk104_fifo *fifo, int engn)
 	schedule_work(&fifo->recover.work);
 }
 
+static void
+gk104_fifo_fault(struct nvkm_fifo *base, struct nvkm_fault_data *info)
+{
+	struct gk104_fifo *fifo = gk104_fifo(base);
+	struct nvkm_subdev *subdev = &fifo->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	const struct nvkm_enum *er, *ee, *ec, *ea;
+	struct nvkm_engine *engine = NULL;
+	struct nvkm_fifo_chan *chan;
+	unsigned long flags;
+	char ct[8] = "HUB/", en[16] = "";
+	int engn;
+
+	er = nvkm_enum_find(fifo->func->fault.reason, info->reason);
+	ee = nvkm_enum_find(fifo->func->fault.engine, info->engine);
+	if (info->hub) {
+		ec = nvkm_enum_find(fifo->func->fault.hubclient, info->client);
+	} else {
+		ec = nvkm_enum_find(fifo->func->fault.gpcclient, info->client);
+		snprintf(ct, sizeof(ct), "GPC%d/", info->gpc);
+	}
+	ea = nvkm_enum_find(fifo->func->fault.access, info->access);
+
+	if (ee && ee->data2) {
+		switch (ee->data2) {
+		case NVKM_SUBDEV_BAR:
+			nvkm_mask(device, 0x001704, 0x00000000, 0x00000000);
+			break;
+		case NVKM_SUBDEV_INSTMEM:
+			nvkm_mask(device, 0x001714, 0x00000000, 0x00000000);
+			break;
+		case NVKM_ENGINE_IFB:
+			nvkm_mask(device, 0x001718, 0x00000000, 0x00000000);
+			break;
+		default:
+			engine = nvkm_device_engine(device, ee->data2);
+			break;
+		}
+	}
+
+	if (ee == NULL) {
+		enum nvkm_devidx engidx = nvkm_top_fault(device, info->engine);
+		if (engidx < NVKM_SUBDEV_NR) {
+			const char *src = nvkm_subdev_name[engidx];
+			char *dst = en;
+			do {
+				*dst++ = toupper(*src++);
+			} while(*src);
+			engine = nvkm_device_engine(device, engidx);
+		}
+	} else {
+		snprintf(en, sizeof(en), "%s", ee->name);
+	}
+
+	spin_lock_irqsave(&fifo->base.lock, flags);
+	chan = nvkm_fifo_chan_inst_locked(&fifo->base, info->inst);
+
+	nvkm_error(subdev,
+		   "fault %02x [%s] at %016llx engine %02x [%s] client %02x "
+		   "[%s%s] reason %02x [%s] on channel %d [%010llx %s]\n",
+		   info->access, ea ? ea->name : "", info->addr,
+		   info->engine, ee ? ee->name : en,
+		   info->client, ct, ec ? ec->name : "",
+		   info->reason, er ? er->name : "", chan ? chan->chid : -1,
+		   info->inst, chan ? chan->object.client->name : "unknown");
+
+	/* Kill the channel that caused the fault. */
+	if (chan)
+		gk104_fifo_recover_chan(&fifo->base, chan->chid);
+
+	/* Channel recovery will probably have already done this for the
+	 * correct engine(s), but just in case we can't find the channel
+	 * information...
+	 */
+	for (engn = 0; engn < fifo->engine_nr && engine; engn++) {
+		if (fifo->engine[engn].engine == engine) {
+			gk104_fifo_recover_engn(fifo, engn);
+			break;
+		}
+	}
+
+	spin_unlock_irqrestore(&fifo->base.lock, flags);
+}
+
 static const struct nvkm_enum
 gk104_fifo_bind_reason[] = {
 	{ 0x01, "BIND_NOT_UNBOUND" },
@@ -456,88 +623,21 @@ gk104_fifo_intr_fault(struct gk104_fifo *fifo, int unit)
 	u32 inst = nvkm_rd32(device, 0x002800 + (unit * 0x10));
 	u32 valo = nvkm_rd32(device, 0x002804 + (unit * 0x10));
 	u32 vahi = nvkm_rd32(device, 0x002808 + (unit * 0x10));
-	u32 stat = nvkm_rd32(device, 0x00280c + (unit * 0x10));
-	u32 gpc    = (stat & 0x1f000000) >> 24;
-	u32 client = (stat & 0x00001f00) >> 8;
-	u32 write  = (stat & 0x00000080);
-	u32 hub    = (stat & 0x00000040);
-	u32 reason = (stat & 0x0000000f);
-	const struct nvkm_enum *er, *eu, *ec;
-	struct nvkm_engine *engine = NULL;
-	struct nvkm_fifo_chan *chan;
-	unsigned long flags;
-	char gpcid[8] = "", en[16] = "";
-	int engn;
-
-	er = nvkm_enum_find(fifo->func->fault.reason, reason);
-	eu = nvkm_enum_find(fifo->func->fault.engine, unit);
-	if (hub) {
-		ec = nvkm_enum_find(fifo->func->fault.hubclient, client);
-	} else {
-		ec = nvkm_enum_find(fifo->func->fault.gpcclient, client);
-		snprintf(gpcid, sizeof(gpcid), "GPC%d/", gpc);
-	}
-
-	if (eu && eu->data2) {
-		switch (eu->data2) {
-		case NVKM_SUBDEV_BAR:
-			nvkm_mask(device, 0x001704, 0x00000000, 0x00000000);
-			break;
-		case NVKM_SUBDEV_INSTMEM:
-			nvkm_mask(device, 0x001714, 0x00000000, 0x00000000);
-			break;
-		case NVKM_ENGINE_IFB:
-			nvkm_mask(device, 0x001718, 0x00000000, 0x00000000);
-			break;
-		default:
-			engine = nvkm_device_engine(device, eu->data2);
-			break;
-		}
-	}
-
-	if (eu == NULL) {
-		enum nvkm_devidx engidx = nvkm_top_fault(device, unit);
-		if (engidx < NVKM_SUBDEV_NR) {
-			const char *src = nvkm_subdev_name[engidx];
-			char *dst = en;
-			do {
-				*dst++ = toupper(*src++);
-			} while(*src);
-			engine = nvkm_device_engine(device, engidx);
-		}
-	} else {
-		snprintf(en, sizeof(en), "%s", eu->name);
-	}
-
-	spin_lock_irqsave(&fifo->base.lock, flags);
-	chan = nvkm_fifo_chan_inst_locked(&fifo->base, (u64)inst << 12);
-
-	nvkm_error(subdev,
-		   "%s fault at %010llx engine %02x [%s] client %02x [%s%s] "
-		   "reason %02x [%s] on channel %d [%010llx %s]\n",
-		   write ? "write" : "read", (u64)vahi << 32 | valo,
-		   unit, en, client, gpcid, ec ? ec->name : "",
-		   reason, er ? er->name : "", chan ? chan->chid : -1,
-		   (u64)inst << 12,
-		   chan ? chan->object.client->name : "unknown");
-
-
-	/* Kill the channel that caused the fault. */
-	if (chan)
-		gk104_fifo_recover_chan(&fifo->base, chan->chid);
-
-	/* Channel recovery will probably have already done this for the
-	 * correct engine(s), but just in case we can't find the channel
-	 * information...
-	 */
-	for (engn = 0; engn < fifo->engine_nr && engine; engn++) {
-		if (fifo->engine[engn].engine == engine) {
-			gk104_fifo_recover_engn(fifo, engn);
-			break;
-		}
-	}
-
-	spin_unlock_irqrestore(&fifo->base.lock, flags);
+	u32 type = nvkm_rd32(device, 0x00280c + (unit * 0x10));
+	struct nvkm_fault_data info;
+
+	info.inst   =  (u64)inst << 12;
+	info.addr   = ((u64)vahi << 32) | valo;
+	info.time   = 0;
+	info.engine = unit;
+	info.valid  = 1;
+	info.gpc    = (type & 0x1f000000) >> 24;
+	info.client = (type & 0x00001f00) >> 8;
+	info.access = (type & 0x00000080) >> 7;
+	info.hub    = (type & 0x00000040) >> 6;
+	info.reason = (type & 0x000000ff);
+
+	nvkm_fifo_fault(&fifo->base, &info);
 }
 
 static const struct nvkm_bitfield gk104_fifo_pbdma_intr_0[] = {
@@ -766,6 +866,34 @@ gk104_fifo_fini(struct nvkm_fifo *base)
 }
 
 static int
+gk104_fifo_info(struct nvkm_fifo *base, u64 mthd, u64 *data)
+{
+	struct gk104_fifo *fifo = gk104_fifo(base);
+	switch (mthd) {
+	case NV_DEVICE_FIFO_RUNLISTS:
+		*data = (1ULL << fifo->runlist_nr) - 1;
+		return 0;
+	case NV_DEVICE_FIFO_RUNLIST_ENGINES(0)...
+	     NV_DEVICE_FIFO_RUNLIST_ENGINES(63): {
+		int runl = mthd - NV_DEVICE_FIFO_RUNLIST_ENGINES(0), engn;
+		if (runl < fifo->runlist_nr) {
+			unsigned long engm = fifo->runlist[runl].engm;
+			struct nvkm_engine *engine;
+			*data = 0;
+			for_each_set_bit(engn, &engm, fifo->engine_nr) {
+				if ((engine = fifo->engine[engn].engine))
+					*data |= BIT_ULL(engine->subdev.index);
+			}
+			return 0;
+		}
+	}
+		return -EINVAL;
+	default:
+		return -EINVAL;
+	}
+}
+
+static int
 gk104_fifo_oneinit(struct nvkm_fifo *base)
 {
 	struct gk104_fifo *fifo = gk104_fifo(base);
@@ -782,7 +910,7 @@ gk104_fifo_oneinit(struct nvkm_fifo *base)
 	nvkm_debug(subdev, "%d PBDMA(s)\n", fifo->pbdma_nr);
 
 	/* Read PBDMA->runlist(s) mapping from HW. */
-	if (!(map = kzalloc(sizeof(*map) * fifo->pbdma_nr, GFP_KERNEL)))
+	if (!(map = kcalloc(fifo->pbdma_nr, sizeof(*map), GFP_KERNEL)))
 		return -ENOMEM;
 
 	for (i = 0; i < fifo->pbdma_nr; i++)
@@ -813,19 +941,18 @@ gk104_fifo_oneinit(struct nvkm_fifo *base)
 	kfree(map);
 
 	for (i = 0; i < fifo->runlist_nr; i++) {
-		ret = nvkm_memory_new(device, NVKM_MEM_TARGET_INST,
-				      0x8000, 0x1000, false,
-				      &fifo->runlist[i].mem[0]);
-		if (ret)
-			return ret;
-
-		ret = nvkm_memory_new(device, NVKM_MEM_TARGET_INST,
-				      0x8000, 0x1000, false,
-				      &fifo->runlist[i].mem[1]);
-		if (ret)
-			return ret;
+		for (j = 0; j < ARRAY_SIZE(fifo->runlist[i].mem); j++) {
+			ret = nvkm_memory_new(device, NVKM_MEM_TARGET_INST,
+					      fifo->base.nr * 2/* TSG+chan */ *
+					      fifo->func->runlist->size,
+					      0x1000, false,
+					      &fifo->runlist[i].mem[j]);
+			if (ret)
+				return ret;
+		}
 
 		init_waitqueue_head(&fifo->runlist[i].wait);
+		INIT_LIST_HEAD(&fifo->runlist[i].cgrp);
 		INIT_LIST_HEAD(&fifo->runlist[i].chan);
 	}
 
@@ -868,6 +995,9 @@ gk104_fifo_init(struct nvkm_fifo *base)
 
 	nvkm_wr32(device, 0x002254, 0x10000000 | fifo->user.bar->addr >> 12);
 
+	if (fifo->func->init_pbdma_timeout)
+		fifo->func->init_pbdma_timeout(fifo);
+
 	nvkm_wr32(device, 0x002100, 0xffffffff);
 	nvkm_wr32(device, 0x002140, 0x7fffffff);
 }
@@ -894,13 +1024,16 @@ static const struct nvkm_fifo_func
 gk104_fifo_ = {
 	.dtor = gk104_fifo_dtor,
 	.oneinit = gk104_fifo_oneinit,
+	.info = gk104_fifo_info,
 	.init = gk104_fifo_init,
 	.fini = gk104_fifo_fini,
 	.intr = gk104_fifo_intr,
+	.fault = gk104_fifo_fault,
 	.uevent_init = gk104_fifo_uevent_init,
 	.uevent_fini = gk104_fifo_uevent_fini,
 	.recover_chan = gk104_fifo_recover_chan,
 	.class_get = gk104_fifo_class_get,
+	.class_new = gk104_fifo_class_new,
 };
 
 int
@@ -919,6 +1052,13 @@ gk104_fifo_new_(const struct gk104_fifo_func *func, struct nvkm_device *device,
 }
 
 const struct nvkm_enum
+gk104_fifo_fault_access[] = {
+	{ 0x0, "READ" },
+	{ 0x1, "WRITE" },
+	{}
+};
+
+const struct nvkm_enum
 gk104_fifo_fault_engine[] = {
 	{ 0x00, "GR", NULL, NVKM_ENGINE_GR },
 	{ 0x01, "DISPLAY" },
@@ -1035,14 +1175,13 @@ gk104_fifo_fault_gpcclient[] = {
 
 static const struct gk104_fifo_func
 gk104_fifo = {
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gk104_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gk104_fifo_gpfifo_oclass,
-		NULL
-	},
+	.runlist = &gk104_fifo_runlist,
+	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
index 1579785cf941..d295b81e18d6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk104.h
@@ -3,6 +3,7 @@
 #define __GK104_FIFO_H__
 #define gk104_fifo(p) container_of((p), struct gk104_fifo, base)
 #include "priv.h"
+struct nvkm_fifo_cgrp;
 
 #include <core/enum.h>
 #include <subdev/mmu.h>
@@ -31,6 +32,7 @@ struct gk104_fifo {
 		struct nvkm_memory *mem[2];
 		int next;
 		wait_queue_head_t wait;
+		struct list_head cgrp;
 		struct list_head chan;
 		u32 engm;
 	} runlist[16];
@@ -43,14 +45,36 @@ struct gk104_fifo {
 };
 
 struct gk104_fifo_func {
+	void (*init_pbdma_timeout)(struct gk104_fifo *);
+
 	struct {
+		const struct nvkm_enum *access;
 		const struct nvkm_enum *engine;
 		const struct nvkm_enum *reason;
 		const struct nvkm_enum *hubclient;
 		const struct nvkm_enum *gpcclient;
 	} fault;
 
-	const struct nvkm_fifo_chan_oclass *chan[];
+	const struct gk104_fifo_runlist_func {
+		u8 size;
+		void (*cgrp)(struct nvkm_fifo_cgrp *,
+			     struct nvkm_memory *, u32 offset);
+		void (*chan)(struct gk104_fifo_chan *,
+			     struct nvkm_memory *, u32 offset);
+	} *runlist;
+
+	struct gk104_fifo_user_user {
+		struct nvkm_sclass user;
+		int (*ctor)(const struct nvkm_oclass *, void *, u32,
+			    struct nvkm_object **);
+	} user;
+
+	struct gk104_fifo_chan_user {
+		struct nvkm_sclass user;
+		int (*ctor)(struct gk104_fifo *, const struct nvkm_oclass *,
+			    void *, u32, struct nvkm_object **);
+	} chan;
+	bool cgrp_force;
 };
 
 int gk104_fifo_new_(const struct gk104_fifo_func *, struct nvkm_device *,
@@ -59,30 +83,23 @@ void gk104_fifo_runlist_insert(struct gk104_fifo *, struct gk104_fifo_chan *);
 void gk104_fifo_runlist_remove(struct gk104_fifo *, struct gk104_fifo_chan *);
 void gk104_fifo_runlist_commit(struct gk104_fifo *, int runl);
 
-static inline u64
-gk104_fifo_engine_subdev(int engine)
-{
-	switch (engine) {
-	case 0: return (1ULL << NVKM_ENGINE_GR) |
-		       (1ULL << NVKM_ENGINE_SW) |
-		       (1ULL << NVKM_ENGINE_CE2);
-	case 1: return (1ULL << NVKM_ENGINE_MSPDEC);
-	case 2: return (1ULL << NVKM_ENGINE_MSPPP);
-	case 3: return (1ULL << NVKM_ENGINE_MSVLD);
-	case 4: return (1ULL << NVKM_ENGINE_CE0);
-	case 5: return (1ULL << NVKM_ENGINE_CE1);
-	case 6: return (1ULL << NVKM_ENGINE_MSENC);
-	default:
-		WARN_ON(1);
-		return 0;
-	}
-}
-
+extern const struct nvkm_enum gk104_fifo_fault_access[];
 extern const struct nvkm_enum gk104_fifo_fault_engine[];
 extern const struct nvkm_enum gk104_fifo_fault_reason[];
 extern const struct nvkm_enum gk104_fifo_fault_hubclient[];
 extern const struct nvkm_enum gk104_fifo_fault_gpcclient[];
+extern const struct gk104_fifo_runlist_func gk104_fifo_runlist;
+void gk104_fifo_runlist_chan(struct gk104_fifo_chan *,
+			     struct nvkm_memory *, u32);
+
+extern const struct gk104_fifo_runlist_func gk110_fifo_runlist;
+void gk110_fifo_runlist_cgrp(struct nvkm_fifo_cgrp *,
+			     struct nvkm_memory *, u32);
+
+void gk208_fifo_init_pbdma_timeout(struct gk104_fifo *);
 
 extern const struct nvkm_enum gm107_fifo_fault_engine[];
+extern const struct gk104_fifo_runlist_func gm107_fifo_runlist;
+
 extern const struct nvkm_enum gp100_fifo_fault_engine[];
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
index b2f8ab7bf847..ac7655a130fb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk110.c
@@ -22,18 +22,38 @@
  * Authors: Ben Skeggs
  */
 #include "gk104.h"
+#include "cgrp.h"
 #include "changk104.h"
 
+#include <core/memory.h>
+
+#include <nvif/class.h>
+
+void
+gk110_fifo_runlist_cgrp(struct nvkm_fifo_cgrp *cgrp,
+			struct nvkm_memory *memory, u32 offset)
+{
+	nvkm_wo32(memory, offset + 0, (cgrp->chan_nr << 26) | (128 << 18) |
+				      (3 << 14) | 0x00002000 | cgrp->id);
+	nvkm_wo32(memory, offset + 4, 0x00000000);
+}
+
+const struct gk104_fifo_runlist_func
+gk110_fifo_runlist = {
+	.size = 8,
+	.cgrp = gk110_fifo_runlist_cgrp,
+	.chan = gk104_fifo_runlist_chan,
+};
+
 static const struct gk104_fifo_func
 gk110_fifo = {
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gk104_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gk110_fifo_gpfifo_oclass,
-		NULL
-	},
+	.runlist = &gk110_fifo_runlist,
+	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_B}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
index 160617d376e4..5ea7e452cc66 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk208.c
@@ -24,16 +24,28 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <nvif/class.h>
+
+void
+gk208_fifo_init_pbdma_timeout(struct gk104_fifo *fifo)
+{
+	struct nvkm_device *device = fifo->base.engine.subdev.device;
+	int i;
+
+	for (i = 0; i < fifo->pbdma_nr; i++)
+		nvkm_wr32(device, 0x04012c + (i * 0x2000), 0x0000ffff);
+}
+
 static const struct gk104_fifo_func
 gk208_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gk104_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gk104_fifo_gpfifo_oclass,
-		NULL
-	},
+	.runlist = &gk110_fifo_runlist,
+	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
index be9f5c16ed7d..535a0eb67a5f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gk20a.c
@@ -22,16 +22,18 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <nvif/class.h>
+
 static const struct gk104_fifo_func
 gk20a_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gk104_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gk104_fifo_gpfifo_oclass,
-		NULL
-	},
+	.runlist = &gk110_fifo_runlist,
+	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
index 29c080683b32..79ae19b1db67 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm107.c
@@ -24,6 +24,25 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <core/gpuobj.h>
+
+#include <nvif/class.h>
+
+static void
+gm107_fifo_runlist_chan(struct gk104_fifo_chan *chan,
+			struct nvkm_memory *memory, u32 offset)
+{
+	nvkm_wo32(memory, offset + 0, chan->base.chid);
+	nvkm_wo32(memory, offset + 4, chan->base.inst->addr >> 12);
+}
+
+const struct gk104_fifo_runlist_func
+gm107_fifo_runlist = {
+	.size = 8,
+	.cgrp = gk110_fifo_runlist_cgrp,
+	.chan = gm107_fifo_runlist_chan,
+};
+
 const struct nvkm_enum
 gm107_fifo_fault_engine[] = {
 	{ 0x01, "DISPLAY" },
@@ -49,14 +68,14 @@ gm107_fifo_fault_engine[] = {
 
 static const struct gk104_fifo_func
 gm107_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gm107_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gk110_fifo_gpfifo_oclass,
-		NULL
-	},
+	.runlist = &gm107_fifo_runlist,
+	.chan = {{0,0,KEPLER_CHANNEL_GPFIFO_B}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
index b069f785c5d8..49565faa854d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm200.c
@@ -24,16 +24,18 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <nvif/class.h>
+
 static const struct gk104_fifo_func
 gm200_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gm107_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gm200_fifo_gpfifo_oclass,
-		NULL
-	},
+	.runlist = &gm107_fifo_runlist,
+	.chan = {{0,0,MAXWELL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
index 2ed87c2e8299..46736513bd11 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gm20b.c
@@ -22,16 +22,18 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <nvif/class.h>
+
 static const struct gk104_fifo_func
 gm20b_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gm107_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gm200_fifo_gpfifo_oclass,
-		NULL
-	},
+	.runlist = &gm107_fifo_runlist,
+	.chan = {{0,0,MAXWELL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
index 41f16cf5a918..e2f8f9087d7c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp100.c
@@ -24,6 +24,8 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <nvif/class.h>
+
 const struct nvkm_enum
 gp100_fifo_fault_engine[] = {
 	{ 0x01, "DISPLAY" },
@@ -50,14 +52,15 @@ gp100_fifo_fault_engine[] = {
 
 static const struct gk104_fifo_func
 gp100_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gp100_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gp100_fifo_gpfifo_oclass,
-		NULL
-	},
+	.runlist = &gm107_fifo_runlist,
+	.chan = {{0,0,PASCAL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
+	.cgrp_force = true,
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
index 4af96c3e69ff..7733bf7c6545 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gp10b.c
@@ -22,16 +22,19 @@
 #include "gk104.h"
 #include "changk104.h"
 
+#include <nvif/class.h>
+
 static const struct gk104_fifo_func
 gp10b_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
+	.fault.access = gk104_fifo_fault_access,
 	.fault.engine = gp100_fifo_fault_engine,
 	.fault.reason = gk104_fifo_fault_reason,
 	.fault.hubclient = gk104_fifo_fault_hubclient,
 	.fault.gpcclient = gk104_fifo_fault_gpcclient,
-	.chan = {
-		&gp100_fifo_gpfifo_oclass,
-		NULL
-	},
+	.runlist = &gm107_fifo_runlist,
+	.chan = {{0,0,PASCAL_CHANNEL_GPFIFO_A}, gk104_fifo_gpfifo_new },
+	.cgrp_force = true,
 };
 
 int
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
index 80c87521bebe..118b37aea318 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogk104.c
@@ -22,6 +22,7 @@
  * Authors: Ben Skeggs
  */
 #include "changk104.h"
+#include "cgrp.h"
 
 #include <core/client.h>
 #include <core/gpuobj.h>
@@ -33,27 +34,40 @@
 #include <nvif/cla06f.h>
 #include <nvif/unpack.h>
 
-static int
-gk104_fifo_gpfifo_kick(struct gk104_fifo_chan *chan)
+int
+gk104_fifo_gpfifo_kick_locked(struct gk104_fifo_chan *chan)
 {
 	struct gk104_fifo *fifo = chan->fifo;
 	struct nvkm_subdev *subdev = &fifo->base.engine.subdev;
 	struct nvkm_device *device = subdev->device;
 	struct nvkm_client *client = chan->base.object.client;
+	struct nvkm_fifo_cgrp *cgrp = chan->cgrp;
 	int ret = 0;
 
-	mutex_lock(&subdev->mutex);
-	nvkm_wr32(device, 0x002634, chan->base.chid);
+	if (cgrp)
+		nvkm_wr32(device, 0x002634, cgrp->id | 0x01000000);
+	else
+		nvkm_wr32(device, 0x002634, chan->base.chid);
 	if (nvkm_msec(device, 2000,
 		if (!(nvkm_rd32(device, 0x002634) & 0x00100000))
 			break;
 	) < 0) {
-		nvkm_error(subdev, "channel %d [%s] kick timeout\n",
-			   chan->base.chid, client->name);
+		nvkm_error(subdev, "%s %d [%s] kick timeout\n",
+			   cgrp ? "tsg" : "channel",
+			   cgrp ? cgrp->id : chan->base.chid, client->name);
 		nvkm_fifo_recover_chan(&fifo->base, chan->base.chid);
 		ret = -ETIMEDOUT;
 	}
-	mutex_unlock(&subdev->mutex);
+	return ret;
+}
+
+int
+gk104_fifo_gpfifo_kick(struct gk104_fifo_chan *chan)
+{
+	int ret;
+	mutex_lock(&chan->base.fifo->engine.subdev.mutex);
+	ret = gk104_fifo_gpfifo_kick_locked(chan);
+	mutex_unlock(&chan->base.fifo->engine.subdev.mutex);
 	return ret;
 }
 
@@ -62,9 +76,8 @@ gk104_fifo_gpfifo_engine_addr(struct nvkm_engine *engine)
 {
 	switch (engine->subdev.index) {
 	case NVKM_ENGINE_SW    :
-	case NVKM_ENGINE_CE0   :
-	case NVKM_ENGINE_CE1   :
-	case NVKM_ENGINE_CE2   : return 0x0000;
+	case NVKM_ENGINE_CE0...NVKM_ENGINE_CE_LAST:
+		return 0;
 	case NVKM_ENGINE_GR    : return 0x0210;
 	case NVKM_ENGINE_SEC   : return 0x0220;
 	case NVKM_ENGINE_MSPDEC: return 0x0250;
@@ -133,7 +146,7 @@ gk104_fifo_gpfifo_engine_init(struct nvkm_fifo_chan *base,
 	return 0;
 }
 
-static void
+void
 gk104_fifo_gpfifo_engine_dtor(struct nvkm_fifo_chan *base,
 			      struct nvkm_engine *engine)
 {
@@ -142,7 +155,7 @@ gk104_fifo_gpfifo_engine_dtor(struct nvkm_fifo_chan *base,
 	nvkm_gpuobj_del(&chan->engn[engine->subdev.index].inst);
 }
 
-static int
+int
 gk104_fifo_gpfifo_engine_ctor(struct nvkm_fifo_chan *base,
 			      struct nvkm_engine *engine,
 			      struct nvkm_object *object)
@@ -167,7 +180,7 @@ gk104_fifo_gpfifo_engine_ctor(struct nvkm_fifo_chan *base,
 			       chan->engn[engn].vma, NULL, 0);
 }
 
-static void
+void
 gk104_fifo_gpfifo_fini(struct nvkm_fifo_chan *base)
 {
 	struct gk104_fifo_chan *chan = gk104_fifo_chan(base);
@@ -185,7 +198,7 @@ gk104_fifo_gpfifo_fini(struct nvkm_fifo_chan *base)
 	nvkm_wr32(device, 0x800000 + coff, 0x00000000);
 }
 
-static void
+void
 gk104_fifo_gpfifo_init(struct nvkm_fifo_chan *base)
 {
 	struct gk104_fifo_chan *chan = gk104_fifo_chan(base);
@@ -205,13 +218,15 @@ gk104_fifo_gpfifo_init(struct nvkm_fifo_chan *base)
 	}
 }
 
-static void *
+void *
 gk104_fifo_gpfifo_dtor(struct nvkm_fifo_chan *base)
 {
-	return gk104_fifo_chan(base);
+	struct gk104_fifo_chan *chan = gk104_fifo_chan(base);
+	kfree(chan->cgrp);
+	return chan;
 }
 
-static const struct nvkm_fifo_chan_func
+const struct nvkm_fifo_chan_func
 gk104_fifo_gpfifo_func = {
 	.dtor = gk104_fifo_gpfifo_dtor,
 	.init = gk104_fifo_gpfifo_init,
@@ -223,62 +238,30 @@ gk104_fifo_gpfifo_func = {
 	.engine_fini = gk104_fifo_gpfifo_engine_fini,
 };
 
-struct gk104_fifo_chan_func {
-	u32 engine;
-	u64 subdev;
-};
-
 static int
-gk104_fifo_gpfifo_new_(const struct gk104_fifo_chan_func *func,
-		       struct gk104_fifo *fifo, u32 *engmask, u16 *chid,
+gk104_fifo_gpfifo_new_(struct gk104_fifo *fifo, u64 *runlists, u16 *chid,
 		       u64 vmm, u64 ioffset, u64 ilength,
 		       const struct nvkm_oclass *oclass,
 		       struct nvkm_object **pobject)
 {
 	struct gk104_fifo_chan *chan;
-	int runlist = -1, ret = -ENOSYS, i, j;
-	u32 engines = 0, present = 0;
+	int runlist = ffs(*runlists) -1, ret, i;
+	unsigned long engm;
 	u64 subdevs = 0;
 	u64 usermem;
 
-	if (!vmm)
+	if (!vmm || runlist < 0 || runlist >= fifo->runlist_nr)
 		return -EINVAL;
+	*runlists = BIT_ULL(runlist);
 
-	/* Determine which downstream engines are present */
-	for (i = 0; i < fifo->engine_nr; i++) {
-		struct nvkm_engine *engine = fifo->engine[i].engine;
-		if (engine) {
-			u64 submask = BIT_ULL(engine->subdev.index);
-			for (j = 0; func[j].subdev; j++) {
-				if (func[j].subdev & submask) {
-					present |= func[j].engine;
-					break;
-				}
-			}
-
-			if (!func[j].subdev)
-				continue;
-
-			if (runlist < 0 && (*engmask & present))
-				runlist = fifo->engine[i].runl;
-			if (runlist == fifo->engine[i].runl) {
-				engines |= func[j].engine;
-				subdevs |= func[j].subdev;
-			}
-		}
-	}
-
-	/* Just an engine mask query?  All done here! */
-	if (!*engmask) {
-		*engmask = present;
-		return nvkm_object_new(oclass, NULL, 0, pobject);
+	engm = fifo->runlist[runlist].engm;
+	for_each_set_bit(i, &engm, fifo->engine_nr) {
+		if (fifo->engine[i].engine)
+			subdevs |= BIT_ULL(fifo->engine[i].engine->subdev.index);
 	}
 
-	/* No runlist?  No supported engines. */
-	*engmask = present;
-	if (runlist < 0)
-		return -ENODEV;
-	*engmask = engines;
+	if (subdevs & BIT_ULL(NVKM_ENGINE_GR))
+		subdevs |= BIT_ULL(NVKM_ENGINE_SW);
 
 	/* Allocate the channel. */
 	if (!(chan = kzalloc(sizeof(*chan), GFP_KERNEL)))
@@ -297,6 +280,18 @@ gk104_fifo_gpfifo_new_(const struct gk104_fifo_chan_func *func,
 
 	*chid = chan->base.chid;
 
+	/* Hack to support GPUs where even individual channels should be
+	 * part of a channel group.
+	 */
+	if (fifo->func->cgrp_force) {
+		if (!(chan->cgrp = kmalloc(sizeof(*chan->cgrp), GFP_KERNEL)))
+			return -ENOMEM;
+		chan->cgrp->id = chan->base.chid;
+		INIT_LIST_HEAD(&chan->cgrp->head);
+		INIT_LIST_HEAD(&chan->cgrp->chan);
+		chan->cgrp->chan_nr = 0;
+	}
+
 	/* Clear channel control registers. */
 	usermem = chan->base.chid * 0x200;
 	ilength = order_base_2(ilength / 8);
@@ -328,45 +323,25 @@ gk104_fifo_gpfifo_new_(const struct gk104_fifo_chan_func *func,
 	return 0;
 }
 
-static const struct gk104_fifo_chan_func
-gk104_fifo_gpfifo[] = {
-	{ NVA06F_V0_ENGINE_SW | NVA06F_V0_ENGINE_GR,
-		BIT_ULL(NVKM_ENGINE_SW) | BIT_ULL(NVKM_ENGINE_GR)
-	},
-	{ NVA06F_V0_ENGINE_SEC   , BIT_ULL(NVKM_ENGINE_SEC   ) },
-	{ NVA06F_V0_ENGINE_MSVLD , BIT_ULL(NVKM_ENGINE_MSVLD ) },
-	{ NVA06F_V0_ENGINE_MSPDEC, BIT_ULL(NVKM_ENGINE_MSPDEC) },
-	{ NVA06F_V0_ENGINE_MSPPP , BIT_ULL(NVKM_ENGINE_MSPPP ) },
-	{ NVA06F_V0_ENGINE_MSENC , BIT_ULL(NVKM_ENGINE_MSENC ) },
-	{ NVA06F_V0_ENGINE_VIC   , BIT_ULL(NVKM_ENGINE_VIC   ) },
-	{ NVA06F_V0_ENGINE_NVDEC , BIT_ULL(NVKM_ENGINE_NVDEC ) },
-	{ NVA06F_V0_ENGINE_NVENC0, BIT_ULL(NVKM_ENGINE_NVENC0) },
-	{ NVA06F_V0_ENGINE_NVENC1, BIT_ULL(NVKM_ENGINE_NVENC1) },
-	{ NVA06F_V0_ENGINE_CE0   , BIT_ULL(NVKM_ENGINE_CE0   ) },
-	{ NVA06F_V0_ENGINE_CE1   , BIT_ULL(NVKM_ENGINE_CE1   ) },
-	{ NVA06F_V0_ENGINE_CE2   , BIT_ULL(NVKM_ENGINE_CE2   ) },
-	{}
-};
-
 int
-gk104_fifo_gpfifo_new(struct nvkm_fifo *base, const struct nvkm_oclass *oclass,
+gk104_fifo_gpfifo_new(struct gk104_fifo *fifo, const struct nvkm_oclass *oclass,
 		      void *data, u32 size, struct nvkm_object **pobject)
 {
 	struct nvkm_object *parent = oclass->parent;
 	union {
 		struct kepler_channel_gpfifo_a_v0 v0;
 	} *args = data;
-	struct gk104_fifo *fifo = gk104_fifo(base);
 	int ret = -ENOSYS;
 
 	nvif_ioctl(parent, "create channel gpfifo size %d\n", size);
 	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
 		nvif_ioctl(parent, "create channel gpfifo vers %d vmm %llx "
-				   "ioffset %016llx ilength %08x engine %08x\n",
+				   "ioffset %016llx ilength %08x "
+				   "runlist %016llx\n",
 			   args->v0.version, args->v0.vmm, args->v0.ioffset,
-			   args->v0.ilength, args->v0.engines);
-		return gk104_fifo_gpfifo_new_(gk104_fifo_gpfifo, fifo,
-					      &args->v0.engines,
+			   args->v0.ilength, args->v0.runlist);
+		return gk104_fifo_gpfifo_new_(fifo,
+					      &args->v0.runlist,
 					      &args->v0.chid,
 					       args->v0.vmm,
 					       args->v0.ioffset,
@@ -376,11 +351,3 @@ gk104_fifo_gpfifo_new(struct nvkm_fifo *base, const struct nvkm_oclass *oclass,
 
 	return ret;
 }
-
-const struct nvkm_fifo_chan_oclass
-gk104_fifo_gpfifo_oclass = {
-	.base.oclass = KEPLER_CHANNEL_GPFIFO_A,
-	.base.minver = 0,
-	.base.maxver = 0,
-	.ctor = gk104_fifo_gpfifo_new,
-};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogv100.c
new file mode 100644
index 000000000000..9598853ced56
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gpfifogv100.c
@@ -0,0 +1,225 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "changk104.h"
+#include "cgrp.h"
+
+#include <core/client.h>
+#include <core/gpuobj.h>
+
+#include <nvif/cla06f.h>
+#include <nvif/unpack.h>
+
+static int
+gv100_fifo_gpfifo_engine_valid(struct gk104_fifo_chan *chan, bool ce, bool valid)
+{
+	struct nvkm_subdev *subdev = &chan->base.fifo->engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	const u32 mask = ce ? 0x00020000 : 0x00010000;
+	const u32 data = valid ? mask : 0x00000000;
+	int ret;
+
+	/* Block runlist to prevent the channel from being rescheduled. */
+	mutex_lock(&subdev->mutex);
+	nvkm_mask(device, 0x002630, BIT(chan->runl), BIT(chan->runl));
+
+	/* Preempt the channel. */
+	ret = gk104_fifo_gpfifo_kick_locked(chan);
+	if (ret == 0) {
+		/* Update engine context validity. */
+		nvkm_kmap(chan->base.inst);
+		nvkm_mo32(chan->base.inst, 0x0ac, mask, data);
+		nvkm_done(chan->base.inst);
+	}
+
+	/* Resume runlist. */
+	nvkm_mask(device, 0x002630, BIT(chan->runl), 0);
+	mutex_unlock(&subdev->mutex);
+	return ret;
+}
+
+static int
+gv100_fifo_gpfifo_engine_fini(struct nvkm_fifo_chan *base,
+			      struct nvkm_engine *engine, bool suspend)
+{
+	struct gk104_fifo_chan *chan = gk104_fifo_chan(base);
+	struct nvkm_gpuobj *inst = chan->base.inst;
+	int ret;
+
+	if (engine->subdev.index >= NVKM_ENGINE_CE0 &&
+	    engine->subdev.index <= NVKM_ENGINE_CE_LAST)
+		return gk104_fifo_gpfifo_kick(chan);
+
+	ret = gv100_fifo_gpfifo_engine_valid(chan, false, false);
+	if (ret && suspend)
+		return ret;
+
+	nvkm_kmap(inst);
+	nvkm_wo32(inst, 0x0210, 0x00000000);
+	nvkm_wo32(inst, 0x0214, 0x00000000);
+	nvkm_done(inst);
+	return ret;
+}
+
+static int
+gv100_fifo_gpfifo_engine_init(struct nvkm_fifo_chan *base,
+			      struct nvkm_engine *engine)
+{
+	struct gk104_fifo_chan *chan = gk104_fifo_chan(base);
+	struct nvkm_gpuobj *inst = chan->base.inst;
+	u64 addr;
+
+	if (engine->subdev.index >= NVKM_ENGINE_CE0 &&
+	    engine->subdev.index <= NVKM_ENGINE_CE_LAST)
+		return 0;
+
+	addr = chan->engn[engine->subdev.index].vma->addr;
+	nvkm_kmap(inst);
+	nvkm_wo32(inst, 0x210, lower_32_bits(addr) | 0x00000004);
+	nvkm_wo32(inst, 0x214, upper_32_bits(addr));
+	nvkm_done(inst);
+
+	return gv100_fifo_gpfifo_engine_valid(chan, false, true);
+}
+
+const struct nvkm_fifo_chan_func
+gv100_fifo_gpfifo_func = {
+	.dtor = gk104_fifo_gpfifo_dtor,
+	.init = gk104_fifo_gpfifo_init,
+	.fini = gk104_fifo_gpfifo_fini,
+	.ntfy = gf100_fifo_chan_ntfy,
+	.engine_ctor = gk104_fifo_gpfifo_engine_ctor,
+	.engine_dtor = gk104_fifo_gpfifo_engine_dtor,
+	.engine_init = gv100_fifo_gpfifo_engine_init,
+	.engine_fini = gv100_fifo_gpfifo_engine_fini,
+};
+
+static int
+gv100_fifo_gpfifo_new_(struct gk104_fifo *fifo, u64 *runlists, u16 *chid,
+		       u64 vmm, u64 ioffset, u64 ilength,
+		       const struct nvkm_oclass *oclass,
+		       struct nvkm_object **pobject)
+{
+	struct gk104_fifo_chan *chan;
+	int runlist = ffs(*runlists) -1, ret, i;
+	unsigned long engm;
+	u64 subdevs = 0;
+	u64 usermem;
+
+	if (!vmm || runlist < 0 || runlist >= fifo->runlist_nr)
+		return -EINVAL;
+	*runlists = BIT_ULL(runlist);
+
+	engm = fifo->runlist[runlist].engm;
+	for_each_set_bit(i, &engm, fifo->engine_nr) {
+		if (fifo->engine[i].engine)
+			subdevs |= BIT_ULL(fifo->engine[i].engine->subdev.index);
+	}
+
+	/* Allocate the channel. */
+	if (!(chan = kzalloc(sizeof(*chan), GFP_KERNEL)))
+		return -ENOMEM;
+	*pobject = &chan->base.object;
+	chan->fifo = fifo;
+	chan->runl = runlist;
+	INIT_LIST_HEAD(&chan->head);
+
+	ret = nvkm_fifo_chan_ctor(&gv100_fifo_gpfifo_func, &fifo->base,
+				  0x1000, 0x1000, true, vmm, 0, subdevs,
+				  1, fifo->user.bar->addr, 0x200,
+				  oclass, &chan->base);
+	if (ret)
+		return ret;
+
+	*chid = chan->base.chid;
+
+	/* Hack to support GPUs where even individual channels should be
+	 * part of a channel group.
+	 */
+	if (fifo->func->cgrp_force) {
+		if (!(chan->cgrp = kmalloc(sizeof(*chan->cgrp), GFP_KERNEL)))
+			return -ENOMEM;
+		chan->cgrp->id = chan->base.chid;
+		INIT_LIST_HEAD(&chan->cgrp->head);
+		INIT_LIST_HEAD(&chan->cgrp->chan);
+		chan->cgrp->chan_nr = 0;
+	}
+
+	/* Clear channel control registers. */
+	usermem = chan->base.chid * 0x200;
+	ilength = order_base_2(ilength / 8);
+
+	nvkm_kmap(fifo->user.mem);
+	for (i = 0; i < 0x200; i += 4)
+		nvkm_wo32(fifo->user.mem, usermem + i, 0x00000000);
+	nvkm_done(fifo->user.mem);
+	usermem = nvkm_memory_addr(fifo->user.mem) + usermem;
+
+	/* RAMFC */
+	nvkm_kmap(chan->base.inst);
+	nvkm_wo32(chan->base.inst, 0x008, lower_32_bits(usermem));
+	nvkm_wo32(chan->base.inst, 0x00c, upper_32_bits(usermem));
+	nvkm_wo32(chan->base.inst, 0x010, 0x0000face);
+	nvkm_wo32(chan->base.inst, 0x030, 0x7ffff902);
+	nvkm_wo32(chan->base.inst, 0x048, lower_32_bits(ioffset));
+	nvkm_wo32(chan->base.inst, 0x04c, upper_32_bits(ioffset) |
+					  (ilength << 16));
+	nvkm_wo32(chan->base.inst, 0x084, 0x20400000);
+	nvkm_wo32(chan->base.inst, 0x094, 0x30000001);
+	nvkm_wo32(chan->base.inst, 0x0e4, 0x00000020);
+	nvkm_wo32(chan->base.inst, 0x0e8, chan->base.chid);
+	nvkm_wo32(chan->base.inst, 0x0f4, 0x00001100);
+	nvkm_wo32(chan->base.inst, 0x0f8, 0x10003080);
+	nvkm_mo32(chan->base.inst, 0x218, 0x00000000, 0x00000000);
+	nvkm_wo32(chan->base.inst, 0x220, 0x020a1000);
+	nvkm_wo32(chan->base.inst, 0x224, 0x00000000);
+	nvkm_done(chan->base.inst);
+	return gv100_fifo_gpfifo_engine_valid(chan, true, true);
+}
+
+int
+gv100_fifo_gpfifo_new(struct gk104_fifo *fifo, const struct nvkm_oclass *oclass,
+		      void *data, u32 size, struct nvkm_object **pobject)
+{
+	struct nvkm_object *parent = oclass->parent;
+	union {
+		struct kepler_channel_gpfifo_a_v0 v0;
+	} *args = data;
+	int ret = -ENOSYS;
+
+	nvif_ioctl(parent, "create channel gpfifo size %d\n", size);
+	if (!(ret = nvif_unpack(ret, &data, &size, args->v0, 0, 0, false))) {
+		nvif_ioctl(parent, "create channel gpfifo vers %d vmm %llx "
+				   "ioffset %016llx ilength %08x "
+				   "runlist %016llx\n",
+			   args->v0.version, args->v0.vmm, args->v0.ioffset,
+			   args->v0.ilength, args->v0.runlist);
+		return gv100_fifo_gpfifo_new_(fifo,
+					      &args->v0.runlist,
+					      &args->v0.chid,
+					       args->v0.vmm,
+					       args->v0.ioffset,
+					       args->v0.ilength,
+					      oclass, pobject);
+	}
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gv100.c
new file mode 100644
index 000000000000..4e1d159c0ae7
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/gv100.c
@@ -0,0 +1,306 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "gk104.h"
+#include "cgrp.h"
+#include "changk104.h"
+#include "user.h"
+
+#include <core/gpuobj.h>
+
+#include <nvif/class.h>
+
+static void
+gv100_fifo_runlist_chan(struct gk104_fifo_chan *chan,
+			struct nvkm_memory *memory, u32 offset)
+{
+	struct nvkm_memory *usermem = chan->fifo->user.mem;
+	const u64 user = nvkm_memory_addr(usermem) + (chan->base.chid * 0x200);
+	const u64 inst = chan->base.inst->addr;
+
+	nvkm_wo32(memory, offset + 0x0, lower_32_bits(user));
+	nvkm_wo32(memory, offset + 0x4, upper_32_bits(user));
+	nvkm_wo32(memory, offset + 0x8, lower_32_bits(inst) | chan->base.chid);
+	nvkm_wo32(memory, offset + 0xc, upper_32_bits(inst));
+}
+
+static void
+gv100_fifo_runlist_cgrp(struct nvkm_fifo_cgrp *cgrp,
+			struct nvkm_memory *memory, u32 offset)
+{
+	nvkm_wo32(memory, offset + 0x0, (128 << 24) | (3 << 16) | 0x00000001);
+	nvkm_wo32(memory, offset + 0x4, cgrp->chan_nr);
+	nvkm_wo32(memory, offset + 0x8, cgrp->id);
+	nvkm_wo32(memory, offset + 0xc, 0x00000000);
+}
+
+const struct gk104_fifo_runlist_func
+gv100_fifo_runlist = {
+	.size = 16,
+	.cgrp = gv100_fifo_runlist_cgrp,
+	.chan = gv100_fifo_runlist_chan,
+};
+
+static const struct nvkm_enum
+gv100_fifo_fault_gpcclient[] = {
+	{ 0x00, "T1_0" },
+	{ 0x01, "T1_1" },
+	{ 0x02, "T1_2" },
+	{ 0x03, "T1_3" },
+	{ 0x04, "T1_4" },
+	{ 0x05, "T1_5" },
+	{ 0x06, "T1_6" },
+	{ 0x07, "T1_7" },
+	{ 0x08, "PE_0" },
+	{ 0x09, "PE_1" },
+	{ 0x0a, "PE_2" },
+	{ 0x0b, "PE_3" },
+	{ 0x0c, "PE_4" },
+	{ 0x0d, "PE_5" },
+	{ 0x0e, "PE_6" },
+	{ 0x0f, "PE_7" },
+	{ 0x10, "RAST" },
+	{ 0x11, "GCC" },
+	{ 0x12, "GPCCS" },
+	{ 0x13, "PROP_0" },
+	{ 0x14, "PROP_1" },
+	{ 0x15, "PROP_2" },
+	{ 0x16, "PROP_3" },
+	{ 0x17, "GPM" },
+	{ 0x18, "LTP_UTLB_0" },
+	{ 0x19, "LTP_UTLB_1" },
+	{ 0x1a, "LTP_UTLB_2" },
+	{ 0x1b, "LTP_UTLB_3" },
+	{ 0x1c, "LTP_UTLB_4" },
+	{ 0x1d, "LTP_UTLB_5" },
+	{ 0x1e, "LTP_UTLB_6" },
+	{ 0x1f, "LTP_UTLB_7" },
+	{ 0x20, "RGG_UTLB" },
+	{ 0x21, "T1_8" },
+	{ 0x22, "T1_9" },
+	{ 0x23, "T1_10" },
+	{ 0x24, "T1_11" },
+	{ 0x25, "T1_12" },
+	{ 0x26, "T1_13" },
+	{ 0x27, "T1_14" },
+	{ 0x28, "T1_15" },
+	{ 0x29, "TPCCS_0" },
+	{ 0x2a, "TPCCS_1" },
+	{ 0x2b, "TPCCS_2" },
+	{ 0x2c, "TPCCS_3" },
+	{ 0x2d, "TPCCS_4" },
+	{ 0x2e, "TPCCS_5" },
+	{ 0x2f, "TPCCS_6" },
+	{ 0x30, "TPCCS_7" },
+	{ 0x31, "PE_8" },
+	{ 0x32, "PE_9" },
+	{ 0x33, "TPCCS_8" },
+	{ 0x34, "TPCCS_9" },
+	{ 0x35, "T1_16" },
+	{ 0x36, "T1_17" },
+	{ 0x37, "T1_18" },
+	{ 0x38, "T1_19" },
+	{ 0x39, "PE_10" },
+	{ 0x3a, "PE_11" },
+	{ 0x3b, "TPCCS_10" },
+	{ 0x3c, "TPCCS_11" },
+	{ 0x3d, "T1_20" },
+	{ 0x3e, "T1_21" },
+	{ 0x3f, "T1_22" },
+	{ 0x40, "T1_23" },
+	{ 0x41, "PE_12" },
+	{ 0x42, "PE_13" },
+	{ 0x43, "TPCCS_12" },
+	{ 0x44, "TPCCS_13" },
+	{ 0x45, "T1_24" },
+	{ 0x46, "T1_25" },
+	{ 0x47, "T1_26" },
+	{ 0x48, "T1_27" },
+	{ 0x49, "PE_14" },
+	{ 0x4a, "PE_15" },
+	{ 0x4b, "TPCCS_14" },
+	{ 0x4c, "TPCCS_15" },
+	{ 0x4d, "T1_28" },
+	{ 0x4e, "T1_29" },
+	{ 0x4f, "T1_30" },
+	{ 0x50, "T1_31" },
+	{ 0x51, "PE_16" },
+	{ 0x52, "PE_17" },
+	{ 0x53, "TPCCS_16" },
+	{ 0x54, "TPCCS_17" },
+	{ 0x55, "T1_32" },
+	{ 0x56, "T1_33" },
+	{ 0x57, "T1_34" },
+	{ 0x58, "T1_35" },
+	{ 0x59, "PE_18" },
+	{ 0x5a, "PE_19" },
+	{ 0x5b, "TPCCS_18" },
+	{ 0x5c, "TPCCS_19" },
+	{ 0x5d, "T1_36" },
+	{ 0x5e, "T1_37" },
+	{ 0x5f, "T1_38" },
+	{ 0x60, "T1_39" },
+	{}
+};
+
+static const struct nvkm_enum
+gv100_fifo_fault_hubclient[] = {
+	{ 0x00, "VIP" },
+	{ 0x01, "CE0" },
+	{ 0x02, "CE1" },
+	{ 0x03, "DNISO" },
+	{ 0x04, "FE" },
+	{ 0x05, "FECS" },
+	{ 0x06, "HOST" },
+	{ 0x07, "HOST_CPU" },
+	{ 0x08, "HOST_CPU_NB" },
+	{ 0x09, "ISO" },
+	{ 0x0a, "MMU" },
+	{ 0x0b, "NVDEC" },
+	{ 0x0d, "NVENC1" },
+	{ 0x0e, "NISO" },
+	{ 0x0f, "P2P" },
+	{ 0x10, "PD" },
+	{ 0x11, "PERF" },
+	{ 0x12, "PMU" },
+	{ 0x13, "RASTERTWOD" },
+	{ 0x14, "SCC" },
+	{ 0x15, "SCC_NB" },
+	{ 0x16, "SEC" },
+	{ 0x17, "SSYNC" },
+	{ 0x18, "CE2" },
+	{ 0x19, "XV" },
+	{ 0x1a, "MMU_NB" },
+	{ 0x1b, "NVENC0" },
+	{ 0x1c, "DFALCON" },
+	{ 0x1d, "SKED" },
+	{ 0x1e, "AFALCON" },
+	{ 0x1f, "DONT_CARE" },
+	{ 0x20, "HSCE0" },
+	{ 0x21, "HSCE1" },
+	{ 0x22, "HSCE2" },
+	{ 0x23, "HSCE3" },
+	{ 0x24, "HSCE4" },
+	{ 0x25, "HSCE5" },
+	{ 0x26, "HSCE6" },
+	{ 0x27, "HSCE7" },
+	{ 0x28, "HSCE8" },
+	{ 0x29, "HSCE9" },
+	{ 0x2a, "HSHUB" },
+	{ 0x2b, "PTP_X0" },
+	{ 0x2c, "PTP_X1" },
+	{ 0x2d, "PTP_X2" },
+	{ 0x2e, "PTP_X3" },
+	{ 0x2f, "PTP_X4" },
+	{ 0x30, "PTP_X5" },
+	{ 0x31, "PTP_X6" },
+	{ 0x32, "PTP_X7" },
+	{ 0x33, "NVENC2" },
+	{ 0x34, "VPR_SCRUBBER0" },
+	{ 0x35, "VPR_SCRUBBER1" },
+	{ 0x36, "DWBIF" },
+	{ 0x37, "FBFALCON" },
+	{ 0x38, "CE_SHIM" },
+	{ 0x39, "GSP" },
+	{}
+};
+
+static const struct nvkm_enum
+gv100_fifo_fault_reason[] = {
+	{ 0x00, "PDE" },
+	{ 0x01, "PDE_SIZE" },
+	{ 0x02, "PTE" },
+	{ 0x03, "VA_LIMIT_VIOLATION" },
+	{ 0x04, "UNBOUND_INST_BLOCK" },
+	{ 0x05, "PRIV_VIOLATION" },
+	{ 0x06, "RO_VIOLATION" },
+	{ 0x07, "WO_VIOLATION" },
+	{ 0x08, "PITCH_MASK_VIOLATION" },
+	{ 0x09, "WORK_CREATION" },
+	{ 0x0a, "UNSUPPORTED_APERTURE" },
+	{ 0x0b, "COMPRESSION_FAILURE" },
+	{ 0x0c, "UNSUPPORTED_KIND" },
+	{ 0x0d, "REGION_VIOLATION" },
+	{ 0x0e, "POISONED" },
+	{ 0x0f, "ATOMIC_VIOLATION" },
+	{}
+};
+
+static const struct nvkm_enum
+gv100_fifo_fault_engine[] = {
+	{ 0x01, "DISPLAY" },
+	{ 0x03, "PTP" },
+	{ 0x04, "BAR1", NULL, NVKM_SUBDEV_BAR },
+	{ 0x05, "BAR2", NULL, NVKM_SUBDEV_INSTMEM },
+	{ 0x06, "PWR_PMU" },
+	{ 0x08, "IFB", NULL, NVKM_ENGINE_IFB },
+	{ 0x09, "PERF" },
+	{ 0x1f, "PHYSICAL" },
+	{ 0x20, "HOST0" },
+	{ 0x21, "HOST1" },
+	{ 0x22, "HOST2" },
+	{ 0x23, "HOST3" },
+	{ 0x24, "HOST4" },
+	{ 0x25, "HOST5" },
+	{ 0x26, "HOST6" },
+	{ 0x27, "HOST7" },
+	{ 0x28, "HOST8" },
+	{ 0x29, "HOST9" },
+	{ 0x2a, "HOST10" },
+	{ 0x2b, "HOST11" },
+	{ 0x2c, "HOST12" },
+	{ 0x2d, "HOST13" },
+	{}
+};
+
+static const struct nvkm_enum
+gv100_fifo_fault_access[] = {
+	{ 0x0, "VIRT_READ" },
+	{ 0x1, "VIRT_WRITE" },
+	{ 0x2, "VIRT_ATOMIC" },
+	{ 0x3, "VIRT_PREFETCH" },
+	{ 0x4, "VIRT_ATOMIC_WEAK" },
+	{ 0x8, "PHYS_READ" },
+	{ 0x9, "PHYS_WRITE" },
+	{ 0xa, "PHYS_ATOMIC" },
+	{ 0xb, "PHYS_PREFETCH" },
+	{}
+};
+
+static const struct gk104_fifo_func
+gv100_fifo = {
+	.init_pbdma_timeout = gk208_fifo_init_pbdma_timeout,
+	.fault.access = gv100_fifo_fault_access,
+	.fault.engine = gv100_fifo_fault_engine,
+	.fault.reason = gv100_fifo_fault_reason,
+	.fault.hubclient = gv100_fifo_fault_hubclient,
+	.fault.gpcclient = gv100_fifo_fault_gpcclient,
+	.runlist = &gv100_fifo_runlist,
+	.user = {{-1,-1,VOLTA_USERMODE_A      }, gv100_fifo_user_new   },
+	.chan = {{ 0, 0,VOLTA_CHANNEL_GPFIFO_A}, gv100_fifo_gpfifo_new },
+	.cgrp_force = true,
+};
+
+int
+gv100_fifo_new(struct nvkm_device *device, int index, struct nvkm_fifo **pfifo)
+{
+	return gk104_fifo_new_(&gv100_fifo, device, index, 4096, pfifo);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
index ae76b1aaccd4..d5acbba293f4 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/priv.h
@@ -18,16 +18,19 @@ struct nvkm_fifo_chan_oclass;
 struct nvkm_fifo_func {
 	void *(*dtor)(struct nvkm_fifo *);
 	int (*oneinit)(struct nvkm_fifo *);
+	int (*info)(struct nvkm_fifo *, u64 mthd, u64 *data);
 	void (*init)(struct nvkm_fifo *);
 	void (*fini)(struct nvkm_fifo *);
 	void (*intr)(struct nvkm_fifo *);
+	void (*fault)(struct nvkm_fifo *, struct nvkm_fault_data *);
 	void (*pause)(struct nvkm_fifo *, unsigned long *);
 	void (*start)(struct nvkm_fifo *, unsigned long *);
 	void (*uevent_init)(struct nvkm_fifo *);
 	void (*uevent_fini)(struct nvkm_fifo *);
 	void (*recover_chan)(struct nvkm_fifo *, int chid);
-	int (*class_get)(struct nvkm_fifo *, int index,
-			 const struct nvkm_fifo_chan_oclass **);
+	int (*class_get)(struct nvkm_fifo *, int index, struct nvkm_oclass *);
+	int (*class_new)(struct nvkm_fifo *, const struct nvkm_oclass *,
+			 void *, u32, struct nvkm_object **);
 	const struct nvkm_fifo_chan_oclass *chan[];
 };
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/user.h b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/user.h
new file mode 100644
index 000000000000..ed840921ebe8
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/user.h
@@ -0,0 +1,6 @@
+#ifndef __NVKM_FIFO_USER_H__
+#define __NVKM_FIFO_USER_H__
+#include "priv.h"
+int gv100_fifo_user_new(const struct nvkm_oclass *, void *, u32,
+			struct nvkm_object **);
+#endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/usergv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/usergv100.c
new file mode 100644
index 000000000000..3dc3b8b312de
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/usergv100.c
@@ -0,0 +1,45 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "user.h"
+
+static int
+gv100_fifo_user_map(struct nvkm_object *object, void *argv, u32 argc,
+		    enum nvkm_object_map *type, u64 *addr, u64 *size)
+{
+	struct nvkm_device *device = object->engine->subdev.device;
+	*addr = 0x810000 + device->func->resource_addr(device, 0);
+	*size = 0x010000;
+	*type = NVKM_OBJECT_MAP_IO;
+	return 0;
+}
+
+static const struct nvkm_object_func
+gv100_fifo_user = {
+	.map = gv100_fifo_user_map,
+};
+
+int
+gv100_fifo_user_new(const struct nvkm_oclass *oclass, void *argv, u32 argc,
+		    struct nvkm_object **pobject)
+{
+	return nvkm_object_new_(&gv100_fifo_user, oclass, argv, argc, pobject);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild b/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild
index 8a22558b7b52..93e3733f54e2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/Kbuild
@@ -33,8 +33,10 @@ nvkm-y += nvkm/engine/gr/gm200.o
 nvkm-y += nvkm/engine/gr/gm20b.o
 nvkm-y += nvkm/engine/gr/gp100.o
 nvkm-y += nvkm/engine/gr/gp102.o
+nvkm-y += nvkm/engine/gr/gp104.o
 nvkm-y += nvkm/engine/gr/gp107.o
 nvkm-y += nvkm/engine/gr/gp10b.o
+nvkm-y += nvkm/engine/gr/gv100.o
 
 nvkm-y += nvkm/engine/gr/ctxnv40.o
 nvkm-y += nvkm/engine/gr/ctxnv50.o
@@ -54,4 +56,6 @@ nvkm-y += nvkm/engine/gr/ctxgm200.o
 nvkm-y += nvkm/engine/gr/ctxgm20b.o
 nvkm-y += nvkm/engine/gr/ctxgp100.o
 nvkm-y += nvkm/engine/gr/ctxgp102.o
+nvkm-y += nvkm/engine/gr/ctxgp104.o
 nvkm-y += nvkm/engine/gr/ctxgp107.o
+nvkm-y += nvkm/engine/gr/ctxgv100.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
index 881015080d83..e813a3f8ea93 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c
@@ -850,12 +850,17 @@ gf100_grctx_init_gcc_0[] = {
 };
 
 const struct gf100_gr_pack
-gf100_grctx_pack_gpc[] = {
+gf100_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gf100_grctx_init_prop_0 },
 	{ gf100_grctx_init_gpc_unk_1 },
 	{ gf100_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
+	{}
+};
+
+const struct gf100_gr_pack
+gf100_grctx_pack_gpc_1[] = {
 	{ gf100_grctx_init_crstr_0 },
 	{ gf100_grctx_init_gpm_0 },
 	{ gf100_grctx_init_gcc_0 },
@@ -1025,6 +1030,13 @@ gf100_grctx_mmio_item(struct gf100_grctx *info, u32 addr, u32 data,
 }
 
 void
+gf100_grctx_generate_r419cb8(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419cb8, 0x00007c00, 0x00000000);
+}
+
+void
 gf100_grctx_generate_bundle(struct gf100_grctx *info)
 {
 	const struct gf100_grctx_func *grctx = info->gr->func->grctx;
@@ -1080,89 +1092,38 @@ gf100_grctx_generate_unkn(struct gf100_gr *gr)
 }
 
 void
-gf100_grctx_generate_tpcid(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	int gpc, tpc, id;
-
-	for (tpc = 0, id = 0; tpc < 4; tpc++) {
-		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-			if (tpc < gr->tpc_nr[gpc]) {
-				nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x698), id);
-				nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x4e8), id);
-				nvkm_wr32(device, GPC_UNIT(gpc, 0x0c10 + tpc * 4), id);
-				nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x088), id);
-				id++;
-			}
-
-			nvkm_wr32(device, GPC_UNIT(gpc, 0x0c08), gr->tpc_nr[gpc]);
-			nvkm_wr32(device, GPC_UNIT(gpc, 0x0c8c), gr->tpc_nr[gpc]);
-		}
-	}
-}
-
-void
-gf100_grctx_generate_r406028(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	u32 tmp[GPC_MAX / 8] = {}, i = 0;
-	for (i = 0; i < gr->gpc_nr; i++)
-		tmp[i / 8] |= gr->tpc_nr[i] << ((i % 8) * 4);
-	for (i = 0; i < 4; i++) {
-		nvkm_wr32(device, 0x406028 + (i * 4), tmp[i]);
-		nvkm_wr32(device, 0x405870 + (i * 4), tmp[i]);
-	}
-}
-
-void
 gf100_grctx_generate_r4060a8(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	u8  tpcnr[GPC_MAX], data[TPC_MAX];
-	int gpc, tpc, i;
-
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	memset(data, 0x1f, sizeof(data));
-
-	gpc = -1;
-	for (tpc = 0; tpc < gr->tpc_total; tpc++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpcnr[gpc]--;
-		data[tpc] = gpc;
+	const u8 gpcmax = nvkm_rd32(device, 0x022430);
+	const u8 tpcmax = nvkm_rd32(device, 0x022434) * gpcmax;
+	int i, j, sm = 0;
+	u32 data;
+
+	for (i = 0; i < DIV_ROUND_UP(tpcmax, 4); i++) {
+		for (data = 0, j = 0; j < 4; j++) {
+			if (sm < gr->sm_nr)
+				data |= gr->sm[sm++].gpc << (j * 8);
+			else
+				data |= 0x1f << (j * 8);
+		}
+		nvkm_wr32(device, 0x4060a8 + (i * 4), data);
 	}
-
-	for (i = 0; i < 4; i++)
-		nvkm_wr32(device, 0x4060a8 + (i * 4), ((u32 *)data)[i]);
 }
 
 void
-gf100_grctx_generate_r418bb8(struct gf100_gr *gr)
+gf100_grctx_generate_rop_mapping(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	u32 data[6] = {}, data2[2] = {};
-	u8  tpcnr[GPC_MAX];
 	u8  shift, ntpcv;
-	int gpc, tpc, i;
-
-	/* calculate first set of magics */
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-
-	gpc = -1;
-	for (tpc = 0; tpc < gr->tpc_total; tpc++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpcnr[gpc]--;
-
-		data[tpc / 6] |= gpc << ((tpc % 6) * 5);
-	}
+	int i;
 
-	for (; tpc < 32; tpc++)
-		data[tpc / 6] |= 7 << ((tpc % 6) * 5);
+	/* Pack tile map into register format. */
+	for (i = 0; i < 32; i++)
+		data[i / 6] |= (gr->tile[i] & 0x07) << ((i % 6) * 5);
 
-	/* and the second... */
+	/* Magic. */
 	shift = 0;
 	ntpcv = gr->tpc_total;
 	while (!(ntpcv & (1 << 4))) {
@@ -1197,40 +1158,214 @@ gf100_grctx_generate_r418bb8(struct gf100_gr *gr)
 }
 
 void
-gf100_grctx_generate_r406800(struct gf100_gr *gr)
+gf100_grctx_generate_max_ways_evict(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	u64 tpc_mask = 0, tpc_set = 0;
-	u8  tpcnr[GPC_MAX];
-	int gpc, tpc;
-	int i, a, b;
-
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++)
-		tpc_mask |= ((1ULL << gr->tpc_nr[gpc]) - 1) << (gpc * 8);
-
-	for (i = 0, gpc = -1, b = -1; i < 32; i++) {
-		a = (i * (gr->tpc_total - 1)) / 32;
-		if (a != b) {
-			b = a;
-			do {
-				gpc = (gpc + 1) % gr->gpc_nr;
-			} while (!tpcnr[gpc]);
-			tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-			tpc_set |= 1ULL << ((gpc * 8) + tpc);
+	u32 fbps = nvkm_rd32(device, 0x121c74);
+	if (fbps == 1)
+		nvkm_mask(device, 0x17e91c, 0x001f0000, 0x00090000);
+}
+
+static const u32
+gf100_grctx_alpha_beta_map[17][32] = {
+	[1] = {
+		1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+		1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+	},
+	[2] = {
+		1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+		1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
+	},
+	//XXX: 3
+	[4] = {
+		1, 1, 1, 1, 1, 1, 1, 1,
+		2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
+		3, 3, 3, 3, 3, 3, 3, 3,
+	},
+	//XXX: 5
+	//XXX: 6
+	[7] = {
+		1, 1, 1, 1,
+		2, 2, 2, 2, 2, 2,
+		3, 3, 3, 3, 3, 3,
+		4, 4, 4, 4, 4, 4,
+		5, 5, 5, 5, 5, 5,
+		6, 6, 6, 6,
+	},
+	[8] = {
+		1, 1, 1,
+		2, 2, 2, 2, 2,
+		3, 3, 3, 3, 3,
+		4, 4, 4, 4, 4, 4,
+		5, 5, 5, 5, 5,
+		6, 6, 6, 6, 6,
+		7, 7, 7,
+	},
+	//XXX: 9
+	//XXX: 10
+	[11] = {
+		1, 1,
+		2, 2, 2, 2,
+		3, 3, 3,
+		4, 4, 4, 4,
+		5, 5, 5,
+		6, 6, 6,
+		7, 7, 7, 7,
+		8, 8, 8,
+		9, 9, 9, 9,
+		10, 10,
+	},
+	//XXX: 12
+	//XXX: 13
+	[14] = {
+		1, 1,
+		2, 2,
+		3, 3, 3,
+		4, 4, 4,
+		5, 5,
+		6, 6, 6,
+		7, 7,
+		8, 8, 8,
+		9, 9,
+		10, 10, 10,
+		11, 11, 11,
+		12, 12,
+		13, 13,
+	},
+	[15] = {
+		1, 1,
+		2, 2,
+		3, 3,
+		4, 4, 4,
+		5, 5,
+		6, 6, 6,
+		7, 7,
+		8, 8,
+		9, 9, 9,
+		10, 10,
+		11, 11, 11,
+		12, 12,
+		13, 13,
+		14, 14,
+	},
+	[16] = {
+		1, 1,
+		2, 2,
+		3, 3,
+		4, 4,
+		5, 5,
+		6, 6, 6,
+		7, 7,
+		8, 8,
+		9, 9,
+		10, 10, 10,
+		11, 11,
+		12, 12,
+		13, 13,
+		14, 14,
+		15, 15,
+	},
+};
+
+void
+gf100_grctx_generate_alpha_beta_tables(struct gf100_gr *gr)
+{
+	struct nvkm_subdev *subdev = &gr->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	int i, gpc;
+
+	for (i = 0; i < 32; i++) {
+		u32 atarget = gf100_grctx_alpha_beta_map[gr->tpc_total][i];
+		u32 abits[GPC_MAX] = {}, amask = 0, bmask = 0;
+
+		if (!atarget) {
+			nvkm_warn(subdev, "missing alpha/beta mapping table\n");
+			atarget = max_t(u32, gr->tpc_total * i / 32, 1);
+		}
+
+		while (atarget) {
+			for (gpc = 0; atarget && gpc < gr->gpc_nr; gpc++) {
+				if (abits[gpc] < gr->tpc_nr[gpc]) {
+					abits[gpc]++;
+					atarget--;
+				}
+			}
 		}
 
-		nvkm_wr32(device, 0x406800 + (i * 0x20), lower_32_bits(tpc_set));
-		nvkm_wr32(device, 0x406c00 + (i * 0x20), lower_32_bits(tpc_set ^ tpc_mask));
-		if (gr->gpc_nr > 4) {
-			nvkm_wr32(device, 0x406804 + (i * 0x20), upper_32_bits(tpc_set));
-			nvkm_wr32(device, 0x406c04 + (i * 0x20), upper_32_bits(tpc_set ^ tpc_mask));
+		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+			u32 bbits = gr->tpc_nr[gpc] - abits[gpc];
+			amask |= ((1 << abits[gpc]) - 1) << (gpc * 8);
+			bmask |= ((1 << bbits) - 1) << abits[gpc] << (gpc * 8);
 		}
+
+		nvkm_wr32(device, 0x406800 + (i * 0x20), amask);
+		nvkm_wr32(device, 0x406c00 + (i * 0x20), bmask);
 	}
 }
 
 void
+gf100_grctx_generate_tpc_nr(struct gf100_gr *gr, int gpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, GPC_UNIT(gpc, 0x0c08), gr->tpc_nr[gpc]);
+	nvkm_wr32(device, GPC_UNIT(gpc, 0x0c8c), gr->tpc_nr[gpc]);
+}
+
+void
+gf100_grctx_generate_sm_id(struct gf100_gr *gr, int gpc, int tpc, int sm)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x698), sm);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x4e8), sm);
+	nvkm_wr32(device, GPC_UNIT(gpc, 0x0c10 + tpc * 4), sm);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x088), sm);
+}
+
+void
+gf100_grctx_generate_floorsweep(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	const struct gf100_grctx_func *func = gr->func->grctx;
+	int gpc, sm, i, j;
+	u32 data;
+
+	for (sm = 0; sm < gr->sm_nr; sm++) {
+		func->sm_id(gr, gr->sm[sm].gpc, gr->sm[sm].tpc, sm);
+		if (func->tpc_nr)
+			func->tpc_nr(gr, gr->sm[sm].gpc);
+	}
+
+	for (gpc = 0, i = 0; i < 4; i++) {
+		for (data = 0, j = 0; j < 8 && gpc < gr->gpc_nr; j++, gpc++)
+			data |= gr->tpc_nr[gpc] << (j * 4);
+		nvkm_wr32(device, 0x406028 + (i * 4), data);
+		nvkm_wr32(device, 0x405870 + (i * 4), data);
+	}
+
+	if (func->r4060a8)
+		func->r4060a8(gr);
+
+	func->rop_mapping(gr);
+
+	if (func->alpha_beta_tables)
+		func->alpha_beta_tables(gr);
+	if (func->max_ways_evict)
+		func->max_ways_evict(gr);
+	if (func->dist_skip_table)
+		func->dist_skip_table(gr);
+	if (func->r406500)
+		func->r406500(gr);
+	if (func->gpc_tpc_nr)
+		func->gpc_tpc_nr(gr);
+	if (func->r419f78)
+		func->r419f78(gr);
+	if (func->tpc_mask)
+		func->tpc_mask(gr);
+	if (func->smid_config)
+		func->smid_config(gr);
+}
+
+void
 gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -1239,29 +1374,63 @@ gf100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	nvkm_mc_unk260(device, 0);
 
-	gf100_gr_mmio(gr, grctx->hub);
-	gf100_gr_mmio(gr, grctx->gpc);
-	gf100_gr_mmio(gr, grctx->zcull);
-	gf100_gr_mmio(gr, grctx->tpc);
-	gf100_gr_mmio(gr, grctx->ppc);
+	if (!gr->fuc_sw_ctx) {
+		gf100_gr_mmio(gr, grctx->hub);
+		gf100_gr_mmio(gr, grctx->gpc_0);
+		gf100_gr_mmio(gr, grctx->zcull);
+		gf100_gr_mmio(gr, grctx->gpc_1);
+		gf100_gr_mmio(gr, grctx->tpc);
+		gf100_gr_mmio(gr, grctx->ppc);
+	} else {
+		gf100_gr_mmio(gr, gr->fuc_sw_ctx);
+	}
+
+	gf100_gr_wait_idle(gr);
 
 	idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
 
-	grctx->bundle(info);
 	grctx->pagepool(info);
+	grctx->bundle(info);
 	grctx->attrib(info);
+	if (grctx->patch_ltc)
+		grctx->patch_ltc(info);
 	grctx->unkn(gr);
 
-	gf100_grctx_generate_tpcid(gr);
-	gf100_grctx_generate_r406028(gr);
-	gf100_grctx_generate_r4060a8(gr);
-	gf100_grctx_generate_r418bb8(gr);
-	gf100_grctx_generate_r406800(gr);
+	gf100_grctx_generate_floorsweep(gr);
+
+	gf100_gr_wait_idle(gr);
+
+	if (grctx->r400088) grctx->r400088(gr, false);
+	if (gr->fuc_bundle)
+		gf100_gr_icmd(gr, gr->fuc_bundle);
+	else
+		gf100_gr_icmd(gr, grctx->icmd);
+	if (grctx->sw_veid_bundle_init)
+		gf100_gr_icmd(gr, grctx->sw_veid_bundle_init);
+	if (grctx->r400088) grctx->r400088(gr, true);
 
-	gf100_gr_icmd(gr, grctx->icmd);
 	nvkm_wr32(device, 0x404154, idle_timeout);
-	gf100_gr_mthd(gr, grctx->mthd);
+
+	if (gr->fuc_method)
+		gf100_gr_mthd(gr, gr->fuc_method);
+	else
+		gf100_gr_mthd(gr, grctx->mthd);
 	nvkm_mc_unk260(device, 1);
+
+	if (grctx->r419cb8)
+		grctx->r419cb8(gr);
+	if (grctx->r418800)
+		grctx->r418800(gr);
+	if (grctx->r419eb0)
+		grctx->r419eb0(gr);
+	if (grctx->r419e00)
+		grctx->r419e00(gr);
+	if (grctx->r418e94)
+		grctx->r418e94(gr);
+	if (grctx->r419a3c)
+		grctx->r419a3c(gr);
+	if (grctx->r408840)
+		grctx->r408840(gr);
 }
 
 #define CB_RESERVED 0x80000
@@ -1280,6 +1449,32 @@ gf100_grctx_generate(struct gf100_gr *gr)
 	int ret, i;
 	u64 addr;
 
+	/* NV_PGRAPH_FE_PWR_MODE_FORCE_ON. */
+	nvkm_wr32(device, 0x404170, 0x00000012);
+	nvkm_msec(device, 2000,
+		if (!(nvkm_rd32(device, 0x404170) & 0x00000010))
+			break;
+	);
+
+	if (grctx->unkn88c)
+		grctx->unkn88c(gr, true);
+
+	/* Reset FECS. */
+	nvkm_wr32(device, 0x409614, 0x00000070);
+	nvkm_usec(device, 10, NVKM_DELAY);
+	nvkm_mask(device, 0x409614, 0x00000700, 0x00000700);
+	nvkm_usec(device, 10, NVKM_DELAY);
+	nvkm_rd32(device, 0x409614);
+
+	if (grctx->unkn88c)
+		grctx->unkn88c(gr, false);
+
+	/* NV_PGRAPH_FE_PWR_MODE_AUTO. */
+	nvkm_wr32(device, 0x404170, 0x00000010);
+
+	/* Init SCC RAM. */
+	nvkm_wr32(device, 0x40802c, 0x00000001);
+
 	/* Allocate memory to for a "channel", which we'll use to generate
 	 * the default context values.
 	 */
@@ -1392,7 +1587,8 @@ gf100_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gf100_grctx_generate_unkn,
 	.hub   = gf100_grctx_pack_hub,
-	.gpc   = gf100_grctx_pack_gpc,
+	.gpc_0 = gf100_grctx_pack_gpc_0,
+	.gpc_1 = gf100_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gf100_grctx_pack_tpc,
 	.icmd  = gf100_grctx_pack_icmd,
@@ -1404,4 +1600,11 @@ gf100_grctx = {
 	.attrib = gf100_grctx_generate_attrib,
 	.attrib_nr_max = 0x324,
 	.attrib_nr = 0x218,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.r4060a8 = gf100_grctx_generate_r4060a8,
+	.rop_mapping = gf100_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
+	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
+	.r419cb8 = gf100_grctx_generate_r419cb8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
index 5199e5aa0cb7..33e932bd73b1 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.h
@@ -21,19 +21,22 @@ void gf100_grctx_mmio_item(struct gf100_grctx *, u32 addr, u32 data, int s, int)
 #define mmio_wr32(a,b,c) mmio_refn((a), (b), (c),  0, -1)
 
 struct gf100_grctx_func {
+	void (*unkn88c)(struct gf100_gr *, bool on);
 	/* main context generation function */
 	void  (*main)(struct gf100_gr *, struct gf100_grctx *);
 	/* context-specific modify-on-first-load list generation function */
 	void  (*unkn)(struct gf100_gr *);
 	/* mmio context data */
 	const struct gf100_gr_pack *hub;
-	const struct gf100_gr_pack *gpc;
+	const struct gf100_gr_pack *gpc_0;
+	const struct gf100_gr_pack *gpc_1;
 	const struct gf100_gr_pack *zcull;
 	const struct gf100_gr_pack *tpc;
 	const struct gf100_gr_pack *ppc;
 	/* indirect context data, generated with icmds/mthds */
 	const struct gf100_gr_pack *icmd;
 	const struct gf100_gr_pack *mthd;
+	const struct gf100_gr_pack *sw_veid_bundle_init;
 	/* bundle circular buffer */
 	void (*bundle)(struct gf100_grctx *);
 	u32 bundle_size;
@@ -48,6 +51,31 @@ struct gf100_grctx_func {
 	u32 attrib_nr;
 	u32 alpha_nr_max;
 	u32 alpha_nr;
+	u32 gfxp_nr;
+	/* other patch buffer stuff */
+	void (*patch_ltc)(struct gf100_grctx *);
+	/* floorsweeping */
+	void (*sm_id)(struct gf100_gr *, int gpc, int tpc, int sm);
+	void (*tpc_nr)(struct gf100_gr *, int gpc);
+	void (*r4060a8)(struct gf100_gr *);
+	void (*rop_mapping)(struct gf100_gr *);
+	void (*alpha_beta_tables)(struct gf100_gr *);
+	void (*max_ways_evict)(struct gf100_gr *);
+	void (*dist_skip_table)(struct gf100_gr *);
+	void (*r406500)(struct gf100_gr *);
+	void (*gpc_tpc_nr)(struct gf100_gr *);
+	void (*r419f78)(struct gf100_gr *);
+	void (*tpc_mask)(struct gf100_gr *);
+	void (*smid_config)(struct gf100_gr *);
+	/* misc other things */
+	void (*r400088)(struct gf100_gr *, bool);
+	void (*r419cb8)(struct gf100_gr *);
+	void (*r418800)(struct gf100_gr *);
+	void (*r419eb0)(struct gf100_gr *);
+	void (*r419e00)(struct gf100_gr *);
+	void (*r418e94)(struct gf100_gr *);
+	void (*r419a3c)(struct gf100_gr *);
+	void (*r408840)(struct gf100_gr *);
 };
 
 extern const struct gf100_grctx_func gf100_grctx;
@@ -57,11 +85,14 @@ void gf100_grctx_generate_bundle(struct gf100_grctx *);
 void gf100_grctx_generate_pagepool(struct gf100_grctx *);
 void gf100_grctx_generate_attrib(struct gf100_grctx *);
 void gf100_grctx_generate_unkn(struct gf100_gr *);
-void gf100_grctx_generate_tpcid(struct gf100_gr *);
-void gf100_grctx_generate_r406028(struct gf100_gr *);
+void gf100_grctx_generate_floorsweep(struct gf100_gr *);
+void gf100_grctx_generate_sm_id(struct gf100_gr *, int, int, int);
+void gf100_grctx_generate_tpc_nr(struct gf100_gr *, int);
 void gf100_grctx_generate_r4060a8(struct gf100_gr *);
-void gf100_grctx_generate_r418bb8(struct gf100_gr *);
-void gf100_grctx_generate_r406800(struct gf100_gr *);
+void gf100_grctx_generate_rop_mapping(struct gf100_gr *);
+void gf100_grctx_generate_alpha_beta_tables(struct gf100_gr *);
+void gf100_grctx_generate_max_ways_evict(struct gf100_gr *);
+void gf100_grctx_generate_r419cb8(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gf108_grctx;
 void gf108_grctx_generate_attrib(struct gf100_grctx *);
@@ -72,22 +103,25 @@ extern const struct gf100_grctx_func gf110_grctx;
 
 extern const struct gf100_grctx_func gf117_grctx;
 void gf117_grctx_generate_attrib(struct gf100_grctx *);
+void gf117_grctx_generate_rop_mapping(struct gf100_gr *);
+void gf117_grctx_generate_dist_skip_table(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gf119_grctx;
 
 extern const struct gf100_grctx_func gk104_grctx;
+void gk104_grctx_generate_alpha_beta_tables(struct gf100_gr *);
+void gk104_grctx_generate_gpc_tpc_nr(struct gf100_gr *);
+
 extern const struct gf100_grctx_func gk20a_grctx;
-void gk104_grctx_generate_main(struct gf100_gr *, struct gf100_grctx *);
 void gk104_grctx_generate_bundle(struct gf100_grctx *);
 void gk104_grctx_generate_pagepool(struct gf100_grctx *);
+void gk104_grctx_generate_patch_ltc(struct gf100_grctx *);
 void gk104_grctx_generate_unkn(struct gf100_gr *);
-void gk104_grctx_generate_r418bb8(struct gf100_gr *);
-
-void gm107_grctx_generate_bundle(struct gf100_grctx *);
-void gm107_grctx_generate_pagepool(struct gf100_grctx *);
-void gm107_grctx_generate_attrib(struct gf100_grctx *);
+void gk104_grctx_generate_r418800(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gk110_grctx;
+void gk110_grctx_generate_r419eb0(struct gf100_gr *);
+
 extern const struct gf100_grctx_func gk110b_grctx;
 extern const struct gf100_grctx_func gk208_grctx;
 
@@ -95,22 +129,30 @@ extern const struct gf100_grctx_func gm107_grctx;
 void gm107_grctx_generate_bundle(struct gf100_grctx *);
 void gm107_grctx_generate_pagepool(struct gf100_grctx *);
 void gm107_grctx_generate_attrib(struct gf100_grctx *);
+void gm107_grctx_generate_sm_id(struct gf100_gr *, int, int, int);
 
 extern const struct gf100_grctx_func gm200_grctx;
-void gm200_grctx_generate_tpcid(struct gf100_gr *);
-void gm200_grctx_generate_405b60(struct gf100_gr *);
+void gm200_grctx_generate_dist_skip_table(struct gf100_gr *);
+void gm200_grctx_generate_r406500(struct gf100_gr *);
+void gm200_grctx_generate_tpc_mask(struct gf100_gr *);
+void gm200_grctx_generate_smid_config(struct gf100_gr *);
+void gm200_grctx_generate_r419a3c(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gm20b_grctx;
 
 extern const struct gf100_grctx_func gp100_grctx;
-void gp100_grctx_generate_main(struct gf100_gr *, struct gf100_grctx *);
 void gp100_grctx_generate_pagepool(struct gf100_grctx *);
+void gp100_grctx_generate_smid_config(struct gf100_gr *);
 
 extern const struct gf100_grctx_func gp102_grctx;
 void gp102_grctx_generate_attrib(struct gf100_grctx *);
 
+extern const struct gf100_grctx_func gp104_grctx;
+
 extern const struct gf100_grctx_func gp107_grctx;
 
+extern const struct gf100_grctx_func gv100_grctx;
+
 /* context init value lists */
 
 extern const struct gf100_gr_pack gf100_grctx_pack_icmd[];
@@ -128,7 +170,8 @@ extern const struct gf100_gr_init gf100_grctx_init_memfmt_0[];
 extern const struct gf100_gr_init gf100_grctx_init_rstr2d_0[];
 extern const struct gf100_gr_init gf100_grctx_init_scc_0[];
 
-extern const struct gf100_gr_pack gf100_grctx_pack_gpc[];
+extern const struct gf100_gr_pack gf100_grctx_pack_gpc_0[];
+extern const struct gf100_gr_pack gf100_grctx_pack_gpc_1[];
 extern const struct gf100_gr_init gf100_grctx_init_gpc_unk_0[];
 extern const struct gf100_gr_init gf100_grctx_init_prop_0[];
 extern const struct gf100_gr_init gf100_grctx_init_gpc_unk_1[];
@@ -177,6 +220,8 @@ extern const struct gf100_gr_init gf117_grctx_init_pe_0[];
 
 extern const struct gf100_gr_init gf117_grctx_init_wwdx_0[];
 
+extern const struct gf100_gr_pack gf117_grctx_pack_gpc_1[];
+
 extern const struct gf100_gr_init gk104_grctx_init_memfmt_0[];
 extern const struct gf100_gr_init gk104_grctx_init_ds_0[];
 extern const struct gf100_gr_init gk104_grctx_init_scc_0[];
@@ -186,7 +231,6 @@ extern const struct gf100_gr_init gk104_grctx_init_gpm_0[];
 extern const struct gf100_gr_init gk104_grctx_init_pes_0[];
 
 extern const struct gf100_gr_pack gk104_grctx_pack_hub[];
-extern const struct gf100_gr_pack gk104_grctx_pack_gpc[];
 extern const struct gf100_gr_pack gk104_grctx_pack_tpc[];
 extern const struct gf100_gr_pack gk104_grctx_pack_ppc[];
 extern const struct gf100_gr_pack gk104_grctx_pack_icmd[];
@@ -200,7 +244,8 @@ extern const struct gf100_gr_pack gk110_grctx_pack_hub[];
 extern const struct gf100_gr_init gk110_grctx_init_pri_0[];
 extern const struct gf100_gr_init gk110_grctx_init_cwd_0[];
 
-extern const struct gf100_gr_pack gk110_grctx_pack_gpc[];
+extern const struct gf100_gr_pack gk110_grctx_pack_gpc_0[];
+extern const struct gf100_gr_pack gk110_grctx_pack_gpc_1[];
 extern const struct gf100_gr_init gk110_grctx_init_gpc_unk_2[];
 
 extern const struct gf100_gr_init gk110_grctx_init_tex_0[];
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
index 54fd74e9cca0..7a0564b6e3c7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf104.c
@@ -84,7 +84,8 @@ gf104_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gf100_grctx_generate_unkn,
 	.hub   = gf100_grctx_pack_hub,
-	.gpc   = gf100_grctx_pack_gpc,
+	.gpc_0 = gf100_grctx_pack_gpc_0,
+	.gpc_1 = gf100_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gf104_grctx_pack_tpc,
 	.icmd  = gf100_grctx_pack_icmd,
@@ -96,4 +97,11 @@ gf104_grctx = {
 	.attrib = gf100_grctx_generate_attrib,
 	.attrib_nr_max = 0x324,
 	.attrib_nr = 0x218,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.r4060a8 = gf100_grctx_generate_r4060a8,
+	.rop_mapping = gf100_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
+	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
+	.r419cb8 = gf100_grctx_generate_r419cb8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
index 82f71b10c06e..dda2c32e6232 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf108.c
@@ -667,12 +667,17 @@ gf108_grctx_init_gpm_0[] = {
 };
 
 static const struct gf100_gr_pack
-gf108_grctx_pack_gpc[] = {
+gf108_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gf100_grctx_init_prop_0 },
 	{ gf100_grctx_init_gpc_unk_1 },
 	{ gf108_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
+	{}
+};
+
+static const struct gf100_gr_pack
+gf108_grctx_pack_gpc_1[] = {
 	{ gf100_grctx_init_crstr_0 },
 	{ gf108_grctx_init_gpm_0 },
 	{ gf100_grctx_init_gcc_0 },
@@ -780,7 +785,8 @@ gf108_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gf108_grctx_generate_unkn,
 	.hub   = gf108_grctx_pack_hub,
-	.gpc   = gf108_grctx_pack_gpc,
+	.gpc_0 = gf108_grctx_pack_gpc_0,
+	.gpc_1 = gf108_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gf108_grctx_pack_tpc,
 	.icmd  = gf108_grctx_pack_icmd,
@@ -794,4 +800,11 @@ gf108_grctx = {
 	.attrib_nr = 0x218,
 	.alpha_nr_max = 0x324,
 	.alpha_nr = 0x218,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.r4060a8 = gf100_grctx_generate_r4060a8,
+	.rop_mapping = gf100_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
+	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
+	.r419cb8 = gf100_grctx_generate_r419cb8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
index 7df398b53f8f..f5cca5e6a4f2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf110.c
@@ -314,15 +314,12 @@ gf110_grctx_init_setup_0[] = {
 };
 
 static const struct gf100_gr_pack
-gf110_grctx_pack_gpc[] = {
+gf110_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gf100_grctx_init_prop_0 },
 	{ gf100_grctx_init_gpc_unk_1 },
 	{ gf110_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
-	{ gf100_grctx_init_crstr_0 },
-	{ gf100_grctx_init_gpm_0 },
-	{ gf100_grctx_init_gcc_0 },
 	{}
 };
 
@@ -335,7 +332,8 @@ gf110_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gf100_grctx_generate_unkn,
 	.hub   = gf100_grctx_pack_hub,
-	.gpc   = gf110_grctx_pack_gpc,
+	.gpc_0 = gf110_grctx_pack_gpc_0,
+	.gpc_1 = gf100_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gf100_grctx_pack_tpc,
 	.icmd  = gf110_grctx_pack_icmd,
@@ -347,4 +345,11 @@ gf110_grctx = {
 	.attrib = gf100_grctx_generate_attrib,
 	.attrib_nr_max = 0x324,
 	.attrib_nr = 0x218,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.r4060a8 = gf100_grctx_generate_r4060a8,
+	.rop_mapping = gf100_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
+	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
+	.r419cb8 = gf100_grctx_generate_r419cb8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
index 19301d88577d..276c282d19aa 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf117.c
@@ -84,12 +84,17 @@ gf117_grctx_init_setup_0[] = {
 };
 
 static const struct gf100_gr_pack
-gf117_grctx_pack_gpc[] = {
+gf117_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gf119_grctx_init_prop_0 },
 	{ gf119_grctx_init_gpc_unk_1 },
 	{ gf117_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
+	{}
+};
+
+const struct gf100_gr_pack
+gf117_grctx_pack_gpc_1[] = {
 	{ gf119_grctx_init_crstr_0 },
 	{ gf108_grctx_init_gpm_0 },
 	{ gf100_grctx_init_gcc_0 },
@@ -180,6 +185,62 @@ gf117_grctx_pack_ppc[] = {
  ******************************************************************************/
 
 void
+gf117_grctx_generate_dist_skip_table(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	int i;
+
+	for (i = 0; i < 8; i++)
+		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
+}
+
+void
+gf117_grctx_generate_rop_mapping(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	u32 data[6] = {}, data2[2] = {};
+	u8  shift, ntpcv;
+	int i;
+
+	/* Pack tile map into register format. */
+	for (i = 0; i < 32; i++)
+		data[i / 6] |= (gr->tile[i] & 0x07) << ((i % 6) * 5);
+
+	/* Magic. */
+	shift = 0;
+	ntpcv = gr->tpc_total;
+	while (!(ntpcv & (1 << 4))) {
+		ntpcv <<= 1;
+		shift++;
+	}
+
+	data2[0]  = (ntpcv << 16);
+	data2[0] |= (shift << 21);
+	data2[0] |= (((1 << (0 + 5)) % ntpcv) << 24);
+	for (i = 1; i < 7; i++)
+		data2[1] |= ((1 << (i + 5)) % ntpcv) << ((i - 1) * 5);
+
+	/* GPC_BROADCAST */
+	nvkm_wr32(device, 0x418bb8, (gr->tpc_total << 8) |
+				     gr->screen_tile_row_offset);
+	for (i = 0; i < 6; i++)
+		nvkm_wr32(device, 0x418b08 + (i * 4), data[i]);
+
+	/* GPC_BROADCAST.TP_BROADCAST */
+	nvkm_wr32(device, 0x41bfd0, (gr->tpc_total << 8) |
+				     gr->screen_tile_row_offset | data2[0]);
+	nvkm_wr32(device, 0x41bfe4, data2[1]);
+	for (i = 0; i < 6; i++)
+		nvkm_wr32(device, 0x41bf00 + (i * 4), data[i]);
+
+	/* UNK78xx */
+	nvkm_wr32(device, 0x4078bc, (gr->tpc_total << 8) |
+				     gr->screen_tile_row_offset);
+	for (i = 0; i < 6; i++)
+		nvkm_wr32(device, 0x40780c + (i * 4), data[i]);
+}
+
+void
 gf117_grctx_generate_attrib(struct gf100_grctx *info)
 {
 	struct gf100_gr *gr = info->gr;
@@ -217,50 +278,13 @@ gf117_grctx_generate_attrib(struct gf100_grctx *info)
 	}
 }
 
-static void
-gf117_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const struct gf100_grctx_func *grctx = gr->func->grctx;
-	u32 idle_timeout;
-	int i;
-
-	nvkm_mc_unk260(device, 0);
-
-	gf100_gr_mmio(gr, grctx->hub);
-	gf100_gr_mmio(gr, grctx->gpc);
-	gf100_gr_mmio(gr, grctx->zcull);
-	gf100_gr_mmio(gr, grctx->tpc);
-	gf100_gr_mmio(gr, grctx->ppc);
-
-	idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
-
-	grctx->bundle(info);
-	grctx->pagepool(info);
-	grctx->attrib(info);
-	grctx->unkn(gr);
-
-	gf100_grctx_generate_tpcid(gr);
-	gf100_grctx_generate_r406028(gr);
-	gf100_grctx_generate_r4060a8(gr);
-	gk104_grctx_generate_r418bb8(gr);
-	gf100_grctx_generate_r406800(gr);
-
-	for (i = 0; i < 8; i++)
-		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
-
-	gf100_gr_icmd(gr, grctx->icmd);
-	nvkm_wr32(device, 0x404154, idle_timeout);
-	gf100_gr_mthd(gr, grctx->mthd);
-	nvkm_mc_unk260(device, 1);
-}
-
 const struct gf100_grctx_func
 gf117_grctx = {
-	.main  = gf117_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gf117_grctx_pack_hub,
-	.gpc   = gf117_grctx_pack_gpc,
+	.gpc_0 = gf117_grctx_pack_gpc_0,
+	.gpc_1 = gf117_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gf117_grctx_pack_tpc,
 	.ppc   = gf117_grctx_pack_ppc,
@@ -275,4 +299,12 @@ gf117_grctx = {
 	.attrib_nr = 0x218,
 	.alpha_nr_max = 0x7ff,
 	.alpha_nr = 0x324,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.r4060a8 = gf100_grctx_generate_r4060a8,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
+	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
+	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
+	.r419cb8 = gf100_grctx_generate_r419cb8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
index 605185b078be..0cfe46366af6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf119.c
@@ -431,15 +431,12 @@ gf119_grctx_init_crstr_0[] = {
 };
 
 static const struct gf100_gr_pack
-gf119_grctx_pack_gpc[] = {
+gf119_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gf119_grctx_init_prop_0 },
 	{ gf119_grctx_init_gpc_unk_1 },
 	{ gf119_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
-	{ gf119_grctx_init_crstr_0 },
-	{ gf108_grctx_init_gpm_0 },
-	{ gf100_grctx_init_gcc_0 },
 	{}
 };
 
@@ -503,7 +500,8 @@ gf119_grctx = {
 	.main  = gf100_grctx_generate_main,
 	.unkn  = gf108_grctx_generate_unkn,
 	.hub   = gf119_grctx_pack_hub,
-	.gpc   = gf119_grctx_pack_gpc,
+	.gpc_0 = gf119_grctx_pack_gpc_0,
+	.gpc_1 = gf117_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gf119_grctx_pack_tpc,
 	.icmd  = gf119_grctx_pack_icmd,
@@ -517,4 +515,11 @@ gf119_grctx = {
 	.attrib_nr = 0x218,
 	.alpha_nr_max = 0x324,
 	.alpha_nr = 0x218,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.r4060a8 = gf100_grctx_generate_r4060a8,
+	.rop_mapping = gf100_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gf100_grctx_generate_alpha_beta_tables,
+	.max_ways_evict = gf100_grctx_generate_max_ways_evict,
+	.r419cb8 = gf100_grctx_generate_r419cb8,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
index 825c8fd500bc..304e9d268bad 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c
@@ -739,13 +739,18 @@ gk104_grctx_init_gpm_0[] = {
 	{}
 };
 
-const struct gf100_gr_pack
-gk104_grctx_pack_gpc[] = {
+static const struct gf100_gr_pack
+gk104_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gf119_grctx_init_prop_0 },
 	{ gf119_grctx_init_gpc_unk_1 },
 	{ gk104_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
+	{}
+};
+
+static const struct gf100_gr_pack
+gk104_grctx_pack_gpc_1[] = {
 	{ gf119_grctx_init_crstr_0 },
 	{ gk104_grctx_init_gpm_0 },
 	{ gf100_grctx_init_gcc_0 },
@@ -841,6 +846,32 @@ gk104_grctx_pack_ppc[] = {
  ******************************************************************************/
 
 void
+gk104_grctx_generate_r418800(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	/*XXX: Not real sure where to apply these, there doesn't seem
+	 *     to be any pattern to which chipsets it's done on.
+	 *
+	 *     Perhaps a VBIOS tweak?
+	 */
+	if (0) {
+		nvkm_mask(device, 0x418800, 0x00200000, 0x00200000);
+		nvkm_mask(device, 0x41be10, 0x00800000, 0x00800000);
+	}
+}
+
+void
+gk104_grctx_generate_patch_ltc(struct gf100_grctx *info)
+{
+	struct nvkm_device *device = info->gr->base.engine.subdev.device;
+	u32 data0 = nvkm_rd32(device, 0x17e91c);
+	u32 data1 = nvkm_rd32(device, 0x17e920);
+	/*XXX: Figure out how to modify this correctly! */
+	mmio_wr32(info, 0x17e91c, data0);
+	mmio_wr32(info, 0x17e920, data1);
+}
+
+void
 gk104_grctx_generate_bundle(struct gf100_grctx *info)
 {
 	const struct gf100_grctx_func *grctx = info->gr->func->grctx;
@@ -881,114 +912,74 @@ gk104_grctx_generate_unkn(struct gf100_gr *gr)
 	nvkm_mask(device, 0x419c00, 0x00000008, 0x00000008);
 }
 
-void
-gk104_grctx_generate_r418bb8(struct gf100_gr *gr)
+static void
+gk104_grctx_generate_r419f78(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	u32 data[6] = {}, data2[2] = {};
-	u8  tpcnr[GPC_MAX];
-	u8  shift, ntpcv;
-	int gpc, tpc, i;
-
-	/* calculate first set of magics */
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-
-	gpc = -1;
-	for (tpc = 0; tpc < gr->tpc_total; tpc++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpcnr[gpc]--;
-
-		data[tpc / 6] |= gpc << ((tpc % 6) * 5);
-	}
-
-	for (; tpc < 32; tpc++)
-		data[tpc / 6] |= 7 << ((tpc % 6) * 5);
-
-	/* and the second... */
-	shift = 0;
-	ntpcv = gr->tpc_total;
-	while (!(ntpcv & (1 << 4))) {
-		ntpcv <<= 1;
-		shift++;
-	}
-
-	data2[0]  = (ntpcv << 16);
-	data2[0] |= (shift << 21);
-	data2[0] |= (((1 << (0 + 5)) % ntpcv) << 24);
-	for (i = 1; i < 7; i++)
-		data2[1] |= ((1 << (i + 5)) % ntpcv) << ((i - 1) * 5);
-
-	/* GPC_BROADCAST */
-	nvkm_wr32(device, 0x418bb8, (gr->tpc_total << 8) |
-				     gr->screen_tile_row_offset);
-	for (i = 0; i < 6; i++)
-		nvkm_wr32(device, 0x418b08 + (i * 4), data[i]);
-
-	/* GPC_BROADCAST.TP_BROADCAST */
-	nvkm_wr32(device, 0x41bfd0, (gr->tpc_total << 8) |
-				     gr->screen_tile_row_offset | data2[0]);
-	nvkm_wr32(device, 0x41bfe4, data2[1]);
-	for (i = 0; i < 6; i++)
-		nvkm_wr32(device, 0x41bf00 + (i * 4), data[i]);
-
-	/* UNK78xx */
-	nvkm_wr32(device, 0x4078bc, (gr->tpc_total << 8) |
-				     gr->screen_tile_row_offset);
-	for (i = 0; i < 6; i++)
-		nvkm_wr32(device, 0x40780c + (i * 4), data[i]);
+	nvkm_mask(device, 0x419f78, 0x00000001, 0x00000000);
 }
 
 void
-gk104_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
+gk104_grctx_generate_gpc_tpc_nr(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const struct gf100_grctx_func *grctx = gr->func->grctx;
-	u32 idle_timeout;
-	int i;
-
-	nvkm_mc_unk260(device, 0);
-
-	gf100_gr_mmio(gr, grctx->hub);
-	gf100_gr_mmio(gr, grctx->gpc);
-	gf100_gr_mmio(gr, grctx->zcull);
-	gf100_gr_mmio(gr, grctx->tpc);
-	gf100_gr_mmio(gr, grctx->ppc);
-
-	idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
-
-	grctx->bundle(info);
-	grctx->pagepool(info);
-	grctx->attrib(info);
-	grctx->unkn(gr);
-
-	gf100_grctx_generate_tpcid(gr);
-	gf100_grctx_generate_r406028(gr);
-	gk104_grctx_generate_r418bb8(gr);
-	gf100_grctx_generate_r406800(gr);
-
-	for (i = 0; i < 8; i++)
-		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
-
 	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
-	nvkm_mask(device, 0x419f78, 0x00000001, 0x00000000);
-
-	gf100_gr_icmd(gr, grctx->icmd);
-	nvkm_wr32(device, 0x404154, idle_timeout);
-	gf100_gr_mthd(gr, grctx->mthd);
-	nvkm_mc_unk260(device, 1);
+}
 
-	nvkm_mask(device, 0x418800, 0x00200000, 0x00200000);
-	nvkm_mask(device, 0x41be10, 0x00800000, 0x00800000);
+void
+gk104_grctx_generate_alpha_beta_tables(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	int i, j, gpc, ppc;
+
+	for (i = 0; i < 32; i++) {
+		u32 atarget = max_t(u32, gr->tpc_total * i / 32, 1);
+		u32 btarget = gr->tpc_total - atarget;
+		bool alpha = atarget < btarget;
+		u64 amask = 0, bmask = 0;
+
+		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+			for (ppc = 0; ppc < gr->func->ppc_nr; ppc++) {
+				u32 ppc_tpcs = gr->ppc_tpc_nr[gpc][ppc];
+				u32 abits, bbits, pmask;
+
+				if (alpha) {
+					abits = atarget ? ppc_tpcs : 0;
+					bbits = ppc_tpcs - abits;
+				} else {
+					bbits = btarget ? ppc_tpcs : 0;
+					abits = ppc_tpcs - bbits;
+				}
+
+				pmask = gr->ppc_tpc_mask[gpc][ppc];
+				while (ppc_tpcs-- > abits)
+					pmask &= pmask - 1;
+				amask |= (u64)pmask << (gpc * 8);
+
+				pmask ^= gr->ppc_tpc_mask[gpc][ppc];
+				bmask |= (u64)pmask << (gpc * 8);
+
+				atarget -= min(abits, atarget);
+				btarget -= min(bbits, btarget);
+				if ((abits > 0) || (bbits > 0))
+					alpha = !alpha;
+			}
+		}
+
+		for (j = 0; j < gr->gpc_nr; j += 4, amask >>= 32, bmask >>= 32) {
+			nvkm_wr32(device, 0x406800 + (i * 0x20) + j, amask);
+			nvkm_wr32(device, 0x406c00 + (i * 0x20) + j, bmask);
+		}
+	}
 }
 
 const struct gf100_grctx_func
 gk104_grctx = {
-	.main  = gk104_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gk104_grctx_pack_hub,
-	.gpc   = gk104_grctx_pack_gpc,
+	.gpc_0 = gk104_grctx_pack_gpc_0,
+	.gpc_1 = gk104_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gk104_grctx_pack_tpc,
 	.ppc   = gk104_grctx_pack_ppc,
@@ -1005,4 +996,13 @@ gk104_grctx = {
 	.attrib_nr = 0x218,
 	.alpha_nr_max = 0x7ff,
 	.alpha_nr = 0x648,
+	.patch_ltc = gk104_grctx_generate_patch_ltc,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
+	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.r419f78 = gk104_grctx_generate_r419f78,
+	.r418800 = gk104_grctx_generate_r418800,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
index 7b95ec2fe453..86547cfc38dc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110.c
@@ -704,12 +704,17 @@ gk110_grctx_init_gpc_unk_2[] = {
 };
 
 const struct gf100_gr_pack
-gk110_grctx_pack_gpc[] = {
+gk110_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gf119_grctx_init_prop_0 },
 	{ gf119_grctx_init_gpc_unk_1 },
 	{ gk110_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
+	{}
+};
+
+const struct gf100_gr_pack
+gk110_grctx_pack_gpc_1[] = {
 	{ gf119_grctx_init_crstr_0 },
 	{ gk104_grctx_init_gpm_0 },
 	{ gk110_grctx_init_gpc_unk_2 },
@@ -808,12 +813,20 @@ gk110_grctx_pack_ppc[] = {
  * PGRAPH context implementation
  ******************************************************************************/
 
+void
+gk110_grctx_generate_r419eb0(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419eb0, 0x00001000, 0x00001000);
+}
+
 const struct gf100_grctx_func
 gk110_grctx = {
-	.main  = gk104_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gk110_grctx_pack_hub,
-	.gpc   = gk110_grctx_pack_gpc,
+	.gpc_0 = gk110_grctx_pack_gpc_0,
+	.gpc_1 = gk110_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gk110_grctx_pack_tpc,
 	.ppc   = gk110_grctx_pack_ppc,
@@ -830,4 +843,13 @@ gk110_grctx = {
 	.attrib_nr = 0x218,
 	.alpha_nr_max = 0x7ff,
 	.alpha_nr = 0x648,
+	.patch_ltc = gk104_grctx_generate_patch_ltc,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
+	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.r418800 = gk104_grctx_generate_r418800,
+	.r419eb0 = gk110_grctx_generate_r419eb0,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
index 048b1152da44..ebb947bd1446 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk110b.c
@@ -71,10 +71,11 @@ gk110b_grctx_pack_tpc[] = {
 
 const struct gf100_grctx_func
 gk110b_grctx = {
-	.main  = gk104_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gk110_grctx_pack_hub,
-	.gpc   = gk110_grctx_pack_gpc,
+	.gpc_0 = gk110_grctx_pack_gpc_0,
+	.gpc_1 = gk110_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gk110b_grctx_pack_tpc,
 	.ppc   = gk110_grctx_pack_ppc,
@@ -91,4 +92,13 @@ gk110b_grctx = {
 	.attrib_nr = 0x218,
 	.alpha_nr_max = 0x7ff,
 	.alpha_nr = 0x648,
+	.patch_ltc = gk104_grctx_generate_patch_ltc,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
+	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.r418800 = gk104_grctx_generate_r418800,
+	.r419eb0 = gk110_grctx_generate_r419eb0,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
index 67b7a1b43617..4d40512b5c99 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk208.c
@@ -443,12 +443,17 @@ gk208_grctx_init_gpm_0[] = {
 };
 
 static const struct gf100_gr_pack
-gk208_grctx_pack_gpc[] = {
+gk208_grctx_pack_gpc_0[] = {
 	{ gf100_grctx_init_gpc_unk_0 },
 	{ gk208_grctx_init_prop_0 },
 	{ gk208_grctx_init_gpc_unk_1 },
 	{ gk208_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
+	{}
+};
+
+static const struct gf100_gr_pack
+gk208_grctx_pack_gpc_1[] = {
 	{ gk208_grctx_init_crstr_0 },
 	{ gk208_grctx_init_gpm_0 },
 	{ gk110_grctx_init_gpc_unk_2 },
@@ -532,10 +537,11 @@ gk208_grctx_pack_ppc[] = {
 
 const struct gf100_grctx_func
 gk208_grctx = {
-	.main  = gk104_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gk208_grctx_pack_hub,
-	.gpc   = gk208_grctx_pack_gpc,
+	.gpc_0 = gk208_grctx_pack_gpc_0,
+	.gpc_1 = gk208_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gk208_grctx_pack_tpc,
 	.ppc   = gk208_grctx_pack_ppc,
@@ -552,4 +558,12 @@ gk208_grctx = {
 	.attrib_nr = 0x218,
 	.alpha_nr_max = 0x7ff,
 	.alpha_nr = 0x648,
+	.patch_ltc = gk104_grctx_generate_patch_ltc,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
+	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.r418800 = gk104_grctx_generate_r418800,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
index da7c35a6a3d2..896d473dcc0f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk20a.c
@@ -42,10 +42,7 @@ gk20a_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	grctx->unkn(gr);
 
-	gf100_grctx_generate_tpcid(gr);
-	gf100_grctx_generate_r406028(gr);
-	gk104_grctx_generate_r418bb8(gr);
-	gf100_grctx_generate_r406800(gr);
+	gf100_grctx_generate_floorsweep(gr);
 
 	for (i = 0; i < 8; i++)
 		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
@@ -82,4 +79,8 @@ gk20a_grctx = {
 	.attrib_nr = 0x240,
 	.alpha_nr_max = 0x648 + (0x648 / 2),
 	.alpha_nr = 0x648,
+	.sm_id = gf100_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
index 9b43d4ce3eaa..0b3964e6b36e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c
@@ -744,12 +744,17 @@ gm107_grctx_init_gpc_unk_2[] = {
 };
 
 static const struct gf100_gr_pack
-gm107_grctx_pack_gpc[] = {
+gm107_grctx_pack_gpc_0[] = {
 	{ gm107_grctx_init_gpc_unk_0 },
 	{ gk208_grctx_init_prop_0 },
 	{ gm107_grctx_init_gpc_unk_1 },
 	{ gm107_grctx_init_setup_0 },
 	{ gf100_grctx_init_zcull_0 },
+	{}
+};
+
+static const struct gf100_gr_pack
+gm107_grctx_pack_gpc_1[] = {
 	{ gk208_grctx_init_crstr_0 },
 	{ gk104_grctx_init_gpm_0 },
 	{ gm107_grctx_init_gpc_unk_2 },
@@ -860,6 +865,16 @@ gm107_grctx_pack_ppc[] = {
  * PGRAPH context implementation
  ******************************************************************************/
 
+static void
+gm107_grctx_generate_r419e00(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419e00, 0x00808080, 0x00808080);
+	nvkm_mask(device, 0x419ccc, 0x80000000, 0x80000000);
+	nvkm_mask(device, 0x419f80, 0x80000000, 0x80000000);
+	nvkm_mask(device, 0x419f88, 0x80000000, 0x80000000);
+}
+
 void
 gm107_grctx_generate_bundle(struct gf100_grctx *info)
 {
@@ -931,75 +946,27 @@ gm107_grctx_generate_attrib(struct gf100_grctx *info)
 }
 
 static void
-gm107_grctx_generate_tpcid(struct gf100_gr *gr)
+gm107_grctx_generate_r406500(struct gf100_gr *gr)
 {
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	int gpc, tpc, id;
-
-	for (tpc = 0, id = 0; tpc < 4; tpc++) {
-		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-			if (tpc < gr->tpc_nr[gpc]) {
-				nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x698), id);
-				nvkm_wr32(device, GPC_UNIT(gpc, 0x0c10 + tpc * 4), id);
-				nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x088), id);
-				id++;
-			}
-
-			nvkm_wr32(device, GPC_UNIT(gpc, 0x0c08), gr->tpc_nr[gpc]);
-			nvkm_wr32(device, GPC_UNIT(gpc, 0x0c8c), gr->tpc_nr[gpc]);
-		}
-	}
+	nvkm_wr32(gr->base.engine.subdev.device, 0x406500, 0x00000001);
 }
 
-static void
-gm107_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
+void
+gm107_grctx_generate_sm_id(struct gf100_gr *gr, int gpc, int tpc, int sm)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const struct gf100_grctx_func *grctx = gr->func->grctx;
-	u32 idle_timeout;
-	int i;
-
-	gf100_gr_mmio(gr, grctx->hub);
-	gf100_gr_mmio(gr, grctx->gpc);
-	gf100_gr_mmio(gr, grctx->zcull);
-	gf100_gr_mmio(gr, grctx->tpc);
-	gf100_gr_mmio(gr, grctx->ppc);
-
-	idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
-
-	grctx->bundle(info);
-	grctx->pagepool(info);
-	grctx->attrib(info);
-	grctx->unkn(gr);
-
-	gm107_grctx_generate_tpcid(gr);
-	gf100_grctx_generate_r406028(gr);
-	gk104_grctx_generate_r418bb8(gr);
-	gf100_grctx_generate_r406800(gr);
-
-	nvkm_wr32(device, 0x4064d0, 0x00000001);
-	for (i = 1; i < 8; i++)
-		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
-	nvkm_wr32(device, 0x406500, 0x00000001);
-
-	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
-
-	gf100_gr_icmd(gr, grctx->icmd);
-	nvkm_wr32(device, 0x404154, idle_timeout);
-	gf100_gr_mthd(gr, grctx->mthd);
-
-	nvkm_mask(device, 0x419e00, 0x00808080, 0x00808080);
-	nvkm_mask(device, 0x419ccc, 0x80000000, 0x80000000);
-	nvkm_mask(device, 0x419f80, 0x80000000, 0x80000000);
-	nvkm_mask(device, 0x419f88, 0x80000000, 0x80000000);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x698), sm);
+	nvkm_wr32(device, GPC_UNIT(gpc, 0x0c10 + tpc * 4), sm);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x088), sm);
 }
 
 const struct gf100_grctx_func
 gm107_grctx = {
-	.main  = gm107_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.hub   = gm107_grctx_pack_hub,
-	.gpc   = gm107_grctx_pack_gpc,
+	.gpc_0 = gm107_grctx_pack_gpc_0,
+	.gpc_1 = gm107_grctx_pack_gpc_1,
 	.zcull = gf100_grctx_pack_zcull,
 	.tpc   = gm107_grctx_pack_tpc,
 	.ppc   = gm107_grctx_pack_ppc,
@@ -1016,4 +983,12 @@ gm107_grctx = {
 	.attrib_nr = 0xaa0,
 	.alpha_nr_max = 0x1800,
 	.alpha_nr = 0x1000,
+	.sm_id = gm107_grctx_generate_sm_id,
+	.tpc_nr = gf100_grctx_generate_tpc_nr,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.alpha_beta_tables = gk104_grctx_generate_alpha_beta_tables,
+	.dist_skip_table = gf117_grctx_generate_dist_skip_table,
+	.r406500 = gm107_grctx_generate_r406500,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.r419e00 = gm107_grctx_generate_r419e00,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
index db209d33f486..013d05a0f0f6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm200.c
@@ -28,47 +28,34 @@
  ******************************************************************************/
 
 void
-gm200_grctx_generate_tpcid(struct gf100_gr *gr)
+gm200_grctx_generate_r419a3c(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	int gpc, tpc, id;
+	nvkm_mask(device, 0x419a3c, 0x00000014, 0x00000000);
+}
 
-	for (tpc = 0, id = 0; tpc < TPC_MAX_PER_GPC; tpc++) {
-		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-			if (tpc < gr->tpc_nr[gpc]) {
-				nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x698), id);
-				nvkm_wr32(device, GPC_UNIT(gpc, 0x0c10 + tpc * 4), id);
-				nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x088), id);
-				id++;
-			}
-		}
-	}
+static void
+gm200_grctx_generate_r418e94(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x418e94, 0xffffffff, 0xc4230000);
+	nvkm_mask(device, 0x418e4c, 0xffffffff, 0x70000000);
 }
 
 void
-gm200_grctx_generate_405b60(struct gf100_gr *gr)
+gm200_grctx_generate_smid_config(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const u32 dist_nr = DIV_ROUND_UP(gr->tpc_total, 4);
 	u32 dist[TPC_MAX / 4] = {};
 	u32 gpcs[GPC_MAX] = {};
-	u8  tpcnr[GPC_MAX];
-	int tpc, gpc, i;
+	u8  sm, i;
 
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-
-	/* won't result in the same distribution as the binary driver where
-	 * some of the gpcs have more tpcs than others, but this shall do
-	 * for the moment.  the code for earlier gpus has this issue too.
-	 */
-	for (gpc = -1, i = 0; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while(!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		dist[i / 4] |= ((gpc << 4) | tpc) << ((i % 4) * 8);
-		gpcs[gpc] |= i << (tpc * 8);
+	for (sm = 0; sm < gr->sm_nr; sm++) {
+		const u8 gpc = gr->sm[sm].gpc;
+		const u8 tpc = gr->sm[sm].tpc;
+		dist[sm / 4] |= ((gpc << 4) | tpc) << ((sm % 4) * 8);
+		gpcs[gpc] |= sm << (tpc * 8);
 	}
 
 	for (i = 0; i < dist_nr; i++)
@@ -77,50 +64,46 @@ gm200_grctx_generate_405b60(struct gf100_gr *gr)
 		nvkm_wr32(device, 0x405ba0 + (i * 4), gpcs[i]);
 }
 
-static void
-gm200_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
+void
+gm200_grctx_generate_tpc_mask(struct gf100_gr *gr)
 {
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const struct gf100_grctx_func *grctx = gr->func->grctx;
-	u32 idle_timeout, tmp;
-	int i;
-
-	gf100_gr_mmio(gr, gr->fuc_sw_ctx);
-
-	idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
-
-	grctx->bundle(info);
-	grctx->pagepool(info);
-	grctx->attrib(info);
-	grctx->unkn(gr);
-
-	gm200_grctx_generate_tpcid(gr);
-	gf100_grctx_generate_r406028(gr);
-	gk104_grctx_generate_r418bb8(gr);
-
-	for (i = 0; i < 8; i++)
-		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
-	nvkm_wr32(device, 0x406500, 0x00000000);
-
-	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
-
+	u32 tmp, i;
 	for (tmp = 0, i = 0; i < gr->gpc_nr; i++)
-		tmp |= ((1 << gr->tpc_nr[i]) - 1) << (i * 4);
-	nvkm_wr32(device, 0x4041c4, tmp);
+		tmp |= ((1 << gr->tpc_nr[i]) - 1) << (i * gr->func->tpc_nr);
+	nvkm_wr32(gr->base.engine.subdev.device, 0x4041c4, tmp);
+}
 
-	gm200_grctx_generate_405b60(gr);
+void
+gm200_grctx_generate_r406500(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x406500, 0x00000000);
+}
 
-	gf100_gr_icmd(gr, gr->fuc_bundle);
-	nvkm_wr32(device, 0x404154, idle_timeout);
-	gf100_gr_mthd(gr, gr->fuc_method);
+void
+gm200_grctx_generate_dist_skip_table(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	u32 data[8] = {};
+	int gpc, ppc, i;
+
+	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+		for (ppc = 0; ppc < gr->ppc_nr[gpc]; ppc++) {
+			u8 ppc_tpcs = gr->ppc_tpc_nr[gpc][ppc];
+			u8 ppc_tpcm = gr->ppc_tpc_mask[gpc][ppc];
+			while (ppc_tpcs-- > gr->ppc_tpc_min)
+				ppc_tpcm &= ppc_tpcm - 1;
+			ppc_tpcm ^= gr->ppc_tpc_mask[gpc][ppc];
+			((u8 *)data)[gpc] |= ppc_tpcm;
+		}
+	}
 
-	nvkm_mask(device, 0x418e94, 0xffffffff, 0xc4230000);
-	nvkm_mask(device, 0x418e4c, 0xffffffff, 0x70000000);
+	for (i = 0; i < ARRAY_SIZE(data); i++)
+		nvkm_wr32(device, 0x4064d0 + (i * 0x04), data[i]);
 }
 
 const struct gf100_grctx_func
 gm200_grctx = {
-	.main  = gm200_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.bundle = gm107_grctx_generate_bundle,
 	.bundle_size = 0x3000,
@@ -133,4 +116,13 @@ gm200_grctx = {
 	.attrib_nr = 0x400,
 	.alpha_nr_max = 0x1800,
 	.alpha_nr = 0x1000,
+	.sm_id = gm107_grctx_generate_sm_id,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
+	.r406500 = gm200_grctx_generate_r406500,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.tpc_mask = gm200_grctx_generate_tpc_mask,
+	.smid_config = gm200_grctx_generate_smid_config,
+	.r418e94 = gm200_grctx_generate_r418e94,
+	.r419a3c = gm200_grctx_generate_r419a3c,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
index e5702e3e0a5a..a1d9e114ebeb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm20b.c
@@ -22,20 +22,6 @@
 #include "ctxgf100.h"
 
 static void
-gm20b_grctx_generate_r406028(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	u32 tpc_per_gpc = 0;
-	int i;
-
-	for (i = 0; i < gr->gpc_nr; i++)
-		tpc_per_gpc |= gr->tpc_nr[i] << (4 * i);
-
-	nvkm_wr32(device, 0x406028, tpc_per_gpc);
-	nvkm_wr32(device, 0x405870, tpc_per_gpc);
-}
-
-static void
 gm20b_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -53,9 +39,7 @@ gm20b_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 
 	grctx->unkn(gr);
 
-	gm200_grctx_generate_tpcid(gr);
-	gm20b_grctx_generate_r406028(gr);
-	gk104_grctx_generate_r418bb8(gr);
+	gf100_grctx_generate_floorsweep(gr);
 
 	for (i = 0; i < 8; i++)
 		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
@@ -68,7 +52,7 @@ gm20b_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
 		tmp |= ((1 << gr->tpc_nr[i]) - 1) << (i * 4);
 	nvkm_wr32(device, 0x4041c4, tmp);
 
-	gm200_grctx_generate_405b60(gr);
+	gm200_grctx_generate_smid_config(gr);
 
 	gf100_gr_wait_idle(gr);
 
@@ -98,4 +82,6 @@ gm20b_grctx = {
 	.attrib_nr = 0x400,
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
+	.sm_id = gm107_grctx_generate_sm_id,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
index 88ea322d956c..0b3326262e12 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp100.c
@@ -36,7 +36,7 @@ gp100_grctx_generate_pagepool(struct gf100_grctx *info)
 	const int s = 8;
 	const int b = mmio_vram(info, grctx->pagepool_size, (1 << s), true);
 	mmio_refn(info, 0x40800c, 0x00000000, s, b);
-	mmio_wr32(info, 0x408010, 0x80000000);
+	mmio_wr32(info, 0x408010, 0x8007d800);
 	mmio_refn(info, 0x419004, 0x00000000, s, b);
 	mmio_wr32(info, 0x419008, 0x00000000);
 }
@@ -48,14 +48,17 @@ gp100_grctx_generate_attrib(struct gf100_grctx *info)
 	const struct gf100_grctx_func *grctx = gr->func->grctx;
 	const u32  alpha = grctx->alpha_nr;
 	const u32 attrib = grctx->attrib_nr;
-	const u32 pertpc = 0x20 * (grctx->attrib_nr_max + grctx->alpha_nr_max);
-	const u32   size = roundup(gr->tpc_total * pertpc, 0x80);
 	const int s = 12;
-	const int b = mmio_vram(info, size, (1 << s), false);
 	const int max_batches = 0xffff;
+	u32 size = grctx->alpha_nr_max * gr->tpc_total;
 	u32 ao = 0;
-	u32 bo = ao + grctx->alpha_nr_max * gr->tpc_total;
-	int gpc, ppc, n = 0;
+	u32 bo = ao + size;
+	int gpc, ppc, b, n = 0;
+
+	for (gpc = 0; gpc < gr->gpc_nr; gpc++)
+		size += grctx->attrib_nr_max * gr->ppc_nr[gpc] * gr->ppc_tpc_max;
+	size = ((size * 0x20) + 128) & ~127;
+	b = mmio_vram(info, size, (1 << s), false);
 
 	mmio_refn(info, 0x418810, 0x80000000, s, b);
 	mmio_refn(info, 0x419848, 0x10000000, s, b);
@@ -69,7 +72,7 @@ gp100_grctx_generate_attrib(struct gf100_grctx *info)
 	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
 		for (ppc = 0; ppc < gr->ppc_nr[gpc]; ppc++, n++) {
 			const u32 as =  alpha * gr->ppc_tpc_nr[gpc][ppc];
-			const u32 bs = attrib * gr->ppc_tpc_nr[gpc][ppc];
+			const u32 bs = attrib * gr->ppc_tpc_max;
 			const u32 u = 0x418ea0 + (n * 0x04);
 			const u32 o = PPC_UNIT(gpc, ppc, 0);
 			if (!(gr->ppc_mask[gpc] & (1 << ppc)))
@@ -77,7 +80,7 @@ gp100_grctx_generate_attrib(struct gf100_grctx *info)
 			mmio_wr32(info, o + 0xc0, bs);
 			mmio_wr32(info, o + 0xf4, bo);
 			mmio_wr32(info, o + 0xf0, bs);
-			bo += grctx->attrib_nr_max * gr->ppc_tpc_nr[gpc][ppc];
+			bo += grctx->attrib_nr_max * gr->ppc_tpc_max;
 			mmio_wr32(info, o + 0xe4, as);
 			mmio_wr32(info, o + 0xf8, ao);
 			ao += grctx->alpha_nr_max * gr->ppc_tpc_nr[gpc][ppc];
@@ -89,79 +92,30 @@ gp100_grctx_generate_attrib(struct gf100_grctx *info)
 	mmio_wr32(info, 0x41befc, 0x00000000);
 }
 
-static void
-gp100_grctx_generate_405b60(struct gf100_gr *gr)
+void
+gp100_grctx_generate_smid_config(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	const u32 dist_nr = DIV_ROUND_UP(gr->tpc_total, 4);
-	u32 dist[TPC_MAX / 4] = {};
-	u32 gpcs[GPC_MAX * 2] = {};
-	u8  tpcnr[GPC_MAX];
-	int tpc, gpc, i;
-
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-
-	/* won't result in the same distribution as the binary driver where
-	 * some of the gpcs have more tpcs than others, but this shall do
-	 * for the moment.  the code for earlier gpus has this issue too.
-	 */
-	for (gpc = -1, i = 0; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while(!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		dist[i / 4] |= ((gpc << 4) | tpc) << ((i % 4) * 8);
-		gpcs[gpc + (gr->gpc_nr * (tpc / 4))] |= i << (tpc * 8);
+	u32 dist[TPC_MAX / 4] = {}, gpcs[16] = {};
+	u8  sm, i;
+
+	for (sm = 0; sm < gr->sm_nr; sm++) {
+		const u8 gpc = gr->sm[sm].gpc;
+		const u8 tpc = gr->sm[sm].tpc;
+		dist[sm / 4] |= ((gpc << 4) | tpc) << ((sm % 4) * 8);
+		gpcs[gpc + (gr->func->gpc_nr * (tpc / 4))] |= sm << ((tpc % 4) * 8);
 	}
 
 	for (i = 0; i < dist_nr; i++)
 		nvkm_wr32(device, 0x405b60 + (i * 4), dist[i]);
-	for (i = 0; i < gr->gpc_nr * 2; i++)
+	for (i = 0; i < ARRAY_SIZE(gpcs); i++)
 		nvkm_wr32(device, 0x405ba0 + (i * 4), gpcs[i]);
 }
 
-void
-gp100_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const struct gf100_grctx_func *grctx = gr->func->grctx;
-	u32 idle_timeout, tmp;
-	int i;
-
-	gf100_gr_mmio(gr, gr->fuc_sw_ctx);
-
-	idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
-
-	grctx->pagepool(info);
-	grctx->bundle(info);
-	grctx->attrib(info);
-	grctx->unkn(gr);
-
-	gm200_grctx_generate_tpcid(gr);
-	gf100_grctx_generate_r406028(gr);
-	gk104_grctx_generate_r418bb8(gr);
-
-	for (i = 0; i < 8; i++)
-		nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
-	nvkm_wr32(device, 0x406500, 0x00000000);
-
-	nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
-
-	for (tmp = 0, i = 0; i < gr->gpc_nr; i++)
-		tmp |= ((1 << gr->tpc_nr[i]) - 1) << (i * 5);
-	nvkm_wr32(device, 0x4041c4, tmp);
-
-	gp100_grctx_generate_405b60(gr);
-
-	gf100_gr_icmd(gr, gr->fuc_bundle);
-	nvkm_wr32(device, 0x404154, idle_timeout);
-	gf100_gr_mthd(gr, gr->fuc_method);
-}
-
 const struct gf100_grctx_func
 gp100_grctx = {
-	.main  = gp100_grctx_generate_main,
+	.main  = gf100_grctx_generate_main,
 	.unkn  = gk104_grctx_generate_unkn,
 	.bundle = gm107_grctx_generate_bundle,
 	.bundle_size = 0x3000,
@@ -174,4 +128,12 @@ gp100_grctx = {
 	.attrib_nr = 0x440,
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
+	.sm_id = gm107_grctx_generate_sm_id,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
+	.r406500 = gm200_grctx_generate_r406500,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.tpc_mask = gm200_grctx_generate_tpc_mask,
+	.smid_config = gp100_grctx_generate_smid_config,
+	.r419a3c = gm200_grctx_generate_r419a3c,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
index 7a66b4c2eb18..daee17bf7d0d 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp102.c
@@ -29,6 +29,13 @@
  * PGRAPH context implementation
  ******************************************************************************/
 
+static void
+gp102_grctx_generate_r408840(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x408840, 0x00000003, 0x00000000);
+}
+
 void
 gp102_grctx_generate_attrib(struct gf100_grctx *info)
 {
@@ -36,14 +43,18 @@ gp102_grctx_generate_attrib(struct gf100_grctx *info)
 	const struct gf100_grctx_func *grctx = gr->func->grctx;
 	const u32  alpha = grctx->alpha_nr;
 	const u32 attrib = grctx->attrib_nr;
-	const u32 pertpc = 0x20 * (grctx->attrib_nr_max + grctx->alpha_nr_max);
-	const u32   size = roundup(gr->tpc_total * pertpc, 0x80);
+	const u32   gfxp = grctx->gfxp_nr;
 	const int s = 12;
-	const int b = mmio_vram(info, size, (1 << s), false);
 	const int max_batches = 0xffff;
+	u32 size = grctx->alpha_nr_max * gr->tpc_total;
 	u32 ao = 0;
-	u32 bo = ao + grctx->alpha_nr_max * gr->tpc_total;
-	int gpc, ppc, n = 0;
+	u32 bo = ao + size;
+	int gpc, ppc, b, n = 0;
+
+	for (gpc = 0; gpc < gr->gpc_nr; gpc++)
+		size += grctx->gfxp_nr * gr->ppc_nr[gpc] * gr->ppc_tpc_max;
+	size = ((size * 0x20) + 128) & ~127;
+	b = mmio_vram(info, size, (1 << s), false);
 
 	mmio_refn(info, 0x418810, 0x80000000, s, b);
 	mmio_refn(info, 0x419848, 0x10000000, s, b);
@@ -57,17 +68,18 @@ gp102_grctx_generate_attrib(struct gf100_grctx *info)
 	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
 		for (ppc = 0; ppc < gr->ppc_nr[gpc]; ppc++, n++) {
 			const u32 as =  alpha * gr->ppc_tpc_nr[gpc][ppc];
-			const u32 bs = attrib * gr->ppc_tpc_nr[gpc][ppc];
+			const u32 bs = attrib * gr->ppc_tpc_max;
+			const u32 gs =   gfxp * gr->ppc_tpc_max;
 			const u32 u = 0x418ea0 + (n * 0x04);
 			const u32 o = PPC_UNIT(gpc, ppc, 0);
 			const u32 p = GPC_UNIT(gpc, 0xc44 + (ppc * 4));
 			if (!(gr->ppc_mask[gpc] & (1 << ppc)))
 				continue;
-			mmio_wr32(info, o + 0xc0, bs);
+			mmio_wr32(info, o + 0xc0, gs);
 			mmio_wr32(info, p, bs);
 			mmio_wr32(info, o + 0xf4, bo);
 			mmio_wr32(info, o + 0xf0, bs);
-			bo += grctx->attrib_nr_max * gr->ppc_tpc_nr[gpc][ppc];
+			bo += gs;
 			mmio_wr32(info, o + 0xe4, as);
 			mmio_wr32(info, o + 0xf8, ao);
 			ao += grctx->alpha_nr_max * gr->ppc_tpc_nr[gpc][ppc];
@@ -81,7 +93,7 @@ gp102_grctx_generate_attrib(struct gf100_grctx *info)
 
 const struct gf100_grctx_func
 gp102_grctx = {
-	.main = gp100_grctx_generate_main,
+	.main = gf100_grctx_generate_main,
 	.unkn = gk104_grctx_generate_unkn,
 	.bundle = gm107_grctx_generate_bundle,
 	.bundle_size = 0x3000,
@@ -90,8 +102,18 @@ gp102_grctx = {
 	.pagepool = gp100_grctx_generate_pagepool,
 	.pagepool_size = 0x20000,
 	.attrib = gp102_grctx_generate_attrib,
-	.attrib_nr_max = 0x5d4,
+	.attrib_nr_max = 0x4b0,
 	.attrib_nr = 0x320,
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
+	.gfxp_nr = 0xba8,
+	.sm_id = gm107_grctx_generate_sm_id,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
+	.r406500 = gm200_grctx_generate_r406500,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.tpc_mask = gm200_grctx_generate_tpc_mask,
+	.smid_config = gp100_grctx_generate_smid_config,
+	.r419a3c = gm200_grctx_generate_r419a3c,
+	.r408840 = gp102_grctx_generate_r408840,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c
new file mode 100644
index 000000000000..3b85e3d326b2
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp104.c
@@ -0,0 +1,48 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ctxgf100.h"
+
+const struct gf100_grctx_func
+gp104_grctx = {
+	.main = gf100_grctx_generate_main,
+	.unkn = gk104_grctx_generate_unkn,
+	.bundle = gm107_grctx_generate_bundle,
+	.bundle_size = 0x3000,
+	.bundle_min_gpm_fifo_depth = 0x180,
+	.bundle_token_limit = 0x900,
+	.pagepool = gp100_grctx_generate_pagepool,
+	.pagepool_size = 0x20000,
+	.attrib = gp102_grctx_generate_attrib,
+	.attrib_nr_max = 0x4b0,
+	.attrib_nr = 0x320,
+	.alpha_nr_max = 0xc00,
+	.alpha_nr = 0x800,
+	.gfxp_nr = 0xba8,
+	.sm_id = gm107_grctx_generate_sm_id,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
+	.r406500 = gm200_grctx_generate_r406500,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.tpc_mask = gm200_grctx_generate_tpc_mask,
+	.smid_config = gp100_grctx_generate_smid_config,
+	.r419a3c = gm200_grctx_generate_r419a3c,
+};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
index 8da91a0b3bd2..5060c5ee5ce0 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgp107.c
@@ -31,7 +31,7 @@
 
 const struct gf100_grctx_func
 gp107_grctx = {
-	.main = gp100_grctx_generate_main,
+	.main = gf100_grctx_generate_main,
 	.unkn = gk104_grctx_generate_unkn,
 	.bundle = gm107_grctx_generate_bundle,
 	.bundle_size = 0x3000,
@@ -44,4 +44,13 @@ gp107_grctx = {
 	.attrib_nr = 0x540,
 	.alpha_nr_max = 0xc00,
 	.alpha_nr = 0x800,
+	.gfxp_nr = 0xe94,
+	.sm_id = gm107_grctx_generate_sm_id,
+	.rop_mapping = gf117_grctx_generate_rop_mapping,
+	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
+	.r406500 = gm200_grctx_generate_r406500,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.tpc_mask = gm200_grctx_generate_tpc_mask,
+	.smid_config = gp100_grctx_generate_smid_config,
+	.r419a3c = gm200_grctx_generate_r419a3c,
 };
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgv100.c
new file mode 100644
index 000000000000..0990765ef191
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgv100.c
@@ -0,0 +1,215 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "ctxgf100.h"
+
+/*******************************************************************************
+ * PGRAPH context implementation
+ ******************************************************************************/
+
+static const struct gf100_gr_init
+gv100_grctx_init_sw_veid_bundle_init_0[] = {
+	{ 0x00001000, 64, 0x00100000, 0x00000008 },
+	{ 0x00000941, 64, 0x00100000, 0x00000000 },
+	{ 0x0000097e, 64, 0x00100000, 0x00000000 },
+	{ 0x0000097f, 64, 0x00100000, 0x00000100 },
+	{ 0x0000035c, 64, 0x00100000, 0x00000000 },
+	{ 0x0000035d, 64, 0x00100000, 0x00000000 },
+	{ 0x00000a08, 64, 0x00100000, 0x00000000 },
+	{ 0x00000a09, 64, 0x00100000, 0x00000000 },
+	{ 0x00000a0a, 64, 0x00100000, 0x00000000 },
+	{ 0x00000352, 64, 0x00100000, 0x00000000 },
+	{ 0x00000353, 64, 0x00100000, 0x00000000 },
+	{ 0x00000358, 64, 0x00100000, 0x00000000 },
+	{ 0x00000359, 64, 0x00100000, 0x00000000 },
+	{ 0x00000370, 64, 0x00100000, 0x00000000 },
+	{ 0x00000371, 64, 0x00100000, 0x00000000 },
+	{ 0x00000372, 64, 0x00100000, 0x000fffff },
+	{ 0x00000366, 64, 0x00100000, 0x00000000 },
+	{ 0x00000367, 64, 0x00100000, 0x00000000 },
+	{ 0x00000368, 64, 0x00100000, 0x00000fff },
+	{ 0x00000623, 64, 0x00100000, 0x00000000 },
+	{ 0x00000624, 64, 0x00100000, 0x00000000 },
+	{ 0x0001e100,  1, 0x00000001, 0x02000001 },
+	{}
+};
+
+static const struct gf100_gr_pack
+gv100_grctx_pack_sw_veid_bundle_init[] = {
+	{ gv100_grctx_init_sw_veid_bundle_init_0 },
+	{}
+};
+
+static void
+gv100_grctx_generate_attrib(struct gf100_grctx *info)
+{
+	struct gf100_gr *gr = info->gr;
+	const struct gf100_grctx_func *grctx = gr->func->grctx;
+	const u32  alpha = grctx->alpha_nr;
+	const u32 attrib = grctx->attrib_nr;
+	const u32   gfxp = grctx->gfxp_nr;
+	const int s = 12;
+	const int max_batches = 0xffff;
+	u32 size = grctx->alpha_nr_max * gr->tpc_total;
+	u32 ao = 0;
+	u32 bo = ao + size;
+	int gpc, ppc, b, n = 0;
+
+	size += grctx->gfxp_nr * gr->tpc_total;
+	size = ((size * 0x20) + 128) & ~127;
+	b = mmio_vram(info, size, (1 << s), false);
+
+	mmio_refn(info, 0x418810, 0x80000000, s, b);
+	mmio_refn(info, 0x419848, 0x10000000, s, b);
+	mmio_refn(info, 0x419c2c, 0x10000000, s, b);
+	mmio_refn(info, 0x419e00, 0x00000000, s, b);
+	mmio_wr32(info, 0x419e04, 0x80000000 | size >> 7);
+	mmio_wr32(info, 0x405830, attrib);
+	mmio_wr32(info, 0x40585c, alpha);
+	mmio_wr32(info, 0x4064c4, ((alpha / 4) << 16) | max_batches);
+
+	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+		for (ppc = 0; ppc < gr->ppc_nr[gpc]; ppc++, n++) {
+			const u32 as =  alpha * gr->ppc_tpc_nr[gpc][ppc];
+			const u32 bs = attrib * gr->ppc_tpc_nr[gpc][ppc];
+			const u32 gs =   gfxp * gr->ppc_tpc_nr[gpc][ppc];
+			const u32 u = 0x418ea0 + (n * 0x04);
+			const u32 o = PPC_UNIT(gpc, ppc, 0);
+			if (!(gr->ppc_mask[gpc] & (1 << ppc)))
+				continue;
+			mmio_wr32(info, o + 0xc0, gs);
+			mmio_wr32(info, o + 0xf4, bo);
+			mmio_wr32(info, o + 0xf0, bs);
+			bo += gs;
+			mmio_wr32(info, o + 0xe4, as);
+			mmio_wr32(info, o + 0xf8, ao);
+			ao += grctx->alpha_nr_max * gr->ppc_tpc_nr[gpc][ppc];
+			mmio_wr32(info, u, bs);
+		}
+	}
+
+	mmio_wr32(info, 0x4181e4, 0x00000100);
+	mmio_wr32(info, 0x41befc, 0x00000100);
+}
+
+static void
+gv100_grctx_generate_rop_mapping(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	u32 data;
+	int i, j;
+
+	/* Pack tile map into register format. */
+	nvkm_wr32(device, 0x418bb8, (gr->tpc_total << 8) |
+				     gr->screen_tile_row_offset);
+	for (i = 0; i < 11; i++) {
+		for (data = 0, j = 0; j < 6; j++)
+			data |= (gr->tile[i * 6 + j] & 0x1f) << (j * 5);
+		nvkm_wr32(device, 0x418b08 + (i * 4), data);
+		nvkm_wr32(device, 0x41bf00 + (i * 4), data);
+		nvkm_wr32(device, 0x40780c + (i * 4), data);
+	}
+
+	/* GPC_BROADCAST.TP_BROADCAST */
+	nvkm_wr32(device, 0x41bfd0, (gr->tpc_total << 8) |
+				     gr->screen_tile_row_offset);
+	for (i = 0, j = 1; i < 5; i++, j += 4) {
+		u8 v19 = (1 << (j + 0)) % gr->tpc_total;
+		u8 v20 = (1 << (j + 1)) % gr->tpc_total;
+		u8 v21 = (1 << (j + 2)) % gr->tpc_total;
+		u8 v22 = (1 << (j + 3)) % gr->tpc_total;
+		nvkm_wr32(device, 0x41bfb0 + (i * 4), (v22 << 24) |
+						      (v21 << 16) |
+						      (v20 <<  8) |
+						       v19);
+	}
+
+	/* UNK78xx */
+	nvkm_wr32(device, 0x4078bc, (gr->tpc_total << 8) |
+				     gr->screen_tile_row_offset);
+}
+
+static void
+gv100_grctx_generate_r400088(struct gf100_gr *gr, bool on)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x400088, 0x00060000, on ? 0x00060000 : 0x00000000);
+}
+
+static void
+gv100_grctx_generate_sm_id(struct gf100_gr *gr, int gpc, int tpc, int sm)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x608), sm);
+	nvkm_wr32(device, GPC_UNIT(gpc, 0x0c10 + tpc * 4), sm);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x088), sm);
+}
+
+static void
+gv100_grctx_generate_unkn(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x41980c, 0x00000010, 0x00000010);
+	nvkm_mask(device, 0x41be08, 0x00000004, 0x00000004);
+	nvkm_mask(device, 0x4064c0, 0x80000000, 0x80000000);
+	nvkm_mask(device, 0x405800, 0x08000000, 0x08000000);
+	nvkm_mask(device, 0x419c00, 0x00000008, 0x00000008);
+}
+
+static void
+gv100_grctx_unkn88c(struct gf100_gr *gr, bool on)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	const u32 mask = 0x00000010, data = on ? mask : 0x00000000;
+	nvkm_mask(device, 0x40988c, mask, data);
+	nvkm_rd32(device, 0x40988c);
+	nvkm_mask(device, 0x41a88c, mask, data);
+	nvkm_rd32(device, 0x41a88c);
+	nvkm_mask(device, 0x408a14, mask, data);
+	nvkm_rd32(device, 0x408a14);
+}
+
+const struct gf100_grctx_func
+gv100_grctx = {
+	.unkn88c = gv100_grctx_unkn88c,
+	.main = gf100_grctx_generate_main,
+	.unkn = gv100_grctx_generate_unkn,
+	.sw_veid_bundle_init = gv100_grctx_pack_sw_veid_bundle_init,
+	.bundle = gm107_grctx_generate_bundle,
+	.bundle_size = 0x3000,
+	.bundle_min_gpm_fifo_depth = 0x180,
+	.bundle_token_limit = 0x1680,
+	.pagepool = gp100_grctx_generate_pagepool,
+	.pagepool_size = 0x20000,
+	.attrib = gv100_grctx_generate_attrib,
+	.attrib_nr_max = 0x6c0,
+	.attrib_nr = 0x480,
+	.alpha_nr_max = 0xc00,
+	.alpha_nr = 0x800,
+	.gfxp_nr = 0xd10,
+	.sm_id = gv100_grctx_generate_sm_id,
+	.rop_mapping = gv100_grctx_generate_rop_mapping,
+	.dist_skip_table = gm200_grctx_generate_dist_skip_table,
+	.r406500 = gm200_grctx_generate_r406500,
+	.gpc_tpc_nr = gk104_grctx_generate_gpc_tpc_nr,
+	.smid_config = gp100_grctx_generate_smid_config,
+	.r400088 = gv100_grctx_generate_r400088,
+};
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
index 2f8dc107047d..70d3d41e616c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.c
@@ -32,6 +32,7 @@
 #include <subdev/fb.h>
 #include <subdev/mc.h>
 #include <subdev/pmu.h>
+#include <subdev/therm.h>
 #include <subdev/timer.h>
 #include <engine/fifo.h>
 
@@ -91,7 +92,7 @@ gf100_gr_zbc_color_get(struct gf100_gr *gr, int format,
 	memcpy(gr->zbc_color[zbc].l2, l2, sizeof(gr->zbc_color[zbc].l2));
 	gr->zbc_color[zbc].format = format;
 	nvkm_ltc_zbc_color_get(ltc, zbc, l2);
-	gf100_gr_zbc_clear_color(gr, zbc);
+	gr->func->zbc->clear_color(gr, zbc);
 	return zbc;
 }
 
@@ -136,10 +137,16 @@ gf100_gr_zbc_depth_get(struct gf100_gr *gr, int format,
 	gr->zbc_depth[zbc].ds = ds;
 	gr->zbc_depth[zbc].l2 = l2;
 	nvkm_ltc_zbc_depth_get(ltc, zbc, l2);
-	gf100_gr_zbc_clear_depth(gr, zbc);
+	gr->func->zbc->clear_depth(gr, zbc);
 	return zbc;
 }
 
+const struct gf100_gr_func_zbc
+gf100_gr_zbc = {
+	.clear_color = gf100_gr_zbc_clear_color,
+	.clear_depth = gf100_gr_zbc_clear_depth,
+};
+
 /*******************************************************************************
  * Graphics object classes
  ******************************************************************************/
@@ -743,21 +750,31 @@ gf100_gr_zbc_init(struct gf100_gr *gr)
 	const u32 f32_1[] = { 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000,
 			      0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000 };
 	struct nvkm_ltc *ltc = gr->base.engine.subdev.device->ltc;
-	int index;
+	int index, c = ltc->zbc_min, d = ltc->zbc_min, s = ltc->zbc_min;
 
 	if (!gr->zbc_color[0].format) {
-		gf100_gr_zbc_color_get(gr, 1,  & zero[0],   &zero[4]);
-		gf100_gr_zbc_color_get(gr, 2,  &  one[0],    &one[4]);
-		gf100_gr_zbc_color_get(gr, 4,  &f32_0[0],  &f32_0[4]);
-		gf100_gr_zbc_color_get(gr, 4,  &f32_1[0],  &f32_1[4]);
-		gf100_gr_zbc_depth_get(gr, 1, 0x00000000, 0x00000000);
-		gf100_gr_zbc_depth_get(gr, 1, 0x3f800000, 0x3f800000);
-	}
-
-	for (index = ltc->zbc_min; index <= ltc->zbc_max; index++)
-		gf100_gr_zbc_clear_color(gr, index);
-	for (index = ltc->zbc_min; index <= ltc->zbc_max; index++)
-		gf100_gr_zbc_clear_depth(gr, index);
+		gf100_gr_zbc_color_get(gr, 1,  & zero[0],   &zero[4]); c++;
+		gf100_gr_zbc_color_get(gr, 2,  &  one[0],    &one[4]); c++;
+		gf100_gr_zbc_color_get(gr, 4,  &f32_0[0],  &f32_0[4]); c++;
+		gf100_gr_zbc_color_get(gr, 4,  &f32_1[0],  &f32_1[4]); c++;
+		gf100_gr_zbc_depth_get(gr, 1, 0x00000000, 0x00000000); d++;
+		gf100_gr_zbc_depth_get(gr, 1, 0x3f800000, 0x3f800000); d++;
+		if (gr->func->zbc->stencil_get) {
+			gr->func->zbc->stencil_get(gr, 1, 0x00, 0x00); s++;
+			gr->func->zbc->stencil_get(gr, 1, 0x01, 0x01); s++;
+			gr->func->zbc->stencil_get(gr, 1, 0xff, 0xff); s++;
+		}
+	}
+
+	for (index = c; index <= ltc->zbc_max; index++)
+		gr->func->zbc->clear_color(gr, index);
+	for (index = d; index <= ltc->zbc_max; index++)
+		gr->func->zbc->clear_depth(gr, index);
+
+	if (gr->func->zbc->clear_stencil) {
+		for (index = s; index <= ltc->zbc_max; index++)
+			gr->func->zbc->clear_stencil(gr, index);
+	}
 }
 
 /**
@@ -970,7 +987,7 @@ gf100_gr_trap_gpc_rop(struct gf100_gr *gr, int gpc)
 	nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000);
 }
 
-static const struct nvkm_enum gf100_mp_warp_error[] = {
+const struct nvkm_enum gf100_mp_warp_error[] = {
 	{ 0x01, "STACK_ERROR" },
 	{ 0x02, "API_STACK_ERROR" },
 	{ 0x03, "RET_EMPTY_STACK_ERROR" },
@@ -995,7 +1012,7 @@ static const struct nvkm_enum gf100_mp_warp_error[] = {
 	{}
 };
 
-static const struct nvkm_bitfield gf100_mp_global_error[] = {
+const struct nvkm_bitfield gf100_mp_global_error[] = {
 	{ 0x00000001, "SM_TO_SM_FAULT" },
 	{ 0x00000002, "L1_ERROR" },
 	{ 0x00000004, "MULTIPLE_WARP_ERRORS" },
@@ -1009,7 +1026,7 @@ static const struct nvkm_bitfield gf100_mp_global_error[] = {
 	{}
 };
 
-static void
+void
 gf100_gr_trap_mp(struct gf100_gr *gr, int gpc, int tpc)
 {
 	struct nvkm_subdev *subdev = &gr->base.engine.subdev;
@@ -1045,7 +1062,7 @@ gf100_gr_trap_tpc(struct gf100_gr *gr, int gpc, int tpc)
 	}
 
 	if (stat & 0x00000002) {
-		gf100_gr_trap_mp(gr, gpc, tpc);
+		gr->func->trap_mp(gr, gpc, tpc);
 		stat &= ~0x00000002;
 	}
 
@@ -1611,7 +1628,8 @@ gf100_gr_init_ctxctl_int(struct gf100_gr *gr)
 
 	/* load register lists */
 	gf100_gr_init_csdata(gr, grctx->hub, 0x409000, 0x000, 0x000000);
-	gf100_gr_init_csdata(gr, grctx->gpc, 0x41a000, 0x000, 0x418000);
+	gf100_gr_init_csdata(gr, grctx->gpc_0, 0x41a000, 0x000, 0x418000);
+	gf100_gr_init_csdata(gr, grctx->gpc_1, 0x41a000, 0x000, 0x418000);
 	gf100_gr_init_csdata(gr, grctx->tpc, 0x41a000, 0x004, 0x419800);
 	gf100_gr_init_csdata(gr, grctx->ppc, 0x41a000, 0x008, 0x41be00);
 
@@ -1651,6 +1669,97 @@ gf100_gr_init_ctxctl(struct gf100_gr *gr)
 	return ret;
 }
 
+void
+gf100_gr_oneinit_sm_id(struct gf100_gr *gr)
+{
+	int tpc, gpc;
+	for (tpc = 0; tpc < gr->tpc_max; tpc++) {
+		for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+			if (tpc < gr->tpc_nr[gpc]) {
+				gr->sm[gr->sm_nr].gpc = gpc;
+				gr->sm[gr->sm_nr].tpc = tpc;
+				gr->sm_nr++;
+			}
+		}
+	}
+}
+
+void
+gf100_gr_oneinit_tiles(struct gf100_gr *gr)
+{
+	static const u8 primes[] = {
+		3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47, 53, 59, 61
+	};
+	int init_frac[GPC_MAX], init_err[GPC_MAX], run_err[GPC_MAX], i, j;
+	u32 mul_factor, comm_denom;
+	u8  gpc_map[GPC_MAX];
+	bool sorted;
+
+	switch (gr->tpc_total) {
+	case 15: gr->screen_tile_row_offset = 0x06; break;
+	case 14: gr->screen_tile_row_offset = 0x05; break;
+	case 13: gr->screen_tile_row_offset = 0x02; break;
+	case 11: gr->screen_tile_row_offset = 0x07; break;
+	case 10: gr->screen_tile_row_offset = 0x06; break;
+	case  7:
+	case  5: gr->screen_tile_row_offset = 0x01; break;
+	case  3: gr->screen_tile_row_offset = 0x02; break;
+	case  2:
+	case  1: gr->screen_tile_row_offset = 0x01; break;
+	default: gr->screen_tile_row_offset = 0x03;
+		for (i = 0; i < ARRAY_SIZE(primes); i++) {
+			if (gr->tpc_total % primes[i]) {
+				gr->screen_tile_row_offset = primes[i];
+				break;
+			}
+		}
+		break;
+	}
+
+	/* Sort GPCs by TPC count, highest-to-lowest. */
+	for (i = 0; i < gr->gpc_nr; i++)
+		gpc_map[i] = i;
+	sorted = false;
+
+	while (!sorted) {
+		for (sorted = true, i = 0; i < gr->gpc_nr - 1; i++) {
+			if (gr->tpc_nr[gpc_map[i + 1]] >
+			    gr->tpc_nr[gpc_map[i + 0]]) {
+				u8 swap = gpc_map[i];
+				gpc_map[i + 0] = gpc_map[i + 1];
+				gpc_map[i + 1] = swap;
+				sorted = false;
+			}
+		}
+	}
+
+	/* Determine tile->GPC mapping */
+	mul_factor = gr->gpc_nr * gr->tpc_max;
+	if (mul_factor & 1)
+		mul_factor = 2;
+	else
+		mul_factor = 1;
+
+	comm_denom = gr->gpc_nr * gr->tpc_max * mul_factor;
+
+	for (i = 0; i < gr->gpc_nr; i++) {
+		init_frac[i] = gr->tpc_nr[gpc_map[i]] * gr->gpc_nr * mul_factor;
+		 init_err[i] = i * gr->tpc_max * mul_factor - comm_denom/2;
+		  run_err[i] = init_frac[i] + init_err[i];
+	}
+
+	for (i = 0; i < gr->tpc_total;) {
+		for (j = 0; j < gr->gpc_nr; j++) {
+			if ((run_err[j] * 2) >= comm_denom) {
+				gr->tile[i++] = gpc_map[j];
+				run_err[j] += init_frac[j] - comm_denom;
+			} else {
+				run_err[j] += init_frac[j];
+			}
+		}
+	}
+}
+
 static int
 gf100_gr_oneinit(struct nvkm_gr *base)
 {
@@ -1674,55 +1783,27 @@ gf100_gr_oneinit(struct nvkm_gr *base)
 	gr->gpc_nr = nvkm_rd32(device, 0x409604) & 0x0000001f;
 	for (i = 0; i < gr->gpc_nr; i++) {
 		gr->tpc_nr[i]  = nvkm_rd32(device, GPC_UNIT(i, 0x2608));
+		gr->tpc_max = max(gr->tpc_max, gr->tpc_nr[i]);
 		gr->tpc_total += gr->tpc_nr[i];
 		gr->ppc_nr[i]  = gr->func->ppc_nr;
 		for (j = 0; j < gr->ppc_nr[i]; j++) {
-			u8 mask = nvkm_rd32(device, GPC_UNIT(i, 0x0c30 + (j * 4)));
-			if (mask)
-				gr->ppc_mask[i] |= (1 << j);
-			gr->ppc_tpc_nr[i][j] = hweight8(mask);
-		}
-	}
-
-	/*XXX: these need figuring out... though it might not even matter */
-	switch (device->chipset) {
-	case 0xc0:
-		if (gr->tpc_total == 11) { /* 465, 3/4/4/0, 4 */
-			gr->screen_tile_row_offset = 0x07;
-		} else
-		if (gr->tpc_total == 14) { /* 470, 3/3/4/4, 5 */
-			gr->screen_tile_row_offset = 0x05;
-		} else
-		if (gr->tpc_total == 15) { /* 480, 3/4/4/4, 6 */
-			gr->screen_tile_row_offset = 0x06;
+			gr->ppc_tpc_mask[i][j] =
+				nvkm_rd32(device, GPC_UNIT(i, 0x0c30 + (j * 4)));
+			if (gr->ppc_tpc_mask[i][j] == 0)
+				continue;
+			gr->ppc_mask[i] |= (1 << j);
+			gr->ppc_tpc_nr[i][j] = hweight8(gr->ppc_tpc_mask[i][j]);
+			if (gr->ppc_tpc_min == 0 ||
+			    gr->ppc_tpc_min > gr->ppc_tpc_nr[i][j])
+				gr->ppc_tpc_min = gr->ppc_tpc_nr[i][j];
+			if (gr->ppc_tpc_max < gr->ppc_tpc_nr[i][j])
+				gr->ppc_tpc_max = gr->ppc_tpc_nr[i][j];
 		}
-		break;
-	case 0xc3: /* 450, 4/0/0/0, 2 */
-		gr->screen_tile_row_offset = 0x03;
-		break;
-	case 0xc4: /* 460, 3/4/0/0, 4 */
-		gr->screen_tile_row_offset = 0x01;
-		break;
-	case 0xc1: /* 2/0/0/0, 1 */
-		gr->screen_tile_row_offset = 0x01;
-		break;
-	case 0xc8: /* 4/4/3/4, 5 */
-		gr->screen_tile_row_offset = 0x06;
-		break;
-	case 0xce: /* 4/4/0/0, 4 */
-		gr->screen_tile_row_offset = 0x03;
-		break;
-	case 0xcf: /* 4/0/0/0, 3 */
-		gr->screen_tile_row_offset = 0x03;
-		break;
-	case 0xd7:
-	case 0xd9: /* 1/0/0/0, 1 */
-	case 0xea: /* gk20a */
-	case 0x12b: /* gm20b */
-		gr->screen_tile_row_offset = 0x01;
-		break;
 	}
 
+	memset(gr->tile, 0xff, sizeof(gr->tile));
+	gr->func->oneinit_tiles(gr);
+	gr->func->oneinit_sm_id(gr);
 	return 0;
 }
 
@@ -1914,13 +1995,68 @@ gf100_gr_new_(const struct gf100_gr_func *func, struct nvkm_device *device,
 }
 
 void
+gf100_gr_init_400054(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x400054, 0x34ce3464);
+}
+
+void
+gf100_gr_init_shader_exceptions(struct gf100_gr *gr, int gpc, int tpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x001ffffe);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x0000000f);
+}
+
+void
+gf100_gr_init_tex_hww_esr(struct gf100_gr *gr, int gpc, int tpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
+}
+
+void
+gf100_gr_init_419eb4(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419eb4, 0x00001000, 0x00001000);
+}
+
+void
+gf100_gr_init_419cc0(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	int gpc, tpc;
+
+	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
+
+	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++)
+			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x48c), 0xc0000000);
+	}
+}
+
+void
+gf100_gr_init_40601c(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x40601c, 0xc0000000);
+}
+
+void
+gf100_gr_init_fecs_exceptions(struct gf100_gr *gr)
+{
+	const u32 data = gr->firmware ? 0x000e0000 : 0x000e0001;
+	nvkm_wr32(gr->base.engine.subdev.device, 0x409c24, data);
+}
+
+void
 gf100_gr_init_gpc_mmu(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	struct nvkm_fb *fb = device->fb;
 
 	nvkm_wr32(device, 0x418880, nvkm_rd32(device, 0x100c80) & 0x00000001);
-	nvkm_wr32(device, 0x4188a4, 0x00000000);
+	nvkm_wr32(device, 0x4188a4, 0x03000000);
 	nvkm_wr32(device, 0x418888, 0x00000000);
 	nvkm_wr32(device, 0x41888c, 0x00000000);
 	nvkm_wr32(device, 0x418890, 0x00000000);
@@ -1929,37 +2065,30 @@ gf100_gr_init_gpc_mmu(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x4188b8, nvkm_memory_addr(fb->mmu_rd) >> 8);
 }
 
-int
-gf100_gr_init(struct gf100_gr *gr)
+void
+gf100_gr_init_num_active_ltcs(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
-	u32 data[TPC_MAX / 8] = {};
-	u8  tpcnr[GPC_MAX];
-	int gpc, tpc, rop;
-	int i;
-
-	gr->func->init_gpc_mmu(gr);
-
-	gf100_gr_mmio(gr, gr->func->mmio);
-
-	nvkm_mask(device, TPC_UNIT(0, 0, 0x05c), 0x00000001, 0x00000001);
-
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
+	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
+}
 
-		data[i / 8] |= tpc << ((i % 8) * 4);
+void
+gf100_gr_init_zcull(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
+	const u8 tile_nr = ALIGN(gr->tpc_total, 32);
+	u8 bank[GPC_MAX] = {}, gpc, i, j;
+	u32 data;
+
+	for (i = 0; i < tile_nr; i += 8) {
+		for (data = 0, j = 0; j < 8 && i + j < gr->tpc_total; j++) {
+			data |= bank[gr->tile[i + j]] << (j * 4);
+			bank[gr->tile[i + j]]++;
+		}
+		nvkm_wr32(device, GPC_BCAST(0x0980 + ((i / 8) * 4)), data);
 	}
 
-	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
-	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
-	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
-	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
-
 	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
 			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
@@ -1968,29 +2097,88 @@ gf100_gr_init(struct gf100_gr *gr)
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
 	}
 
-	if (device->chipset != 0xd7)
-		nvkm_wr32(device, GPC_BCAST(0x1bd4), magicgpc918);
+	nvkm_wr32(device, GPC_BCAST(0x1bd4), magicgpc918);
+}
+
+void
+gf100_gr_init_vsc_stream_master(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, TPC_UNIT(0, 0, 0x05c), 0x00000001, 0x00000001);
+}
+
+int
+gf100_gr_init(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	int gpc, tpc, rop;
+
+	if (gr->func->init_419bd8)
+		gr->func->init_419bd8(gr);
+
+	gr->func->init_gpc_mmu(gr);
+
+	if (gr->fuc_sw_nonctx)
+		gf100_gr_mmio(gr, gr->fuc_sw_nonctx);
 	else
-		nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
+		gf100_gr_mmio(gr, gr->func->mmio);
 
-	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
+	gf100_gr_wait_idle(gr);
+
+	if (gr->func->init_r405a14)
+		gr->func->init_r405a14(gr);
+
+	if (gr->func->clkgate_pack)
+		nvkm_therm_clkgate_init(device->therm, gr->func->clkgate_pack);
+
+	if (gr->func->init_bios)
+		gr->func->init_bios(gr);
+
+	gr->func->init_vsc_stream_master(gr);
+	gr->func->init_zcull(gr);
+	gr->func->init_num_active_ltcs(gr);
+	if (gr->func->init_rop_active_fbps)
+		gr->func->init_rop_active_fbps(gr);
+	if (gr->func->init_bios_2)
+		gr->func->init_bios_2(gr);
+	if (gr->func->init_swdx_pes_mask)
+		gr->func->init_swdx_pes_mask(gr);
 
 	nvkm_wr32(device, 0x400500, 0x00010001);
 
 	nvkm_wr32(device, 0x400100, 0xffffffff);
 	nvkm_wr32(device, 0x40013c, 0xffffffff);
+	nvkm_wr32(device, 0x400124, 0x00000002);
+
+	gr->func->init_fecs_exceptions(gr);
+	if (gr->func->init_ds_hww_esr_2)
+		gr->func->init_ds_hww_esr_2(gr);
 
-	nvkm_wr32(device, 0x409c24, 0x000f0000);
 	nvkm_wr32(device, 0x404000, 0xc0000000);
 	nvkm_wr32(device, 0x404600, 0xc0000000);
 	nvkm_wr32(device, 0x408030, 0xc0000000);
-	nvkm_wr32(device, 0x40601c, 0xc0000000);
+
+	if (gr->func->init_40601c)
+		gr->func->init_40601c(gr);
+
 	nvkm_wr32(device, 0x404490, 0xc0000000);
 	nvkm_wr32(device, 0x406018, 0xc0000000);
+
+	if (gr->func->init_sked_hww_esr)
+		gr->func->init_sked_hww_esr(gr);
+
 	nvkm_wr32(device, 0x405840, 0xc0000000);
 	nvkm_wr32(device, 0x405844, 0x00ffffff);
-	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
-	nvkm_mask(device, 0x419eb4, 0x00001000, 0x00001000);
+
+	if (gr->func->init_419cc0)
+		gr->func->init_419cc0(gr);
+	if (gr->func->init_419eb4)
+		gr->func->init_419eb4(gr);
+	if (gr->func->init_419c9c)
+		gr->func->init_419c9c(gr);
+
+	if (gr->func->init_ppc_exceptions)
+		gr->func->init_ppc_exceptions(gr);
 
 	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000);
@@ -2000,19 +2188,20 @@ gf100_gr_init(struct gf100_gr *gr)
 		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) {
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x48c), 0xc0000000);
+			if (gr->func->init_tex_hww_esr)
+				gr->func->init_tex_hww_esr(gr, gpc, tpc);
 			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x001ffffe);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x0000000f);
+			if (gr->func->init_504430)
+				gr->func->init_504430(gr, gpc, tpc);
+			gr->func->init_shader_exceptions(gr, gpc, tpc);
 		}
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
 		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
 	}
 
 	for (rop = 0; rop < gr->rop_nr; rop++) {
-		nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0xc0000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0xc0000000);
+		nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0x40000000);
+		nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0x40000000);
 		nvkm_wr32(device, ROP_UNIT(rop, 0x204), 0xffffffff);
 		nvkm_wr32(device, ROP_UNIT(rop, 0x208), 0xffffffff);
 	}
@@ -2024,10 +2213,14 @@ gf100_gr_init(struct gf100_gr *gr)
 	nvkm_wr32(device, 0x40011c, 0xffffffff);
 	nvkm_wr32(device, 0x400134, 0xffffffff);
 
-	nvkm_wr32(device, 0x400054, 0x34ce3464);
+	if (gr->func->init_400054)
+		gr->func->init_400054(gr);
 
 	gf100_gr_zbc_init(gr);
 
+	if (gr->func->init_4188a4)
+		gr->func->init_4188a4(gr);
+
 	return gf100_gr_init_ctxctl(gr);
 }
 
@@ -2053,13 +2246,27 @@ gf100_gr_gpccs_ucode = {
 
 static const struct gf100_gr_func
 gf100_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
+	.init_zcull = gf100_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_40601c = gf100_gr_init_40601c,
+	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gf100_gr_init_419eb4,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gf100_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
 	.rops = gf100_gr_rops,
 	.grctx = &gf100_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, FERMI_MEMORY_TO_MEMORY_FORMAT_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
index c8ec3fd97155..dc46cf0131db 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf100.h
@@ -72,6 +72,12 @@ struct gf100_gr_zbc_depth {
 	u32 l2;
 };
 
+struct gf100_gr_zbc_stencil {
+	u32 format;
+	u32 ds;
+	u32 l2;
+};
+
 struct gf100_gr {
 	const struct gf100_gr_func *func;
 	struct nvkm_gr base;
@@ -95,21 +101,33 @@ struct gf100_gr {
 
 	struct gf100_gr_zbc_color zbc_color[NVKM_LTC_MAX_ZBC_CNT];
 	struct gf100_gr_zbc_depth zbc_depth[NVKM_LTC_MAX_ZBC_CNT];
+	struct gf100_gr_zbc_stencil zbc_stencil[NVKM_LTC_MAX_ZBC_CNT];
 
 	u8 rop_nr;
 	u8 gpc_nr;
 	u8 tpc_nr[GPC_MAX];
+	u8 tpc_max;
 	u8 tpc_total;
 	u8 ppc_nr[GPC_MAX];
 	u8 ppc_mask[GPC_MAX];
+	u8 ppc_tpc_mask[GPC_MAX][4];
 	u8 ppc_tpc_nr[GPC_MAX][4];
+	u8 ppc_tpc_min;
+	u8 ppc_tpc_max;
+
+	u8 screen_tile_row_offset;
+	u8 tile[TPC_MAX];
+
+	struct {
+		u8 gpc;
+		u8 tpc;
+	} sm[TPC_MAX];
+	u8 sm_nr;
 
 	struct gf100_gr_data mmio_data[4];
 	struct gf100_gr_mmio mmio_list[4096/8];
 	u32  size;
 	u32 *data;
-
-	u8 screen_tile_row_offset;
 };
 
 int gf100_gr_ctor(const struct gf100_gr_func *, struct nvkm_device *,
@@ -118,14 +136,43 @@ int gf100_gr_new_(const struct gf100_gr_func *, struct nvkm_device *,
 		  int, struct nvkm_gr **);
 void *gf100_gr_dtor(struct nvkm_gr *);
 
+struct gf100_gr_func_zbc {
+	void (*clear_color)(struct gf100_gr *, int zbc);
+	void (*clear_depth)(struct gf100_gr *, int zbc);
+	int (*stencil_get)(struct gf100_gr *, int format,
+			   const u32 ds, const u32 l2);
+	void (*clear_stencil)(struct gf100_gr *, int zbc);
+};
+
 struct gf100_gr_func {
 	void (*dtor)(struct gf100_gr *);
+	void (*oneinit_tiles)(struct gf100_gr *);
+	void (*oneinit_sm_id)(struct gf100_gr *);
 	int (*init)(struct gf100_gr *);
+	void (*init_419bd8)(struct gf100_gr *);
 	void (*init_gpc_mmu)(struct gf100_gr *);
+	void (*init_r405a14)(struct gf100_gr *);
+	void (*init_bios)(struct gf100_gr *);
+	void (*init_vsc_stream_master)(struct gf100_gr *);
+	void (*init_zcull)(struct gf100_gr *);
+	void (*init_num_active_ltcs)(struct gf100_gr *);
 	void (*init_rop_active_fbps)(struct gf100_gr *);
-	void (*init_ppc_exceptions)(struct gf100_gr *);
+	void (*init_bios_2)(struct gf100_gr *);
 	void (*init_swdx_pes_mask)(struct gf100_gr *);
-	void (*init_num_active_ltcs)(struct gf100_gr *);
+	void (*init_fecs_exceptions)(struct gf100_gr *);
+	void (*init_ds_hww_esr_2)(struct gf100_gr *);
+	void (*init_40601c)(struct gf100_gr *);
+	void (*init_sked_hww_esr)(struct gf100_gr *);
+	void (*init_419cc0)(struct gf100_gr *);
+	void (*init_419eb4)(struct gf100_gr *);
+	void (*init_419c9c)(struct gf100_gr *);
+	void (*init_ppc_exceptions)(struct gf100_gr *);
+	void (*init_tex_hww_esr)(struct gf100_gr *, int gpc, int tpc);
+	void (*init_504430)(struct gf100_gr *, int gpc, int tpc);
+	void (*init_shader_exceptions)(struct gf100_gr *, int gpc, int tpc);
+	void (*init_400054)(struct gf100_gr *);
+	void (*init_4188a4)(struct gf100_gr *);
+	void (*trap_mp)(struct gf100_gr *, int gpc, int tpc);
 	void (*set_hww_esr_report_mask)(struct gf100_gr *);
 	const struct gf100_gr_pack *mmio;
 	struct {
@@ -135,26 +182,60 @@ struct gf100_gr_func {
 		struct gf100_gr_ucode *ucode;
 	} gpccs;
 	int (*rops)(struct gf100_gr *);
+	int gpc_nr;
+	int tpc_nr;
 	int ppc_nr;
 	const struct gf100_grctx_func *grctx;
 	const struct nvkm_therm_clkgate_pack *clkgate_pack;
+	const struct gf100_gr_func_zbc *zbc;
 	struct nvkm_sclass sclass[];
 };
 
-int gf100_gr_init(struct gf100_gr *);
 int gf100_gr_rops(struct gf100_gr *);
-
-int gk104_gr_init(struct gf100_gr *);
+void gf100_gr_oneinit_tiles(struct gf100_gr *);
+void gf100_gr_oneinit_sm_id(struct gf100_gr *);
+int gf100_gr_init(struct gf100_gr *);
+void gf100_gr_init_vsc_stream_master(struct gf100_gr *);
+void gf100_gr_init_zcull(struct gf100_gr *);
+void gf100_gr_init_num_active_ltcs(struct gf100_gr *);
+void gf100_gr_init_fecs_exceptions(struct gf100_gr *);
+void gf100_gr_init_40601c(struct gf100_gr *);
+void gf100_gr_init_419cc0(struct gf100_gr *);
+void gf100_gr_init_419eb4(struct gf100_gr *);
+void gf100_gr_init_tex_hww_esr(struct gf100_gr *, int, int);
+void gf100_gr_init_shader_exceptions(struct gf100_gr *, int, int);
+void gf100_gr_init_400054(struct gf100_gr *);
+extern const struct gf100_gr_func_zbc gf100_gr_zbc;
+
+void gf117_gr_init_zcull(struct gf100_gr *);
+
+void gk104_gr_init_vsc_stream_master(struct gf100_gr *);
 void gk104_gr_init_rop_active_fbps(struct gf100_gr *);
 void gk104_gr_init_ppc_exceptions(struct gf100_gr *);
+void gk104_gr_init_sked_hww_esr(struct gf100_gr *);
+
+void gk110_gr_init_419eb4(struct gf100_gr *);
+
+void gm107_gr_init_504430(struct gf100_gr *, int, int);
+void gm107_gr_init_shader_exceptions(struct gf100_gr *, int, int);
+void gm107_gr_init_400054(struct gf100_gr *);
 
 int gk20a_gr_init(struct gf100_gr *);
 
-int gm200_gr_init(struct gf100_gr *);
+void gm200_gr_oneinit_tiles(struct gf100_gr *);
+void gm200_gr_oneinit_sm_id(struct gf100_gr *);
 int gm200_gr_rops(struct gf100_gr *);
+void gm200_gr_init_num_active_ltcs(struct gf100_gr *);
+void gm200_gr_init_ds_hww_esr_2(struct gf100_gr *);
 
-int gp100_gr_init(struct gf100_gr *);
 void gp100_gr_init_rop_active_fbps(struct gf100_gr *);
+void gp100_gr_init_fecs_exceptions(struct gf100_gr *);
+void gp100_gr_init_shader_exceptions(struct gf100_gr *, int, int);
+void gp100_gr_zbc_clear_color(struct gf100_gr *, int);
+void gp100_gr_zbc_clear_depth(struct gf100_gr *, int);
+
+void gp102_gr_init_swdx_pes_mask(struct gf100_gr *);
+extern const struct gf100_gr_func_zbc gp102_gr_zbc;
 
 #define gf100_gr_chan(p) container_of((p), struct gf100_gr_chan, object)
 #include <core/object.h>
@@ -187,7 +268,7 @@ extern const struct nvkm_object_func gf100_fermi;
 struct gf100_gr_init {
 	u32 addr;
 	u8  count;
-	u8  pitch;
+	u32 pitch;
 	u32 data;
 };
 
@@ -257,6 +338,9 @@ extern const struct gf100_gr_init gf100_gr_init_be_0[];
 extern const struct gf100_gr_init gf100_gr_init_fe_1[];
 extern const struct gf100_gr_init gf100_gr_init_pe_1[];
 void gf100_gr_init_gpc_mmu(struct gf100_gr *);
+void gf100_gr_trap_mp(struct gf100_gr *, int, int);
+extern const struct nvkm_bitfield gf100_mp_global_error[];
+extern const struct nvkm_enum gf100_mp_warp_error[];
 
 extern const struct gf100_gr_init gf104_gr_init_ds_0[];
 extern const struct gf100_gr_init gf104_gr_init_tex_0[];
@@ -279,6 +363,7 @@ extern const struct gf100_gr_init gf117_gr_init_wwdx_0[];
 extern const struct gf100_gr_init gf117_gr_init_cbm_0[];
 
 extern const struct gf100_gr_init gk104_gr_init_main_0[];
+extern const struct gf100_gr_init gk104_gr_init_gpc_unk_2[];
 extern const struct gf100_gr_init gk104_gr_init_tpccs_0[];
 extern const struct gf100_gr_init gk104_gr_init_pe_0[];
 extern const struct gf100_gr_init gk104_gr_init_be_0[];
@@ -306,8 +391,4 @@ extern const struct gf100_gr_init gm107_gr_init_cbm_0[];
 void gm107_gr_init_bios(struct gf100_gr *);
 
 void gm200_gr_init_gpc_mmu(struct gf100_gr *);
-
-void gp100_gr_init_num_active_ltcs(struct gf100_gr *gr);
-
-void gp102_gr_init_swdx_pes_mask(struct gf100_gr *);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
index ec0f11983b23..42c2fd9fc04e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf104.c
@@ -114,13 +114,27 @@ gf104_gr_pack_mmio[] = {
 
 static const struct gf100_gr_func
 gf104_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
+	.init_zcull = gf100_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_40601c = gf100_gr_init_40601c,
+	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gf100_gr_init_419eb4,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gf104_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
 	.rops = gf100_gr_rops,
 	.grctx = &gf104_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, FERMI_MEMORY_TO_MEMORY_FORMAT_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
index cc152eb74123..4731a460adc7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf108.c
@@ -103,15 +103,36 @@ gf108_gr_pack_mmio[] = {
  * PGRAPH engine/subdev functions
  ******************************************************************************/
 
+static void
+gf108_gr_init_r405a14(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x405a14, 0x80000000);
+}
+
 static const struct gf100_gr_func
 gf108_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_r405a14 = gf108_gr_init_r405a14,
+	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
+	.init_zcull = gf100_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_40601c = gf100_gr_init_40601c,
+	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gf100_gr_init_419eb4,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gf108_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
 	.rops = gf100_gr_rops,
 	.grctx = &gf108_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, FERMI_MEMORY_TO_MEMORY_FORMAT_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
index 10d2d73ca8c3..cdf759c8cd7f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf110.c
@@ -86,13 +86,27 @@ gf110_gr_pack_mmio[] = {
 
 static const struct gf100_gr_func
 gf110_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
+	.init_zcull = gf100_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_40601c = gf100_gr_init_40601c,
+	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gf100_gr_init_419eb4,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gf110_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
 	.rops = gf100_gr_rops,
 	.grctx = &gf110_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, FERMI_MEMORY_TO_MEMORY_FORMAT_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
index ac09a07c4150..a4158f84c649 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf117.c
@@ -120,16 +120,58 @@ gf117_gr_gpccs_ucode = {
 	.data.size = sizeof(gf117_grgpc_data),
 };
 
+void
+gf117_gr_init_zcull(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
+	const u8 tile_nr = ALIGN(gr->tpc_total, 32);
+	u8 bank[GPC_MAX] = {}, gpc, i, j;
+	u32 data;
+
+	for (i = 0; i < tile_nr; i += 8) {
+		for (data = 0, j = 0; j < 8 && i + j < gr->tpc_total; j++) {
+			data |= bank[gr->tile[i + j]] << (j * 4);
+			bank[gr->tile[i + j]]++;
+		}
+		nvkm_wr32(device, GPC_BCAST(0x0980 + ((i / 8) * 4)), data);
+	}
+
+	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
+		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
+			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
+		nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 |
+							 gr->tpc_total);
+		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
+	}
+
+	nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
+}
+
 static const struct gf100_gr_func
 gf117_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_40601c = gf100_gr_init_40601c,
+	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gf100_gr_init_419eb4,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gf117_gr_pack_mmio,
 	.fecs.ucode = &gf117_gr_fecs_ucode,
 	.gpccs.ucode = &gf117_gr_gpccs_ucode,
 	.rops = gf100_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gf117_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, FERMI_MEMORY_TO_MEMORY_FORMAT_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
index 7f449ec6f760..4197844870b3 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gf119.c
@@ -177,13 +177,27 @@ gf119_gr_pack_mmio[] = {
 
 static const struct gf100_gr_func
 gf119_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gf100_gr_init_vsc_stream_master,
+	.init_zcull = gf100_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_40601c = gf100_gr_init_40601c,
+	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gf100_gr_init_419eb4,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gf119_gr_pack_mmio,
 	.fecs.ucode = &gf100_gr_fecs_ucode,
 	.gpccs.ucode = &gf100_gr_gpccs_ucode,
 	.rops = gf100_gr_rops,
 	.grctx = &gf119_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, FERMI_MEMORY_TO_MEMORY_FORMAT_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
index 1b52fcb2c49a..477fee3e3715 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c
@@ -83,6 +83,12 @@ gk104_gr_init_gpc_unk_1[] = {
 };
 
 const struct gf100_gr_init
+gk104_gr_init_gpc_unk_2[] = {
+	{ 0x418884,   1, 0x04, 0x00000000 },
+	{}
+};
+
+const struct gf100_gr_init
 gk104_gr_init_tpccs_0[] = {
 	{ 0x419d0c,   1, 0x04, 0x00000000 },
 	{ 0x419d10,   1, 0x04, 0x00000014 },
@@ -160,6 +166,7 @@ gk104_gr_pack_mmio[] = {
 	{ gf119_gr_init_gpm_0 },
 	{ gk104_gr_init_gpc_unk_1 },
 	{ gf100_gr_init_gcc_0 },
+	{ gk104_gr_init_gpc_unk_2 },
 	{ gk104_gr_init_tpccs_0 },
 	{ gf119_gr_init_tex_0 },
 	{ gk104_gr_init_pe_0 },
@@ -381,6 +388,21 @@ gk104_clkgate_pack[] = {
  ******************************************************************************/
 
 void
+gk104_gr_init_sked_hww_esr(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x407020, 0x40000000);
+}
+
+static void
+gk104_gr_init_fecs_exceptions(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, 0x409ffc, 0x00000000);
+	nvkm_wr32(device, 0x409c14, 0x00003e3e);
+	nvkm_wr32(device, 0x409c24, 0x000f0001);
+}
+
+void
 gk104_gr_init_rop_active_fbps(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -404,112 +426,11 @@ gk104_gr_init_ppc_exceptions(struct gf100_gr *gr)
 	}
 }
 
-int
-gk104_gr_init(struct gf100_gr *gr)
+void
+gk104_gr_init_vsc_stream_master(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
-	u32 data[TPC_MAX / 8] = {};
-	u8  tpcnr[GPC_MAX];
-	int gpc, tpc, rop;
-	int i;
-
-	gr->func->init_gpc_mmu(gr);
-
-	gf100_gr_mmio(gr, gr->func->mmio);
-	if (gr->func->clkgate_pack)
-		nvkm_therm_clkgate_init(gr->base.engine.subdev.device->therm,
-					gr->func->clkgate_pack);
-
 	nvkm_wr32(device, GPC_UNIT(0, 0x3018), 0x00000001);
-
-	memset(data, 0x00, sizeof(data));
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		data[i / 8] |= tpc << ((i % 8) * 4);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
-	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
-	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
-	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
-			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 |
-							 gr->tpc_total);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
-	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
-
-	gr->func->init_rop_active_fbps(gr);
-
-	nvkm_wr32(device, 0x400500, 0x00010001);
-
-	nvkm_wr32(device, 0x400100, 0xffffffff);
-	nvkm_wr32(device, 0x40013c, 0xffffffff);
-
-	nvkm_wr32(device, 0x409ffc, 0x00000000);
-	nvkm_wr32(device, 0x409c14, 0x00003e3e);
-	nvkm_wr32(device, 0x409c24, 0x000f0001);
-	nvkm_wr32(device, 0x404000, 0xc0000000);
-	nvkm_wr32(device, 0x404600, 0xc0000000);
-	nvkm_wr32(device, 0x408030, 0xc0000000);
-	nvkm_wr32(device, 0x404490, 0xc0000000);
-	nvkm_wr32(device, 0x406018, 0xc0000000);
-	nvkm_wr32(device, 0x407020, 0x40000000);
-	nvkm_wr32(device, 0x405840, 0xc0000000);
-	nvkm_wr32(device, 0x405844, 0x00ffffff);
-	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
-	nvkm_mask(device, 0x419eb4, 0x00001000, 0x00001000);
-
-	gr->func->init_ppc_exceptions(gr);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0900), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x1028), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0824), 0xc0000000);
-		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) {
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x48c), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x001ffffe);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x0000000f);
-		}
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
-	}
-
-	for (rop = 0; rop < gr->rop_nr; rop++) {
-		nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0xc0000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0xc0000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x204), 0xffffffff);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x208), 0xffffffff);
-	}
-
-	nvkm_wr32(device, 0x400108, 0xffffffff);
-	nvkm_wr32(device, 0x400138, 0xffffffff);
-	nvkm_wr32(device, 0x400118, 0xffffffff);
-	nvkm_wr32(device, 0x400130, 0xffffffff);
-	nvkm_wr32(device, 0x40011c, 0xffffffff);
-	nvkm_wr32(device, 0x400134, 0xffffffff);
-
-	nvkm_wr32(device, 0x400054, 0x34ce3464);
-
-	gf100_gr_zbc_init(gr);
-
-	return gf100_gr_init_ctxctl(gr);
 }
 
 #include "fuc/hubgk104.fuc3.h"
@@ -534,10 +455,23 @@ gk104_gr_gpccs_ucode = {
 
 static const struct gf100_gr_func
 gk104_gr = {
-	.init = gk104_gr_init,
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
+	.init_fecs_exceptions = gk104_gr_init_fecs_exceptions,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gf100_gr_init_419eb4,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gk104_gr_pack_mmio,
 	.fecs.ucode = &gk104_gr_fecs_ucode,
 	.gpccs.ucode = &gk104_gr_gpccs_ucode,
@@ -545,6 +479,7 @@ gk104_gr = {
 	.ppc_nr = 1,
 	.grctx = &gk104_grctx,
 	.clkgate_pack = gk104_clkgate_pack,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
index 4da916a9fc73..7cd628c84e07 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110.c
@@ -143,6 +143,7 @@ gk110_gr_pack_mmio[] = {
 	{ gf119_gr_init_gpm_0 },
 	{ gk110_gr_init_gpc_unk_1 },
 	{ gf100_gr_init_gcc_0 },
+	{ gk104_gr_init_gpc_unk_2 },
 	{ gk104_gr_init_tpccs_0 },
 	{ gk110_gr_init_tex_0 },
 	{ gk104_gr_init_pe_0 },
@@ -334,12 +335,39 @@ gk110_gr_gpccs_ucode = {
 	.data.size = sizeof(gk110_grgpc_data),
 };
 
+void
+gk110_gr_init_419eb4(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419eb4, 0x00001000, 0x00001000);
+	nvkm_mask(device, 0x419eb4, 0x00002000, 0x00002000);
+	nvkm_mask(device, 0x419eb4, 0x00004000, 0x00004000);
+	nvkm_mask(device, 0x419eb4, 0x00008000, 0x00008000);
+	nvkm_mask(device, 0x419eb4, 0x00001000, 0x00000000);
+	nvkm_mask(device, 0x419eb4, 0x00002000, 0x00000000);
+	nvkm_mask(device, 0x419eb4, 0x00004000, 0x00000000);
+	nvkm_mask(device, 0x419eb4, 0x00008000, 0x00000000);
+}
+
 static const struct gf100_gr_func
 gk110_gr = {
-	.init = gk104_gr_init,
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gk110_gr_init_419eb4,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gk110_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
 	.gpccs.ucode = &gk110_gr_gpccs_ucode,
@@ -347,6 +375,7 @@ gk110_gr = {
 	.ppc_nr = 2,
 	.grctx = &gk110_grctx,
 	.clkgate_pack = gk110_clkgate_pack,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
index 1912c0bfd7ee..a38faa215635 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk110b.c
@@ -82,6 +82,7 @@ gk110b_gr_pack_mmio[] = {
 	{ gf119_gr_init_gpm_0 },
 	{ gk110_gr_init_gpc_unk_1 },
 	{ gf100_gr_init_gcc_0 },
+	{ gk104_gr_init_gpc_unk_2 },
 	{ gk104_gr_init_tpccs_0 },
 	{ gk110_gr_init_tex_0 },
 	{ gk104_gr_init_pe_0 },
@@ -102,16 +103,30 @@ gk110b_gr_pack_mmio[] = {
 
 static const struct gf100_gr_func
 gk110b_gr = {
-	.init = gk104_gr_init,
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419eb4 = gk110_gr_init_419eb4,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gk110b_gr_pack_mmio,
 	.fecs.ucode = &gk110_gr_fecs_ucode,
 	.gpccs.ucode = &gk110_gr_gpccs_ucode,
 	.rops = gf100_gr_rops,
 	.ppc_nr = 2,
 	.grctx = &gk110b_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
index 1fc258163f25..58456660e603 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk208.c
@@ -121,6 +121,7 @@ gk208_gr_pack_mmio[] = {
 	{ gf119_gr_init_gpm_0 },
 	{ gk110_gr_init_gpc_unk_1 },
 	{ gf100_gr_init_gcc_0 },
+	{ gk104_gr_init_gpc_unk_2 },
 	{ gk104_gr_init_tpccs_0 },
 	{ gk208_gr_init_tex_0 },
 	{ gk104_gr_init_pe_0 },
@@ -161,16 +162,29 @@ gk208_gr_gpccs_ucode = {
 
 static const struct gf100_gr_func
 gk208_gr = {
-	.init = gk104_gr_init,
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
+	.init_400054 = gf100_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gk208_gr_pack_mmio,
 	.fecs.ucode = &gk208_gr_fecs_ucode,
 	.gpccs.ucode = &gk208_gr_gpccs_ucode,
 	.rops = gf100_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gk208_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
index de8b806b88fd..500cb08dd608 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk20a.c
@@ -219,11 +219,7 @@ int
 gk20a_gr_init(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
-	u32 data[TPC_MAX / 8] = {};
-	u8  tpcnr[GPC_MAX];
-	int gpc, tpc;
-	int ret, i;
+	int ret;
 
 	/* Clear SCC RAM */
 	nvkm_wr32(device, 0x40802c, 0x1);
@@ -246,31 +242,7 @@ gk20a_gr_init(struct gf100_gr *gr)
 	nvkm_mask(device, 0x503018, 0x1, 0x1);
 
 	/* Zcull init */
-	memset(data, 0x00, sizeof(data));
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		data[i / 8] |= tpc << ((i % 8) * 4);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
-	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
-	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
-	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
-			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 |
-			  gr->tpc_total);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
+	gr->func->init_zcull(gr);
 
 	gr->func->init_rop_active_fbps(gr);
 
@@ -310,12 +282,17 @@ gk20a_gr_init(struct gf100_gr *gr)
 
 static const struct gf100_gr_func
 gk20a_gr = {
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
 	.init = gk20a_gr_init,
+	.init_zcull = gf117_gr_init_zcull,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
+	.trap_mp = gf100_gr_trap_mp,
 	.set_hww_esr_report_mask = gk20a_gr_set_hww_esr_report_mask,
 	.rops = gf100_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gk20a_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_A },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
index 2c67fac576d1..92e31d397207 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm107.c
@@ -25,6 +25,8 @@
 #include "ctxgf100.h"
 
 #include <subdev/bios.h>
+#include <subdev/bios/bit.h>
+#include <subdev/bios/init.h>
 #include <subdev/bios/P0260.h>
 #include <subdev/fb.h>
 
@@ -36,6 +38,10 @@
 
 static const struct gf100_gr_init
 gm107_gr_init_main_0[] = {
+	{ 0x40880c,   1, 0x04, 0x00000000 },
+	{ 0x408910,   1, 0x04, 0x00000000 },
+	{ 0x408984,   1, 0x04, 0x00000000 },
+	{ 0x41a8a0,   1, 0x04, 0x00000000 },
 	{ 0x400080,   1, 0x04, 0x003003c2 },
 	{ 0x400088,   1, 0x04, 0x0001bfe7 },
 	{ 0x40008c,   1, 0x04, 0x00060000 },
@@ -210,14 +216,13 @@ gm107_gr_init_cbm_0[] = {
 static const struct gf100_gr_init
 gm107_gr_init_be_0[] = {
 	{ 0x408890,   1, 0x04, 0x000000ff },
-	{ 0x40880c,   1, 0x04, 0x00000000 },
 	{ 0x408850,   1, 0x04, 0x00000004 },
 	{ 0x408878,   1, 0x04, 0x00c81603 },
 	{ 0x40887c,   1, 0x04, 0x80543432 },
 	{ 0x408880,   1, 0x04, 0x0010581e },
 	{ 0x408884,   1, 0x04, 0x00001205 },
 	{ 0x408974,   1, 0x04, 0x000000ff },
-	{ 0x408910,   9, 0x04, 0x00000000 },
+	{ 0x408914,   8, 0x04, 0x00000000 },
 	{ 0x408950,   1, 0x04, 0x00000000 },
 	{ 0x408954,   1, 0x04, 0x0000ffff },
 	{ 0x408958,   1, 0x04, 0x00000034 },
@@ -227,7 +232,6 @@ gm107_gr_init_be_0[] = {
 	{ 0x408968,   1, 0x04, 0x02808833 },
 	{ 0x40896c,   1, 0x04, 0x01f02438 },
 	{ 0x408970,   1, 0x04, 0x00012c00 },
-	{ 0x408984,   1, 0x04, 0x00000000 },
 	{ 0x408988,   1, 0x04, 0x08040201 },
 	{ 0x40898c,   1, 0x04, 0x80402010 },
 	{}
@@ -260,6 +264,7 @@ gm107_gr_pack_mmio[] = {
 	{ gf100_gr_init_gpm_0 },
 	{ gm107_gr_init_gpc_unk_1 },
 	{ gf100_gr_init_gcc_0 },
+	{ gk104_gr_init_gpc_unk_2 },
 	{ gm107_gr_init_tpccs_0 },
 	{ gm107_gr_init_tex_0 },
 	{ gm107_gr_init_pe_0 },
@@ -280,6 +285,52 @@ gm107_gr_pack_mmio[] = {
  ******************************************************************************/
 
 void
+gm107_gr_init_400054(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x400054, 0x2c350f63);
+}
+
+void
+gm107_gr_init_shader_exceptions(struct gf100_gr *gr, int gpc, int tpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x00000005);
+}
+
+void
+gm107_gr_init_504430(struct gf100_gr *gr, int gpc, int tpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
+}
+
+static void
+gm107_gr_init_bios_2(struct gf100_gr *gr)
+{
+	struct nvkm_subdev *subdev = &gr->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	struct nvkm_bios *bios = device->bios;
+	struct bit_entry bit_P;
+	if (!bit_entry(bios, 'P', &bit_P) &&
+	    bit_P.version == 2 && bit_P.length >= 0x2c) {
+		u32 data = nvbios_rd32(bios, bit_P.offset + 0x28);
+		if (data) {
+			u8 ver = nvbios_rd08(bios, data + 0x00);
+			u8 hdr = nvbios_rd08(bios, data + 0x01);
+			if (ver == 0x20 && hdr >= 8) {
+				data = nvbios_rd32(bios, data + 0x04);
+				if (data) {
+					u32 save = nvkm_rd32(device, 0x619444);
+					nvbios_init(subdev, data);
+					nvkm_wr32(device, 0x619444, save);
+				}
+			}
+		}
+	}
+}
+
+void
 gm107_gr_init_bios(struct gf100_gr *gr)
 {
 	static const struct {
@@ -308,115 +359,17 @@ gm107_gr_init_bios(struct gf100_gr *gr)
 	}
 }
 
-static int
-gm107_gr_init(struct gf100_gr *gr)
+static void
+gm107_gr_init_gpc_mmu(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
 	struct nvkm_fb *fb = device->fb;
-	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
-	u32 data[TPC_MAX / 8] = {};
-	u8  tpcnr[GPC_MAX];
-	int gpc, tpc, rop;
-	int i;
 
 	nvkm_wr32(device, GPC_BCAST(0x0880), 0x00000000);
 	nvkm_wr32(device, GPC_BCAST(0x0890), 0x00000000);
 	nvkm_wr32(device, GPC_BCAST(0x0894), 0x00000000);
 	nvkm_wr32(device, GPC_BCAST(0x08b4), nvkm_memory_addr(fb->mmu_wr) >> 8);
 	nvkm_wr32(device, GPC_BCAST(0x08b8), nvkm_memory_addr(fb->mmu_rd) >> 8);
-
-	gf100_gr_mmio(gr, gr->func->mmio);
-
-	gm107_gr_init_bios(gr);
-
-	nvkm_wr32(device, GPC_UNIT(0, 0x3018), 0x00000001);
-
-	memset(data, 0x00, sizeof(data));
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		data[i / 8] |= tpc << ((i % 8) * 4);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
-	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
-	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
-	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
-			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 |
-							 gr->tpc_total);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
-	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
-
-	gr->func->init_rop_active_fbps(gr);
-
-	nvkm_wr32(device, 0x400500, 0x00010001);
-
-	nvkm_wr32(device, 0x400100, 0xffffffff);
-	nvkm_wr32(device, 0x40013c, 0xffffffff);
-	nvkm_wr32(device, 0x400124, 0x00000002);
-	nvkm_wr32(device, 0x409c24, 0x000e0000);
-
-	nvkm_wr32(device, 0x404000, 0xc0000000);
-	nvkm_wr32(device, 0x404600, 0xc0000000);
-	nvkm_wr32(device, 0x408030, 0xc0000000);
-	nvkm_wr32(device, 0x404490, 0xc0000000);
-	nvkm_wr32(device, 0x406018, 0xc0000000);
-	nvkm_wr32(device, 0x407020, 0x40000000);
-	nvkm_wr32(device, 0x405840, 0xc0000000);
-	nvkm_wr32(device, 0x405844, 0x00ffffff);
-	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
-
-	gr->func->init_ppc_exceptions(gr);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0900), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x1028), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0824), 0xc0000000);
-		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) {
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x48c), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x00000005);
-		}
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
-	}
-
-	for (rop = 0; rop < gr->rop_nr; rop++) {
-		nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0x40000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0x40000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x204), 0xffffffff);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x208), 0xffffffff);
-	}
-
-	nvkm_wr32(device, 0x400108, 0xffffffff);
-	nvkm_wr32(device, 0x400138, 0xffffffff);
-	nvkm_wr32(device, 0x400118, 0xffffffff);
-	nvkm_wr32(device, 0x400130, 0xffffffff);
-	nvkm_wr32(device, 0x40011c, 0xffffffff);
-	nvkm_wr32(device, 0x400134, 0xffffffff);
-
-	nvkm_wr32(device, 0x400054, 0x2c350f63);
-
-	gf100_gr_zbc_init(gr);
-
-	return gf100_gr_init_ctxctl(gr);
 }
 
 #include "fuc/hubgm107.fuc5.h"
@@ -441,15 +394,32 @@ gm107_gr_gpccs_ucode = {
 
 static const struct gf100_gr_func
 gm107_gr = {
-	.init = gm107_gr_init,
+	.oneinit_tiles = gf100_gr_oneinit_tiles,
+	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
+	.init = gf100_gr_init,
+	.init_gpc_mmu = gm107_gr_init_gpc_mmu,
+	.init_bios = gm107_gr_init_bios,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
+	.init_bios_2 = gm107_gr_init_bios_2,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_504430 = gm107_gr_init_504430,
+	.init_shader_exceptions = gm107_gr_init_shader_exceptions,
+	.init_400054 = gm107_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.mmio = gm107_gr_pack_mmio,
 	.fecs.ucode = &gm107_gr_fecs_ucode,
 	.gpccs.ucode = &gm107_gr_gpccs_ucode,
 	.rops = gf100_gr_rops,
 	.ppc_nr = 2,
 	.grctx = &gm107_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
index 6435f1257572..eff30662b984 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm200.c
@@ -39,6 +39,22 @@ gm200_gr_rops(struct gf100_gr *gr)
 }
 
 void
+gm200_gr_init_ds_hww_esr_2(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, 0x405848, 0xc0000000);
+	nvkm_mask(device, 0x40584c, 0x00000001, 0x00000001);
+}
+
+void
+gm200_gr_init_num_active_ltcs(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
+	nvkm_wr32(device, GPC_BCAST(0x033c), nvkm_rd32(device, 0x100804));
+}
+
+void
 gm200_gr_init_gpc_mmu(struct gf100_gr *gr)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
@@ -61,111 +77,51 @@ gm200_gr_init_rop_active_fbps(struct gf100_gr *gr)
 	nvkm_mask(device, 0x408958, 0x0000000f, fbp_count); /* crop */
 }
 
-int
-gm200_gr_init(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
-	u32 data[TPC_MAX / 8] = {};
-	u8  tpcnr[GPC_MAX];
-	int gpc, tpc, rop;
-	int i;
-
-	gr->func->init_gpc_mmu(gr);
-
-	gf100_gr_mmio(gr, gr->fuc_sw_nonctx);
-
-	gm107_gr_init_bios(gr);
-
-	nvkm_wr32(device, GPC_UNIT(0, 0x3018), 0x00000001);
-
-	memset(data, 0x00, sizeof(data));
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		data[i / 8] |= tpc << ((i % 8) * 4);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
-	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
-	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
-	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
-			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 |
-							 gr->tpc_total);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
-	}
+static u8
+gm200_gr_tile_map_6_24[] = {
+	0, 1, 2, 3, 4, 5, 3, 4, 5, 0, 1, 2, 0, 1, 2, 3, 4, 5, 3, 4, 5, 0, 1, 2,
+};
 
-	nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
-	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
-	nvkm_wr32(device, GPC_BCAST(0x033c), nvkm_rd32(device, 0x100804));
+static u8
+gm200_gr_tile_map_4_16[] = {
+	0, 1, 2, 3, 2, 3, 0, 1, 3, 0, 1, 2, 1, 2, 3, 0,
+};
 
-	gr->func->init_rop_active_fbps(gr);
+static u8
+gm200_gr_tile_map_2_8[] = {
+	0, 1, 1, 0, 0, 1, 1, 0,
+};
 
-	nvkm_wr32(device, 0x400500, 0x00010001);
-	nvkm_wr32(device, 0x400100, 0xffffffff);
-	nvkm_wr32(device, 0x40013c, 0xffffffff);
-	nvkm_wr32(device, 0x400124, 0x00000002);
-	nvkm_wr32(device, 0x409c24, 0x000e0000);
-	nvkm_wr32(device, 0x405848, 0xc0000000);
-	nvkm_wr32(device, 0x40584c, 0x00000001);
-	nvkm_wr32(device, 0x404000, 0xc0000000);
-	nvkm_wr32(device, 0x404600, 0xc0000000);
-	nvkm_wr32(device, 0x408030, 0xc0000000);
-	nvkm_wr32(device, 0x404490, 0xc0000000);
-	nvkm_wr32(device, 0x406018, 0xc0000000);
-	nvkm_wr32(device, 0x407020, 0x40000000);
-	nvkm_wr32(device, 0x405840, 0xc0000000);
-	nvkm_wr32(device, 0x405844, 0x00ffffff);
-	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
-
-	gr->func->init_ppc_exceptions(gr);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0900), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x1028), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0824), 0xc0000000);
-		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) {
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x48c), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x00000005);
-		}
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
-	}
+void
+gm200_gr_oneinit_sm_id(struct gf100_gr *gr)
+{
+	/*XXX: There's a different algorithm here I've not yet figured out. */
+	gf100_gr_oneinit_sm_id(gr);
+}
 
-	for (rop = 0; rop < gr->rop_nr; rop++) {
-		nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0x40000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0x40000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x204), 0xffffffff);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x208), 0xffffffff);
+void
+gm200_gr_oneinit_tiles(struct gf100_gr *gr)
+{
+	/*XXX: Not sure what this is about.  The algorithm from NVGPU
+	 *     seems to work for all boards I tried from earlier (and
+	 *     later) GPUs except in these specific configurations.
+	 *
+	 *     Let's just hardcode them for now.
+	 */
+	if (gr->gpc_nr == 2 && gr->tpc_total == 8) {
+		memcpy(gr->tile, gm200_gr_tile_map_2_8, gr->tpc_total);
+		gr->screen_tile_row_offset = 1;
+	} else
+	if (gr->gpc_nr == 4 && gr->tpc_total == 16) {
+		memcpy(gr->tile, gm200_gr_tile_map_4_16, gr->tpc_total);
+		gr->screen_tile_row_offset = 4;
+	} else
+	if (gr->gpc_nr == 6 && gr->tpc_total == 24) {
+		memcpy(gr->tile, gm200_gr_tile_map_6_24, gr->tpc_total);
+		gr->screen_tile_row_offset = 5;
+	} else {
+		gf100_gr_oneinit_tiles(gr);
 	}
-
-	nvkm_wr32(device, 0x400108, 0xffffffff);
-	nvkm_wr32(device, 0x400138, 0xffffffff);
-	nvkm_wr32(device, 0x400118, 0xffffffff);
-	nvkm_wr32(device, 0x400130, 0xffffffff);
-	nvkm_wr32(device, 0x40011c, 0xffffffff);
-	nvkm_wr32(device, 0x400134, 0xffffffff);
-
-	nvkm_wr32(device, 0x400054, 0x2c350f63);
-
-	gf100_gr_zbc_init(gr);
-
-	return gf100_gr_init_ctxctl(gr);
 }
 
 int
@@ -208,13 +164,30 @@ gm200_gr_new_(const struct gf100_gr_func *func, struct nvkm_device *device,
 
 static const struct gf100_gr_func
 gm200_gr = {
-	.init = gm200_gr_init,
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
+	.init_bios = gm107_gr_init_bios,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gm200_gr_init_rop_active_fbps,
+	.init_fecs_exceptions = gf100_gr_init_fecs_exceptions,
+	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_504430 = gm107_gr_init_504430,
+	.init_shader_exceptions = gm107_gr_init_shader_exceptions,
+	.init_400054 = gm107_gr_init_400054,
+	.trap_mp = gf100_gr_trap_mp,
 	.rops = gm200_gr_rops,
+	.tpc_nr = 4,
 	.ppc_nr = 2,
 	.grctx = &gm200_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
index 69479af1d829..a667770ce3cb 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gm20b.c
@@ -64,13 +64,18 @@ gm20b_gr_set_hww_esr_report_mask(struct gf100_gr *gr)
 
 static const struct gf100_gr_func
 gm20b_gr = {
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
 	.init = gk20a_gr_init,
+	.init_zcull = gf117_gr_init_zcull,
 	.init_gpc_mmu = gm20b_gr_init_gpc_mmu,
 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
+	.trap_mp = gf100_gr_trap_mp,
 	.set_hww_esr_report_mask = gm20b_gr_set_hww_esr_report_mask,
 	.rops = gm200_gr_rops,
 	.ppc_nr = 1,
 	.grctx = &gm20b_grctx,
+	.zbc = &gf100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
index 867a5f7cc5bc..9d0521ce309a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp100.c
@@ -29,143 +29,103 @@
 /*******************************************************************************
  * PGRAPH engine/subdev functions
  ******************************************************************************/
-
 void
-gp100_gr_init_rop_active_fbps(struct gf100_gr *gr)
+gp100_gr_zbc_clear_color(struct gf100_gr *gr, int zbc)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	/*XXX: otherwise identical to gm200 aside from mask.. do everywhere? */
-	const u32 fbp_count = nvkm_rd32(device, 0x12006c) & 0x0000000f;
-	nvkm_mask(device, 0x408850, 0x0000000f, fbp_count); /* zrop */
-	nvkm_mask(device, 0x408958, 0x0000000f, fbp_count); /* crop */
+	const int znum =  zbc - 1;
+	const u32 zoff = znum * 4;
+
+	if (gr->zbc_color[zbc].format) {
+		nvkm_wr32(device, 0x418010 + zoff, gr->zbc_color[zbc].ds[0]);
+		nvkm_wr32(device, 0x41804c + zoff, gr->zbc_color[zbc].ds[1]);
+		nvkm_wr32(device, 0x418088 + zoff, gr->zbc_color[zbc].ds[2]);
+		nvkm_wr32(device, 0x4180c4 + zoff, gr->zbc_color[zbc].ds[3]);
+	}
+
+	nvkm_mask(device, 0x418100 + ((znum / 4) * 4),
+			  0x0000007f << ((znum % 4) * 7),
+			  gr->zbc_color[zbc].format << ((znum % 4) * 7));
 }
 
 void
-gp100_gr_init_num_active_ltcs(struct gf100_gr *gr)
+gp100_gr_zbc_clear_depth(struct gf100_gr *gr, int zbc)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-
-	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
-	nvkm_wr32(device, GPC_BCAST(0x033c), nvkm_rd32(device, 0x100804));
+	const int znum =  zbc - 1;
+	const u32 zoff = znum * 4;
+
+	if (gr->zbc_depth[zbc].format)
+		nvkm_wr32(device, 0x418110 + zoff, gr->zbc_depth[zbc].ds);
+	nvkm_mask(device, 0x41814c + ((znum / 4) * 4),
+			  0x0000007f << ((znum % 4) * 7),
+			  gr->zbc_depth[zbc].format << ((znum % 4) * 7));
 }
 
-int
-gp100_gr_init(struct gf100_gr *gr)
+static const struct gf100_gr_func_zbc
+gp100_gr_zbc = {
+	.clear_color = gp100_gr_zbc_clear_color,
+	.clear_depth = gp100_gr_zbc_clear_depth,
+};
+
+void
+gp100_gr_init_shader_exceptions(struct gf100_gr *gr, int gpc, int tpc)
 {
 	struct nvkm_device *device = gr->base.engine.subdev.device;
-	const u32 magicgpc918 = DIV_ROUND_UP(0x00800000, gr->tpc_total);
-	u32 data[TPC_MAX / 8] = {};
-	u8  tpcnr[GPC_MAX];
-	int gpc, tpc, rop;
-	int i;
-
-	gr->func->init_gpc_mmu(gr);
-
-	gf100_gr_mmio(gr, gr->fuc_sw_nonctx);
-
-	nvkm_wr32(device, GPC_UNIT(0, 0x3018), 0x00000001);
-
-	memset(data, 0x00, sizeof(data));
-	memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
-	for (i = 0, gpc = -1; i < gr->tpc_total; i++) {
-		do {
-			gpc = (gpc + 1) % gr->gpc_nr;
-		} while (!tpcnr[gpc]);
-		tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
-
-		data[i / 8] |= tpc << ((i % 8) * 4);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x0980), data[0]);
-	nvkm_wr32(device, GPC_BCAST(0x0984), data[1]);
-	nvkm_wr32(device, GPC_BCAST(0x0988), data[2]);
-	nvkm_wr32(device, GPC_BCAST(0x098c), data[3]);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0914),
-			  gr->screen_tile_row_offset << 8 | gr->tpc_nr[gpc]);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0910), 0x00040000 |
-							 gr->tpc_total);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0918), magicgpc918);
-	}
-
-	nvkm_wr32(device, GPC_BCAST(0x3fd4), magicgpc918);
-	gr->func->init_num_active_ltcs(gr);
-
-	gr->func->init_rop_active_fbps(gr);
-	if (gr->func->init_swdx_pes_mask)
-		gr->func->init_swdx_pes_mask(gr);
-
-	nvkm_wr32(device, 0x400500, 0x00010001);
-	nvkm_wr32(device, 0x400100, 0xffffffff);
-	nvkm_wr32(device, 0x40013c, 0xffffffff);
-	nvkm_wr32(device, 0x400124, 0x00000002);
-	nvkm_wr32(device, 0x409c24, 0x000f0002);
-	nvkm_wr32(device, 0x405848, 0xc0000000);
-	nvkm_mask(device, 0x40584c, 0x00000000, 0x00000001);
-	nvkm_wr32(device, 0x404000, 0xc0000000);
-	nvkm_wr32(device, 0x404600, 0xc0000000);
-	nvkm_wr32(device, 0x408030, 0xc0000000);
-	nvkm_wr32(device, 0x404490, 0xc0000000);
-	nvkm_wr32(device, 0x406018, 0xc0000000);
-	nvkm_wr32(device, 0x407020, 0x40000000);
-	nvkm_wr32(device, 0x405840, 0xc0000000);
-	nvkm_wr32(device, 0x405844, 0x00ffffff);
-	nvkm_mask(device, 0x419cc0, 0x00000008, 0x00000008);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x00000105);
+}
 
+static void
+gp100_gr_init_419c9c(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
 	nvkm_mask(device, 0x419c9c, 0x00010000, 0x00010000);
 	nvkm_mask(device, 0x419c9c, 0x00020000, 0x00020000);
+}
 
-	gr->func->init_ppc_exceptions(gr);
-
-	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0420), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0900), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x1028), 0xc0000000);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x0824), 0xc0000000);
-		for (tpc = 0; tpc < gr->tpc_nr[gpc]; tpc++) {
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x508), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x50c), 0xffffffff);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x224), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x48c), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x084), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0xc0000000);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x644), 0x00dffffe);
-			nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x64c), 0x00000105);
-		}
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c90), 0xffffffff);
-		nvkm_wr32(device, GPC_UNIT(gpc, 0x2c94), 0xffffffff);
-	}
-
-	for (rop = 0; rop < gr->rop_nr; rop++) {
-		nvkm_wr32(device, ROP_UNIT(rop, 0x144), 0x40000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x070), 0x40000000);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x204), 0xffffffff);
-		nvkm_wr32(device, ROP_UNIT(rop, 0x208), 0xffffffff);
-	}
-
-	nvkm_wr32(device, 0x400108, 0xffffffff);
-	nvkm_wr32(device, 0x400138, 0xffffffff);
-	nvkm_wr32(device, 0x400118, 0xffffffff);
-	nvkm_wr32(device, 0x400130, 0xffffffff);
-	nvkm_wr32(device, 0x40011c, 0xffffffff);
-	nvkm_wr32(device, 0x400134, 0xffffffff);
-
-	gf100_gr_zbc_init(gr);
+void
+gp100_gr_init_fecs_exceptions(struct gf100_gr *gr)
+{
+	nvkm_wr32(gr->base.engine.subdev.device, 0x409c24, 0x000f0002);
+}
 
-	return gf100_gr_init_ctxctl(gr);
+void
+gp100_gr_init_rop_active_fbps(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	/*XXX: otherwise identical to gm200 aside from mask.. do everywhere? */
+	const u32 fbp_count = nvkm_rd32(device, 0x12006c) & 0x0000000f;
+	nvkm_mask(device, 0x408850, 0x0000000f, fbp_count); /* zrop */
+	nvkm_mask(device, 0x408958, 0x0000000f, fbp_count); /* crop */
 }
 
 static const struct gf100_gr_func
 gp100_gr = {
-	.init = gp100_gr_init,
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
+	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
+	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_419c9c = gp100_gr_init_419c9c,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
-	.init_num_active_ltcs = gp100_gr_init_num_active_ltcs,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_504430 = gm107_gr_init_504430,
+	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
+	.trap_mp = gf100_gr_trap_mp,
 	.rops = gm200_gr_rops,
+	.gpc_nr = 6,
+	.tpc_nr = 5,
 	.ppc_nr = 2,
 	.grctx = &gp100_grctx,
+	.zbc = &gp100_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
index 61e3a0b08559..37f7d739bf80 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp102.c
@@ -26,6 +26,62 @@
 
 #include <nvif/class.h>
 
+static void
+gp102_gr_zbc_clear_stencil(struct gf100_gr *gr, int zbc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	const int znum =  zbc - 1;
+	const u32 zoff = znum * 4;
+
+	if (gr->zbc_stencil[zbc].format)
+		nvkm_wr32(device, 0x41815c + zoff, gr->zbc_stencil[zbc].ds);
+	nvkm_mask(device, 0x418198 + ((znum / 4) * 4),
+			  0x0000007f << ((znum % 4) * 7),
+			  gr->zbc_stencil[zbc].format << ((znum % 4) * 7));
+}
+
+static int
+gp102_gr_zbc_stencil_get(struct gf100_gr *gr, int format,
+			 const u32 ds, const u32 l2)
+{
+	struct nvkm_ltc *ltc = gr->base.engine.subdev.device->ltc;
+	int zbc = -ENOSPC, i;
+
+	for (i = ltc->zbc_min; i <= ltc->zbc_max; i++) {
+		if (gr->zbc_stencil[i].format) {
+			if (gr->zbc_stencil[i].format != format)
+				continue;
+			if (gr->zbc_stencil[i].ds != ds)
+				continue;
+			if (gr->zbc_stencil[i].l2 != l2) {
+				WARN_ON(1);
+				return -EINVAL;
+			}
+			return i;
+		} else {
+			zbc = (zbc < 0) ? i : zbc;
+		}
+	}
+
+	if (zbc < 0)
+		return zbc;
+
+	gr->zbc_stencil[zbc].format = format;
+	gr->zbc_stencil[zbc].ds = ds;
+	gr->zbc_stencil[zbc].l2 = l2;
+	nvkm_ltc_zbc_stencil_get(ltc, zbc, l2);
+	gr->func->zbc->clear_stencil(gr, zbc);
+	return zbc;
+}
+
+const struct gf100_gr_func_zbc
+gp102_gr_zbc = {
+	.clear_color = gp100_gr_zbc_clear_color,
+	.clear_depth = gp100_gr_zbc_clear_depth,
+	.stencil_get = gp102_gr_zbc_stencil_get,
+	.clear_stencil = gp102_gr_zbc_clear_stencil,
+};
+
 void
 gp102_gr_init_swdx_pes_mask(struct gf100_gr *gr)
 {
@@ -42,15 +98,30 @@ gp102_gr_init_swdx_pes_mask(struct gf100_gr *gr)
 
 static const struct gf100_gr_func
 gp102_gr = {
-	.init = gp100_gr_init,
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
-	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
-	.init_num_active_ltcs = gp100_gr_init_num_active_ltcs,
+	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
+	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_504430 = gm107_gr_init_504430,
+	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
+	.trap_mp = gf100_gr_trap_mp,
 	.rops = gm200_gr_rops,
+	.gpc_nr = 6,
+	.tpc_nr = 5,
 	.ppc_nr = 3,
 	.grctx = &gp102_grctx,
+	.zbc = &gp102_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
new file mode 100644
index 000000000000..4573c914c021
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp104.c
@@ -0,0 +1,66 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "gf100.h"
+#include "ctxgf100.h"
+
+#include <nvif/class.h>
+
+static const struct gf100_gr_func
+gp104_gr = {
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
+	.init = gf100_gr_init,
+	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
+	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
+	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
+	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
+	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_504430 = gm107_gr_init_504430,
+	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
+	.trap_mp = gf100_gr_trap_mp,
+	.rops = gm200_gr_rops,
+	.gpc_nr = 6,
+	.tpc_nr = 5,
+	.ppc_nr = 3,
+	.grctx = &gp104_grctx,
+	.zbc = &gp102_gr_zbc,
+	.sclass = {
+		{ -1, -1, FERMI_TWOD_A },
+		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
+		{ -1, -1, PASCAL_B, &gf100_fermi },
+		{ -1, -1, PASCAL_COMPUTE_B },
+		{}
+	}
+};
+
+int
+gp104_gr_new(struct nvkm_device *device, int index, struct nvkm_gr **pgr)
+{
+	return gm200_gr_new_(&gp104_gr, device, index, pgr);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
index f7272323f694..812aba91653f 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp107.c
@@ -28,15 +28,30 @@
 
 static const struct gf100_gr_func
 gp107_gr = {
-	.init = gp100_gr_init,
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
-	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
 	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
-	.init_num_active_ltcs = gp100_gr_init_num_active_ltcs,
+	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
+	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
+	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_504430 = gm107_gr_init_504430,
+	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
+	.trap_mp = gf100_gr_trap_mp,
 	.rops = gm200_gr_rops,
+	.gpc_nr = 2,
+	.tpc_nr = 3,
 	.ppc_nr = 1,
 	.grctx = &gp107_grctx,
+	.zbc = &gp102_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
index 5f3d161a0842..303dceddd4a8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gp10b.c
@@ -25,24 +25,31 @@
 
 #include <nvif/class.h>
 
-static void
-gp10b_gr_init_num_active_ltcs(struct gf100_gr *gr)
-{
-	struct nvkm_device *device = gr->base.engine.subdev.device;
-
-	nvkm_wr32(device, GPC_BCAST(0x08ac), nvkm_rd32(device, 0x100800));
-}
-
 static const struct gf100_gr_func
 gp10b_gr = {
-	.init = gp100_gr_init,
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
+	.init = gf100_gr_init,
 	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
 	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
+	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
+	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_419cc0 = gf100_gr_init_419cc0,
 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
-	.init_num_active_ltcs = gp10b_gr_init_num_active_ltcs,
+	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
+	.init_504430 = gm107_gr_init_504430,
+	.init_shader_exceptions = gp100_gr_init_shader_exceptions,
+	.trap_mp = gf100_gr_trap_mp,
 	.rops = gm200_gr_rops,
+	.gpc_nr = 1,
+	.tpc_nr = 2,
 	.ppc_nr = 1,
 	.grctx = &gp102_grctx,
+	.zbc = &gp102_gr_zbc,
 	.sclass = {
 		{ -1, -1, FERMI_TWOD_A },
 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/gv100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gv100.c
new file mode 100644
index 000000000000..3b3327789ae7
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/gv100.c
@@ -0,0 +1,127 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "gf100.h"
+#include "ctxgf100.h"
+
+#include <nvif/class.h>
+
+static void
+gv100_gr_trap_sm(struct gf100_gr *gr, int gpc, int tpc, int sm)
+{
+	struct nvkm_subdev *subdev = &gr->base.engine.subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 werr = nvkm_rd32(device, TPC_UNIT(gpc, tpc, 0x730 + (sm * 0x80)));
+	u32 gerr = nvkm_rd32(device, TPC_UNIT(gpc, tpc, 0x734 + (sm * 0x80)));
+	const struct nvkm_enum *warp;
+	char glob[128];
+
+	nvkm_snprintbf(glob, sizeof(glob), gf100_mp_global_error, gerr);
+	warp = nvkm_enum_find(gf100_mp_warp_error, werr & 0xffff);
+
+	nvkm_error(subdev, "GPC%i/TPC%i/SM%d trap: "
+			   "global %08x [%s] warp %04x [%s]\n",
+		   gpc, tpc, sm, gerr, glob, werr, warp ? warp->name : "");
+
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x730 + sm * 0x80), 0x00000000);
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x734 + sm * 0x80), gerr);
+}
+
+static void
+gv100_gr_trap_mp(struct gf100_gr *gr, int gpc, int tpc)
+{
+	gv100_gr_trap_sm(gr, gpc, tpc, 0);
+	gv100_gr_trap_sm(gr, gpc, tpc, 1);
+}
+
+static void
+gv100_gr_init_4188a4(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x4188a4, 0x03000000, 0x03000000);
+}
+
+static void
+gv100_gr_init_shader_exceptions(struct gf100_gr *gr, int gpc, int tpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	int sm;
+	for (sm = 0; sm < 0x100; sm += 0x80) {
+		nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x728 + sm), 0x0085eb64);
+		nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x610), 0x00000001);
+		nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x72c + sm), 0x00000004);
+	}
+}
+
+static void
+gv100_gr_init_504430(struct gf100_gr *gr, int gpc, int tpc)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x430), 0x403f0000);
+}
+
+static void
+gv100_gr_init_419bd8(struct gf100_gr *gr)
+{
+	struct nvkm_device *device = gr->base.engine.subdev.device;
+	nvkm_mask(device, 0x419bd8, 0x00000700, 0x00000000);
+}
+
+static const struct gf100_gr_func
+gv100_gr = {
+	.oneinit_tiles = gm200_gr_oneinit_tiles,
+	.oneinit_sm_id = gm200_gr_oneinit_sm_id,
+	.init = gf100_gr_init,
+	.init_419bd8 = gv100_gr_init_419bd8,
+	.init_gpc_mmu = gm200_gr_init_gpc_mmu,
+	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
+	.init_zcull = gf117_gr_init_zcull,
+	.init_num_active_ltcs = gm200_gr_init_num_active_ltcs,
+	.init_rop_active_fbps = gp100_gr_init_rop_active_fbps,
+	.init_swdx_pes_mask = gp102_gr_init_swdx_pes_mask,
+	.init_fecs_exceptions = gp100_gr_init_fecs_exceptions,
+	.init_ds_hww_esr_2 = gm200_gr_init_ds_hww_esr_2,
+	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
+	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
+	.init_504430 = gv100_gr_init_504430,
+	.init_shader_exceptions = gv100_gr_init_shader_exceptions,
+	.init_4188a4 = gv100_gr_init_4188a4,
+	.trap_mp = gv100_gr_trap_mp,
+	.rops = gm200_gr_rops,
+	.gpc_nr = 6,
+	.tpc_nr = 5,
+	.ppc_nr = 3,
+	.grctx = &gv100_grctx,
+	.zbc = &gp102_gr_zbc,
+	.sclass = {
+		{ -1, -1, FERMI_TWOD_A },
+		{ -1, -1, KEPLER_INLINE_TO_MEMORY_B },
+		{ -1, -1, VOLTA_A, &gf100_fermi },
+		{ -1, -1, VOLTA_COMPUTE_A },
+		{}
+	}
+};
+
+int
+gv100_gr_new(struct nvkm_device *device, int index, struct nvkm_gr **pgr)
+{
+	return gm200_gr_new_(&gv100_gr, device, index, pgr);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/pm/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/pm/base.c
index 53859b6254d6..b2785bee418e 100644
--- a/drivers/gpu/drm/nouveau/nvkm/engine/pm/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/engine/pm/base.c
@@ -779,8 +779,8 @@ nvkm_perfdom_new(struct nvkm_pm *pm, const char *name, u32 mask,
 
 		sdom = spec;
 		while (sdom->signal_nr) {
-			dom = kzalloc(sizeof(*dom) + sdom->signal_nr *
-				      sizeof(*dom->signal), GFP_KERNEL);
+			dom = kzalloc(struct_size(dom, signal, sdom->signal_nr),
+				      GFP_KERNEL);
 			if (!dom)
 				return -ENOMEM;
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.c b/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.c
index 58a59b7db2e5..771e16a16267 100644
--- a/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.c
+++ b/drivers/gpu/drm/nouveau/nvkm/falcon/msgqueue.c
@@ -506,6 +506,7 @@ nvkm_msgqueue_new(u32 version, struct nvkm_falcon *falcon,
 		break;
 	case 0x0148cdec:
 	case 0x015ccf3e:
+	case 0x0167d263:
 		ret = msgqueue_0148cdec_new(falcon, sb, queue);
 		break;
 	default:
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/Kbuild
index 3f5d38d74fba..cfdffef1afb9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/Kbuild
@@ -3,6 +3,7 @@ include $(src)/nvkm/subdev/bios/Kbuild
 include $(src)/nvkm/subdev/bus/Kbuild
 include $(src)/nvkm/subdev/clk/Kbuild
 include $(src)/nvkm/subdev/devinit/Kbuild
+include $(src)/nvkm/subdev/fault/Kbuild
 include $(src)/nvkm/subdev/fb/Kbuild
 include $(src)/nvkm/subdev/fuse/Kbuild
 include $(src)/nvkm/subdev/gpio/Kbuild
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/dp.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/dp.c
index 7c7efa4ea0d0..3133b28f849c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/dp.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/dp.c
@@ -25,7 +25,7 @@
 #include <subdev/bios/bit.h>
 #include <subdev/bios/dp.h>
 
-static u16
+u16
 nvbios_dp_table(struct nvkm_bios *bios, u8 *ver, u8 *hdr, u8 *cnt, u8 *len)
 {
 	struct bit_entry d;
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/iccsense.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/iccsense.c
index 73e463ed55c3..dea444d48f94 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/iccsense.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/iccsense.c
@@ -73,7 +73,8 @@ nvbios_iccsense_parse(struct nvkm_bios *bios, struct nvbios_iccsense *iccsense)
 	}
 
 	iccsense->nr_entry = cnt;
-	iccsense->rail = kmalloc(sizeof(struct pwr_rail_t) * cnt, GFP_KERNEL);
+	iccsense->rail = kmalloc_array(cnt, sizeof(struct pwr_rail_t),
+				       GFP_KERNEL);
 	if (!iccsense->rail)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pll.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pll.c
index 2ca23a9157ab..e6e804cee2bc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pll.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/pll.c
@@ -193,7 +193,10 @@ pll_map_type(struct nvkm_bios *bios, u8 type, u32 *reg, u8 *ver, u8 *len)
 		data += hdr;
 		while (cnt--) {
 			if (nvbios_rd08(bios, data + 0) == type) {
-				*reg = nvbios_rd32(bios, data + 3);
+				if (*ver < 0x50)
+					*reg = nvbios_rd32(bios, data + 3);
+				else
+					*reg = 0;
 				return data;
 			}
 			data += *len;
@@ -361,6 +364,20 @@ nvbios_pll_parse(struct nvkm_bios *bios, u32 type, struct nvbios_pll *info)
 		info->min_p = nvbios_rd08(bios, data + 12);
 		info->max_p = nvbios_rd08(bios, data + 13);
 		break;
+	case 0x50:
+		info->refclk = nvbios_rd16(bios, data + 1) * 1000;
+		/* info->refclk_alt = nvbios_rd16(bios, data + 3) * 1000; */
+		info->vco1.min_freq = nvbios_rd16(bios, data + 5) * 1000;
+		info->vco1.max_freq = nvbios_rd16(bios, data + 7) * 1000;
+		info->vco1.min_inputfreq = nvbios_rd16(bios, data + 9) * 1000;
+		info->vco1.max_inputfreq = nvbios_rd16(bios, data + 11) * 1000;
+		info->vco1.min_m = nvbios_rd08(bios, data + 13);
+		info->vco1.max_m = nvbios_rd08(bios, data + 14);
+		info->vco1.min_n = nvbios_rd08(bios, data + 15);
+		info->vco1.max_n = nvbios_rd08(bios, data + 16);
+		info->min_p = nvbios_rd08(bios, data + 17);
+		info->max_p = nvbios_rd08(bios, data + 18);
+		break;
 	default:
 		nvkm_error(subdev, "unknown pll limits version 0x%02x\n", ver);
 		return -EINVAL;
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowramin.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowramin.c
index 0f537c22804c..3634cd0630b8 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowramin.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/shadowramin.c
@@ -78,7 +78,10 @@ pramin_init(struct nvkm_bios *bios, const char *name)
 	 * important as we don't want to be touching vram on an
 	 * uninitialised board
 	 */
-	addr = nvkm_rd32(device, 0x619f04);
+	if (device->card_type >= GV100)
+		addr = nvkm_rd32(device, 0x625f04);
+	else
+		addr = nvkm_rd32(device, 0x619f04);
 	if (!(addr & 0x00000008)) {
 		nvkm_debug(subdev, "... not enabled\n");
 		return ERR_PTR(-ENODEV);
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/vpstate.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/vpstate.c
index 20b6fc8243e0..71524548de32 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/vpstate.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/vpstate.c
@@ -58,8 +58,14 @@ nvbios_vpstate_parse(struct nvkm_bios *b, struct nvbios_vpstate_header *h)
 		h->ecount   = nvbios_rd08(b, h->offset + 0x5);
 
 		h->base_id  = nvbios_rd08(b, h->offset + 0x0f);
-		h->boost_id = nvbios_rd08(b, h->offset + 0x10);
-		h->tdp_id   = nvbios_rd08(b, h->offset + 0x11);
+		if (h->hlen > 0x10)
+			h->boost_id = nvbios_rd08(b, h->offset + 0x10);
+		else
+			h->boost_id = 0xff;
+		if (h->hlen > 0x11)
+			h->tdp_id = nvbios_rd08(b, h->offset + 0x11);
+		else
+			h->tdp_id = 0xff;
 		return 0;
 	default:
 		return -EINVAL;
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c
index 81c3567d4e67..ba6a868d4c95 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/clk/base.c
@@ -109,18 +109,17 @@ nvkm_cstate_valid(struct nvkm_clk *clk, struct nvkm_cstate *cstate,
 
 static struct nvkm_cstate *
 nvkm_cstate_find_best(struct nvkm_clk *clk, struct nvkm_pstate *pstate,
-		      struct nvkm_cstate *start)
+		      struct nvkm_cstate *cstate)
 {
 	struct nvkm_device *device = clk->subdev.device;
 	struct nvkm_volt *volt = device->volt;
-	struct nvkm_cstate *cstate;
 	int max_volt;
 
-	if (!pstate || !start)
+	if (!pstate || !cstate)
 		return NULL;
 
 	if (!volt)
-		return start;
+		return cstate;
 
 	max_volt = volt->max_uv;
 	if (volt->max0_id != 0xff)
@@ -133,8 +132,7 @@ nvkm_cstate_find_best(struct nvkm_clk *clk, struct nvkm_pstate *pstate,
 		max_volt = min(max_volt,
 			       nvkm_volt_map(volt, volt->max2_id, clk->temp));
 
-	for (cstate = start; &cstate->head != &pstate->list;
-	     cstate = list_prev_entry(cstate, head)) {
+	list_for_each_entry_from_reverse(cstate, &pstate->list, head) {
 		if (nvkm_cstate_valid(clk, cstate, max_volt, clk->temp))
 			break;
 	}
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/Kbuild
index eac88e3dc6e5..50a436926484 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/Kbuild
@@ -12,3 +12,4 @@ nvkm-y += nvkm/subdev/devinit/mcp89.o
 nvkm-y += nvkm/subdev/devinit/gf100.o
 nvkm-y += nvkm/subdev/devinit/gm107.o
 nvkm-y += nvkm/subdev/devinit/gm200.o
+nvkm-y += nvkm/subdev/devinit/gv100.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gm200.c b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gm200.c
index 1730371933df..b80618e35491 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gm200.c
@@ -107,7 +107,7 @@ pmu_load(struct nv50_devinit *init, u8 type, bool post,
 	return pmu_exec(init, pmu.init_addr_pmu), 0;
 }
 
-static int
+int
 gm200_devinit_post(struct nvkm_devinit *base, bool post)
 {
 	struct nv50_devinit *init = nv50_devinit(base);
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gv100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gv100.c
new file mode 100644
index 000000000000..fbde6828bd38
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/gv100.c
@@ -0,0 +1,79 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "nv50.h"
+
+#include <subdev/bios.h>
+#include <subdev/bios/pll.h>
+#include <subdev/clk/pll.h>
+
+static int
+gv100_devinit_pll_set(struct nvkm_devinit *init, u32 type, u32 freq)
+{
+	struct nvkm_subdev *subdev = &init->subdev;
+	struct nvkm_device *device = subdev->device;
+	struct nvbios_pll info;
+	int head = type - PLL_VPLL0;
+	int N, fN, M, P;
+	int ret;
+
+	ret = nvbios_pll_parse(device->bios, type, &info);
+	if (ret)
+		return ret;
+
+	ret = gt215_pll_calc(subdev, &info, freq, &N, &fN, &M, &P);
+	if (ret < 0)
+		return ret;
+
+	switch (info.type) {
+	case PLL_VPLL0:
+	case PLL_VPLL1:
+	case PLL_VPLL2:
+	case PLL_VPLL3:
+		nvkm_wr32(device, 0x00ef10 + (head * 0x40), fN << 16);
+		nvkm_wr32(device, 0x00ef04 + (head * 0x40), (P << 16) |
+							    (N <<  8) |
+							    (M <<  0));
+		break;
+	default:
+		nvkm_warn(subdev, "%08x/%dKhz unimplemented\n", type, freq);
+		ret = -EINVAL;
+		break;
+	}
+
+	return ret;
+}
+
+static const struct nvkm_devinit_func
+gv100_devinit = {
+	.preinit = gf100_devinit_preinit,
+	.init = nv50_devinit_init,
+	.post = gm200_devinit_post,
+	.pll_set = gv100_devinit_pll_set,
+	.disable = gm107_devinit_disable,
+};
+
+int
+gv100_devinit_new(struct nvkm_device *device, int index,
+		struct nvkm_devinit **pinit)
+{
+	return nv50_devinit_new_(&gv100_devinit, device, index, pinit);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/nv50.h b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/nv50.h
index 315ebaff1165..9b9f0dc1e192 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/nv50.h
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/devinit/nv50.h
@@ -24,4 +24,6 @@ int  gf100_devinit_pll_set(struct nvkm_devinit *, u32, u32);
 void gf100_devinit_preinit(struct nvkm_devinit *);
 
 u64  gm107_devinit_disable(struct nvkm_devinit *);
+
+int gm200_devinit_post(struct nvkm_devinit *, bool);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild
new file mode 100644
index 000000000000..45bb46fb0929
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/Kbuild
@@ -0,0 +1,3 @@
+nvkm-y += nvkm/subdev/fault/base.o
+nvkm-y += nvkm/subdev/fault/gp100.o
+nvkm-y += nvkm/subdev/fault/gv100.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/base.c
new file mode 100644
index 000000000000..16ad91c91a7b
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/base.c
@@ -0,0 +1,185 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "priv.h"
+
+#include <core/memory.h>
+#include <core/notify.h>
+#include <subdev/bar.h>
+#include <subdev/mmu.h>
+
+static void
+nvkm_fault_ntfy_fini(struct nvkm_event *event, int type, int index)
+{
+	struct nvkm_fault *fault = container_of(event, typeof(*fault), event);
+	fault->func->buffer.fini(fault->buffer[index]);
+}
+
+static void
+nvkm_fault_ntfy_init(struct nvkm_event *event, int type, int index)
+{
+	struct nvkm_fault *fault = container_of(event, typeof(*fault), event);
+	fault->func->buffer.init(fault->buffer[index]);
+}
+
+static int
+nvkm_fault_ntfy_ctor(struct nvkm_object *object, void *argv, u32 argc,
+		     struct nvkm_notify *notify)
+{
+	struct nvkm_fault_buffer *buffer = nvkm_fault_buffer(object);
+	if (argc == 0) {
+		notify->size  = 0;
+		notify->types = 1;
+		notify->index = buffer->id;
+		return 0;
+	}
+	return -ENOSYS;
+}
+
+static const struct nvkm_event_func
+nvkm_fault_ntfy = {
+	.ctor = nvkm_fault_ntfy_ctor,
+	.init = nvkm_fault_ntfy_init,
+	.fini = nvkm_fault_ntfy_fini,
+};
+
+static void
+nvkm_fault_intr(struct nvkm_subdev *subdev)
+{
+	struct nvkm_fault *fault = nvkm_fault(subdev);
+	return fault->func->intr(fault);
+}
+
+static int
+nvkm_fault_fini(struct nvkm_subdev *subdev, bool suspend)
+{
+	struct nvkm_fault *fault = nvkm_fault(subdev);
+	if (fault->func->fini)
+		fault->func->fini(fault);
+	return 0;
+}
+
+static int
+nvkm_fault_init(struct nvkm_subdev *subdev)
+{
+	struct nvkm_fault *fault = nvkm_fault(subdev);
+	if (fault->func->init)
+		fault->func->init(fault);
+	return 0;
+}
+
+static int
+nvkm_fault_oneinit_buffer(struct nvkm_fault *fault, int id)
+{
+	struct nvkm_subdev *subdev = &fault->subdev;
+	struct nvkm_device *device = subdev->device;
+	struct nvkm_vmm *bar2 = nvkm_bar_bar2_vmm(device);
+	struct nvkm_fault_buffer *buffer;
+	int ret;
+
+	if (!(buffer = kzalloc(sizeof(*buffer), GFP_KERNEL)))
+		return -ENOMEM;
+	buffer->fault = fault;
+	buffer->id = id;
+	buffer->entries = fault->func->buffer.entries(buffer);
+	fault->buffer[id] = buffer;
+
+	nvkm_debug(subdev, "buffer %d: %d entries\n", id, buffer->entries);
+
+	ret = nvkm_memory_new(device, NVKM_MEM_TARGET_INST, buffer->entries *
+			      fault->func->buffer.entry_size, 0x1000, true,
+			      &buffer->mem);
+	if (ret)
+		return ret;
+
+	ret = nvkm_vmm_get(bar2, 12, nvkm_memory_size(buffer->mem),
+			   &buffer->vma);
+	if (ret)
+		return ret;
+
+	return nvkm_memory_map(buffer->mem, 0, bar2, buffer->vma, NULL, 0);
+}
+
+static int
+nvkm_fault_oneinit(struct nvkm_subdev *subdev)
+{
+	struct nvkm_fault *fault = nvkm_fault(subdev);
+	int ret, i;
+
+	for (i = 0; i < ARRAY_SIZE(fault->buffer); i++) {
+		if (i < fault->func->buffer.nr) {
+			ret = nvkm_fault_oneinit_buffer(fault, i);
+			if (ret)
+				return ret;
+			fault->buffer_nr = i + 1;
+		}
+	}
+
+	ret = nvkm_event_init(&nvkm_fault_ntfy, 1, fault->buffer_nr,
+			      &fault->event);
+	if (ret)
+		return ret;
+
+	if (fault->func->oneinit)
+		ret = fault->func->oneinit(fault);
+	return ret;
+}
+
+static void *
+nvkm_fault_dtor(struct nvkm_subdev *subdev)
+{
+	struct nvkm_vmm *bar2 = nvkm_bar_bar2_vmm(subdev->device);
+	struct nvkm_fault *fault = nvkm_fault(subdev);
+	int i;
+
+	nvkm_event_fini(&fault->event);
+
+	for (i = 0; i < fault->buffer_nr; i++) {
+		if (fault->buffer[i]) {
+			nvkm_vmm_put(bar2, &fault->buffer[i]->vma);
+			nvkm_memory_unref(&fault->buffer[i]->mem);
+			kfree(fault->buffer[i]);
+		}
+	}
+
+	return fault;
+}
+
+static const struct nvkm_subdev_func
+nvkm_fault = {
+	.dtor = nvkm_fault_dtor,
+	.oneinit = nvkm_fault_oneinit,
+	.init = nvkm_fault_init,
+	.fini = nvkm_fault_fini,
+	.intr = nvkm_fault_intr,
+};
+
+int
+nvkm_fault_new_(const struct nvkm_fault_func *func, struct nvkm_device *device,
+		int index, struct nvkm_fault **pfault)
+{
+	struct nvkm_fault *fault;
+	if (!(fault = *pfault = kzalloc(sizeof(*fault), GFP_KERNEL)))
+		return -ENOMEM;
+	nvkm_subdev_ctor(&nvkm_fault, device, index, &fault->subdev);
+	fault->func = func;
+	return 0;
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/gp100.c
new file mode 100644
index 000000000000..5e71db2e8d75
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/gp100.c
@@ -0,0 +1,69 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "priv.h"
+
+#include <subdev/mmu.h>
+
+static void
+gp100_fault_buffer_fini(struct nvkm_fault_buffer *buffer)
+{
+	struct nvkm_device *device = buffer->fault->subdev.device;
+	nvkm_mask(device, 0x002a70, 0x00000001, 0x00000000);
+}
+
+static void
+gp100_fault_buffer_init(struct nvkm_fault_buffer *buffer)
+{
+	struct nvkm_device *device = buffer->fault->subdev.device;
+	nvkm_wr32(device, 0x002a74, upper_32_bits(buffer->vma->addr));
+	nvkm_wr32(device, 0x002a70, lower_32_bits(buffer->vma->addr));
+	nvkm_mask(device, 0x002a70, 0x00000001, 0x00000001);
+}
+
+static u32
+gp100_fault_buffer_entries(struct nvkm_fault_buffer *buffer)
+{
+	return nvkm_rd32(buffer->fault->subdev.device, 0x002a78);
+}
+
+static void
+gp100_fault_intr(struct nvkm_fault *fault)
+{
+	nvkm_event_send(&fault->event, 1, 0, NULL, 0);
+}
+
+static const struct nvkm_fault_func
+gp100_fault = {
+	.intr = gp100_fault_intr,
+	.buffer.nr = 1,
+	.buffer.entry_size = 32,
+	.buffer.entries = gp100_fault_buffer_entries,
+	.buffer.init = gp100_fault_buffer_init,
+	.buffer.fini = gp100_fault_buffer_fini,
+};
+
+int
+gp100_fault_new(struct nvkm_device *device, int index,
+		struct nvkm_fault **pfault)
+{
+	return nvkm_fault_new_(&gp100_fault, device, index, pfault);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/gv100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/gv100.c
new file mode 100644
index 000000000000..3cd610d7deb5
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/gv100.c
@@ -0,0 +1,205 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "priv.h"
+
+#include <core/memory.h>
+#include <subdev/mmu.h>
+#include <engine/fifo.h>
+
+static void
+gv100_fault_buffer_process(struct nvkm_fault_buffer *buffer)
+{
+	struct nvkm_device *device = buffer->fault->subdev.device;
+	struct nvkm_memory *mem = buffer->mem;
+	const u32 foff = buffer->id * 0x14;
+	u32 get = nvkm_rd32(device, 0x100e2c + foff);
+	u32 put = nvkm_rd32(device, 0x100e30 + foff);
+	if (put == get)
+		return;
+
+	nvkm_kmap(mem);
+	while (get != put) {
+		const u32   base = get * buffer->fault->func->buffer.entry_size;
+		const u32 instlo = nvkm_ro32(mem, base + 0x00);
+		const u32 insthi = nvkm_ro32(mem, base + 0x04);
+		const u32 addrlo = nvkm_ro32(mem, base + 0x08);
+		const u32 addrhi = nvkm_ro32(mem, base + 0x0c);
+		const u32 timelo = nvkm_ro32(mem, base + 0x10);
+		const u32 timehi = nvkm_ro32(mem, base + 0x14);
+		const u32  info0 = nvkm_ro32(mem, base + 0x18);
+		const u32  info1 = nvkm_ro32(mem, base + 0x1c);
+		struct nvkm_fault_data info;
+
+		if (++get == buffer->entries)
+			get = 0;
+		nvkm_wr32(device, 0x100e2c + foff, get);
+
+		info.addr   = ((u64)addrhi << 32) | addrlo;
+		info.inst   = ((u64)insthi << 32) | instlo;
+		info.time   = ((u64)timehi << 32) | timelo;
+		info.engine = (info0 & 0x000000ff);
+		info.valid  = (info1 & 0x80000000) >> 31;
+		info.gpc    = (info1 & 0x1f000000) >> 24;
+		info.hub    = (info1 & 0x00100000) >> 20;
+		info.access = (info1 & 0x000f0000) >> 16;
+		info.client = (info1 & 0x00007f00) >> 8;
+		info.reason = (info1 & 0x0000001f);
+
+		nvkm_fifo_fault(device->fifo, &info);
+	}
+	nvkm_done(mem);
+}
+
+static void
+gv100_fault_buffer_fini(struct nvkm_fault_buffer *buffer)
+{
+	struct nvkm_device *device = buffer->fault->subdev.device;
+	const u32 intr = buffer->id ? 0x08000000 : 0x20000000;
+	const u32 foff = buffer->id * 0x14;
+
+	nvkm_mask(device, 0x100a34, intr, intr);
+	nvkm_mask(device, 0x100e34 + foff, 0x80000000, 0x00000000);
+}
+
+static void
+gv100_fault_buffer_init(struct nvkm_fault_buffer *buffer)
+{
+	struct nvkm_device *device = buffer->fault->subdev.device;
+	const u32 intr = buffer->id ? 0x08000000 : 0x20000000;
+	const u32 foff = buffer->id * 0x14;
+
+	nvkm_mask(device, 0x100e34 + foff, 0xc0000000, 0x40000000);
+	nvkm_wr32(device, 0x100e28 + foff, upper_32_bits(buffer->vma->addr));
+	nvkm_wr32(device, 0x100e24 + foff, lower_32_bits(buffer->vma->addr));
+	nvkm_mask(device, 0x100e34 + foff, 0x80000000, 0x80000000);
+	nvkm_mask(device, 0x100a2c, intr, intr);
+}
+
+static u32
+gv100_fault_buffer_entries(struct nvkm_fault_buffer *buffer)
+{
+	struct nvkm_device *device = buffer->fault->subdev.device;
+	const u32 foff = buffer->id * 0x14;
+	nvkm_mask(device, 0x100e34 + foff, 0x40000000, 0x40000000);
+	return nvkm_rd32(device, 0x100e34 + foff) & 0x000fffff;
+}
+
+static int
+gv100_fault_ntfy_nrpfb(struct nvkm_notify *notify)
+{
+	struct nvkm_fault *fault = container_of(notify, typeof(*fault), nrpfb);
+	gv100_fault_buffer_process(fault->buffer[0]);
+	return NVKM_NOTIFY_KEEP;
+}
+
+static void
+gv100_fault_intr_fault(struct nvkm_fault *fault)
+{
+	struct nvkm_subdev *subdev = &fault->subdev;
+	struct nvkm_device *device = subdev->device;
+	struct nvkm_fault_data info;
+	const u32 addrlo = nvkm_rd32(device, 0x100e4c);
+	const u32 addrhi = nvkm_rd32(device, 0x100e50);
+	const u32  info0 = nvkm_rd32(device, 0x100e54);
+	const u32 insthi = nvkm_rd32(device, 0x100e58);
+	const u32  info1 = nvkm_rd32(device, 0x100e5c);
+
+	info.addr = ((u64)addrhi << 32) | addrlo;
+	info.inst = ((u64)insthi << 32) | (info0 & 0xfffff000);
+	info.time = 0;
+	info.engine = (info0 & 0x000000ff);
+	info.valid  = (info1 & 0x80000000) >> 31;
+	info.gpc    = (info1 & 0x1f000000) >> 24;
+	info.hub    = (info1 & 0x00100000) >> 20;
+	info.access = (info1 & 0x000f0000) >> 16;
+	info.client = (info1 & 0x00007f00) >> 8;
+	info.reason = (info1 & 0x0000001f);
+
+	nvkm_fifo_fault(device->fifo, &info);
+}
+
+static void
+gv100_fault_intr(struct nvkm_fault *fault)
+{
+	struct nvkm_subdev *subdev = &fault->subdev;
+	struct nvkm_device *device = subdev->device;
+	u32 stat = nvkm_rd32(device, 0x100a20);
+
+	if (stat & 0x80000000) {
+		gv100_fault_intr_fault(fault);
+		nvkm_wr32(device, 0x100e60, 0x80000000);
+		stat &= ~0x80000000;
+	}
+
+	if (stat & 0x20000000) {
+		if (fault->buffer[0]) {
+			nvkm_event_send(&fault->event, 1, 0, NULL, 0);
+			stat &= ~0x20000000;
+		}
+	}
+
+	if (stat) {
+		nvkm_debug(subdev, "intr %08x\n", stat);
+	}
+}
+
+static void
+gv100_fault_fini(struct nvkm_fault *fault)
+{
+	nvkm_notify_put(&fault->nrpfb);
+	nvkm_mask(fault->subdev.device, 0x100a34, 0x80000000, 0x80000000);
+}
+
+static void
+gv100_fault_init(struct nvkm_fault *fault)
+{
+	nvkm_mask(fault->subdev.device, 0x100a2c, 0x80000000, 0x80000000);
+	nvkm_notify_get(&fault->nrpfb);
+}
+
+static int
+gv100_fault_oneinit(struct nvkm_fault *fault)
+{
+	return nvkm_notify_init(&fault->buffer[0]->object, &fault->event,
+				gv100_fault_ntfy_nrpfb, false, NULL, 0, 0,
+				&fault->nrpfb);
+}
+
+static const struct nvkm_fault_func
+gv100_fault = {
+	.oneinit = gv100_fault_oneinit,
+	.init = gv100_fault_init,
+	.fini = gv100_fault_fini,
+	.intr = gv100_fault_intr,
+	.buffer.nr = 2,
+	.buffer.entry_size = 32,
+	.buffer.entries = gv100_fault_buffer_entries,
+	.buffer.init = gv100_fault_buffer_init,
+	.buffer.fini = gv100_fault_buffer_fini,
+};
+
+int
+gv100_fault_new(struct nvkm_device *device, int index,
+		struct nvkm_fault **pfault)
+{
+	return nvkm_fault_new_(&gv100_fault, device, index, pfault);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fault/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/priv.h
new file mode 100644
index 000000000000..e4d2f5234fd1
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fault/priv.h
@@ -0,0 +1,35 @@
+#ifndef __NVKM_FAULT_PRIV_H__
+#define __NVKM_FAULT_PRIV_H__
+#define nvkm_fault_buffer(p) container_of((p), struct nvkm_fault_buffer, object)
+#define nvkm_fault(p) container_of((p), struct nvkm_fault, subdev)
+#include <subdev/fault.h>
+
+#include <core/event.h>
+#include <core/object.h>
+
+struct nvkm_fault_buffer {
+	struct nvkm_object object;
+	struct nvkm_fault *fault;
+	int id;
+	int entries;
+	struct nvkm_memory *mem;
+	struct nvkm_vma *vma;
+};
+
+int nvkm_fault_new_(const struct nvkm_fault_func *, struct nvkm_device *,
+		    int index, struct nvkm_fault **);
+
+struct nvkm_fault_func {
+	int (*oneinit)(struct nvkm_fault *);
+	void (*init)(struct nvkm_fault *);
+	void (*fini)(struct nvkm_fault *);
+	void (*intr)(struct nvkm_fault *);
+	struct {
+		int nr;
+		u32 entry_size;
+		u32 (*entries)(struct nvkm_fault_buffer *);
+		void (*init)(struct nvkm_fault_buffer *);
+		void (*fini)(struct nvkm_fault_buffer *);
+	} buffer;
+};
+#endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild
index b4f22cce5d43..969610951263 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/Kbuild
@@ -30,6 +30,7 @@ nvkm-y += nvkm/subdev/fb/gm20b.o
 nvkm-y += nvkm/subdev/fb/gp100.o
 nvkm-y += nvkm/subdev/fb/gp102.o
 nvkm-y += nvkm/subdev/fb/gp10b.o
+nvkm-y += nvkm/subdev/fb/gv100.o
 
 nvkm-y += nvkm/subdev/fb/ram.o
 nvkm-y += nvkm/subdev/fb/ramnv04.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.c
index cdc4e0a2cc6b..e8dc4e913494 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gf100.c
@@ -46,10 +46,10 @@ gf100_fb_oneinit(struct nvkm_fb *base)
 {
 	struct gf100_fb *fb = gf100_fb(base);
 	struct nvkm_device *device = fb->base.subdev.device;
-	int ret, size = 0x1000;
+	int ret, size = 1 << (fb->base.page ? fb->base.page : 17);
 
 	size = nvkm_longopt(device->cfgopt, "MmuDebugBufferSize", size);
-	size = min(size, 0x1000);
+	size = max(size, 0x1000);
 
 	ret = nvkm_memory_new(device, NVKM_MEM_TARGET_INST, size, 0x1000,
 			      true, &fb->base.mmu_rd);
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm200.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm200.c
index 8137e19d3292..d3b8c3367152 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm200.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gm200.c
@@ -49,8 +49,6 @@ gm200_fb_init(struct nvkm_fb *base)
 	if (fb->r100c10_page)
 		nvkm_wr32(device, 0x100c10, fb->r100c10 >> 8);
 
-	nvkm_mask(device, 0x100c80, 0x00000001, 0x00000000); /* 128KiB lpg */
-
 	nvkm_wr32(device, 0x100cc8, nvkm_memory_addr(fb->base.mmu_wr) >> 8);
 	nvkm_wr32(device, 0x100ccc, nvkm_memory_addr(fb->base.mmu_rd) >> 8);
 	nvkm_mask(device, 0x100cc4, 0x00060000,
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c
index 147f69b30cd8..dffe1f5e1071 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gp100.c
@@ -26,7 +26,7 @@
 
 #include <core/memory.h>
 
-static void
+void
 gp100_fb_init_unkn(struct nvkm_fb *base)
 {
 	struct nvkm_device *device = gf100_fb(base)->base.subdev.device;
@@ -48,7 +48,7 @@ gp100_fb_init(struct nvkm_fb *base)
 	nvkm_wr32(device, 0x100cc8, nvkm_memory_addr(fb->base.mmu_wr) >> 8);
 	nvkm_wr32(device, 0x100ccc, nvkm_memory_addr(fb->base.mmu_rd) >> 8);
 	nvkm_mask(device, 0x100cc4, 0x00060000,
-		  max(nvkm_memory_size(fb->base.mmu_rd) >> 16, (u64)2) << 17);
+		  min(nvkm_memory_size(fb->base.mmu_rd) >> 16, (u64)2) << 17);
 }
 
 static const struct nvkm_fb_func
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gv100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gv100.c
new file mode 100644
index 000000000000..3c5e02e9794a
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/gv100.c
@@ -0,0 +1,46 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "gf100.h"
+#include "ram.h"
+
+static int
+gv100_fb_init_page(struct nvkm_fb *fb)
+{
+	return (fb->page == 16) ? 0 : -EINVAL;
+}
+
+static const struct nvkm_fb_func
+gv100_fb = {
+	.dtor = gf100_fb_dtor,
+	.oneinit = gf100_fb_oneinit,
+	.init = gp100_fb_init,
+	.init_page = gv100_fb_init_page,
+	.init_unkn = gp100_fb_init_unkn,
+	.ram_new = gp100_ram_new,
+	.default_bigpage = 16,
+};
+
+int
+gv100_fb_new(struct nvkm_device *device, int index, struct nvkm_fb **pfb)
+{
+	return gf100_fb_new_(&gv100_fb, device, index, pfb);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h
index 414a423e0e55..2857f31466bf 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/priv.h
@@ -68,4 +68,6 @@ int gf100_fb_oneinit(struct nvkm_fb *);
 int gf100_fb_init_page(struct nvkm_fb *);
 
 int gm200_fb_init_page(struct nvkm_fb *);
+
+void gp100_fb_init_unkn(struct nvkm_fb *);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgt215.c b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgt215.c
index 920b3d347803..bbfde1cb3a17 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgt215.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/fb/ramgt215.c
@@ -171,7 +171,7 @@ gt215_link_train(struct gt215_ram *ram)
 		return -ENOSYS;
 
 	/* XXX: Multiple partitions? */
-	result = kmalloc(64 * sizeof(u32), GFP_KERNEL);
+	result = kmalloc_array(64, sizeof(u32), GFP_KERNEL);
 	if (!result)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/Kbuild
index 12d6f4f102cb..290ff1c425a9 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/Kbuild
@@ -4,3 +4,4 @@ nvkm-y += nvkm/subdev/ltc/gk104.o
 nvkm-y += nvkm/subdev/ltc/gm107.o
 nvkm-y += nvkm/subdev/ltc/gm200.o
 nvkm-y += nvkm/subdev/ltc/gp100.o
+nvkm-y += nvkm/subdev/ltc/gp102.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/base.c
index 1f185274d3e6..23242179e600 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/base.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/base.c
@@ -55,6 +55,14 @@ nvkm_ltc_zbc_depth_get(struct nvkm_ltc *ltc, int index, const u32 depth)
 	return index;
 }
 
+int
+nvkm_ltc_zbc_stencil_get(struct nvkm_ltc *ltc, int index, const u32 stencil)
+{
+	ltc->zbc_stencil[index] = stencil;
+	ltc->func->zbc_clear_stencil(ltc, index, stencil);
+	return index;
+}
+
 void
 nvkm_ltc_invalidate(struct nvkm_ltc *ltc)
 {
@@ -92,6 +100,8 @@ nvkm_ltc_init(struct nvkm_subdev *subdev)
 	for (i = ltc->zbc_min; i <= ltc->zbc_max; i++) {
 		ltc->func->zbc_clear_color(ltc, i, ltc->zbc_color[i]);
 		ltc->func->zbc_clear_depth(ltc, i, ltc->zbc_depth[i]);
+		if (ltc->func->zbc_clear_stencil)
+			ltc->func->zbc_clear_stencil(ltc, i, ltc->zbc_stencil[i]);
 	}
 
 	ltc->func->init(ltc);
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp100.c
index e34d42108019..e923ed76d37a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp100.c
@@ -23,7 +23,7 @@
  */
 #include "priv.h"
 
-static void
+void
 gp100_ltc_intr(struct nvkm_ltc *ltc)
 {
 	struct nvkm_device *device = ltc->subdev.device;
@@ -38,7 +38,7 @@ gp100_ltc_intr(struct nvkm_ltc *ltc)
 	}
 }
 
-static int
+int
 gp100_ltc_oneinit(struct nvkm_ltc *ltc)
 {
 	struct nvkm_device *device = ltc->subdev.device;
@@ -48,7 +48,7 @@ gp100_ltc_oneinit(struct nvkm_ltc *ltc)
 	return 0;
 }
 
-static void
+void
 gp100_ltc_init(struct nvkm_ltc *ltc)
 {
 	/*XXX: PMU LS call to setup tagram address */
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp102.c b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp102.c
new file mode 100644
index 000000000000..601747ada655
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/gp102.c
@@ -0,0 +1,51 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "priv.h"
+
+void
+gp102_ltc_zbc_clear_stencil(struct nvkm_ltc *ltc, int i, const u32 stencil)
+{
+	struct nvkm_device *device = ltc->subdev.device;
+	nvkm_mask(device, 0x17e338, 0x0000000f, i);
+	nvkm_wr32(device, 0x17e204, stencil);
+}
+
+static const struct nvkm_ltc_func
+gp102_ltc = {
+	.oneinit = gp100_ltc_oneinit,
+	.init = gp100_ltc_init,
+	.intr = gp100_ltc_intr,
+	.cbc_clear = gm107_ltc_cbc_clear,
+	.cbc_wait = gm107_ltc_cbc_wait,
+	.zbc = 16,
+	.zbc_clear_color = gm107_ltc_zbc_clear_color,
+	.zbc_clear_depth = gm107_ltc_zbc_clear_depth,
+	.zbc_clear_stencil = gp102_ltc_zbc_clear_stencil,
+	.invalidate = gf100_ltc_invalidate,
+	.flush = gf100_ltc_flush,
+};
+
+int
+gp102_ltc_new(struct nvkm_device *device, int index, struct nvkm_ltc **pltc)
+{
+	return nvkm_ltc_new_(&gp102_ltc, device, index, pltc);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/priv.h
index e71cc25cc775..9dcde43c0f3c 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/ltc/priv.h
@@ -19,6 +19,7 @@ struct nvkm_ltc_func {
 	int zbc;
 	void (*zbc_clear_color)(struct nvkm_ltc *, int, const u32[4]);
 	void (*zbc_clear_depth)(struct nvkm_ltc *, int, const u32);
+	void (*zbc_clear_stencil)(struct nvkm_ltc *, int, const u32);
 
 	void (*invalidate)(struct nvkm_ltc *);
 	void (*flush)(struct nvkm_ltc *);
@@ -41,4 +42,8 @@ void gm107_ltc_cbc_clear(struct nvkm_ltc *, u32, u32);
 void gm107_ltc_cbc_wait(struct nvkm_ltc *);
 void gm107_ltc_zbc_clear_color(struct nvkm_ltc *, int, const u32[4]);
 void gm107_ltc_zbc_clear_depth(struct nvkm_ltc *, int, const u32);
+
+int gp100_ltc_oneinit(struct nvkm_ltc *);
+void gp100_ltc_init(struct nvkm_ltc *);
+void gp100_ltc_intr(struct nvkm_ltc *);
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c
index 7321ad3758c3..43db245eec9a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp100.c
@@ -75,10 +75,28 @@ gp100_mc_intr_mask(struct nvkm_mc *base, u32 mask, u32 intr)
 	spin_unlock_irqrestore(&mc->lock, flags);
 }
 
+const struct nvkm_mc_map
+gp100_mc_intr[] = {
+	{ 0x04000000, NVKM_ENGINE_DISP },
+	{ 0x00000100, NVKM_ENGINE_FIFO },
+	{ 0x00000200, NVKM_SUBDEV_FAULT },
+	{ 0x40000000, NVKM_SUBDEV_IBUS },
+	{ 0x10000000, NVKM_SUBDEV_BUS },
+	{ 0x08000000, NVKM_SUBDEV_FB },
+	{ 0x02000000, NVKM_SUBDEV_LTC },
+	{ 0x01000000, NVKM_SUBDEV_PMU },
+	{ 0x00200000, NVKM_SUBDEV_GPIO },
+	{ 0x00200000, NVKM_SUBDEV_I2C },
+	{ 0x00100000, NVKM_SUBDEV_TIMER },
+	{ 0x00040000, NVKM_SUBDEV_THERM },
+	{ 0x00002000, NVKM_SUBDEV_FB },
+	{},
+};
+
 static const struct nvkm_mc_func
 gp100_mc = {
 	.init = nv50_mc_init,
-	.intr = gk104_mc_intr,
+	.intr = gp100_mc_intr,
 	.intr_unarm = gp100_mc_intr_unarm,
 	.intr_rearm = gp100_mc_intr_rearm,
 	.intr_mask = gp100_mc_intr_mask,
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp10b.c
index 2283e3b74277..ff8629de97d6 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/gp10b.c
@@ -34,7 +34,7 @@ gp10b_mc_init(struct nvkm_mc *mc)
 static const struct nvkm_mc_func
 gp10b_mc = {
 	.init = gp10b_mc_init,
-	.intr = gk104_mc_intr,
+	.intr = gp100_mc_intr,
 	.intr_unarm = gp100_mc_intr_unarm,
 	.intr_rearm = gp100_mc_intr_rearm,
 	.intr_mask = gp100_mc_intr_mask,
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h
index 8869d79c2b59..d9e3691d45b7 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mc/priv.h
@@ -57,4 +57,6 @@ int gp100_mc_new_(const struct nvkm_mc_func *, struct nvkm_device *, int,
 
 extern const struct nvkm_mc_map gk104_mc_intr[];
 extern const struct nvkm_mc_map gk104_mc_reset[];
+
+extern const struct nvkm_mc_map gp100_mc_intr[];
 #endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/Kbuild b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/Kbuild
index 67ee983bb026..58a24e3a0598 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/Kbuild
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/Kbuild
@@ -12,6 +12,7 @@ nvkm-y += nvkm/subdev/mmu/gm200.o
 nvkm-y += nvkm/subdev/mmu/gm20b.o
 nvkm-y += nvkm/subdev/mmu/gp100.o
 nvkm-y += nvkm/subdev/mmu/gp10b.o
+nvkm-y += nvkm/subdev/mmu/gv100.o
 
 nvkm-y += nvkm/subdev/mmu/mem.o
 nvkm-y += nvkm/subdev/mmu/memnv04.o
@@ -31,6 +32,7 @@ nvkm-y += nvkm/subdev/mmu/vmmgm200.o
 nvkm-y += nvkm/subdev/mmu/vmmgm20b.o
 nvkm-y += nvkm/subdev/mmu/vmmgp100.o
 nvkm-y += nvkm/subdev/mmu/vmmgp10b.o
+nvkm-y += nvkm/subdev/mmu/vmmgv100.o
 
 nvkm-y += nvkm/subdev/mmu/umem.o
 nvkm-y += nvkm/subdev/mmu/ummu.o
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/gp10b. b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/gp10b.
deleted file mode 100644
index e69de29bb2d1..000000000000
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/gp10b.
+++ /dev/null
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/gv100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/gv100.c
new file mode 100644
index 000000000000..f666cb57f69e
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/gv100.c
@@ -0,0 +1,43 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "mem.h"
+#include "vmm.h"
+
+#include <core/option.h>
+
+#include <nvif/class.h>
+
+static const struct nvkm_mmu_func
+gv100_mmu = {
+	.dma_bits = 47,
+	.mmu = {{ -1, -1, NVIF_CLASS_MMU_GF100}},
+	.mem = {{ -1,  0, NVIF_CLASS_MEM_GF100}, gf100_mem_new, gf100_mem_map },
+	.vmm = {{ -1, -1, NVIF_CLASS_VMM_GP100}, gv100_vmm_new },
+	.kind = gm200_mmu_kind,
+	.kind_sys = true,
+};
+
+int
+gv100_mmu_new(struct nvkm_device *device, int index, struct nvkm_mmu **pmmu)
+{
+	return nvkm_mmu_new_(&gv100_mmu, device, index, pmmu);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/mem.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/mem.c
index 39808489f21d..92e363dbbc5a 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/mem.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/mem.c
@@ -191,9 +191,9 @@ nvkm_mem_new_host(struct nvkm_mmu *mmu, int type, u8 page, u64 size,
 	nvkm_memory_ctor(&nvkm_mem_dma, &mem->memory);
 	size = ALIGN(size, PAGE_SIZE) >> PAGE_SHIFT;
 
-	if (!(mem->mem = kvmalloc(sizeof(*mem->mem) * size, GFP_KERNEL)))
+	if (!(mem->mem = kvmalloc_array(size, sizeof(*mem->mem), GFP_KERNEL)))
 		return -ENOMEM;
-	if (!(mem->dma = kvmalloc(sizeof(*mem->dma) * size, GFP_KERNEL)))
+	if (!(mem->dma = kvmalloc_array(size, sizeof(*mem->dma), GFP_KERNEL)))
 		return -ENOMEM;
 
 	if (mmu->dma_bits > 32)
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.c
index 1c12e58f44c2..de269eb482dd 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.c
@@ -59,7 +59,7 @@ nvkm_vmm_pt_new(const struct nvkm_vmm_desc *desc, bool sparse,
 	pgt->sparse = sparse;
 
 	if (desc->type == PGD) {
-		pgt->pde = kvzalloc(sizeof(*pgt->pde) * pten, GFP_KERNEL);
+		pgt->pde = kvcalloc(pten, sizeof(*pgt->pde), GFP_KERNEL);
 		if (!pgt->pde) {
 			kfree(pgt);
 			return NULL;
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.h b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.h
index da06e64d8a7d..1a3b0a3724ca 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.h
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmm.h
@@ -236,6 +236,9 @@ int gp100_vmm_new(struct nvkm_mmu *, u64, u64, void *, u32,
 int gp10b_vmm_new(struct nvkm_mmu *, u64, u64, void *, u32,
 		  struct lock_class_key *, const char *,
 		  struct nvkm_vmm **);
+int gv100_vmm_new(struct nvkm_mmu *, u64, u64, void *, u32,
+		  struct lock_class_key *, const char *,
+		  struct nvkm_vmm **);
 
 #define VMM_PRINT(l,v,p,f,a...) do {                                           \
 	struct nvkm_vmm *_vmm = (v);                                           \
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmmgv100.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmmgv100.c
new file mode 100644
index 000000000000..2fa40c16e6d2
--- /dev/null
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mmu/vmmgv100.c
@@ -0,0 +1,87 @@
+/*
+ * Copyright 2018 Red Hat Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ */
+#include "vmm.h"
+
+#include <subdev/fb.h>
+#include <subdev/ltc.h>
+
+#include <nvif/ifc00d.h>
+#include <nvif/unpack.h>
+
+int
+gv100_vmm_join(struct nvkm_vmm *vmm, struct nvkm_memory *inst)
+{
+	u64 data[2], mask;
+	int ret = gp100_vmm_join(vmm, inst), i;
+	if (ret)
+		return ret;
+
+	nvkm_kmap(inst);
+	data[0] = nvkm_ro32(inst, 0x200);
+	data[1] = nvkm_ro32(inst, 0x204);
+	mask = BIT_ULL(0);
+
+	nvkm_wo32(inst, 0x21c, 0x00000000);
+
+	for (i = 0; i < 64; i++) {
+		if (mask & BIT_ULL(i)) {
+			nvkm_wo32(inst, 0x2a4 + (i * 0x10), data[1]);
+			nvkm_wo32(inst, 0x2a0 + (i * 0x10), data[0]);
+		} else {
+			nvkm_wo32(inst, 0x2a4 + (i * 0x10), 0x00000001);
+			nvkm_wo32(inst, 0x2a0 + (i * 0x10), 0x00000001);
+		}
+		nvkm_wo32(inst, 0x2a8 + (i * 0x10), 0x00000000);
+	}
+
+	nvkm_wo32(inst, 0x298, lower_32_bits(mask));
+	nvkm_wo32(inst, 0x29c, upper_32_bits(mask));
+	nvkm_done(inst);
+	return 0;
+}
+
+static const struct nvkm_vmm_func
+gv100_vmm = {
+	.join = gv100_vmm_join,
+	.part = gf100_vmm_part,
+	.aper = gf100_vmm_aper,
+	.valid = gp100_vmm_valid,
+	.flush = gp100_vmm_flush,
+	.page = {
+		{ 47, &gp100_vmm_desc_16[4], NVKM_VMM_PAGE_Sxxx },
+		{ 38, &gp100_vmm_desc_16[3], NVKM_VMM_PAGE_Sxxx },
+		{ 29, &gp100_vmm_desc_16[2], NVKM_VMM_PAGE_Sxxx },
+		{ 21, &gp100_vmm_desc_16[1], NVKM_VMM_PAGE_SVxC },
+		{ 16, &gp100_vmm_desc_16[0], NVKM_VMM_PAGE_SVxC },
+		{ 12, &gp100_vmm_desc_12[0], NVKM_VMM_PAGE_SVHx },
+		{}
+	}
+};
+
+int
+gv100_vmm_new(struct nvkm_mmu *mmu, u64 addr, u64 size, void *argv, u32 argc,
+	      struct lock_class_key *key, const char *name,
+	      struct nvkm_vmm **pvmm)
+{
+	return nv04_vmm_new_(&gv100_vmm, mmu, 0, addr, size,
+			     argv, argc, key, name, pvmm);
+}
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.c
index a721354249ce..d02e183717dc 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r352.c
@@ -414,6 +414,20 @@ acr_r352_ls_write_wpr(struct acr_r352 *acr, struct list_head *imgs,
 {
 	struct ls_ucode_img *_img;
 	u32 pos = 0;
+	u32 max_desc_size = 0;
+	u8 *gdesc;
+
+	/* Figure out how large we need gdesc to be. */
+	list_for_each_entry(_img, imgs, node) {
+		const struct acr_r352_ls_func *ls_func =
+					    acr->func->ls_func[_img->falcon_id];
+
+		max_desc_size = max(max_desc_size, ls_func->bl_desc_size);
+	}
+
+	gdesc = kmalloc(max_desc_size, GFP_KERNEL);
+	if (!gdesc)
+		return -ENOMEM;
 
 	nvkm_kmap(wpr_blob);
 
@@ -421,7 +435,6 @@ acr_r352_ls_write_wpr(struct acr_r352 *acr, struct list_head *imgs,
 		struct ls_ucode_img_r352 *img = ls_ucode_img_r352(_img);
 		const struct acr_r352_ls_func *ls_func =
 					    acr->func->ls_func[_img->falcon_id];
-		u8 gdesc[ls_func->bl_desc_size];
 
 		nvkm_gpuobj_memcpy_to(wpr_blob, pos, &img->wpr_header,
 				      sizeof(img->wpr_header));
@@ -447,6 +460,8 @@ acr_r352_ls_write_wpr(struct acr_r352 *acr, struct list_head *imgs,
 
 	nvkm_done(wpr_blob);
 
+	kfree(gdesc);
+
 	return 0;
 }
 
@@ -771,7 +786,11 @@ acr_r352_load(struct nvkm_acr *_acr, struct nvkm_falcon *falcon,
 	struct fw_bl_desc *hsbl_desc;
 	void *bl, *blob_data, *hsbl_code, *hsbl_data;
 	u32 code_size;
-	u8 bl_desc[bl_desc_size];
+	u8 *bl_desc;
+
+	bl_desc = kzalloc(bl_desc_size, GFP_KERNEL);
+	if (!bl_desc)
+		return -ENOMEM;
 
 	/* Find the bootloader descriptor for our blob and copy it */
 	if (blob == acr->load_blob) {
@@ -802,7 +821,6 @@ acr_r352_load(struct nvkm_acr *_acr, struct nvkm_falcon *falcon,
 			      code_size, hsbl_desc->start_tag, 0, false);
 
 	/* Generate the BL header */
-	memset(bl_desc, 0, bl_desc_size);
 	acr->func->generate_hs_bl_desc(load_hdr, bl_desc, offset);
 
 	/*
@@ -811,6 +829,7 @@ acr_r352_load(struct nvkm_acr *_acr, struct nvkm_falcon *falcon,
 	nvkm_falcon_load_dmem(falcon, bl_desc, hsbl_desc->dmem_load_off,
 			      bl_desc_size, 0);
 
+	kfree(bl_desc);
 	return hsbl_desc->start_tag << 8;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r367.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r367.c
index 866877b88797..978ad0790367 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r367.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/acr_r367.c
@@ -265,6 +265,19 @@ acr_r367_ls_write_wpr(struct acr_r352 *acr, struct list_head *imgs,
 {
 	struct ls_ucode_img *_img;
 	u32 pos = 0;
+	u32 max_desc_size = 0;
+	u8 *gdesc;
+
+	list_for_each_entry(_img, imgs, node) {
+		const struct acr_r352_ls_func *ls_func =
+					    acr->func->ls_func[_img->falcon_id];
+
+		max_desc_size = max(max_desc_size, ls_func->bl_desc_size);
+	}
+
+	gdesc = kmalloc(max_desc_size, GFP_KERNEL);
+	if (!gdesc)
+		return -ENOMEM;
 
 	nvkm_kmap(wpr_blob);
 
@@ -272,7 +285,6 @@ acr_r367_ls_write_wpr(struct acr_r352 *acr, struct list_head *imgs,
 		struct ls_ucode_img_r367 *img = ls_ucode_img_r367(_img);
 		const struct acr_r352_ls_func *ls_func =
 					    acr->func->ls_func[_img->falcon_id];
-		u8 gdesc[ls_func->bl_desc_size];
 
 		nvkm_gpuobj_memcpy_to(wpr_blob, pos, &img->wpr_header,
 				      sizeof(img->wpr_header));
@@ -298,6 +310,8 @@ acr_r367_ls_write_wpr(struct acr_r352 *acr, struct list_head *imgs,
 
 	nvkm_done(wpr_blob);
 
+	kfree(gdesc);
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c
index 30491d132d59..df8b919dcf09 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gm20b.c
@@ -129,6 +129,7 @@ gm20b_secboot_new(struct nvkm_device *device, int index,
 	return 0;
 }
 
+#if IS_ENABLED(CONFIG_ARCH_TEGRA_210_SOC)
 MODULE_FIRMWARE("nvidia/gm20b/acr/bl.bin");
 MODULE_FIRMWARE("nvidia/gm20b/acr/ucode_load.bin");
 MODULE_FIRMWARE("nvidia/gm20b/gr/fecs_bl.bin");
@@ -144,3 +145,4 @@ MODULE_FIRMWARE("nvidia/gm20b/gr/sw_method_init.bin");
 MODULE_FIRMWARE("nvidia/gm20b/pmu/desc.bin");
 MODULE_FIRMWARE("nvidia/gm20b/pmu/image.bin");
 MODULE_FIRMWARE("nvidia/gm20b/pmu/sig.bin");
+#endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp108.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp108.c
index e8c27ec700de..737a8d50a1f2 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp108.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp108.c
@@ -65,3 +65,24 @@ MODULE_FIRMWARE("nvidia/gp108/nvdec/scrubber.bin");
 MODULE_FIRMWARE("nvidia/gp108/sec2/desc.bin");
 MODULE_FIRMWARE("nvidia/gp108/sec2/image.bin");
 MODULE_FIRMWARE("nvidia/gp108/sec2/sig.bin");
+
+MODULE_FIRMWARE("nvidia/gv100/acr/bl.bin");
+MODULE_FIRMWARE("nvidia/gv100/acr/unload_bl.bin");
+MODULE_FIRMWARE("nvidia/gv100/acr/ucode_load.bin");
+MODULE_FIRMWARE("nvidia/gv100/acr/ucode_unload.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/fecs_bl.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/fecs_inst.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/fecs_data.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/fecs_sig.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/gpccs_bl.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/gpccs_inst.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/gpccs_data.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/gpccs_sig.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/sw_ctx.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/sw_nonctx.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/sw_bundle_init.bin");
+MODULE_FIRMWARE("nvidia/gv100/gr/sw_method_init.bin");
+MODULE_FIRMWARE("nvidia/gv100/nvdec/scrubber.bin");
+MODULE_FIRMWARE("nvidia/gv100/sec2/desc.bin");
+MODULE_FIRMWARE("nvidia/gv100/sec2/image.bin");
+MODULE_FIRMWARE("nvidia/gv100/sec2/sig.bin");
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp10b.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp10b.c
index 632e9545e292..28ca29d0eeee 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp10b.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/gp10b.c
@@ -74,6 +74,7 @@ gp10b_secboot_new(struct nvkm_device *device, int index,
 	return 0;
 }
 
+#if IS_ENABLED(CONFIG_ARCH_TEGRA_186_SOC)
 MODULE_FIRMWARE("nvidia/gp10b/acr/bl.bin");
 MODULE_FIRMWARE("nvidia/gp10b/acr/ucode_load.bin");
 MODULE_FIRMWARE("nvidia/gp10b/gr/fecs_bl.bin");
@@ -91,3 +92,4 @@ MODULE_FIRMWARE("nvidia/gp10b/gr/sw_method_init.bin");
 MODULE_FIRMWARE("nvidia/gp10b/pmu/desc.bin");
 MODULE_FIRMWARE("nvidia/gp10b/pmu/image.bin");
 MODULE_FIRMWARE("nvidia/gp10b/pmu/sig.bin");
+#endif
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_msgqueue.c b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_msgqueue.c
index 6f10b098676c..1e1f1c635cab 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_msgqueue.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/secboot/ls_ucode_msgqueue.c
@@ -80,12 +80,11 @@ acr_ls_msgqueue_post_run(struct nvkm_msgqueue *queue,
 			 struct nvkm_falcon *falcon, u32 addr_args)
 {
 	struct nvkm_device *device = falcon->owner->device;
-	u32 cmdline_size = NVKM_MSGQUEUE_CMDLINE_SIZE;
-	u8 buf[cmdline_size];
+	u8 buf[NVKM_MSGQUEUE_CMDLINE_SIZE];
 
-	memset(buf, 0, cmdline_size);
+	memset(buf, 0, sizeof(buf));
 	nvkm_msgqueue_write_cmdline(queue, buf);
-	nvkm_falcon_load_dmem(falcon, buf, addr_args, cmdline_size, 0);
+	nvkm_falcon_load_dmem(falcon, buf, addr_args, sizeof(buf), 0);
 	/* rearm the queue so it will wait for the init message */
 	nvkm_msgqueue_reinit(queue);
 
diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.c b/drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.c
index fea4957291da..4f1f3e890650 100644
--- a/drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.c
+++ b/drivers/gpu/drm/nouveau/nvkm/subdev/top/gk104.c
@@ -48,7 +48,8 @@ gk104_top_oneinit(struct nvkm_top *top)
 		case 0x00000001: /* DATA */
 			inst        = (data & 0x3c000000) >> 26;
 			info->addr  = (data & 0x00fff000);
-			info->fault = (data & 0x000000f8) >> 3;
+			if (data & 0x00000004)
+				info->fault = (data & 0x000003f8) >> 3;
 			break;
 		case 0x00000002: /* ENUM */
 			if (data & 0x00000020)
diff --git a/drivers/gpu/drm/omapdrm/displays/panel-sony-acx565akm.c b/drivers/gpu/drm/omapdrm/displays/panel-sony-acx565akm.c
index 92fe125ce22e..f34c06bb5bd7 100644
--- a/drivers/gpu/drm/omapdrm/displays/panel-sony-acx565akm.c
+++ b/drivers/gpu/drm/omapdrm/displays/panel-sony-acx565akm.c
@@ -4,7 +4,7 @@
  * Copyright (C) 2010 Nokia Corporation
  *
  * Original Driver Author: Imre Deak <imre.deak@nokia.com>
- * Based on panel-generic.c by Tomi Valkeinen <tomi.valkeinen@nokia.com>
+ * Based on panel-generic.c by Tomi Valkeinen <tomi.valkeinen@ti.com>
  * Adapted to new DSS2 framework: Roger Quadros <roger.quadros@nokia.com>
  *
  * This program is free software; you can redistribute it and/or modify it
diff --git a/drivers/gpu/drm/omapdrm/displays/panel-tpo-td028ttec1.c b/drivers/gpu/drm/omapdrm/displays/panel-tpo-td028ttec1.c
index b5d8a00df811..a1f1dc18407a 100644
--- a/drivers/gpu/drm/omapdrm/displays/panel-tpo-td028ttec1.c
+++ b/drivers/gpu/drm/omapdrm/displays/panel-tpo-td028ttec1.c
@@ -2,7 +2,7 @@
  * Toppoly TD028TTEC1 panel support
  *
  * Copyright (C) 2008 Nokia Corporation
- * Author: Tomi Valkeinen <tomi.valkeinen@nokia.com>
+ * Author: Tomi Valkeinen <tomi.valkeinen@ti.com>
  *
  * Neo 1973 code (jbt6k74.c):
  * Copyright (C) 2006-2007 by OpenMoko, Inc.
diff --git a/drivers/gpu/drm/omapdrm/dss/core.c b/drivers/gpu/drm/omapdrm/dss/core.c
index acef7ece5783..07d00a186f15 100644
--- a/drivers/gpu/drm/omapdrm/dss/core.c
+++ b/drivers/gpu/drm/omapdrm/dss/core.c
@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2009 Nokia Corporation
- * Author: Tomi Valkeinen <tomi.valkeinen@nokia.com>
+ * Author: Tomi Valkeinen <tomi.valkeinen@ti.com>
  *
  * Some code and ideas taken from drivers/video/omap/ driver
  * by Imre Deak.
@@ -82,7 +82,7 @@ static void __exit omap_dss_exit(void)
 module_init(omap_dss_init);
 module_exit(omap_dss_exit);
 
-MODULE_AUTHOR("Tomi Valkeinen <tomi.valkeinen@nokia.com>");
+MODULE_AUTHOR("Tomi Valkeinen <tomi.valkeinen@ti.com>");
 MODULE_DESCRIPTION("OMAP2/3 Display Subsystem");
 MODULE_LICENSE("GPL v2");
 
diff --git a/drivers/gpu/drm/omapdrm/dss/dispc.c b/drivers/gpu/drm/omapdrm/dss/dispc.c
index 5e2e65e88847..84f274c4a4cb 100644
--- a/drivers/gpu/drm/omapdrm/dss/dispc.c
+++ b/drivers/gpu/drm/omapdrm/dss/dispc.c
@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2009 Nokia Corporation
- * Author: Tomi Valkeinen <tomi.valkeinen@nokia.com>
+ * Author: Tomi Valkeinen <tomi.valkeinen@ti.com>
  *
  * Some code and ideas taken from drivers/video/omap/ driver
  * by Imre Deak.
@@ -828,6 +828,12 @@ static void dispc_ovl_set_scale_coef(struct dispc_device *dispc,
 	h_coef = dispc_ovl_get_scale_coef(fir_hinc, true);
 	v_coef = dispc_ovl_get_scale_coef(fir_vinc, five_taps);
 
+	if (!h_coef || !v_coef) {
+		dev_err(&dispc->pdev->dev, "%s: failed to find scale coefs\n",
+			__func__);
+		return;
+	}
+
 	for (i = 0; i < 8; i++) {
 		u32 h, hv;
 
@@ -2342,7 +2348,7 @@ static int dispc_ovl_calc_scaling_24xx(struct dispc_device *dispc,
 	}
 
 	if (in_width > maxsinglelinewidth) {
-		DSSERR("Cannot scale max input width exceeded");
+		DSSERR("Cannot scale max input width exceeded\n");
 		return -EINVAL;
 	}
 	return 0;
@@ -2424,13 +2430,13 @@ again:
 	}
 
 	if (in_width > (maxsinglelinewidth * 2)) {
-		DSSERR("Cannot setup scaling");
-		DSSERR("width exceeds maximum width possible");
+		DSSERR("Cannot setup scaling\n");
+		DSSERR("width exceeds maximum width possible\n");
 		return -EINVAL;
 	}
 
 	if (in_width > maxsinglelinewidth && *five_taps) {
-		DSSERR("cannot setup scaling with five taps");
+		DSSERR("cannot setup scaling with five taps\n");
 		return -EINVAL;
 	}
 	return 0;
@@ -2472,7 +2478,7 @@ static int dispc_ovl_calc_scaling_44xx(struct dispc_device *dispc,
 			in_width > maxsinglelinewidth && ++*decim_x);
 
 	if (in_width > maxsinglelinewidth) {
-		DSSERR("Cannot scale width exceeds max line width");
+		DSSERR("Cannot scale width exceeds max line width\n");
 		return -EINVAL;
 	}
 
@@ -2490,7 +2496,7 @@ static int dispc_ovl_calc_scaling_44xx(struct dispc_device *dispc,
 		 * bandwidth. Despite what theory says this appears to
 		 * be true also for 16-bit color formats.
 		 */
-		DSSERR("Not enough bandwidth, too much downscaling (x-decimation factor %d > 4)", *decim_x);
+		DSSERR("Not enough bandwidth, too much downscaling (x-decimation factor %d > 4)\n", *decim_x);
 
 		return -EINVAL;
 	}
@@ -4633,7 +4639,7 @@ static int dispc_errata_i734_wa_init(struct dispc_device *dispc)
 						i734_buf.size, &i734_buf.paddr,
 						GFP_KERNEL);
 	if (!i734_buf.vaddr) {
-		dev_err(&dispc->pdev->dev, "%s: dma_alloc_writecombine failed",
+		dev_err(&dispc->pdev->dev, "%s: dma_alloc_writecombine failed\n",
 			__func__);
 		return -ENOMEM;
 	}
diff --git a/drivers/gpu/drm/omapdrm/dss/display.c b/drivers/gpu/drm/omapdrm/dss/display.c
index 424143128cd4..9e7fcbd57e52 100644
--- a/drivers/gpu/drm/omapdrm/dss/display.c
+++ b/drivers/gpu/drm/omapdrm/dss/display.c
@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2009 Nokia Corporation
- * Author: Tomi Valkeinen <tomi.valkeinen@nokia.com>
+ * Author: Tomi Valkeinen <tomi.valkeinen@ti.com>
  *
  * Some code and ideas taken from drivers/video/omap/ driver
  * by Imre Deak.
diff --git a/drivers/gpu/drm/omapdrm/dss/dpi.c b/drivers/gpu/drm/omapdrm/dss/dpi.c
index 3d662e6805eb..9fcc50217133 100644
--- a/drivers/gpu/drm/omapdrm/dss/dpi.c
+++ b/drivers/gpu/drm/omapdrm/dss/dpi.c
@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2009 Nokia Corporation
- * Author: Tomi Valkeinen <tomi.valkeinen@nokia.com>
+ * Author: Tomi Valkeinen <tomi.valkeinen@ti.com>
  *
  * Some code and ideas taken from drivers/video/omap/ driver
  * by Imre Deak.
diff --git a/drivers/gpu/drm/omapdrm/dss/dsi.c b/drivers/gpu/drm/omapdrm/dss/dsi.c
index d4a680629825..74467b308721 100644
--- a/drivers/gpu/drm/omapdrm/dss/dsi.c
+++ b/drivers/gpu/drm/omapdrm/dss/dsi.c
@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2009 Nokia Corporation
- * Author: Tomi Valkeinen <tomi.valkeinen@nokia.com>
+ * Author: Tomi Valkeinen <tomi.valkeinen@ti.com>
  *
  * This program is free software; you can redistribute it and/or modify it
  * under the terms of the GNU General Public License version 2 as published by
diff --git a/drivers/gpu/drm/omapdrm/dss/dss.c b/drivers/gpu/drm/omapdrm/dss/dss.c
index 0b908e9de792..cb80ddaa19d2 100644
--- a/drivers/gpu/drm/omapdrm/dss/dss.c
+++ b/drivers/gpu/drm/omapdrm/dss/dss.c
@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2009 Nokia Corporation
- * Author: Tomi Valkeinen <tomi.valkeinen@nokia.com>
+ * Author: Tomi Valkeinen <tomi.valkeinen@ti.com>
  *
  * Some code and ideas taken from drivers/video/omap/ driver
  * by Imre Deak.
diff --git a/drivers/gpu/drm/omapdrm/dss/dss.h b/drivers/gpu/drm/omapdrm/dss/dss.h
index 847c78ade024..38302631b64b 100644
--- a/drivers/gpu/drm/omapdrm/dss/dss.h
+++ b/drivers/gpu/drm/omapdrm/dss/dss.h
@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2009 Nokia Corporation
- * Author: Tomi Valkeinen <tomi.valkeinen@nokia.com>
+ * Author: Tomi Valkeinen <tomi.valkeinen@ti.com>
  *
  * Some code and ideas taken from drivers/video/omap/ driver
  * by Imre Deak.
@@ -180,6 +180,9 @@ struct dss_pll_hw {
 
 	/* DRA7 errata i886: use high N & M to avoid jitter */
 	bool errata_i886;
+
+	/* DRA7 errata i932: retry pll lock on failure */
+	bool errata_i932;
 };
 
 struct dss_pll {
diff --git a/drivers/gpu/drm/omapdrm/dss/hdmi4.c b/drivers/gpu/drm/omapdrm/dss/hdmi4.c
index 97c88861d67a..5879f45f6fc9 100644
--- a/drivers/gpu/drm/omapdrm/dss/hdmi4.c
+++ b/drivers/gpu/drm/omapdrm/dss/hdmi4.c
@@ -679,7 +679,7 @@ static int hdmi_audio_config(struct device *dev,
 			     struct omap_dss_audio *dss_audio)
 {
 	struct omap_hdmi *hd = dev_get_drvdata(dev);
-	int ret;
+	int ret = 0;
 
 	mutex_lock(&hd->lock);
 
diff --git a/drivers/gpu/drm/omapdrm/dss/hdmi4_core.c b/drivers/gpu/drm/omapdrm/dss/hdmi4_core.c
index 35ed2add6189..813ba42f2753 100644
--- a/drivers/gpu/drm/omapdrm/dss/hdmi4_core.c
+++ b/drivers/gpu/drm/omapdrm/dss/hdmi4_core.c
@@ -922,8 +922,13 @@ int hdmi4_core_init(struct platform_device *pdev, struct hdmi_core_data *core)
 {
 	const struct hdmi4_features *features;
 	struct resource *res;
+	const struct soc_device_attribute *soc;
 
-	features = soc_device_match(hdmi4_soc_devices)->data;
+	soc = soc_device_match(hdmi4_soc_devices);
+	if (!soc)
+		return -ENODEV;
+
+	features = soc->data;
 	core->cts_swmode = features->cts_swmode;
 	core->audio_use_mclk = features->audio_use_mclk;
 
diff --git a/drivers/gpu/drm/omapdrm/dss/hdmi5.c b/drivers/gpu/drm/omapdrm/dss/hdmi5.c
index d28da9ac3e90..ae1a001d1b83 100644
--- a/drivers/gpu/drm/omapdrm/dss/hdmi5.c
+++ b/drivers/gpu/drm/omapdrm/dss/hdmi5.c
@@ -671,7 +671,7 @@ static int hdmi_audio_config(struct device *dev,
 			     struct omap_dss_audio *dss_audio)
 {
 	struct omap_hdmi *hd = dev_get_drvdata(dev);
-	int ret;
+	int ret = 0;
 
 	mutex_lock(&hd->lock);
 
diff --git a/drivers/gpu/drm/omapdrm/dss/pll.c b/drivers/gpu/drm/omapdrm/dss/pll.c
index 078b0e8216c3..ff362b38bf0d 100644
--- a/drivers/gpu/drm/omapdrm/dss/pll.c
+++ b/drivers/gpu/drm/omapdrm/dss/pll.c
@@ -16,6 +16,7 @@
 
 #define DSS_SUBSYS_NAME "PLL"
 
+#include <linux/delay.h>
 #include <linux/clk.h>
 #include <linux/io.h>
 #include <linux/kernel.h>
@@ -381,6 +382,22 @@ static int dss_wait_hsdiv_ack(struct dss_pll *pll, u32 hsdiv_ack_mask)
 	return -ETIMEDOUT;
 }
 
+static bool pll_is_locked(u32 stat)
+{
+	/*
+	 * Required value for each bitfield listed below
+	 *
+	 * PLL_STATUS[6] = 0  PLL_BYPASS
+	 * PLL_STATUS[5] = 0  PLL_HIGHJITTER
+	 *
+	 * PLL_STATUS[3] = 0  PLL_LOSSREF
+	 * PLL_STATUS[2] = 0  PLL_RECAL
+	 * PLL_STATUS[1] = 1  PLL_LOCK
+	 * PLL_STATUS[0] = 1  PLL_CTRL_RESET_DONE
+	 */
+	return ((stat & 0x6f) == 0x3);
+}
+
 int dss_pll_write_config_type_a(struct dss_pll *pll,
 		const struct dss_pll_clock_info *cinfo)
 {
@@ -436,18 +453,54 @@ int dss_pll_write_config_type_a(struct dss_pll *pll,
 	l = FLD_MOD(l, 0, 25, 25);		/* M7_CLOCK_EN */
 	writel_relaxed(l, base + PLL_CONFIGURATION2);
 
-	writel_relaxed(1, base + PLL_GO);	/* PLL_GO */
+	if (hw->errata_i932) {
+		int cnt = 0;
+		u32 sleep_time;
+		const u32 max_lock_retries = 20;
 
-	if (wait_for_bit_change(base + PLL_GO, 0, 0) != 0) {
-		DSSERR("DSS DPLL GO bit not going down.\n");
-		r = -EIO;
-		goto err;
-	}
+		/*
+		 * Calculate wait time for PLL LOCK
+		 * 1000 REFCLK cycles in us.
+		 */
+		sleep_time = DIV_ROUND_UP(1000*1000*1000, cinfo->fint);
 
-	if (wait_for_bit_change(base + PLL_STATUS, 1, 1) != 1) {
-		DSSERR("cannot lock DSS DPLL\n");
-		r = -EIO;
-		goto err;
+		for (cnt = 0; cnt < max_lock_retries; cnt++) {
+			writel_relaxed(1, base + PLL_GO);	/* PLL_GO */
+
+			/**
+			 * read the register back to ensure the write is
+			 * flushed
+			 */
+			readl_relaxed(base + PLL_GO);
+
+			usleep_range(sleep_time, sleep_time + 5);
+			l = readl_relaxed(base + PLL_STATUS);
+
+			if (pll_is_locked(l) &&
+			    !(readl_relaxed(base + PLL_GO) & 0x1))
+				break;
+
+		}
+
+		if (cnt == max_lock_retries) {
+			DSSERR("cannot lock PLL\n");
+			r = -EIO;
+			goto err;
+		}
+	} else {
+		writel_relaxed(1, base + PLL_GO);	/* PLL_GO */
+
+		if (wait_for_bit_change(base + PLL_GO, 0, 0) != 0) {
+			DSSERR("DSS DPLL GO bit not going down.\n");
+			r = -EIO;
+			goto err;
+		}
+
+		if (wait_for_bit_change(base + PLL_STATUS, 1, 1) != 1) {
+			DSSERR("cannot lock DSS DPLL\n");
+			r = -EIO;
+			goto err;
+		}
 	}
 
 	l = readl_relaxed(base + PLL_CONFIGURATION2);
diff --git a/drivers/gpu/drm/omapdrm/dss/sdi.c b/drivers/gpu/drm/omapdrm/dss/sdi.c
index 68a40ae26f5b..69c3b7a3d5c7 100644
--- a/drivers/gpu/drm/omapdrm/dss/sdi.c
+++ b/drivers/gpu/drm/omapdrm/dss/sdi.c
@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2009 Nokia Corporation
- * Author: Tomi Valkeinen <tomi.valkeinen@nokia.com>
+ * Author: Tomi Valkeinen <tomi.valkeinen@ti.com>
  *
  * This program is free software; you can redistribute it and/or modify it
  * under the terms of the GNU General Public License version 2 as published by
@@ -82,7 +82,7 @@ static int sdi_calc_clock_div(struct sdi_device *sdi, unsigned long pclk,
 			      struct dispc_clock_info *dispc_cinfo)
 {
 	int i;
-	struct sdi_clk_calc_ctx ctx = { .sdi = sdi };
+	struct sdi_clk_calc_ctx ctx;
 
 	/*
 	 * DSS fclk gives us very few possibilities, so finding a good pixel
@@ -95,6 +95,9 @@ static int sdi_calc_clock_div(struct sdi_device *sdi, unsigned long pclk,
 		bool ok;
 
 		memset(&ctx, 0, sizeof(ctx));
+
+		ctx.sdi = sdi;
+
 		if (pclk > 1000 * i * i * i)
 			ctx.pck_min = max(pclk - 1000 * i * i * i, 0lu);
 		else
diff --git a/drivers/gpu/drm/omapdrm/dss/venc.c b/drivers/gpu/drm/omapdrm/dss/venc.c
index 24d1ced210bd..ac01907dcc34 100644
--- a/drivers/gpu/drm/omapdrm/dss/venc.c
+++ b/drivers/gpu/drm/omapdrm/dss/venc.c
@@ -1,6 +1,6 @@
 /*
  * Copyright (C) 2009 Nokia Corporation
- * Author: Tomi Valkeinen <tomi.valkeinen@nokia.com>
+ * Author: Tomi Valkeinen <tomi.valkeinen@ti.com>
  *
  * VENC settings from TI's DSS driver
  *
diff --git a/drivers/gpu/drm/omapdrm/dss/video-pll.c b/drivers/gpu/drm/omapdrm/dss/video-pll.c
index 585ed94ccf17..cb46311f92c9 100644
--- a/drivers/gpu/drm/omapdrm/dss/video-pll.c
+++ b/drivers/gpu/drm/omapdrm/dss/video-pll.c
@@ -134,6 +134,7 @@ static const struct dss_pll_hw dss_dra7_video_pll_hw = {
 	.has_refsel = true,
 
 	.errata_i886 = true,
+	.errata_i932 = true,
 };
 
 struct dss_pll *dss_video_pll_init(struct dss_device *dss,
diff --git a/drivers/gpu/drm/omapdrm/omap_connector.c b/drivers/gpu/drm/omapdrm/omap_connector.c
index a0d7b1d905e8..2ddb856666c4 100644
--- a/drivers/gpu/drm/omapdrm/omap_connector.c
+++ b/drivers/gpu/drm/omapdrm/omap_connector.c
@@ -121,16 +121,19 @@ static int omap_connector_get_modes(struct drm_connector *connector)
 	if (dssdrv->read_edid) {
 		void *edid = kzalloc(MAX_EDID, GFP_KERNEL);
 
+		if (!edid)
+			return 0;
+
 		if ((dssdrv->read_edid(dssdev, edid, MAX_EDID) > 0) &&
 				drm_edid_is_valid(edid)) {
-			drm_mode_connector_update_edid_property(
+			drm_connector_update_edid_property(
 					connector, edid);
 			n = drm_add_edid_modes(connector, edid);
 
 			omap_connector->hdmi_mode =
 				drm_detect_hdmi_monitor(edid);
 		} else {
-			drm_mode_connector_update_edid_property(
+			drm_connector_update_edid_property(
 					connector, NULL);
 		}
 
@@ -139,6 +142,9 @@ static int omap_connector_get_modes(struct drm_connector *connector)
 		struct drm_display_mode *mode = drm_mode_create(dev);
 		struct videomode vm = {0};
 
+		if (!mode)
+			return 0;
+
 		dssdrv->get_timings(dssdev, &vm);
 
 		drm_display_mode_from_videomode(&vm, mode);
@@ -200,6 +206,10 @@ static int omap_connector_mode_valid(struct drm_connector *connector,
 	if (!r) {
 		/* check if vrefresh is still valid */
 		new_mode = drm_mode_duplicate(dev, mode);
+
+		if (!new_mode)
+			return MODE_BAD;
+
 		new_mode->clock = vm.pixelclock / 1000;
 		new_mode->vrefresh = 0;
 		if (mode->vrefresh == drm_mode_vrefresh(new_mode))
diff --git a/drivers/gpu/drm/omapdrm/omap_debugfs.c b/drivers/gpu/drm/omapdrm/omap_debugfs.c
index b42e286616b0..91cf043f2b6b 100644
--- a/drivers/gpu/drm/omapdrm/omap_debugfs.c
+++ b/drivers/gpu/drm/omapdrm/omap_debugfs.c
@@ -30,16 +30,11 @@ static int gem_show(struct seq_file *m, void *arg)
 	struct drm_info_node *node = (struct drm_info_node *) m->private;
 	struct drm_device *dev = node->minor->dev;
 	struct omap_drm_private *priv = dev->dev_private;
-	int ret;
-
-	ret = mutex_lock_interruptible(&dev->struct_mutex);
-	if (ret)
-		return ret;
 
 	seq_printf(m, "All Objects:\n");
+	mutex_lock(&priv->list_lock);
 	omap_gem_describe_objects(&priv->obj_list, m);
-
-	mutex_unlock(&dev->struct_mutex);
+	mutex_unlock(&priv->list_lock);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c b/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c
index f9fa1c90b35c..f92fe205550b 100644
--- a/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c
+++ b/drivers/gpu/drm/omapdrm/omap_dmm_tiler.c
@@ -401,12 +401,16 @@ int tiler_unpin(struct tiler_block *block)
 struct tiler_block *tiler_reserve_2d(enum tiler_fmt fmt, u16 w,
 		u16 h, u16 align)
 {
-	struct tiler_block *block = kzalloc(sizeof(*block), GFP_KERNEL);
+	struct tiler_block *block;
 	u32 min_align = 128;
 	int ret;
 	unsigned long flags;
 	u32 slot_bytes;
 
+	block = kzalloc(sizeof(*block), GFP_KERNEL);
+	if (!block)
+		return ERR_PTR(-ENOMEM);
+
 	BUG_ON(!validfmt(fmt));
 
 	/* convert width/height to slots */
@@ -936,8 +940,8 @@ int tiler_map_show(struct seq_file *s, void *arg)
 	h_adj = omap_dmm->container_height / ydiv;
 	w_adj = omap_dmm->container_width / xdiv;
 
-	map = kmalloc(h_adj * sizeof(*map), GFP_KERNEL);
-	global_map = kmalloc((w_adj + 1) * h_adj, GFP_KERNEL);
+	map = kmalloc_array(h_adj, sizeof(*map), GFP_KERNEL);
+	global_map = kmalloc_array(w_adj + 1, h_adj, GFP_KERNEL);
 
 	if (!map || !global_map)
 		goto error;
diff --git a/drivers/gpu/drm/omapdrm/omap_drv.c b/drivers/gpu/drm/omapdrm/omap_drv.c
index ef3b0e3571ec..1b6601e9b107 100644
--- a/drivers/gpu/drm/omapdrm/omap_drv.c
+++ b/drivers/gpu/drm/omapdrm/omap_drv.c
@@ -274,7 +274,7 @@ static int omap_modeset_init(struct drm_device *dev)
 		if (IS_ERR(crtc))
 			return PTR_ERR(crtc);
 
-		drm_mode_connector_attach_encoder(connector, encoder);
+		drm_connector_attach_encoder(connector, encoder);
 		encoder->possible_crtcs = (1 << crtc_idx);
 
 		priv->crtcs[priv->num_crtcs++] = crtc;
@@ -493,7 +493,7 @@ static struct drm_driver omap_drm_driver = {
 	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
 	.gem_prime_export = omap_gem_prime_export,
 	.gem_prime_import = omap_gem_prime_import,
-	.gem_free_object = omap_gem_free_object,
+	.gem_free_object_unlocked = omap_gem_free_object,
 	.gem_vm_ops = &omap_gem_vm_ops,
 	.dumb_create = omap_gem_dumb_create,
 	.dumb_map_offset = omap_gem_dumb_map_offset,
@@ -540,7 +540,7 @@ static int omapdrm_init(struct omap_drm_private *priv, struct device *dev)
 	priv->omaprev = soc ? (unsigned int)soc->data : 0;
 	priv->wq = alloc_ordered_workqueue("omapdrm", 0);
 
-	spin_lock_init(&priv->list_lock);
+	mutex_init(&priv->list_lock);
 	INIT_LIST_HEAD(&priv->obj_list);
 
 	/* Allocate and initialize the DRM device. */
diff --git a/drivers/gpu/drm/omapdrm/omap_drv.h b/drivers/gpu/drm/omapdrm/omap_drv.h
index 6eaee4df4559..f27c8e216adf 100644
--- a/drivers/gpu/drm/omapdrm/omap_drv.h
+++ b/drivers/gpu/drm/omapdrm/omap_drv.h
@@ -71,7 +71,7 @@ struct omap_drm_private {
 	struct workqueue_struct *wq;
 
 	/* lock for obj_list below */
-	spinlock_t list_lock;
+	struct mutex list_lock;
 
 	/* list of GEM objects: */
 	struct list_head obj_list;
diff --git a/drivers/gpu/drm/omapdrm/omap_fb.c b/drivers/gpu/drm/omapdrm/omap_fb.c
index 5fd22ca73913..9f1e3d8f8488 100644
--- a/drivers/gpu/drm/omapdrm/omap_fb.c
+++ b/drivers/gpu/drm/omapdrm/omap_fb.c
@@ -19,6 +19,7 @@
 
 #include <drm/drm_crtc.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 
 #include "omap_dmm_tiler.h"
 #include "omap_drv.h"
@@ -51,9 +52,6 @@ static const u32 formats[] = {
 
 /* per-plane info for the fb: */
 struct plane {
-	struct drm_gem_object *bo;
-	u32 pitch;
-	u32 offset;
 	dma_addr_t dma_addr;
 };
 
@@ -68,56 +66,28 @@ struct omap_framebuffer {
 	struct mutex lock;
 };
 
-static int omap_framebuffer_create_handle(struct drm_framebuffer *fb,
-		struct drm_file *file_priv,
-		unsigned int *handle)
-{
-	struct omap_framebuffer *omap_fb = to_omap_framebuffer(fb);
-	return drm_gem_handle_create(file_priv,
-			omap_fb->planes[0].bo, handle);
-}
-
-static void omap_framebuffer_destroy(struct drm_framebuffer *fb)
-{
-	struct omap_framebuffer *omap_fb = to_omap_framebuffer(fb);
-	int i, n = fb->format->num_planes;
-
-	DBG("destroy: FB ID: %d (%p)", fb->base.id, fb);
-
-	drm_framebuffer_cleanup(fb);
-
-	for (i = 0; i < n; i++) {
-		struct plane *plane = &omap_fb->planes[i];
-
-		drm_gem_object_unreference_unlocked(plane->bo);
-	}
-
-	kfree(omap_fb);
-}
-
 static const struct drm_framebuffer_funcs omap_framebuffer_funcs = {
-	.create_handle = omap_framebuffer_create_handle,
-	.destroy = omap_framebuffer_destroy,
+	.create_handle = drm_gem_fb_create_handle,
+	.destroy = drm_gem_fb_destroy,
 };
 
-static u32 get_linear_addr(struct plane *plane,
+static u32 get_linear_addr(struct drm_framebuffer *fb,
 		const struct drm_format_info *format, int n, int x, int y)
 {
+	struct omap_framebuffer *omap_fb = to_omap_framebuffer(fb);
+	struct plane *plane = &omap_fb->planes[n];
 	u32 offset;
 
-	offset = plane->offset
+	offset = fb->offsets[n]
 	       + (x * format->cpp[n] / (n == 0 ? 1 : format->hsub))
-	       + (y * plane->pitch / (n == 0 ? 1 : format->vsub));
+	       + (y * fb->pitches[n] / (n == 0 ? 1 : format->vsub));
 
 	return plane->dma_addr + offset;
 }
 
 bool omap_framebuffer_supports_rotation(struct drm_framebuffer *fb)
 {
-	struct omap_framebuffer *omap_fb = to_omap_framebuffer(fb);
-	struct plane *plane = &omap_fb->planes[0];
-
-	return omap_gem_flags(plane->bo) & OMAP_BO_TILED;
+	return omap_gem_flags(fb->obj[0]) & OMAP_BO_TILED;
 }
 
 /* Note: DRM rotates counter-clockwise, TILER & DSS rotates clockwise */
@@ -176,7 +146,7 @@ void omap_framebuffer_update_scanout(struct drm_framebuffer *fb,
 	x = state->src_x >> 16;
 	y = state->src_y >> 16;
 
-	if (omap_gem_flags(plane->bo) & OMAP_BO_TILED) {
+	if (omap_gem_flags(fb->obj[0]) & OMAP_BO_TILED) {
 		u32 w = state->src_w >> 16;
 		u32 h = state->src_h >> 16;
 
@@ -201,12 +171,12 @@ void omap_framebuffer_update_scanout(struct drm_framebuffer *fb,
 			x += w - 1;
 
 		/* Note: x and y are in TILER units, not pixels */
-		omap_gem_rotated_dma_addr(plane->bo, orient, x, y,
+		omap_gem_rotated_dma_addr(fb->obj[0], orient, x, y,
 					  &info->paddr);
 		info->rotation_type = OMAP_DSS_ROT_TILER;
 		info->rotation = state->rotation ?: DRM_MODE_ROTATE_0;
 		/* Note: stride in TILER units, not pixels */
-		info->screen_width  = omap_gem_tiled_stride(plane->bo, orient);
+		info->screen_width  = omap_gem_tiled_stride(fb->obj[0], orient);
 	} else {
 		switch (state->rotation & DRM_MODE_ROTATE_MASK) {
 		case 0:
@@ -221,10 +191,10 @@ void omap_framebuffer_update_scanout(struct drm_framebuffer *fb,
 			break;
 		}
 
-		info->paddr         = get_linear_addr(plane, format, 0, x, y);
+		info->paddr         = get_linear_addr(fb, format, 0, x, y);
 		info->rotation_type = OMAP_DSS_ROT_NONE;
 		info->rotation      = DRM_MODE_ROTATE_0;
-		info->screen_width  = plane->pitch;
+		info->screen_width  = fb->pitches[0];
 	}
 
 	/* convert to pixels: */
@@ -234,11 +204,11 @@ void omap_framebuffer_update_scanout(struct drm_framebuffer *fb,
 		plane = &omap_fb->planes[1];
 
 		if (info->rotation_type == OMAP_DSS_ROT_TILER) {
-			WARN_ON(!(omap_gem_flags(plane->bo) & OMAP_BO_TILED));
-			omap_gem_rotated_dma_addr(plane->bo, orient, x/2, y/2,
+			WARN_ON(!(omap_gem_flags(fb->obj[1]) & OMAP_BO_TILED));
+			omap_gem_rotated_dma_addr(fb->obj[1], orient, x/2, y/2,
 						  &info->p_uv_addr);
 		} else {
-			info->p_uv_addr = get_linear_addr(plane, format, 1, x, y);
+			info->p_uv_addr = get_linear_addr(fb, format, 1, x, y);
 		}
 	} else {
 		info->p_uv_addr = 0;
@@ -261,10 +231,10 @@ int omap_framebuffer_pin(struct drm_framebuffer *fb)
 
 	for (i = 0; i < n; i++) {
 		struct plane *plane = &omap_fb->planes[i];
-		ret = omap_gem_pin(plane->bo, &plane->dma_addr);
+		ret = omap_gem_pin(fb->obj[i], &plane->dma_addr);
 		if (ret)
 			goto fail;
-		omap_gem_dma_sync_buffer(plane->bo, DMA_TO_DEVICE);
+		omap_gem_dma_sync_buffer(fb->obj[i], DMA_TO_DEVICE);
 	}
 
 	omap_fb->pin_count++;
@@ -276,7 +246,7 @@ int omap_framebuffer_pin(struct drm_framebuffer *fb)
 fail:
 	for (i--; i >= 0; i--) {
 		struct plane *plane = &omap_fb->planes[i];
-		omap_gem_unpin(plane->bo);
+		omap_gem_unpin(fb->obj[i]);
 		plane->dma_addr = 0;
 	}
 
@@ -302,54 +272,25 @@ void omap_framebuffer_unpin(struct drm_framebuffer *fb)
 
 	for (i = 0; i < n; i++) {
 		struct plane *plane = &omap_fb->planes[i];
-		omap_gem_unpin(plane->bo);
+		omap_gem_unpin(fb->obj[i]);
 		plane->dma_addr = 0;
 	}
 
 	mutex_unlock(&omap_fb->lock);
 }
 
-/* iterate thru all the connectors, returning ones that are attached
- * to the same fb..
- */
-struct drm_connector *omap_framebuffer_get_next_connector(
-		struct drm_framebuffer *fb, struct drm_connector *from)
-{
-	struct drm_device *dev = fb->dev;
-	struct list_head *connector_list = &dev->mode_config.connector_list;
-	struct drm_connector *connector = from;
-
-	if (!from)
-		return list_first_entry_or_null(connector_list, typeof(*from),
-						head);
-
-	list_for_each_entry_from(connector, connector_list, head) {
-		if (connector != from) {
-			struct drm_encoder *encoder = connector->encoder;
-			struct drm_crtc *crtc = encoder ? encoder->crtc : NULL;
-			if (crtc && crtc->primary->fb == fb)
-				return connector;
-
-		}
-	}
-
-	return NULL;
-}
-
 #ifdef CONFIG_DEBUG_FS
 void omap_framebuffer_describe(struct drm_framebuffer *fb, struct seq_file *m)
 {
-	struct omap_framebuffer *omap_fb = to_omap_framebuffer(fb);
 	int i, n = fb->format->num_planes;
 
 	seq_printf(m, "fb: %dx%d@%4.4s\n", fb->width, fb->height,
 			(char *)&fb->format->format);
 
 	for (i = 0; i < n; i++) {
-		struct plane *plane = &omap_fb->planes[i];
 		seq_printf(m, "   %d: offset=%d pitch=%d, obj: ",
-				i, plane->offset, plane->pitch);
-		omap_gem_describe(plane->bo, m);
+				i, fb->offsets[n], fb->pitches[i]);
+		omap_gem_describe(fb->obj[i], m);
 	}
 }
 #endif
@@ -454,9 +395,7 @@ struct drm_framebuffer *omap_framebuffer_init(struct drm_device *dev,
 			goto fail;
 		}
 
-		plane->bo     = bos[i];
-		plane->offset = mode_cmd->offsets[i];
-		plane->pitch  = pitch;
+		fb->obj[i]    = bos[i];
 		plane->dma_addr  = 0;
 	}
 
diff --git a/drivers/gpu/drm/omapdrm/omap_fb.h b/drivers/gpu/drm/omapdrm/omap_fb.h
index 94ad5f9e4404..c20cb4bc714d 100644
--- a/drivers/gpu/drm/omapdrm/omap_fb.h
+++ b/drivers/gpu/drm/omapdrm/omap_fb.h
@@ -38,8 +38,6 @@ int omap_framebuffer_pin(struct drm_framebuffer *fb);
 void omap_framebuffer_unpin(struct drm_framebuffer *fb);
 void omap_framebuffer_update_scanout(struct drm_framebuffer *fb,
 		struct drm_plane_state *state, struct omap_overlay_info *info);
-struct drm_connector *omap_framebuffer_get_next_connector(
-		struct drm_framebuffer *fb, struct drm_connector *from);
 bool omap_framebuffer_supports_rotation(struct drm_framebuffer *fb);
 void omap_framebuffer_describe(struct drm_framebuffer *fb, struct seq_file *m);
 
diff --git a/drivers/gpu/drm/omapdrm/omap_fbdev.c b/drivers/gpu/drm/omapdrm/omap_fbdev.c
index 0f66c74a54b0..d958cc813a94 100644
--- a/drivers/gpu/drm/omapdrm/omap_fbdev.c
+++ b/drivers/gpu/drm/omapdrm/omap_fbdev.c
@@ -170,13 +170,11 @@ static int omap_fbdev_create(struct drm_fb_helper *helper,
 		goto fail;
 	}
 
-	mutex_lock(&dev->struct_mutex);
-
 	fbi = drm_fb_helper_alloc_fbi(helper);
 	if (IS_ERR(fbi)) {
 		dev_err(dev->dev, "failed to allocate fb info\n");
 		ret = PTR_ERR(fbi);
-		goto fail_unlock;
+		goto fail;
 	}
 
 	DBG("fbi=%p, dev=%p", fbi, dev);
@@ -212,12 +210,8 @@ static int omap_fbdev_create(struct drm_fb_helper *helper,
 	DBG("par=%p, %dx%d", fbi->par, fbi->var.xres, fbi->var.yres);
 	DBG("allocated %dx%d fb", fbdev->fb->width, fbdev->fb->height);
 
-	mutex_unlock(&dev->struct_mutex);
-
 	return 0;
 
-fail_unlock:
-	mutex_unlock(&dev->struct_mutex);
 fail:
 
 	if (ret) {
diff --git a/drivers/gpu/drm/omapdrm/omap_gem.c b/drivers/gpu/drm/omapdrm/omap_gem.c
index 0faf042b82e1..4ba5d035c590 100644
--- a/drivers/gpu/drm/omapdrm/omap_gem.c
+++ b/drivers/gpu/drm/omapdrm/omap_gem.c
@@ -47,6 +47,9 @@ struct omap_gem_object {
 	/** roll applied when mapping to DMM */
 	u32 roll;
 
+	/** protects dma_addr_cnt, block, pages, dma_addrs and vaddr */
+	struct mutex lock;
+
 	/**
 	 * dma_addr contains the buffer DMA address. It is valid for
 	 *
@@ -137,14 +140,12 @@ struct omap_drm_usergart {
  */
 
 /** get mmap offset */
-static u64 mmap_offset(struct drm_gem_object *obj)
+u64 omap_gem_mmap_offset(struct drm_gem_object *obj)
 {
 	struct drm_device *dev = obj->dev;
 	int ret;
 	size_t size;
 
-	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
-
 	/* Make it mmapable */
 	size = omap_gem_mmap_size(obj);
 	ret = drm_gem_create_mmap_offset_size(obj, size);
@@ -156,7 +157,7 @@ static u64 mmap_offset(struct drm_gem_object *obj)
 	return drm_vma_node_offset_addr(&obj->vma_node);
 }
 
-static bool is_contiguous(struct omap_gem_object *omap_obj)
+static bool omap_gem_is_contiguous(struct omap_gem_object *omap_obj)
 {
 	if (omap_obj->flags & OMAP_BO_MEM_DMA_API)
 		return true;
@@ -171,14 +172,14 @@ static bool is_contiguous(struct omap_gem_object *omap_obj)
  * Eviction
  */
 
-static void evict_entry(struct drm_gem_object *obj,
+static void omap_gem_evict_entry(struct drm_gem_object *obj,
 		enum tiler_fmt fmt, struct omap_drm_usergart_entry *entry)
 {
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
 	struct omap_drm_private *priv = obj->dev->dev_private;
 	int n = priv->usergart[fmt].height;
 	size_t size = PAGE_SIZE * n;
-	loff_t off = mmap_offset(obj) +
+	loff_t off = omap_gem_mmap_offset(obj) +
 			(entry->obj_pgoff << PAGE_SHIFT);
 	const int m = DIV_ROUND_UP(omap_obj->width << fmt, PAGE_SIZE);
 
@@ -199,7 +200,7 @@ static void evict_entry(struct drm_gem_object *obj,
 }
 
 /* Evict a buffer from usergart, if it is mapped there */
-static void evict(struct drm_gem_object *obj)
+static void omap_gem_evict(struct drm_gem_object *obj)
 {
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
 	struct omap_drm_private *priv = obj->dev->dev_private;
@@ -213,7 +214,7 @@ static void evict(struct drm_gem_object *obj)
 				&priv->usergart[fmt].entry[i];
 
 			if (entry->obj == obj)
-				evict_entry(obj, fmt, entry);
+				omap_gem_evict_entry(obj, fmt, entry);
 		}
 	}
 }
@@ -222,7 +223,10 @@ static void evict(struct drm_gem_object *obj)
  * Page Management
  */
 
-/** ensure backing pages are allocated */
+/*
+ * Ensure backing pages are allocated. Must be called with the omap_obj.lock
+ * held.
+ */
 static int omap_gem_attach_pages(struct drm_gem_object *obj)
 {
 	struct drm_device *dev = obj->dev;
@@ -232,7 +236,14 @@ static int omap_gem_attach_pages(struct drm_gem_object *obj)
 	int i, ret;
 	dma_addr_t *addrs;
 
-	WARN_ON(omap_obj->pages);
+	lockdep_assert_held(&omap_obj->lock);
+
+	/*
+	 * If not using shmem (in which case backing pages don't need to be
+	 * allocated) or if pages are already allocated we're done.
+	 */
+	if (!(omap_obj->flags & OMAP_BO_MEM_SHMEM) || omap_obj->pages)
+		return 0;
 
 	pages = drm_gem_get_pages(obj);
 	if (IS_ERR(pages)) {
@@ -244,7 +255,7 @@ static int omap_gem_attach_pages(struct drm_gem_object *obj)
 	 * DSS, GPU, etc. are not cache coherent:
 	 */
 	if (omap_obj->flags & (OMAP_BO_WC|OMAP_BO_UNCACHED)) {
-		addrs = kmalloc(npages * sizeof(*addrs), GFP_KERNEL);
+		addrs = kmalloc_array(npages, sizeof(*addrs), GFP_KERNEL);
 		if (!addrs) {
 			ret = -ENOMEM;
 			goto free_pages;
@@ -268,7 +279,7 @@ static int omap_gem_attach_pages(struct drm_gem_object *obj)
 			}
 		}
 	} else {
-		addrs = kzalloc(npages * sizeof(*addrs), GFP_KERNEL);
+		addrs = kcalloc(npages, sizeof(*addrs), GFP_KERNEL);
 		if (!addrs) {
 			ret = -ENOMEM;
 			goto free_pages;
@@ -288,35 +299,15 @@ free_pages:
 	return ret;
 }
 
-/* acquire pages when needed (for example, for DMA where physically
- * contiguous buffer is not required
- */
-static int get_pages(struct drm_gem_object *obj, struct page ***pages)
-{
-	struct omap_gem_object *omap_obj = to_omap_bo(obj);
-	int ret = 0;
-
-	if ((omap_obj->flags & OMAP_BO_MEM_SHMEM) && !omap_obj->pages) {
-		ret = omap_gem_attach_pages(obj);
-		if (ret) {
-			dev_err(obj->dev->dev, "could not attach pages\n");
-			return ret;
-		}
-	}
-
-	/* TODO: even phys-contig.. we should have a list of pages? */
-	*pages = omap_obj->pages;
-
-	return 0;
-}
-
-/** release backing pages */
+/* Release backing pages. Must be called with the omap_obj.lock held. */
 static void omap_gem_detach_pages(struct drm_gem_object *obj)
 {
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
 	unsigned int npages = obj->size >> PAGE_SHIFT;
 	unsigned int i;
 
+	lockdep_assert_held(&omap_obj->lock);
+
 	for (i = 0; i < npages; i++) {
 		if (omap_obj->dma_addrs[i])
 			dma_unmap_page(obj->dev->dev, omap_obj->dma_addrs[i],
@@ -336,16 +327,6 @@ u32 omap_gem_flags(struct drm_gem_object *obj)
 	return to_omap_bo(obj)->flags;
 }
 
-u64 omap_gem_mmap_offset(struct drm_gem_object *obj)
-{
-	u64 offset;
-
-	mutex_lock(&obj->dev->struct_mutex);
-	offset = mmap_offset(obj);
-	mutex_unlock(&obj->dev->struct_mutex);
-	return offset;
-}
-
 /** get mmap size */
 size_t omap_gem_mmap_size(struct drm_gem_object *obj)
 {
@@ -371,7 +352,7 @@ size_t omap_gem_mmap_size(struct drm_gem_object *obj)
  */
 
 /* Normal handling for the case of faulting in non-tiled buffers */
-static int fault_1d(struct drm_gem_object *obj,
+static vm_fault_t omap_gem_fault_1d(struct drm_gem_object *obj,
 		struct vm_area_struct *vma, struct vm_fault *vmf)
 {
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
@@ -385,18 +366,19 @@ static int fault_1d(struct drm_gem_object *obj,
 		omap_gem_cpu_sync_page(obj, pgoff);
 		pfn = page_to_pfn(omap_obj->pages[pgoff]);
 	} else {
-		BUG_ON(!is_contiguous(omap_obj));
+		BUG_ON(!omap_gem_is_contiguous(omap_obj));
 		pfn = (omap_obj->dma_addr >> PAGE_SHIFT) + pgoff;
 	}
 
 	VERB("Inserting %p pfn %lx, pa %lx", (void *)vmf->address,
 			pfn, pfn << PAGE_SHIFT);
 
-	return vm_insert_mixed(vma, vmf->address, __pfn_to_pfn_t(pfn, PFN_DEV));
+	return vmf_insert_mixed(vma, vmf->address,
+			__pfn_to_pfn_t(pfn, PFN_DEV));
 }
 
 /* Special handling for the case of faulting in 2d tiled buffers */
-static int fault_2d(struct drm_gem_object *obj,
+static vm_fault_t omap_gem_fault_2d(struct drm_gem_object *obj,
 		struct vm_area_struct *vma, struct vm_fault *vmf)
 {
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
@@ -407,7 +389,8 @@ static int fault_2d(struct drm_gem_object *obj,
 	unsigned long pfn;
 	pgoff_t pgoff, base_pgoff;
 	unsigned long vaddr;
-	int i, ret, slots;
+	int i, err, slots;
+	vm_fault_t ret = VM_FAULT_NOPAGE;
 
 	/*
 	 * Note the height of the slot is also equal to the number of pages
@@ -443,7 +426,7 @@ static int fault_2d(struct drm_gem_object *obj,
 
 	/* evict previous buffer using this usergart entry, if any: */
 	if (entry->obj)
-		evict_entry(entry->obj, fmt, entry);
+		omap_gem_evict_entry(entry->obj, fmt, entry);
 
 	entry->obj = obj;
 	entry->obj_pgoff = base_pgoff;
@@ -473,9 +456,10 @@ static int fault_2d(struct drm_gem_object *obj,
 	memset(pages + slots, 0,
 			sizeof(struct page *) * (n - slots));
 
-	ret = tiler_pin(entry->block, pages, ARRAY_SIZE(pages), 0, true);
-	if (ret) {
-		dev_err(obj->dev->dev, "failed to pin: %d\n", ret);
+	err = tiler_pin(entry->block, pages, ARRAY_SIZE(pages), 0, true);
+	if (err) {
+		ret = vmf_error(err);
+		dev_err(obj->dev->dev, "failed to pin: %d\n", err);
 		return ret;
 	}
 
@@ -485,7 +469,10 @@ static int fault_2d(struct drm_gem_object *obj,
 			pfn, pfn << PAGE_SHIFT);
 
 	for (i = n; i > 0; i--) {
-		vm_insert_mixed(vma, vaddr, __pfn_to_pfn_t(pfn, PFN_DEV));
+		ret = vmf_insert_mixed(vma,
+			vaddr, __pfn_to_pfn_t(pfn, PFN_DEV));
+		if (ret & VM_FAULT_ERROR)
+			break;
 		pfn += priv->usergart[fmt].stride_pfn;
 		vaddr += PAGE_SIZE * m;
 	}
@@ -494,7 +481,7 @@ static int fault_2d(struct drm_gem_object *obj,
 	priv->usergart[fmt].last = (priv->usergart[fmt].last + 1)
 				 % NUM_USERGART_ENTRIES;
 
-	return 0;
+	return ret;
 }
 
 /**
@@ -509,24 +496,25 @@ static int fault_2d(struct drm_gem_object *obj,
  * vma->vm_private_data points to the GEM object that is backing this
  * mapping.
  */
-int omap_gem_fault(struct vm_fault *vmf)
+vm_fault_t omap_gem_fault(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 	struct drm_gem_object *obj = vma->vm_private_data;
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
-	struct drm_device *dev = obj->dev;
-	struct page **pages;
-	int ret;
+	int err;
+	vm_fault_t ret;
 
 	/* Make sure we don't parallel update on a fault, nor move or remove
 	 * something from beneath our feet
 	 */
-	mutex_lock(&dev->struct_mutex);
+	mutex_lock(&omap_obj->lock);
 
 	/* if a shmem backed object, make sure we have pages attached now */
-	ret = get_pages(obj, &pages);
-	if (ret)
+	err = omap_gem_attach_pages(obj);
+	if (err) {
+		ret = vmf_error(err);
 		goto fail;
+	}
 
 	/* where should we do corresponding put_pages().. we are mapping
 	 * the original page, rather than thru a GART, so we can't rely
@@ -535,28 +523,14 @@ int omap_gem_fault(struct vm_fault *vmf)
 	 */
 
 	if (omap_obj->flags & OMAP_BO_TILED)
-		ret = fault_2d(obj, vma, vmf);
+		ret = omap_gem_fault_2d(obj, vma, vmf);
 	else
-		ret = fault_1d(obj, vma, vmf);
+		ret = omap_gem_fault_1d(obj, vma, vmf);
 
 
 fail:
-	mutex_unlock(&dev->struct_mutex);
-	switch (ret) {
-	case 0:
-	case -ERESTARTSYS:
-	case -EINTR:
-	case -EBUSY:
-		/*
-		 * EBUSY is ok: this just means that another thread
-		 * already did the job.
-		 */
-		return VM_FAULT_NOPAGE;
-	case -ENOMEM:
-		return VM_FAULT_OOM;
-	default:
-		return VM_FAULT_SIGBUS;
-	}
+	mutex_unlock(&omap_obj->lock);
+	return ret;
 }
 
 /** We override mainly to fix up some of the vm mapping flags.. */
@@ -689,21 +663,22 @@ int omap_gem_roll(struct drm_gem_object *obj, u32 roll)
 
 	omap_obj->roll = roll;
 
-	mutex_lock(&obj->dev->struct_mutex);
+	mutex_lock(&omap_obj->lock);
 
 	/* if we aren't mapped yet, we don't need to do anything */
 	if (omap_obj->block) {
-		struct page **pages;
-		ret = get_pages(obj, &pages);
+		ret = omap_gem_attach_pages(obj);
 		if (ret)
 			goto fail;
-		ret = tiler_pin(omap_obj->block, pages, npages, roll, true);
+
+		ret = tiler_pin(omap_obj->block, omap_obj->pages, npages,
+				roll, true);
 		if (ret)
 			dev_err(obj->dev->dev, "could not repin: %d\n", ret);
 	}
 
 fail:
-	mutex_unlock(&obj->dev->struct_mutex);
+	mutex_unlock(&omap_obj->lock);
 
 	return ret;
 }
@@ -722,7 +697,7 @@ fail:
  * the omap_obj->dma_addrs[i] is set to the DMA address, and the page is
  * unmapped from the CPU.
  */
-static inline bool is_cached_coherent(struct drm_gem_object *obj)
+static inline bool omap_gem_is_cached_coherent(struct drm_gem_object *obj)
 {
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
 
@@ -738,7 +713,7 @@ void omap_gem_cpu_sync_page(struct drm_gem_object *obj, int pgoff)
 	struct drm_device *dev = obj->dev;
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
 
-	if (is_cached_coherent(obj))
+	if (omap_gem_is_cached_coherent(obj))
 		return;
 
 	if (omap_obj->dma_addrs[pgoff]) {
@@ -758,7 +733,7 @@ void omap_gem_dma_sync_buffer(struct drm_gem_object *obj,
 	struct page **pages = omap_obj->pages;
 	bool dirty = false;
 
-	if (is_cached_coherent(obj))
+	if (omap_gem_is_cached_coherent(obj))
 		return;
 
 	for (i = 0; i < npages; i++) {
@@ -804,18 +779,17 @@ int omap_gem_pin(struct drm_gem_object *obj, dma_addr_t *dma_addr)
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
 	int ret = 0;
 
-	mutex_lock(&obj->dev->struct_mutex);
+	mutex_lock(&omap_obj->lock);
 
-	if (!is_contiguous(omap_obj) && priv->has_dmm) {
+	if (!omap_gem_is_contiguous(omap_obj) && priv->has_dmm) {
 		if (omap_obj->dma_addr_cnt == 0) {
-			struct page **pages;
 			u32 npages = obj->size >> PAGE_SHIFT;
 			enum tiler_fmt fmt = gem2fmt(omap_obj->flags);
 			struct tiler_block *block;
 
 			BUG_ON(omap_obj->block);
 
-			ret = get_pages(obj, &pages);
+			ret = omap_gem_attach_pages(obj);
 			if (ret)
 				goto fail;
 
@@ -835,7 +809,7 @@ int omap_gem_pin(struct drm_gem_object *obj, dma_addr_t *dma_addr)
 			}
 
 			/* TODO: enable async refill.. */
-			ret = tiler_pin(block, pages, npages,
+			ret = tiler_pin(block, omap_obj->pages, npages,
 					omap_obj->roll, true);
 			if (ret) {
 				tiler_release(block);
@@ -853,7 +827,7 @@ int omap_gem_pin(struct drm_gem_object *obj, dma_addr_t *dma_addr)
 		omap_obj->dma_addr_cnt++;
 
 		*dma_addr = omap_obj->dma_addr;
-	} else if (is_contiguous(omap_obj)) {
+	} else if (omap_gem_is_contiguous(omap_obj)) {
 		*dma_addr = omap_obj->dma_addr;
 	} else {
 		ret = -EINVAL;
@@ -861,7 +835,7 @@ int omap_gem_pin(struct drm_gem_object *obj, dma_addr_t *dma_addr)
 	}
 
 fail:
-	mutex_unlock(&obj->dev->struct_mutex);
+	mutex_unlock(&omap_obj->lock);
 
 	return ret;
 }
@@ -879,7 +853,8 @@ void omap_gem_unpin(struct drm_gem_object *obj)
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
 	int ret;
 
-	mutex_lock(&obj->dev->struct_mutex);
+	mutex_lock(&omap_obj->lock);
+
 	if (omap_obj->dma_addr_cnt > 0) {
 		omap_obj->dma_addr_cnt--;
 		if (omap_obj->dma_addr_cnt == 0) {
@@ -898,7 +873,7 @@ void omap_gem_unpin(struct drm_gem_object *obj)
 		}
 	}
 
-	mutex_unlock(&obj->dev->struct_mutex);
+	mutex_unlock(&omap_obj->lock);
 }
 
 /* Get rotated scanout address (only valid if already pinned), at the
@@ -911,13 +886,16 @@ int omap_gem_rotated_dma_addr(struct drm_gem_object *obj, u32 orient,
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
 	int ret = -EINVAL;
 
-	mutex_lock(&obj->dev->struct_mutex);
+	mutex_lock(&omap_obj->lock);
+
 	if ((omap_obj->dma_addr_cnt > 0) && omap_obj->block &&
 			(omap_obj->flags & OMAP_BO_TILED)) {
 		*dma_addr = tiler_tsptr(omap_obj->block, orient, x, y);
 		ret = 0;
 	}
-	mutex_unlock(&obj->dev->struct_mutex);
+
+	mutex_unlock(&omap_obj->lock);
+
 	return ret;
 }
 
@@ -944,17 +922,27 @@ int omap_gem_tiled_stride(struct drm_gem_object *obj, u32 orient)
 int omap_gem_get_pages(struct drm_gem_object *obj, struct page ***pages,
 		bool remap)
 {
-	int ret;
-	if (!remap) {
-		struct omap_gem_object *omap_obj = to_omap_bo(obj);
-		if (!omap_obj->pages)
-			return -ENOMEM;
-		*pages = omap_obj->pages;
-		return 0;
+	struct omap_gem_object *omap_obj = to_omap_bo(obj);
+	int ret = 0;
+
+	mutex_lock(&omap_obj->lock);
+
+	if (remap) {
+		ret = omap_gem_attach_pages(obj);
+		if (ret)
+			goto unlock;
 	}
-	mutex_lock(&obj->dev->struct_mutex);
-	ret = get_pages(obj, pages);
-	mutex_unlock(&obj->dev->struct_mutex);
+
+	if (!omap_obj->pages) {
+		ret = -ENOMEM;
+		goto unlock;
+	}
+
+	*pages = omap_obj->pages;
+
+unlock:
+	mutex_unlock(&omap_obj->lock);
+
 	return ret;
 }
 
@@ -969,23 +957,34 @@ int omap_gem_put_pages(struct drm_gem_object *obj)
 }
 
 #ifdef CONFIG_DRM_FBDEV_EMULATION
-/* Get kernel virtual address for CPU access.. this more or less only
- * exists for omap_fbdev.  This should be called with struct_mutex
- * held.
+/*
+ * Get kernel virtual address for CPU access.. this more or less only
+ * exists for omap_fbdev.
  */
 void *omap_gem_vaddr(struct drm_gem_object *obj)
 {
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
-	WARN_ON(!mutex_is_locked(&obj->dev->struct_mutex));
+	void *vaddr;
+	int ret;
+
+	mutex_lock(&omap_obj->lock);
+
 	if (!omap_obj->vaddr) {
-		struct page **pages;
-		int ret = get_pages(obj, &pages);
-		if (ret)
-			return ERR_PTR(ret);
-		omap_obj->vaddr = vmap(pages, obj->size >> PAGE_SHIFT,
+		ret = omap_gem_attach_pages(obj);
+		if (ret) {
+			vaddr = ERR_PTR(ret);
+			goto unlock;
+		}
+
+		omap_obj->vaddr = vmap(omap_obj->pages, obj->size >> PAGE_SHIFT,
 				VM_MAP, pgprot_writecombine(PAGE_KERNEL));
 	}
-	return omap_obj->vaddr;
+
+	vaddr = omap_obj->vaddr;
+
+unlock:
+	mutex_unlock(&omap_obj->lock);
+	return vaddr;
 }
 #endif
 
@@ -1001,6 +1000,7 @@ int omap_gem_resume(struct drm_device *dev)
 	struct omap_gem_object *omap_obj;
 	int ret = 0;
 
+	mutex_lock(&priv->list_lock);
 	list_for_each_entry(omap_obj, &priv->obj_list, mm_list) {
 		if (omap_obj->block) {
 			struct drm_gem_object *obj = &omap_obj->base;
@@ -1012,12 +1012,14 @@ int omap_gem_resume(struct drm_device *dev)
 					omap_obj->roll, true);
 			if (ret) {
 				dev_err(dev->dev, "could not repin: %d\n", ret);
-				return ret;
+				goto done;
 			}
 		}
 	}
 
-	return 0;
+done:
+	mutex_unlock(&priv->list_lock);
+	return ret;
 }
 #endif
 
@@ -1033,6 +1035,8 @@ void omap_gem_describe(struct drm_gem_object *obj, struct seq_file *m)
 
 	off = drm_vma_node_start(&obj->vma_node);
 
+	mutex_lock(&omap_obj->lock);
+
 	seq_printf(m, "%08x: %2d (%2d) %08llx %pad (%2d) %p %4d",
 			omap_obj->flags, obj->name, kref_read(&obj->refcount),
 			off, &omap_obj->dma_addr, omap_obj->dma_addr_cnt,
@@ -1050,6 +1054,8 @@ void omap_gem_describe(struct drm_gem_object *obj, struct seq_file *m)
 		seq_printf(m, " %zu", obj->size);
 	}
 
+	mutex_unlock(&omap_obj->lock);
+
 	seq_printf(m, "\n");
 }
 
@@ -1081,17 +1087,21 @@ void omap_gem_free_object(struct drm_gem_object *obj)
 	struct omap_drm_private *priv = dev->dev_private;
 	struct omap_gem_object *omap_obj = to_omap_bo(obj);
 
-	evict(obj);
-
-	WARN_ON(!mutex_is_locked(&dev->struct_mutex));
+	omap_gem_evict(obj);
 
-	spin_lock(&priv->list_lock);
+	mutex_lock(&priv->list_lock);
 	list_del(&omap_obj->mm_list);
-	spin_unlock(&priv->list_lock);
+	mutex_unlock(&priv->list_lock);
 
-	/* this means the object is still pinned.. which really should
-	 * not happen.  I think..
+	/*
+	 * We own the sole reference to the object at this point, but to keep
+	 * lockdep happy, we must still take the omap_obj_lock to call
+	 * omap_gem_detach_pages(). This should hardly make any difference as
+	 * there can't be any lock contention.
 	 */
+	mutex_lock(&omap_obj->lock);
+
+	/* The object should not be pinned. */
 	WARN_ON(omap_obj->dma_addr_cnt > 0);
 
 	if (omap_obj->pages) {
@@ -1110,8 +1120,12 @@ void omap_gem_free_object(struct drm_gem_object *obj)
 		drm_prime_gem_destroy(obj, omap_obj->sgt);
 	}
 
+	mutex_unlock(&omap_obj->lock);
+
 	drm_gem_object_release(obj);
 
+	mutex_destroy(&omap_obj->lock);
+
 	kfree(omap_obj);
 }
 
@@ -1167,6 +1181,7 @@ struct drm_gem_object *omap_gem_new(struct drm_device *dev,
 
 	obj = &omap_obj->base;
 	omap_obj->flags = flags;
+	mutex_init(&omap_obj->lock);
 
 	if (flags & OMAP_BO_TILED) {
 		/*
@@ -1206,9 +1221,9 @@ struct drm_gem_object *omap_gem_new(struct drm_device *dev,
 			goto err_release;
 	}
 
-	spin_lock(&priv->list_lock);
+	mutex_lock(&priv->list_lock);
 	list_add(&omap_obj->mm_list, &priv->obj_list);
-	spin_unlock(&priv->list_lock);
+	mutex_unlock(&priv->list_lock);
 
 	return obj;
 
@@ -1231,16 +1246,15 @@ struct drm_gem_object *omap_gem_new_dmabuf(struct drm_device *dev, size_t size,
 	if (sgt->orig_nents != 1 && !priv->has_dmm)
 		return ERR_PTR(-EINVAL);
 
-	mutex_lock(&dev->struct_mutex);
-
 	gsize.bytes = PAGE_ALIGN(size);
 	obj = omap_gem_new(dev, gsize, OMAP_BO_MEM_DMABUF | OMAP_BO_WC);
-	if (!obj) {
-		obj = ERR_PTR(-ENOMEM);
-		goto done;
-	}
+	if (!obj)
+		return ERR_PTR(-ENOMEM);
 
 	omap_obj = to_omap_bo(obj);
+
+	mutex_lock(&omap_obj->lock);
+
 	omap_obj->sgt = sgt;
 
 	if (sgt->orig_nents == 1) {
@@ -1276,7 +1290,7 @@ struct drm_gem_object *omap_gem_new_dmabuf(struct drm_device *dev, size_t size,
 	}
 
 done:
-	mutex_unlock(&dev->struct_mutex);
+	mutex_unlock(&omap_obj->lock);
 	return obj;
 }
 
diff --git a/drivers/gpu/drm/omapdrm/omap_gem.h b/drivers/gpu/drm/omapdrm/omap_gem.h
index a78bde05193a..c1c45fbde155 100644
--- a/drivers/gpu/drm/omapdrm/omap_gem.h
+++ b/drivers/gpu/drm/omapdrm/omap_gem.h
@@ -21,6 +21,7 @@
 #define __OMAPDRM_GEM_H__
 
 #include <linux/types.h>
+#include <linux/mm_types.h>
 
 enum dma_data_direction;
 
@@ -80,7 +81,7 @@ struct dma_buf *omap_gem_prime_export(struct drm_device *dev,
 struct drm_gem_object *omap_gem_prime_import(struct drm_device *dev,
 		struct dma_buf *buffer);
 
-int omap_gem_fault(struct vm_fault *vmf);
+vm_fault_t omap_gem_fault(struct vm_fault *vmf);
 int omap_gem_roll(struct drm_gem_object *obj, u32 roll);
 void omap_gem_cpu_sync_page(struct drm_gem_object *obj, int pgoff);
 void omap_gem_dma_sync_buffer(struct drm_gem_object *obj,
diff --git a/drivers/gpu/drm/omapdrm/omap_gem_dmabuf.c b/drivers/gpu/drm/omapdrm/omap_gem_dmabuf.c
index 8e41d649e248..ec04a69ade46 100644
--- a/drivers/gpu/drm/omapdrm/omap_gem_dmabuf.c
+++ b/drivers/gpu/drm/omapdrm/omap_gem_dmabuf.c
@@ -93,23 +93,6 @@ static int omap_gem_dmabuf_end_cpu_access(struct dma_buf *buffer,
 	return 0;
 }
 
-
-static void *omap_gem_dmabuf_kmap_atomic(struct dma_buf *buffer,
-		unsigned long page_num)
-{
-	struct drm_gem_object *obj = buffer->priv;
-	struct page **pages;
-	omap_gem_get_pages(obj, &pages, false);
-	omap_gem_cpu_sync_page(obj, page_num);
-	return kmap_atomic(pages[page_num]);
-}
-
-static void omap_gem_dmabuf_kunmap_atomic(struct dma_buf *buffer,
-		unsigned long page_num, void *addr)
-{
-	kunmap_atomic(addr);
-}
-
 static void *omap_gem_dmabuf_kmap(struct dma_buf *buffer,
 		unsigned long page_num)
 {
@@ -148,8 +131,6 @@ static const struct dma_buf_ops omap_dmabuf_ops = {
 	.release = drm_gem_dmabuf_release,
 	.begin_cpu_access = omap_gem_dmabuf_begin_cpu_access,
 	.end_cpu_access = omap_gem_dmabuf_end_cpu_access,
-	.map_atomic = omap_gem_dmabuf_kmap_atomic,
-	.unmap_atomic = omap_gem_dmabuf_kunmap_atomic,
 	.map = omap_gem_dmabuf_kmap,
 	.unmap = omap_gem_dmabuf_kunmap,
 	.mmap = omap_gem_dmabuf_mmap,
diff --git a/drivers/gpu/drm/omapdrm/tcm-sita.c b/drivers/gpu/drm/omapdrm/tcm-sita.c
index d7f7bc9f061a..817be3c41863 100644
--- a/drivers/gpu/drm/omapdrm/tcm-sita.c
+++ b/drivers/gpu/drm/omapdrm/tcm-sita.c
@@ -90,7 +90,7 @@ static int l2r_t2b(u16 w, u16 h, u16 a, s16 offset,
 {
 	int i;
 	unsigned long index;
-	bool area_free;
+	bool area_free = false;
 	unsigned long slots_per_band = PAGE_SIZE / slot_bytes;
 	unsigned long bit_offset = (offset > 0) ? offset / slot_bytes : 0;
 	unsigned long curr_bit = bit_offset;
diff --git a/drivers/gpu/drm/panel/Kconfig b/drivers/gpu/drm/panel/Kconfig
index 25682ff3449a..6020c30a33b3 100644
--- a/drivers/gpu/drm/panel/Kconfig
+++ b/drivers/gpu/drm/panel/Kconfig
@@ -46,6 +46,15 @@ config DRM_PANEL_ILITEK_IL9322
 	  Say Y here if you want to enable support for Ilitek IL9322
 	  QVGA (320x240) RGB, YUV and ITU-T BT.656 panels.
 
+config DRM_PANEL_ILITEK_ILI9881C
+	tristate "Ilitek ILI9881C-based panels"
+	depends on OF
+	depends on DRM_MIPI_DSI
+	depends on BACKLIGHT_CLASS_DEVICE
+	help
+	  Say Y if you want to enable support for panels based on the
+	  Ilitek ILI9881c controller.
+
 config DRM_PANEL_INNOLUX_P079ZCA
 	tristate "Innolux P079ZCA panel"
 	depends on OF
diff --git a/drivers/gpu/drm/panel/Makefile b/drivers/gpu/drm/panel/Makefile
index f26efc11d746..5ccaaa9d13af 100644
--- a/drivers/gpu/drm/panel/Makefile
+++ b/drivers/gpu/drm/panel/Makefile
@@ -3,6 +3,7 @@ obj-$(CONFIG_DRM_PANEL_ARM_VERSATILE) += panel-arm-versatile.o
 obj-$(CONFIG_DRM_PANEL_LVDS) += panel-lvds.o
 obj-$(CONFIG_DRM_PANEL_SIMPLE) += panel-simple.o
 obj-$(CONFIG_DRM_PANEL_ILITEK_IL9322) += panel-ilitek-ili9322.o
+obj-$(CONFIG_DRM_PANEL_ILITEK_ILI9881C) += panel-ilitek-ili9881c.o
 obj-$(CONFIG_DRM_PANEL_INNOLUX_P079ZCA) += panel-innolux-p079zca.o
 obj-$(CONFIG_DRM_PANEL_JDI_LT070ME05000) += panel-jdi-lt070me05000.o
 obj-$(CONFIG_DRM_PANEL_LG_LG4573) += panel-lg-lg4573.o
diff --git a/drivers/gpu/drm/panel/panel-ilitek-ili9881c.c b/drivers/gpu/drm/panel/panel-ilitek-ili9881c.c
new file mode 100644
index 000000000000..3ad4a46c4e94
--- /dev/null
+++ b/drivers/gpu/drm/panel/panel-ilitek-ili9881c.c
@@ -0,0 +1,503 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2017-2018, Bootlin
+ */
+
+#include <linux/backlight.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/err.h>
+#include <linux/errno.h>
+#include <linux/fb.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+
+#include <linux/gpio/consumer.h>
+#include <linux/regulator/consumer.h>
+
+#include <drm/drm_mipi_dsi.h>
+#include <drm/drm_modes.h>
+#include <drm/drm_panel.h>
+
+#include <video/mipi_display.h>
+
+struct ili9881c {
+	struct drm_panel	panel;
+	struct mipi_dsi_device	*dsi;
+
+	struct backlight_device *backlight;
+	struct regulator	*power;
+	struct gpio_desc	*reset;
+};
+
+enum ili9881c_op {
+	ILI9881C_SWITCH_PAGE,
+	ILI9881C_COMMAND,
+};
+
+struct ili9881c_instr {
+	enum ili9881c_op	op;
+
+	union arg {
+		struct cmd {
+			u8	cmd;
+			u8	data;
+		} cmd;
+		u8	page;
+	} arg;
+};
+
+#define ILI9881C_SWITCH_PAGE_INSTR(_page)	\
+	{					\
+		.op = ILI9881C_SWITCH_PAGE,	\
+		.arg = {			\
+			.page = (_page),	\
+		},				\
+	}
+
+#define ILI9881C_COMMAND_INSTR(_cmd, _data)		\
+	{						\
+		.op = ILI9881C_COMMAND,		\
+		.arg = {				\
+			.cmd = {			\
+				.cmd = (_cmd),		\
+				.data = (_data),	\
+			},				\
+		},					\
+	}
+
+static const struct ili9881c_instr ili9881c_init[] = {
+	ILI9881C_SWITCH_PAGE_INSTR(3),
+	ILI9881C_COMMAND_INSTR(0x01, 0x00),
+	ILI9881C_COMMAND_INSTR(0x02, 0x00),
+	ILI9881C_COMMAND_INSTR(0x03, 0x73),
+	ILI9881C_COMMAND_INSTR(0x04, 0x03),
+	ILI9881C_COMMAND_INSTR(0x05, 0x00),
+	ILI9881C_COMMAND_INSTR(0x06, 0x06),
+	ILI9881C_COMMAND_INSTR(0x07, 0x06),
+	ILI9881C_COMMAND_INSTR(0x08, 0x00),
+	ILI9881C_COMMAND_INSTR(0x09, 0x18),
+	ILI9881C_COMMAND_INSTR(0x0a, 0x04),
+	ILI9881C_COMMAND_INSTR(0x0b, 0x00),
+	ILI9881C_COMMAND_INSTR(0x0c, 0x02),
+	ILI9881C_COMMAND_INSTR(0x0d, 0x03),
+	ILI9881C_COMMAND_INSTR(0x0e, 0x00),
+	ILI9881C_COMMAND_INSTR(0x0f, 0x25),
+	ILI9881C_COMMAND_INSTR(0x10, 0x25),
+	ILI9881C_COMMAND_INSTR(0x11, 0x00),
+	ILI9881C_COMMAND_INSTR(0x12, 0x00),
+	ILI9881C_COMMAND_INSTR(0x13, 0x00),
+	ILI9881C_COMMAND_INSTR(0x14, 0x00),
+	ILI9881C_COMMAND_INSTR(0x15, 0x00),
+	ILI9881C_COMMAND_INSTR(0x16, 0x0C),
+	ILI9881C_COMMAND_INSTR(0x17, 0x00),
+	ILI9881C_COMMAND_INSTR(0x18, 0x00),
+	ILI9881C_COMMAND_INSTR(0x19, 0x00),
+	ILI9881C_COMMAND_INSTR(0x1a, 0x00),
+	ILI9881C_COMMAND_INSTR(0x1b, 0x00),
+	ILI9881C_COMMAND_INSTR(0x1c, 0x00),
+	ILI9881C_COMMAND_INSTR(0x1d, 0x00),
+	ILI9881C_COMMAND_INSTR(0x1e, 0xC0),
+	ILI9881C_COMMAND_INSTR(0x1f, 0x80),
+	ILI9881C_COMMAND_INSTR(0x20, 0x04),
+	ILI9881C_COMMAND_INSTR(0x21, 0x01),
+	ILI9881C_COMMAND_INSTR(0x22, 0x00),
+	ILI9881C_COMMAND_INSTR(0x23, 0x00),
+	ILI9881C_COMMAND_INSTR(0x24, 0x00),
+	ILI9881C_COMMAND_INSTR(0x25, 0x00),
+	ILI9881C_COMMAND_INSTR(0x26, 0x00),
+	ILI9881C_COMMAND_INSTR(0x27, 0x00),
+	ILI9881C_COMMAND_INSTR(0x28, 0x33),
+	ILI9881C_COMMAND_INSTR(0x29, 0x03),
+	ILI9881C_COMMAND_INSTR(0x2a, 0x00),
+	ILI9881C_COMMAND_INSTR(0x2b, 0x00),
+	ILI9881C_COMMAND_INSTR(0x2c, 0x00),
+	ILI9881C_COMMAND_INSTR(0x2d, 0x00),
+	ILI9881C_COMMAND_INSTR(0x2e, 0x00),
+	ILI9881C_COMMAND_INSTR(0x2f, 0x00),
+	ILI9881C_COMMAND_INSTR(0x30, 0x00),
+	ILI9881C_COMMAND_INSTR(0x31, 0x00),
+	ILI9881C_COMMAND_INSTR(0x32, 0x00),
+	ILI9881C_COMMAND_INSTR(0x33, 0x00),
+	ILI9881C_COMMAND_INSTR(0x34, 0x04),
+	ILI9881C_COMMAND_INSTR(0x35, 0x00),
+	ILI9881C_COMMAND_INSTR(0x36, 0x00),
+	ILI9881C_COMMAND_INSTR(0x37, 0x00),
+	ILI9881C_COMMAND_INSTR(0x38, 0x3C),
+	ILI9881C_COMMAND_INSTR(0x39, 0x00),
+	ILI9881C_COMMAND_INSTR(0x3a, 0x00),
+	ILI9881C_COMMAND_INSTR(0x3b, 0x00),
+	ILI9881C_COMMAND_INSTR(0x3c, 0x00),
+	ILI9881C_COMMAND_INSTR(0x3d, 0x00),
+	ILI9881C_COMMAND_INSTR(0x3e, 0x00),
+	ILI9881C_COMMAND_INSTR(0x3f, 0x00),
+	ILI9881C_COMMAND_INSTR(0x40, 0x00),
+	ILI9881C_COMMAND_INSTR(0x41, 0x00),
+	ILI9881C_COMMAND_INSTR(0x42, 0x00),
+	ILI9881C_COMMAND_INSTR(0x43, 0x00),
+	ILI9881C_COMMAND_INSTR(0x44, 0x00),
+	ILI9881C_COMMAND_INSTR(0x50, 0x01),
+	ILI9881C_COMMAND_INSTR(0x51, 0x23),
+	ILI9881C_COMMAND_INSTR(0x52, 0x45),
+	ILI9881C_COMMAND_INSTR(0x53, 0x67),
+	ILI9881C_COMMAND_INSTR(0x54, 0x89),
+	ILI9881C_COMMAND_INSTR(0x55, 0xab),
+	ILI9881C_COMMAND_INSTR(0x56, 0x01),
+	ILI9881C_COMMAND_INSTR(0x57, 0x23),
+	ILI9881C_COMMAND_INSTR(0x58, 0x45),
+	ILI9881C_COMMAND_INSTR(0x59, 0x67),
+	ILI9881C_COMMAND_INSTR(0x5a, 0x89),
+	ILI9881C_COMMAND_INSTR(0x5b, 0xab),
+	ILI9881C_COMMAND_INSTR(0x5c, 0xcd),
+	ILI9881C_COMMAND_INSTR(0x5d, 0xef),
+	ILI9881C_COMMAND_INSTR(0x5e, 0x11),
+	ILI9881C_COMMAND_INSTR(0x5f, 0x02),
+	ILI9881C_COMMAND_INSTR(0x60, 0x02),
+	ILI9881C_COMMAND_INSTR(0x61, 0x02),
+	ILI9881C_COMMAND_INSTR(0x62, 0x02),
+	ILI9881C_COMMAND_INSTR(0x63, 0x02),
+	ILI9881C_COMMAND_INSTR(0x64, 0x02),
+	ILI9881C_COMMAND_INSTR(0x65, 0x02),
+	ILI9881C_COMMAND_INSTR(0x66, 0x02),
+	ILI9881C_COMMAND_INSTR(0x67, 0x02),
+	ILI9881C_COMMAND_INSTR(0x68, 0x02),
+	ILI9881C_COMMAND_INSTR(0x69, 0x02),
+	ILI9881C_COMMAND_INSTR(0x6a, 0x0C),
+	ILI9881C_COMMAND_INSTR(0x6b, 0x02),
+	ILI9881C_COMMAND_INSTR(0x6c, 0x0F),
+	ILI9881C_COMMAND_INSTR(0x6d, 0x0E),
+	ILI9881C_COMMAND_INSTR(0x6e, 0x0D),
+	ILI9881C_COMMAND_INSTR(0x6f, 0x06),
+	ILI9881C_COMMAND_INSTR(0x70, 0x07),
+	ILI9881C_COMMAND_INSTR(0x71, 0x02),
+	ILI9881C_COMMAND_INSTR(0x72, 0x02),
+	ILI9881C_COMMAND_INSTR(0x73, 0x02),
+	ILI9881C_COMMAND_INSTR(0x74, 0x02),
+	ILI9881C_COMMAND_INSTR(0x75, 0x02),
+	ILI9881C_COMMAND_INSTR(0x76, 0x02),
+	ILI9881C_COMMAND_INSTR(0x77, 0x02),
+	ILI9881C_COMMAND_INSTR(0x78, 0x02),
+	ILI9881C_COMMAND_INSTR(0x79, 0x02),
+	ILI9881C_COMMAND_INSTR(0x7a, 0x02),
+	ILI9881C_COMMAND_INSTR(0x7b, 0x02),
+	ILI9881C_COMMAND_INSTR(0x7c, 0x02),
+	ILI9881C_COMMAND_INSTR(0x7d, 0x02),
+	ILI9881C_COMMAND_INSTR(0x7e, 0x02),
+	ILI9881C_COMMAND_INSTR(0x7f, 0x02),
+	ILI9881C_COMMAND_INSTR(0x80, 0x0C),
+	ILI9881C_COMMAND_INSTR(0x81, 0x02),
+	ILI9881C_COMMAND_INSTR(0x82, 0x0F),
+	ILI9881C_COMMAND_INSTR(0x83, 0x0E),
+	ILI9881C_COMMAND_INSTR(0x84, 0x0D),
+	ILI9881C_COMMAND_INSTR(0x85, 0x06),
+	ILI9881C_COMMAND_INSTR(0x86, 0x07),
+	ILI9881C_COMMAND_INSTR(0x87, 0x02),
+	ILI9881C_COMMAND_INSTR(0x88, 0x02),
+	ILI9881C_COMMAND_INSTR(0x89, 0x02),
+	ILI9881C_COMMAND_INSTR(0x8A, 0x02),
+	ILI9881C_SWITCH_PAGE_INSTR(4),
+	ILI9881C_COMMAND_INSTR(0x6C, 0x15),
+	ILI9881C_COMMAND_INSTR(0x6E, 0x22),
+	ILI9881C_COMMAND_INSTR(0x6F, 0x33),
+	ILI9881C_COMMAND_INSTR(0x3A, 0xA4),
+	ILI9881C_COMMAND_INSTR(0x8D, 0x0D),
+	ILI9881C_COMMAND_INSTR(0x87, 0xBA),
+	ILI9881C_COMMAND_INSTR(0x26, 0x76),
+	ILI9881C_COMMAND_INSTR(0xB2, 0xD1),
+	ILI9881C_SWITCH_PAGE_INSTR(1),
+	ILI9881C_COMMAND_INSTR(0x22, 0x0A),
+	ILI9881C_COMMAND_INSTR(0x53, 0xDC),
+	ILI9881C_COMMAND_INSTR(0x55, 0xA7),
+	ILI9881C_COMMAND_INSTR(0x50, 0x78),
+	ILI9881C_COMMAND_INSTR(0x51, 0x78),
+	ILI9881C_COMMAND_INSTR(0x31, 0x02),
+	ILI9881C_COMMAND_INSTR(0x60, 0x14),
+	ILI9881C_COMMAND_INSTR(0xA0, 0x2A),
+	ILI9881C_COMMAND_INSTR(0xA1, 0x39),
+	ILI9881C_COMMAND_INSTR(0xA2, 0x46),
+	ILI9881C_COMMAND_INSTR(0xA3, 0x0e),
+	ILI9881C_COMMAND_INSTR(0xA4, 0x12),
+	ILI9881C_COMMAND_INSTR(0xA5, 0x25),
+	ILI9881C_COMMAND_INSTR(0xA6, 0x19),
+	ILI9881C_COMMAND_INSTR(0xA7, 0x1d),
+	ILI9881C_COMMAND_INSTR(0xA8, 0xa6),
+	ILI9881C_COMMAND_INSTR(0xA9, 0x1C),
+	ILI9881C_COMMAND_INSTR(0xAA, 0x29),
+	ILI9881C_COMMAND_INSTR(0xAB, 0x85),
+	ILI9881C_COMMAND_INSTR(0xAC, 0x1C),
+	ILI9881C_COMMAND_INSTR(0xAD, 0x1B),
+	ILI9881C_COMMAND_INSTR(0xAE, 0x51),
+	ILI9881C_COMMAND_INSTR(0xAF, 0x22),
+	ILI9881C_COMMAND_INSTR(0xB0, 0x2d),
+	ILI9881C_COMMAND_INSTR(0xB1, 0x4f),
+	ILI9881C_COMMAND_INSTR(0xB2, 0x59),
+	ILI9881C_COMMAND_INSTR(0xB3, 0x3F),
+	ILI9881C_COMMAND_INSTR(0xC0, 0x2A),
+	ILI9881C_COMMAND_INSTR(0xC1, 0x3a),
+	ILI9881C_COMMAND_INSTR(0xC2, 0x45),
+	ILI9881C_COMMAND_INSTR(0xC3, 0x0e),
+	ILI9881C_COMMAND_INSTR(0xC4, 0x11),
+	ILI9881C_COMMAND_INSTR(0xC5, 0x24),
+	ILI9881C_COMMAND_INSTR(0xC6, 0x1a),
+	ILI9881C_COMMAND_INSTR(0xC7, 0x1c),
+	ILI9881C_COMMAND_INSTR(0xC8, 0xaa),
+	ILI9881C_COMMAND_INSTR(0xC9, 0x1C),
+	ILI9881C_COMMAND_INSTR(0xCA, 0x29),
+	ILI9881C_COMMAND_INSTR(0xCB, 0x96),
+	ILI9881C_COMMAND_INSTR(0xCC, 0x1C),
+	ILI9881C_COMMAND_INSTR(0xCD, 0x1B),
+	ILI9881C_COMMAND_INSTR(0xCE, 0x51),
+	ILI9881C_COMMAND_INSTR(0xCF, 0x22),
+	ILI9881C_COMMAND_INSTR(0xD0, 0x2b),
+	ILI9881C_COMMAND_INSTR(0xD1, 0x4b),
+	ILI9881C_COMMAND_INSTR(0xD2, 0x59),
+	ILI9881C_COMMAND_INSTR(0xD3, 0x3F),
+};
+
+static inline struct ili9881c *panel_to_ili9881c(struct drm_panel *panel)
+{
+	return container_of(panel, struct ili9881c, panel);
+}
+
+/*
+ * The panel seems to accept some private DCS commands that map
+ * directly to registers.
+ *
+ * It is organised by page, with each page having its own set of
+ * registers, and the first page looks like it's holding the standard
+ * DCS commands.
+ *
+ * So before any attempt at sending a command or data, we have to be
+ * sure if we're in the right page or not.
+ */
+static int ili9881c_switch_page(struct ili9881c *ctx, u8 page)
+{
+	u8 buf[4] = { 0xff, 0x98, 0x81, page };
+	int ret;
+
+	ret = mipi_dsi_dcs_write_buffer(ctx->dsi, buf, sizeof(buf));
+	if (ret < 0)
+		return ret;
+
+	return 0;
+}
+
+static int ili9881c_send_cmd_data(struct ili9881c *ctx, u8 cmd, u8 data)
+{
+	u8 buf[2] = { cmd, data };
+	int ret;
+
+	ret = mipi_dsi_dcs_write_buffer(ctx->dsi, buf, sizeof(buf));
+	if (ret < 0)
+		return ret;
+
+	return 0;
+}
+
+static int ili9881c_prepare(struct drm_panel *panel)
+{
+	struct ili9881c *ctx = panel_to_ili9881c(panel);
+	unsigned int i;
+	int ret;
+
+	/* Power the panel */
+	ret = regulator_enable(ctx->power);
+	if (ret)
+		return ret;
+	msleep(5);
+
+	/* And reset it */
+	gpiod_set_value(ctx->reset, 1);
+	msleep(20);
+
+	gpiod_set_value(ctx->reset, 0);
+	msleep(20);
+
+	for (i = 0; i < ARRAY_SIZE(ili9881c_init); i++) {
+		const struct ili9881c_instr *instr = &ili9881c_init[i];
+
+		if (instr->op == ILI9881C_SWITCH_PAGE)
+			ret = ili9881c_switch_page(ctx, instr->arg.page);
+		else if (instr->op == ILI9881C_COMMAND)
+			ret = ili9881c_send_cmd_data(ctx, instr->arg.cmd.cmd,
+						      instr->arg.cmd.data);
+
+		if (ret)
+			return ret;
+	}
+
+	ret = ili9881c_switch_page(ctx, 0);
+	if (ret)
+		return ret;
+
+	ret = mipi_dsi_dcs_set_tear_on(ctx->dsi, MIPI_DSI_DCS_TEAR_MODE_VBLANK);
+	if (ret)
+		return ret;
+
+	ret = mipi_dsi_dcs_exit_sleep_mode(ctx->dsi);
+	if (ret)
+		return ret;
+
+	return 0;
+}
+
+static int ili9881c_enable(struct drm_panel *panel)
+{
+	struct ili9881c *ctx = panel_to_ili9881c(panel);
+
+	msleep(120);
+
+	mipi_dsi_dcs_set_display_on(ctx->dsi);
+	backlight_enable(ctx->backlight);
+
+	return 0;
+}
+
+static int ili9881c_disable(struct drm_panel *panel)
+{
+	struct ili9881c *ctx = panel_to_ili9881c(panel);
+
+	backlight_disable(ctx->backlight);
+	return mipi_dsi_dcs_set_display_off(ctx->dsi);
+}
+
+static int ili9881c_unprepare(struct drm_panel *panel)
+{
+	struct ili9881c *ctx = panel_to_ili9881c(panel);
+
+	mipi_dsi_dcs_enter_sleep_mode(ctx->dsi);
+	regulator_disable(ctx->power);
+	gpiod_set_value(ctx->reset, 1);
+
+	return 0;
+}
+
+static const struct drm_display_mode bananapi_default_mode = {
+	.clock		= 62000,
+	.vrefresh	= 60,
+
+	.hdisplay	= 720,
+	.hsync_start	= 720 + 10,
+	.hsync_end	= 720 + 10 + 20,
+	.htotal		= 720 + 10 + 20 + 30,
+
+	.vdisplay	= 1280,
+	.vsync_start	= 1280 + 10,
+	.vsync_end	= 1280 + 10 + 10,
+	.vtotal		= 1280 + 10 + 10 + 20,
+};
+
+static int ili9881c_get_modes(struct drm_panel *panel)
+{
+	struct drm_connector *connector = panel->connector;
+	struct ili9881c *ctx = panel_to_ili9881c(panel);
+	struct drm_display_mode *mode;
+
+	mode = drm_mode_duplicate(panel->drm, &bananapi_default_mode);
+	if (!mode) {
+		dev_err(&ctx->dsi->dev, "failed to add mode %ux%ux@%u\n",
+			bananapi_default_mode.hdisplay,
+			bananapi_default_mode.vdisplay,
+			bananapi_default_mode.vrefresh);
+		return -ENOMEM;
+	}
+
+	drm_mode_set_name(mode);
+
+	mode->type = DRM_MODE_TYPE_DRIVER | DRM_MODE_TYPE_PREFERRED;
+	drm_mode_probed_add(connector, mode);
+
+	panel->connector->display_info.width_mm = 62;
+	panel->connector->display_info.height_mm = 110;
+
+	return 1;
+}
+
+static const struct drm_panel_funcs ili9881c_funcs = {
+	.prepare	= ili9881c_prepare,
+	.unprepare	= ili9881c_unprepare,
+	.enable		= ili9881c_enable,
+	.disable	= ili9881c_disable,
+	.get_modes	= ili9881c_get_modes,
+};
+
+static int ili9881c_dsi_probe(struct mipi_dsi_device *dsi)
+{
+	struct device_node *np;
+	struct ili9881c *ctx;
+	int ret;
+
+	ctx = devm_kzalloc(&dsi->dev, sizeof(*ctx), GFP_KERNEL);
+	if (!ctx)
+		return -ENOMEM;
+	mipi_dsi_set_drvdata(dsi, ctx);
+	ctx->dsi = dsi;
+
+	drm_panel_init(&ctx->panel);
+	ctx->panel.dev = &dsi->dev;
+	ctx->panel.funcs = &ili9881c_funcs;
+
+	ctx->power = devm_regulator_get(&dsi->dev, "power");
+	if (IS_ERR(ctx->power)) {
+		dev_err(&dsi->dev, "Couldn't get our power regulator\n");
+		return PTR_ERR(ctx->power);
+	}
+
+	ctx->reset = devm_gpiod_get(&dsi->dev, "reset", GPIOD_OUT_LOW);
+	if (IS_ERR(ctx->reset)) {
+		dev_err(&dsi->dev, "Couldn't get our reset GPIO\n");
+		return PTR_ERR(ctx->reset);
+	}
+
+	np = of_parse_phandle(dsi->dev.of_node, "backlight", 0);
+	if (np) {
+		ctx->backlight = of_find_backlight_by_node(np);
+		of_node_put(np);
+
+		if (!ctx->backlight)
+			return -EPROBE_DEFER;
+	}
+
+	ret = drm_panel_add(&ctx->panel);
+	if (ret < 0)
+		return ret;
+
+	dsi->mode_flags = MIPI_DSI_MODE_VIDEO_SYNC_PULSE;
+	dsi->format = MIPI_DSI_FMT_RGB888;
+	dsi->lanes = 4;
+
+	return mipi_dsi_attach(dsi);
+}
+
+static int ili9881c_dsi_remove(struct mipi_dsi_device *dsi)
+{
+	struct ili9881c *ctx = mipi_dsi_get_drvdata(dsi);
+
+	mipi_dsi_detach(dsi);
+	drm_panel_remove(&ctx->panel);
+
+	if (ctx->backlight)
+		put_device(&ctx->backlight->dev);
+
+	return 0;
+}
+
+static const struct of_device_id ili9881c_of_match[] = {
+	{ .compatible = "bananapi,lhr050h41" },
+	{ }
+};
+MODULE_DEVICE_TABLE(of, ili9881c_of_match);
+
+static struct mipi_dsi_driver ili9881c_dsi_driver = {
+	.probe		= ili9881c_dsi_probe,
+	.remove		= ili9881c_dsi_remove,
+	.driver = {
+		.name		= "ili9881c-dsi",
+		.of_match_table	= ili9881c_of_match,
+	},
+};
+module_mipi_dsi_driver(ili9881c_dsi_driver);
+
+MODULE_AUTHOR("Maxime Ripard <maxime.ripard@free-electrons.com>");
+MODULE_DESCRIPTION("Ilitek ILI9881C Controller Driver");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/panel/panel-innolux-p079zca.c b/drivers/gpu/drm/panel/panel-innolux-p079zca.c
index 57df39b5c589..72edb334d997 100644
--- a/drivers/gpu/drm/panel/panel-innolux-p079zca.c
+++ b/drivers/gpu/drm/panel/panel-innolux-p079zca.c
@@ -11,6 +11,7 @@
 #include <linux/gpio/consumer.h>
 #include <linux/module.h>
 #include <linux/of.h>
+#include <linux/of_device.h>
 #include <linux/regulator/consumer.h>
 
 #include <drm/drmP.h>
@@ -20,12 +21,41 @@
 
 #include <video/mipi_display.h>
 
+struct panel_init_cmd {
+	size_t len;
+	const char *data;
+};
+
+#define _INIT_CMD(...) { \
+	.len = sizeof((char[]){__VA_ARGS__}), \
+	.data = (char[]){__VA_ARGS__} }
+
+struct panel_desc {
+	const struct drm_display_mode *mode;
+	unsigned int bpc;
+	struct {
+		unsigned int width;
+		unsigned int height;
+	} size;
+
+	unsigned long flags;
+	enum mipi_dsi_pixel_format format;
+	const struct panel_init_cmd *init_cmds;
+	unsigned int lanes;
+	const char * const *supply_names;
+	unsigned int num_supplies;
+	unsigned int sleep_mode_delay;
+	unsigned int power_down_delay;
+};
+
 struct innolux_panel {
 	struct drm_panel base;
 	struct mipi_dsi_device *link;
+	const struct panel_desc *desc;
 
 	struct backlight_device *backlight;
-	struct regulator *supply;
+	struct regulator_bulk_data *supplies;
+	unsigned int num_supplies;
 	struct gpio_desc *enable_gpio;
 
 	bool prepared;
@@ -72,12 +102,16 @@ static int innolux_panel_unprepare(struct drm_panel *panel)
 		return err;
 	}
 
+	if (innolux->desc->sleep_mode_delay)
+		msleep(innolux->desc->sleep_mode_delay);
+
 	gpiod_set_value_cansleep(innolux->enable_gpio, 0);
 
-	/* T8: 80ms - 1000ms */
-	msleep(80);
+	if (innolux->desc->power_down_delay)
+		msleep(innolux->desc->power_down_delay);
 
-	err = regulator_disable(innolux->supply);
+	err = regulator_bulk_disable(innolux->desc->num_supplies,
+				     innolux->supplies);
 	if (err < 0)
 		return err;
 
@@ -89,24 +123,55 @@ static int innolux_panel_unprepare(struct drm_panel *panel)
 static int innolux_panel_prepare(struct drm_panel *panel)
 {
 	struct innolux_panel *innolux = to_innolux_panel(panel);
-	int err, regulator_err;
+	int err;
 
 	if (innolux->prepared)
 		return 0;
 
 	gpiod_set_value_cansleep(innolux->enable_gpio, 0);
 
-	err = regulator_enable(innolux->supply);
+	err = regulator_bulk_enable(innolux->desc->num_supplies,
+				    innolux->supplies);
 	if (err < 0)
 		return err;
 
-	/* T2: 15ms - 1000ms */
-	usleep_range(15000, 16000);
+	/* p079zca: t2 (20ms), p097pfg: t4 (15ms) */
+	usleep_range(20000, 21000);
 
 	gpiod_set_value_cansleep(innolux->enable_gpio, 1);
 
-	/* T4: 15ms - 1000ms */
-	usleep_range(15000, 16000);
+	/* p079zca: t4, p097pfg: t5 */
+	usleep_range(20000, 21000);
+
+	if (innolux->desc->init_cmds) {
+		const struct panel_init_cmd *cmds =
+					innolux->desc->init_cmds;
+		unsigned int i;
+
+		for (i = 0; cmds[i].len != 0; i++) {
+			const struct panel_init_cmd *cmd = &cmds[i];
+
+			err = mipi_dsi_generic_write(innolux->link, cmd->data,
+						     cmd->len);
+			if (err < 0) {
+				dev_err(panel->dev,
+					"failed to write command %u\n", i);
+				goto poweroff;
+			}
+
+			/*
+			 * Included by random guessing, because without this
+			 * (or at least, some delay), the panel sometimes
+			 * didn't appear to pick up the command sequence.
+			 */
+			err = mipi_dsi_dcs_nop(innolux->link);
+			if (err < 0) {
+				dev_err(panel->dev,
+					"failed to send DCS nop: %d\n", err);
+				goto poweroff;
+			}
+		}
+	}
 
 	err = mipi_dsi_dcs_exit_sleep_mode(innolux->link);
 	if (err < 0) {
@@ -133,12 +198,9 @@ static int innolux_panel_prepare(struct drm_panel *panel)
 	return 0;
 
 poweroff:
-	regulator_err = regulator_disable(innolux->supply);
-	if (regulator_err)
-		DRM_DEV_ERROR(panel->dev, "failed to disable regulator: %d\n",
-			      regulator_err);
-
 	gpiod_set_value_cansleep(innolux->enable_gpio, 0);
+	regulator_bulk_disable(innolux->desc->num_supplies, innolux->supplies);
+
 	return err;
 }
 
@@ -162,7 +224,11 @@ static int innolux_panel_enable(struct drm_panel *panel)
 	return 0;
 }
 
-static const struct drm_display_mode default_mode = {
+static const char * const innolux_p079zca_supply_names[] = {
+	"power",
+};
+
+static const struct drm_display_mode innolux_p079zca_mode = {
 	.clock = 56900,
 	.hdisplay = 768,
 	.hsync_start = 768 + 40,
@@ -175,15 +241,181 @@ static const struct drm_display_mode default_mode = {
 	.vrefresh = 60,
 };
 
+static const struct panel_desc innolux_p079zca_panel_desc = {
+	.mode = &innolux_p079zca_mode,
+	.bpc = 8,
+	.size = {
+		.width = 120,
+		.height = 160,
+	},
+	.flags = MIPI_DSI_MODE_VIDEO | MIPI_DSI_MODE_VIDEO_SYNC_PULSE |
+		 MIPI_DSI_MODE_LPM,
+	.format = MIPI_DSI_FMT_RGB888,
+	.lanes = 4,
+	.supply_names = innolux_p079zca_supply_names,
+	.num_supplies = ARRAY_SIZE(innolux_p079zca_supply_names),
+	.power_down_delay = 80, /* T8: 80ms - 1000ms */
+};
+
+static const char * const innolux_p097pfg_supply_names[] = {
+	"avdd",
+	"avee",
+};
+
+static const struct drm_display_mode innolux_p097pfg_mode = {
+	.clock = 229000,
+	.hdisplay = 1536,
+	.hsync_start = 1536 + 100,
+	.hsync_end = 1536 + 100 + 24,
+	.htotal = 1536 + 100 + 24 + 100,
+	.vdisplay = 2048,
+	.vsync_start = 2048 + 100,
+	.vsync_end = 2048 + 100 + 2,
+	.vtotal = 2048 + 100 + 2 + 18,
+	.vrefresh = 60,
+};
+
+/*
+ * Display manufacturer failed to provide init sequencing according to
+ * https://chromium-review.googlesource.com/c/chromiumos/third_party/coreboot/+/892065/
+ * so the init sequence stems from a register dump of a working panel.
+ */
+static const struct panel_init_cmd innolux_p097pfg_init_cmds[] = {
+	/* page 0 */
+	_INIT_CMD(0xF0, 0x55, 0xAA, 0x52, 0x08, 0x00),
+	_INIT_CMD(0xB1, 0xE8, 0x11),
+	_INIT_CMD(0xB2, 0x25, 0x02),
+	_INIT_CMD(0xB5, 0x08, 0x00),
+	_INIT_CMD(0xBC, 0x0F, 0x00),
+	_INIT_CMD(0xB8, 0x03, 0x06, 0x00, 0x00),
+	_INIT_CMD(0xBD, 0x01, 0x90, 0x14, 0x14),
+	_INIT_CMD(0x6F, 0x01),
+	_INIT_CMD(0xC0, 0x03),
+	_INIT_CMD(0x6F, 0x02),
+	_INIT_CMD(0xC1, 0x0D),
+	_INIT_CMD(0xD9, 0x01, 0x09, 0x70),
+	_INIT_CMD(0xC5, 0x12, 0x21, 0x00),
+	_INIT_CMD(0xBB, 0x93, 0x93),
+
+	/* page 1 */
+	_INIT_CMD(0xF0, 0x55, 0xAA, 0x52, 0x08, 0x01),
+	_INIT_CMD(0xB3, 0x3C, 0x3C),
+	_INIT_CMD(0xB4, 0x0F, 0x0F),
+	_INIT_CMD(0xB9, 0x45, 0x45),
+	_INIT_CMD(0xBA, 0x14, 0x14),
+	_INIT_CMD(0xCA, 0x02),
+	_INIT_CMD(0xCE, 0x04),
+	_INIT_CMD(0xC3, 0x9B, 0x9B),
+	_INIT_CMD(0xD8, 0xC0, 0x03),
+	_INIT_CMD(0xBC, 0x82, 0x01),
+	_INIT_CMD(0xBD, 0x9E, 0x01),
+
+	/* page 2 */
+	_INIT_CMD(0xF0, 0x55, 0xAA, 0x52, 0x08, 0x02),
+	_INIT_CMD(0xB0, 0x82),
+	_INIT_CMD(0xD1, 0x00, 0x00, 0x00, 0x3E, 0x00, 0x82, 0x00, 0xA5,
+		  0x00, 0xC1, 0x00, 0xEA, 0x01, 0x0D, 0x01, 0x40),
+	_INIT_CMD(0xD2, 0x01, 0x6A, 0x01, 0xA8, 0x01, 0xDC, 0x02, 0x29,
+		  0x02, 0x67, 0x02, 0x68, 0x02, 0xA8, 0x02, 0xF0),
+	_INIT_CMD(0xD3, 0x03, 0x19, 0x03, 0x49, 0x03, 0x67, 0x03, 0x8C,
+		  0x03, 0xA6, 0x03, 0xC7, 0x03, 0xDE, 0x03, 0xEC),
+	_INIT_CMD(0xD4, 0x03, 0xFF, 0x03, 0xFF),
+	_INIT_CMD(0xE0, 0x00, 0x00, 0x00, 0x86, 0x00, 0xC5, 0x00, 0xE5,
+		  0x00, 0xFF, 0x01, 0x26, 0x01, 0x45, 0x01, 0x75),
+	_INIT_CMD(0xE1, 0x01, 0x9C, 0x01, 0xD5, 0x02, 0x05, 0x02, 0x4D,
+		  0x02, 0x86, 0x02, 0x87, 0x02, 0xC3, 0x03, 0x03),
+	_INIT_CMD(0xE2, 0x03, 0x2A, 0x03, 0x56, 0x03, 0x72, 0x03, 0x94,
+		  0x03, 0xAC, 0x03, 0xCB, 0x03, 0xE0, 0x03, 0xED),
+	_INIT_CMD(0xE3, 0x03, 0xFF, 0x03, 0xFF),
+
+	/* page 3 */
+	_INIT_CMD(0xF0, 0x55, 0xAA, 0x52, 0x08, 0x03),
+	_INIT_CMD(0xB0, 0x00, 0x00, 0x00, 0x00),
+	_INIT_CMD(0xB1, 0x00, 0x00, 0x00, 0x00),
+	_INIT_CMD(0xB2, 0x00, 0x00, 0x06, 0x04, 0x01, 0x40, 0x85),
+	_INIT_CMD(0xB3, 0x10, 0x07, 0xFC, 0x04, 0x01, 0x40, 0x80),
+	_INIT_CMD(0xB6, 0xF0, 0x08, 0x00, 0x04, 0x00, 0x00, 0x00, 0x01,
+		  0x40, 0x80),
+	_INIT_CMD(0xBA, 0xC5, 0x07, 0x00, 0x04, 0x11, 0x25, 0x8C),
+	_INIT_CMD(0xBB, 0xC5, 0x07, 0x00, 0x03, 0x11, 0x25, 0x8C),
+	_INIT_CMD(0xC0, 0x00, 0x3C, 0x00, 0x00, 0x00, 0x80, 0x80),
+	_INIT_CMD(0xC1, 0x00, 0x3C, 0x00, 0x00, 0x00, 0x80, 0x80),
+	_INIT_CMD(0xC4, 0x00, 0x00),
+	_INIT_CMD(0xEF, 0x41),
+
+	/* page 4 */
+	_INIT_CMD(0xF0, 0x55, 0xAA, 0x52, 0x08, 0x04),
+	_INIT_CMD(0xEC, 0x4C),
+
+	/* page 5 */
+	_INIT_CMD(0xF0, 0x55, 0xAA, 0x52, 0x08, 0x05),
+	_INIT_CMD(0xB0, 0x13, 0x03, 0x03, 0x01),
+	_INIT_CMD(0xB1, 0x30, 0x00),
+	_INIT_CMD(0xB2, 0x02, 0x02, 0x00),
+	_INIT_CMD(0xB3, 0x82, 0x23, 0x82, 0x9D),
+	_INIT_CMD(0xB4, 0xC5, 0x75, 0x24, 0x57),
+	_INIT_CMD(0xB5, 0x00, 0xD4, 0x72, 0x11, 0x11, 0xAB, 0x0A),
+	_INIT_CMD(0xB6, 0x00, 0x00, 0xD5, 0x72, 0x24, 0x56),
+	_INIT_CMD(0xB7, 0x5C, 0xDC, 0x5C, 0x5C),
+	_INIT_CMD(0xB9, 0x0C, 0x00, 0x00, 0x01, 0x00),
+	_INIT_CMD(0xC0, 0x75, 0x11, 0x11, 0x54, 0x05),
+	_INIT_CMD(0xC6, 0x00, 0x00, 0x00, 0x00),
+	_INIT_CMD(0xD0, 0x00, 0x48, 0x08, 0x00, 0x00),
+	_INIT_CMD(0xD1, 0x00, 0x48, 0x09, 0x00, 0x00),
+
+	/* page 6 */
+	_INIT_CMD(0xF0, 0x55, 0xAA, 0x52, 0x08, 0x06),
+	_INIT_CMD(0xB0, 0x02, 0x32, 0x32, 0x08, 0x2F),
+	_INIT_CMD(0xB1, 0x2E, 0x15, 0x14, 0x13, 0x12),
+	_INIT_CMD(0xB2, 0x11, 0x10, 0x00, 0x3D, 0x3D),
+	_INIT_CMD(0xB3, 0x3D, 0x3D, 0x3D, 0x3D, 0x3D),
+	_INIT_CMD(0xB4, 0x3D, 0x32),
+	_INIT_CMD(0xB5, 0x03, 0x32, 0x32, 0x09, 0x2F),
+	_INIT_CMD(0xB6, 0x2E, 0x1B, 0x1A, 0x19, 0x18),
+	_INIT_CMD(0xB7, 0x17, 0x16, 0x01, 0x3D, 0x3D),
+	_INIT_CMD(0xB8, 0x3D, 0x3D, 0x3D, 0x3D, 0x3D),
+	_INIT_CMD(0xB9, 0x3D, 0x32),
+	_INIT_CMD(0xC0, 0x01, 0x32, 0x32, 0x09, 0x2F),
+	_INIT_CMD(0xC1, 0x2E, 0x1A, 0x1B, 0x16, 0x17),
+	_INIT_CMD(0xC2, 0x18, 0x19, 0x03, 0x3D, 0x3D),
+	_INIT_CMD(0xC3, 0x3D, 0x3D, 0x3D, 0x3D, 0x3D),
+	_INIT_CMD(0xC4, 0x3D, 0x32),
+	_INIT_CMD(0xC5, 0x00, 0x32, 0x32, 0x08, 0x2F),
+	_INIT_CMD(0xC6, 0x2E, 0x14, 0x15, 0x10, 0x11),
+	_INIT_CMD(0xC7, 0x12, 0x13, 0x02, 0x3D, 0x3D),
+	_INIT_CMD(0xC8, 0x3D, 0x3D, 0x3D, 0x3D, 0x3D),
+	_INIT_CMD(0xC9, 0x3D, 0x32),
+
+	{},
+};
+
+static const struct panel_desc innolux_p097pfg_panel_desc = {
+	.mode = &innolux_p097pfg_mode,
+	.bpc = 8,
+	.size = {
+		.width = 147,
+		.height = 196,
+	},
+	.flags = MIPI_DSI_MODE_VIDEO | MIPI_DSI_MODE_VIDEO_SYNC_PULSE |
+		 MIPI_DSI_MODE_LPM,
+	.format = MIPI_DSI_FMT_RGB888,
+	.init_cmds = innolux_p097pfg_init_cmds,
+	.lanes = 4,
+	.supply_names = innolux_p097pfg_supply_names,
+	.num_supplies = ARRAY_SIZE(innolux_p097pfg_supply_names),
+	.sleep_mode_delay = 100, /* T15 */
+};
+
 static int innolux_panel_get_modes(struct drm_panel *panel)
 {
+	struct innolux_panel *innolux = to_innolux_panel(panel);
+	const struct drm_display_mode *m = innolux->desc->mode;
 	struct drm_display_mode *mode;
 
-	mode = drm_mode_duplicate(panel->drm, &default_mode);
+	mode = drm_mode_duplicate(panel->drm, m);
 	if (!mode) {
 		DRM_DEV_ERROR(panel->drm->dev, "failed to add mode %ux%ux@%u\n",
-			      default_mode.hdisplay, default_mode.vdisplay,
-			      default_mode.vrefresh);
+			      m->hdisplay, m->vdisplay, m->vrefresh);
 		return -ENOMEM;
 	}
 
@@ -191,9 +423,11 @@ static int innolux_panel_get_modes(struct drm_panel *panel)
 
 	drm_mode_probed_add(panel->connector, mode);
 
-	panel->connector->display_info.width_mm = 120;
-	panel->connector->display_info.height_mm = 160;
-	panel->connector->display_info.bpc = 8;
+	panel->connector->display_info.width_mm =
+			innolux->desc->size.width;
+	panel->connector->display_info.height_mm =
+			innolux->desc->size.height;
+	panel->connector->display_info.bpc = innolux->desc->bpc;
 
 	return 1;
 }
@@ -207,19 +441,42 @@ static const struct drm_panel_funcs innolux_panel_funcs = {
 };
 
 static const struct of_device_id innolux_of_match[] = {
-	{ .compatible = "innolux,p079zca", },
+	{ .compatible = "innolux,p079zca",
+	  .data = &innolux_p079zca_panel_desc
+	},
+	{ .compatible = "innolux,p097pfg",
+	  .data = &innolux_p097pfg_panel_desc
+	},
 	{ }
 };
 MODULE_DEVICE_TABLE(of, innolux_of_match);
 
-static int innolux_panel_add(struct innolux_panel *innolux)
+static int innolux_panel_add(struct mipi_dsi_device *dsi,
+			     const struct panel_desc *desc)
 {
-	struct device *dev = &innolux->link->dev;
-	int err;
+	struct innolux_panel *innolux;
+	struct device *dev = &dsi->dev;
+	int err, i;
+
+	innolux = devm_kzalloc(dev, sizeof(*innolux), GFP_KERNEL);
+	if (!innolux)
+		return -ENOMEM;
+
+	innolux->desc = desc;
+
+	innolux->supplies = devm_kcalloc(dev, desc->num_supplies,
+					 sizeof(*innolux->supplies),
+					 GFP_KERNEL);
+	if (!innolux->supplies)
+		return -ENOMEM;
+
+	for (i = 0; i < desc->num_supplies; i++)
+		innolux->supplies[i].supply = desc->supply_names[i];
 
-	innolux->supply = devm_regulator_get(dev, "power");
-	if (IS_ERR(innolux->supply))
-		return PTR_ERR(innolux->supply);
+	err = devm_regulator_bulk_get(dev, desc->num_supplies,
+				      innolux->supplies);
+	if (err < 0)
+		return err;
 
 	innolux->enable_gpio = devm_gpiod_get_optional(dev, "enable",
 						       GPIOD_OUT_HIGH);
@@ -230,15 +487,21 @@ static int innolux_panel_add(struct innolux_panel *innolux)
 	}
 
 	innolux->backlight = devm_of_find_backlight(dev);
-
 	if (IS_ERR(innolux->backlight))
 		return PTR_ERR(innolux->backlight);
 
 	drm_panel_init(&innolux->base);
 	innolux->base.funcs = &innolux_panel_funcs;
-	innolux->base.dev = &innolux->link->dev;
+	innolux->base.dev = dev;
+
+	err = drm_panel_add(&innolux->base);
+	if (err < 0)
+		return err;
+
+	mipi_dsi_set_drvdata(dsi, innolux);
+	innolux->link = dsi;
 
-	return drm_panel_add(&innolux->base);
+	return 0;
 }
 
 static void innolux_panel_del(struct innolux_panel *innolux)
@@ -249,28 +512,19 @@ static void innolux_panel_del(struct innolux_panel *innolux)
 
 static int innolux_panel_probe(struct mipi_dsi_device *dsi)
 {
-	struct innolux_panel *innolux;
+	const struct panel_desc *desc;
 	int err;
 
-	dsi->lanes = 4;
-	dsi->format = MIPI_DSI_FMT_RGB888;
-	dsi->mode_flags = MIPI_DSI_MODE_VIDEO | MIPI_DSI_MODE_VIDEO_SYNC_PULSE |
-			  MIPI_DSI_MODE_LPM;
-
-	innolux = devm_kzalloc(&dsi->dev, sizeof(*innolux), GFP_KERNEL);
-	if (!innolux)
-		return -ENOMEM;
-
-	mipi_dsi_set_drvdata(dsi, innolux);
+	desc = of_device_get_match_data(&dsi->dev);
+	dsi->mode_flags = desc->flags;
+	dsi->format = desc->format;
+	dsi->lanes = desc->lanes;
 
-	innolux->link = dsi;
-
-	err = innolux_panel_add(innolux);
+	err = innolux_panel_add(dsi, desc);
 	if (err < 0)
 		return err;
 
-	err = mipi_dsi_attach(dsi);
-	return err;
+	return mipi_dsi_attach(dsi);
 }
 
 static int innolux_panel_remove(struct mipi_dsi_device *dsi)
@@ -292,7 +546,6 @@ static int innolux_panel_remove(struct mipi_dsi_device *dsi)
 		DRM_DEV_ERROR(&dsi->dev, "failed to detach from DSI host: %d\n",
 			      err);
 
-	drm_panel_detach(&innolux->base);
 	innolux_panel_del(innolux);
 
 	return 0;
@@ -318,5 +571,6 @@ static struct mipi_dsi_driver innolux_panel_driver = {
 module_mipi_dsi_driver(innolux_panel_driver);
 
 MODULE_AUTHOR("Chris Zhong <zyw@rock-chips.com>");
+MODULE_AUTHOR("Lin Huang <hl@rock-chips.com>");
 MODULE_DESCRIPTION("Innolux P079ZCA panel driver");
 MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/panel/panel-jdi-lt070me05000.c b/drivers/gpu/drm/panel/panel-jdi-lt070me05000.c
index 0a94ab79a6c0..99caa7835e7b 100644
--- a/drivers/gpu/drm/panel/panel-jdi-lt070me05000.c
+++ b/drivers/gpu/drm/panel/panel-jdi-lt070me05000.c
@@ -500,7 +500,6 @@ static int jdi_panel_remove(struct mipi_dsi_device *dsi)
 		dev_err(&dsi->dev, "failed to detach from DSI host: %d\n",
 			ret);
 
-	drm_panel_detach(&jdi->base);
 	jdi_panel_del(jdi);
 
 	return 0;
diff --git a/drivers/gpu/drm/panel/panel-lvds.c b/drivers/gpu/drm/panel/panel-lvds.c
index 5185819c5b79..8a1687887ae9 100644
--- a/drivers/gpu/drm/panel/panel-lvds.c
+++ b/drivers/gpu/drm/panel/panel-lvds.c
@@ -282,7 +282,6 @@ static int panel_lvds_remove(struct platform_device *pdev)
 {
 	struct panel_lvds *lvds = dev_get_drvdata(&pdev->dev);
 
-	drm_panel_detach(&lvds->panel);
 	drm_panel_remove(&lvds->panel);
 
 	panel_lvds_disable(&lvds->panel);
diff --git a/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c b/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c
index 90f1ae4af93c..87fa316e1d7b 100644
--- a/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c
+++ b/drivers/gpu/drm/panel/panel-orisetech-otm8009a.c
@@ -14,8 +14,6 @@
 #include <linux/regulator/consumer.h>
 #include <video/mipi_display.h>
 
-#define DRV_NAME "orisetech_otm8009a"
-
 #define OTM8009A_BACKLIGHT_DEFAULT	240
 #define OTM8009A_BACKLIGHT_MAX		255
 
@@ -98,6 +96,20 @@ static void otm8009a_dcs_write_buf(struct otm8009a *ctx, const void *data,
 		DRM_WARN("mipi dsi dcs write buffer failed\n");
 }
 
+static void otm8009a_dcs_write_buf_hs(struct otm8009a *ctx, const void *data,
+				      size_t len)
+{
+	struct mipi_dsi_device *dsi = to_mipi_dsi_device(ctx->dev);
+
+	/* data will be sent in dsi hs mode (ie. no lpm) */
+	dsi->mode_flags &= ~MIPI_DSI_MODE_LPM;
+
+	otm8009a_dcs_write_buf(ctx, data, len);
+
+	/* restore back the dsi lpm mode */
+	dsi->mode_flags |= MIPI_DSI_MODE_LPM;
+}
+
 #define dcs_write_seq(ctx, seq...)			\
 ({							\
 	static const u8 d[] = { seq };			\
@@ -248,11 +260,7 @@ static int otm8009a_disable(struct drm_panel *panel)
 	if (!ctx->enabled)
 		return 0; /* This is not an issue so we return 0 here */
 
-	/* Power off the backlight. Note: end-user still controls brightness */
-	ctx->bl_dev->props.power = FB_BLANK_POWERDOWN;
-	ret = backlight_update_status(ctx->bl_dev);
-	if (ret)
-		return ret;
+	backlight_disable(ctx->bl_dev);
 
 	ret = mipi_dsi_dcs_set_display_off(dsi);
 	if (ret)
@@ -316,13 +324,6 @@ static int otm8009a_prepare(struct drm_panel *panel)
 
 	ctx->prepared = true;
 
-	/*
-	 * Power on the backlight. Note: end-user still controls brightness
-	 * Note: ctx->prepared must be true before updating the backlight.
-	 */
-	ctx->bl_dev->props.power = FB_BLANK_UNBLANK;
-	backlight_update_status(ctx->bl_dev);
-
 	return 0;
 }
 
@@ -330,6 +331,11 @@ static int otm8009a_enable(struct drm_panel *panel)
 {
 	struct otm8009a *ctx = panel_to_otm8009a(panel);
 
+	if (ctx->enabled)
+		return 0;
+
+	backlight_enable(ctx->bl_dev);
+
 	ctx->enabled = true;
 
 	return 0;
@@ -387,7 +393,7 @@ static int otm8009a_backlight_update_status(struct backlight_device *bd)
 		 */
 		data[0] = MIPI_DCS_SET_DISPLAY_BRIGHTNESS;
 		data[1] = bd->props.brightness;
-		otm8009a_dcs_write_buf(ctx, data, ARRAY_SIZE(data));
+		otm8009a_dcs_write_buf_hs(ctx, data, ARRAY_SIZE(data));
 
 		/* set Brightness Control & Backlight on */
 		data[1] = 0x24;
@@ -399,7 +405,7 @@ static int otm8009a_backlight_update_status(struct backlight_device *bd)
 
 	/* Update Brightness Control & Backlight */
 	data[0] = MIPI_DCS_WRITE_CONTROL_DISPLAY;
-	otm8009a_dcs_write_buf(ctx, data, ARRAY_SIZE(data));
+	otm8009a_dcs_write_buf_hs(ctx, data, ARRAY_SIZE(data));
 
 	return 0;
 }
@@ -444,11 +450,14 @@ static int otm8009a_probe(struct mipi_dsi_device *dsi)
 	ctx->panel.dev = dev;
 	ctx->panel.funcs = &otm8009a_drm_funcs;
 
-	ctx->bl_dev = backlight_device_register(DRV_NAME "_backlight", dev, ctx,
-						&otm8009a_backlight_ops, NULL);
+	ctx->bl_dev = devm_backlight_device_register(dev, dev_name(dev),
+						     dsi->host->dev, ctx,
+						     &otm8009a_backlight_ops,
+						     NULL);
 	if (IS_ERR(ctx->bl_dev)) {
-		dev_err(dev, "failed to register backlight device\n");
-		return PTR_ERR(ctx->bl_dev);
+		ret = PTR_ERR(ctx->bl_dev);
+		dev_err(dev, "failed to register backlight: %d\n", ret);
+		return ret;
 	}
 
 	ctx->bl_dev->props.max_brightness = OTM8009A_BACKLIGHT_MAX;
@@ -466,11 +475,6 @@ static int otm8009a_probe(struct mipi_dsi_device *dsi)
 		return ret;
 	}
 
-	DRM_INFO(DRV_NAME "_panel %ux%u@%u %ubpp dsi %udl - ready\n",
-		 default_mode.hdisplay, default_mode.vdisplay,
-		 default_mode.vrefresh,
-		 mipi_dsi_pixel_format_to_bpp(dsi->format), dsi->lanes);
-
 	return 0;
 }
 
@@ -481,8 +485,6 @@ static int otm8009a_remove(struct mipi_dsi_device *dsi)
 	mipi_dsi_detach(dsi);
 	drm_panel_remove(&ctx->panel);
 
-	backlight_device_unregister(ctx->bl_dev);
-
 	return 0;
 }
 
@@ -496,7 +498,7 @@ static struct mipi_dsi_driver orisetech_otm8009a_driver = {
 	.probe  = otm8009a_probe,
 	.remove = otm8009a_remove,
 	.driver = {
-		.name = DRV_NAME "_panel",
+		.name = "panel-orisetech-otm8009a",
 		.of_match_table = orisetech_otm8009a_of_match,
 	},
 };
diff --git a/drivers/gpu/drm/panel/panel-panasonic-vvx10f034n00.c b/drivers/gpu/drm/panel/panel-panasonic-vvx10f034n00.c
index 74a806121f80..cb4dfb98be0f 100644
--- a/drivers/gpu/drm/panel/panel-panasonic-vvx10f034n00.c
+++ b/drivers/gpu/drm/panel/panel-panasonic-vvx10f034n00.c
@@ -299,7 +299,6 @@ static int wuxga_nt_panel_remove(struct mipi_dsi_device *dsi)
 	if (ret < 0)
 		dev_err(&dsi->dev, "failed to detach from DSI host: %d\n", ret);
 
-	drm_panel_detach(&wuxga_nt->base);
 	wuxga_nt_panel_del(wuxga_nt);
 
 	return 0;
diff --git a/drivers/gpu/drm/panel/panel-raspberrypi-touchscreen.c b/drivers/gpu/drm/panel/panel-raspberrypi-touchscreen.c
index d964d454e4ae..2c9c9722734f 100644
--- a/drivers/gpu/drm/panel/panel-raspberrypi-touchscreen.c
+++ b/drivers/gpu/drm/panel/panel-raspberrypi-touchscreen.c
@@ -238,12 +238,6 @@ static void rpi_touchscreen_i2c_write(struct rpi_touchscreen *ts,
 
 static int rpi_touchscreen_write(struct rpi_touchscreen *ts, u16 reg, u32 val)
 {
-#if 0
-	/* The firmware uses LP DSI transactions like this to bring up
-	 * the hardware, which should be faster than using I2C to then
-	 * pass to the Toshiba.  However, I was unable to get it to
-	 * work.
-	 */
 	u8 msg[] = {
 		reg,
 		reg >> 8,
@@ -253,13 +247,7 @@ static int rpi_touchscreen_write(struct rpi_touchscreen *ts, u16 reg, u32 val)
 		val >> 24,
 	};
 
-	mipi_dsi_dcs_write_buffer(ts->dsi, msg, sizeof(msg));
-#else
-	rpi_touchscreen_i2c_write(ts, REG_WR_ADDRH, reg >> 8);
-	rpi_touchscreen_i2c_write(ts, REG_WR_ADDRL, reg);
-	rpi_touchscreen_i2c_write(ts, REG_WRITEH, val >> 8);
-	rpi_touchscreen_i2c_write(ts, REG_WRITEL, val);
-#endif
+	mipi_dsi_generic_write(ts->dsi, msg, sizeof(msg));
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/panel/panel-samsung-s6e8aa0.c b/drivers/gpu/drm/panel/panel-samsung-s6e8aa0.c
index a188a3959f1a..6ad827b93ae1 100644
--- a/drivers/gpu/drm/panel/panel-samsung-s6e8aa0.c
+++ b/drivers/gpu/drm/panel/panel-samsung-s6e8aa0.c
@@ -823,7 +823,7 @@ static void s6e8aa0_read_mtp_id(struct s6e8aa0 *ctx)
 	int ret, i;
 
 	ret = s6e8aa0_dcs_read(ctx, 0xd1, id, ARRAY_SIZE(id));
-	if (ret < ARRAY_SIZE(id) || id[0] == 0x00) {
+	if (ret < 0 || ret < ARRAY_SIZE(id) || id[0] == 0x00) {
 		dev_err(ctx->dev, "read id failed\n");
 		ctx->error = -EIO;
 		return;
diff --git a/drivers/gpu/drm/panel/panel-seiko-43wvf1g.c b/drivers/gpu/drm/panel/panel-seiko-43wvf1g.c
index 71c09ed436ae..75f925390551 100644
--- a/drivers/gpu/drm/panel/panel-seiko-43wvf1g.c
+++ b/drivers/gpu/drm/panel/panel-seiko-43wvf1g.c
@@ -292,7 +292,6 @@ static int seiko_panel_remove(struct platform_device *pdev)
 {
 	struct seiko_panel *panel = dev_get_drvdata(&pdev->dev);
 
-	drm_panel_detach(&panel->base);
 	drm_panel_remove(&panel->base);
 
 	seiko_panel_disable(&panel->base);
diff --git a/drivers/gpu/drm/panel/panel-sharp-lq101r1sx01.c b/drivers/gpu/drm/panel/panel-sharp-lq101r1sx01.c
index 6bf8730f1a21..02fc0f5423d4 100644
--- a/drivers/gpu/drm/panel/panel-sharp-lq101r1sx01.c
+++ b/drivers/gpu/drm/panel/panel-sharp-lq101r1sx01.c
@@ -418,7 +418,6 @@ static int sharp_panel_remove(struct mipi_dsi_device *dsi)
 	if (err < 0)
 		dev_err(&dsi->dev, "failed to detach from DSI host: %d\n", err);
 
-	drm_panel_detach(&sharp->base);
 	sharp_panel_del(sharp);
 
 	return 0;
diff --git a/drivers/gpu/drm/panel/panel-sharp-ls043t1le01.c b/drivers/gpu/drm/panel/panel-sharp-ls043t1le01.c
index 494aa9b1628a..e5cae0050f52 100644
--- a/drivers/gpu/drm/panel/panel-sharp-ls043t1le01.c
+++ b/drivers/gpu/drm/panel/panel-sharp-ls043t1le01.c
@@ -327,7 +327,6 @@ static int sharp_nt_panel_remove(struct mipi_dsi_device *dsi)
 	if (ret < 0)
 		dev_err(&dsi->dev, "failed to detach from DSI host: %d\n", ret);
 
-	drm_panel_detach(&sharp_nt->base);
 	sharp_nt_panel_del(sharp_nt);
 
 	return 0;
diff --git a/drivers/gpu/drm/panel/panel-simple.c b/drivers/gpu/drm/panel/panel-simple.c
index cbf1ab404ee7..5b5d0a24e713 100644
--- a/drivers/gpu/drm/panel/panel-simple.c
+++ b/drivers/gpu/drm/panel/panel-simple.c
@@ -252,7 +252,7 @@ static int panel_simple_get_modes(struct drm_panel *panel)
 	/* probe EDID if a DDC bus is available */
 	if (p->ddc) {
 		struct edid *edid = drm_get_edid(panel->connector, p->ddc);
-		drm_mode_connector_update_edid_property(panel->connector, edid);
+		drm_connector_update_edid_property(panel->connector, edid);
 		if (edid) {
 			num += drm_add_edid_modes(panel->connector, edid);
 			kfree(edid);
@@ -364,7 +364,6 @@ static int panel_simple_remove(struct device *dev)
 {
 	struct panel_simple *panel = dev_get_drvdata(dev);
 
-	drm_panel_detach(&panel->base);
 	drm_panel_remove(&panel->base);
 
 	panel_simple_disable(&panel->base);
@@ -581,6 +580,34 @@ static const struct panel_desc auo_b133htn01 = {
 	},
 };
 
+static const struct display_timing auo_g070vvn01_timings = {
+	.pixelclock = { 33300000, 34209000, 45000000 },
+	.hactive = { 800, 800, 800 },
+	.hfront_porch = { 20, 40, 200 },
+	.hback_porch = { 87, 40, 1 },
+	.hsync_len = { 1, 48, 87 },
+	.vactive = { 480, 480, 480 },
+	.vfront_porch = { 5, 13, 200 },
+	.vback_porch = { 31, 31, 29 },
+	.vsync_len = { 1, 1, 3 },
+};
+
+static const struct panel_desc auo_g070vvn01 = {
+	.timings = &auo_g070vvn01_timings,
+	.num_timings = 1,
+	.bpc = 8,
+	.size = {
+		.width = 152,
+		.height = 91,
+	},
+	.delay = {
+		.prepare = 200,
+		.enable = 50,
+		.disable = 50,
+		.unprepare = 1000,
+	},
+};
+
 static const struct drm_display_mode auo_g104sn02_mode = {
 	.clock = 40000,
 	.hdisplay = 800,
@@ -687,7 +714,7 @@ static const struct panel_desc auo_p320hvn03 = {
 		.enable = 450,
 		.unprepare = 500,
 	},
-	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_JEIDA,
+	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
 };
 
 static const struct drm_display_mode auo_t215hvn01_mode = {
@@ -745,6 +772,28 @@ static const struct panel_desc avic_tm070ddh03 = {
 	},
 };
 
+static const struct drm_display_mode boe_hv070wsa_mode = {
+	.clock = 40800,
+	.hdisplay = 1024,
+	.hsync_start = 1024 + 90,
+	.hsync_end = 1024 + 90 + 90,
+	.htotal = 1024 + 90 + 90 + 90,
+	.vdisplay = 600,
+	.vsync_start = 600 + 3,
+	.vsync_end = 600 + 3 + 4,
+	.vtotal = 600 + 3 + 4 + 3,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc boe_hv070wsa = {
+	.modes = &boe_hv070wsa_mode,
+	.num_modes = 1,
+	.size = {
+		.width = 154,
+		.height = 90,
+	},
+};
+
 static const struct drm_display_mode boe_nv101wxmn51_modes[] = {
 	{
 		.clock = 71900,
@@ -857,6 +906,61 @@ static const struct panel_desc chunghwa_claa101wb01 = {
 	},
 };
 
+static const struct drm_display_mode dataimage_scf0700c48ggu18_mode = {
+	.clock = 33260,
+	.hdisplay = 800,
+	.hsync_start = 800 + 40,
+	.hsync_end = 800 + 40 + 128,
+	.htotal = 800 + 40 + 128 + 88,
+	.vdisplay = 480,
+	.vsync_start = 480 + 10,
+	.vsync_end = 480 + 10 + 2,
+	.vtotal = 480 + 10 + 2 + 33,
+	.vrefresh = 60,
+	.flags = DRM_MODE_FLAG_NVSYNC | DRM_MODE_FLAG_NHSYNC,
+};
+
+static const struct panel_desc dataimage_scf0700c48ggu18 = {
+	.modes = &dataimage_scf0700c48ggu18_mode,
+	.num_modes = 1,
+	.bpc = 8,
+	.size = {
+		.width = 152,
+		.height = 91,
+	},
+	.bus_format = MEDIA_BUS_FMT_RGB888_1X24,
+	.bus_flags = DRM_BUS_FLAG_DE_HIGH | DRM_BUS_FLAG_PIXDATA_POSEDGE,
+};
+
+static const struct display_timing dlc_dlc0700yzg_1_timing = {
+	.pixelclock = { 45000000, 51200000, 57000000 },
+	.hactive = { 1024, 1024, 1024 },
+	.hfront_porch = { 100, 106, 113 },
+	.hback_porch = { 100, 106, 113 },
+	.hsync_len = { 100, 108, 114 },
+	.vactive = { 600, 600, 600 },
+	.vfront_porch = { 8, 11, 15 },
+	.vback_porch = { 8, 11, 15 },
+	.vsync_len = { 9, 13, 15 },
+	.flags = DISPLAY_FLAGS_DE_HIGH,
+};
+
+static const struct panel_desc dlc_dlc0700yzg_1 = {
+	.timings = &dlc_dlc0700yzg_1_timing,
+	.num_timings = 1,
+	.bpc = 6,
+	.size = {
+		.width = 154,
+		.height = 86,
+	},
+	.delay = {
+		.prepare = 30,
+		.enable = 200,
+		.disable = 200,
+	},
+	.bus_format = MEDIA_BUS_FMT_RGB666_1X7X3_SPWG,
+};
+
 static const struct drm_display_mode edt_et057090dhu_mode = {
 	.clock = 25175,
 	.hdisplay = 640,
@@ -909,6 +1013,18 @@ static const struct panel_desc edt_etm0700g0dh6 = {
 	.bus_flags = DRM_BUS_FLAG_DE_HIGH | DRM_BUS_FLAG_PIXDATA_NEGEDGE,
 };
 
+static const struct panel_desc edt_etm0700g0bdh6 = {
+	.modes = &edt_etm0700g0dh6_mode,
+	.num_modes = 1,
+	.bpc = 6,
+	.size = {
+		.width = 152,
+		.height = 91,
+	},
+	.bus_format = MEDIA_BUS_FMT_RGB666_1X18,
+	.bus_flags = DRM_BUS_FLAG_DE_HIGH | DRM_BUS_FLAG_PIXDATA_POSEDGE,
+};
+
 static const struct drm_display_mode foxlink_fl500wvr00_a0t_mode = {
 	.clock = 32260,
 	.hdisplay = 800,
@@ -1086,6 +1202,36 @@ static const struct panel_desc innolux_at070tn92 = {
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X24,
 };
 
+static const struct display_timing innolux_g070y2_l01_timing = {
+	.pixelclock = { 28000000, 29500000, 32000000 },
+	.hactive = { 800, 800, 800 },
+	.hfront_porch = { 61, 91, 141 },
+	.hback_porch = { 60, 90, 140 },
+	.hsync_len = { 12, 12, 12 },
+	.vactive = { 480, 480, 480 },
+	.vfront_porch = { 4, 9, 30 },
+	.vback_porch = { 4, 8, 28 },
+	.vsync_len = { 2, 2, 2 },
+	.flags = DISPLAY_FLAGS_DE_HIGH,
+};
+
+static const struct panel_desc innolux_g070y2_l01 = {
+	.timings = &innolux_g070y2_l01_timing,
+	.num_timings = 1,
+	.bpc = 6,
+	.size = {
+		.width = 152,
+		.height = 91,
+	},
+	.delay = {
+		.prepare = 10,
+		.enable = 100,
+		.disable = 100,
+		.unprepare = 800,
+	},
+	.bus_format = MEDIA_BUS_FMT_RGB888_1X7X4_SPWG,
+};
+
 static const struct display_timing innolux_g101ice_l01_timing = {
 	.pixelclock = { 60400000, 71100000, 74700000 },
 	.hactive = { 1280, 1280, 1280 },
@@ -1217,6 +1363,30 @@ static const struct panel_desc innolux_n156bge_l21 = {
 	},
 };
 
+static const struct drm_display_mode innolux_tv123wam_mode = {
+	.clock = 206016,
+	.hdisplay = 2160,
+	.hsync_start = 2160 + 48,
+	.hsync_end = 2160 + 48 + 32,
+	.htotal = 2160 + 48 + 32 + 80,
+	.vdisplay = 1440,
+	.vsync_start = 1440 + 3,
+	.vsync_end = 1440 + 3 + 10,
+	.vtotal = 1440 + 3 + 10 + 27,
+	.vrefresh = 60,
+	.flags = DRM_MODE_FLAG_PHSYNC | DRM_MODE_FLAG_PVSYNC,
+};
+
+static const struct panel_desc innolux_tv123wam = {
+	.modes = &innolux_tv123wam_mode,
+	.num_modes = 1,
+	.bpc = 8,
+	.size = {
+		.width = 259,
+		.height = 173,
+	},
+};
+
 static const struct drm_display_mode innolux_zj070na_01p_mode = {
 	.clock = 51501,
 	.hdisplay = 1024,
@@ -1247,8 +1417,8 @@ static const struct display_timing koe_tx31d200vm0baa_timing = {
 	.hback_porch = { 16, 36, 56 },
 	.hsync_len = { 8, 8, 8 },
 	.vactive = { 480, 480, 480 },
-	.vfront_porch = { 6, 21, 33.5 },
-	.vback_porch = { 6, 21, 33.5 },
+	.vfront_porch = { 6, 21, 33 },
+	.vback_porch = { 6, 21, 33 },
 	.vsync_len = { 8, 8, 8 },
 	.flags = DISPLAY_FLAGS_DE_HIGH,
 };
@@ -1511,6 +1681,33 @@ static const struct panel_desc netron_dy_e231732 = {
 	.bus_format = MEDIA_BUS_FMT_RGB666_1X18,
 };
 
+static const struct drm_display_mode newhaven_nhd_43_480272ef_atxl_mode = {
+	.clock = 9000,
+	.hdisplay = 480,
+	.hsync_start = 480 + 2,
+	.hsync_end = 480 + 2 + 41,
+	.htotal = 480 + 2 + 41 + 2,
+	.vdisplay = 272,
+	.vsync_start = 272 + 2,
+	.vsync_end = 272 + 2 + 10,
+	.vtotal = 272 + 2 + 10 + 2,
+	.vrefresh = 60,
+	.flags = DRM_MODE_FLAG_NVSYNC | DRM_MODE_FLAG_NHSYNC,
+};
+
+static const struct panel_desc newhaven_nhd_43_480272ef_atxl = {
+	.modes = &newhaven_nhd_43_480272ef_atxl_mode,
+	.num_modes = 1,
+	.bpc = 8,
+	.size = {
+		.width = 95,
+		.height = 54,
+	},
+	.bus_format = MEDIA_BUS_FMT_RGB888_1X24,
+	.bus_flags = DRM_BUS_FLAG_DE_HIGH | DRM_BUS_FLAG_PIXDATA_POSEDGE |
+		     DRM_BUS_FLAG_SYNC_POSEDGE,
+};
+
 static const struct display_timing nlt_nl192108ac18_02d_timing = {
 	.pixelclock = { 130000000, 148350000, 163000000 },
 	.hactive = { 1920, 1920, 1920 },
@@ -1696,6 +1893,36 @@ static const struct panel_desc qd43003c0_40 = {
 	.bus_format = MEDIA_BUS_FMT_RGB888_1X24,
 };
 
+static const struct display_timing rocktech_rk070er9427_timing = {
+	.pixelclock = { 26400000, 33300000, 46800000 },
+	.hactive = { 800, 800, 800 },
+	.hfront_porch = { 16, 210, 354 },
+	.hback_porch = { 46, 46, 46 },
+	.hsync_len = { 1, 1, 1 },
+	.vactive = { 480, 480, 480 },
+	.vfront_porch = { 7, 22, 147 },
+	.vback_porch = { 23, 23, 23 },
+	.vsync_len = { 1, 1, 1 },
+	.flags = DISPLAY_FLAGS_DE_HIGH,
+};
+
+static const struct panel_desc rocktech_rk070er9427 = {
+	.timings = &rocktech_rk070er9427_timing,
+	.num_timings = 1,
+	.bpc = 6,
+	.size = {
+		.width = 154,
+		.height = 86,
+	},
+	.delay = {
+		.prepare = 41,
+		.enable = 50,
+		.unprepare = 41,
+		.disable = 50,
+	},
+	.bus_format = MEDIA_BUS_FMT_RGB666_1X18,
+};
+
 static const struct drm_display_mode samsung_lsn122dl01_c01_mode = {
 	.clock = 271560,
 	.hdisplay = 2560,
@@ -1764,6 +1991,30 @@ static const struct panel_desc samsung_ltn140at29_301 = {
 	},
 };
 
+static const struct drm_display_mode sharp_lq035q7db03_mode = {
+	.clock = 5500,
+	.hdisplay = 240,
+	.hsync_start = 240 + 16,
+	.hsync_end = 240 + 16 + 7,
+	.htotal = 240 + 16 + 7 + 5,
+	.vdisplay = 320,
+	.vsync_start = 320 + 9,
+	.vsync_end = 320 + 9 + 1,
+	.vtotal = 320 + 9 + 1 + 7,
+	.vrefresh = 60,
+};
+
+static const struct panel_desc sharp_lq035q7db03 = {
+	.modes = &sharp_lq035q7db03_mode,
+	.num_modes = 1,
+	.bpc = 6,
+	.size = {
+		.width = 54,
+		.height = 72,
+	},
+	.bus_format = MEDIA_BUS_FMT_RGB666_1X18,
+};
+
 static const struct display_timing sharp_lq101k1ly04_timing = {
 	.pixelclock = { 60000000, 65000000, 80000000 },
 	.hactive = { 1280, 1280, 1280 },
@@ -2095,6 +2346,9 @@ static const struct of_device_id platform_of_match[] = {
 		.compatible = "auo,b133xtn01",
 		.data = &auo_b133xtn01,
 	}, {
+		.compatible = "auo,g070vvn01",
+		.data = &auo_g070vvn01,
+	}, {
 		.compatible = "auo,g104sn02",
 		.data = &auo_g104sn02,
 	}, {
@@ -2113,6 +2367,9 @@ static const struct of_device_id platform_of_match[] = {
 		.compatible = "avic,tm070ddh03",
 		.data = &avic_tm070ddh03,
 	}, {
+		.compatible = "boe,hv070wsa-100",
+		.data = &boe_hv070wsa
+	}, {
 		.compatible = "boe,nv101wxmn51",
 		.data = &boe_nv101wxmn51,
 	}, {
@@ -2125,6 +2382,12 @@ static const struct of_device_id platform_of_match[] = {
 		.compatible = "chunghwa,claa101wb01",
 		.data = &chunghwa_claa101wb01
 	}, {
+		.compatible = "dataimage,scf0700c48ggu18",
+		.data = &dataimage_scf0700c48ggu18,
+	}, {
+		.compatible = "dlc,dlc0700yzg-1",
+		.data = &dlc_dlc0700yzg_1,
+	}, {
 		.compatible = "edt,et057090dhu",
 		.data = &edt_et057090dhu,
 	}, {
@@ -2134,6 +2397,12 @@ static const struct of_device_id platform_of_match[] = {
 		.compatible = "edt,etm0700g0dh6",
 		.data = &edt_etm0700g0dh6,
 	}, {
+		.compatible = "edt,etm0700g0bdh6",
+		.data = &edt_etm0700g0bdh6,
+	}, {
+		.compatible = "edt,etm0700g0edh6",
+		.data = &edt_etm0700g0bdh6,
+	}, {
 		.compatible = "foxlink,fl500wvr00-a0t",
 		.data = &foxlink_fl500wvr00_a0t,
 	}, {
@@ -2155,10 +2424,13 @@ static const struct of_device_id platform_of_match[] = {
 		.compatible = "innolux,at070tn92",
 		.data = &innolux_at070tn92,
 	}, {
-		.compatible ="innolux,g101ice-l01",
+		.compatible = "innolux,g070y2-l01",
+		.data = &innolux_g070y2_l01,
+	}, {
+		.compatible = "innolux,g101ice-l01",
 		.data = &innolux_g101ice_l01
 	}, {
-		.compatible ="innolux,g121i1-l01",
+		.compatible = "innolux,g121i1-l01",
 		.data = &innolux_g121i1_l01
 	}, {
 		.compatible = "innolux,g121x1-l03",
@@ -2170,6 +2442,9 @@ static const struct of_device_id platform_of_match[] = {
 		.compatible = "innolux,n156bge-l21",
 		.data = &innolux_n156bge_l21,
 	}, {
+		.compatible = "innolux,tv123wam",
+		.data = &innolux_tv123wam,
+	}, {
 		.compatible = "innolux,zj070na-01p",
 		.data = &innolux_zj070na_01p,
 	}, {
@@ -2206,6 +2481,9 @@ static const struct of_device_id platform_of_match[] = {
 		.compatible = "netron-dy,e231732",
 		.data = &netron_dy_e231732,
 	}, {
+		.compatible = "newhaven,nhd-4.3-480272ef-atxl",
+		.data = &newhaven_nhd_43_480272ef_atxl,
+	}, {
 		.compatible = "nlt,nl192108ac18-02d",
 		.data = &nlt_nl192108ac18_02d,
 	}, {
@@ -2227,6 +2505,9 @@ static const struct of_device_id platform_of_match[] = {
 		.compatible = "qiaodian,qd43003c0-40",
 		.data = &qd43003c0_40,
 	}, {
+		.compatible = "rocktech,rk070er9427",
+		.data = &rocktech_rk070er9427,
+	}, {
 		.compatible = "samsung,lsn122dl01-c01",
 		.data = &samsung_lsn122dl01_c01,
 	}, {
@@ -2236,6 +2517,9 @@ static const struct of_device_id platform_of_match[] = {
 		.compatible = "samsung,ltn140at29-301",
 		.data = &samsung_ltn140at29_301,
 	}, {
+		.compatible = "sharp,lq035q7db03",
+		.data = &sharp_lq035q7db03,
+	}, {
 		.compatible = "sharp,lq101k1ly04",
 		.data = &sharp_lq101k1ly04,
 	}, {
diff --git a/drivers/gpu/drm/panel/panel-sitronix-st7789v.c b/drivers/gpu/drm/panel/panel-sitronix-st7789v.c
index 358c64ef1922..74284e5afc5d 100644
--- a/drivers/gpu/drm/panel/panel-sitronix-st7789v.c
+++ b/drivers/gpu/drm/panel/panel-sitronix-st7789v.c
@@ -419,7 +419,6 @@ static int st7789v_remove(struct spi_device *spi)
 {
 	struct st7789v *ctx = spi_get_drvdata(spi);
 
-	drm_panel_detach(&ctx->panel);
 	drm_panel_remove(&ctx->panel);
 
 	if (ctx->backlight)
diff --git a/drivers/gpu/drm/pl111/Makefile b/drivers/gpu/drm/pl111/Makefile
index 9c5e8dba8ac6..0c70f0e91d21 100644
--- a/drivers/gpu/drm/pl111/Makefile
+++ b/drivers/gpu/drm/pl111/Makefile
@@ -3,6 +3,8 @@ pl111_drm-y +=	pl111_display.o \
 		pl111_versatile.o \
 		pl111_drv.o
 
+pl111_drm-$(CONFIG_ARCH_VEXPRESS) += pl111_vexpress.o
+pl111_drm-$(CONFIG_ARCH_NOMADIK) += pl111_nomadik.o
 pl111_drm-$(CONFIG_DEBUG_FS) += pl111_debugfs.o
 
 obj-$(CONFIG_DRM_PL111) += pl111_drm.o
diff --git a/drivers/gpu/drm/pl111/pl111_display.c b/drivers/gpu/drm/pl111/pl111_display.c
index 19b0d006a54a..754f6b25f265 100644
--- a/drivers/gpu/drm/pl111/pl111_display.c
+++ b/drivers/gpu/drm/pl111/pl111_display.c
@@ -63,7 +63,7 @@ pl111_mode_valid(struct drm_crtc *crtc,
 	 * We use the pixelclock to also account for interlaced modes, the
 	 * resulting bandwidth is in bytes per second.
 	 */
-	bw = mode->clock * 1000; /* In Hz */
+	bw = mode->clock * 1000ULL; /* In Hz */
 	bw = bw * mode->hdisplay * mode->vdisplay * cpp;
 	bw = div_u64(bw, mode->htotal * mode->vtotal);
 
@@ -223,48 +223,84 @@ static void pl111_display_enable(struct drm_simple_display_pipe *pipe,
 
 	/* Hard-code TFT panel */
 	cntl = CNTL_LCDEN | CNTL_LCDTFT | CNTL_LCDVCOMP(1);
+	/* On the ST Micro variant, assume all 24 bits are connected */
+	if (priv->variant->st_bitmux_control)
+		cntl |= CNTL_ST_CDWID_24;
 
-	/* Note that the the hardware's format reader takes 'r' from
+	/*
+	 * Note that the the ARM hardware's format reader takes 'r' from
 	 * the low bit, while DRM formats list channels from high bit
-	 * to low bit as you read left to right.
+	 * to low bit as you read left to right. The ST Micro version of
+	 * the PL110 (LCDC) however uses the standard DRM format.
 	 */
 	switch (fb->format->format) {
+	case DRM_FORMAT_BGR888:
+		/* Only supported on the ST Micro variant */
+		if (priv->variant->st_bitmux_control)
+			cntl |= CNTL_ST_LCDBPP24_PACKED | CNTL_BGR;
+		break;
+	case DRM_FORMAT_RGB888:
+		/* Only supported on the ST Micro variant */
+		if (priv->variant->st_bitmux_control)
+			cntl |= CNTL_ST_LCDBPP24_PACKED;
+		break;
 	case DRM_FORMAT_ABGR8888:
 	case DRM_FORMAT_XBGR8888:
-		cntl |= CNTL_LCDBPP24;
+		if (priv->variant->st_bitmux_control)
+			cntl |= CNTL_LCDBPP24 | CNTL_BGR;
+		else
+			cntl |= CNTL_LCDBPP24;
 		break;
 	case DRM_FORMAT_ARGB8888:
 	case DRM_FORMAT_XRGB8888:
-		cntl |= CNTL_LCDBPP24 | CNTL_BGR;
+		if (priv->variant->st_bitmux_control)
+			cntl |= CNTL_LCDBPP24;
+		else
+			cntl |= CNTL_LCDBPP24 | CNTL_BGR;
 		break;
 	case DRM_FORMAT_BGR565:
 		if (priv->variant->is_pl110)
 			cntl |= CNTL_LCDBPP16;
+		else if (priv->variant->st_bitmux_control)
+			cntl |= CNTL_LCDBPP16 | CNTL_ST_1XBPP_565 | CNTL_BGR;
 		else
 			cntl |= CNTL_LCDBPP16_565;
 		break;
 	case DRM_FORMAT_RGB565:
 		if (priv->variant->is_pl110)
-			cntl |= CNTL_LCDBPP16;
+			cntl |= CNTL_LCDBPP16 | CNTL_BGR;
+		else if (priv->variant->st_bitmux_control)
+			cntl |= CNTL_LCDBPP16 | CNTL_ST_1XBPP_565;
 		else
-			cntl |= CNTL_LCDBPP16_565;
-		cntl |= CNTL_BGR;
+			cntl |= CNTL_LCDBPP16_565 | CNTL_BGR;
 		break;
 	case DRM_FORMAT_ABGR1555:
 	case DRM_FORMAT_XBGR1555:
 		cntl |= CNTL_LCDBPP16;
+		if (priv->variant->st_bitmux_control)
+			cntl |= CNTL_ST_1XBPP_5551 | CNTL_BGR;
 		break;
 	case DRM_FORMAT_ARGB1555:
 	case DRM_FORMAT_XRGB1555:
-		cntl |= CNTL_LCDBPP16 | CNTL_BGR;
+		cntl |= CNTL_LCDBPP16;
+		if (priv->variant->st_bitmux_control)
+			cntl |= CNTL_ST_1XBPP_5551;
+		else
+			cntl |= CNTL_BGR;
 		break;
 	case DRM_FORMAT_ABGR4444:
 	case DRM_FORMAT_XBGR4444:
 		cntl |= CNTL_LCDBPP16_444;
+		if (priv->variant->st_bitmux_control)
+			cntl |= CNTL_ST_1XBPP_444 | CNTL_BGR;
 		break;
 	case DRM_FORMAT_ARGB4444:
 	case DRM_FORMAT_XRGB4444:
-		cntl |= CNTL_LCDBPP16_444 | CNTL_BGR;
+		cntl |= CNTL_LCDBPP16_444;
+		if (priv->variant->st_bitmux_control)
+			cntl |= CNTL_ST_1XBPP_444;
+		else
+			cntl |= CNTL_BGR;
 		break;
 	default:
 		WARN_ONCE(true, "Unknown FB format 0x%08x\n",
diff --git a/drivers/gpu/drm/pl111/pl111_drm.h b/drivers/gpu/drm/pl111/pl111_drm.h
index 8639b2d4ddf7..1aa015ccacef 100644
--- a/drivers/gpu/drm/pl111/pl111_drm.h
+++ b/drivers/gpu/drm/pl111/pl111_drm.h
@@ -36,11 +36,14 @@ struct drm_minor;
  * struct pl111_variant_data - encodes IP differences
  * @name: the name of this variant
  * @is_pl110: this is the early PL110 variant
+ * @is_lcdc: this is the ST Microelectronics Nomadik LCDC variant
  * @external_bgr: this is the Versatile Pl110 variant with external
  *	BGR/RGB routing
  * @broken_clockdivider: the clock divider is broken and we need to
  *	use the supplied clock directly
  * @broken_vblank: the vblank IRQ is broken on this variant
+ * @st_bitmux_control: this variant is using the ST Micro bitmux
+ *	extensions to the control register
  * @formats: array of supported pixel formats on this variant
  * @nformats: the length of the array of supported pixel formats
  * @fb_bpp: desired bits per pixel on the default framebuffer
@@ -48,9 +51,11 @@ struct drm_minor;
 struct pl111_variant_data {
 	const char *name;
 	bool is_pl110;
+	bool is_lcdc;
 	bool external_bgr;
 	bool broken_clockdivider;
 	bool broken_vblank;
+	bool st_bitmux_control;
 	const u32 *formats;
 	unsigned int nformats;
 	unsigned int fb_bpp;
@@ -79,6 +84,7 @@ struct pl111_drm_dev_private {
 	const struct pl111_variant_data *variant;
 	void (*variant_display_enable) (struct drm_device *drm, u32 format);
 	void (*variant_display_disable) (struct drm_device *drm);
+	bool use_device_memory;
 };
 
 int pl111_display_init(struct drm_device *dev);
diff --git a/drivers/gpu/drm/pl111/pl111_drv.c b/drivers/gpu/drm/pl111/pl111_drv.c
index 4621259d5387..47fe30223444 100644
--- a/drivers/gpu/drm/pl111/pl111_drv.c
+++ b/drivers/gpu/drm/pl111/pl111_drv.c
@@ -60,6 +60,7 @@
 #include <linux/slab.h>
 #include <linux/of.h>
 #include <linux/of_graph.h>
+#include <linux/of_reserved_mem.h>
 
 #include <drm/drmP.h>
 #include <drm/drm_atomic_helper.h>
@@ -74,6 +75,7 @@
 
 #include "pl111_drm.h"
 #include "pl111_versatile.h"
+#include "pl111_nomadik.h"
 
 #define DRIVER_DESC      "DRM module for PL111"
 
@@ -207,6 +209,24 @@ finish:
 	return ret;
 }
 
+static struct drm_gem_object *
+pl111_gem_import_sg_table(struct drm_device *dev,
+			  struct dma_buf_attachment *attach,
+			  struct sg_table *sgt)
+{
+	struct pl111_drm_dev_private *priv = dev->dev_private;
+
+	/*
+	 * When using device-specific reserved memory we can't import
+	 * DMA buffers: those are passed by reference in any global
+	 * memory and we can only handle a specific range of memory.
+	 */
+	if (priv->use_device_memory)
+		return ERR_PTR(-EINVAL);
+
+	return drm_gem_cma_prime_import_sg_table(dev, attach, sgt);
+}
+
 DEFINE_DRM_GEM_CMA_FOPS(drm_fops);
 
 static struct drm_driver pl111_drm_driver = {
@@ -227,9 +247,11 @@ static struct drm_driver pl111_drm_driver = {
 	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
 	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
 	.gem_prime_import = drm_gem_prime_import,
-	.gem_prime_import_sg_table = drm_gem_cma_prime_import_sg_table,
+	.gem_prime_import_sg_table = pl111_gem_import_sg_table,
 	.gem_prime_export = drm_gem_prime_export,
 	.gem_prime_get_sg_table	= drm_gem_cma_prime_get_sg_table,
+	.gem_prime_mmap = drm_gem_cma_prime_mmap,
+	.gem_prime_vmap = drm_gem_cma_prime_vmap,
 
 #if defined(CONFIG_DEBUG_FS)
 	.debugfs_init = pl111_debugfs_init,
@@ -257,14 +279,20 @@ static int pl111_amba_probe(struct amba_device *amba_dev,
 	drm->dev_private = priv;
 	priv->variant = variant;
 
+	ret = of_reserved_mem_device_init(dev);
+	if (!ret) {
+		dev_info(dev, "using device-specific reserved memory\n");
+		priv->use_device_memory = true;
+	}
+
 	if (of_property_read_u32(dev->of_node, "max-memory-bandwidth",
 				 &priv->memory_bw)) {
 		dev_info(dev, "no max memory bandwidth specified, assume unlimited\n");
 		priv->memory_bw = 0;
 	}
 
-	/* The two variants swap this register */
-	if (variant->is_pl110) {
+	/* The two main variants swap this register */
+	if (variant->is_pl110 || variant->is_lcdc) {
 		priv->ienb = CLCD_PL110_IENB;
 		priv->ctrl = CLCD_PL110_CNTL;
 	} else {
@@ -275,13 +303,16 @@ static int pl111_amba_probe(struct amba_device *amba_dev,
 	priv->regs = devm_ioremap_resource(dev, &amba_dev->res);
 	if (IS_ERR(priv->regs)) {
 		dev_err(dev, "%s failed mmio\n", __func__);
-		return PTR_ERR(priv->regs);
+		ret = PTR_ERR(priv->regs);
+		goto dev_put;
 	}
 
 	/* This may override some variant settings */
 	ret = pl111_versatile_init(dev, priv);
 	if (ret)
-		goto dev_unref;
+		goto dev_put;
+
+	pl111_nomadik_init(dev);
 
 	/* turn off interrupts before requesting the irq */
 	writel(0, priv->regs + priv->ienb);
@@ -295,21 +326,24 @@ static int pl111_amba_probe(struct amba_device *amba_dev,
 
 	ret = pl111_modeset_init(drm);
 	if (ret != 0)
-		goto dev_unref;
+		goto dev_put;
 
 	ret = drm_dev_register(drm, 0);
 	if (ret < 0)
-		goto dev_unref;
+		goto dev_put;
 
 	return 0;
 
-dev_unref:
-	drm_dev_unref(drm);
+dev_put:
+	drm_dev_put(drm);
+	of_reserved_mem_device_release(dev);
+
 	return ret;
 }
 
 static int pl111_amba_remove(struct amba_device *amba_dev)
 {
+	struct device *dev = &amba_dev->dev;
 	struct drm_device *drm = amba_get_drvdata(amba_dev);
 	struct pl111_drm_dev_private *priv = drm->dev_private;
 
@@ -318,7 +352,8 @@ static int pl111_amba_remove(struct amba_device *amba_dev)
 	if (priv->panel)
 		drm_panel_bridge_remove(priv->bridge);
 	drm_mode_config_cleanup(drm);
-	drm_dev_unref(drm);
+	drm_dev_put(drm);
+	of_reserved_mem_device_release(dev);
 
 	return 0;
 }
@@ -370,16 +405,50 @@ static const struct pl111_variant_data pl111_variant = {
 	.fb_bpp = 32,
 };
 
+static const u32 pl110_nomadik_pixel_formats[] = {
+	DRM_FORMAT_RGB888,
+	DRM_FORMAT_BGR888,
+	DRM_FORMAT_ABGR8888,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_BGR565,
+	DRM_FORMAT_RGB565,
+	DRM_FORMAT_ABGR1555,
+	DRM_FORMAT_XBGR1555,
+	DRM_FORMAT_ARGB1555,
+	DRM_FORMAT_XRGB1555,
+	DRM_FORMAT_ABGR4444,
+	DRM_FORMAT_XBGR4444,
+	DRM_FORMAT_ARGB4444,
+	DRM_FORMAT_XRGB4444,
+};
+
+static const struct pl111_variant_data pl110_nomadik_variant = {
+	.name = "LCDC (PL110 Nomadik)",
+	.formats = pl110_nomadik_pixel_formats,
+	.nformats = ARRAY_SIZE(pl110_nomadik_pixel_formats),
+	.is_lcdc = true,
+	.st_bitmux_control = true,
+	.broken_vblank = true,
+	.fb_bpp = 16,
+};
+
 static const struct amba_id pl111_id_table[] = {
 	{
 		.id = 0x00041110,
 		.mask = 0x000fffff,
-		.data = (void*)&pl110_variant,
+		.data = (void *)&pl110_variant,
+	},
+	{
+		.id = 0x00180110,
+		.mask = 0x00fffffe,
+		.data = (void *)&pl110_nomadik_variant,
 	},
 	{
 		.id = 0x00041111,
 		.mask = 0x000fffff,
-		.data = (void*)&pl111_variant,
+		.data = (void *)&pl111_variant,
 	},
 	{0, 0},
 };
diff --git a/drivers/gpu/drm/pl111/pl111_nomadik.c b/drivers/gpu/drm/pl111/pl111_nomadik.c
new file mode 100644
index 000000000000..6f385e59be22
--- /dev/null
+++ b/drivers/gpu/drm/pl111/pl111_nomadik.c
@@ -0,0 +1,36 @@
+// SPDX-License-Identifier: GPL-2.0+
+#include <linux/device.h>
+#include <linux/regmap.h>
+#include <linux/mfd/syscon.h>
+#include <linux/bitops.h>
+#include <linux/module.h>
+#include "pl111_nomadik.h"
+
+#define PMU_CTRL_OFFSET 0x0000
+#define PMU_CTRL_LCDNDIF BIT(26)
+
+void pl111_nomadik_init(struct device *dev)
+{
+	struct regmap *pmu_regmap;
+
+	/*
+	 * Just bail out of this is not found, we could be running
+	 * multiplatform on something else than Nomadik.
+	 */
+	pmu_regmap =
+		syscon_regmap_lookup_by_compatible("stericsson,nomadik-pmu");
+	if (IS_ERR(pmu_regmap))
+		return;
+
+	/*
+	 * This bit in the PMU controller multiplexes the two graphics
+	 * blocks found in the Nomadik STn8815. The other one is called
+	 * MDIF (Master Display Interface) and gets muxed out here.
+	 */
+	regmap_update_bits(pmu_regmap,
+			   PMU_CTRL_OFFSET,
+			   PMU_CTRL_LCDNDIF,
+			   0);
+	dev_info(dev, "set Nomadik PMU mux to CLCD mode\n");
+}
+EXPORT_SYMBOL_GPL(pl111_nomadik_init);
diff --git a/drivers/gpu/drm/pl111/pl111_nomadik.h b/drivers/gpu/drm/pl111/pl111_nomadik.h
new file mode 100644
index 000000000000..19d663d46353
--- /dev/null
+++ b/drivers/gpu/drm/pl111/pl111_nomadik.h
@@ -0,0 +1,18 @@
+// SPDX-License-Identifier: GPL-2.0+
+#include <linux/device.h>
+
+#ifndef PL111_NOMADIK_H
+#define PL111_NOMADIK_H
+#endif
+
+#ifdef CONFIG_ARCH_NOMADIK
+
+void pl111_nomadik_init(struct device *dev);
+
+#else
+
+static inline void pl111_nomadik_init(struct device *dev)
+{
+}
+
+#endif
diff --git a/drivers/gpu/drm/pl111/pl111_versatile.c b/drivers/gpu/drm/pl111/pl111_versatile.c
index 9302f516045e..b9baefdba38a 100644
--- a/drivers/gpu/drm/pl111/pl111_versatile.c
+++ b/drivers/gpu/drm/pl111/pl111_versatile.c
@@ -1,12 +1,14 @@
 #include <linux/amba/clcd-regs.h>
 #include <linux/device.h>
 #include <linux/of.h>
+#include <linux/of_platform.h>
 #include <linux/regmap.h>
 #include <linux/mfd/syscon.h>
 #include <linux/bitops.h>
 #include <linux/module.h>
 #include <drm/drmP.h>
 #include "pl111_versatile.h"
+#include "pl111_vexpress.h"
 #include "pl111_drm.h"
 
 static struct regmap *versatile_syscon_map;
@@ -22,6 +24,7 @@ enum versatile_clcd {
 	REALVIEW_CLCD_PB11MP,
 	REALVIEW_CLCD_PBA8,
 	REALVIEW_CLCD_PBX,
+	VEXPRESS_CLCD_V2M,
 };
 
 static const struct of_device_id versatile_clcd_of_match[] = {
@@ -53,6 +56,10 @@ static const struct of_device_id versatile_clcd_of_match[] = {
 		.compatible = "arm,realview-pbx-syscon",
 		.data = (void *)REALVIEW_CLCD_PBX,
 	},
+	{
+		.compatible = "arm,vexpress-muxfpga",
+		.data = (void *)VEXPRESS_CLCD_V2M,
+	},
 	{},
 };
 
@@ -286,12 +293,26 @@ static const struct pl111_variant_data pl111_realview = {
 	.fb_bpp = 16,
 };
 
+/*
+ * Versatile Express PL111 variant, again we just push the maximum
+ * BPP to 16 to be able to get 1024x768 without saturating the memory
+ * bus. The clockdivider also seems broken on the Versatile Express.
+ */
+static const struct pl111_variant_data pl111_vexpress = {
+	.name = "PL111 Versatile Express",
+	.formats = pl111_realview_pixel_formats,
+	.nformats = ARRAY_SIZE(pl111_realview_pixel_formats),
+	.fb_bpp = 16,
+	.broken_clockdivider = true,
+};
+
 int pl111_versatile_init(struct device *dev, struct pl111_drm_dev_private *priv)
 {
 	const struct of_device_id *clcd_id;
 	enum versatile_clcd versatile_clcd_type;
 	struct device_node *np;
 	struct regmap *map;
+	int ret;
 
 	np = of_find_matching_node_and_match(NULL, versatile_clcd_of_match,
 					     &clcd_id);
@@ -301,7 +322,33 @@ int pl111_versatile_init(struct device *dev, struct pl111_drm_dev_private *priv)
 	}
 	versatile_clcd_type = (enum versatile_clcd)clcd_id->data;
 
-	map = syscon_node_to_regmap(np);
+	/* Versatile Express special handling */
+	if (versatile_clcd_type == VEXPRESS_CLCD_V2M) {
+		struct platform_device *pdev;
+
+		/* Registers a driver for the muxfpga */
+		ret = vexpress_muxfpga_init();
+		if (ret) {
+			dev_err(dev, "unable to initialize muxfpga driver\n");
+			return ret;
+		}
+
+		/* Call into deep Vexpress configuration API */
+		pdev = of_find_device_by_node(np);
+		if (!pdev) {
+			dev_err(dev, "can't find the sysreg device, deferring\n");
+			return -EPROBE_DEFER;
+		}
+		map = dev_get_drvdata(&pdev->dev);
+		if (!map) {
+			dev_err(dev, "sysreg has not yet probed\n");
+			platform_device_put(pdev);
+			return -EPROBE_DEFER;
+		}
+	} else {
+		map = syscon_node_to_regmap(np);
+	}
+
 	if (IS_ERR(map)) {
 		dev_err(dev, "no Versatile syscon regmap\n");
 		return PTR_ERR(map);
@@ -340,6 +387,13 @@ int pl111_versatile_init(struct device *dev, struct pl111_drm_dev_private *priv)
 		priv->variant_display_disable = pl111_realview_clcd_disable;
 		dev_info(dev, "set up callbacks for RealView PL111\n");
 		break;
+	case VEXPRESS_CLCD_V2M:
+		priv->variant = &pl111_vexpress;
+		dev_info(dev, "initializing Versatile Express PL111\n");
+		ret = pl111_vexpress_clcd_init(dev, priv, map);
+		if (ret)
+			return ret;
+		break;
 	default:
 		dev_info(dev, "unknown Versatile system controller\n");
 		break;
diff --git a/drivers/gpu/drm/pl111/pl111_vexpress.c b/drivers/gpu/drm/pl111/pl111_vexpress.c
new file mode 100644
index 000000000000..a534b225e31b
--- /dev/null
+++ b/drivers/gpu/drm/pl111/pl111_vexpress.c
@@ -0,0 +1,134 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Versatile Express PL111 handling
+ * Copyright (C) 2018 Linus Walleij
+ *
+ * This module binds to the "arm,vexpress-muxfpga" device on the
+ * Versatile Express configuration bus and sets up which CLCD instance
+ * gets muxed out on the DVI bridge.
+ */
+#include <linux/device.h>
+#include <linux/module.h>
+#include <linux/regmap.h>
+#include <linux/vexpress.h>
+#include <linux/platform_device.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_platform.h>
+#include "pl111_drm.h"
+#include "pl111_vexpress.h"
+
+#define VEXPRESS_FPGAMUX_MOTHERBOARD		0x00
+#define VEXPRESS_FPGAMUX_DAUGHTERBOARD_1	0x01
+#define VEXPRESS_FPGAMUX_DAUGHTERBOARD_2	0x02
+
+int pl111_vexpress_clcd_init(struct device *dev,
+			     struct pl111_drm_dev_private *priv,
+			     struct regmap *map)
+{
+	struct device_node *root;
+	struct device_node *child;
+	struct device_node *ct_clcd = NULL;
+	bool has_coretile_clcd = false;
+	bool has_coretile_hdlcd = false;
+	bool mux_motherboard = true;
+	u32 val;
+	int ret;
+
+	/*
+	 * Check if we have a CLCD or HDLCD on the core tile by checking if a
+	 * CLCD or HDLCD is available in the root of the device tree.
+	 */
+	root = of_find_node_by_path("/");
+	if (!root)
+		return -EINVAL;
+
+	for_each_available_child_of_node(root, child) {
+		if (of_device_is_compatible(child, "arm,pl111")) {
+			has_coretile_clcd = true;
+			ct_clcd = child;
+			break;
+		}
+		if (of_device_is_compatible(child, "arm,hdlcd")) {
+			has_coretile_hdlcd = true;
+			break;
+		}
+	}
+
+	/*
+	 * If there is a coretile HDLCD and it has a driver,
+	 * do not mux the CLCD on the motherboard to the DVI.
+	 */
+	if (has_coretile_hdlcd && IS_ENABLED(CONFIG_DRM_HDLCD))
+		mux_motherboard = false;
+
+	/*
+	 * On the Vexpress CA9 we let the CLCD on the coretile
+	 * take precedence, so also in this case do not mux the
+	 * motherboard to the DVI.
+	 */
+	if (has_coretile_clcd)
+		mux_motherboard = false;
+
+	if (mux_motherboard) {
+		dev_info(dev, "DVI muxed to motherboard CLCD\n");
+		val = VEXPRESS_FPGAMUX_MOTHERBOARD;
+	} else if (ct_clcd == dev->of_node) {
+		dev_info(dev,
+			 "DVI muxed to daughterboard 1 (core tile) CLCD\n");
+		val = VEXPRESS_FPGAMUX_DAUGHTERBOARD_1;
+	} else {
+		dev_info(dev, "core tile graphics present\n");
+		dev_info(dev, "this device will be deactivated\n");
+		return -ENODEV;
+	}
+
+	ret = regmap_write(map, 0, val);
+	if (ret) {
+		dev_err(dev, "error setting DVI muxmode\n");
+		return -ENODEV;
+	}
+
+	return 0;
+}
+
+/*
+ * This sets up the regmap pointer that will then be retrieved by
+ * the detection code in pl111_versatile.c and passed in to the
+ * pl111_vexpress_clcd_init() function above.
+ */
+static int vexpress_muxfpga_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct regmap *map;
+
+	map = devm_regmap_init_vexpress_config(&pdev->dev);
+	if (IS_ERR(map))
+		return PTR_ERR(map);
+	dev_set_drvdata(dev, map);
+
+	return 0;
+}
+
+static const struct of_device_id vexpress_muxfpga_match[] = {
+	{ .compatible = "arm,vexpress-muxfpga", }
+};
+
+static struct platform_driver vexpress_muxfpga_driver = {
+	.driver = {
+		.name = "vexpress-muxfpga",
+		.of_match_table = of_match_ptr(vexpress_muxfpga_match),
+	},
+	.probe = vexpress_muxfpga_probe,
+};
+
+int vexpress_muxfpga_init(void)
+{
+	int ret;
+
+	ret = platform_driver_register(&vexpress_muxfpga_driver);
+	/* -EBUSY just means this driver is already registered */
+	if (ret == -EBUSY)
+		ret = 0;
+	return ret;
+}
diff --git a/drivers/gpu/drm/pl111/pl111_vexpress.h b/drivers/gpu/drm/pl111/pl111_vexpress.h
new file mode 100644
index 000000000000..5d3681bb4c00
--- /dev/null
+++ b/drivers/gpu/drm/pl111/pl111_vexpress.h
@@ -0,0 +1,29 @@
+// SPDX-License-Identifier: GPL-2.0
+
+struct device;
+struct pl111_drm_dev_private;
+struct regmap;
+
+#ifdef CONFIG_ARCH_VEXPRESS
+
+int pl111_vexpress_clcd_init(struct device *dev,
+			     struct pl111_drm_dev_private *priv,
+			     struct regmap *map);
+
+int vexpress_muxfpga_init(void);
+
+#else
+
+static inline int pl111_vexpress_clcd_init(struct device *dev,
+					   struct pl111_drm_dev_private *priv,
+					   struct regmap *map)
+{
+	return -ENODEV;
+}
+
+static inline int vexpress_muxfpga_init(void)
+{
+	return 0;
+}
+
+#endif
diff --git a/drivers/gpu/drm/qxl/qxl_cmd.c b/drivers/gpu/drm/qxl/qxl_cmd.c
index 01665b98c57e..208af9f37914 100644
--- a/drivers/gpu/drm/qxl/qxl_cmd.c
+++ b/drivers/gpu/drm/qxl/qxl_cmd.c
@@ -339,12 +339,9 @@ int qxl_io_update_area(struct qxl_device *qdev, struct qxl_bo *surf,
 	surface_height = surf->surf.height;
 
 	if (area->left < 0 || area->top < 0 ||
-	    area->right > surface_width || area->bottom > surface_height) {
-		qxl_io_log(qdev, "%s: not doing area update for "
-			   "%d, (%d,%d,%d,%d) (%d,%d)\n", __func__, surface_id, area->left,
-			   area->top, area->right, area->bottom, surface_width, surface_height);
+	    area->right > surface_width || area->bottom > surface_height)
 		return -EINVAL;
-	}
+
 	mutex_lock(&qdev->update_area_mutex);
 	qdev->ram_header->update_area = *area;
 	qdev->ram_header->update_surface = surface_id;
@@ -372,6 +369,7 @@ void qxl_io_flush_surfaces(struct qxl_device *qdev)
 void qxl_io_destroy_primary(struct qxl_device *qdev)
 {
 	wait_for_io_cmd(qdev, 0, QXL_IO_DESTROY_PRIMARY_ASYNC);
+	qdev->primary_created = false;
 }
 
 void qxl_io_create_primary(struct qxl_device *qdev,
@@ -397,6 +395,7 @@ void qxl_io_create_primary(struct qxl_device *qdev,
 	create->type = QXL_SURF_TYPE_PRIMARY;
 
 	wait_for_io_cmd(qdev, 0, QXL_IO_CREATE_PRIMARY_ASYNC);
+	qdev->primary_created = true;
 }
 
 void qxl_io_memslot_add(struct qxl_device *qdev, uint8_t id)
@@ -405,20 +404,6 @@ void qxl_io_memslot_add(struct qxl_device *qdev, uint8_t id)
 	wait_for_io_cmd(qdev, id, QXL_IO_MEMSLOT_ADD_ASYNC);
 }
 
-void qxl_io_log(struct qxl_device *qdev, const char *fmt, ...)
-{
-	va_list args;
-
-	va_start(args, fmt);
-	vsnprintf(qdev->ram_header->log_buf, QXL_LOG_BUF_SIZE, fmt, args);
-	va_end(args);
-	/*
-	 * DO not do a DRM output here - this will call printk, which will
-	 * call back into qxl for rendering (qxl_fb)
-	 */
-	outb(0, qdev->io_base + QXL_IO_LOG);
-}
-
 void qxl_io_reset(struct qxl_device *qdev)
 {
 	outb(0, qdev->io_base + QXL_IO_RESET);
@@ -426,19 +411,6 @@ void qxl_io_reset(struct qxl_device *qdev)
 
 void qxl_io_monitors_config(struct qxl_device *qdev)
 {
-	qxl_io_log(qdev, "%s: %d [%dx%d+%d+%d]\n", __func__,
-		   qdev->monitors_config ?
-		   qdev->monitors_config->count : -1,
-		   qdev->monitors_config && qdev->monitors_config->count ?
-		   qdev->monitors_config->heads[0].width : -1,
-		   qdev->monitors_config && qdev->monitors_config->count ?
-		   qdev->monitors_config->heads[0].height : -1,
-		   qdev->monitors_config && qdev->monitors_config->count ?
-		   qdev->monitors_config->heads[0].x : -1,
-		   qdev->monitors_config && qdev->monitors_config->count ?
-		   qdev->monitors_config->heads[0].y : -1
-		   );
-
 	wait_for_io_cmd(qdev, 0, QXL_IO_MONITORS_CONFIG_ASYNC);
 }
 
diff --git a/drivers/gpu/drm/qxl/qxl_display.c b/drivers/gpu/drm/qxl/qxl_display.c
index 820cbca3bf6e..0570c6826bff 100644
--- a/drivers/gpu/drm/qxl/qxl_display.c
+++ b/drivers/gpu/drm/qxl/qxl_display.c
@@ -48,12 +48,8 @@ static void qxl_alloc_client_monitors_config(struct qxl_device *qdev, unsigned c
 		qdev->client_monitors_config = kzalloc(
 				sizeof(struct qxl_monitors_config) +
 				sizeof(struct qxl_head) * count, GFP_KERNEL);
-		if (!qdev->client_monitors_config) {
-			qxl_io_log(qdev,
-				   "%s: allocation failure for %u heads\n",
-				   __func__, count);
+		if (!qdev->client_monitors_config)
 			return;
-		}
 	}
 	qdev->client_monitors_config->count = count;
 }
@@ -74,12 +70,8 @@ static int qxl_display_copy_rom_client_monitors_config(struct qxl_device *qdev)
 	num_monitors = qdev->rom->client_monitors_config.count;
 	crc = crc32(0, (const uint8_t *)&qdev->rom->client_monitors_config,
 		  sizeof(qdev->rom->client_monitors_config));
-	if (crc != qdev->rom->client_monitors_config_crc) {
-		qxl_io_log(qdev, "crc mismatch: have %X (%zd) != %X\n", crc,
-			   sizeof(qdev->rom->client_monitors_config),
-			   qdev->rom->client_monitors_config_crc);
+	if (crc != qdev->rom->client_monitors_config_crc)
 		return MONITORS_CONFIG_BAD_CRC;
-	}
 	if (!num_monitors) {
 		DRM_DEBUG_KMS("no client monitors configured\n");
 		return status;
@@ -170,12 +162,10 @@ void qxl_display_read_client_monitors_config(struct qxl_device *qdev)
 		udelay(5);
 	}
 	if (status == MONITORS_CONFIG_BAD_CRC) {
-		qxl_io_log(qdev, "config: bad crc\n");
 		DRM_DEBUG_KMS("ignoring client monitors config: bad crc");
 		return;
 	}
 	if (status == MONITORS_CONFIG_UNCHANGED) {
-		qxl_io_log(qdev, "config: unchanged\n");
 		DRM_DEBUG_KMS("ignoring client monitors config: unchanged");
 		return;
 	}
@@ -268,6 +258,89 @@ static int qxl_add_common_modes(struct drm_connector *connector,
 	return i - 1;
 }
 
+static void qxl_send_monitors_config(struct qxl_device *qdev)
+{
+	int i;
+
+	BUG_ON(!qdev->ram_header->monitors_config);
+
+	if (qdev->monitors_config->count == 0)
+		return;
+
+	for (i = 0 ; i < qdev->monitors_config->count ; ++i) {
+		struct qxl_head *head = &qdev->monitors_config->heads[i];
+
+		if (head->y > 8192 || head->x > 8192 ||
+		    head->width > 8192 || head->height > 8192) {
+			DRM_ERROR("head %d wrong: %dx%d+%d+%d\n",
+				  i, head->width, head->height,
+				  head->x, head->y);
+			return;
+		}
+	}
+	qxl_io_monitors_config(qdev);
+}
+
+static void qxl_crtc_update_monitors_config(struct drm_crtc *crtc,
+					    const char *reason)
+{
+	struct drm_device *dev = crtc->dev;
+	struct qxl_device *qdev = dev->dev_private;
+	struct qxl_crtc *qcrtc = to_qxl_crtc(crtc);
+	struct qxl_head head;
+	int oldcount, i = qcrtc->index;
+
+	if (!qdev->primary_created) {
+		DRM_DEBUG_KMS("no primary surface, skip (%s)\n", reason);
+		return;
+	}
+
+	if (!qdev->monitors_config ||
+	    qdev->monitors_config->max_allowed <= i)
+		return;
+
+	head.id = i;
+	head.flags = 0;
+	oldcount = qdev->monitors_config->count;
+	if (crtc->state->active) {
+		struct drm_display_mode *mode = &crtc->mode;
+		head.width = mode->hdisplay;
+		head.height = mode->vdisplay;
+		head.x = crtc->x;
+		head.y = crtc->y;
+		if (qdev->monitors_config->count < i + 1)
+			qdev->monitors_config->count = i + 1;
+	} else if (i > 0) {
+		head.width = 0;
+		head.height = 0;
+		head.x = 0;
+		head.y = 0;
+		if (qdev->monitors_config->count == i + 1)
+			qdev->monitors_config->count = i;
+	} else {
+		DRM_DEBUG_KMS("inactive head 0, skip (%s)\n", reason);
+		return;
+	}
+
+	if (head.width  == qdev->monitors_config->heads[i].width  &&
+	    head.height == qdev->monitors_config->heads[i].height &&
+	    head.x      == qdev->monitors_config->heads[i].x      &&
+	    head.y      == qdev->monitors_config->heads[i].y      &&
+	    oldcount    == qdev->monitors_config->count)
+		return;
+
+	DRM_DEBUG_KMS("head %d, %dx%d, at +%d+%d, %s (%s)\n",
+		      i, head.width, head.height, head.x, head.y,
+		      crtc->state->active ? "on" : "off", reason);
+	if (oldcount != qdev->monitors_config->count)
+		DRM_DEBUG_KMS("active heads %d -> %d (%d total)\n",
+			      oldcount, qdev->monitors_config->count,
+			      qdev->monitors_config->max_allowed);
+
+	qdev->monitors_config->heads[i] = head;
+	qxl_send_monitors_config(qdev);
+}
+
 static void qxl_crtc_atomic_flush(struct drm_crtc *crtc,
 				  struct drm_crtc_state *old_crtc_state)
 {
@@ -283,6 +356,8 @@ static void qxl_crtc_atomic_flush(struct drm_crtc *crtc,
 		drm_crtc_send_vblank_event(crtc, event);
 		spin_unlock_irqrestore(&dev->event_lock, flags);
 	}
+
+	qxl_crtc_update_monitors_config(crtc, "flush");
 }
 
 static void qxl_crtc_destroy(struct drm_crtc *crtc)
@@ -381,95 +456,19 @@ qxl_framebuffer_init(struct drm_device *dev,
 	return 0;
 }
 
-static bool qxl_crtc_mode_fixup(struct drm_crtc *crtc,
-				  const struct drm_display_mode *mode,
-				  struct drm_display_mode *adjusted_mode)
-{
-	struct drm_device *dev = crtc->dev;
-	struct qxl_device *qdev = dev->dev_private;
-
-	qxl_io_log(qdev, "%s: (%d,%d) => (%d,%d)\n",
-		   __func__,
-		   mode->hdisplay, mode->vdisplay,
-		   adjusted_mode->hdisplay,
-		   adjusted_mode->vdisplay);
-	return true;
-}
-
-static void
-qxl_send_monitors_config(struct qxl_device *qdev)
-{
-	int i;
-
-	BUG_ON(!qdev->ram_header->monitors_config);
-
-	if (qdev->monitors_config->count == 0) {
-		qxl_io_log(qdev, "%s: 0 monitors??\n", __func__);
-		return;
-	}
-	for (i = 0 ; i < qdev->monitors_config->count ; ++i) {
-		struct qxl_head *head = &qdev->monitors_config->heads[i];
-
-		if (head->y > 8192 || head->x > 8192 ||
-		    head->width > 8192 || head->height > 8192) {
-			DRM_ERROR("head %d wrong: %dx%d+%d+%d\n",
-				  i, head->width, head->height,
-				  head->x, head->y);
-			return;
-		}
-	}
-	qxl_io_monitors_config(qdev);
-}
-
-static void qxl_monitors_config_set(struct qxl_device *qdev,
-				    int index,
-				    unsigned x, unsigned y,
-				    unsigned width, unsigned height,
-				    unsigned surf_id)
-{
-	DRM_DEBUG_KMS("%d:%dx%d+%d+%d\n", index, width, height, x, y);
-	qdev->monitors_config->heads[index].x = x;
-	qdev->monitors_config->heads[index].y = y;
-	qdev->monitors_config->heads[index].width = width;
-	qdev->monitors_config->heads[index].height = height;
-	qdev->monitors_config->heads[index].surface_id = surf_id;
-
-}
-
-static void qxl_mode_set_nofb(struct drm_crtc *crtc)
-{
-	struct qxl_device *qdev = crtc->dev->dev_private;
-	struct qxl_crtc *qcrtc = to_qxl_crtc(crtc);
-	struct drm_display_mode *mode = &crtc->mode;
-
-	DRM_DEBUG("Mode set (%d,%d)\n",
-		  mode->hdisplay, mode->vdisplay);
-
-	qxl_monitors_config_set(qdev, qcrtc->index, 0, 0,
-				mode->hdisplay,	mode->vdisplay, 0);
-
-}
-
 static void qxl_crtc_atomic_enable(struct drm_crtc *crtc,
 				   struct drm_crtc_state *old_state)
 {
-	DRM_DEBUG("\n");
+	qxl_crtc_update_monitors_config(crtc, "enable");
 }
 
 static void qxl_crtc_atomic_disable(struct drm_crtc *crtc,
 				    struct drm_crtc_state *old_state)
 {
-	struct qxl_crtc *qcrtc = to_qxl_crtc(crtc);
-	struct qxl_device *qdev = crtc->dev->dev_private;
-
-	qxl_monitors_config_set(qdev, qcrtc->index, 0, 0, 0, 0, 0);
-
-	qxl_send_monitors_config(qdev);
+	qxl_crtc_update_monitors_config(crtc, "disable");
 }
 
 static const struct drm_crtc_helper_funcs qxl_crtc_helper_funcs = {
-	.mode_fixup = qxl_crtc_mode_fixup,
-	.mode_set_nofb = qxl_mode_set_nofb,
 	.atomic_flush = qxl_crtc_atomic_flush,
 	.atomic_enable = qxl_crtc_atomic_enable,
 	.atomic_disable = qxl_crtc_atomic_disable,
@@ -613,12 +612,6 @@ static void qxl_primary_atomic_disable(struct drm_plane *plane,
 	}
 }
 
-static int qxl_plane_atomic_check(struct drm_plane *plane,
-				  struct drm_plane_state *state)
-{
-	return 0;
-}
-
 static void qxl_cursor_atomic_update(struct drm_plane *plane,
 				     struct drm_plane_state *old_state)
 {
@@ -630,7 +623,7 @@ static void qxl_cursor_atomic_update(struct drm_plane *plane,
 	struct qxl_cursor_cmd *cmd;
 	struct qxl_cursor *cursor;
 	struct drm_gem_object *obj;
-	struct qxl_bo *cursor_bo = NULL, *user_bo = NULL;
+	struct qxl_bo *cursor_bo = NULL, *user_bo = NULL, *old_cursor_bo = NULL;
 	int ret;
 	void *user_ptr;
 	int size = 64*64*4;
@@ -684,7 +677,7 @@ static void qxl_cursor_atomic_update(struct drm_plane *plane,
 							   cursor_bo, 0);
 		cmd->type = QXL_CURSOR_SET;
 
-		qxl_bo_unref(&qcrtc->cursor_bo);
+		old_cursor_bo = qcrtc->cursor_bo;
 		qcrtc->cursor_bo = cursor_bo;
 		cursor_bo = NULL;
 	} else {
@@ -704,6 +697,9 @@ static void qxl_cursor_atomic_update(struct drm_plane *plane,
 	qxl_push_cursor_ring_release(qdev, release, QXL_CMD_CURSOR, false);
 	qxl_release_fence_buffer_objects(release);
 
+	if (old_cursor_bo)
+		qxl_bo_unref(&old_cursor_bo);
+
 	qxl_bo_unref(&cursor_bo);
 
 	return;
@@ -824,7 +820,6 @@ static const uint32_t qxl_cursor_plane_formats[] = {
 };
 
 static const struct drm_plane_helper_funcs qxl_cursor_helper_funcs = {
-	.atomic_check = qxl_plane_atomic_check,
 	.atomic_update = qxl_cursor_atomic_update,
 	.atomic_disable = qxl_cursor_atomic_disable,
 	.prepare_fb = qxl_plane_prepare_fb,
@@ -949,81 +944,6 @@ free_mem:
 	return r;
 }
 
-static void qxl_enc_dpms(struct drm_encoder *encoder, int mode)
-{
-	DRM_DEBUG("\n");
-}
-
-static void qxl_enc_prepare(struct drm_encoder *encoder)
-{
-	DRM_DEBUG("\n");
-}
-
-static void qxl_write_monitors_config_for_encoder(struct qxl_device *qdev,
-		struct drm_encoder *encoder)
-{
-	int i;
-	struct qxl_output *output = drm_encoder_to_qxl_output(encoder);
-	struct qxl_head *head;
-	struct drm_display_mode *mode;
-
-	BUG_ON(!encoder);
-	/* TODO: ugly, do better */
-	i = output->index;
-	if (!qdev->monitors_config ||
-	    qdev->monitors_config->max_allowed <= i) {
-		DRM_ERROR(
-		"head number too large or missing monitors config: %p, %d",
-		qdev->monitors_config,
-		qdev->monitors_config ?
-			qdev->monitors_config->max_allowed : -1);
-		return;
-	}
-	if (!encoder->crtc) {
-		DRM_ERROR("missing crtc on encoder %p\n", encoder);
-		return;
-	}
-	if (i != 0)
-		DRM_DEBUG("missing for multiple monitors: no head holes\n");
-	head = &qdev->monitors_config->heads[i];
-	head->id = i;
-	if (encoder->crtc->enabled) {
-		mode = &encoder->crtc->mode;
-		head->width = mode->hdisplay;
-		head->height = mode->vdisplay;
-		head->x = encoder->crtc->x;
-		head->y = encoder->crtc->y;
-		if (qdev->monitors_config->count < i + 1)
-			qdev->monitors_config->count = i + 1;
-	} else {
-		head->width = 0;
-		head->height = 0;
-		head->x = 0;
-		head->y = 0;
-	}
-	DRM_DEBUG_KMS("setting head %d to +%d+%d %dx%d out of %d\n",
-		      i, head->x, head->y, head->width, head->height, qdev->monitors_config->count);
-	head->flags = 0;
-	/* TODO - somewhere else to call this for multiple monitors
-	 * (config_commit?) */
-	qxl_send_monitors_config(qdev);
-}
-
-static void qxl_enc_commit(struct drm_encoder *encoder)
-{
-	struct qxl_device *qdev = encoder->dev->dev_private;
-
-	qxl_write_monitors_config_for_encoder(qdev, encoder);
-	DRM_DEBUG("\n");
-}
-
-static void qxl_enc_mode_set(struct drm_encoder *encoder,
-				struct drm_display_mode *mode,
-				struct drm_display_mode *adjusted_mode)
-{
-	DRM_DEBUG("\n");
-}
-
 static int qxl_conn_get_modes(struct drm_connector *connector)
 {
 	unsigned pwidth = 1024;
@@ -1069,10 +989,6 @@ static struct drm_encoder *qxl_best_encoder(struct drm_connector *connector)
 
 
 static const struct drm_encoder_helper_funcs qxl_enc_helper_funcs = {
-	.dpms = qxl_enc_dpms,
-	.prepare = qxl_enc_prepare,
-	.mode_set = qxl_enc_mode_set,
-	.commit = qxl_enc_commit,
 };
 
 static const struct drm_connector_helper_funcs qxl_connector_helper_funcs = {
@@ -1100,21 +1016,11 @@ static enum drm_connector_status qxl_conn_detect(
 		     qxl_head_enabled(&qdev->client_monitors_config->heads[output->index]);
 
 	DRM_DEBUG("#%d connected: %d\n", output->index, connected);
-	if (!connected)
-		qxl_monitors_config_set(qdev, output->index, 0, 0, 0, 0, 0);
 
 	return connected ? connector_status_connected
 			 : connector_status_disconnected;
 }
 
-static int qxl_conn_set_property(struct drm_connector *connector,
-				   struct drm_property *property,
-				   uint64_t value)
-{
-	DRM_DEBUG("\n");
-	return 0;
-}
-
 static void qxl_conn_destroy(struct drm_connector *connector)
 {
 	struct qxl_output *qxl_output =
@@ -1129,7 +1035,6 @@ static const struct drm_connector_funcs qxl_connector_funcs = {
 	.dpms = drm_helper_connector_dpms,
 	.detect = qxl_conn_detect,
 	.fill_modes = drm_helper_probe_single_connector_modes,
-	.set_property = qxl_conn_set_property,
 	.destroy = qxl_conn_destroy,
 	.reset = drm_atomic_helper_connector_reset,
 	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
@@ -1181,7 +1086,7 @@ static int qdev_output_init(struct drm_device *dev, int num_output)
 	/* we get HPD via client monitors config */
 	connector->polled = DRM_CONNECTOR_POLL_HPD;
 	encoder->possible_crtcs = 1 << num_output;
-	drm_mode_connector_attach_encoder(&qxl_output->base,
+	drm_connector_attach_encoder(&qxl_output->base,
 					  &qxl_output->enc);
 	drm_encoder_helper_add(encoder, &qxl_enc_helper_funcs);
 	drm_connector_helper_add(connector, &qxl_connector_helper_funcs);
diff --git a/drivers/gpu/drm/qxl/qxl_drv.h b/drivers/gpu/drm/qxl/qxl_drv.h
index 864b456080c4..01220d386b0a 100644
--- a/drivers/gpu/drm/qxl/qxl_drv.h
+++ b/drivers/gpu/drm/qxl/qxl_drv.h
@@ -299,9 +299,6 @@ struct qxl_device {
 	int monitors_config_height;
 };
 
-/* forward declaration for QXL_INFO_IO */
-__printf(2,3) void qxl_io_log(struct qxl_device *qdev, const char *fmt, ...);
-
 extern const struct drm_ioctl_desc qxl_ioctls[];
 extern int qxl_max_ioctl;
 
diff --git a/drivers/gpu/drm/qxl/qxl_fb.c b/drivers/gpu/drm/qxl/qxl_fb.c
index 338891401f35..ca465c0d49fa 100644
--- a/drivers/gpu/drm/qxl/qxl_fb.c
+++ b/drivers/gpu/drm/qxl/qxl_fb.c
@@ -185,8 +185,6 @@ static int qxlfb_framebuffer_dirty(struct drm_framebuffer *fb,
 	/*
 	 * we are using a shadow draw buffer, at qdev->surface0_shadow
 	 */
-	qxl_io_log(qdev, "dirty x[%d, %d], y[%d, %d]\n", clips->x1, clips->x2,
-		   clips->y1, clips->y2);
 	image->dx = clips->x1;
 	image->dy = clips->y1;
 	image->width = clips->x2 - clips->x1;
@@ -243,7 +241,7 @@ static int qxlfb_create(struct qxl_fbdev *qfbdev,
 	DRM_DEBUG_DRIVER("%dx%d %d\n", mode_cmd.width,
 			 mode_cmd.height, mode_cmd.pitches[0]);
 
-	shadow = vmalloc(mode_cmd.pitches[0] * mode_cmd.height);
+	shadow = vmalloc(array_size(mode_cmd.pitches[0], mode_cmd.height));
 	/* TODO: what's the usual response to memory allocation errors? */
 	BUG_ON(!shadow);
 	DRM_DEBUG_DRIVER("surface0 at gpu offset %lld, mmap_offset %lld (virt %p, shadow %p)\n",
diff --git a/drivers/gpu/drm/qxl/qxl_irq.c b/drivers/gpu/drm/qxl/qxl_irq.c
index 23a40106ab53..3bb31add6350 100644
--- a/drivers/gpu/drm/qxl/qxl_irq.c
+++ b/drivers/gpu/drm/qxl/qxl_irq.c
@@ -57,10 +57,9 @@ irqreturn_t qxl_irq_handler(int irq, void *arg)
 		 * to avoid endless loops).
 		 */
 		qdev->irq_received_error++;
-		qxl_io_log(qdev, "%s: driver is in bug mode.\n", __func__);
+		DRM_WARN("driver is in bug mode\n");
 	}
 	if (pending & QXL_INTERRUPT_CLIENT_MONITORS_CONFIG) {
-		qxl_io_log(qdev, "QXL_INTERRUPT_CLIENT_MONITORS_CONFIG\n");
 		schedule_work(&qdev->client_monitors_config_work);
 	}
 	qdev->ram_header->int_mask = QXL_INTERRUPT_MASK;
diff --git a/drivers/gpu/drm/qxl/qxl_kms.c b/drivers/gpu/drm/qxl/qxl_kms.c
index c5716a0ca3b8..771250aed78d 100644
--- a/drivers/gpu/drm/qxl/qxl_kms.c
+++ b/drivers/gpu/drm/qxl/qxl_kms.c
@@ -200,8 +200,8 @@ int qxl_device_init(struct qxl_device *qdev,
 		(~(uint64_t)0) >> (qdev->slot_id_bits + qdev->slot_gen_bits);
 
 	qdev->mem_slots =
-		kmalloc(qdev->n_mem_slots * sizeof(struct qxl_memslot),
-			GFP_KERNEL);
+		kmalloc_array(qdev->n_mem_slots, sizeof(struct qxl_memslot),
+			      GFP_KERNEL);
 
 	idr_init(&qdev->release_idr);
 	spin_lock_init(&qdev->release_idr_lock);
diff --git a/drivers/gpu/drm/qxl/qxl_release.c b/drivers/gpu/drm/qxl/qxl_release.c
index 7cb214577275..e37f0097f744 100644
--- a/drivers/gpu/drm/qxl/qxl_release.c
+++ b/drivers/gpu/drm/qxl/qxl_release.c
@@ -50,12 +50,6 @@ static const char *qxl_get_timeline_name(struct dma_fence *fence)
 	return "release";
 }
 
-static bool qxl_nop_signaling(struct dma_fence *fence)
-{
-	/* fences are always automatically signaled, so just pretend we did this.. */
-	return true;
-}
-
 static long qxl_fence_wait(struct dma_fence *fence, bool intr,
 			   signed long timeout)
 {
@@ -119,7 +113,6 @@ signaled:
 static const struct dma_fence_ops qxl_fence_ops = {
 	.get_driver_name = qxl_get_driver_name,
 	.get_timeline_name = qxl_get_timeline_name,
-	.enable_signaling = qxl_nop_signaling,
 	.wait = qxl_fence_wait,
 };
 
diff --git a/drivers/gpu/drm/qxl/qxl_ttm.c b/drivers/gpu/drm/qxl/qxl_ttm.c
index ee2340e31f06..86a1fb32f6db 100644
--- a/drivers/gpu/drm/qxl/qxl_ttm.c
+++ b/drivers/gpu/drm/qxl/qxl_ttm.c
@@ -105,16 +105,16 @@ static void qxl_ttm_global_fini(struct qxl_device *qdev)
 static struct vm_operations_struct qxl_ttm_vm_ops;
 static const struct vm_operations_struct *ttm_vm_ops;
 
-static int qxl_ttm_fault(struct vm_fault *vmf)
+static vm_fault_t qxl_ttm_fault(struct vm_fault *vmf)
 {
 	struct ttm_buffer_object *bo;
-	int r;
+	vm_fault_t ret;
 
 	bo = (struct ttm_buffer_object *)vmf->vma->vm_private_data;
 	if (bo == NULL)
 		return VM_FAULT_NOPAGE;
-	r = ttm_vm_ops->fault(vmf);
-	return r;
+	ret = ttm_vm_ops->fault(vmf);
+	return ret;
 }
 
 int qxl_mmap(struct file *filp, struct vm_area_struct *vma)
diff --git a/drivers/gpu/drm/radeon/atom.c b/drivers/gpu/drm/radeon/atom.c
index 6a2e091aa7b6..e55cbeee7a53 100644
--- a/drivers/gpu/drm/radeon/atom.c
+++ b/drivers/gpu/drm/radeon/atom.c
@@ -1176,7 +1176,7 @@ static int atom_execute_table_locked(struct atom_context *ctx, int index, uint32
 	ectx.abort = false;
 	ectx.last_jump = 0;
 	if (ws)
-		ectx.ws = kzalloc(4 * ws, GFP_KERNEL);
+		ectx.ws = kcalloc(4, ws, GFP_KERNEL);
 	else
 		ectx.ws = NULL;
 
diff --git a/drivers/gpu/drm/radeon/atombios_crtc.c b/drivers/gpu/drm/radeon/atombios_crtc.c
index 02baaaf20e9d..efbd5816082d 100644
--- a/drivers/gpu/drm/radeon/atombios_crtc.c
+++ b/drivers/gpu/drm/radeon/atombios_crtc.c
@@ -1145,7 +1145,6 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
-	struct radeon_framebuffer *radeon_fb;
 	struct drm_framebuffer *target_fb;
 	struct drm_gem_object *obj;
 	struct radeon_bo *rbo;
@@ -1164,19 +1163,15 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		radeon_fb = to_radeon_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	}
-	else {
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
 	/* If atomic, assume fb object is pinned & idle & fenced and
 	 * just update base pointers
 	 */
-	obj = radeon_fb->obj;
+	obj = target_fb->obj[0];
 	rbo = gem_to_radeon_bo(obj);
 	r = radeon_bo_reserve(rbo, false);
 	if (unlikely(r != 0))
@@ -1441,8 +1436,7 @@ static int dce4_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(EVERGREEN_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 0);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		radeon_fb = to_radeon_framebuffer(fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r != 0))
 			return r;
@@ -1463,7 +1457,6 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
-	struct radeon_framebuffer *radeon_fb;
 	struct drm_gem_object *obj;
 	struct radeon_bo *rbo;
 	struct drm_framebuffer *target_fb;
@@ -1481,16 +1474,12 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		radeon_fb = to_radeon_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	}
-	else {
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
-	obj = radeon_fb->obj;
+	obj = target_fb->obj[0];
 	rbo = gem_to_radeon_bo(obj);
 	r = radeon_bo_reserve(rbo, false);
 	if (unlikely(r != 0))
@@ -1641,8 +1630,7 @@ static int avivo_crtc_do_set_base(struct drm_crtc *crtc,
 	WREG32(AVIVO_D1MODE_MASTER_UPDATE_MODE + radeon_crtc->crtc_offset, 3);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		radeon_fb = to_radeon_framebuffer(fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r != 0))
 			return r;
@@ -2149,11 +2137,9 @@ static void atombios_crtc_disable(struct drm_crtc *crtc)
 	atombios_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct radeon_framebuffer *radeon_fb;
 		struct radeon_bo *rbo;
 
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(crtc->primary->fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve rbo before unpin\n");
diff --git a/drivers/gpu/drm/radeon/btc_dpm.c b/drivers/gpu/drm/radeon/btc_dpm.c
index 95652e643da1..0aef4937c901 100644
--- a/drivers/gpu/drm/radeon/btc_dpm.c
+++ b/drivers/gpu/drm/radeon/btc_dpm.c
@@ -2581,7 +2581,9 @@ int btc_dpm_init(struct radeon_device *rdev)
 		return ret;
 
 	rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries =
-		kzalloc(4 * sizeof(struct radeon_clock_voltage_dependency_entry), GFP_KERNEL);
+		kcalloc(4,
+			sizeof(struct radeon_clock_voltage_dependency_entry),
+			GFP_KERNEL);
 	if (!rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries) {
 		r600_free_extended_power_table(rdev);
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/radeon/ci_dpm.c b/drivers/gpu/drm/radeon/ci_dpm.c
index 7e1b04dc5593..d587779a80b4 100644
--- a/drivers/gpu/drm/radeon/ci_dpm.c
+++ b/drivers/gpu/drm/radeon/ci_dpm.c
@@ -5568,8 +5568,9 @@ static int ci_parse_power_table(struct radeon_device *rdev)
 		(mode_info->atom_context->bios + data_offset +
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  state_array->ucNumEntries, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(state_array->ucNumEntries,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
@@ -5675,19 +5676,29 @@ int ci_dpm_init(struct radeon_device *rdev)
 	u16 data_offset, size;
 	u8 frev, crev;
 	struct ci_power_info *pi;
+	enum pci_bus_speed speed_cap;
+	struct pci_dev *root = rdev->pdev->bus->self;
 	int ret;
-	u32 mask;
 
 	pi = kzalloc(sizeof(struct ci_power_info), GFP_KERNEL);
 	if (pi == NULL)
 		return -ENOMEM;
 	rdev->pm.dpm.priv = pi;
 
-	ret = drm_pcie_get_speed_cap_mask(rdev->ddev, &mask);
-	if (ret)
+	speed_cap = pcie_get_speed_cap(root);
+	if (speed_cap == PCI_SPEED_UNKNOWN) {
 		pi->sys_pcie_mask = 0;
-	else
-		pi->sys_pcie_mask = mask;
+	} else {
+		if (speed_cap == PCIE_SPEED_8_0GT)
+			pi->sys_pcie_mask = RADEON_PCIE_SPEED_25 |
+				RADEON_PCIE_SPEED_50 |
+				RADEON_PCIE_SPEED_80;
+		else if (speed_cap == PCIE_SPEED_5_0GT)
+			pi->sys_pcie_mask = RADEON_PCIE_SPEED_25 |
+				RADEON_PCIE_SPEED_50;
+		else
+			pi->sys_pcie_mask = RADEON_PCIE_SPEED_25;
+	}
 	pi->force_pcie_gen = RADEON_PCIE_GEN_INVALID;
 
 	pi->pcie_gen_performance.max = RADEON_PCIE_GEN1;
@@ -5770,7 +5781,9 @@ int ci_dpm_init(struct radeon_device *rdev)
 	ci_set_private_data_variables_based_on_pptable(rdev);
 
 	rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries =
-		kzalloc(4 * sizeof(struct radeon_clock_voltage_dependency_entry), GFP_KERNEL);
+		kcalloc(4,
+			sizeof(struct radeon_clock_voltage_dependency_entry),
+			GFP_KERNEL);
 	if (!rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries) {
 		ci_dpm_fini(rdev);
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c
index 7c73bc7e2f85..ebce4601a305 100644
--- a/drivers/gpu/drm/radeon/cik.c
+++ b/drivers/gpu/drm/radeon/cik.c
@@ -9499,9 +9499,10 @@ int cik_set_vce_clocks(struct radeon_device *rdev, u32 evclk, u32 ecclk)
 static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 {
 	struct pci_dev *root = rdev->pdev->bus->self;
+	enum pci_bus_speed speed_cap;
 	int bridge_pos, gpu_pos;
-	u32 speed_cntl, mask, current_data_rate;
-	int ret, i;
+	u32 speed_cntl, current_data_rate;
+	int i;
 	u16 tmp16;
 
 	if (pci_is_root_bus(rdev->pdev->bus))
@@ -9516,23 +9517,24 @@ static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 	if (!(rdev->flags & RADEON_IS_PCIE))
 		return;
 
-	ret = drm_pcie_get_speed_cap_mask(rdev->ddev, &mask);
-	if (ret != 0)
+	speed_cap = pcie_get_speed_cap(root);
+	if (speed_cap == PCI_SPEED_UNKNOWN)
 		return;
 
-	if (!(mask & (DRM_PCIE_SPEED_50 | DRM_PCIE_SPEED_80)))
+	if ((speed_cap != PCIE_SPEED_8_0GT) &&
+	    (speed_cap != PCIE_SPEED_5_0GT))
 		return;
 
 	speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
 	current_data_rate = (speed_cntl & LC_CURRENT_DATA_RATE_MASK) >>
 		LC_CURRENT_DATA_RATE_SHIFT;
-	if (mask & DRM_PCIE_SPEED_80) {
+	if (speed_cap == PCIE_SPEED_8_0GT) {
 		if (current_data_rate == 2) {
 			DRM_INFO("PCIE gen 3 link speeds already enabled\n");
 			return;
 		}
 		DRM_INFO("enabling PCIE gen 3 link speeds, disable with radeon.pcie_gen2=0\n");
-	} else if (mask & DRM_PCIE_SPEED_50) {
+	} else if (speed_cap == PCIE_SPEED_5_0GT) {
 		if (current_data_rate == 1) {
 			DRM_INFO("PCIE gen 2 link speeds already enabled\n");
 			return;
@@ -9548,7 +9550,7 @@ static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 	if (!gpu_pos)
 		return;
 
-	if (mask & DRM_PCIE_SPEED_80) {
+	if (speed_cap == PCIE_SPEED_8_0GT) {
 		/* re-try equalization if gen3 is not already enabled */
 		if (current_data_rate != 2) {
 			u16 bridge_cfg, gpu_cfg;
@@ -9636,9 +9638,9 @@ static void cik_pcie_gen3_enable(struct radeon_device *rdev)
 
 	pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
 	tmp16 &= ~0xf;
-	if (mask & DRM_PCIE_SPEED_80)
+	if (speed_cap == PCIE_SPEED_8_0GT)
 		tmp16 |= 3; /* gen3 */
-	else if (mask & DRM_PCIE_SPEED_50)
+	else if (speed_cap == PCIE_SPEED_5_0GT)
 		tmp16 |= 2; /* gen2 */
 	else
 		tmp16 |= 1; /* gen1 */
diff --git a/drivers/gpu/drm/radeon/kv_dpm.c b/drivers/gpu/drm/radeon/kv_dpm.c
index ae1529b0ef6f..f055d6ea3522 100644
--- a/drivers/gpu/drm/radeon/kv_dpm.c
+++ b/drivers/gpu/drm/radeon/kv_dpm.c
@@ -2660,8 +2660,9 @@ static int kv_parse_power_table(struct radeon_device *rdev)
 		(mode_info->atom_context->bios + data_offset +
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  state_array->ucNumEntries, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(state_array->ucNumEntries,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
diff --git a/drivers/gpu/drm/radeon/ni_dpm.c b/drivers/gpu/drm/radeon/ni_dpm.c
index 9416e72f86aa..0fd8d6ba9828 100644
--- a/drivers/gpu/drm/radeon/ni_dpm.c
+++ b/drivers/gpu/drm/radeon/ni_dpm.c
@@ -3998,8 +3998,9 @@ static int ni_parse_power_table(struct radeon_device *rdev)
 		return -EINVAL;
 	power_info = (union power_info *)(mode_info->atom_context->bios + data_offset);
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  power_info->pplib.ucNumStates, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(power_info->pplib.ucNumStates,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 
@@ -4075,7 +4076,9 @@ int ni_dpm_init(struct radeon_device *rdev)
 		return ret;
 
 	rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries =
-		kzalloc(4 * sizeof(struct radeon_clock_voltage_dependency_entry), GFP_KERNEL);
+		kcalloc(4,
+			sizeof(struct radeon_clock_voltage_dependency_entry),
+			GFP_KERNEL);
 	if (!rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries) {
 		r600_free_extended_power_table(rdev);
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/radeon/r600_dpm.c b/drivers/gpu/drm/radeon/r600_dpm.c
index 31d1b4710844..5e044c98fca2 100644
--- a/drivers/gpu/drm/radeon/r600_dpm.c
+++ b/drivers/gpu/drm/radeon/r600_dpm.c
@@ -991,7 +991,7 @@ int r600_parse_extended_power_table(struct radeon_device *rdev)
 			ATOM_PPLIB_PhaseSheddingLimits_Record *entry;
 
 			rdev->pm.dpm.dyn_state.phase_shedding_limits_table.entries =
-				kzalloc(psl->ucNumEntries *
+				kcalloc(psl->ucNumEntries,
 					sizeof(struct radeon_phase_shedding_limits_entry),
 					GFP_KERNEL);
 			if (!rdev->pm.dpm.dyn_state.phase_shedding_limits_table.entries) {
@@ -1327,9 +1327,9 @@ enum radeon_pcie_gen r600_get_pcie_gen_support(struct radeon_device *rdev,
 	case RADEON_PCIE_GEN3:
 		return RADEON_PCIE_GEN3;
 	default:
-		if ((sys_mask & DRM_PCIE_SPEED_80) && (default_gen == RADEON_PCIE_GEN3))
+		if ((sys_mask & RADEON_PCIE_SPEED_80) && (default_gen == RADEON_PCIE_GEN3))
 			return RADEON_PCIE_GEN3;
-		else if ((sys_mask & DRM_PCIE_SPEED_50) && (default_gen == RADEON_PCIE_GEN2))
+		else if ((sys_mask & RADEON_PCIE_SPEED_50) && (default_gen == RADEON_PCIE_GEN2))
 			return RADEON_PCIE_GEN2;
 		else
 			return RADEON_PCIE_GEN1;
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index 4a2eb409aacc..1a6f6edb3515 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -1653,6 +1653,10 @@ struct radeon_pm {
 	struct radeon_dpm       dpm;
 };
 
+#define RADEON_PCIE_SPEED_25 1
+#define RADEON_PCIE_SPEED_50 2
+#define RADEON_PCIE_SPEED_80 4
+
 int radeon_pm_get_type_index(struct radeon_device *rdev,
 			     enum radeon_pm_state_type ps_type,
 			     int instance);
diff --git a/drivers/gpu/drm/radeon/radeon_atombios.c b/drivers/gpu/drm/radeon/radeon_atombios.c
index 4134759a6823..f422a8d6aec4 100644
--- a/drivers/gpu/drm/radeon/radeon_atombios.c
+++ b/drivers/gpu/drm/radeon/radeon_atombios.c
@@ -2126,13 +2126,16 @@ static int radeon_atombios_parse_power_table_1_3(struct radeon_device *rdev)
 		num_modes = ATOM_MAX_NUMBEROF_POWER_BLOCK;
 	if (num_modes == 0)
 		return state_index;
-	rdev->pm.power_state = kzalloc(sizeof(struct radeon_power_state) * num_modes, GFP_KERNEL);
+	rdev->pm.power_state = kcalloc(num_modes,
+				       sizeof(struct radeon_power_state),
+				       GFP_KERNEL);
 	if (!rdev->pm.power_state)
 		return state_index;
 	/* last mode is usually default, array is low to high */
 	for (i = 0; i < num_modes; i++) {
 		rdev->pm.power_state[state_index].clock_info =
-			kzalloc(sizeof(struct radeon_pm_clock_info) * 1, GFP_KERNEL);
+			kcalloc(1, sizeof(struct radeon_pm_clock_info),
+				GFP_KERNEL);
 		if (!rdev->pm.power_state[state_index].clock_info)
 			return state_index;
 		rdev->pm.power_state[state_index].num_clock_modes = 1;
@@ -2587,8 +2590,9 @@ static int radeon_atombios_parse_power_table_4_5(struct radeon_device *rdev)
 	radeon_atombios_add_pplib_thermal_controller(rdev, &power_info->pplib.sThermalController);
 	if (power_info->pplib.ucNumStates == 0)
 		return state_index;
-	rdev->pm.power_state = kzalloc(sizeof(struct radeon_power_state) *
-				       power_info->pplib.ucNumStates, GFP_KERNEL);
+	rdev->pm.power_state = kcalloc(power_info->pplib.ucNumStates,
+				       sizeof(struct radeon_power_state),
+				       GFP_KERNEL);
 	if (!rdev->pm.power_state)
 		return state_index;
 	/* first mode is usually default, followed by low to high */
@@ -2603,10 +2607,11 @@ static int radeon_atombios_parse_power_table_4_5(struct radeon_device *rdev)
 			 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset) +
 			 (power_state->v1.ucNonClockStateIndex *
 			  power_info->pplib.ucNonClockSize));
-		rdev->pm.power_state[i].clock_info = kzalloc(sizeof(struct radeon_pm_clock_info) *
-							     ((power_info->pplib.ucStateEntrySize - 1) ?
-							      (power_info->pplib.ucStateEntrySize - 1) : 1),
-							     GFP_KERNEL);
+		rdev->pm.power_state[i].clock_info =
+			kcalloc((power_info->pplib.ucStateEntrySize - 1) ?
+				(power_info->pplib.ucStateEntrySize - 1) : 1,
+				sizeof(struct radeon_pm_clock_info),
+				GFP_KERNEL);
 		if (!rdev->pm.power_state[i].clock_info)
 			return state_index;
 		if (power_info->pplib.ucStateEntrySize - 1) {
@@ -2688,8 +2693,9 @@ static int radeon_atombios_parse_power_table_6(struct radeon_device *rdev)
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 	if (state_array->ucNumEntries == 0)
 		return state_index;
-	rdev->pm.power_state = kzalloc(sizeof(struct radeon_power_state) *
-				       state_array->ucNumEntries, GFP_KERNEL);
+	rdev->pm.power_state = kcalloc(state_array->ucNumEntries,
+				       sizeof(struct radeon_power_state),
+				       GFP_KERNEL);
 	if (!rdev->pm.power_state)
 		return state_index;
 	power_state_offset = (u8 *)state_array->states;
@@ -2699,10 +2705,11 @@ static int radeon_atombios_parse_power_table_6(struct radeon_device *rdev)
 		non_clock_array_index = power_state->v2.nonClockInfoIndex;
 		non_clock_info = (struct _ATOM_PPLIB_NONCLOCK_INFO *)
 			&non_clock_info_array->nonClockInfo[non_clock_array_index];
-		rdev->pm.power_state[i].clock_info = kzalloc(sizeof(struct radeon_pm_clock_info) *
-							     (power_state->v2.ucNumDPMLevels ?
-							      power_state->v2.ucNumDPMLevels : 1),
-							     GFP_KERNEL);
+		rdev->pm.power_state[i].clock_info =
+			kcalloc(power_state->v2.ucNumDPMLevels ?
+				power_state->v2.ucNumDPMLevels : 1,
+				sizeof(struct radeon_pm_clock_info),
+				GFP_KERNEL);
 		if (!rdev->pm.power_state[i].clock_info)
 			return state_index;
 		if (power_state->v2.ucNumDPMLevels) {
@@ -2782,7 +2789,9 @@ void radeon_atombios_get_power_modes(struct radeon_device *rdev)
 		rdev->pm.power_state = kzalloc(sizeof(struct radeon_power_state), GFP_KERNEL);
 		if (rdev->pm.power_state) {
 			rdev->pm.power_state[0].clock_info =
-				kzalloc(sizeof(struct radeon_pm_clock_info) * 1, GFP_KERNEL);
+				kcalloc(1,
+				        sizeof(struct radeon_pm_clock_info),
+				        GFP_KERNEL);
 			if (rdev->pm.power_state[0].clock_info) {
 				/* add the default mode */
 				rdev->pm.power_state[state_index].type =
diff --git a/drivers/gpu/drm/radeon/radeon_atpx_handler.c b/drivers/gpu/drm/radeon/radeon_atpx_handler.c
index 40be4068ca69..fa5fadaa9bbb 100644
--- a/drivers/gpu/drm/radeon/radeon_atpx_handler.c
+++ b/drivers/gpu/drm/radeon/radeon_atpx_handler.c
@@ -526,7 +526,7 @@ static int radeon_atpx_init(void)
  * look up whether we are the integrated or discrete GPU (all asics).
  * Returns the client id.
  */
-static int radeon_atpx_get_client_id(struct pci_dev *pdev)
+static enum vga_switcheroo_client_id radeon_atpx_get_client_id(struct pci_dev *pdev)
 {
 	if (radeon_atpx_priv.dhandle == ACPI_HANDLE(&pdev->dev))
 		return VGA_SWITCHEROO_IGD;
diff --git a/drivers/gpu/drm/radeon/radeon_combios.c b/drivers/gpu/drm/radeon/radeon_combios.c
index 3178ba0c537c..60a61d33f607 100644
--- a/drivers/gpu/drm/radeon/radeon_combios.c
+++ b/drivers/gpu/drm/radeon/radeon_combios.c
@@ -2642,13 +2642,16 @@ void radeon_combios_get_power_modes(struct radeon_device *rdev)
 	rdev->pm.default_power_state_index = -1;
 
 	/* allocate 2 power states */
-	rdev->pm.power_state = kzalloc(sizeof(struct radeon_power_state) * 2, GFP_KERNEL);
+	rdev->pm.power_state = kcalloc(2, sizeof(struct radeon_power_state),
+				       GFP_KERNEL);
 	if (rdev->pm.power_state) {
 		/* allocate 1 clock mode per state */
 		rdev->pm.power_state[0].clock_info =
-			kzalloc(sizeof(struct radeon_pm_clock_info) * 1, GFP_KERNEL);
+			kcalloc(1, sizeof(struct radeon_pm_clock_info),
+				GFP_KERNEL);
 		rdev->pm.power_state[1].clock_info =
-			kzalloc(sizeof(struct radeon_pm_clock_info) * 1, GFP_KERNEL);
+			kcalloc(1, sizeof(struct radeon_pm_clock_info),
+				GFP_KERNEL);
 		if (!rdev->pm.power_state[0].clock_info ||
 		    !rdev->pm.power_state[1].clock_info)
 			goto pm_failed;
diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c
index df9469a8fdb1..414642e5b7a3 100644
--- a/drivers/gpu/drm/radeon/radeon_connectors.c
+++ b/drivers/gpu/drm/radeon/radeon_connectors.c
@@ -244,23 +244,15 @@ radeon_connector_update_scratch_regs(struct drm_connector *connector, enum drm_c
 {
 	struct drm_device *dev = connector->dev;
 	struct radeon_device *rdev = dev->dev_private;
-	struct drm_encoder *best_encoder = NULL;
-	struct drm_encoder *encoder = NULL;
+	struct drm_encoder *best_encoder;
+	struct drm_encoder *encoder;
 	const struct drm_connector_helper_funcs *connector_funcs = connector->helper_private;
 	bool connected;
 	int i;
 
 	best_encoder = connector_funcs->best_encoder(connector);
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-
-		encoder = drm_encoder_find(connector->dev, NULL,
-					   connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
-
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		if ((encoder == best_encoder) && (status == connector_status_connected))
 			connected = true;
 		else
@@ -270,7 +262,6 @@ radeon_connector_update_scratch_regs(struct drm_connector *connector, enum drm_c
 			radeon_atombios_connected_scratch_regs(connector, encoder, connected);
 		else
 			radeon_combios_connected_scratch_regs(connector, encoder, connected);
-
 	}
 }
 
@@ -279,17 +270,11 @@ static struct drm_encoder *radeon_find_encoder(struct drm_connector *connector,
 	struct drm_encoder *encoder;
 	int i;
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-
-		encoder = drm_encoder_find(connector->dev, NULL, connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
-
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		if (encoder->encoder_type == encoder_type)
 			return encoder;
 	}
+
 	return NULL;
 }
 
@@ -383,20 +368,23 @@ static int radeon_ddc_get_modes(struct drm_connector *connector)
 	int ret;
 
 	if (radeon_connector->edid) {
-		drm_mode_connector_update_edid_property(connector, radeon_connector->edid);
+		drm_connector_update_edid_property(connector, radeon_connector->edid);
 		ret = drm_add_edid_modes(connector, radeon_connector->edid);
 		return ret;
 	}
-	drm_mode_connector_update_edid_property(connector, NULL);
+	drm_connector_update_edid_property(connector, NULL);
 	return 0;
 }
 
 static struct drm_encoder *radeon_best_single_encoder(struct drm_connector *connector)
 {
-	int enc_id = connector->encoder_ids[0];
-	/* pick the encoder ids */
-	if (enc_id)
-		return drm_encoder_find(connector->dev, NULL, enc_id);
+	struct drm_encoder *encoder;
+	int i;
+
+	/* pick the first one */
+	drm_connector_for_each_possible_encoder(connector, encoder, i)
+		return encoder;
+
 	return NULL;
 }
 
@@ -436,19 +424,19 @@ radeon_connector_analog_encoder_conflict_solve(struct drm_connector *connector,
 	struct drm_device *dev = connector->dev;
 	struct drm_connector *conflict;
 	struct radeon_connector *radeon_conflict;
-	int i;
 
 	list_for_each_entry(conflict, &dev->mode_config.connector_list, head) {
+		struct drm_encoder *enc;
+		int i;
+
 		if (conflict == connector)
 			continue;
 
 		radeon_conflict = to_radeon_connector(conflict);
-		for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-			if (conflict->encoder_ids[i] == 0)
-				break;
 
+		drm_connector_for_each_possible_encoder(conflict, enc, i) {
 			/* if the IDs match */
-			if (conflict->encoder_ids[i] == encoder->base.id) {
+			if (enc == encoder) {
 				if (conflict->status != connector_status_connected)
 					continue;
 
@@ -852,7 +840,7 @@ static int radeon_lvds_get_modes(struct drm_connector *connector)
 	return ret;
 }
 
-static int radeon_lvds_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status radeon_lvds_mode_valid(struct drm_connector *connector,
 				  struct drm_display_mode *mode)
 {
 	struct drm_encoder *encoder = radeon_best_single_encoder(connector);
@@ -1012,7 +1000,7 @@ static int radeon_vga_get_modes(struct drm_connector *connector)
 	return ret;
 }
 
-static int radeon_vga_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status radeon_vga_mode_valid(struct drm_connector *connector,
 				  struct drm_display_mode *mode)
 {
 	struct drm_device *dev = connector->dev;
@@ -1156,7 +1144,7 @@ static int radeon_tv_get_modes(struct drm_connector *connector)
 	return 1;
 }
 
-static int radeon_tv_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status radeon_tv_mode_valid(struct drm_connector *connector,
 				struct drm_display_mode *mode)
 {
 	if ((mode->hdisplay > 1024) || (mode->vdisplay > 768))
@@ -1256,7 +1244,7 @@ radeon_dvi_detect(struct drm_connector *connector, bool force)
 	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
 	struct drm_encoder *encoder = NULL;
 	const struct drm_encoder_helper_funcs *encoder_funcs;
-	int i, r;
+	int r;
 	enum drm_connector_status ret = connector_status_disconnected;
 	bool dret = false, broken_edid = false;
 
@@ -1374,15 +1362,9 @@ radeon_dvi_detect(struct drm_connector *connector, bool force)
 
 	/* find analog encoder */
 	if (radeon_connector->dac_load_detect) {
-		for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-			if (connector->encoder_ids[i] == 0)
-				break;
-
-			encoder = drm_encoder_find(connector->dev, NULL,
-						   connector->encoder_ids[i]);
-			if (!encoder)
-				continue;
+		int i;
 
+		drm_connector_for_each_possible_encoder(connector, encoder, i) {
 			if (encoder->encoder_type != DRM_MODE_ENCODER_DAC &&
 			    encoder->encoder_type != DRM_MODE_ENCODER_TVDAC)
 				continue;
@@ -1458,18 +1440,11 @@ exit:
 /* okay need to be smart in here about which encoder to pick */
 static struct drm_encoder *radeon_dvi_encoder(struct drm_connector *connector)
 {
-	int enc_id = connector->encoder_ids[0];
 	struct radeon_connector *radeon_connector = to_radeon_connector(connector);
 	struct drm_encoder *encoder;
 	int i;
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-
-		encoder = drm_encoder_find(connector->dev, NULL, connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
 
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		if (radeon_connector->use_digital == true) {
 			if (encoder->encoder_type == DRM_MODE_ENCODER_TMDS)
 				return encoder;
@@ -1484,8 +1459,9 @@ static struct drm_encoder *radeon_dvi_encoder(struct drm_connector *connector)
 
 	/* then check use digitial */
 	/* pick the first one */
-	if (enc_id)
-		return drm_encoder_find(connector->dev, NULL, enc_id);
+	drm_connector_for_each_possible_encoder(connector, encoder, i)
+		return encoder;
+
 	return NULL;
 }
 
@@ -1498,7 +1474,7 @@ static void radeon_dvi_force(struct drm_connector *connector)
 		radeon_connector->use_digital = true;
 }
 
-static int radeon_dvi_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status radeon_dvi_mode_valid(struct drm_connector *connector,
 				  struct drm_display_mode *mode)
 {
 	struct drm_device *dev = connector->dev;
@@ -1628,14 +1604,7 @@ u16 radeon_connector_encoder_get_dp_bridge_encoder_id(struct drm_connector *conn
 	struct radeon_encoder *radeon_encoder;
 	int i;
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-
-		encoder = drm_encoder_find(connector->dev, NULL, connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
-
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		radeon_encoder = to_radeon_encoder(encoder);
 
 		switch (radeon_encoder->encoder_id) {
@@ -1657,14 +1626,7 @@ static bool radeon_connector_encoder_is_hbr2(struct drm_connector *connector)
 	int i;
 	bool found = false;
 
-	for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
-		if (connector->encoder_ids[i] == 0)
-			break;
-
-		encoder = drm_encoder_find(connector->dev, NULL, connector->encoder_ids[i]);
-		if (!encoder)
-			continue;
-
+	drm_connector_for_each_possible_encoder(connector, encoder, i) {
 		radeon_encoder = to_radeon_encoder(encoder);
 		if (radeon_encoder->caps & ATOM_ENCODER_CAP_RECORD_HBR2)
 			found = true;
@@ -1800,7 +1762,7 @@ out:
 	return ret;
 }
 
-static int radeon_dp_mode_valid(struct drm_connector *connector,
+static enum drm_mode_status radeon_dp_mode_valid(struct drm_connector *connector,
 				  struct drm_display_mode *mode)
 {
 	struct drm_device *dev = connector->dev;
diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index 48d0e6bd0508..59c8a6647ff2 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -1591,7 +1591,7 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend,
 	/* unpin the front buffers and cursors */
 	list_for_each_entry(crtc, &dev->mode_config.crtc_list, head) {
 		struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-		struct radeon_framebuffer *rfb = to_radeon_framebuffer(crtc->primary->fb);
+		struct drm_framebuffer *fb = crtc->primary->fb;
 		struct radeon_bo *robj;
 
 		if (radeon_crtc->cursor_bo) {
@@ -1603,10 +1603,10 @@ int radeon_suspend_kms(struct drm_device *dev, bool suspend,
 			}
 		}
 
-		if (rfb == NULL || rfb->obj == NULL) {
+		if (fb == NULL || fb->obj[0] == NULL) {
 			continue;
 		}
-		robj = gem_to_radeon_bo(rfb->obj);
+		robj = gem_to_radeon_bo(fb->obj[0]);
 		/* don't unpin kernel fb objects */
 		if (!radeon_fbdev_robj_is_fb(rdev, robj)) {
 			r = radeon_bo_reserve(robj, false);
diff --git a/drivers/gpu/drm/radeon/radeon_display.c b/drivers/gpu/drm/radeon/radeon_display.c
index 26129b2b082d..9d3ac8b981da 100644
--- a/drivers/gpu/drm/radeon/radeon_display.c
+++ b/drivers/gpu/drm/radeon/radeon_display.c
@@ -32,6 +32,7 @@
 
 #include <linux/pm_runtime.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_plane_helper.h>
 #include <drm/drm_edid.h>
@@ -478,8 +479,6 @@ static int radeon_crtc_page_flip_target(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-	struct radeon_framebuffer *old_radeon_fb;
-	struct radeon_framebuffer *new_radeon_fb;
 	struct drm_gem_object *obj;
 	struct radeon_flip_work *work;
 	struct radeon_bo *new_rbo;
@@ -501,15 +500,13 @@ static int radeon_crtc_page_flip_target(struct drm_crtc *crtc,
 	work->async = (page_flip_flags & DRM_MODE_PAGE_FLIP_ASYNC) != 0;
 
 	/* schedule unpin of the old buffer */
-	old_radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
-	obj = old_radeon_fb->obj;
+	obj = crtc->primary->fb->obj[0];
 
 	/* take a reference to the old object */
 	drm_gem_object_get(obj);
 	work->old_rbo = gem_to_radeon_bo(obj);
 
-	new_radeon_fb = to_radeon_framebuffer(fb);
-	obj = new_radeon_fb->obj;
+	obj = fb->obj[0];
 	new_rbo = gem_to_radeon_bo(obj);
 
 	/* pin the new buffer */
@@ -1285,41 +1282,23 @@ void radeon_compute_pll_legacy(struct radeon_pll *pll,
 
 }
 
-static void radeon_user_framebuffer_destroy(struct drm_framebuffer *fb)
-{
-	struct radeon_framebuffer *radeon_fb = to_radeon_framebuffer(fb);
-
-	drm_gem_object_put_unlocked(radeon_fb->obj);
-	drm_framebuffer_cleanup(fb);
-	kfree(radeon_fb);
-}
-
-static int radeon_user_framebuffer_create_handle(struct drm_framebuffer *fb,
-						  struct drm_file *file_priv,
-						  unsigned int *handle)
-{
-	struct radeon_framebuffer *radeon_fb = to_radeon_framebuffer(fb);
-
-	return drm_gem_handle_create(file_priv, radeon_fb->obj, handle);
-}
-
 static const struct drm_framebuffer_funcs radeon_fb_funcs = {
-	.destroy = radeon_user_framebuffer_destroy,
-	.create_handle = radeon_user_framebuffer_create_handle,
+	.destroy = drm_gem_fb_destroy,
+	.create_handle = drm_gem_fb_create_handle,
 };
 
 int
 radeon_framebuffer_init(struct drm_device *dev,
-			struct radeon_framebuffer *rfb,
+			struct drm_framebuffer *fb,
 			const struct drm_mode_fb_cmd2 *mode_cmd,
 			struct drm_gem_object *obj)
 {
 	int ret;
-	rfb->obj = obj;
-	drm_helper_mode_fill_fb_struct(dev, &rfb->base, mode_cmd);
-	ret = drm_framebuffer_init(dev, &rfb->base, &radeon_fb_funcs);
+	fb->obj[0] = obj;
+	drm_helper_mode_fill_fb_struct(dev, fb, mode_cmd);
+	ret = drm_framebuffer_init(dev, fb, &radeon_fb_funcs);
 	if (ret) {
-		rfb->obj = NULL;
+		fb->obj[0] = NULL;
 		return ret;
 	}
 	return 0;
@@ -1331,7 +1310,7 @@ radeon_user_framebuffer_create(struct drm_device *dev,
 			       const struct drm_mode_fb_cmd2 *mode_cmd)
 {
 	struct drm_gem_object *obj;
-	struct radeon_framebuffer *radeon_fb;
+	struct drm_framebuffer *fb;
 	int ret;
 
 	obj = drm_gem_object_lookup(file_priv, mode_cmd->handles[0]);
@@ -1347,20 +1326,20 @@ radeon_user_framebuffer_create(struct drm_device *dev,
 		return ERR_PTR(-EINVAL);
 	}
 
-	radeon_fb = kzalloc(sizeof(*radeon_fb), GFP_KERNEL);
-	if (radeon_fb == NULL) {
+	fb = kzalloc(sizeof(*fb), GFP_KERNEL);
+	if (fb == NULL) {
 		drm_gem_object_put_unlocked(obj);
 		return ERR_PTR(-ENOMEM);
 	}
 
-	ret = radeon_framebuffer_init(dev, radeon_fb, mode_cmd, obj);
+	ret = radeon_framebuffer_init(dev, fb, mode_cmd, obj);
 	if (ret) {
-		kfree(radeon_fb);
+		kfree(fb);
 		drm_gem_object_put_unlocked(obj);
 		return ERR_PTR(ret);
 	}
 
-	return &radeon_fb->base;
+	return fb;
 }
 
 static const struct drm_mode_config_funcs radeon_mode_funcs = {
diff --git a/drivers/gpu/drm/radeon/radeon_dp_mst.c b/drivers/gpu/drm/radeon/radeon_dp_mst.c
index cd8a3ee16649..f920be236cc9 100644
--- a/drivers/gpu/drm/radeon/radeon_dp_mst.c
+++ b/drivers/gpu/drm/radeon/radeon_dp_mst.c
@@ -195,11 +195,11 @@ static int radeon_dp_mst_get_ddc_modes(struct drm_connector *connector)
 	radeon_connector->edid = edid;
 	DRM_DEBUG_KMS("edid retrieved %p\n", edid);
 	if (radeon_connector->edid) {
-		drm_mode_connector_update_edid_property(&radeon_connector->base, radeon_connector->edid);
+		drm_connector_update_edid_property(&radeon_connector->base, radeon_connector->edid);
 		ret = drm_add_edid_modes(&radeon_connector->base, radeon_connector->edid);
 		return ret;
 	}
-	drm_mode_connector_update_edid_property(&radeon_connector->base, NULL);
+	drm_connector_update_edid_property(&radeon_connector->base, NULL);
 
 	return ret;
 }
@@ -290,7 +290,7 @@ static struct drm_connector *radeon_dp_add_mst_connector(struct drm_dp_mst_topol
 
 	drm_object_attach_property(&connector->base, dev->mode_config.path_property, 0);
 	drm_object_attach_property(&connector->base, dev->mode_config.tile_property, 0);
-	drm_mode_connector_set_path_property(connector, pathprop);
+	drm_connector_set_path_property(connector, pathprop);
 
 	return connector;
 }
diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c
index b28288a781ef..2a7977a23b31 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.c
+++ b/drivers/gpu/drm/radeon/radeon_drv.c
@@ -168,7 +168,12 @@ int radeon_no_wb;
 int radeon_modeset = -1;
 int radeon_dynclks = -1;
 int radeon_r4xx_atom = 0;
+#ifdef __powerpc__
+/* Default to PCI on PowerPC (fdo #95017) */
+int radeon_agpmode = -1;
+#else
 int radeon_agpmode = 0;
+#endif
 int radeon_vram_limit = 0;
 int radeon_gart_size = -1; /* auto */
 int radeon_benchmarking = 0;
diff --git a/drivers/gpu/drm/radeon/radeon_encoders.c b/drivers/gpu/drm/radeon/radeon_encoders.c
index c6ee80216cf4..c341fb2a5b56 100644
--- a/drivers/gpu/drm/radeon/radeon_encoders.c
+++ b/drivers/gpu/drm/radeon/radeon_encoders.c
@@ -211,7 +211,7 @@ radeon_link_encoder_connector(struct drm_device *dev)
 		list_for_each_entry(encoder, &dev->mode_config.encoder_list, head) {
 			radeon_encoder = to_radeon_encoder(encoder);
 			if (radeon_encoder->devices & radeon_connector->devices) {
-				drm_mode_connector_attach_encoder(connector, encoder);
+				drm_connector_attach_encoder(connector, encoder);
 				if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT))
 					radeon_encoder_add_backlight(radeon_encoder, connector);
 			}
diff --git a/drivers/gpu/drm/radeon/radeon_fb.c b/drivers/gpu/drm/radeon/radeon_fb.c
index 57c5404a1654..1179034024ae 100644
--- a/drivers/gpu/drm/radeon/radeon_fb.c
+++ b/drivers/gpu/drm/radeon/radeon_fb.c
@@ -43,7 +43,7 @@
  */
 struct radeon_fbdev {
 	struct drm_fb_helper helper;
-	struct radeon_framebuffer rfb;
+	struct drm_framebuffer fb;
 	struct radeon_device *rdev;
 };
 
@@ -246,13 +246,13 @@ static int radeonfb_create(struct drm_fb_helper *helper,
 
 	info->par = rfbdev;
 
-	ret = radeon_framebuffer_init(rdev->ddev, &rfbdev->rfb, &mode_cmd, gobj);
+	ret = radeon_framebuffer_init(rdev->ddev, &rfbdev->fb, &mode_cmd, gobj);
 	if (ret) {
 		DRM_ERROR("failed to initialize framebuffer %d\n", ret);
 		goto out;
 	}
 
-	fb = &rfbdev->rfb.base;
+	fb = &rfbdev->fb;
 
 	/* setup helper */
 	rfbdev->helper.fb = fb;
@@ -308,15 +308,15 @@ out:
 
 static int radeon_fbdev_destroy(struct drm_device *dev, struct radeon_fbdev *rfbdev)
 {
-	struct radeon_framebuffer *rfb = &rfbdev->rfb;
+	struct drm_framebuffer *fb = &rfbdev->fb;
 
 	drm_fb_helper_unregister_fbi(&rfbdev->helper);
 
-	if (rfb->obj) {
-		radeonfb_destroy_pinned_object(rfb->obj);
-		rfb->obj = NULL;
-		drm_framebuffer_unregister_private(&rfb->base);
-		drm_framebuffer_cleanup(&rfb->base);
+	if (fb->obj[0]) {
+		radeonfb_destroy_pinned_object(fb->obj[0]);
+		fb->obj[0] = NULL;
+		drm_framebuffer_unregister_private(fb);
+		drm_framebuffer_cleanup(fb);
 	}
 	drm_fb_helper_fini(&rfbdev->helper);
 
@@ -400,7 +400,7 @@ bool radeon_fbdev_robj_is_fb(struct radeon_device *rdev, struct radeon_bo *robj)
 	if (!rdev->mode_info.rfbdev)
 		return false;
 
-	if (robj == gem_to_radeon_bo(rdev->mode_info.rfbdev->rfb.obj))
+	if (robj == gem_to_radeon_bo(rdev->mode_info.rfbdev->fb.obj[0]))
 		return true;
 	return false;
 }
diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c
index 0b3ec35515f3..1cef155cc933 100644
--- a/drivers/gpu/drm/radeon/radeon_gart.c
+++ b/drivers/gpu/drm/radeon/radeon_gart.c
@@ -347,13 +347,14 @@ int radeon_gart_init(struct radeon_device *rdev)
 	DRM_INFO("GART: num cpu pages %u, num gpu pages %u\n",
 		 rdev->gart.num_cpu_pages, rdev->gart.num_gpu_pages);
 	/* Allocate pages table */
-	rdev->gart.pages = vzalloc(sizeof(void *) * rdev->gart.num_cpu_pages);
+	rdev->gart.pages = vzalloc(array_size(sizeof(void *),
+				   rdev->gart.num_cpu_pages));
 	if (rdev->gart.pages == NULL) {
 		radeon_gart_fini(rdev);
 		return -ENOMEM;
 	}
-	rdev->gart.pages_entry = vmalloc(sizeof(uint64_t) *
-					 rdev->gart.num_gpu_pages);
+	rdev->gart.pages_entry = vmalloc(array_size(sizeof(uint64_t),
+						    rdev->gart.num_gpu_pages));
 	if (rdev->gart.pages_entry == NULL) {
 		radeon_gart_fini(rdev);
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
index 1f1856e0b1e0..35a205ae4318 100644
--- a/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
+++ b/drivers/gpu/drm/radeon/radeon_legacy_crtc.c
@@ -374,7 +374,6 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct radeon_device *rdev = dev->dev_private;
 	struct radeon_crtc *radeon_crtc = to_radeon_crtc(crtc);
-	struct radeon_framebuffer *radeon_fb;
 	struct drm_framebuffer *target_fb;
 	struct drm_gem_object *obj;
 	struct radeon_bo *rbo;
@@ -393,14 +392,10 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc,
 		return 0;
 	}
 
-	if (atomic) {
-		radeon_fb = to_radeon_framebuffer(fb);
+	if (atomic)
 		target_fb = fb;
-	}
-	else {
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
+	else
 		target_fb = crtc->primary->fb;
-	}
 
 	switch (target_fb->format->cpp[0] * 8) {
 	case 8:
@@ -423,7 +418,7 @@ int radeon_crtc_do_set_base(struct drm_crtc *crtc,
 	}
 
 	/* Pin framebuffer & get tilling informations */
-	obj = radeon_fb->obj;
+	obj = target_fb->obj[0];
 	rbo = gem_to_radeon_bo(obj);
 retry:
 	r = radeon_bo_reserve(rbo, false);
@@ -451,7 +446,7 @@ retry:
 			struct radeon_bo *old_rbo;
 			unsigned long nsize, osize;
 
-			old_rbo = gem_to_radeon_bo(to_radeon_framebuffer(fb)->obj);
+			old_rbo = gem_to_radeon_bo(fb->obj[0]);
 			osize = radeon_bo_size(old_rbo);
 			nsize = radeon_bo_size(rbo);
 			if (nsize <= osize && !radeon_bo_reserve(old_rbo, false)) {
@@ -558,8 +553,7 @@ retry:
 	WREG32(RADEON_CRTC_PITCH + radeon_crtc->crtc_offset, crtc_pitch);
 
 	if (!atomic && fb && fb != crtc->primary->fb) {
-		radeon_fb = to_radeon_framebuffer(fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r != 0))
 			return r;
@@ -1093,11 +1087,9 @@ static void radeon_crtc_disable(struct drm_crtc *crtc)
 	radeon_crtc_dpms(crtc, DRM_MODE_DPMS_OFF);
 	if (crtc->primary->fb) {
 		int r;
-		struct radeon_framebuffer *radeon_fb;
 		struct radeon_bo *rbo;
 
-		radeon_fb = to_radeon_framebuffer(crtc->primary->fb);
-		rbo = gem_to_radeon_bo(radeon_fb->obj);
+		rbo = gem_to_radeon_bo(crtc->primary->fb->obj[0]);
 		r = radeon_bo_reserve(rbo, false);
 		if (unlikely(r))
 			DRM_ERROR("failed to reserve rbo before unpin\n");
diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h
index 3243e5e01432..fd470d6bf3f4 100644
--- a/drivers/gpu/drm/radeon/radeon_mode.h
+++ b/drivers/gpu/drm/radeon/radeon_mode.h
@@ -46,7 +46,6 @@ struct radeon_device;
 #define to_radeon_crtc(x) container_of(x, struct radeon_crtc, base)
 #define to_radeon_connector(x) container_of(x, struct radeon_connector, base)
 #define to_radeon_encoder(x) container_of(x, struct radeon_encoder, base)
-#define to_radeon_framebuffer(x) container_of(x, struct radeon_framebuffer, base)
 
 #define RADEON_MAX_HPD_PINS 7
 #define RADEON_MAX_CRTCS 6
@@ -574,11 +573,6 @@ struct radeon_connector {
 	int enabled_attribs;
 };
 
-struct radeon_framebuffer {
-	struct drm_framebuffer base;
-	struct drm_gem_object *obj;
-};
-
 #define ENCODER_MODE_IS_DP(em) (((em) == ATOM_ENCODER_MODE_DP) || \
 				((em) == ATOM_ENCODER_MODE_DP_MST))
 
@@ -932,7 +926,7 @@ radeon_combios_encoder_crtc_scratch_regs(struct drm_encoder *encoder, int crtc);
 extern void
 radeon_combios_encoder_dpms_scratch_regs(struct drm_encoder *encoder, bool on);
 int radeon_framebuffer_init(struct drm_device *dev,
-			     struct radeon_framebuffer *rfb,
+			     struct drm_framebuffer *rfb,
 			     const struct drm_mode_fb_cmd2 *mode_cmd,
 			     struct drm_gem_object *obj);
 
diff --git a/drivers/gpu/drm/radeon/radeon_test.c b/drivers/gpu/drm/radeon/radeon_test.c
index f5e9abfadb56..0c7f228db6e3 100644
--- a/drivers/gpu/drm/radeon/radeon_test.c
+++ b/drivers/gpu/drm/radeon/radeon_test.c
@@ -1,3 +1,4 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /*
  * Copyright 2009 VMware, Inc.
  *
@@ -59,7 +60,7 @@ static void radeon_do_test_moves(struct radeon_device *rdev, int flag)
 	n = rdev->mc.gtt_size - rdev->gart_pin_size;
 	n /= size;
 
-	gtt_obj = kzalloc(n * sizeof(*gtt_obj), GFP_KERNEL);
+	gtt_obj = kcalloc(n, sizeof(*gtt_obj), GFP_KERNEL);
 	if (!gtt_obj) {
 		DRM_ERROR("Failed to allocate %d pointers\n", n);
 		r = 1;
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
index 8689fcca051c..cbb67e9ffb3a 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -947,11 +947,11 @@ void radeon_ttm_set_active_vram_size(struct radeon_device *rdev, u64 size)
 static struct vm_operations_struct radeon_ttm_vm_ops;
 static const struct vm_operations_struct *ttm_vm_ops = NULL;
 
-static int radeon_ttm_fault(struct vm_fault *vmf)
+static vm_fault_t radeon_ttm_fault(struct vm_fault *vmf)
 {
 	struct ttm_buffer_object *bo;
 	struct radeon_device *rdev;
-	int r;
+	vm_fault_t ret;
 
 	bo = (struct ttm_buffer_object *)vmf->vma->vm_private_data;
 	if (bo == NULL) {
@@ -959,9 +959,9 @@ static int radeon_ttm_fault(struct vm_fault *vmf)
 	}
 	rdev = radeon_get_rdev(bo->bdev);
 	down_read(&rdev->pm.mclk_lock);
-	r = ttm_vm_ops->fault(vmf);
+	ret = ttm_vm_ops->fault(vmf);
 	up_read(&rdev->pm.mclk_lock);
-	return r;
+	return ret;
 }
 
 int radeon_mmap(struct file *filp, struct vm_area_struct *vma)
diff --git a/drivers/gpu/drm/radeon/rs780_dpm.c b/drivers/gpu/drm/radeon/rs780_dpm.c
index b5e4e09a8996..694b7b3e9799 100644
--- a/drivers/gpu/drm/radeon/rs780_dpm.c
+++ b/drivers/gpu/drm/radeon/rs780_dpm.c
@@ -804,8 +804,9 @@ static int rs780_parse_power_table(struct radeon_device *rdev)
 		return -EINVAL;
 	power_info = (union power_info *)(mode_info->atom_context->bios + data_offset);
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  power_info->pplib.ucNumStates, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(power_info->pplib.ucNumStates,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/radeon/rv6xx_dpm.c b/drivers/gpu/drm/radeon/rv6xx_dpm.c
index d91aa3944593..6986051fbb89 100644
--- a/drivers/gpu/drm/radeon/rv6xx_dpm.c
+++ b/drivers/gpu/drm/radeon/rv6xx_dpm.c
@@ -1888,8 +1888,9 @@ static int rv6xx_parse_power_table(struct radeon_device *rdev)
 		return -EINVAL;
 	power_info = (union power_info *)(mode_info->atom_context->bios + data_offset);
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  power_info->pplib.ucNumStates, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(power_info->pplib.ucNumStates,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/radeon/rv770_dpm.c b/drivers/gpu/drm/radeon/rv770_dpm.c
index cb2a7ec4e217..c765ae7ea806 100644
--- a/drivers/gpu/drm/radeon/rv770_dpm.c
+++ b/drivers/gpu/drm/radeon/rv770_dpm.c
@@ -2282,8 +2282,9 @@ int rv7xx_parse_power_table(struct radeon_device *rdev)
 		return -EINVAL;
 	power_info = (union power_info *)(mode_info->atom_context->bios + data_offset);
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  power_info->pplib.ucNumStates, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(power_info->pplib.ucNumStates,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/radeon/si.c b/drivers/gpu/drm/radeon/si.c
index 1907c950d76f..85c604d29235 100644
--- a/drivers/gpu/drm/radeon/si.c
+++ b/drivers/gpu/drm/radeon/si.c
@@ -7082,9 +7082,10 @@ int si_set_uvd_clocks(struct radeon_device *rdev, u32 vclk, u32 dclk)
 static void si_pcie_gen3_enable(struct radeon_device *rdev)
 {
 	struct pci_dev *root = rdev->pdev->bus->self;
+	enum pci_bus_speed speed_cap;
 	int bridge_pos, gpu_pos;
-	u32 speed_cntl, mask, current_data_rate;
-	int ret, i;
+	u32 speed_cntl, current_data_rate;
+	int i;
 	u16 tmp16;
 
 	if (pci_is_root_bus(rdev->pdev->bus))
@@ -7099,23 +7100,24 @@ static void si_pcie_gen3_enable(struct radeon_device *rdev)
 	if (!(rdev->flags & RADEON_IS_PCIE))
 		return;
 
-	ret = drm_pcie_get_speed_cap_mask(rdev->ddev, &mask);
-	if (ret != 0)
+	speed_cap = pcie_get_speed_cap(root);
+	if (speed_cap == PCI_SPEED_UNKNOWN)
 		return;
 
-	if (!(mask & (DRM_PCIE_SPEED_50 | DRM_PCIE_SPEED_80)))
+	if ((speed_cap != PCIE_SPEED_8_0GT) &&
+	    (speed_cap != PCIE_SPEED_5_0GT))
 		return;
 
 	speed_cntl = RREG32_PCIE_PORT(PCIE_LC_SPEED_CNTL);
 	current_data_rate = (speed_cntl & LC_CURRENT_DATA_RATE_MASK) >>
 		LC_CURRENT_DATA_RATE_SHIFT;
-	if (mask & DRM_PCIE_SPEED_80) {
+	if (speed_cap == PCIE_SPEED_8_0GT) {
 		if (current_data_rate == 2) {
 			DRM_INFO("PCIE gen 3 link speeds already enabled\n");
 			return;
 		}
 		DRM_INFO("enabling PCIE gen 3 link speeds, disable with radeon.pcie_gen2=0\n");
-	} else if (mask & DRM_PCIE_SPEED_50) {
+	} else if (speed_cap == PCIE_SPEED_5_0GT) {
 		if (current_data_rate == 1) {
 			DRM_INFO("PCIE gen 2 link speeds already enabled\n");
 			return;
@@ -7131,7 +7133,7 @@ static void si_pcie_gen3_enable(struct radeon_device *rdev)
 	if (!gpu_pos)
 		return;
 
-	if (mask & DRM_PCIE_SPEED_80) {
+	if (speed_cap == PCIE_SPEED_8_0GT) {
 		/* re-try equalization if gen3 is not already enabled */
 		if (current_data_rate != 2) {
 			u16 bridge_cfg, gpu_cfg;
@@ -7219,9 +7221,9 @@ static void si_pcie_gen3_enable(struct radeon_device *rdev)
 
 	pci_read_config_word(rdev->pdev, gpu_pos + PCI_EXP_LNKCTL2, &tmp16);
 	tmp16 &= ~0xf;
-	if (mask & DRM_PCIE_SPEED_80)
+	if (speed_cap == PCIE_SPEED_8_0GT)
 		tmp16 |= 3; /* gen3 */
-	else if (mask & DRM_PCIE_SPEED_50)
+	else if (speed_cap == PCIE_SPEED_5_0GT)
 		tmp16 |= 2; /* gen2 */
 	else
 		tmp16 |= 1; /* gen1 */
diff --git a/drivers/gpu/drm/radeon/si_dpm.c b/drivers/gpu/drm/radeon/si_dpm.c
index 90d5b41007bf..8fb60b3af015 100644
--- a/drivers/gpu/drm/radeon/si_dpm.c
+++ b/drivers/gpu/drm/radeon/si_dpm.c
@@ -6832,8 +6832,9 @@ static int si_parse_power_table(struct radeon_device *rdev)
 		(mode_info->atom_context->bios + data_offset +
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  state_array->ucNumEntries, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(state_array->ucNumEntries,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
@@ -6898,8 +6899,9 @@ int si_dpm_init(struct radeon_device *rdev)
 	struct ni_power_info *ni_pi;
 	struct si_power_info *si_pi;
 	struct atom_clock_dividers dividers;
+	enum pci_bus_speed speed_cap;
+	struct pci_dev *root = rdev->pdev->bus->self;
 	int ret;
-	u32 mask;
 
 	si_pi = kzalloc(sizeof(struct si_power_info), GFP_KERNEL);
 	if (si_pi == NULL)
@@ -6909,11 +6911,20 @@ int si_dpm_init(struct radeon_device *rdev)
 	eg_pi = &ni_pi->eg;
 	pi = &eg_pi->rv7xx;
 
-	ret = drm_pcie_get_speed_cap_mask(rdev->ddev, &mask);
-	if (ret)
+	speed_cap = pcie_get_speed_cap(root);
+	if (speed_cap == PCI_SPEED_UNKNOWN) {
 		si_pi->sys_pcie_mask = 0;
-	else
-		si_pi->sys_pcie_mask = mask;
+	} else {
+		if (speed_cap == PCIE_SPEED_8_0GT)
+			si_pi->sys_pcie_mask = RADEON_PCIE_SPEED_25 |
+				RADEON_PCIE_SPEED_50 |
+				RADEON_PCIE_SPEED_80;
+		else if (speed_cap == PCIE_SPEED_5_0GT)
+			si_pi->sys_pcie_mask = RADEON_PCIE_SPEED_25 |
+				RADEON_PCIE_SPEED_50;
+		else
+			si_pi->sys_pcie_mask = RADEON_PCIE_SPEED_25;
+	}
 	si_pi->force_pcie_gen = RADEON_PCIE_GEN_INVALID;
 	si_pi->boot_pcie_gen = si_get_current_pcie_speed(rdev);
 
@@ -6941,7 +6952,9 @@ int si_dpm_init(struct radeon_device *rdev)
 		return ret;
 
 	rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries =
-		kzalloc(4 * sizeof(struct radeon_clock_voltage_dependency_entry), GFP_KERNEL);
+		kcalloc(4,
+			sizeof(struct radeon_clock_voltage_dependency_entry),
+			GFP_KERNEL);
 	if (!rdev->pm.dpm.dyn_state.vddc_dependency_on_dispclk.entries) {
 		r600_free_extended_power_table(rdev);
 		return -ENOMEM;
diff --git a/drivers/gpu/drm/radeon/sumo_dpm.c b/drivers/gpu/drm/radeon/sumo_dpm.c
index fd4804829e46..1e4975f3374c 100644
--- a/drivers/gpu/drm/radeon/sumo_dpm.c
+++ b/drivers/gpu/drm/radeon/sumo_dpm.c
@@ -1482,8 +1482,9 @@ static int sumo_parse_power_table(struct radeon_device *rdev)
 		(mode_info->atom_context->bios + data_offset +
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  state_array->ucNumEntries, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(state_array->ucNumEntries,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
diff --git a/drivers/gpu/drm/radeon/trinity_dpm.c b/drivers/gpu/drm/radeon/trinity_dpm.c
index 2ef7c4e5e495..5d317f763eea 100644
--- a/drivers/gpu/drm/radeon/trinity_dpm.c
+++ b/drivers/gpu/drm/radeon/trinity_dpm.c
@@ -1757,8 +1757,9 @@ static int trinity_parse_power_table(struct radeon_device *rdev)
 		(mode_info->atom_context->bios + data_offset +
 		 le16_to_cpu(power_info->pplib.usNonClockInfoArrayOffset));
 
-	rdev->pm.dpm.ps = kzalloc(sizeof(struct radeon_ps) *
-				  state_array->ucNumEntries, GFP_KERNEL);
+	rdev->pm.dpm.ps = kcalloc(state_array->ucNumEntries,
+				  sizeof(struct radeon_ps),
+				  GFP_KERNEL);
 	if (!rdev->pm.dpm.ps)
 		return -ENOMEM;
 	power_state_offset = (u8 *)state_array->states;
diff --git a/drivers/gpu/drm/rcar-du/Makefile b/drivers/gpu/drm/rcar-du/Makefile
index 3e58ed93d5b1..2a3b8d7972b5 100644
--- a/drivers/gpu/drm/rcar-du/Makefile
+++ b/drivers/gpu/drm/rcar-du/Makefile
@@ -17,3 +17,10 @@ rcar-du-drm-$(CONFIG_DRM_RCAR_VSP)	+= rcar_du_vsp.o
 obj-$(CONFIG_DRM_RCAR_DU)		+= rcar-du-drm.o
 obj-$(CONFIG_DRM_RCAR_DW_HDMI)		+= rcar_dw_hdmi.o
 obj-$(CONFIG_DRM_RCAR_LVDS)		+= rcar_lvds.o
+
+# 'remote-endpoint' is fixed up at run-time
+DTC_FLAGS_rcar_du_of_lvds_r8a7790 += -Wno-graph_endpoint
+DTC_FLAGS_rcar_du_of_lvds_r8a7791 += -Wno-graph_endpoint
+DTC_FLAGS_rcar_du_of_lvds_r8a7793 += -Wno-graph_endpoint
+DTC_FLAGS_rcar_du_of_lvds_r8a7795 += -Wno-graph_endpoint
+DTC_FLAGS_rcar_du_of_lvds_r8a7796 += -Wno-graph_endpoint
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
index c4420538ec85..15dc9caa128b 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.c
@@ -691,6 +691,52 @@ static const struct drm_crtc_helper_funcs crtc_helper_funcs = {
 	.atomic_disable = rcar_du_crtc_atomic_disable,
 };
 
+static struct drm_crtc_state *
+rcar_du_crtc_atomic_duplicate_state(struct drm_crtc *crtc)
+{
+	struct rcar_du_crtc_state *state;
+	struct rcar_du_crtc_state *copy;
+
+	if (WARN_ON(!crtc->state))
+		return NULL;
+
+	state = to_rcar_crtc_state(crtc->state);
+	copy = kmemdup(state, sizeof(*state), GFP_KERNEL);
+	if (copy == NULL)
+		return NULL;
+
+	__drm_atomic_helper_crtc_duplicate_state(crtc, &copy->state);
+
+	return &copy->state;
+}
+
+static void rcar_du_crtc_atomic_destroy_state(struct drm_crtc *crtc,
+					      struct drm_crtc_state *state)
+{
+	__drm_atomic_helper_crtc_destroy_state(state);
+	kfree(to_rcar_crtc_state(state));
+}
+
+static void rcar_du_crtc_reset(struct drm_crtc *crtc)
+{
+	struct rcar_du_crtc_state *state;
+
+	if (crtc->state) {
+		rcar_du_crtc_atomic_destroy_state(crtc, crtc->state);
+		crtc->state = NULL;
+	}
+
+	state = kzalloc(sizeof(*state), GFP_KERNEL);
+	if (state == NULL)
+		return;
+
+	state->crc.source = VSP1_DU_CRC_NONE;
+	state->crc.index = 0;
+
+	crtc->state = &state->state;
+	crtc->state->crtc = crtc;
+}
+
 static int rcar_du_crtc_enable_vblank(struct drm_crtc *crtc)
 {
 	struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc);
@@ -710,15 +756,111 @@ static void rcar_du_crtc_disable_vblank(struct drm_crtc *crtc)
 	rcrtc->vblank_enable = false;
 }
 
-static const struct drm_crtc_funcs crtc_funcs = {
-	.reset = drm_atomic_helper_crtc_reset,
+static int rcar_du_crtc_set_crc_source(struct drm_crtc *crtc,
+				       const char *source_name,
+				       size_t *values_cnt)
+{
+	struct rcar_du_crtc *rcrtc = to_rcar_crtc(crtc);
+	struct drm_modeset_acquire_ctx ctx;
+	struct drm_crtc_state *crtc_state;
+	struct drm_atomic_state *state;
+	enum vsp1_du_crc_source source;
+	unsigned int index = 0;
+	unsigned int i;
+	int ret;
+
+	/*
+	 * Parse the source name. Supported values are "plane%u" to compute the
+	 * CRC on an input plane (%u is the plane ID), and "auto" to compute the
+	 * CRC on the composer (VSP) output.
+	 */
+	if (!source_name) {
+		source = VSP1_DU_CRC_NONE;
+	} else if (!strcmp(source_name, "auto")) {
+		source = VSP1_DU_CRC_OUTPUT;
+	} else if (strstarts(source_name, "plane")) {
+		source = VSP1_DU_CRC_PLANE;
+
+		ret = kstrtouint(source_name + strlen("plane"), 10, &index);
+		if (ret < 0)
+			return ret;
+
+		for (i = 0; i < rcrtc->vsp->num_planes; ++i) {
+			if (index == rcrtc->vsp->planes[i].plane.base.id) {
+				index = i;
+				break;
+			}
+		}
+
+		if (i >= rcrtc->vsp->num_planes)
+			return -EINVAL;
+	} else {
+		return -EINVAL;
+	}
+
+	*values_cnt = 1;
+
+	/* Perform an atomic commit to set the CRC source. */
+	drm_modeset_acquire_init(&ctx, 0);
+
+	state = drm_atomic_state_alloc(crtc->dev);
+	if (!state) {
+		ret = -ENOMEM;
+		goto unlock;
+	}
+
+	state->acquire_ctx = &ctx;
+
+retry:
+	crtc_state = drm_atomic_get_crtc_state(state, crtc);
+	if (!IS_ERR(crtc_state)) {
+		struct rcar_du_crtc_state *rcrtc_state;
+
+		rcrtc_state = to_rcar_crtc_state(crtc_state);
+		rcrtc_state->crc.source = source;
+		rcrtc_state->crc.index = index;
+
+		ret = drm_atomic_commit(state);
+	} else {
+		ret = PTR_ERR(crtc_state);
+	}
+
+	if (ret == -EDEADLK) {
+		drm_atomic_state_clear(state);
+		drm_modeset_backoff(&ctx);
+		goto retry;
+	}
+
+	drm_atomic_state_put(state);
+
+unlock:
+	drm_modeset_drop_locks(&ctx);
+	drm_modeset_acquire_fini(&ctx);
+
+	return 0;
+}
+
+static const struct drm_crtc_funcs crtc_funcs_gen2 = {
+	.reset = rcar_du_crtc_reset,
+	.destroy = drm_crtc_cleanup,
+	.set_config = drm_atomic_helper_set_config,
+	.page_flip = drm_atomic_helper_page_flip,
+	.atomic_duplicate_state = rcar_du_crtc_atomic_duplicate_state,
+	.atomic_destroy_state = rcar_du_crtc_atomic_destroy_state,
+	.enable_vblank = rcar_du_crtc_enable_vblank,
+	.disable_vblank = rcar_du_crtc_disable_vblank,
+};
+
+static const struct drm_crtc_funcs crtc_funcs_gen3 = {
+	.reset = rcar_du_crtc_reset,
 	.destroy = drm_crtc_cleanup,
 	.set_config = drm_atomic_helper_set_config,
 	.page_flip = drm_atomic_helper_page_flip,
-	.atomic_duplicate_state = drm_atomic_helper_crtc_duplicate_state,
-	.atomic_destroy_state = drm_atomic_helper_crtc_destroy_state,
+	.atomic_duplicate_state = rcar_du_crtc_atomic_duplicate_state,
+	.atomic_destroy_state = rcar_du_crtc_atomic_destroy_state,
 	.enable_vblank = rcar_du_crtc_enable_vblank,
 	.disable_vblank = rcar_du_crtc_disable_vblank,
+	.set_crc_source = rcar_du_crtc_set_crc_source,
 };
 
 /* -----------------------------------------------------------------------------
@@ -767,7 +909,8 @@ static irqreturn_t rcar_du_crtc_irq(int irq, void *arg)
  * Initialization
  */
 
-int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
+int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int swindex,
+			unsigned int hwindex)
 {
 	static const unsigned int mmio_offsets[] = {
 		DU0_REG_OFFSET, DU1_REG_OFFSET, DU2_REG_OFFSET, DU3_REG_OFFSET
@@ -775,7 +918,7 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 
 	struct rcar_du_device *rcdu = rgrp->dev;
 	struct platform_device *pdev = to_platform_device(rcdu->dev);
-	struct rcar_du_crtc *rcrtc = &rcdu->crtcs[index];
+	struct rcar_du_crtc *rcrtc = &rcdu->crtcs[swindex];
 	struct drm_crtc *crtc = &rcrtc->crtc;
 	struct drm_plane *primary;
 	unsigned int irqflags;
@@ -787,7 +930,7 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 
 	/* Get the CRTC clock and the optional external clock. */
 	if (rcar_du_has(rcdu, RCAR_DU_FEATURE_CRTC_IRQ_CLOCK)) {
-		sprintf(clk_name, "du.%u", index);
+		sprintf(clk_name, "du.%u", hwindex);
 		name = clk_name;
 	} else {
 		name = NULL;
@@ -795,16 +938,16 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 
 	rcrtc->clock = devm_clk_get(rcdu->dev, name);
 	if (IS_ERR(rcrtc->clock)) {
-		dev_err(rcdu->dev, "no clock for CRTC %u\n", index);
+		dev_err(rcdu->dev, "no clock for DU channel %u\n", hwindex);
 		return PTR_ERR(rcrtc->clock);
 	}
 
-	sprintf(clk_name, "dclkin.%u", index);
+	sprintf(clk_name, "dclkin.%u", hwindex);
 	clk = devm_clk_get(rcdu->dev, clk_name);
 	if (!IS_ERR(clk)) {
 		rcrtc->extclock = clk;
 	} else if (PTR_ERR(rcrtc->clock) == -EPROBE_DEFER) {
-		dev_info(rcdu->dev, "can't get external clock %u\n", index);
+		dev_info(rcdu->dev, "can't get external clock %u\n", hwindex);
 		return -EPROBE_DEFER;
 	}
 
@@ -813,16 +956,18 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 	spin_lock_init(&rcrtc->vblank_lock);
 
 	rcrtc->group = rgrp;
-	rcrtc->mmio_offset = mmio_offsets[index];
-	rcrtc->index = index;
+	rcrtc->mmio_offset = mmio_offsets[hwindex];
+	rcrtc->index = hwindex;
 
 	if (rcar_du_has(rcdu, RCAR_DU_FEATURE_VSP1_SOURCE))
 		primary = &rcrtc->vsp->planes[rcrtc->vsp_pipe].plane;
 	else
-		primary = &rgrp->planes[index % 2].plane;
+		primary = &rgrp->planes[swindex % 2].plane;
 
-	ret = drm_crtc_init_with_planes(rcdu->ddev, crtc, primary,
-					NULL, &crtc_funcs, NULL);
+	ret = drm_crtc_init_with_planes(rcdu->ddev, crtc, primary, NULL,
+					rcdu->info->gen <= 2 ?
+					&crtc_funcs_gen2 : &crtc_funcs_gen3,
+					NULL);
 	if (ret < 0)
 		return ret;
 
@@ -833,7 +978,8 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 
 	/* Register the interrupt handler. */
 	if (rcar_du_has(rcdu, RCAR_DU_FEATURE_CRTC_IRQ_CLOCK)) {
-		irq = platform_get_irq(pdev, index);
+		/* The IRQ's are associated with the CRTC (sw)index. */
+		irq = platform_get_irq(pdev, swindex);
 		irqflags = 0;
 	} else {
 		irq = platform_get_irq(pdev, 0);
@@ -841,7 +987,7 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 	}
 
 	if (irq < 0) {
-		dev_err(rcdu->dev, "no IRQ for CRTC %u\n", index);
+		dev_err(rcdu->dev, "no IRQ for CRTC %u\n", swindex);
 		return irq;
 	}
 
@@ -849,7 +995,7 @@ int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index)
 			       dev_name(rcdu->dev), rcrtc);
 	if (ret < 0) {
 		dev_err(rcdu->dev,
-			"failed to register IRQ for CRTC %u\n", index);
+			"failed to register IRQ for CRTC %u\n", swindex);
 		return ret;
 	}
 
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_crtc.h b/drivers/gpu/drm/rcar-du/rcar_du_crtc.h
index fdc2bf99bda1..7680cb2636c8 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_crtc.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_crtc.h
@@ -21,6 +21,8 @@
 #include <drm/drmP.h>
 #include <drm/drm_crtc.h>
 
+#include <media/vsp1.h>
+
 struct rcar_du_group;
 struct rcar_du_vsp;
 
@@ -69,6 +71,19 @@ struct rcar_du_crtc {
 
 #define to_rcar_crtc(c)	container_of(c, struct rcar_du_crtc, crtc)
 
+/**
+ * struct rcar_du_crtc_state - Driver-specific CRTC state
+ * @state: base DRM CRTC state
+ * @crc: CRC computation configuration
+ */
+struct rcar_du_crtc_state {
+	struct drm_crtc_state state;
+
+	struct vsp1_du_crc_config crc;
+};
+
+#define to_rcar_crtc_state(s) container_of(s, struct rcar_du_crtc_state, state)
+
 enum rcar_du_output {
 	RCAR_DU_OUTPUT_DPAD0,
 	RCAR_DU_OUTPUT_DPAD1,
@@ -80,7 +95,8 @@ enum rcar_du_output {
 	RCAR_DU_OUTPUT_MAX,
 };
 
-int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int index);
+int rcar_du_crtc_create(struct rcar_du_group *rgrp, unsigned int swindex,
+			unsigned int hwindex);
 void rcar_du_crtc_suspend(struct rcar_du_crtc *rcrtc);
 void rcar_du_crtc_resume(struct rcar_du_crtc *rcrtc);
 
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.c b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
index 3917d839c04c..02aee6cb0e53 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.c
@@ -40,7 +40,7 @@ static const struct rcar_du_device_info rzg1_du_r8a7743_info = {
 	.gen = 2,
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS,
-	.num_crtcs = 2,
+	.channels_mask = BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A7743 has one RGB output and one LVDS output
@@ -61,7 +61,7 @@ static const struct rcar_du_device_info rzg1_du_r8a7745_info = {
 	.gen = 2,
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS,
-	.num_crtcs = 2,
+	.channels_mask = BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A7745 has two RGB outputs
@@ -80,7 +80,7 @@ static const struct rcar_du_device_info rzg1_du_r8a7745_info = {
 static const struct rcar_du_device_info rcar_du_r8a7779_info = {
 	.gen = 2,
 	.features = 0,
-	.num_crtcs = 2,
+	.channels_mask = BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A7779 has two RGB outputs and one (currently unsupported)
@@ -102,7 +102,7 @@ static const struct rcar_du_device_info rcar_du_r8a7790_info = {
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS,
 	.quirks = RCAR_DU_QUIRK_ALIGN_128B,
-	.num_crtcs = 3,
+	.channels_mask = BIT(2) | BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A7790 has one RGB output, two LVDS outputs and one
@@ -129,7 +129,7 @@ static const struct rcar_du_device_info rcar_du_r8a7791_info = {
 	.gen = 2,
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS,
-	.num_crtcs = 2,
+	.channels_mask = BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A779[13] has one RGB output, one LVDS output and one
@@ -151,7 +151,7 @@ static const struct rcar_du_device_info rcar_du_r8a7792_info = {
 	.gen = 2,
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS,
-	.num_crtcs = 2,
+	.channels_mask = BIT(1) | BIT(0),
 	.routes = {
 		/* R8A7792 has two RGB outputs. */
 		[RCAR_DU_OUTPUT_DPAD0] = {
@@ -169,7 +169,7 @@ static const struct rcar_du_device_info rcar_du_r8a7794_info = {
 	.gen = 2,
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS,
-	.num_crtcs = 2,
+	.channels_mask = BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A7794 has two RGB outputs and one (currently unsupported)
@@ -191,7 +191,7 @@ static const struct rcar_du_device_info rcar_du_r8a7795_info = {
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS
 		  | RCAR_DU_FEATURE_VSP1_SOURCE,
-	.num_crtcs = 4,
+	.channels_mask = BIT(3) | BIT(2) | BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A7795 has one RGB output, two HDMI outputs and one
@@ -215,7 +215,7 @@ static const struct rcar_du_device_info rcar_du_r8a7795_info = {
 		},
 	},
 	.num_lvds = 1,
-	.dpll_ch =  BIT(1) | BIT(2),
+	.dpll_ch =  BIT(2) | BIT(1),
 };
 
 static const struct rcar_du_device_info rcar_du_r8a7796_info = {
@@ -223,7 +223,7 @@ static const struct rcar_du_device_info rcar_du_r8a7796_info = {
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS
 		  | RCAR_DU_FEATURE_VSP1_SOURCE,
-	.num_crtcs = 3,
+	.channels_mask = BIT(2) | BIT(1) | BIT(0),
 	.routes = {
 		/*
 		 * R8A7796 has one RGB output, one LVDS output and one HDMI
@@ -246,12 +246,40 @@ static const struct rcar_du_device_info rcar_du_r8a7796_info = {
 	.dpll_ch =  BIT(1),
 };
 
+static const struct rcar_du_device_info rcar_du_r8a77965_info = {
+	.gen = 3,
+	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
+		  | RCAR_DU_FEATURE_EXT_CTRL_REGS
+		  | RCAR_DU_FEATURE_VSP1_SOURCE,
+	.channels_mask = BIT(3) | BIT(1) | BIT(0),
+	.routes = {
+		/*
+		 * R8A77965 has one RGB output, one LVDS output and one HDMI
+		 * output.
+		 */
+		[RCAR_DU_OUTPUT_DPAD0] = {
+			.possible_crtcs = BIT(2),
+			.port = 0,
+		},
+		[RCAR_DU_OUTPUT_HDMI0] = {
+			.possible_crtcs = BIT(1),
+			.port = 1,
+		},
+		[RCAR_DU_OUTPUT_LVDS0] = {
+			.possible_crtcs = BIT(0),
+			.port = 2,
+		},
+	},
+	.num_lvds = 1,
+	.dpll_ch =  BIT(1),
+};
+
 static const struct rcar_du_device_info rcar_du_r8a77970_info = {
 	.gen = 3,
 	.features = RCAR_DU_FEATURE_CRTC_IRQ_CLOCK
 		  | RCAR_DU_FEATURE_EXT_CTRL_REGS
 		  | RCAR_DU_FEATURE_VSP1_SOURCE,
-	.num_crtcs = 1,
+	.channels_mask = BIT(0),
 	.routes = {
 		/* R8A77970 has one RGB output and one LVDS output. */
 		[RCAR_DU_OUTPUT_DPAD0] = {
@@ -277,6 +305,7 @@ static const struct of_device_id rcar_du_of_table[] = {
 	{ .compatible = "renesas,du-r8a7794", .data = &rcar_du_r8a7794_info },
 	{ .compatible = "renesas,du-r8a7795", .data = &rcar_du_r8a7795_info },
 	{ .compatible = "renesas,du-r8a7796", .data = &rcar_du_r8a7796_info },
+	{ .compatible = "renesas,du-r8a77965", .data = &rcar_du_r8a77965_info },
 	{ .compatible = "renesas,du-r8a77970", .data = &rcar_du_r8a77970_info },
 	{ }
 };
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_drv.h b/drivers/gpu/drm/rcar-du/rcar_du_drv.h
index 131d8e88b06c..b3a25e8e07d0 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_drv.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_drv.h
@@ -52,7 +52,7 @@ struct rcar_du_output_routing {
  * @gen: device generation (2 or 3)
  * @features: device features (RCAR_DU_FEATURE_*)
  * @quirks: device quirks (RCAR_DU_QUIRK_*)
- * @num_crtcs: total number of CRTCs
+ * @channels_mask: bit mask of available DU channels
  * @routes: array of CRTC to output routes, indexed by output (RCAR_DU_OUTPUT_*)
  * @num_lvds: number of internal LVDS encoders
  */
@@ -60,7 +60,7 @@ struct rcar_du_device_info {
 	unsigned int gen;
 	unsigned int features;
 	unsigned int quirks;
-	unsigned int num_crtcs;
+	unsigned int channels_mask;
 	struct rcar_du_output_routing routes[RCAR_DU_OUTPUT_MAX];
 	unsigned int num_lvds;
 	unsigned int dpll_ch;
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_group.c b/drivers/gpu/drm/rcar-du/rcar_du_group.c
index 2f37ea901873..d539cb290a35 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_group.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_group.c
@@ -46,10 +46,13 @@ void rcar_du_group_write(struct rcar_du_group *rgrp, u32 reg, u32 data)
 
 static void rcar_du_group_setup_pins(struct rcar_du_group *rgrp)
 {
-	u32 defr6 = DEFR6_CODE | DEFR6_ODPM12_DISP;
+	u32 defr6 = DEFR6_CODE;
 
-	if (rgrp->num_crtcs > 1)
-		defr6 |= DEFR6_ODPM22_DISP;
+	if (rgrp->channels_mask & BIT(0))
+		defr6 |= DEFR6_ODPM02_DISP;
+
+	if (rgrp->channels_mask & BIT(1))
+		defr6 |= DEFR6_ODPM12_DISP;
 
 	rcar_du_group_write(rgrp, DEFR6, defr6);
 }
@@ -80,10 +83,11 @@ static void rcar_du_group_setup_defr8(struct rcar_du_group *rgrp)
 		 * On Gen3 VSPD routing can't be configured, but DPAD routing
 		 * needs to be set despite having a single option available.
 		 */
-		u32 crtc = ffs(possible_crtcs) - 1;
+		unsigned int rgb_crtc = ffs(possible_crtcs) - 1;
+		struct rcar_du_crtc *crtc = &rcdu->crtcs[rgb_crtc];
 
-		if (crtc / 2 == rgrp->index)
-			defr8 |= DEFR8_DRGBS_DU(crtc);
+		if (crtc->index / 2 == rgrp->index)
+			defr8 |= DEFR8_DRGBS_DU(crtc->index);
 	}
 
 	rcar_du_group_write(rgrp, DEFR8, defr8);
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_group.h b/drivers/gpu/drm/rcar-du/rcar_du_group.h
index 5e3adc6b31b5..42105aedecc8 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_group.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_group.h
@@ -25,6 +25,7 @@ struct rcar_du_device;
  * @dev: the DU device
  * @mmio_offset: registers offset in the device memory map
  * @index: group index
+ * @channels_mask: bitmask of populated DU channels in this group
  * @num_crtcs: number of CRTCs in this group (1 or 2)
  * @use_count: number of users of the group (rcar_du_group_(get|put))
  * @used_crtcs: number of CRTCs currently in use
@@ -39,6 +40,7 @@ struct rcar_du_group {
 	unsigned int mmio_offset;
 	unsigned int index;
 
+	unsigned int channels_mask;
 	unsigned int num_crtcs;
 	unsigned int use_count;
 	unsigned int used_crtcs;
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_kms.c b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
index f4ac0f884f00..f0bc7cc0e913 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_kms.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_kms.c
@@ -428,7 +428,7 @@ static int rcar_du_vsps_init(struct rcar_du_device *rcdu)
 	struct {
 		struct device_node *np;
 		unsigned int crtcs_mask;
-	} vsps[RCAR_DU_MAX_VSPS] = { { 0, }, };
+	} vsps[RCAR_DU_MAX_VSPS] = { { NULL, }, };
 	unsigned int vsps_count = 0;
 	unsigned int cells;
 	unsigned int i;
@@ -507,6 +507,8 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 	struct drm_fbdev_cma *fbdev;
 	unsigned int num_encoders;
 	unsigned int num_groups;
+	unsigned int swindex;
+	unsigned int hwindex;
 	unsigned int i;
 	int ret;
 
@@ -520,7 +522,7 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 	dev->mode_config.funcs = &rcar_du_mode_config_funcs;
 	dev->mode_config.helper_private = &rcar_du_mode_config_helper;
 
-	rcdu->num_crtcs = rcdu->info->num_crtcs;
+	rcdu->num_crtcs = hweight8(rcdu->info->channels_mask);
 
 	ret = rcar_du_properties_init(rcdu);
 	if (ret < 0)
@@ -530,7 +532,7 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 	 * Initialize vertical blanking interrupts handling. Start with vblank
 	 * disabled for all CRTCs.
 	 */
-	ret = drm_vblank_init(dev, (1 << rcdu->info->num_crtcs) - 1);
+	ret = drm_vblank_init(dev, (1 << rcdu->num_crtcs) - 1);
 	if (ret < 0)
 		return ret;
 
@@ -545,7 +547,10 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 		rgrp->dev = rcdu;
 		rgrp->mmio_offset = mmio_offsets[i];
 		rgrp->index = i;
-		rgrp->num_crtcs = min(rcdu->num_crtcs - 2 * i, 2U);
+		/* Extract the channel mask for this group only. */
+		rgrp->channels_mask = (rcdu->info->channels_mask >> (2 * i))
+				   & GENMASK(1, 0);
+		rgrp->num_crtcs = hweight8(rgrp->channels_mask);
 
 		/*
 		 * If we have more than one CRTCs in this group pre-associate
@@ -572,10 +577,16 @@ int rcar_du_modeset_init(struct rcar_du_device *rcdu)
 	}
 
 	/* Create the CRTCs. */
-	for (i = 0; i < rcdu->num_crtcs; ++i) {
-		struct rcar_du_group *rgrp = &rcdu->groups[i / 2];
+	for (swindex = 0, hwindex = 0; swindex < rcdu->num_crtcs; ++hwindex) {
+		struct rcar_du_group *rgrp;
+
+		/* Skip unpopulated DU channels. */
+		if (!(rcdu->info->channels_mask & BIT(hwindex)))
+			continue;
+
+		rgrp = &rcdu->groups[hwindex / 2];
 
-		ret = rcar_du_crtc_create(rgrp, i);
+		ret = rcar_du_crtc_create(rgrp, swindex++, hwindex);
 		if (ret < 0)
 			return ret;
 	}
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_of.c b/drivers/gpu/drm/rcar-du/rcar_du_of.c
index 68a0b82cb17e..afef69669bb4 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_of.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_of.c
@@ -18,6 +18,7 @@
 
 #include "rcar_du_crtc.h"
 #include "rcar_du_drv.h"
+#include "rcar_du_of.h"
 
 /* -----------------------------------------------------------------------------
  * Generic Overlay Handling
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_of.h b/drivers/gpu/drm/rcar-du/rcar_du_of.h
index c2e65a727e91..8dd3fbe96650 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_of.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_of.h
@@ -11,7 +11,7 @@
 
 struct of_device_id;
 
-#ifdef CONFIG_DRM_RCAR_LVDS
+#if IS_ENABLED(CONFIG_DRM_RCAR_LVDS)
 void __init rcar_du_of_init(const struct of_device_id *of_ids);
 #else
 static inline void rcar_du_of_init(const struct of_device_id *of_ids) { }
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_regs.h b/drivers/gpu/drm/rcar-du/rcar_du_regs.h
index d5bae99d3cfe..9dfd220ceda1 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_regs.h
+++ b/drivers/gpu/drm/rcar-du/rcar_du_regs.h
@@ -187,14 +187,14 @@
 
 #define DEFR6			0x000e8
 #define DEFR6_CODE		(0x7778 << 16)
-#define DEFR6_ODPM22_DSMR	(0 << 10)
-#define DEFR6_ODPM22_DISP	(2 << 10)
-#define DEFR6_ODPM22_CDE	(3 << 10)
-#define DEFR6_ODPM22_MASK	(3 << 10)
-#define DEFR6_ODPM12_DSMR	(0 << 8)
-#define DEFR6_ODPM12_DISP	(2 << 8)
-#define DEFR6_ODPM12_CDE	(3 << 8)
-#define DEFR6_ODPM12_MASK	(3 << 8)
+#define DEFR6_ODPM12_DSMR	(0 << 10)
+#define DEFR6_ODPM12_DISP	(2 << 10)
+#define DEFR6_ODPM12_CDE	(3 << 10)
+#define DEFR6_ODPM12_MASK	(3 << 10)
+#define DEFR6_ODPM02_DSMR	(0 << 8)
+#define DEFR6_ODPM02_DISP	(2 << 8)
+#define DEFR6_ODPM02_CDE	(3 << 8)
+#define DEFR6_ODPM02_MASK	(3 << 8)
 #define DEFR6_TCNE1		(1 << 6)
 #define DEFR6_TCNE0		(1 << 4)
 #define DEFR6_MLOS1		(1 << 2)
diff --git a/drivers/gpu/drm/rcar-du/rcar_du_vsp.c b/drivers/gpu/drm/rcar-du/rcar_du_vsp.c
index b3bec0125696..72eebeda518e 100644
--- a/drivers/gpu/drm/rcar-du/rcar_du_vsp.c
+++ b/drivers/gpu/drm/rcar-du/rcar_du_vsp.c
@@ -17,6 +17,7 @@
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_fb_cma_helper.h>
 #include <drm/drm_gem_cma_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 #include <drm/drm_plane_helper.h>
 
 #include <linux/bitops.h>
@@ -31,7 +32,7 @@
 #include "rcar_du_kms.h"
 #include "rcar_du_vsp.h"
 
-static void rcar_du_vsp_complete(void *private, bool completed)
+static void rcar_du_vsp_complete(void *private, bool completed, u32 crc)
 {
 	struct rcar_du_crtc *crtc = private;
 
@@ -40,6 +41,8 @@ static void rcar_du_vsp_complete(void *private, bool completed)
 
 	if (completed)
 		rcar_du_crtc_finish_page_flip(crtc);
+
+	drm_crtc_add_crc_entry(&crtc->crtc, false, 0, &crc);
 }
 
 void rcar_du_vsp_enable(struct rcar_du_crtc *crtc)
@@ -102,7 +105,13 @@ void rcar_du_vsp_atomic_begin(struct rcar_du_crtc *crtc)
 
 void rcar_du_vsp_atomic_flush(struct rcar_du_crtc *crtc)
 {
-	vsp1_du_atomic_flush(crtc->vsp->vsp, crtc->vsp_pipe);
+	struct vsp1_du_atomic_pipe_config cfg = { { 0, } };
+	struct rcar_du_crtc_state *state;
+
+	state = to_rcar_crtc_state(crtc->crtc.state);
+	cfg.crc = state->crc;
+
+	vsp1_du_atomic_flush(crtc->vsp->vsp, crtc->vsp_pipe, &cfg);
 }
 
 /* Keep the two tables in sync. */
@@ -237,6 +246,10 @@ static int rcar_du_vsp_plane_prepare_fb(struct drm_plane *plane,
 		}
 	}
 
+	ret = drm_gem_fb_prepare_fb(plane, state);
+	if (ret)
+		goto fail;
+
 	return 0;
 
 fail:
@@ -299,14 +312,12 @@ static const struct drm_plane_helper_funcs rcar_du_vsp_plane_helper_funcs = {
 static struct drm_plane_state *
 rcar_du_vsp_plane_atomic_duplicate_state(struct drm_plane *plane)
 {
-	struct rcar_du_vsp_plane_state *state;
 	struct rcar_du_vsp_plane_state *copy;
 
 	if (WARN_ON(!plane->state))
 		return NULL;
 
-	state = to_rcar_vsp_plane_state(plane->state);
-	copy = kmemdup(state, sizeof(*state), GFP_KERNEL);
+	copy = kzalloc(sizeof(*copy), GFP_KERNEL);
 	if (copy == NULL)
 		return NULL;
 
diff --git a/drivers/gpu/drm/rcar-du/rcar_lvds.c b/drivers/gpu/drm/rcar-du/rcar_lvds.c
index 3d2d3bbd1342..4c39de3f4f0f 100644
--- a/drivers/gpu/drm/rcar-du/rcar_lvds.c
+++ b/drivers/gpu/drm/rcar-du/rcar_lvds.c
@@ -88,6 +88,9 @@ static int rcar_lvds_connector_atomic_check(struct drm_connector *connector,
 	const struct drm_display_mode *panel_mode;
 	struct drm_crtc_state *crtc_state;
 
+	if (!state->crtc)
+		return 0;
+
 	if (list_empty(&connector->modes)) {
 		dev_dbg(lvds->dev, "connector: empty modes list\n");
 		return -EINVAL;
@@ -350,7 +353,7 @@ static int rcar_lvds_attach(struct drm_bridge *bridge)
 
 	drm_connector_helper_add(connector, &rcar_lvds_conn_helper_funcs);
 
-	ret = drm_mode_connector_attach_encoder(connector, encoder);
+	ret = drm_connector_attach_encoder(connector, encoder);
 	if (ret < 0)
 		return ret;
 
@@ -431,8 +434,8 @@ static int rcar_lvds_parse_dt(struct rcar_lvds *lvds)
 			ret = -EPROBE_DEFER;
 	} else {
 		lvds->panel = of_drm_find_panel(remote);
-		if (!lvds->panel)
-			ret = -EPROBE_DEFER;
+		if (IS_ERR(lvds->panel))
+			ret = PTR_ERR(lvds->panel);
 	}
 
 done:
diff --git a/drivers/gpu/drm/rockchip/cdn-dp-core.c b/drivers/gpu/drm/rockchip/cdn-dp-core.c
index c6fbdcd87c16..8ad0d773dc33 100644
--- a/drivers/gpu/drm/rockchip/cdn-dp-core.c
+++ b/drivers/gpu/drm/rockchip/cdn-dp-core.c
@@ -275,7 +275,7 @@ static int cdn_dp_connector_get_modes(struct drm_connector *connector)
 		dp->sink_has_audio = drm_detect_monitor_audio(edid);
 		ret = drm_add_edid_modes(connector, edid);
 		if (ret)
-			drm_mode_connector_update_edid_property(connector,
+			drm_connector_update_edid_property(connector,
 								edid);
 	}
 	mutex_unlock(&dp->lock);
@@ -1062,7 +1062,7 @@ static int cdn_dp_bind(struct device *dev, struct device *master, void *data)
 
 	drm_connector_helper_add(connector, &cdn_dp_connector_helper_funcs);
 
-	ret = drm_mode_connector_attach_encoder(connector, encoder);
+	ret = drm_connector_attach_encoder(connector, encoder);
 	if (ret) {
 		DRM_ERROR("failed to attach connector and encoder\n");
 		goto err_free_connector;
diff --git a/drivers/gpu/drm/rockchip/cdn-dp-reg.c b/drivers/gpu/drm/rockchip/cdn-dp-reg.c
index eb3042c6d1b2..3105965fc260 100644
--- a/drivers/gpu/drm/rockchip/cdn-dp-reg.c
+++ b/drivers/gpu/drm/rockchip/cdn-dp-reg.c
@@ -792,7 +792,6 @@ err_config_video:
 
 int cdn_dp_audio_stop(struct cdn_dp_device *dp, struct audio_info *audio)
 {
-	u32 val;
 	int ret;
 
 	ret = cdn_dp_reg_write(dp, AUDIO_PACK_CONTROL, 0);
@@ -801,11 +800,7 @@ int cdn_dp_audio_stop(struct cdn_dp_device *dp, struct audio_info *audio)
 		return ret;
 	}
 
-	val = SPDIF_AVG_SEL | SPDIF_JITTER_BYPASS;
-	val |= SPDIF_FIFO_MID_RANGE(0xe0);
-	val |= SPDIF_JITTER_THRSH(0xe0);
-	val |= SPDIF_JITTER_AVG_WIN(7);
-	writel(val, dp->regs + SPDIF_CTRL_ADDR);
+	writel(0, dp->regs + SPDIF_CTRL_ADDR);
 
 	/* clearn the audio config and reset */
 	writel(0, dp->regs + AUDIO_SRC_CNTL);
@@ -929,12 +924,6 @@ static void cdn_dp_audio_config_spdif(struct cdn_dp_device *dp)
 {
 	u32 val;
 
-	val = SPDIF_AVG_SEL | SPDIF_JITTER_BYPASS;
-	val |= SPDIF_FIFO_MID_RANGE(0xe0);
-	val |= SPDIF_JITTER_THRSH(0xe0);
-	val |= SPDIF_JITTER_AVG_WIN(7);
-	writel(val, dp->regs + SPDIF_CTRL_ADDR);
-
 	writel(SYNC_WR_TO_CH_ZERO, dp->regs + FIFO_CNTL);
 
 	val = MAX_NUM_CH(2) | AUDIO_TYPE_LPCM | CFG_SUB_PCKT_NUM(4);
@@ -942,9 +931,6 @@ static void cdn_dp_audio_config_spdif(struct cdn_dp_device *dp)
 	writel(SMPL2PKT_EN, dp->regs + SMPL2PKT_CNTL);
 
 	val = SPDIF_ENABLE | SPDIF_AVG_SEL | SPDIF_JITTER_BYPASS;
-	val |= SPDIF_FIFO_MID_RANGE(0xe0);
-	val |= SPDIF_JITTER_THRSH(0xe0);
-	val |= SPDIF_JITTER_AVG_WIN(7);
 	writel(val, dp->regs + SPDIF_CTRL_ADDR);
 
 	clk_prepare_enable(dp->spdif_clk);
diff --git a/drivers/gpu/drm/rockchip/dw-mipi-dsi.c b/drivers/gpu/drm/rockchip/dw-mipi-dsi.c
index d53d5a09547f..662b6cb5d3f0 100644
--- a/drivers/gpu/drm/rockchip/dw-mipi-dsi.c
+++ b/drivers/gpu/drm/rockchip/dw-mipi-dsi.c
@@ -595,7 +595,7 @@ static int dw_mipi_dsi_host_attach(struct mipi_dsi_host *host,
 	dsi->format = device->format;
 	dsi->mode_flags = device->mode_flags;
 	dsi->panel = of_drm_find_panel(device->dev.of_node);
-	if (dsi->panel)
+	if (!IS_ERR(dsi->panel))
 		return drm_panel_attach(dsi->panel, &dsi->connector);
 
 	return -EINVAL;
@@ -1129,7 +1129,7 @@ static int dw_mipi_dsi_register(struct drm_device *drm,
 			   &dw_mipi_dsi_atomic_connector_funcs,
 			   DRM_MODE_CONNECTOR_DSI);
 
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/rockchip/inno_hdmi.c b/drivers/gpu/drm/rockchip/inno_hdmi.c
index 88d0774c97bd..1c02b3e61299 100644
--- a/drivers/gpu/drm/rockchip/inno_hdmi.c
+++ b/drivers/gpu/drm/rockchip/inno_hdmi.c
@@ -565,7 +565,7 @@ static int inno_hdmi_connector_get_modes(struct drm_connector *connector)
 	if (edid) {
 		hdmi->hdmi_data.sink_is_hdmi = drm_detect_hdmi_monitor(edid);
 		hdmi->hdmi_data.sink_has_audio = drm_detect_monitor_audio(edid);
-		drm_mode_connector_update_edid_property(connector, edid);
+		drm_connector_update_edid_property(connector, edid);
 		ret = drm_add_edid_modes(connector, edid);
 		kfree(edid);
 	}
@@ -634,7 +634,7 @@ static int inno_hdmi_register(struct drm_device *drm, struct inno_hdmi *hdmi)
 	drm_connector_init(drm, &hdmi->connector, &inno_hdmi_connector_funcs,
 			   DRM_MODE_CONNECTOR_HDMIA);
 
-	drm_mode_connector_attach_encoder(&hdmi->connector, encoder);
+	drm_connector_attach_encoder(&hdmi->connector, encoder);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_fb.c b/drivers/gpu/drm/rockchip/rockchip_drm_fb.c
index d4f4118b482d..ea18cb2a76c0 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_fb.c
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_fb.c
@@ -18,52 +18,13 @@
 #include <drm/drm_atomic.h>
 #include <drm/drm_fb_helper.h>
 #include <drm/drm_crtc_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 
 #include "rockchip_drm_drv.h"
 #include "rockchip_drm_fb.h"
 #include "rockchip_drm_gem.h"
 #include "rockchip_drm_psr.h"
 
-#define to_rockchip_fb(x) container_of(x, struct rockchip_drm_fb, fb)
-
-struct rockchip_drm_fb {
-	struct drm_framebuffer fb;
-	struct drm_gem_object *obj[ROCKCHIP_MAX_FB_BUFFER];
-};
-
-struct drm_gem_object *rockchip_fb_get_gem_obj(struct drm_framebuffer *fb,
-					       unsigned int plane)
-{
-	struct rockchip_drm_fb *rk_fb = to_rockchip_fb(fb);
-
-	if (plane >= ROCKCHIP_MAX_FB_BUFFER)
-		return NULL;
-
-	return rk_fb->obj[plane];
-}
-
-static void rockchip_drm_fb_destroy(struct drm_framebuffer *fb)
-{
-	struct rockchip_drm_fb *rockchip_fb = to_rockchip_fb(fb);
-	int i;
-
-	for (i = 0; i < ROCKCHIP_MAX_FB_BUFFER; i++)
-		drm_gem_object_put_unlocked(rockchip_fb->obj[i]);
-
-	drm_framebuffer_cleanup(fb);
-	kfree(rockchip_fb);
-}
-
-static int rockchip_drm_fb_create_handle(struct drm_framebuffer *fb,
-					 struct drm_file *file_priv,
-					 unsigned int *handle)
-{
-	struct rockchip_drm_fb *rockchip_fb = to_rockchip_fb(fb);
-
-	return drm_gem_handle_create(file_priv,
-				     rockchip_fb->obj[0], handle);
-}
-
 static int rockchip_drm_fb_dirty(struct drm_framebuffer *fb,
 				 struct drm_file *file,
 				 unsigned int flags, unsigned int color,
@@ -75,46 +36,45 @@ static int rockchip_drm_fb_dirty(struct drm_framebuffer *fb,
 }
 
 static const struct drm_framebuffer_funcs rockchip_drm_fb_funcs = {
-	.destroy	= rockchip_drm_fb_destroy,
-	.create_handle	= rockchip_drm_fb_create_handle,
-	.dirty		= rockchip_drm_fb_dirty,
+	.destroy       = drm_gem_fb_destroy,
+	.create_handle = drm_gem_fb_create_handle,
+	.dirty	       = rockchip_drm_fb_dirty,
 };
 
-static struct rockchip_drm_fb *
+static struct drm_framebuffer *
 rockchip_fb_alloc(struct drm_device *dev, const struct drm_mode_fb_cmd2 *mode_cmd,
 		  struct drm_gem_object **obj, unsigned int num_planes)
 {
-	struct rockchip_drm_fb *rockchip_fb;
+	struct drm_framebuffer *fb;
 	int ret;
 	int i;
 
-	rockchip_fb = kzalloc(sizeof(*rockchip_fb), GFP_KERNEL);
-	if (!rockchip_fb)
+	fb = kzalloc(sizeof(*fb), GFP_KERNEL);
+	if (!fb)
 		return ERR_PTR(-ENOMEM);
 
-	drm_helper_mode_fill_fb_struct(dev, &rockchip_fb->fb, mode_cmd);
+	drm_helper_mode_fill_fb_struct(dev, fb, mode_cmd);
 
 	for (i = 0; i < num_planes; i++)
-		rockchip_fb->obj[i] = obj[i];
+		fb->obj[i] = obj[i];
 
-	ret = drm_framebuffer_init(dev, &rockchip_fb->fb,
-				   &rockchip_drm_fb_funcs);
+	ret = drm_framebuffer_init(dev, fb, &rockchip_drm_fb_funcs);
 	if (ret) {
 		DRM_DEV_ERROR(dev->dev,
 			      "Failed to initialize framebuffer: %d\n",
 			      ret);
-		kfree(rockchip_fb);
+		kfree(fb);
 		return ERR_PTR(ret);
 	}
 
-	return rockchip_fb;
+	return fb;
 }
 
 static struct drm_framebuffer *
 rockchip_user_fb_create(struct drm_device *dev, struct drm_file *file_priv,
 			const struct drm_mode_fb_cmd2 *mode_cmd)
 {
-	struct rockchip_drm_fb *rockchip_fb;
+	struct drm_framebuffer *fb;
 	struct drm_gem_object *objs[ROCKCHIP_MAX_FB_BUFFER];
 	struct drm_gem_object *obj;
 	unsigned int hsub;
@@ -153,13 +113,13 @@ rockchip_user_fb_create(struct drm_device *dev, struct drm_file *file_priv,
 		objs[i] = obj;
 	}
 
-	rockchip_fb = rockchip_fb_alloc(dev, mode_cmd, objs, i);
-	if (IS_ERR(rockchip_fb)) {
-		ret = PTR_ERR(rockchip_fb);
+	fb = rockchip_fb_alloc(dev, mode_cmd, objs, i);
+	if (IS_ERR(fb)) {
+		ret = PTR_ERR(fb);
 		goto err_gem_object_unreference;
 	}
 
-	return &rockchip_fb->fb;
+	return fb;
 
 err_gem_object_unreference:
 	for (i--; i >= 0; i--)
@@ -242,13 +202,13 @@ rockchip_drm_framebuffer_init(struct drm_device *dev,
 			      const struct drm_mode_fb_cmd2 *mode_cmd,
 			      struct drm_gem_object *obj)
 {
-	struct rockchip_drm_fb *rockchip_fb;
+	struct drm_framebuffer *fb;
 
-	rockchip_fb = rockchip_fb_alloc(dev, mode_cmd, &obj, 1);
-	if (IS_ERR(rockchip_fb))
-		return ERR_CAST(rockchip_fb);
+	fb = rockchip_fb_alloc(dev, mode_cmd, &obj, 1);
+	if (IS_ERR(fb))
+		return ERR_CAST(fb);
 
-	return &rockchip_fb->fb;
+	return fb;
 }
 
 void rockchip_drm_mode_config_init(struct drm_device *dev)
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_fb.h b/drivers/gpu/drm/rockchip/rockchip_drm_fb.h
index 2fe47f1ee98f..f1265cb1aee8 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_fb.h
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_fb.h
@@ -22,7 +22,4 @@ rockchip_drm_framebuffer_init(struct drm_device *dev,
 void rockchip_drm_framebuffer_fini(struct drm_framebuffer *fb);
 
 void rockchip_drm_mode_config_init(struct drm_device *dev);
-
-struct drm_gem_object *rockchip_fb_get_gem_obj(struct drm_framebuffer *fb,
-					       unsigned int plane);
 #endif /* _ROCKCHIP_DRM_FB_H */
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
index fe3faa7c38d9..1359e5c773e4 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_vop.c
@@ -76,6 +76,9 @@
 #define VOP_WIN_GET_YRGBADDR(vop, win) \
 		vop_readl(vop, win->base + win->phy->yrgb_mst.offset)
 
+#define VOP_WIN_TO_INDEX(vop_win) \
+	((vop_win) - (vop_win)->vop->win)
+
 #define to_vop(x) container_of(x, struct vop, crtc)
 #define to_vop_win(x) container_of(x, struct vop_win, base)
 
@@ -240,18 +243,6 @@ static enum vop_data_format vop_convert_format(uint32_t format)
 	}
 }
 
-static bool is_yuv_support(uint32_t format)
-{
-	switch (format) {
-	case DRM_FORMAT_NV12:
-	case DRM_FORMAT_NV16:
-	case DRM_FORMAT_NV24:
-		return true;
-	default:
-		return false;
-	}
-}
-
 static uint16_t scl_vop_cal_scale(enum scale_mode mode, uint32_t src,
 				  uint32_t dst, bool is_horizontal,
 				  int vsu_mode, int *vskiplines)
@@ -295,7 +286,8 @@ static void scl_vop_cal_scl_fac(struct vop *vop, const struct vop_win_data *win,
 	uint16_t cbcr_ver_scl_mode = SCALE_NONE;
 	int hsub = drm_format_horz_chroma_subsampling(pixel_format);
 	int vsub = drm_format_vert_chroma_subsampling(pixel_format);
-	bool is_yuv = is_yuv_support(pixel_format);
+	const struct drm_format_info *info;
+	bool is_yuv = false;
 	uint16_t cbcr_src_w = src_w / hsub;
 	uint16_t cbcr_src_h = src_h / vsub;
 	uint16_t vsu_mode;
@@ -303,6 +295,11 @@ static void scl_vop_cal_scl_fac(struct vop *vop, const struct vop_win_data *win,
 	uint32_t val;
 	int vskiplines;
 
+	info = drm_format_info(pixel_format);
+
+	if (info->is_yuv)
+		is_yuv = true;
+
 	if (dst_w > 3840) {
 		DRM_DEV_ERROR(vop->dev, "Maximum dst width (3840) exceeded\n");
 		return;
@@ -483,6 +480,31 @@ static void vop_line_flag_irq_disable(struct vop *vop)
 	spin_unlock_irqrestore(&vop->irq_lock, flags);
 }
 
+static int vop_core_clks_enable(struct vop *vop)
+{
+	int ret;
+
+	ret = clk_enable(vop->hclk);
+	if (ret < 0)
+		return ret;
+
+	ret = clk_enable(vop->aclk);
+	if (ret < 0)
+		goto err_disable_hclk;
+
+	return 0;
+
+err_disable_hclk:
+	clk_disable(vop->hclk);
+	return ret;
+}
+
+static void vop_core_clks_disable(struct vop *vop)
+{
+	clk_disable(vop->aclk);
+	clk_disable(vop->hclk);
+}
+
 static int vop_enable(struct drm_crtc *crtc)
 {
 	struct vop *vop = to_vop(crtc);
@@ -494,17 +516,13 @@ static int vop_enable(struct drm_crtc *crtc)
 		return ret;
 	}
 
-	ret = clk_enable(vop->hclk);
+	ret = vop_core_clks_enable(vop);
 	if (WARN_ON(ret < 0))
 		goto err_put_pm_runtime;
 
 	ret = clk_enable(vop->dclk);
 	if (WARN_ON(ret < 0))
-		goto err_disable_hclk;
-
-	ret = clk_enable(vop->aclk);
-	if (WARN_ON(ret < 0))
-		goto err_disable_dclk;
+		goto err_disable_core;
 
 	/*
 	 * Slave iommu shares power, irq and clock with vop.  It was associated
@@ -516,7 +534,7 @@ static int vop_enable(struct drm_crtc *crtc)
 	if (ret) {
 		DRM_DEV_ERROR(vop->dev,
 			      "failed to attach dma mapping, %d\n", ret);
-		goto err_disable_aclk;
+		goto err_disable_dclk;
 	}
 
 	spin_lock(&vop->reg_lock);
@@ -549,18 +567,14 @@ static int vop_enable(struct drm_crtc *crtc)
 
 	spin_unlock(&vop->reg_lock);
 
-	enable_irq(vop->irq);
-
 	drm_crtc_vblank_on(crtc);
 
 	return 0;
 
-err_disable_aclk:
-	clk_disable(vop->aclk);
 err_disable_dclk:
 	clk_disable(vop->dclk);
-err_disable_hclk:
-	clk_disable(vop->hclk);
+err_disable_core:
+	vop_core_clks_disable(vop);
 err_put_pm_runtime:
 	pm_runtime_put_sync(vop->dev);
 	return ret;
@@ -596,8 +610,6 @@ static void vop_crtc_atomic_disable(struct drm_crtc *crtc,
 
 	vop_dsp_hold_valid_irq_disable(vop);
 
-	disable_irq(vop->irq);
-
 	vop->is_enabled = false;
 
 	/*
@@ -606,8 +618,7 @@ static void vop_crtc_atomic_disable(struct drm_crtc *crtc,
 	rockchip_drm_dma_detach_device(vop->drm_dev, vop->dev);
 
 	clk_disable(vop->dclk);
-	clk_disable(vop->aclk);
-	clk_disable(vop->hclk);
+	vop_core_clks_disable(vop);
 	pm_runtime_put(vop->dev);
 	mutex_unlock(&vop->vop_lock);
 
@@ -663,7 +674,7 @@ static int vop_plane_atomic_check(struct drm_plane *plane,
 	 * Src.x1 can be odd when do clip, but yuv plane start point
 	 * need align with 2 pixel.
 	 */
-	if (is_yuv_support(fb->format->format) && ((state->src.x1 >> 16) % 2)) {
+	if (fb->format->is_yuv && ((state->src.x1 >> 16) % 2)) {
 		DRM_ERROR("Invalid Source: Yuv format not support odd xpos\n");
 		return -EINVAL;
 	}
@@ -708,6 +719,7 @@ static void vop_plane_atomic_update(struct drm_plane *plane,
 	dma_addr_t dma_addr;
 	uint32_t val;
 	bool rb_swap;
+	int win_index = VOP_WIN_TO_INDEX(vop_win);
 	int format;
 
 	/*
@@ -724,7 +736,7 @@ static void vop_plane_atomic_update(struct drm_plane *plane,
 		return;
 	}
 
-	obj = rockchip_fb_get_gem_obj(fb, 0);
+	obj = fb->obj[0];
 	rk_obj = to_rockchip_obj(obj);
 
 	actual_w = drm_rect_width(src) >> 16;
@@ -749,12 +761,12 @@ static void vop_plane_atomic_update(struct drm_plane *plane,
 	VOP_WIN_SET(vop, win, format, format);
 	VOP_WIN_SET(vop, win, yrgb_vir, DIV_ROUND_UP(fb->pitches[0], 4));
 	VOP_WIN_SET(vop, win, yrgb_mst, dma_addr);
-	if (is_yuv_support(fb->format->format)) {
+	if (fb->format->is_yuv) {
 		int hsub = drm_format_horz_chroma_subsampling(fb->format->format);
 		int vsub = drm_format_vert_chroma_subsampling(fb->format->format);
 		int bpp = fb->format->cpp[1];
 
-		uv_obj = rockchip_fb_get_gem_obj(fb, 1);
+		uv_obj = fb->obj[1];
 		rk_uv_obj = to_rockchip_obj(uv_obj);
 
 		offset = (src->x1 >> 16) * bpp / hsub;
@@ -777,7 +789,14 @@ static void vop_plane_atomic_update(struct drm_plane *plane,
 	rb_swap = has_rb_swapped(fb->format->format);
 	VOP_WIN_SET(vop, win, rb_swap, rb_swap);
 
-	if (fb->format->has_alpha) {
+	/*
+	 * Blending win0 with the background color doesn't seem to work
+	 * correctly. We only get the background color, no matter the contents
+	 * of the win0 framebuffer.  However, blending pre-multiplied color
+	 * with the default opaque black default background color is a no-op,
+	 * so we can just disable blending to get the correct result.
+	 */
+	if (fb->format->has_alpha && win_index > 0) {
 		VOP_WIN_SET(vop, win, dst_alpha_ctl,
 			    DST_FACTOR_M0(ALPHA_SRC_INVERSE));
 		val = SRC_ALPHA_EN(1) | SRC_COLOR_M0(ALPHA_SRC_PRE_MUL) |
@@ -1167,6 +1186,18 @@ static irqreturn_t vop_isr(int irq, void *data)
 	int ret = IRQ_NONE;
 
 	/*
+	 * The irq is shared with the iommu. If the runtime-pm state of the
+	 * vop-device is disabled the irq has to be targeted at the iommu.
+	 */
+	if (!pm_runtime_get_if_in_use(vop->dev))
+		return IRQ_NONE;
+
+	if (vop_core_clks_enable(vop)) {
+		DRM_DEV_ERROR_RATELIMITED(vop->dev, "couldn't enable clocks\n");
+		goto out;
+	}
+
+	/*
 	 * interrupt register has interrupt status, enable and clear bits, we
 	 * must hold irq_lock to avoid a race with enable/disable_vblank().
 	*/
@@ -1181,7 +1212,7 @@ static irqreturn_t vop_isr(int irq, void *data)
 
 	/* This is expected for vop iommu irqs, since the irq is shared */
 	if (!active_irqs)
-		return IRQ_NONE;
+		goto out_disable;
 
 	if (active_irqs & DSP_HOLD_VALID_INTR) {
 		complete(&vop->dsp_hold_completion);
@@ -1207,6 +1238,10 @@ static irqreturn_t vop_isr(int irq, void *data)
 		DRM_DEV_ERROR(vop->dev, "Unknown VOP IRQs: %#02x\n",
 			      active_irqs);
 
+out_disable:
+	vop_core_clks_disable(vop);
+out:
+	pm_runtime_put(vop->dev);
 	return ret;
 }
 
@@ -1267,7 +1302,7 @@ static int vop_create_crtc(struct vop *vop)
 	for (i = 0; i < vop_data->win_size; i++) {
 		struct vop_win *vop_win = &vop->win[i];
 		const struct vop_win_data *win_data = vop_win->data;
-		unsigned long possible_crtcs = 1 << drm_crtc_index(crtc);
+		unsigned long possible_crtcs = drm_crtc_mask(crtc);
 
 		if (win_data->type != DRM_PLANE_TYPE_OVERLAY)
 			continue;
@@ -1585,9 +1620,6 @@ static int vop_bind(struct device *dev, struct device *master, void *data)
 	if (ret)
 		goto err_disable_pm_runtime;
 
-	/* IRQ is initially disabled; it gets enabled in power_on */
-	disable_irq(vop->irq);
-
 	return 0;
 
 err_disable_pm_runtime:
diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_vop.h b/drivers/gpu/drm/rockchip/rockchip_drm_vop.h
index 084acdd0019a..fcb91041a666 100644
--- a/drivers/gpu/drm/rockchip/rockchip_drm_vop.h
+++ b/drivers/gpu/drm/rockchip/rockchip_drm_vop.h
@@ -331,16 +331,19 @@ static inline int scl_vop_cal_lb_mode(int width, bool is_yuv)
 {
 	int lb_mode;
 
-	if (width > 2560)
-		lb_mode = LB_RGB_3840X2;
-	else if (width > 1920)
-		lb_mode = LB_RGB_2560X4;
-	else if (!is_yuv)
-		lb_mode = LB_RGB_1920X5;
-	else if (width > 1280)
-		lb_mode = LB_YUV_3840X5;
-	else
-		lb_mode = LB_YUV_2560X8;
+	if (is_yuv) {
+		if (width > 1280)
+			lb_mode = LB_YUV_3840X5;
+		else
+			lb_mode = LB_YUV_2560X8;
+	} else {
+		if (width > 2560)
+			lb_mode = LB_RGB_3840X2;
+		else if (width > 1920)
+			lb_mode = LB_RGB_2560X4;
+		else
+			lb_mode = LB_RGB_1920X5;
+	}
 
 	return lb_mode;
 }
diff --git a/drivers/gpu/drm/rockchip/rockchip_lvds.c b/drivers/gpu/drm/rockchip/rockchip_lvds.c
index e67f4ea28c0e..456bd9f13bae 100644
--- a/drivers/gpu/drm/rockchip/rockchip_lvds.c
+++ b/drivers/gpu/drm/rockchip/rockchip_lvds.c
@@ -363,8 +363,10 @@ static int rockchip_lvds_bind(struct device *dev, struct device *master,
 		of_property_read_u32(endpoint, "reg", &endpoint_id);
 		ret = drm_of_find_panel_or_bridge(dev->of_node, 1, endpoint_id,
 						  &lvds->panel, &lvds->bridge);
-		if (!ret)
+		if (!ret) {
+			of_node_put(endpoint);
 			break;
+		}
 	}
 	if (!child_count) {
 		DRM_DEV_ERROR(dev, "lvds port does not have any children\n");
@@ -432,7 +434,7 @@ static int rockchip_lvds_bind(struct device *dev, struct device *master,
 		drm_connector_helper_add(connector,
 					 &rockchip_lvds_connector_helper_funcs);
 
-		ret = drm_mode_connector_attach_encoder(connector, encoder);
+		ret = drm_connector_attach_encoder(connector, encoder);
 		if (ret < 0) {
 			DRM_DEV_ERROR(drm_dev->dev,
 				      "failed to attach encoder: %d\n", ret);
@@ -446,14 +448,12 @@ static int rockchip_lvds_bind(struct device *dev, struct device *master,
 			goto err_free_connector;
 		}
 	} else {
-		lvds->bridge->encoder = encoder;
 		ret = drm_bridge_attach(encoder, lvds->bridge, NULL);
 		if (ret) {
 			DRM_DEV_ERROR(drm_dev->dev,
 				      "failed to attach bridge: %d\n", ret);
 			goto err_free_encoder;
 		}
-		encoder->bridge = lvds->bridge;
 	}
 
 	pm_runtime_enable(dev);
diff --git a/drivers/gpu/drm/savage/savage_bci.c b/drivers/gpu/drm/savage/savage_bci.c
index 2a5b8466d806..35dc74883f83 100644
--- a/drivers/gpu/drm/savage/savage_bci.c
+++ b/drivers/gpu/drm/savage/savage_bci.c
@@ -298,8 +298,9 @@ static int savage_dma_init(drm_savage_private_t * dev_priv)
 
 	dev_priv->nr_dma_pages = dev_priv->cmd_dma->size /
 	    (SAVAGE_DMA_PAGE_SIZE * 4);
-	dev_priv->dma_pages = kmalloc(sizeof(drm_savage_dma_page_t) *
-				      dev_priv->nr_dma_pages, GFP_KERNEL);
+	dev_priv->dma_pages = kmalloc_array(dev_priv->nr_dma_pages,
+					    sizeof(drm_savage_dma_page_t),
+					    GFP_KERNEL);
 	if (dev_priv->dma_pages == NULL)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/savage/savage_state.c b/drivers/gpu/drm/savage/savage_state.c
index 2db89bed52e8..7559a820bd43 100644
--- a/drivers/gpu/drm/savage/savage_state.c
+++ b/drivers/gpu/drm/savage/savage_state.c
@@ -971,7 +971,7 @@ int savage_bci_cmdbuf(struct drm_device *dev, void *data, struct drm_file *file_
 	LOCK_TEST_WITH_RETURN(dev, file_priv);
 
 	if (dma && dma->buflist) {
-		if (cmdbuf->dma_idx > dma->buf_count) {
+		if (cmdbuf->dma_idx >= dma->buf_count) {
 			DRM_ERROR
 			    ("vertex buffer index %u out of range (0-%u)\n",
 			     cmdbuf->dma_idx, dma->buf_count - 1);
diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler.c b/drivers/gpu/drm/scheduler/gpu_scheduler.c
index 0d95888ccc3e..dac71e3b4514 100644
--- a/drivers/gpu/drm/scheduler/gpu_scheduler.c
+++ b/drivers/gpu/drm/scheduler/gpu_scheduler.c
@@ -21,6 +21,29 @@
  *
  */
 
+/**
+ * DOC: Overview
+ *
+ * The GPU scheduler provides entities which allow userspace to push jobs
+ * into software queues which are then scheduled on a hardware run queue.
+ * The software queues have a priority among them. The scheduler selects the entities
+ * from the run queue using a FIFO. The scheduler provides dependency handling
+ * features among jobs. The driver is supposed to provide callback functions for
+ * backend operations to the scheduler like submitting a job to hardware run queue,
+ * returning the dependencies of a job etc.
+ *
+ * The organisation of the scheduler is the following:
+ *
+ * 1. Each hw run queue has one scheduler
+ * 2. Each scheduler has multiple run queues with different priorities
+ *    (e.g., HIGH_HW,HIGH_SW, KERNEL, NORMAL)
+ * 3. Each scheduler run queue has a queue of entities to schedule
+ * 4. Entities themselves maintain a queue of jobs that will be scheduled on
+ *    the hardware.
+ *
+ * The jobs in a entity are always scheduled in the order that they were pushed.
+ */
+
 #include <linux/kthread.h>
 #include <linux/wait.h>
 #include <linux/sched.h>
@@ -30,7 +53,7 @@
 #include <drm/spsc_queue.h>
 
 #define CREATE_TRACE_POINTS
-#include <drm/gpu_scheduler_trace.h>
+#include "gpu_scheduler_trace.h"
 
 #define to_drm_sched_job(sched_job)		\
 		container_of((sched_job), struct drm_sched_job, queue_node)
@@ -39,14 +62,30 @@ static bool drm_sched_entity_is_ready(struct drm_sched_entity *entity);
 static void drm_sched_wakeup(struct drm_gpu_scheduler *sched);
 static void drm_sched_process_job(struct dma_fence *f, struct dma_fence_cb *cb);
 
-/* Initialize a given run queue struct */
-static void drm_sched_rq_init(struct drm_sched_rq *rq)
+/**
+ * drm_sched_rq_init - initialize a given run queue struct
+ *
+ * @rq: scheduler run queue
+ *
+ * Initializes a scheduler runqueue.
+ */
+static void drm_sched_rq_init(struct drm_gpu_scheduler *sched,
+			      struct drm_sched_rq *rq)
 {
 	spin_lock_init(&rq->lock);
 	INIT_LIST_HEAD(&rq->entities);
 	rq->current_entity = NULL;
+	rq->sched = sched;
 }
 
+/**
+ * drm_sched_rq_add_entity - add an entity
+ *
+ * @rq: scheduler run queue
+ * @entity: scheduler entity
+ *
+ * Adds a scheduler entity to the run queue.
+ */
 static void drm_sched_rq_add_entity(struct drm_sched_rq *rq,
 				    struct drm_sched_entity *entity)
 {
@@ -57,6 +96,14 @@ static void drm_sched_rq_add_entity(struct drm_sched_rq *rq,
 	spin_unlock(&rq->lock);
 }
 
+/**
+ * drm_sched_rq_remove_entity - remove an entity
+ *
+ * @rq: scheduler run queue
+ * @entity: scheduler entity
+ *
+ * Removes a scheduler entity from the run queue.
+ */
 static void drm_sched_rq_remove_entity(struct drm_sched_rq *rq,
 				       struct drm_sched_entity *entity)
 {
@@ -70,9 +117,9 @@ static void drm_sched_rq_remove_entity(struct drm_sched_rq *rq,
 }
 
 /**
- * Select an entity which could provide a job to run
+ * drm_sched_rq_select_entity - Select an entity which could provide a job to run
  *
- * @rq		The run queue to check.
+ * @rq: scheduler run queue to check.
  *
  * Try to find a ready entity, returns NULL if none found.
  */
@@ -112,32 +159,37 @@ drm_sched_rq_select_entity(struct drm_sched_rq *rq)
 }
 
 /**
- * Init a context entity used by scheduler when submit to HW ring.
+ * drm_sched_entity_init - Init a context entity used by scheduler when
+ * submit to HW ring.
+ *
+ * @entity: scheduler entity to init
+ * @rq_list: the list of run queue on which jobs from this
+ *           entity can be submitted
+ * @num_rq_list: number of run queue in rq_list
+ * @guilty: atomic_t set to 1 when a job on this queue
+ *          is found to be guilty causing a timeout
  *
- * @sched	The pointer to the scheduler
- * @entity	The pointer to a valid drm_sched_entity
- * @rq		The run queue this entity belongs
- * @kernel	If this is an entity for the kernel
- * @jobs	The max number of jobs in the job queue
+ * Note: the rq_list should have atleast one element to schedule
+ *       the entity
  *
- * return 0 if succeed. negative error code on failure
+ * Returns 0 on success or a negative error code on failure.
 */
-int drm_sched_entity_init(struct drm_gpu_scheduler *sched,
-			  struct drm_sched_entity *entity,
-			  struct drm_sched_rq *rq,
-			  uint32_t jobs, atomic_t *guilty)
+int drm_sched_entity_init(struct drm_sched_entity *entity,
+			  struct drm_sched_rq **rq_list,
+			  unsigned int num_rq_list,
+			  atomic_t *guilty)
 {
-	if (!(sched && entity && rq))
+	if (!(entity && rq_list && num_rq_list > 0 && rq_list[0]))
 		return -EINVAL;
 
 	memset(entity, 0, sizeof(struct drm_sched_entity));
 	INIT_LIST_HEAD(&entity->list);
-	entity->rq = rq;
-	entity->sched = sched;
+	entity->rq = rq_list[0];
+	entity->sched = rq_list[0]->sched;
 	entity->guilty = guilty;
+	entity->last_scheduled = NULL;
 
 	spin_lock_init(&entity->rq_lock);
-	spin_lock_init(&entity->queue_lock);
 	spsc_queue_init(&entity->job_queue);
 
 	atomic_set(&entity->fence_seq, 0);
@@ -148,10 +200,10 @@ int drm_sched_entity_init(struct drm_gpu_scheduler *sched,
 EXPORT_SYMBOL(drm_sched_entity_init);
 
 /**
- * Query if entity is initialized
+ * drm_sched_entity_is_initialized - Query if entity is initialized
  *
- * @sched       Pointer to scheduler instance
- * @entity	The pointer to a valid scheduler entity
+ * @sched: Pointer to scheduler instance
+ * @entity: The pointer to a valid scheduler entity
  *
  * return true if entity is initialized, false otherwise
 */
@@ -163,25 +215,26 @@ static bool drm_sched_entity_is_initialized(struct drm_gpu_scheduler *sched,
 }
 
 /**
- * Check if entity is idle
+ * drm_sched_entity_is_idle - Check if entity is idle
  *
- * @entity	The pointer to a valid scheduler entity
+ * @entity: scheduler entity
  *
- * Return true if entity don't has any unscheduled jobs.
+ * Returns true if the entity does not have any unscheduled jobs.
  */
 static bool drm_sched_entity_is_idle(struct drm_sched_entity *entity)
 {
 	rmb();
-	if (spsc_queue_peek(&entity->job_queue) == NULL)
+
+	if (!entity->rq || spsc_queue_peek(&entity->job_queue) == NULL)
 		return true;
 
 	return false;
 }
 
 /**
- * Check if entity is ready
+ * drm_sched_entity_is_ready - Check if entity is ready
  *
- * @entity	The pointer to a valid scheduler entity
+ * @entity: scheduler entity
  *
  * Return true if entity could provide a job.
  */
@@ -196,33 +249,81 @@ static bool drm_sched_entity_is_ready(struct drm_sched_entity *entity)
 	return true;
 }
 
+static void drm_sched_entity_kill_jobs_cb(struct dma_fence *f,
+				    struct dma_fence_cb *cb)
+{
+	struct drm_sched_job *job = container_of(cb, struct drm_sched_job,
+						 finish_cb);
+	drm_sched_fence_finished(job->s_fence);
+	WARN_ON(job->s_fence->parent);
+	dma_fence_put(&job->s_fence->finished);
+	job->sched->ops->free_job(job);
+}
+
+
 /**
- * Destroy a context entity
+ * drm_sched_entity_flush - Flush a context entity
  *
- * @sched       Pointer to scheduler instance
- * @entity	The pointer to a valid scheduler entity
+ * @sched: scheduler instance
+ * @entity: scheduler entity
+ * @timeout: time to wait in for Q to become empty in jiffies.
  *
- * Cleanup and free the allocated resources.
+ * Splitting drm_sched_entity_fini() into two functions, The first one does the waiting,
+ * removes the entity from the runqueue and returns an error when the process was killed.
+ *
+ * Returns the remaining time in jiffies left from the input timeout
  */
-void drm_sched_entity_fini(struct drm_gpu_scheduler *sched,
-			   struct drm_sched_entity *entity)
+long drm_sched_entity_flush(struct drm_gpu_scheduler *sched,
+			   struct drm_sched_entity *entity, long timeout)
 {
-	int r;
+	long ret = timeout;
 
 	if (!drm_sched_entity_is_initialized(sched, entity))
-		return;
+		return ret;
 	/**
 	 * The client will not queue more IBs during this fini, consume existing
 	 * queued IBs or discard them on SIGKILL
 	*/
-	if ((current->flags & PF_SIGNALED) && current->exit_code == SIGKILL)
-		r = -ERESTARTSYS;
-	else
-		r = wait_event_killable(sched->job_scheduled,
-					drm_sched_entity_is_idle(entity));
+	if (current->flags & PF_EXITING) {
+		if (timeout)
+			ret = wait_event_timeout(
+					sched->job_scheduled,
+					drm_sched_entity_is_idle(entity),
+					timeout);
+	} else
+		wait_event_killable(sched->job_scheduled, drm_sched_entity_is_idle(entity));
+
+
+	/* For killed process disable any more IBs enqueue right now */
+	if ((current->flags & PF_EXITING) && (current->exit_code == SIGKILL))
+		drm_sched_entity_set_rq(entity, NULL);
+
+	return ret;
+}
+EXPORT_SYMBOL(drm_sched_entity_flush);
+
+/**
+ * drm_sched_entity_cleanup - Destroy a context entity
+ *
+ * @sched: scheduler instance
+ * @entity: scheduler entity
+ *
+ * This should be called after @drm_sched_entity_do_release. It goes over the
+ * entity and signals all jobs with an error code if the process was killed.
+ *
+ */
+void drm_sched_entity_fini(struct drm_gpu_scheduler *sched,
+			   struct drm_sched_entity *entity)
+{
+
 	drm_sched_entity_set_rq(entity, NULL);
-	if (r) {
+
+	/* Consumption of existing IBs wasn't completed. Forcefully
+	 * remove them here.
+	 */
+	if (spsc_queue_peek(&entity->job_queue)) {
 		struct drm_sched_job *job;
+		int r;
 
 		/* Park the kernel for a moment to make sure it isn't processing
 		 * our enity.
@@ -240,15 +341,46 @@ void drm_sched_entity_fini(struct drm_gpu_scheduler *sched,
 			struct drm_sched_fence *s_fence = job->s_fence;
 			drm_sched_fence_scheduled(s_fence);
 			dma_fence_set_error(&s_fence->finished, -ESRCH);
-			drm_sched_fence_finished(s_fence);
-			WARN_ON(s_fence->parent);
-			dma_fence_put(&s_fence->finished);
-			sched->ops->free_job(job);
+
+			/*
+			 * When pipe is hanged by older entity, new entity might
+			 * not even have chance to submit it's first job to HW
+			 * and so entity->last_scheduled will remain NULL
+			 */
+			if (!entity->last_scheduled) {
+				drm_sched_entity_kill_jobs_cb(NULL, &job->finish_cb);
+			} else {
+				r = dma_fence_add_callback(entity->last_scheduled, &job->finish_cb,
+								drm_sched_entity_kill_jobs_cb);
+				if (r == -ENOENT)
+					drm_sched_entity_kill_jobs_cb(NULL, &job->finish_cb);
+				else if (r)
+					DRM_ERROR("fence add callback failed (%d)\n", r);
+			}
 		}
 	}
+
+	dma_fence_put(entity->last_scheduled);
+	entity->last_scheduled = NULL;
 }
 EXPORT_SYMBOL(drm_sched_entity_fini);
 
+/**
+ * drm_sched_entity_fini - Destroy a context entity
+ *
+ * @sched: scheduler instance
+ * @entity: scheduler entity
+ *
+ * Calls drm_sched_entity_do_release() and drm_sched_entity_cleanup()
+ */
+void drm_sched_entity_destroy(struct drm_gpu_scheduler *sched,
+				struct drm_sched_entity *entity)
+{
+	drm_sched_entity_flush(sched, entity, MAX_WAIT_SCHED_ENTITY_Q_EMPTY);
+	drm_sched_entity_fini(sched, entity);
+}
+EXPORT_SYMBOL(drm_sched_entity_destroy);
+
 static void drm_sched_entity_wakeup(struct dma_fence *f, struct dma_fence_cb *cb)
 {
 	struct drm_sched_entity *entity =
@@ -266,6 +398,15 @@ static void drm_sched_entity_clear_dep(struct dma_fence *f, struct dma_fence_cb
 	dma_fence_put(f);
 }
 
+/**
+ * drm_sched_entity_set_rq - Sets the run queue for an entity
+ *
+ * @entity: scheduler entity
+ * @rq: scheduler run queue
+ *
+ * Sets the run queue for an entity and removes the entity from the previous
+ * run queue in which was present.
+ */
 void drm_sched_entity_set_rq(struct drm_sched_entity *entity,
 			     struct drm_sched_rq *rq)
 {
@@ -285,6 +426,14 @@ void drm_sched_entity_set_rq(struct drm_sched_entity *entity,
 }
 EXPORT_SYMBOL(drm_sched_entity_set_rq);
 
+/**
+ * drm_sched_dependency_optimized
+ *
+ * @fence: the dependency fence
+ * @entity: the entity which depends on the above fence
+ *
+ * Returns true if the dependency can be optimized and false otherwise
+ */
 bool drm_sched_dependency_optimized(struct dma_fence* fence,
 				    struct drm_sched_entity *entity)
 {
@@ -309,8 +458,13 @@ static bool drm_sched_entity_add_dependency_cb(struct drm_sched_entity *entity)
 	struct dma_fence * fence = entity->dependency;
 	struct drm_sched_fence *s_fence;
 
-	if (fence->context == entity->fence_context) {
-		/* We can ignore fences from ourself */
+	if (fence->context == entity->fence_context ||
+            fence->context == entity->fence_context + 1) {
+                /*
+                 * Fence is a scheduled/finished fence from a job
+                 * which belongs to the same entity, we can ignore
+                 * fences from ourself
+                 */
 		dma_fence_put(entity->dependency);
 		return false;
 	}
@@ -360,14 +514,22 @@ drm_sched_entity_pop_job(struct drm_sched_entity *entity)
 	if (entity->guilty && atomic_read(entity->guilty))
 		dma_fence_set_error(&sched_job->s_fence->finished, -ECANCELED);
 
+	dma_fence_put(entity->last_scheduled);
+	entity->last_scheduled = dma_fence_get(&sched_job->s_fence->finished);
+
 	spsc_queue_pop(&entity->job_queue);
 	return sched_job;
 }
 
 /**
- * Submit a job to the job queue
+ * drm_sched_entity_push_job - Submit a job to the entity's job queue
  *
- * @sched_job		The pointer to job required to submit
+ * @sched_job: job to submit
+ * @entity: scheduler entity
+ *
+ * Note: To guarantee that the order of insertion to queue matches
+ * the job's fence sequence number this function should be
+ * called with drm_sched_job_init under common lock.
  *
  * Returns 0 for success, negative error code otherwise.
  */
@@ -379,15 +541,17 @@ void drm_sched_entity_push_job(struct drm_sched_job *sched_job,
 
 	trace_drm_sched_job(sched_job, entity);
 
-	spin_lock(&entity->queue_lock);
 	first = spsc_queue_push(&entity->job_queue, &sched_job->queue_node);
 
-	spin_unlock(&entity->queue_lock);
-
 	/* first job wakes up scheduler */
 	if (first) {
 		/* Add the entity to the run queue */
 		spin_lock(&entity->rq_lock);
+		if (!entity->rq) {
+			DRM_ERROR("Trying to push to a killed entity\n");
+			spin_unlock(&entity->rq_lock);
+			return;
+		}
 		drm_sched_rq_add_entity(entity->rq, entity);
 		spin_unlock(&entity->rq_lock);
 		drm_sched_wakeup(sched);
@@ -457,6 +621,13 @@ static void drm_sched_job_timedout(struct work_struct *work)
 	job->sched->ops->timedout_job(job);
 }
 
+/**
+ * drm_sched_hw_job_reset - stop the scheduler if it contains the bad job
+ *
+ * @sched: scheduler instance
+ * @bad: bad scheduler job
+ *
+ */
 void drm_sched_hw_job_reset(struct drm_gpu_scheduler *sched, struct drm_sched_job *bad)
 {
 	struct drm_sched_job *s_job;
@@ -501,6 +672,12 @@ void drm_sched_hw_job_reset(struct drm_gpu_scheduler *sched, struct drm_sched_jo
 }
 EXPORT_SYMBOL(drm_sched_hw_job_reset);
 
+/**
+ * drm_sched_job_recovery - recover jobs after a reset
+ *
+ * @sched: scheduler instance
+ *
+ */
 void drm_sched_job_recovery(struct drm_gpu_scheduler *sched)
 {
 	struct drm_sched_job *s_job, *tmp;
@@ -529,6 +706,7 @@ void drm_sched_job_recovery(struct drm_gpu_scheduler *sched)
 		spin_unlock(&sched->job_list_lock);
 		fence = sched->ops->run_job(s_job);
 		atomic_inc(&sched->hw_rq_count);
+
 		if (fence) {
 			s_fence->parent = dma_fence_get(fence);
 			r = dma_fence_add_callback(fence, &s_fence->cb,
@@ -548,13 +726,26 @@ void drm_sched_job_recovery(struct drm_gpu_scheduler *sched)
 }
 EXPORT_SYMBOL(drm_sched_job_recovery);
 
-/* init a sched_job with basic field */
+/**
+ * drm_sched_job_init - init a scheduler job
+ *
+ * @job: scheduler job to init
+ * @sched: scheduler instance
+ * @entity: scheduler entity to use
+ * @owner: job owner for debugging
+ *
+ * Refer to drm_sched_entity_push_job() documentation
+ * for locking considerations.
+ *
+ * Returns 0 for success, negative error code otherwise.
+ */
 int drm_sched_job_init(struct drm_sched_job *job,
 		       struct drm_gpu_scheduler *sched,
 		       struct drm_sched_entity *entity,
 		       void *owner)
 {
 	job->sched = sched;
+	job->entity = entity;
 	job->s_priority = entity->rq - sched->sched_rq;
 	job->s_fence = drm_sched_fence_create(entity, owner);
 	if (!job->s_fence)
@@ -570,7 +761,11 @@ int drm_sched_job_init(struct drm_sched_job *job,
 EXPORT_SYMBOL(drm_sched_job_init);
 
 /**
- * Return ture if we can push more jobs to the hw.
+ * drm_sched_ready - is the scheduler ready
+ *
+ * @sched: scheduler instance
+ *
+ * Return true if we can push more jobs to the hw, otherwise false.
  */
 static bool drm_sched_ready(struct drm_gpu_scheduler *sched)
 {
@@ -579,7 +774,10 @@ static bool drm_sched_ready(struct drm_gpu_scheduler *sched)
 }
 
 /**
- * Wake up the scheduler when it is ready
+ * drm_sched_wakeup - Wake up the scheduler when it is ready
+ *
+ * @sched: scheduler instance
+ *
  */
 static void drm_sched_wakeup(struct drm_gpu_scheduler *sched)
 {
@@ -588,8 +786,12 @@ static void drm_sched_wakeup(struct drm_gpu_scheduler *sched)
 }
 
 /**
- * Select next entity to process
-*/
+ * drm_sched_select_entity - Select next entity to process
+ *
+ * @sched: scheduler instance
+ *
+ * Returns the entity to process or NULL if none are found.
+ */
 static struct drm_sched_entity *
 drm_sched_select_entity(struct drm_gpu_scheduler *sched)
 {
@@ -609,6 +811,14 @@ drm_sched_select_entity(struct drm_gpu_scheduler *sched)
 	return entity;
 }
 
+/**
+ * drm_sched_process_job - process a job
+ *
+ * @f: fence
+ * @cb: fence callbacks
+ *
+ * Called after job has finished execution.
+ */
 static void drm_sched_process_job(struct dma_fence *f, struct dma_fence_cb *cb)
 {
 	struct drm_sched_fence *s_fence =
@@ -624,6 +834,13 @@ static void drm_sched_process_job(struct dma_fence *f, struct dma_fence_cb *cb)
 	wake_up_interruptible(&sched->wake_up_worker);
 }
 
+/**
+ * drm_sched_blocked - check if the scheduler is blocked
+ *
+ * @sched: scheduler instance
+ *
+ * Returns true if blocked, otherwise false.
+ */
 static bool drm_sched_blocked(struct drm_gpu_scheduler *sched)
 {
 	if (kthread_should_park()) {
@@ -634,6 +851,13 @@ static bool drm_sched_blocked(struct drm_gpu_scheduler *sched)
 	return false;
 }
 
+/**
+ * drm_sched_main - main scheduler thread
+ *
+ * @param: scheduler instance
+ *
+ * Returns 0.
+ */
 static int drm_sched_main(void *param)
 {
 	struct sched_param sparam = {.sched_priority = 1};
@@ -688,15 +912,17 @@ static int drm_sched_main(void *param)
 }
 
 /**
- * Init a gpu scheduler instance
+ * drm_sched_init - Init a gpu scheduler instance
  *
- * @sched		The pointer to the scheduler
- * @ops			The backend operations for this scheduler.
- * @hw_submissions	Number of hw submissions to do.
- * @name		Name used for debugging
+ * @sched: scheduler instance
+ * @ops: backend operations for this scheduler
+ * @hw_submission: number of hw submissions that can be in flight
+ * @hang_limit: number of times to allow a job to hang before dropping it
+ * @timeout: timeout value in jiffies for the scheduler
+ * @name: name used for debugging
  *
  * Return 0 on success, otherwise error code.
-*/
+ */
 int drm_sched_init(struct drm_gpu_scheduler *sched,
 		   const struct drm_sched_backend_ops *ops,
 		   unsigned hw_submission,
@@ -711,7 +937,7 @@ int drm_sched_init(struct drm_gpu_scheduler *sched,
 	sched->timeout = timeout;
 	sched->hang_limit = hang_limit;
 	for (i = DRM_SCHED_PRIORITY_MIN; i < DRM_SCHED_PRIORITY_MAX; i++)
-		drm_sched_rq_init(&sched->sched_rq[i]);
+		drm_sched_rq_init(sched, &sched->sched_rq[i]);
 
 	init_waitqueue_head(&sched->wake_up_worker);
 	init_waitqueue_head(&sched->job_scheduled);
@@ -732,9 +958,11 @@ int drm_sched_init(struct drm_gpu_scheduler *sched,
 EXPORT_SYMBOL(drm_sched_init);
 
 /**
- * Destroy a gpu scheduler
+ * drm_sched_fini - Destroy a gpu scheduler
+ *
+ * @sched: scheduler instance
  *
- * @sched	The pointer to the scheduler
+ * Tears down and cleans up the scheduler.
  */
 void drm_sched_fini(struct drm_gpu_scheduler *sched)
 {
diff --git a/drivers/gpu/drm/scheduler/gpu_scheduler_trace.h b/drivers/gpu/drm/scheduler/gpu_scheduler_trace.h
new file mode 100644
index 000000000000..4998ad950a48
--- /dev/null
+++ b/drivers/gpu/drm/scheduler/gpu_scheduler_trace.h
@@ -0,0 +1,82 @@
+/*
+ * Copyright 2017 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ */
+
+#if !defined(_GPU_SCHED_TRACE_H) || defined(TRACE_HEADER_MULTI_READ)
+#define _GPU_SCHED_TRACE_H_
+
+#include <linux/stringify.h>
+#include <linux/types.h>
+#include <linux/tracepoint.h>
+
+#include <drm/drmP.h>
+
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM gpu_scheduler
+#define TRACE_INCLUDE_FILE gpu_scheduler_trace
+
+TRACE_EVENT(drm_sched_job,
+	    TP_PROTO(struct drm_sched_job *sched_job, struct drm_sched_entity *entity),
+	    TP_ARGS(sched_job, entity),
+	    TP_STRUCT__entry(
+			     __field(struct drm_sched_entity *, entity)
+			     __field(struct dma_fence *, fence)
+			     __field(const char *, name)
+			     __field(uint64_t, id)
+			     __field(u32, job_count)
+			     __field(int, hw_job_count)
+			     ),
+
+	    TP_fast_assign(
+			   __entry->entity = entity;
+			   __entry->id = sched_job->id;
+			   __entry->fence = &sched_job->s_fence->finished;
+			   __entry->name = sched_job->sched->name;
+			   __entry->job_count = spsc_queue_count(&entity->job_queue);
+			   __entry->hw_job_count = atomic_read(
+				   &sched_job->sched->hw_rq_count);
+			   ),
+	    TP_printk("entity=%p, id=%llu, fence=%p, ring=%s, job count:%u, hw job count:%d",
+		      __entry->entity, __entry->id,
+		      __entry->fence, __entry->name,
+		      __entry->job_count, __entry->hw_job_count)
+);
+
+TRACE_EVENT(drm_sched_process_job,
+	    TP_PROTO(struct drm_sched_fence *fence),
+	    TP_ARGS(fence),
+	    TP_STRUCT__entry(
+		    __field(struct dma_fence *, fence)
+		    ),
+
+	    TP_fast_assign(
+		    __entry->fence = &fence->finished;
+		    ),
+	    TP_printk("fence=%p signaled", __entry->fence)
+);
+
+#endif
+
+/* This part must be outside protection */
+#undef TRACE_INCLUDE_PATH
+#define TRACE_INCLUDE_PATH ../../drivers/gpu/drm/scheduler
+#include <trace/define_trace.h>
diff --git a/drivers/gpu/drm/scheduler/sched_fence.c b/drivers/gpu/drm/scheduler/sched_fence.c
index 69aab086b913..45d9c3affbea 100644
--- a/drivers/gpu/drm/scheduler/sched_fence.c
+++ b/drivers/gpu/drm/scheduler/sched_fence.c
@@ -81,13 +81,8 @@ static const char *drm_sched_fence_get_timeline_name(struct dma_fence *f)
 	return (const char *)fence->sched->name;
 }
 
-static bool drm_sched_fence_enable_signaling(struct dma_fence *f)
-{
-	return true;
-}
-
 /**
- * amd_sched_fence_free - free up the fence memory
+ * drm_sched_fence_free - free up the fence memory
  *
  * @rcu: RCU callback head
  *
@@ -98,12 +93,11 @@ static void drm_sched_fence_free(struct rcu_head *rcu)
 	struct dma_fence *f = container_of(rcu, struct dma_fence, rcu);
 	struct drm_sched_fence *fence = to_drm_sched_fence(f);
 
-	dma_fence_put(fence->parent);
 	kmem_cache_free(sched_fence_slab, fence);
 }
 
 /**
- * amd_sched_fence_release_scheduled - callback that fence can be freed
+ * drm_sched_fence_release_scheduled - callback that fence can be freed
  *
  * @fence: fence
  *
@@ -114,11 +108,12 @@ static void drm_sched_fence_release_scheduled(struct dma_fence *f)
 {
 	struct drm_sched_fence *fence = to_drm_sched_fence(f);
 
+	dma_fence_put(fence->parent);
 	call_rcu(&fence->finished.rcu, drm_sched_fence_free);
 }
 
 /**
- * amd_sched_fence_release_finished - drop extra reference
+ * drm_sched_fence_release_finished - drop extra reference
  *
  * @f: fence
  *
@@ -134,18 +129,12 @@ static void drm_sched_fence_release_finished(struct dma_fence *f)
 const struct dma_fence_ops drm_sched_fence_ops_scheduled = {
 	.get_driver_name = drm_sched_fence_get_driver_name,
 	.get_timeline_name = drm_sched_fence_get_timeline_name,
-	.enable_signaling = drm_sched_fence_enable_signaling,
-	.signaled = NULL,
-	.wait = dma_fence_default_wait,
 	.release = drm_sched_fence_release_scheduled,
 };
 
 const struct dma_fence_ops drm_sched_fence_ops_finished = {
 	.get_driver_name = drm_sched_fence_get_driver_name,
 	.get_timeline_name = drm_sched_fence_get_timeline_name,
-	.enable_signaling = drm_sched_fence_enable_signaling,
-	.signaled = NULL,
-	.wait = dma_fence_default_wait,
 	.release = drm_sched_fence_release_finished,
 };
 
diff --git a/drivers/gpu/drm/selftests/Makefile b/drivers/gpu/drm/selftests/Makefile
index 4aebfc7f27d4..9fc349fa18e9 100644
--- a/drivers/gpu/drm/selftests/Makefile
+++ b/drivers/gpu/drm/selftests/Makefile
@@ -1 +1 @@
-obj-$(CONFIG_DRM_DEBUG_MM_SELFTEST) += test-drm_mm.o
+obj-$(CONFIG_DRM_DEBUG_SELFTEST) += test-drm_mm.o test-drm-helper.o
diff --git a/drivers/gpu/drm/selftests/drm_helper_selftests.h b/drivers/gpu/drm/selftests/drm_helper_selftests.h
new file mode 100644
index 000000000000..9771290ed228
--- /dev/null
+++ b/drivers/gpu/drm/selftests/drm_helper_selftests.h
@@ -0,0 +1,9 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* List each unit test as selftest(name, function)
+ *
+ * The name is used as both an enum and expanded as igt__name to create
+ * a module parameter. It must be unique and legal for a C identifier.
+ *
+ * Tests are executed in order by igt/drm_selftests_helper
+ */
+selftest(check_plane_state, igt_check_plane_state)
diff --git a/drivers/gpu/drm/selftests/drm_mm_selftests.h b/drivers/gpu/drm/selftests/drm_mm_selftests.h
index 54acc117550c..6b943ea1c57d 100644
--- a/drivers/gpu/drm/selftests/drm_mm_selftests.h
+++ b/drivers/gpu/drm/selftests/drm_mm_selftests.h
@@ -19,7 +19,9 @@ selftest(align64, igt_align64)
 selftest(evict, igt_evict)
 selftest(evict_range, igt_evict_range)
 selftest(bottomup, igt_bottomup)
+selftest(lowest, igt_lowest)
 selftest(topdown, igt_topdown)
+selftest(highest, igt_highest)
 selftest(color, igt_color)
 selftest(color_evict, igt_color_evict)
 selftest(color_evict_range, igt_color_evict_range)
diff --git a/drivers/gpu/drm/selftests/test-drm-helper.c b/drivers/gpu/drm/selftests/test-drm-helper.c
new file mode 100644
index 000000000000..a015712b43e8
--- /dev/null
+++ b/drivers/gpu/drm/selftests/test-drm-helper.c
@@ -0,0 +1,247 @@
+/*
+ * Test cases for the drm_kms_helper functions
+ */
+
+#define pr_fmt(fmt) "drm_kms_helper: " fmt
+
+#include <linux/module.h>
+
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_plane_helper.h>
+#include <drm/drm_modes.h>
+
+#define TESTS "drm_helper_selftests.h"
+#include "drm_selftest.h"
+
+#define FAIL(test, msg, ...) \
+	do { \
+		if (test) { \
+			pr_err("%s/%u: " msg, __FUNCTION__, __LINE__, ##__VA_ARGS__); \
+			return -EINVAL; \
+		} \
+	} while (0)
+
+#define FAIL_ON(x) FAIL((x), "%s", "FAIL_ON(" __stringify(x) ")\n")
+
+static void set_src(struct drm_plane_state *plane_state,
+		    unsigned src_x, unsigned src_y,
+		    unsigned src_w, unsigned src_h)
+{
+	plane_state->src_x = src_x;
+	plane_state->src_y = src_y;
+	plane_state->src_w = src_w;
+	plane_state->src_h = src_h;
+}
+
+static bool check_src_eq(struct drm_plane_state *plane_state,
+			 unsigned src_x, unsigned src_y,
+			 unsigned src_w, unsigned src_h)
+{
+	if (plane_state->src.x1 < 0) {
+		pr_err("src x coordinate %x should never be below 0.\n", plane_state->src.x1);
+		drm_rect_debug_print("src: ", &plane_state->src, true);
+		return false;
+	}
+	if (plane_state->src.y1 < 0) {
+		pr_err("src y coordinate %x should never be below 0.\n", plane_state->src.y1);
+		drm_rect_debug_print("src: ", &plane_state->src, true);
+		return false;
+	}
+
+	if (plane_state->src.x1 != src_x ||
+	    plane_state->src.y1 != src_y ||
+	    drm_rect_width(&plane_state->src) != src_w ||
+	    drm_rect_height(&plane_state->src) != src_h) {
+		drm_rect_debug_print("src: ", &plane_state->src, true);
+		return false;
+	}
+
+	return true;
+}
+
+static void set_crtc(struct drm_plane_state *plane_state,
+		     int crtc_x, int crtc_y,
+		     unsigned crtc_w, unsigned crtc_h)
+{
+	plane_state->crtc_x = crtc_x;
+	plane_state->crtc_y = crtc_y;
+	plane_state->crtc_w = crtc_w;
+	plane_state->crtc_h = crtc_h;
+}
+
+static bool check_crtc_eq(struct drm_plane_state *plane_state,
+			  int crtc_x, int crtc_y,
+			  unsigned crtc_w, unsigned crtc_h)
+{
+	if (plane_state->dst.x1 != crtc_x ||
+	    plane_state->dst.y1 != crtc_y ||
+	    drm_rect_width(&plane_state->dst) != crtc_w ||
+	    drm_rect_height(&plane_state->dst) != crtc_h) {
+		drm_rect_debug_print("dst: ", &plane_state->dst, false);
+
+		return false;
+	}
+
+	return true;
+}
+
+static int igt_check_plane_state(void *ignored)
+{
+	int ret;
+
+	const struct drm_crtc_state crtc_state = {
+		.crtc = ZERO_SIZE_PTR,
+		.enable = true,
+		.active = true,
+		.mode = {
+			DRM_MODE("1024x768", 0, 65000, 1024, 1048,
+				1184, 1344, 0, 768, 771, 777, 806, 0,
+				DRM_MODE_FLAG_NHSYNC | DRM_MODE_FLAG_NVSYNC)
+		},
+	};
+	struct drm_framebuffer fb = {
+		.width = 2048,
+		.height = 2048
+	};
+	struct drm_plane_state plane_state = {
+		.crtc = ZERO_SIZE_PTR,
+		.fb = &fb,
+		.rotation = DRM_MODE_ROTATE_0
+	};
+
+	/* Simple clipping, no scaling. */
+	set_src(&plane_state, 0, 0, fb.width << 16, fb.height << 16);
+	set_crtc(&plane_state, 0, 0, fb.width, fb.height);
+	ret = drm_atomic_helper_check_plane_state(&plane_state, &crtc_state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  false, false);
+	FAIL(ret < 0, "Simple clipping check should pass\n");
+	FAIL_ON(!plane_state.visible);
+	FAIL_ON(!check_src_eq(&plane_state, 0, 0, 1024 << 16, 768 << 16));
+	FAIL_ON(!check_crtc_eq(&plane_state, 0, 0, 1024, 768));
+
+	/* Rotated clipping + reflection, no scaling. */
+	plane_state.rotation = DRM_MODE_ROTATE_90 | DRM_MODE_REFLECT_X;
+	ret = drm_atomic_helper_check_plane_state(&plane_state, &crtc_state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  false, false);
+	FAIL(ret < 0, "Rotated clipping check should pass\n");
+	FAIL_ON(!plane_state.visible);
+	FAIL_ON(!check_src_eq(&plane_state, 0, 0, 768 << 16, 1024 << 16));
+	FAIL_ON(!check_crtc_eq(&plane_state, 0, 0, 1024, 768));
+	plane_state.rotation = DRM_MODE_ROTATE_0;
+
+	/* Check whether positioning works correctly. */
+	set_src(&plane_state, 0, 0, 1023 << 16, 767 << 16);
+	set_crtc(&plane_state, 0, 0, 1023, 767);
+	ret = drm_atomic_helper_check_plane_state(&plane_state, &crtc_state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  false, false);
+	FAIL(!ret, "Should not be able to position on the crtc with can_position=false\n");
+
+	ret = drm_atomic_helper_check_plane_state(&plane_state, &crtc_state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  true, false);
+	FAIL(ret < 0, "Simple positioning should work\n");
+	FAIL_ON(!plane_state.visible);
+	FAIL_ON(!check_src_eq(&plane_state, 0, 0, 1023 << 16, 767 << 16));
+	FAIL_ON(!check_crtc_eq(&plane_state, 0, 0, 1023, 767));
+
+	/* Simple scaling tests. */
+	set_src(&plane_state, 0, 0, 512 << 16, 384 << 16);
+	set_crtc(&plane_state, 0, 0, 1024, 768);
+	ret = drm_atomic_helper_check_plane_state(&plane_state, &crtc_state,
+						  0x8001,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  false, false);
+	FAIL(!ret, "Upscaling out of range should fail.\n");
+	ret = drm_atomic_helper_check_plane_state(&plane_state, &crtc_state,
+						  0x8000,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  false, false);
+	FAIL(ret < 0, "Upscaling exactly 2x should work\n");
+	FAIL_ON(!plane_state.visible);
+	FAIL_ON(!check_src_eq(&plane_state, 0, 0, 512 << 16, 384 << 16));
+	FAIL_ON(!check_crtc_eq(&plane_state, 0, 0, 1024, 768));
+
+	set_src(&plane_state, 0, 0, 2048 << 16, 1536 << 16);
+	ret = drm_atomic_helper_check_plane_state(&plane_state, &crtc_state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  0x1ffff, false, false);
+	FAIL(!ret, "Downscaling out of range should fail.\n");
+	ret = drm_atomic_helper_check_plane_state(&plane_state, &crtc_state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  0x20000, false, false);
+	FAIL(ret < 0, "Should succeed with exact scaling limit\n");
+	FAIL_ON(!plane_state.visible);
+	FAIL_ON(!check_src_eq(&plane_state, 0, 0, 2048 << 16, 1536 << 16));
+	FAIL_ON(!check_crtc_eq(&plane_state, 0, 0, 1024, 768));
+
+	/* Testing rounding errors. */
+	set_src(&plane_state, 0, 0, 0x40001, 0x40001);
+	set_crtc(&plane_state, 1022, 766, 4, 4);
+	ret = drm_atomic_helper_check_plane_state(&plane_state, &crtc_state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  0x10001,
+						  true, false);
+	FAIL(ret < 0, "Should succeed by clipping to exact multiple");
+	FAIL_ON(!plane_state.visible);
+	FAIL_ON(!check_src_eq(&plane_state, 0, 0, 2 << 16, 2 << 16));
+	FAIL_ON(!check_crtc_eq(&plane_state, 1022, 766, 2, 2));
+
+	set_src(&plane_state, 0x20001, 0x20001, 0x4040001, 0x3040001);
+	set_crtc(&plane_state, -2, -2, 1028, 772);
+	ret = drm_atomic_helper_check_plane_state(&plane_state, &crtc_state,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  0x10001,
+						  false, false);
+	FAIL(ret < 0, "Should succeed by clipping to exact multiple");
+	FAIL_ON(!plane_state.visible);
+	FAIL_ON(!check_src_eq(&plane_state, 0x40002, 0x40002, 1024 << 16, 768 << 16));
+	FAIL_ON(!check_crtc_eq(&plane_state, 0, 0, 1024, 768));
+
+	set_src(&plane_state, 0, 0, 0x3ffff, 0x3ffff);
+	set_crtc(&plane_state, 1022, 766, 4, 4);
+	ret = drm_atomic_helper_check_plane_state(&plane_state, &crtc_state,
+						  0xffff,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  true, false);
+	FAIL(ret < 0, "Should succeed by clipping to exact multiple");
+	FAIL_ON(!plane_state.visible);
+	/* Should not be rounded to 0x20001, which would be upscaling. */
+	FAIL_ON(!check_src_eq(&plane_state, 0, 0, 2 << 16, 2 << 16));
+	FAIL_ON(!check_crtc_eq(&plane_state, 1022, 766, 2, 2));
+
+	set_src(&plane_state, 0x1ffff, 0x1ffff, 0x403ffff, 0x303ffff);
+	set_crtc(&plane_state, -2, -2, 1028, 772);
+	ret = drm_atomic_helper_check_plane_state(&plane_state, &crtc_state,
+						  0xffff,
+						  DRM_PLANE_HELPER_NO_SCALING,
+						  false, false);
+	FAIL(ret < 0, "Should succeed by clipping to exact multiple");
+	FAIL_ON(!plane_state.visible);
+	FAIL_ON(!check_src_eq(&plane_state, 0x3fffe, 0x3fffe, 1024 << 16, 768 << 16));
+	FAIL_ON(!check_crtc_eq(&plane_state, 0, 0, 1024, 768));
+
+	return 0;
+}
+
+#include "drm_selftest.c"
+
+static int __init test_drm_helper_init(void)
+{
+	int err;
+
+	err = run_selftests(selftests, ARRAY_SIZE(selftests), NULL);
+
+	return err > 0 ? 0 : err;
+}
+
+module_init(test_drm_helper_init);
+
+MODULE_AUTHOR("Intel Corporation");
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/selftests/test-drm_mm.c b/drivers/gpu/drm/selftests/test-drm_mm.c
index 7cc935d7b7aa..fbed2c90fd51 100644
--- a/drivers/gpu/drm/selftests/test-drm_mm.c
+++ b/drivers/gpu/drm/selftests/test-drm_mm.c
@@ -389,7 +389,7 @@ static int __igt_reserve(unsigned int count, u64 size)
 	if (!order)
 		goto err;
 
-	nodes = vzalloc(sizeof(*nodes) * count);
+	nodes = vzalloc(array_size(count, sizeof(*nodes)));
 	if (!nodes)
 		goto err_order;
 
@@ -579,7 +579,7 @@ static int __igt_insert(unsigned int count, u64 size, bool replace)
 	DRM_MM_BUG_ON(!size);
 
 	ret = -ENOMEM;
-	nodes = vmalloc(count * sizeof(*nodes));
+	nodes = vmalloc(array_size(count, sizeof(*nodes)));
 	if (!nodes)
 		goto err;
 
@@ -889,7 +889,7 @@ static int __igt_insert_range(unsigned int count, u64 size, u64 start, u64 end)
 	 */
 
 	ret = -ENOMEM;
-	nodes = vzalloc(count * sizeof(*nodes));
+	nodes = vzalloc(array_size(count, sizeof(*nodes)));
 	if (!nodes)
 		goto err;
 
@@ -1046,7 +1046,7 @@ static int igt_align(void *ignored)
 	 * meets our requirements.
 	 */
 
-	nodes = vzalloc(max_count * sizeof(*nodes));
+	nodes = vzalloc(array_size(max_count, sizeof(*nodes)));
 	if (!nodes)
 		goto err;
 
@@ -1416,7 +1416,7 @@ static int igt_evict(void *ignored)
 	 */
 
 	ret = -ENOMEM;
-	nodes = vzalloc(size * sizeof(*nodes));
+	nodes = vzalloc(array_size(size, sizeof(*nodes)));
 	if (!nodes)
 		goto err;
 
@@ -1526,7 +1526,7 @@ static int igt_evict_range(void *ignored)
 	 */
 
 	ret = -ENOMEM;
-	nodes = vzalloc(size * sizeof(*nodes));
+	nodes = vzalloc(array_size(size, sizeof(*nodes)));
 	if (!nodes)
 		goto err;
 
@@ -1627,11 +1627,11 @@ static int igt_topdown(void *ignored)
 	 */
 
 	ret = -ENOMEM;
-	nodes = vzalloc(count * sizeof(*nodes));
+	nodes = vzalloc(array_size(count, sizeof(*nodes)));
 	if (!nodes)
 		goto err;
 
-	bitmap = kzalloc(count / BITS_PER_LONG * sizeof(unsigned long),
+	bitmap = kcalloc(count / BITS_PER_LONG, sizeof(unsigned long),
 			 GFP_KERNEL);
 	if (!bitmap)
 		goto err_nodes;
@@ -1741,11 +1741,11 @@ static int igt_bottomup(void *ignored)
 	 */
 
 	ret = -ENOMEM;
-	nodes = vzalloc(count * sizeof(*nodes));
+	nodes = vzalloc(array_size(count, sizeof(*nodes)));
 	if (!nodes)
 		goto err;
 
-	bitmap = kzalloc(count / BITS_PER_LONG * sizeof(unsigned long),
+	bitmap = kcalloc(count / BITS_PER_LONG, sizeof(unsigned long),
 			 GFP_KERNEL);
 	if (!bitmap)
 		goto err_nodes;
@@ -1825,6 +1825,77 @@ err:
 	return ret;
 }
 
+static int __igt_once(unsigned int mode)
+{
+	struct drm_mm mm;
+	struct drm_mm_node rsvd_lo, rsvd_hi, node;
+	int err;
+
+	drm_mm_init(&mm, 0, 7);
+
+	memset(&rsvd_lo, 0, sizeof(rsvd_lo));
+	rsvd_lo.start = 1;
+	rsvd_lo.size = 1;
+	err = drm_mm_reserve_node(&mm, &rsvd_lo);
+	if (err) {
+		pr_err("Could not reserve low node\n");
+		goto err;
+	}
+
+	memset(&rsvd_hi, 0, sizeof(rsvd_hi));
+	rsvd_hi.start = 5;
+	rsvd_hi.size = 1;
+	err = drm_mm_reserve_node(&mm, &rsvd_hi);
+	if (err) {
+		pr_err("Could not reserve low node\n");
+		goto err_lo;
+	}
+
+	if (!drm_mm_hole_follows(&rsvd_lo) || !drm_mm_hole_follows(&rsvd_hi)) {
+		pr_err("Expected a hole after lo and high nodes!\n");
+		err = -EINVAL;
+		goto err_hi;
+	}
+
+	memset(&node, 0, sizeof(node));
+	err = drm_mm_insert_node_generic(&mm, &node,
+					 2, 0, 0,
+					 mode | DRM_MM_INSERT_ONCE);
+	if (!err) {
+		pr_err("Unexpectedly inserted the node into the wrong hole: node.start=%llx\n",
+		       node.start);
+		err = -EINVAL;
+		goto err_node;
+	}
+
+	err = drm_mm_insert_node_generic(&mm, &node, 2, 0, 0, mode);
+	if (err) {
+		pr_err("Could not insert the node into the available hole!\n");
+		err = -EINVAL;
+		goto err_hi;
+	}
+
+err_node:
+	drm_mm_remove_node(&node);
+err_hi:
+	drm_mm_remove_node(&rsvd_hi);
+err_lo:
+	drm_mm_remove_node(&rsvd_lo);
+err:
+	drm_mm_takedown(&mm);
+	return err;
+}
+
+static int igt_lowest(void *ignored)
+{
+	return __igt_once(DRM_MM_INSERT_LOW);
+}
+
+static int igt_highest(void *ignored)
+{
+	return __igt_once(DRM_MM_INSERT_HIGH);
+}
+
 static void separate_adjacent_colors(const struct drm_mm_node *node,
 				     unsigned long color,
 				     u64 *start,
@@ -2098,7 +2169,7 @@ static int igt_color_evict(void *ignored)
 	 */
 
 	ret = -ENOMEM;
-	nodes = vzalloc(total_size * sizeof(*nodes));
+	nodes = vzalloc(array_size(total_size, sizeof(*nodes)));
 	if (!nodes)
 		goto err;
 
@@ -2199,7 +2270,7 @@ static int igt_color_evict_range(void *ignored)
 	 */
 
 	ret = -ENOMEM;
-	nodes = vzalloc(total_size * sizeof(*nodes));
+	nodes = vzalloc(array_size(total_size, sizeof(*nodes)));
 	if (!nodes)
 		goto err;
 
diff --git a/drivers/gpu/drm/shmobile/Kconfig b/drivers/gpu/drm/shmobile/Kconfig
index c987c826daa3..0426d66660d1 100644
--- a/drivers/gpu/drm/shmobile/Kconfig
+++ b/drivers/gpu/drm/shmobile/Kconfig
@@ -2,7 +2,6 @@ config DRM_SHMOBILE
 	tristate "DRM Support for SH Mobile"
 	depends on DRM && ARM
 	depends on ARCH_SHMOBILE || COMPILE_TEST
-	depends on FB_SH_MOBILE_MERAM || !FB_SH_MOBILE_MERAM
 	select BACKLIGHT_CLASS_DEVICE
 	select BACKLIGHT_LCD_SUPPORT
 	select DRM_KMS_HELPER
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
index e7738939a86d..fc66167b0641 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_crtc.c
@@ -21,8 +21,6 @@
 #include <drm/drm_gem_cma_helper.h>
 #include <drm/drm_plane_helper.h>
 
-#include <video/sh_mobile_meram.h>
-
 #include "shmob_drm_backlight.h"
 #include "shmob_drm_crtc.h"
 #include "shmob_drm_drv.h"
@@ -47,20 +45,12 @@ static int shmob_drm_clk_on(struct shmob_drm_device *sdev)
 		if (ret < 0)
 			return ret;
 	}
-#if 0
-	if (sdev->meram_dev && sdev->meram_dev->pdev)
-		pm_runtime_get_sync(&sdev->meram_dev->pdev->dev);
-#endif
 
 	return 0;
 }
 
 static void shmob_drm_clk_off(struct shmob_drm_device *sdev)
 {
-#if 0
-	if (sdev->meram_dev && sdev->meram_dev->pdev)
-		pm_runtime_put_sync(&sdev->meram_dev->pdev->dev);
-#endif
 	if (sdev->clock)
 		clk_disable_unprepare(sdev->clock);
 }
@@ -269,12 +259,6 @@ static void shmob_drm_crtc_stop(struct shmob_drm_crtc *scrtc)
 	if (!scrtc->started)
 		return;
 
-	/* Disable the MERAM cache. */
-	if (scrtc->cache) {
-		sh_mobile_meram_cache_free(sdev->meram, scrtc->cache);
-		scrtc->cache = NULL;
-	}
-
 	/* Stop the LCDC. */
 	shmob_drm_crtc_start_stop(scrtc, false);
 
@@ -305,7 +289,6 @@ static void shmob_drm_crtc_compute_base(struct shmob_drm_crtc *scrtc,
 {
 	struct drm_crtc *crtc = &scrtc->crtc;
 	struct drm_framebuffer *fb = crtc->primary->fb;
-	struct shmob_drm_device *sdev = crtc->dev->dev_private;
 	struct drm_gem_cma_object *gem;
 	unsigned int bpp;
 
@@ -321,11 +304,6 @@ static void shmob_drm_crtc_compute_base(struct shmob_drm_crtc *scrtc,
 			      + y / (bpp == 4 ? 2 : 1) * fb->pitches[1]
 			      + x * (bpp == 16 ? 2 : 1);
 	}
-
-	if (scrtc->cache)
-		sh_mobile_meram_cache_update(sdev->meram, scrtc->cache,
-					     scrtc->dma[0], scrtc->dma[1],
-					     &scrtc->dma[0], &scrtc->dma[1]);
 }
 
 static void shmob_drm_crtc_update_base(struct shmob_drm_crtc *scrtc)
@@ -372,9 +350,7 @@ static int shmob_drm_crtc_mode_set(struct drm_crtc *crtc,
 {
 	struct shmob_drm_crtc *scrtc = to_shmob_crtc(crtc);
 	struct shmob_drm_device *sdev = crtc->dev->dev_private;
-	const struct sh_mobile_meram_cfg *mdata = sdev->pdata->meram;
 	const struct shmob_drm_format_info *format;
-	void *cache;
 
 	format = shmob_drm_format_info(crtc->primary->fb->format->format);
 	if (format == NULL) {
@@ -386,24 +362,6 @@ static int shmob_drm_crtc_mode_set(struct drm_crtc *crtc,
 	scrtc->format = format;
 	scrtc->line_size = crtc->primary->fb->pitches[0];
 
-	if (sdev->meram) {
-		/* Enable MERAM cache if configured. We need to de-init
-		 * configured ICBs before we can re-initialize them.
-		 */
-		if (scrtc->cache) {
-			sh_mobile_meram_cache_free(sdev->meram, scrtc->cache);
-			scrtc->cache = NULL;
-		}
-
-		cache = sh_mobile_meram_cache_alloc(sdev->meram, mdata,
-						    crtc->primary->fb->pitches[0],
-						    adjusted_mode->vdisplay,
-						    format->meram,
-						    &scrtc->line_size);
-		if (!IS_ERR(cache))
-			scrtc->cache = cache;
-	}
-
 	shmob_drm_crtc_compute_base(scrtc, x, y);
 
 	return 0;
@@ -717,7 +675,7 @@ int shmob_drm_connector_create(struct shmob_drm_device *sdev,
 	if (ret < 0)
 		goto err_cleanup;
 
-	ret = drm_mode_connector_attach_encoder(connector, encoder);
+	ret = drm_connector_attach_encoder(connector, encoder);
 	if (ret < 0)
 		goto err_backlight;
 
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_crtc.h b/drivers/gpu/drm/shmobile/shmob_drm_crtc.h
index f152973df11c..c11f421737dc 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_crtc.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_crtc.h
@@ -28,7 +28,6 @@ struct shmob_drm_crtc {
 	int dpms;
 
 	const struct shmob_drm_format_info *format;
-	void *cache;
 	unsigned long dma[2];
 	unsigned int line_size;
 	bool started;
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_drv.h b/drivers/gpu/drm/shmobile/shmob_drm_drv.h
index 02ea315ba69a..088a6e55fa29 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_drv.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_drv.h
@@ -23,7 +23,6 @@
 struct clk;
 struct device;
 struct drm_device;
-struct sh_mobile_meram_info;
 
 struct shmob_drm_device {
 	struct device *dev;
@@ -31,7 +30,6 @@ struct shmob_drm_device {
 
 	void __iomem *mmio;
 	struct clk *clock;
-	struct sh_mobile_meram_info *meram;
 	u32 lddckr;
 	u32 ldmt1r;
 
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_kms.c b/drivers/gpu/drm/shmobile/shmob_drm_kms.c
index d36919b14da7..447638581c08 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_kms.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_kms.c
@@ -18,8 +18,6 @@
 #include <drm/drm_gem_cma_helper.h>
 #include <drm/drm_gem_framebuffer_helper.h>
 
-#include <video/sh_mobile_meram.h>
-
 #include "shmob_drm_crtc.h"
 #include "shmob_drm_drv.h"
 #include "shmob_drm_kms.h"
@@ -35,55 +33,46 @@ static const struct shmob_drm_format_info shmob_drm_format_infos[] = {
 		.bpp = 16,
 		.yuv = false,
 		.lddfr = LDDFR_PKF_RGB16,
-		.meram = SH_MOBILE_MERAM_PF_RGB,
 	}, {
 		.fourcc = DRM_FORMAT_RGB888,
 		.bpp = 24,
 		.yuv = false,
 		.lddfr = LDDFR_PKF_RGB24,
-		.meram = SH_MOBILE_MERAM_PF_RGB,
 	}, {
 		.fourcc = DRM_FORMAT_ARGB8888,
 		.bpp = 32,
 		.yuv = false,
 		.lddfr = LDDFR_PKF_ARGB32,
-		.meram = SH_MOBILE_MERAM_PF_RGB,
 	}, {
 		.fourcc = DRM_FORMAT_NV12,
 		.bpp = 12,
 		.yuv = true,
 		.lddfr = LDDFR_CC | LDDFR_YF_420,
-		.meram = SH_MOBILE_MERAM_PF_NV,
 	}, {
 		.fourcc = DRM_FORMAT_NV21,
 		.bpp = 12,
 		.yuv = true,
 		.lddfr = LDDFR_CC | LDDFR_YF_420,
-		.meram = SH_MOBILE_MERAM_PF_NV,
 	}, {
 		.fourcc = DRM_FORMAT_NV16,
 		.bpp = 16,
 		.yuv = true,
 		.lddfr = LDDFR_CC | LDDFR_YF_422,
-		.meram = SH_MOBILE_MERAM_PF_NV,
 	}, {
 		.fourcc = DRM_FORMAT_NV61,
 		.bpp = 16,
 		.yuv = true,
 		.lddfr = LDDFR_CC | LDDFR_YF_422,
-		.meram = SH_MOBILE_MERAM_PF_NV,
 	}, {
 		.fourcc = DRM_FORMAT_NV24,
 		.bpp = 24,
 		.yuv = true,
 		.lddfr = LDDFR_CC | LDDFR_YF_444,
-		.meram = SH_MOBILE_MERAM_PF_NV24,
 	}, {
 		.fourcc = DRM_FORMAT_NV42,
 		.bpp = 24,
 		.yuv = true,
 		.lddfr = LDDFR_CC | LDDFR_YF_444,
-		.meram = SH_MOBILE_MERAM_PF_NV24,
 	},
 };
 
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_kms.h b/drivers/gpu/drm/shmobile/shmob_drm_kms.h
index 06d5b7caa026..753e2817dc2c 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_kms.h
+++ b/drivers/gpu/drm/shmobile/shmob_drm_kms.h
@@ -24,7 +24,6 @@ struct shmob_drm_format_info {
 	unsigned int bpp;
 	bool yuv;
 	u32 lddfr;
-	unsigned int meram;
 };
 
 const struct shmob_drm_format_info *shmob_drm_format_info(u32 fourcc);
diff --git a/drivers/gpu/drm/shmobile/shmob_drm_plane.c b/drivers/gpu/drm/shmobile/shmob_drm_plane.c
index 97f6e4a3eb0d..1d0359f713ca 100644
--- a/drivers/gpu/drm/shmobile/shmob_drm_plane.c
+++ b/drivers/gpu/drm/shmobile/shmob_drm_plane.c
@@ -17,8 +17,6 @@
 #include <drm/drm_fb_cma_helper.h>
 #include <drm/drm_gem_cma_helper.h>
 
-#include <video/sh_mobile_meram.h>
-
 #include "shmob_drm_drv.h"
 #include "shmob_drm_kms.h"
 #include "shmob_drm_plane.h"
diff --git a/drivers/gpu/drm/sti/sti_crtc.c b/drivers/gpu/drm/sti/sti_crtc.c
index 21e50d7b1f86..5824e6aca8f4 100644
--- a/drivers/gpu/drm/sti/sti_crtc.c
+++ b/drivers/gpu/drm/sti/sti_crtc.c
@@ -357,7 +357,7 @@ int sti_crtc_init(struct drm_device *drm_dev, struct sti_mixer *mixer,
 	res = drm_crtc_init_with_planes(drm_dev, crtc, primary, cursor,
 					&sti_crtc_funcs, NULL);
 	if (res) {
-		DRM_ERROR("Can't initialze CRTC\n");
+		DRM_ERROR("Can't initialize CRTC\n");
 		return -EINVAL;
 	}
 
diff --git a/drivers/gpu/drm/sti/sti_cursor.c b/drivers/gpu/drm/sti/sti_cursor.c
index df0a282b9615..57b870e1e696 100644
--- a/drivers/gpu/drm/sti/sti_cursor.c
+++ b/drivers/gpu/drm/sti/sti_cursor.c
@@ -332,7 +332,7 @@ static void sti_cursor_destroy(struct drm_plane *drm_plane)
 {
 	DRM_DEBUG_DRIVER("\n");
 
-	drm_plane_helper_disable(drm_plane);
+	drm_plane_helper_disable(drm_plane, NULL);
 	drm_plane_cleanup(drm_plane);
 }
 
diff --git a/drivers/gpu/drm/sti/sti_drv.c b/drivers/gpu/drm/sti/sti_drv.c
index 90c46b49c931..832fc43960ee 100644
--- a/drivers/gpu/drm/sti/sti_drv.c
+++ b/drivers/gpu/drm/sti/sti_drv.c
@@ -224,7 +224,7 @@ static int sti_bind(struct device *dev)
 
 	ret = sti_init(ddev);
 	if (ret)
-		goto err_drm_dev_unref;
+		goto err_drm_dev_put;
 
 	ret = component_bind_all(ddev->dev, ddev);
 	if (ret)
@@ -248,8 +248,8 @@ err_register:
 	drm_mode_config_cleanup(ddev);
 err_cleanup:
 	sti_cleanup(ddev);
-err_drm_dev_unref:
-	drm_dev_unref(ddev);
+err_drm_dev_put:
+	drm_dev_put(ddev);
 	return ret;
 }
 
@@ -259,7 +259,7 @@ static void sti_unbind(struct device *dev)
 
 	drm_dev_unregister(ddev);
 	sti_cleanup(ddev);
-	drm_dev_unref(ddev);
+	drm_dev_put(ddev);
 }
 
 static const struct component_master_ops sti_ops = {
diff --git a/drivers/gpu/drm/sti/sti_dvo.c b/drivers/gpu/drm/sti/sti_dvo.c
index a5979cd25cc7..b08376b7611b 100644
--- a/drivers/gpu/drm/sti/sti_dvo.c
+++ b/drivers/gpu/drm/sti/sti_dvo.c
@@ -387,7 +387,9 @@ sti_dvo_connector_detect(struct drm_connector *connector, bool force)
 
 	if (!dvo->panel) {
 		dvo->panel = of_drm_find_panel(dvo->panel_node);
-		if (dvo->panel)
+		if (IS_ERR(dvo->panel))
+			dvo->panel = NULL;
+		else
 			drm_panel_attach(dvo->panel, connector);
 	}
 
@@ -484,7 +486,7 @@ static int sti_dvo_bind(struct device *dev, struct device *master, void *data)
 	drm_connector_helper_add(drm_connector,
 				 &sti_dvo_connector_helper_funcs);
 
-	err = drm_mode_connector_attach_encoder(drm_connector, encoder);
+	err = drm_connector_attach_encoder(drm_connector, encoder);
 	if (err) {
 		DRM_ERROR("Failed to attach a connector to a encoder\n");
 		goto err_sysfs;
diff --git a/drivers/gpu/drm/sti/sti_gdp.c b/drivers/gpu/drm/sti/sti_gdp.c
index 9b2c47051b51..c32de6cbf061 100644
--- a/drivers/gpu/drm/sti/sti_gdp.c
+++ b/drivers/gpu/drm/sti/sti_gdp.c
@@ -211,7 +211,11 @@ static int gdp_dbg_show(struct seq_file *s, void *data)
 	struct drm_info_node *node = s->private;
 	struct sti_gdp *gdp = (struct sti_gdp *)node->info_ent->data;
 	struct drm_plane *drm_plane = &gdp->plane.drm_plane;
-	struct drm_crtc *crtc = drm_plane->crtc;
+	struct drm_crtc *crtc;
+
+	drm_modeset_lock(&drm_plane->mutex, NULL);
+	crtc = drm_plane->state->crtc;
+	drm_modeset_unlock(&drm_plane->mutex);
 
 	seq_printf(s, "%s: (vaddr = 0x%p)",
 		   sti_plane_to_str(&gdp->plane), gdp->regs);
@@ -879,7 +883,7 @@ static void sti_gdp_destroy(struct drm_plane *drm_plane)
 {
 	DRM_DEBUG_DRIVER("\n");
 
-	drm_plane_helper_disable(drm_plane);
+	drm_plane_helper_disable(drm_plane, NULL);
 	drm_plane_cleanup(drm_plane);
 }
 
diff --git a/drivers/gpu/drm/sti/sti_hda.c b/drivers/gpu/drm/sti/sti_hda.c
index 67bbdb49fffc..49438337f70d 100644
--- a/drivers/gpu/drm/sti/sti_hda.c
+++ b/drivers/gpu/drm/sti/sti_hda.c
@@ -709,7 +709,7 @@ static int sti_hda_bind(struct device *dev, struct device *master, void *data)
 	drm_connector_helper_add(drm_connector,
 			&sti_hda_connector_helper_funcs);
 
-	err = drm_mode_connector_attach_encoder(drm_connector, encoder);
+	err = drm_connector_attach_encoder(drm_connector, encoder);
 	if (err) {
 		DRM_ERROR("Failed to attach a connector to a encoder\n");
 		goto err_sysfs;
diff --git a/drivers/gpu/drm/sti/sti_hdmi.c b/drivers/gpu/drm/sti/sti_hdmi.c
index 58f431102512..34cdc4644435 100644
--- a/drivers/gpu/drm/sti/sti_hdmi.c
+++ b/drivers/gpu/drm/sti/sti_hdmi.c
@@ -977,7 +977,7 @@ static int sti_hdmi_connector_get_modes(struct drm_connector *connector)
 	cec_notifier_set_phys_addr_from_edid(hdmi->notifier, edid);
 
 	count = drm_add_edid_modes(connector, edid);
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 
 	kfree(edid);
 	return count;
@@ -1290,7 +1290,7 @@ static int sti_hdmi_bind(struct device *dev, struct device *master, void *data)
 
 	hdmi->drm_connector = drm_connector;
 
-	err = drm_mode_connector_attach_encoder(drm_connector, encoder);
+	err = drm_connector_attach_encoder(drm_connector, encoder);
 	if (err) {
 		DRM_ERROR("Failed to attach a connector to a encoder\n");
 		goto err_sysfs;
diff --git a/drivers/gpu/drm/sti/sti_hqvdp.c b/drivers/gpu/drm/sti/sti_hqvdp.c
index 106be8c4e58b..03ac3b4a4469 100644
--- a/drivers/gpu/drm/sti/sti_hqvdp.c
+++ b/drivers/gpu/drm/sti/sti_hqvdp.c
@@ -1260,7 +1260,7 @@ static void sti_hqvdp_destroy(struct drm_plane *drm_plane)
 {
 	DRM_DEBUG_DRIVER("\n");
 
-	drm_plane_helper_disable(drm_plane);
+	drm_plane_helper_disable(drm_plane, NULL);
 	drm_plane_cleanup(drm_plane);
 }
 
diff --git a/drivers/gpu/drm/stm/drv.c b/drivers/gpu/drm/stm/drv.c
index 8698e08313e1..f2021b23554d 100644
--- a/drivers/gpu/drm/stm/drv.c
+++ b/drivers/gpu/drm/stm/drv.c
@@ -148,16 +148,16 @@ static int stm_drm_platform_probe(struct platform_device *pdev)
 
 	ret = drv_load(ddev);
 	if (ret)
-		goto err_unref;
+		goto err_put;
 
 	ret = drm_dev_register(ddev, 0);
 	if (ret)
-		goto err_unref;
+		goto err_put;
 
 	return 0;
 
-err_unref:
-	drm_dev_unref(ddev);
+err_put:
+	drm_dev_put(ddev);
 
 	return ret;
 }
@@ -170,7 +170,7 @@ static int stm_drm_platform_remove(struct platform_device *pdev)
 
 	drm_dev_unregister(ddev);
 	drv_unload(ddev);
-	drm_dev_unref(ddev);
+	drm_dev_put(ddev);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/stm/ltdc.c b/drivers/gpu/drm/stm/ltdc.c
index e3121d9e4230..808d9fb627e9 100644
--- a/drivers/gpu/drm/stm/ltdc.c
+++ b/drivers/gpu/drm/stm/ltdc.c
@@ -445,6 +445,47 @@ static void ltdc_crtc_atomic_disable(struct drm_crtc *crtc,
 	reg_set(ldev->regs, LTDC_SRCR, SRCR_IMR);
 }
 
+#define CLK_TOLERANCE_HZ 50
+
+static enum drm_mode_status
+ltdc_crtc_mode_valid(struct drm_crtc *crtc,
+		     const struct drm_display_mode *mode)
+{
+	struct ltdc_device *ldev = crtc_to_ltdc(crtc);
+	int target = mode->clock * 1000;
+	int target_min = target - CLK_TOLERANCE_HZ;
+	int target_max = target + CLK_TOLERANCE_HZ;
+	int result;
+
+	result = clk_round_rate(ldev->pixel_clk, target);
+
+	DRM_DEBUG_DRIVER("clk rate target %d, available %d\n", target, result);
+
+	/* Filter modes according to the max frequency supported by the pads */
+	if (result > ldev->caps.pad_max_freq_hz)
+		return MODE_CLOCK_HIGH;
+
+	/*
+	 * Accept all "preferred" modes:
+	 * - this is important for panels because panel clock tolerances are
+	 *   bigger than hdmi ones and there is no reason to not accept them
+	 *   (the fps may vary a little but it is not a problem).
+	 * - the hdmi preferred mode will be accepted too, but userland will
+	 *   be able to use others hdmi "valid" modes if necessary.
+	 */
+	if (mode->type & DRM_MODE_TYPE_PREFERRED)
+		return MODE_OK;
+
+	/*
+	 * Filter modes according to the clock value, particularly useful for
+	 * hdmi modes that require precise pixel clocks.
+	 */
+	if (result < target_min || result > target_max)
+		return MODE_CLOCK_RANGE;
+
+	return MODE_OK;
+}
+
 static bool ltdc_crtc_mode_fixup(struct drm_crtc *crtc,
 				 const struct drm_display_mode *mode,
 				 struct drm_display_mode *adjusted_mode)
@@ -559,6 +600,7 @@ static void ltdc_crtc_atomic_flush(struct drm_crtc *crtc,
 }
 
 static const struct drm_crtc_helper_funcs ltdc_crtc_helper_funcs = {
+	.mode_valid = ltdc_crtc_mode_valid,
 	.mode_fixup = ltdc_crtc_mode_fixup,
 	.mode_set_nofb = ltdc_crtc_mode_set_nofb,
 	.atomic_flush = ltdc_crtc_atomic_flush,
@@ -822,13 +864,13 @@ static struct drm_plane *ltdc_plane_create(struct drm_device *ddev,
 
 	plane = devm_kzalloc(dev, sizeof(*plane), GFP_KERNEL);
 	if (!plane)
-		return 0;
+		return NULL;
 
 	ret = drm_universal_plane_init(ddev, plane, possible_crtcs,
 				       &ltdc_plane_funcs, formats, nb_fmt,
 				       NULL, type, NULL);
 	if (ret < 0)
-		return 0;
+		return NULL;
 
 	drm_plane_helper_add(plane, &ltdc_plane_helper_funcs);
 
@@ -953,11 +995,15 @@ static int ltdc_get_caps(struct drm_device *ddev)
 		 * does not work on 2nd layer.
 		 */
 		ldev->caps.non_alpha_only_l1 = true;
+		ldev->caps.pad_max_freq_hz = 90000000;
+		if (ldev->caps.hw_version == HWVER_10200)
+			ldev->caps.pad_max_freq_hz = 65000000;
 		break;
 	case HWVER_20101:
 		ldev->caps.reg_ofs = REG_OFS_4;
 		ldev->caps.pix_fmt_hw = ltdc_pix_fmt_a1;
 		ldev->caps.non_alpha_only_l1 = false;
+		ldev->caps.pad_max_freq_hz = 150000000;
 		break;
 	default:
 		return -ENODEV;
@@ -987,14 +1033,13 @@ int ltdc_load(struct drm_device *ddev)
 						  &bridge[i]);
 
 		/*
-		 * If at least one endpoint is ready, continue probing,
-		 * else if at least one endpoint is -EPROBE_DEFER and
-		 * there is no previous ready endpoints, defer probing.
+		 * If at least one endpoint is -EPROBE_DEFER, defer probing,
+		 * else if at least one endpoint is ready, continue probing.
 		 */
-		if (!ret)
+		if (ret == -EPROBE_DEFER)
+			return ret;
+		else if (!ret)
 			endpoint_not_ready = 0;
-		else if (ret == -EPROBE_DEFER && endpoint_not_ready)
-			endpoint_not_ready = -EPROBE_DEFER;
 	}
 
 	if (endpoint_not_ready)
@@ -1037,8 +1082,11 @@ int ltdc_load(struct drm_device *ddev)
 		}
 	}
 
-	if (!IS_ERR(rstc))
+	if (!IS_ERR(rstc)) {
+		reset_control_assert(rstc);
+		usleep_range(10, 20);
 		reset_control_deassert(rstc);
+	}
 
 	/* Disable interrupts */
 	reg_clear(ldev->regs, LTDC_IER,
diff --git a/drivers/gpu/drm/stm/ltdc.h b/drivers/gpu/drm/stm/ltdc.h
index 1e16d6afb0d2..d5afb8960867 100644
--- a/drivers/gpu/drm/stm/ltdc.h
+++ b/drivers/gpu/drm/stm/ltdc.h
@@ -18,6 +18,7 @@ struct ltdc_caps {
 	u32 bus_width;		/* bus width (32 or 64 bits) */
 	const u32 *pix_fmt_hw;	/* supported pixel formats */
 	bool non_alpha_only_l1; /* non-native no-alpha formats on layer 1 */
+	int pad_max_freq_hz;	/* max frequency supported by pad */
 };
 
 #define LTDC_MAX_LAYER	4
diff --git a/drivers/gpu/drm/sun4i/Kconfig b/drivers/gpu/drm/sun4i/Kconfig
index 156a865c3e6d..c2c042287c19 100644
--- a/drivers/gpu/drm/sun4i/Kconfig
+++ b/drivers/gpu/drm/sun4i/Kconfig
@@ -68,4 +68,11 @@ config DRM_SUN8I_MIXER
 	  graphics mixture and feed graphics to TCON, If M is
 	  selected the module will be called sun8i-mixer.
 
+config DRM_SUN8I_TCON_TOP
+	tristate
+	default DRM_SUN4I if DRM_SUN8I_MIXER!=n
+	help
+	  TCON TOP is responsible for configuring display pipeline for
+	  HTMI, TVE and LCD.
+
 endif
diff --git a/drivers/gpu/drm/sun4i/Makefile b/drivers/gpu/drm/sun4i/Makefile
index 2589f4acd5ae..b04ea0f3da75 100644
--- a/drivers/gpu/drm/sun4i/Makefile
+++ b/drivers/gpu/drm/sun4i/Makefile
@@ -37,3 +37,4 @@ obj-$(CONFIG_DRM_SUN4I_HDMI)	+= sun4i-drm-hdmi.o
 obj-$(CONFIG_DRM_SUN6I_DSI)	+= sun6i-dsi.o
 obj-$(CONFIG_DRM_SUN8I_DW_HDMI)	+= sun8i-drm-hdmi.o
 obj-$(CONFIG_DRM_SUN8I_MIXER)	+= sun8i-mixer.o
+obj-$(CONFIG_DRM_SUN8I_TCON_TOP) += sun8i_tcon_top.o
diff --git a/drivers/gpu/drm/sun4i/sun4i_backend.c b/drivers/gpu/drm/sun4i/sun4i_backend.c
index de0a76dfa1a2..d7950b52a1fd 100644
--- a/drivers/gpu/drm/sun4i/sun4i_backend.c
+++ b/drivers/gpu/drm/sun4i/sun4i_backend.c
@@ -86,12 +86,6 @@ static inline bool sun4i_backend_format_is_packed_yuv422(uint32_t format)
 	}
 }
 
-static inline bool sun4i_backend_format_is_yuv(uint32_t format)
-{
-	return sun4i_backend_format_is_planar_yuv(format) ||
-		sun4i_backend_format_is_packed_yuv422(format);
-}
-
 static void sun4i_backend_apply_color_correction(struct sunxi_engine *engine)
 {
 	int i;
@@ -304,7 +298,7 @@ int sun4i_backend_update_layer_formats(struct sun4i_backend *backend,
 			   SUN4I_BACKEND_ATTCTL_REG0_LAY_GLBALPHA_EN,
 			   val);
 
-	if (sun4i_backend_format_is_yuv(fb->format->format))
+	if (fb->format->is_yuv)
 		return sun4i_backend_update_yuv_format(backend, layer, plane);
 
 	ret = sun4i_backend_drm_format_to_layer(fb->format->format, &val);
@@ -384,7 +378,7 @@ int sun4i_backend_update_layer_buffer(struct sun4i_backend *backend,
 	 */
 	paddr -= PHYS_OFFSET;
 
-	if (sun4i_backend_format_is_yuv(fb->format->format))
+	if (fb->format->is_yuv)
 		return sun4i_backend_update_yuv_buffer(backend, fb, paddr);
 
 	/* Write the 32 lower bits of the address (in bits) */
@@ -502,7 +496,7 @@ static int sun4i_backend_atomic_check(struct sunxi_engine *engine,
 		if (fb->format->has_alpha || (plane_state->alpha != DRM_BLEND_ALPHA_OPAQUE))
 			num_alpha_planes++;
 
-		if (sun4i_backend_format_is_yuv(fb->format->format)) {
+		if (fb->format->is_yuv) {
 			DRM_DEBUG_DRIVER("Plane FB format is YUV\n");
 			num_yuv_planes++;
 		}
diff --git a/drivers/gpu/drm/sun4i/sun4i_crtc.c b/drivers/gpu/drm/sun4i/sun4i_crtc.c
index 2d7c57406715..3eedf335a935 100644
--- a/drivers/gpu/drm/sun4i/sun4i_crtc.c
+++ b/drivers/gpu/drm/sun4i/sun4i_crtc.c
@@ -242,7 +242,7 @@ struct sun4i_crtc *sun4i_crtc_init(struct drm_device *drm,
 
 	/* Set possible_crtcs to this crtc for overlay planes */
 	for (i = 0; planes[i]; i++) {
-		uint32_t possible_crtcs = BIT(drm_crtc_index(&scrtc->crtc));
+		uint32_t possible_crtcs = drm_crtc_mask(&scrtc->crtc);
 		struct drm_plane *plane = planes[i];
 
 		if (plane->type == DRM_PLANE_TYPE_OVERLAY)
diff --git a/drivers/gpu/drm/sun4i/sun4i_drv.c b/drivers/gpu/drm/sun4i/sun4i_drv.c
index 50d19605c38f..dd19d674055c 100644
--- a/drivers/gpu/drm/sun4i/sun4i_drv.c
+++ b/drivers/gpu/drm/sun4i/sun4i_drv.c
@@ -26,6 +26,7 @@
 #include "sun4i_frontend.h"
 #include "sun4i_framebuffer.h"
 #include "sun4i_tcon.h"
+#include "sun8i_tcon_top.h"
 
 DEFINE_DRM_GEM_CMA_FOPS(sun4i_drv_fops);
 
@@ -143,7 +144,7 @@ cleanup_mode_config:
 	drm_mode_config_cleanup(drm);
 	of_reserved_mem_device_release(dev);
 free_drm:
-	drm_dev_unref(drm);
+	drm_dev_put(drm);
 	return ret;
 }
 
@@ -156,7 +157,7 @@ static void sun4i_drv_unbind(struct device *dev)
 	sun4i_framebuffer_free(drm);
 	drm_mode_config_cleanup(drm);
 	of_reserved_mem_device_release(dev);
-	drm_dev_unref(drm);
+	drm_dev_put(drm);
 }
 
 static const struct component_master_ops sun4i_drv_master_ops = {
@@ -197,6 +198,28 @@ static bool sun4i_drv_node_is_tcon(struct device_node *node)
 	return !!of_match_node(sun4i_tcon_of_table, node);
 }
 
+static bool sun4i_drv_node_is_tcon_with_ch0(struct device_node *node)
+{
+	const struct of_device_id *match;
+
+	match = of_match_node(sun4i_tcon_of_table, node);
+	if (match) {
+		struct sun4i_tcon_quirks *quirks;
+
+		quirks = (struct sun4i_tcon_quirks *)match->data;
+
+		return quirks->has_channel_0;
+	}
+
+	return false;
+}
+
+static bool sun4i_drv_node_is_tcon_top(struct device_node *node)
+{
+	return IS_ENABLED(CONFIG_DRM_SUN8I_TCON_TOP) &&
+		!!of_match_node(sun8i_tcon_top_of_table, node);
+}
+
 static int compare_of(struct device *dev, void *data)
 {
 	DRM_DEBUG_DRIVER("Comparing of node %pOF with %pOF\n",
@@ -231,12 +254,69 @@ struct endpoint_list {
 	DECLARE_KFIFO(fifo, struct device_node *, 16);
 };
 
+static void sun4i_drv_traverse_endpoints(struct endpoint_list *list,
+					 struct device_node *node,
+					 int port_id)
+{
+	struct device_node *ep, *remote, *port;
+
+	port = of_graph_get_port_by_id(node, port_id);
+	if (!port) {
+		DRM_DEBUG_DRIVER("No output to bind on port %d\n", port_id);
+		return;
+	}
+
+	for_each_available_child_of_node(port, ep) {
+		remote = of_graph_get_remote_port_parent(ep);
+		if (!remote) {
+			DRM_DEBUG_DRIVER("Error retrieving the output node\n");
+			continue;
+		}
+
+		if (sun4i_drv_node_is_tcon(node)) {
+			/*
+			 * TCON TOP is always probed before TCON. However, TCON
+			 * points back to TCON TOP when it is source for HDMI.
+			 * We have to skip it here to prevent infinite looping
+			 * between TCON TOP and TCON.
+			 */
+			if (sun4i_drv_node_is_tcon_top(remote)) {
+				DRM_DEBUG_DRIVER("TCON output endpoint is TCON TOP... skipping\n");
+				of_node_put(remote);
+				continue;
+			}
+
+			/*
+			 * If the node is our TCON with channel 0, the first
+			 * port is used for panel or bridges, and will not be
+			 * part of the component framework.
+			 */
+			if (sun4i_drv_node_is_tcon_with_ch0(node)) {
+				struct of_endpoint endpoint;
+
+				if (of_graph_parse_endpoint(ep, &endpoint)) {
+					DRM_DEBUG_DRIVER("Couldn't parse endpoint\n");
+					of_node_put(remote);
+					continue;
+				}
+
+				if (!endpoint.id) {
+					DRM_DEBUG_DRIVER("Endpoint is our panel... skipping\n");
+					of_node_put(remote);
+					continue;
+				}
+			}
+		}
+
+		kfifo_put(&list->fifo, remote);
+	}
+}
+
 static int sun4i_drv_add_endpoints(struct device *dev,
 				   struct endpoint_list *list,
 				   struct component_match **match,
 				   struct device_node *node)
 {
-	struct device_node *port, *ep, *remote;
 	int count = 0;
 
 	/*
@@ -272,41 +352,13 @@ static int sun4i_drv_add_endpoints(struct device *dev,
 		count++;
 	}
 
-	/* Inputs are listed first, then outputs */
-	port = of_graph_get_port_by_id(node, 1);
-	if (!port) {
-		DRM_DEBUG_DRIVER("No output to bind\n");
-		return count;
-	}
+	/* each node has at least one output */
+	sun4i_drv_traverse_endpoints(list, node, 1);
 
-	for_each_available_child_of_node(port, ep) {
-		remote = of_graph_get_remote_port_parent(ep);
-		if (!remote) {
-			DRM_DEBUG_DRIVER("Error retrieving the output node\n");
-			of_node_put(remote);
-			continue;
-		}
-
-		/*
-		 * If the node is our TCON, the first port is used for
-		 * panel or bridges, and will not be part of the
-		 * component framework.
-		 */
-		if (sun4i_drv_node_is_tcon(node)) {
-			struct of_endpoint endpoint;
-
-			if (of_graph_parse_endpoint(ep, &endpoint)) {
-				DRM_DEBUG_DRIVER("Couldn't parse endpoint\n");
-				continue;
-			}
-
-			if (!endpoint.id) {
-				DRM_DEBUG_DRIVER("Endpoint is our panel... skipping\n");
-				continue;
-			}
-		}
-
-		kfifo_put(&list->fifo, remote);
+	/* TCON TOP has second and third output */
+	if (sun4i_drv_node_is_tcon_top(node)) {
+		sun4i_drv_traverse_endpoints(list, node, 3);
+		sun4i_drv_traverse_endpoints(list, node, 5);
 	}
 
 	return count;
@@ -366,6 +418,7 @@ static const struct of_device_id sun4i_drv_of_table[] = {
 	{ .compatible = "allwinner,sun8i-a33-display-engine" },
 	{ .compatible = "allwinner,sun8i-a83t-display-engine" },
 	{ .compatible = "allwinner,sun8i-h3-display-engine" },
+	{ .compatible = "allwinner,sun8i-r40-display-engine" },
 	{ .compatible = "allwinner,sun8i-v3s-display-engine" },
 	{ .compatible = "allwinner,sun9i-a80-display-engine" },
 	{ }
diff --git a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
index fa4bcd092eaf..061d2e0d9011 100644
--- a/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
+++ b/drivers/gpu/drm/sun4i/sun4i_hdmi_enc.c
@@ -220,7 +220,7 @@ static int sun4i_hdmi_get_modes(struct drm_connector *connector)
 	DRM_DEBUG_DRIVER("Monitor is %s monitor\n",
 			 hdmi->hdmi_monitor ? "an HDMI" : "a DVI");
 
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 	cec_s_phys_addr_from_edid(hdmi->cec_adap, edid);
 	ret = drm_add_edid_modes(connector, edid);
 	kfree(edid);
@@ -623,7 +623,7 @@ static int sun4i_hdmi_bind(struct device *dev, struct device *master,
 	ret = cec_register_adapter(hdmi->cec_adap, dev);
 	if (ret < 0)
 		goto err_cleanup_connector;
-	drm_mode_connector_attach_encoder(&hdmi->connector, &hdmi->encoder);
+	drm_connector_attach_encoder(&hdmi->connector, &hdmi->encoder);
 
 	return 0;
 
diff --git a/drivers/gpu/drm/sun4i/sun4i_lvds.c b/drivers/gpu/drm/sun4i/sun4i_lvds.c
index be3f14d7746d..af7dcb6da351 100644
--- a/drivers/gpu/drm/sun4i/sun4i_lvds.c
+++ b/drivers/gpu/drm/sun4i/sun4i_lvds.c
@@ -136,7 +136,7 @@ int sun4i_lvds_init(struct drm_device *drm, struct sun4i_tcon *tcon)
 	}
 
 	/* The LVDS encoder can only work with the TCON channel 0 */
-	lvds->encoder.possible_crtcs = BIT(drm_crtc_index(&tcon->crtc->crtc));
+	lvds->encoder.possible_crtcs = drm_crtc_mask(&tcon->crtc->crtc);
 
 	if (tcon->panel) {
 		drm_connector_helper_add(&lvds->connector,
@@ -149,7 +149,7 @@ int sun4i_lvds_init(struct drm_device *drm, struct sun4i_tcon *tcon)
 			goto err_cleanup_connector;
 		}
 
-		drm_mode_connector_attach_encoder(&lvds->connector,
+		drm_connector_attach_encoder(&lvds->connector,
 						  &lvds->encoder);
 
 		ret = drm_panel_attach(tcon->panel, &lvds->connector);
diff --git a/drivers/gpu/drm/sun4i/sun4i_rgb.c b/drivers/gpu/drm/sun4i/sun4i_rgb.c
index f2fa1f210509..bf068da6b12e 100644
--- a/drivers/gpu/drm/sun4i/sun4i_rgb.c
+++ b/drivers/gpu/drm/sun4i/sun4i_rgb.c
@@ -202,7 +202,7 @@ int sun4i_rgb_init(struct drm_device *drm, struct sun4i_tcon *tcon)
 	}
 
 	/* The RGB encoder can only work with the TCON channel 0 */
-	rgb->encoder.possible_crtcs = BIT(drm_crtc_index(&tcon->crtc->crtc));
+	rgb->encoder.possible_crtcs = drm_crtc_mask(&tcon->crtc->crtc);
 
 	if (tcon->panel) {
 		drm_connector_helper_add(&rgb->connector,
@@ -215,7 +215,7 @@ int sun4i_rgb_init(struct drm_device *drm, struct sun4i_tcon *tcon)
 			goto err_cleanup_connector;
 		}
 
-		drm_mode_connector_attach_encoder(&rgb->connector,
+		drm_connector_attach_encoder(&rgb->connector,
 						  &rgb->encoder);
 
 		ret = drm_panel_attach(tcon->panel, &rgb->connector);
diff --git a/drivers/gpu/drm/sun4i/sun4i_tcon.c b/drivers/gpu/drm/sun4i/sun4i_tcon.c
index 08747fc3ee71..3fb084f802e2 100644
--- a/drivers/gpu/drm/sun4i/sun4i_tcon.c
+++ b/drivers/gpu/drm/sun4i/sun4i_tcon.c
@@ -17,7 +17,6 @@
 #include <drm/drm_encoder.h>
 #include <drm/drm_modes.h>
 #include <drm/drm_of.h>
-#include <drm/drm_panel.h>
 
 #include <uapi/drm/drm_mode.h>
 
@@ -418,9 +417,6 @@ static void sun4i_tcon0_mode_set_lvds(struct sun4i_tcon *tcon,
 static void sun4i_tcon0_mode_set_rgb(struct sun4i_tcon *tcon,
 				     const struct drm_display_mode *mode)
 {
-	struct drm_panel *panel = tcon->panel;
-	struct drm_connector *connector = panel->connector;
-	struct drm_display_info display_info = connector->display_info;
 	unsigned int bp, hsync, vsync;
 	u8 clk_delay;
 	u32 val = 0;
@@ -478,27 +474,6 @@ static void sun4i_tcon0_mode_set_rgb(struct sun4i_tcon *tcon,
 	if (mode->flags & DRM_MODE_FLAG_PVSYNC)
 		val |= SUN4I_TCON0_IO_POL_VSYNC_POSITIVE;
 
-	/*
-	 * On A20 and similar SoCs, the only way to achieve Positive Edge
-	 * (Rising Edge), is setting dclk clock phase to 2/3(240°).
-	 * By default TCON works in Negative Edge(Falling Edge),
-	 * this is why phase is set to 0 in that case.
-	 * Unfortunately there's no way to logically invert dclk through
-	 * IO_POL register.
-	 * The only acceptable way to work, triple checked with scope,
-	 * is using clock phase set to 0° for Negative Edge and set to 240°
-	 * for Positive Edge.
-	 * On A33 and similar SoCs there would be a 90° phase option,
-	 * but it divides also dclk by 2.
-	 * Following code is a way to avoid quirks all around TCON
-	 * and DOTCLOCK drivers.
-	 */
-	if (display_info.bus_flags & DRM_BUS_FLAG_PIXDATA_POSEDGE)
-		clk_set_phase(tcon->dclk, 240);
-
-	if (display_info.bus_flags & DRM_BUS_FLAG_PIXDATA_NEGEDGE)
-		clk_set_phase(tcon->dclk, 0);
-
 	regmap_update_bits(tcon->regs, SUN4I_TCON0_IO_POL_REG,
 			   SUN4I_TCON0_IO_POL_HSYNC_POSITIVE | SUN4I_TCON0_IO_POL_VSYNC_POSITIVE,
 			   val);
@@ -791,12 +766,14 @@ static int sun4i_tcon_init_regmap(struct device *dev,
  */
 static struct sunxi_engine *
 sun4i_tcon_find_engine_traverse(struct sun4i_drv *drv,
-				struct device_node *node)
+				struct device_node *node,
+				u32 port_id)
 {
 	struct device_node *port, *ep, *remote;
 	struct sunxi_engine *engine = ERR_PTR(-EINVAL);
+	u32 reg = 0;
 
-	port = of_graph_get_port_by_id(node, 0);
+	port = of_graph_get_port_by_id(node, port_id);
 	if (!port)
 		return ERR_PTR(-EINVAL);
 
@@ -826,8 +803,21 @@ sun4i_tcon_find_engine_traverse(struct sun4i_drv *drv,
 		if (remote == engine->node)
 			goto out_put_remote;
 
+	/*
+	 * According to device tree binding input ports have even id
+	 * number and output ports have odd id. Since component with
+	 * more than one input and one output (TCON TOP) exits, correct
+	 * remote input id has to be calculated by subtracting 1 from
+	 * remote output id. If this for some reason can't be done, 0
+	 * is used as input port id.
+	 */
+	of_node_put(port);
+	port = of_graph_get_remote_port(ep);
+	if (!of_property_read_u32(port, "reg", &reg) && reg > 0)
+		reg -= 1;
+
 	/* keep looking through upstream ports */
-	engine = sun4i_tcon_find_engine_traverse(drv, remote);
+	engine = sun4i_tcon_find_engine_traverse(drv, remote, reg);
 
 out_put_remote:
 	of_node_put(remote);
@@ -950,7 +940,7 @@ static struct sunxi_engine *sun4i_tcon_find_engine(struct sun4i_drv *drv,
 
 	/* Fallback to old method by traversing input endpoints */
 	of_node_put(port);
-	return sun4i_tcon_find_engine_traverse(drv, node);
+	return sun4i_tcon_find_engine_traverse(drv, node, 0);
 }
 
 static int sun4i_tcon_bind(struct device *dev, struct device *master,
@@ -1092,23 +1082,25 @@ static int sun4i_tcon_bind(struct device *dev, struct device *master,
 		goto err_free_dotclock;
 	}
 
-	/*
-	 * If we have an LVDS panel connected to the TCON, we should
-	 * just probe the LVDS connector. Otherwise, just probe RGB as
-	 * we used to.
-	 */
-	remote = of_graph_get_remote_node(dev->of_node, 1, 0);
-	if (of_device_is_compatible(remote, "panel-lvds"))
-		if (can_lvds)
-			ret = sun4i_lvds_init(drm, tcon);
+	if (tcon->quirks->has_channel_0) {
+		/*
+		 * If we have an LVDS panel connected to the TCON, we should
+		 * just probe the LVDS connector. Otherwise, just probe RGB as
+		 * we used to.
+		 */
+		remote = of_graph_get_remote_node(dev->of_node, 1, 0);
+		if (of_device_is_compatible(remote, "panel-lvds"))
+			if (can_lvds)
+				ret = sun4i_lvds_init(drm, tcon);
+			else
+				ret = -EINVAL;
 		else
-			ret = -EINVAL;
-	else
-		ret = sun4i_rgb_init(drm, tcon);
-	of_node_put(remote);
+			ret = sun4i_rgb_init(drm, tcon);
+		of_node_put(remote);
 
-	if (ret < 0)
-		goto err_free_dotclock;
+		if (ret < 0)
+			goto err_free_dotclock;
+	}
 
 	if (tcon->quirks->needs_de_be_mux) {
 		/*
@@ -1162,13 +1154,19 @@ static const struct component_ops sun4i_tcon_ops = {
 static int sun4i_tcon_probe(struct platform_device *pdev)
 {
 	struct device_node *node = pdev->dev.of_node;
+	const struct sun4i_tcon_quirks *quirks;
 	struct drm_bridge *bridge;
 	struct drm_panel *panel;
 	int ret;
 
-	ret = drm_of_find_panel_or_bridge(node, 1, 0, &panel, &bridge);
-	if (ret == -EPROBE_DEFER)
-		return ret;
+	quirks = of_device_get_match_data(&pdev->dev);
+
+	/* panels and bridges are present only on TCONs with channel 0 */
+	if (quirks->has_channel_0) {
+		ret = drm_of_find_panel_or_bridge(node, 1, 0, &panel, &bridge);
+		if (ret == -EPROBE_DEFER)
+			return ret;
+	}
 
 	return component_add(&pdev->dev, &sun4i_tcon_ops);
 }
diff --git a/drivers/gpu/drm/sun4i/sun4i_tv.c b/drivers/gpu/drm/sun4i/sun4i_tv.c
index b070d522ed8d..1a838d208211 100644
--- a/drivers/gpu/drm/sun4i/sun4i_tv.c
+++ b/drivers/gpu/drm/sun4i/sun4i_tv.c
@@ -623,7 +623,7 @@ static int sun4i_tv_bind(struct device *dev, struct device *master,
 	}
 	tv->connector.interlace_allowed = true;
 
-	drm_mode_connector_attach_encoder(&tv->connector, &tv->encoder);
+	drm_connector_attach_encoder(&tv->connector, &tv->encoder);
 
 	return 0;
 
diff --git a/drivers/gpu/drm/sun4i/sun6i_mipi_dsi.c b/drivers/gpu/drm/sun4i/sun6i_mipi_dsi.c
index bfbf761f0c1d..e3b34a345546 100644
--- a/drivers/gpu/drm/sun4i/sun6i_mipi_dsi.c
+++ b/drivers/gpu/drm/sun4i/sun6i_mipi_dsi.c
@@ -13,6 +13,7 @@
 #include <linux/pm_runtime.h>
 #include <linux/regmap.h>
 #include <linux/reset.h>
+#include <linux/slab.h>
 
 #include <linux/phy/phy.h>
 
@@ -247,10 +248,8 @@ static u16 sun6i_dsi_crc_compute(u8 const *buffer, size_t len)
 	return crc_ccitt(0xffff, buffer, len);
 }
 
-static u16 sun6i_dsi_crc_repeat_compute(u8 pd, size_t len)
+static u16 sun6i_dsi_crc_repeat(u8 pd, u8 *buffer, size_t len)
 {
-	u8 buffer[len];
-
 	memset(buffer, pd, len);
 
 	return sun6i_dsi_crc_compute(buffer, len);
@@ -274,11 +273,11 @@ static u32 sun6i_dsi_build_blk0_pkt(u8 vc, u16 wc)
 					wc & 0xff, wc >> 8);
 }
 
-static u32 sun6i_dsi_build_blk1_pkt(u16 pd, size_t len)
+static u32 sun6i_dsi_build_blk1_pkt(u16 pd, u8 *buffer, size_t len)
 {
 	u32 val = SUN6I_DSI_BLK_PD(pd);
 
-	return val | SUN6I_DSI_BLK_PF(sun6i_dsi_crc_repeat_compute(pd, len));
+	return val | SUN6I_DSI_BLK_PF(sun6i_dsi_crc_repeat(pd, buffer, len));
 }
 
 static void sun6i_dsi_inst_abort(struct sun6i_dsi *dsi)
@@ -452,6 +451,54 @@ static void sun6i_dsi_setup_timings(struct sun6i_dsi *dsi,
 	struct mipi_dsi_device *device = dsi->device;
 	unsigned int Bpp = mipi_dsi_pixel_format_to_bpp(device->format) / 8;
 	u16 hbp, hfp, hsa, hblk, vblk;
+	size_t bytes;
+	u8 *buffer;
+
+	/* Do all timing calculations up front to allocate buffer space */
+
+	/*
+	 * A sync period is composed of a blanking packet (4 bytes +
+	 * payload + 2 bytes) and a sync event packet (4 bytes). Its
+	 * minimal size is therefore 10 bytes
+	 */
+#define HSA_PACKET_OVERHEAD	10
+	hsa = max((unsigned int)HSA_PACKET_OVERHEAD,
+		  (mode->hsync_end - mode->hsync_start) * Bpp - HSA_PACKET_OVERHEAD);
+
+	/*
+	 * The backporch is set using a blanking packet (4 bytes +
+	 * payload + 2 bytes). Its minimal size is therefore 6 bytes
+	 */
+#define HBP_PACKET_OVERHEAD	6
+	hbp = max((unsigned int)HBP_PACKET_OVERHEAD,
+		  (mode->hsync_start - mode->hdisplay) * Bpp - HBP_PACKET_OVERHEAD);
+
+	/*
+	 * The frontporch is set using a blanking packet (4 bytes +
+	 * payload + 2 bytes). Its minimal size is therefore 6 bytes
+	 */
+#define HFP_PACKET_OVERHEAD	6
+	hfp = max((unsigned int)HFP_PACKET_OVERHEAD,
+		  (mode->htotal - mode->hsync_end) * Bpp - HFP_PACKET_OVERHEAD);
+
+	/*
+	 * hblk seems to be the line + porches length.
+	 */
+	hblk = mode->htotal * Bpp - hsa;
+
+	/*
+	 * And I'm not entirely sure what vblk is about. The driver in
+	 * Allwinner BSP is using a rather convoluted calculation
+	 * there only for 4 lanes. However, using 0 (the !4 lanes
+	 * case) even with a 4 lanes screen seems to work...
+	 */
+	vblk = 0;
+
+	/* How many bytes do we need to send all payloads? */
+	bytes = max_t(size_t, max(max(hfp, hblk), max(hsa, hbp)), vblk);
+	buffer = kmalloc(bytes, GFP_KERNEL);
+	if (WARN_ON(!buffer))
+		return;
 
 	regmap_write(dsi->regs, SUN6I_DSI_BASIC_CTL_REG, 0);
 
@@ -485,63 +532,37 @@ static void sun6i_dsi_setup_timings(struct sun6i_dsi *dsi,
 		     SUN6I_DSI_BASIC_SIZE1_VACT(mode->vdisplay) |
 		     SUN6I_DSI_BASIC_SIZE1_VT(mode->vtotal));
 
-	/*
-	 * A sync period is composed of a blanking packet (4 bytes +
-	 * payload + 2 bytes) and a sync event packet (4 bytes). Its
-	 * minimal size is therefore 10 bytes
-	 */
-#define HSA_PACKET_OVERHEAD	10
-	hsa = max((unsigned int)HSA_PACKET_OVERHEAD,
-		  (mode->hsync_end - mode->hsync_start) * Bpp - HSA_PACKET_OVERHEAD);
+	/* sync */
 	regmap_write(dsi->regs, SUN6I_DSI_BLK_HSA0_REG,
 		     sun6i_dsi_build_blk0_pkt(device->channel, hsa));
 	regmap_write(dsi->regs, SUN6I_DSI_BLK_HSA1_REG,
-		     sun6i_dsi_build_blk1_pkt(0, hsa));
+		     sun6i_dsi_build_blk1_pkt(0, buffer, hsa));
 
-	/*
-	 * The backporch is set using a blanking packet (4 bytes +
-	 * payload + 2 bytes). Its minimal size is therefore 6 bytes
-	 */
-#define HBP_PACKET_OVERHEAD	6
-	hbp = max((unsigned int)HBP_PACKET_OVERHEAD,
-		  (mode->hsync_start - mode->hdisplay) * Bpp - HBP_PACKET_OVERHEAD);
+	/* backporch */
 	regmap_write(dsi->regs, SUN6I_DSI_BLK_HBP0_REG,
 		     sun6i_dsi_build_blk0_pkt(device->channel, hbp));
 	regmap_write(dsi->regs, SUN6I_DSI_BLK_HBP1_REG,
-		     sun6i_dsi_build_blk1_pkt(0, hbp));
+		     sun6i_dsi_build_blk1_pkt(0, buffer, hbp));
 
-	/*
-	 * The frontporch is set using a blanking packet (4 bytes +
-	 * payload + 2 bytes). Its minimal size is therefore 6 bytes
-	 */
-#define HFP_PACKET_OVERHEAD	6
-	hfp = max((unsigned int)HFP_PACKET_OVERHEAD,
-		  (mode->htotal - mode->hsync_end) * Bpp - HFP_PACKET_OVERHEAD);
+	/* frontporch */
 	regmap_write(dsi->regs, SUN6I_DSI_BLK_HFP0_REG,
 		     sun6i_dsi_build_blk0_pkt(device->channel, hfp));
 	regmap_write(dsi->regs, SUN6I_DSI_BLK_HFP1_REG,
-		     sun6i_dsi_build_blk1_pkt(0, hfp));
+		     sun6i_dsi_build_blk1_pkt(0, buffer, hfp));
 
-	/*
-	 * hblk seems to be the line + porches length.
-	 */
-	hblk = mode->htotal * Bpp - hsa;
+	/* hblk */
 	regmap_write(dsi->regs, SUN6I_DSI_BLK_HBLK0_REG,
 		     sun6i_dsi_build_blk0_pkt(device->channel, hblk));
 	regmap_write(dsi->regs, SUN6I_DSI_BLK_HBLK1_REG,
-		     sun6i_dsi_build_blk1_pkt(0, hblk));
+		     sun6i_dsi_build_blk1_pkt(0, buffer, hblk));
 
-	/*
-	 * And I'm not entirely sure what vblk is about. The driver in
-	 * Allwinner BSP is using a rather convoluted calculation
-	 * there only for 4 lanes. However, using 0 (the !4 lanes
-	 * case) even with a 4 lanes screen seems to work...
-	 */
-	vblk = 0;
+	/* vblk */
 	regmap_write(dsi->regs, SUN6I_DSI_BLK_VBLK0_REG,
 		     sun6i_dsi_build_blk0_pkt(device->channel, vblk));
 	regmap_write(dsi->regs, SUN6I_DSI_BLK_VBLK1_REG,
-		     sun6i_dsi_build_blk1_pkt(0, vblk));
+		     sun6i_dsi_build_blk1_pkt(0, buffer, vblk));
+
+	kfree(buffer);
 }
 
 static int sun6i_dsi_start(struct sun6i_dsi *dsi,
@@ -812,8 +833,8 @@ static int sun6i_dsi_attach(struct mipi_dsi_host *host,
 
 	dsi->device = device;
 	dsi->panel = of_drm_find_panel(device->dev.of_node);
-	if (!dsi->panel)
-		return -EINVAL;
+	if (IS_ERR(dsi->panel))
+		return PTR_ERR(dsi->panel);
 
 	dev_info(host->dev, "Attached device %s\n", device->name);
 
@@ -920,7 +941,7 @@ static int sun6i_dsi_bind(struct device *dev, struct device *master,
 		goto err_cleanup_connector;
 	}
 
-	drm_mode_connector_attach_encoder(&dsi->connector, &dsi->encoder);
+	drm_connector_attach_encoder(&dsi->connector, &dsi->encoder);
 	drm_panel_attach(dsi->panel, &dsi->connector);
 
 	return 0;
@@ -1040,7 +1061,7 @@ static int sun6i_dsi_remove(struct platform_device *pdev)
 	return 0;
 }
 
-static int sun6i_dsi_runtime_resume(struct device *dev)
+static int __maybe_unused sun6i_dsi_runtime_resume(struct device *dev)
 {
 	struct sun6i_dsi *dsi = dev_get_drvdata(dev);
 
@@ -1069,7 +1090,7 @@ static int sun6i_dsi_runtime_resume(struct device *dev)
 	return 0;
 }
 
-static int sun6i_dsi_runtime_suspend(struct device *dev)
+static int __maybe_unused sun6i_dsi_runtime_suspend(struct device *dev)
 {
 	struct sun6i_dsi *dsi = dev_get_drvdata(dev);
 
diff --git a/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.c b/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.c
index 9f40a44b456b..31875b636434 100644
--- a/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.c
+++ b/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.c
@@ -12,6 +12,7 @@
 #include <drm/drm_crtc_helper.h>
 
 #include "sun8i_dw_hdmi.h"
+#include "sun8i_tcon_top.h"
 
 static void sun8i_dw_hdmi_encoder_mode_set(struct drm_encoder *encoder,
 					   struct drm_display_mode *mode,
@@ -41,6 +42,44 @@ sun8i_dw_hdmi_mode_valid(struct drm_connector *connector,
 	return MODE_OK;
 }
 
+static bool sun8i_dw_hdmi_node_is_tcon_top(struct device_node *node)
+{
+	return IS_ENABLED(CONFIG_DRM_SUN8I_TCON_TOP) &&
+		!!of_match_node(sun8i_tcon_top_of_table, node);
+}
+
+static u32 sun8i_dw_hdmi_find_possible_crtcs(struct drm_device *drm,
+					     struct device_node *node)
+{
+	struct device_node *port, *ep, *remote, *remote_port;
+	u32 crtcs = 0;
+
+	remote = of_graph_get_remote_node(node, 0, -1);
+	if (!remote)
+		return 0;
+
+	if (sun8i_dw_hdmi_node_is_tcon_top(remote)) {
+		port = of_graph_get_port_by_id(remote, 4);
+		if (!port)
+			goto crtcs_exit;
+
+		for_each_child_of_node(port, ep) {
+			remote_port = of_graph_get_remote_port(ep);
+			if (remote_port) {
+				crtcs |= drm_of_crtc_port_mask(drm, remote_port);
+				of_node_put(remote_port);
+			}
+		}
+	} else {
+		crtcs = drm_of_find_possible_crtcs(drm, node);
+	}
+
+crtcs_exit:
+	of_node_put(remote);
+
+	return crtcs;
+}
+
 static int sun8i_dw_hdmi_bind(struct device *dev, struct device *master,
 			      void *data)
 {
@@ -63,7 +102,8 @@ static int sun8i_dw_hdmi_bind(struct device *dev, struct device *master,
 	hdmi->dev = &pdev->dev;
 	encoder = &hdmi->encoder;
 
-	encoder->possible_crtcs = drm_of_find_possible_crtcs(drm, dev->of_node);
+	encoder->possible_crtcs =
+		sun8i_dw_hdmi_find_possible_crtcs(drm, dev->of_node);
 	/*
 	 * If we failed to find the CRTC(s) which this encoder is
 	 * supposed to be connected to, it's because the CRTC has
@@ -181,7 +221,7 @@ static const struct of_device_id sun8i_dw_hdmi_dt_ids[] = {
 };
 MODULE_DEVICE_TABLE(of, sun8i_dw_hdmi_dt_ids);
 
-struct platform_driver sun8i_dw_hdmi_pltfm_driver = {
+static struct platform_driver sun8i_dw_hdmi_pltfm_driver = {
 	.probe  = sun8i_dw_hdmi_probe,
 	.remove = sun8i_dw_hdmi_remove,
 	.driver = {
diff --git a/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.h b/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.h
index 79154f0f674a..aadbe0a10b0c 100644
--- a/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.h
+++ b/drivers/gpu/drm/sun4i/sun8i_dw_hdmi.h
@@ -98,7 +98,8 @@
 #define SUN8I_HDMI_PHY_PLL_CFG1_LDO2_EN		BIT(29)
 #define SUN8I_HDMI_PHY_PLL_CFG1_LDO1_EN		BIT(28)
 #define SUN8I_HDMI_PHY_PLL_CFG1_HV_IS_33	BIT(27)
-#define SUN8I_HDMI_PHY_PLL_CFG1_CKIN_SEL	BIT(26)
+#define SUN8I_HDMI_PHY_PLL_CFG1_CKIN_SEL_MSK	BIT(26)
+#define SUN8I_HDMI_PHY_PLL_CFG1_CKIN_SEL_SHIFT	26
 #define SUN8I_HDMI_PHY_PLL_CFG1_PLLEN		BIT(25)
 #define SUN8I_HDMI_PHY_PLL_CFG1_LDO_VSET(x)	((x) << 22)
 #define SUN8I_HDMI_PHY_PLL_CFG1_UNKNOWN(x)	((x) << 20)
@@ -147,6 +148,7 @@ struct sun8i_hdmi_phy;
 
 struct sun8i_hdmi_phy_variant {
 	bool has_phy_clk;
+	bool has_second_pll;
 	void (*phy_init)(struct sun8i_hdmi_phy *phy);
 	void (*phy_disable)(struct dw_hdmi *hdmi,
 			    struct sun8i_hdmi_phy *phy);
@@ -160,6 +162,7 @@ struct sun8i_hdmi_phy {
 	struct clk			*clk_mod;
 	struct clk			*clk_phy;
 	struct clk			*clk_pll0;
+	struct clk			*clk_pll1;
 	unsigned int			rcal;
 	struct regmap			*regs;
 	struct reset_control		*rst_phy;
@@ -188,6 +191,7 @@ void sun8i_hdmi_phy_remove(struct sun8i_dw_hdmi *hdmi);
 void sun8i_hdmi_phy_init(struct sun8i_hdmi_phy *phy);
 const struct dw_hdmi_phy_ops *sun8i_hdmi_phy_get_ops(void);
 
-int sun8i_phy_clk_create(struct sun8i_hdmi_phy *phy, struct device *dev);
+int sun8i_phy_clk_create(struct sun8i_hdmi_phy *phy, struct device *dev,
+			 bool second_parent);
 
 #endif /* _SUN8I_DW_HDMI_H_ */
diff --git a/drivers/gpu/drm/sun4i/sun8i_hdmi_phy.c b/drivers/gpu/drm/sun4i/sun8i_hdmi_phy.c
index 5a52fc489a9d..82502b351aec 100644
--- a/drivers/gpu/drm/sun4i/sun8i_hdmi_phy.c
+++ b/drivers/gpu/drm/sun4i/sun8i_hdmi_phy.c
@@ -183,7 +183,13 @@ static int sun8i_hdmi_phy_config_h3(struct dw_hdmi *hdmi,
 	regmap_update_bits(phy->regs, SUN8I_HDMI_PHY_ANA_CFG1_REG,
 			   SUN8I_HDMI_PHY_ANA_CFG1_TXEN_MASK, 0);
 
-	regmap_write(phy->regs, SUN8I_HDMI_PHY_PLL_CFG1_REG, pll_cfg1_init);
+	/*
+	 * NOTE: We have to be careful not to overwrite PHY parent
+	 * clock selection bit and clock divider.
+	 */
+	regmap_update_bits(phy->regs, SUN8I_HDMI_PHY_PLL_CFG1_REG,
+			   (u32)~SUN8I_HDMI_PHY_PLL_CFG1_CKIN_SEL_MSK,
+			   pll_cfg1_init);
 	regmap_update_bits(phy->regs, SUN8I_HDMI_PHY_PLL_CFG2_REG,
 			   (u32)~SUN8I_HDMI_PHY_PLL_CFG2_PREDIV_MSK,
 			   pll_cfg2_init);
@@ -352,6 +358,10 @@ static void sun8i_hdmi_phy_init_h3(struct sun8i_hdmi_phy *phy)
 			   SUN8I_HDMI_PHY_ANA_CFG3_SCLEN |
 			   SUN8I_HDMI_PHY_ANA_CFG3_SDAEN);
 
+	/* reset PHY PLL clock parent */
+	regmap_update_bits(phy->regs, SUN8I_HDMI_PHY_PLL_CFG1_REG,
+			   SUN8I_HDMI_PHY_PLL_CFG1_CKIN_SEL_MSK, 0);
+
 	/* set HW control of CEC pins */
 	regmap_write(phy->regs, SUN8I_HDMI_PHY_CEC_REG, 0);
 
@@ -386,6 +396,14 @@ static struct regmap_config sun8i_hdmi_phy_regmap_config = {
 	.name		= "phy"
 };
 
+static const struct sun8i_hdmi_phy_variant sun50i_a64_hdmi_phy = {
+	.has_phy_clk = true,
+	.has_second_pll = true,
+	.phy_init = &sun8i_hdmi_phy_init_h3,
+	.phy_disable = &sun8i_hdmi_phy_disable_h3,
+	.phy_config = &sun8i_hdmi_phy_config_h3,
+};
+
 static const struct sun8i_hdmi_phy_variant sun8i_a83t_hdmi_phy = {
 	.phy_init = &sun8i_hdmi_phy_init_a83t,
 	.phy_disable = &sun8i_hdmi_phy_disable_a83t,
@@ -401,6 +419,10 @@ static const struct sun8i_hdmi_phy_variant sun8i_h3_hdmi_phy = {
 
 static const struct of_device_id sun8i_hdmi_phy_of_table[] = {
 	{
+		.compatible = "allwinner,sun50i-a64-hdmi-phy",
+		.data = &sun50i_a64_hdmi_phy,
+	},
+	{
 		.compatible = "allwinner,sun8i-a83t-hdmi-phy",
 		.data = &sun8i_a83t_hdmi_phy,
 	},
@@ -472,18 +494,30 @@ int sun8i_hdmi_phy_probe(struct sun8i_dw_hdmi *hdmi, struct device_node *node)
 			goto err_put_clk_mod;
 		}
 
-		ret = sun8i_phy_clk_create(phy, dev);
+		if (phy->variant->has_second_pll) {
+			phy->clk_pll1 = of_clk_get_by_name(node, "pll-1");
+			if (IS_ERR(phy->clk_pll1)) {
+				dev_err(dev, "Could not get pll-1 clock\n");
+				ret = PTR_ERR(phy->clk_pll1);
+				goto err_put_clk_pll0;
+			}
+		}
+
+		ret = sun8i_phy_clk_create(phy, dev,
+					   phy->variant->has_second_pll);
 		if (ret) {
 			dev_err(dev, "Couldn't create the PHY clock\n");
-			goto err_put_clk_pll0;
+			goto err_put_clk_pll1;
 		}
+
+		clk_prepare_enable(phy->clk_phy);
 	}
 
 	phy->rst_phy = of_reset_control_get_shared(node, "phy");
 	if (IS_ERR(phy->rst_phy)) {
 		dev_err(dev, "Could not get phy reset control\n");
 		ret = PTR_ERR(phy->rst_phy);
-		goto err_put_clk_pll0;
+		goto err_disable_clk_phy;
 	}
 
 	ret = reset_control_deassert(phy->rst_phy);
@@ -514,9 +548,12 @@ err_deassert_rst_phy:
 	reset_control_assert(phy->rst_phy);
 err_put_rst_phy:
 	reset_control_put(phy->rst_phy);
+err_disable_clk_phy:
+	clk_disable_unprepare(phy->clk_phy);
+err_put_clk_pll1:
+	clk_put(phy->clk_pll1);
 err_put_clk_pll0:
-	if (phy->variant->has_phy_clk)
-		clk_put(phy->clk_pll0);
+	clk_put(phy->clk_pll0);
 err_put_clk_mod:
 	clk_put(phy->clk_mod);
 err_put_clk_bus:
@@ -531,13 +568,14 @@ void sun8i_hdmi_phy_remove(struct sun8i_dw_hdmi *hdmi)
 
 	clk_disable_unprepare(phy->clk_mod);
 	clk_disable_unprepare(phy->clk_bus);
+	clk_disable_unprepare(phy->clk_phy);
 
 	reset_control_assert(phy->rst_phy);
 
 	reset_control_put(phy->rst_phy);
 
-	if (phy->variant->has_phy_clk)
-		clk_put(phy->clk_pll0);
+	clk_put(phy->clk_pll0);
+	clk_put(phy->clk_pll1);
 	clk_put(phy->clk_mod);
 	clk_put(phy->clk_bus);
 }
diff --git a/drivers/gpu/drm/sun4i/sun8i_hdmi_phy_clk.c b/drivers/gpu/drm/sun4i/sun8i_hdmi_phy_clk.c
index faea449812f8..a4d31fe3abff 100644
--- a/drivers/gpu/drm/sun4i/sun8i_hdmi_phy_clk.c
+++ b/drivers/gpu/drm/sun4i/sun8i_hdmi_phy_clk.c
@@ -22,35 +22,45 @@ static int sun8i_phy_clk_determine_rate(struct clk_hw *hw,
 {
 	unsigned long rate = req->rate;
 	unsigned long best_rate = 0;
+	struct clk_hw *best_parent = NULL;
 	struct clk_hw *parent;
 	int best_div = 1;
-	int i;
+	int i, p;
 
-	parent = clk_hw_get_parent(hw);
-
-	for (i = 1; i <= 16; i++) {
-		unsigned long ideal = rate * i;
-		unsigned long rounded;
-
-		rounded = clk_hw_round_rate(parent, ideal);
+	for (p = 0; p < clk_hw_get_num_parents(hw); p++) {
+		parent = clk_hw_get_parent_by_index(hw, p);
+		if (!parent)
+			continue;
 
-		if (rounded == ideal) {
-			best_rate = rounded;
-			best_div = i;
-			break;
+		for (i = 1; i <= 16; i++) {
+			unsigned long ideal = rate * i;
+			unsigned long rounded;
+
+			rounded = clk_hw_round_rate(parent, ideal);
+
+			if (rounded == ideal) {
+				best_rate = rounded;
+				best_div = i;
+				best_parent = parent;
+				break;
+			}
+
+			if (!best_rate ||
+			    abs(rate - rounded / i) <
+			    abs(rate - best_rate / best_div)) {
+				best_rate = rounded;
+				best_div = i;
+				best_parent = parent;
+			}
 		}
 
-		if (!best_rate ||
-		    abs(rate - rounded / i) <
-		    abs(rate - best_rate / best_div)) {
-			best_rate = rounded;
-			best_div = i;
-		}
+		if (best_rate / best_div == rate)
+			break;
 	}
 
 	req->rate = best_rate / best_div;
 	req->best_parent_rate = best_rate;
-	req->best_parent_hw = parent;
+	req->best_parent_hw = best_parent;
 
 	return 0;
 }
@@ -95,22 +105,58 @@ static int sun8i_phy_clk_set_rate(struct clk_hw *hw, unsigned long rate,
 	return 0;
 }
 
+static u8 sun8i_phy_clk_get_parent(struct clk_hw *hw)
+{
+	struct sun8i_phy_clk *priv = hw_to_phy_clk(hw);
+	u32 reg;
+
+	regmap_read(priv->phy->regs, SUN8I_HDMI_PHY_PLL_CFG1_REG, &reg);
+	reg = (reg & SUN8I_HDMI_PHY_PLL_CFG1_CKIN_SEL_MSK) >>
+	      SUN8I_HDMI_PHY_PLL_CFG1_CKIN_SEL_SHIFT;
+
+	return reg;
+}
+
+static int sun8i_phy_clk_set_parent(struct clk_hw *hw, u8 index)
+{
+	struct sun8i_phy_clk *priv = hw_to_phy_clk(hw);
+
+	if (index > 1)
+		return -EINVAL;
+
+	regmap_update_bits(priv->phy->regs, SUN8I_HDMI_PHY_PLL_CFG1_REG,
+			   SUN8I_HDMI_PHY_PLL_CFG1_CKIN_SEL_MSK,
+			   index << SUN8I_HDMI_PHY_PLL_CFG1_CKIN_SEL_SHIFT);
+
+	return 0;
+}
+
 static const struct clk_ops sun8i_phy_clk_ops = {
 	.determine_rate	= sun8i_phy_clk_determine_rate,
 	.recalc_rate	= sun8i_phy_clk_recalc_rate,
 	.set_rate	= sun8i_phy_clk_set_rate,
+
+	.get_parent	= sun8i_phy_clk_get_parent,
+	.set_parent	= sun8i_phy_clk_set_parent,
 };
 
-int sun8i_phy_clk_create(struct sun8i_hdmi_phy *phy, struct device *dev)
+int sun8i_phy_clk_create(struct sun8i_hdmi_phy *phy, struct device *dev,
+			 bool second_parent)
 {
 	struct clk_init_data init;
 	struct sun8i_phy_clk *priv;
-	const char *parents[1];
+	const char *parents[2];
 
 	parents[0] = __clk_get_name(phy->clk_pll0);
 	if (!parents[0])
 		return -ENODEV;
 
+	if (second_parent) {
+		parents[1] = __clk_get_name(phy->clk_pll1);
+		if (!parents[1])
+			return -ENODEV;
+	}
+
 	priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
 	if (!priv)
 		return -ENOMEM;
@@ -118,7 +164,7 @@ int sun8i_phy_clk_create(struct sun8i_hdmi_phy *phy, struct device *dev)
 	init.name = "hdmi-phy-clk";
 	init.ops = &sun8i_phy_clk_ops;
 	init.parent_names = parents;
-	init.num_parents = 1;
+	init.num_parents = second_parent ? 2 : 1;
 	init.flags = CLK_SET_RATE_PARENT;
 
 	priv->phy = phy;
diff --git a/drivers/gpu/drm/sun4i/sun8i_mixer.c b/drivers/gpu/drm/sun4i/sun8i_mixer.c
index 126899d6f0d3..fc3713608f78 100644
--- a/drivers/gpu/drm/sun4i/sun8i_mixer.c
+++ b/drivers/gpu/drm/sun4i/sun8i_mixer.c
@@ -21,8 +21,9 @@
 
 #include <linux/component.h>
 #include <linux/dma-mapping.h>
-#include <linux/reset.h>
 #include <linux/of_device.h>
+#include <linux/of_graph.h>
+#include <linux/reset.h>
 
 #include "sun4i_drv.h"
 #include "sun8i_mixer.h"
@@ -322,6 +323,42 @@ static struct regmap_config sun8i_mixer_regmap_config = {
 	.max_register	= 0xbfffc, /* guessed */
 };
 
+static int sun8i_mixer_of_get_id(struct device_node *node)
+{
+	struct device_node *port, *ep;
+	int ret = -EINVAL;
+
+	/* output is port 1 */
+	port = of_graph_get_port_by_id(node, 1);
+	if (!port)
+		return -EINVAL;
+
+	/* try to find downstream endpoint */
+	for_each_available_child_of_node(port, ep) {
+		struct device_node *remote;
+		u32 reg;
+
+		remote = of_graph_get_remote_endpoint(ep);
+		if (!remote)
+			continue;
+
+		ret = of_property_read_u32(remote, "reg", &reg);
+		if (!ret) {
+			of_node_put(remote);
+			of_node_put(ep);
+			of_node_put(port);
+
+			return reg;
+		}
+
+		of_node_put(remote);
+	}
+
+	of_node_put(port);
+
+	return ret;
+}
+
 static int sun8i_mixer_bind(struct device *dev, struct device *master,
 			      void *data)
 {
@@ -353,8 +390,16 @@ static int sun8i_mixer_bind(struct device *dev, struct device *master,
 	dev_set_drvdata(dev, mixer);
 	mixer->engine.ops = &sun8i_engine_ops;
 	mixer->engine.node = dev->of_node;
-	/* The ID of the mixer currently doesn't matter */
-	mixer->engine.id = -1;
+
+	/*
+	 * While this function can fail, we shouldn't do anything
+	 * if this happens. Some early DE2 DT entries don't provide
+	 * mixer id but work nevertheless because matching between
+	 * TCON and mixer is done by comparing node pointers (old
+	 * way) instead comparing ids. If this function fails and
+	 * id is needed, it will fail during id matching anyway.
+	 */
+	mixer->engine.id = sun8i_mixer_of_get_id(dev->of_node);
 
 	mixer->cfg = of_device_get_match_data(dev);
 	if (!mixer->cfg)
@@ -432,14 +477,14 @@ static int sun8i_mixer_bind(struct device *dev, struct device *master,
 	regmap_write(mixer->engine.regs, SUN8I_MIXER_BLEND_ATTR_FCOLOR(0),
 		     SUN8I_MIXER_BLEND_COLOR_BLACK);
 
-	/* Fixed zpos for now */
-	regmap_write(mixer->engine.regs, SUN8I_MIXER_BLEND_ROUTE, 0x43210);
-
 	plane_cnt = mixer->cfg->vi_num + mixer->cfg->ui_num;
 	for (i = 0; i < plane_cnt; i++)
 		regmap_write(mixer->engine.regs, SUN8I_MIXER_BLEND_MODE(i),
 			     SUN8I_MIXER_BLEND_MODE_DEF);
 
+	regmap_update_bits(mixer->engine.regs, SUN8I_MIXER_BLEND_PIPE_CTL,
+			   SUN8I_MIXER_BLEND_PIPE_CTL_EN_MSK, 0);
+
 	return 0;
 
 err_disable_bus_clk:
@@ -500,6 +545,22 @@ static const struct sun8i_mixer_cfg sun8i_h3_mixer0_cfg = {
 	.vi_num		= 1,
 };
 
+static const struct sun8i_mixer_cfg sun8i_r40_mixer0_cfg = {
+	.ccsc		= 0,
+	.mod_rate	= 297000000,
+	.scaler_mask	= 0xf,
+	.ui_num		= 3,
+	.vi_num		= 1,
+};
+
+static const struct sun8i_mixer_cfg sun8i_r40_mixer1_cfg = {
+	.ccsc		= 1,
+	.mod_rate	= 297000000,
+	.scaler_mask	= 0x3,
+	.ui_num		= 1,
+	.vi_num		= 1,
+};
+
 static const struct sun8i_mixer_cfg sun8i_v3s_mixer_cfg = {
 	.vi_num = 2,
 	.ui_num = 1,
@@ -522,6 +583,14 @@ static const struct of_device_id sun8i_mixer_of_table[] = {
 		.data = &sun8i_h3_mixer0_cfg,
 	},
 	{
+		.compatible = "allwinner,sun8i-r40-de2-mixer-0",
+		.data = &sun8i_r40_mixer0_cfg,
+	},
+	{
+		.compatible = "allwinner,sun8i-r40-de2-mixer-1",
+		.data = &sun8i_r40_mixer1_cfg,
+	},
+	{
 		.compatible = "allwinner,sun8i-v3s-de2-mixer",
 		.data = &sun8i_v3s_mixer_cfg,
 	},
diff --git a/drivers/gpu/drm/sun4i/sun8i_mixer.h b/drivers/gpu/drm/sun4i/sun8i_mixer.h
index f34e70c42adf..406c42e752d7 100644
--- a/drivers/gpu/drm/sun4i/sun8i_mixer.h
+++ b/drivers/gpu/drm/sun4i/sun8i_mixer.h
@@ -44,6 +44,7 @@
 #define SUN8I_MIXER_BLEND_CK_MIN(x)		(0x10e0 + 0x04 * (x))
 #define SUN8I_MIXER_BLEND_OUTCTL		0x10fc
 
+#define SUN8I_MIXER_BLEND_PIPE_CTL_EN_MSK	GENMASK(12, 8)
 #define SUN8I_MIXER_BLEND_PIPE_CTL_EN(pipe)	BIT(8 + pipe)
 #define SUN8I_MIXER_BLEND_PIPE_CTL_FC_EN(pipe)	BIT(pipe)
 /* colors are always in AARRGGBB format */
@@ -51,6 +52,9 @@
 /* The following numbers are some still unknown magic numbers */
 #define SUN8I_MIXER_BLEND_MODE_DEF		0x03010301
 
+#define SUN8I_MIXER_BLEND_ROUTE_PIPE_MSK(n)	(0xf << ((n) << 2))
+#define SUN8I_MIXER_BLEND_ROUTE_PIPE_SHIFT(n)	((n) << 2)
+
 #define SUN8I_MIXER_BLEND_OUTCTL_INTERLACED	BIT(1)
 
 #define SUN8I_MIXER_FBFMT_ARGB8888	0
diff --git a/drivers/gpu/drm/sun4i/sun8i_tcon_top.c b/drivers/gpu/drm/sun4i/sun8i_tcon_top.c
new file mode 100644
index 000000000000..55fe398d8290
--- /dev/null
+++ b/drivers/gpu/drm/sun4i/sun8i_tcon_top.c
@@ -0,0 +1,274 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* Copyright (c) 2018 Jernej Skrabec <jernej.skrabec@siol.net> */
+
+#include <drm/drmP.h>
+
+#include <dt-bindings/clock/sun8i-tcon-top.h>
+
+#include <linux/bitfield.h>
+#include <linux/component.h>
+#include <linux/device.h>
+#include <linux/module.h>
+#include <linux/of_graph.h>
+#include <linux/platform_device.h>
+
+#include "sun8i_tcon_top.h"
+
+static bool sun8i_tcon_top_node_is_tcon_top(struct device_node *node)
+{
+	return !!of_match_node(sun8i_tcon_top_of_table, node);
+}
+
+int sun8i_tcon_top_set_hdmi_src(struct device *dev, int tcon)
+{
+	struct sun8i_tcon_top *tcon_top = dev_get_drvdata(dev);
+	unsigned long flags;
+	u32 val;
+
+	if (!sun8i_tcon_top_node_is_tcon_top(dev->of_node)) {
+		dev_err(dev, "Device is not TCON TOP!\n");
+		return -EINVAL;
+	}
+
+	if (tcon < 2 || tcon > 3) {
+		dev_err(dev, "TCON index must be 2 or 3!\n");
+		return -EINVAL;
+	}
+
+	spin_lock_irqsave(&tcon_top->reg_lock, flags);
+
+	val = readl(tcon_top->regs + TCON_TOP_GATE_SRC_REG);
+	val &= ~TCON_TOP_HDMI_SRC_MSK;
+	val |= FIELD_PREP(TCON_TOP_HDMI_SRC_MSK, tcon - 1);
+	writel(val, tcon_top->regs + TCON_TOP_GATE_SRC_REG);
+
+	spin_unlock_irqrestore(&tcon_top->reg_lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL(sun8i_tcon_top_set_hdmi_src);
+
+int sun8i_tcon_top_de_config(struct device *dev, int mixer, int tcon)
+{
+	struct sun8i_tcon_top *tcon_top = dev_get_drvdata(dev);
+	unsigned long flags;
+	u32 reg;
+
+	if (!sun8i_tcon_top_node_is_tcon_top(dev->of_node)) {
+		dev_err(dev, "Device is not TCON TOP!\n");
+		return -EINVAL;
+	}
+
+	if (mixer > 1) {
+		dev_err(dev, "Mixer index is too high!\n");
+		return -EINVAL;
+	}
+
+	if (tcon > 3) {
+		dev_err(dev, "TCON index is too high!\n");
+		return -EINVAL;
+	}
+
+	spin_lock_irqsave(&tcon_top->reg_lock, flags);
+
+	reg = readl(tcon_top->regs + TCON_TOP_PORT_SEL_REG);
+	if (mixer == 0) {
+		reg &= ~TCON_TOP_PORT_DE0_MSK;
+		reg |= FIELD_PREP(TCON_TOP_PORT_DE0_MSK, tcon);
+	} else {
+		reg &= ~TCON_TOP_PORT_DE1_MSK;
+		reg |= FIELD_PREP(TCON_TOP_PORT_DE1_MSK, tcon);
+	}
+	writel(reg, tcon_top->regs + TCON_TOP_PORT_SEL_REG);
+
+	spin_unlock_irqrestore(&tcon_top->reg_lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL(sun8i_tcon_top_de_config);
+
+
+static struct clk_hw *sun8i_tcon_top_register_gate(struct device *dev,
+						   const char *parent,
+						   void __iomem *regs,
+						   spinlock_t *lock,
+						   u8 bit, int name_index)
+{
+	const char *clk_name, *parent_name;
+	int ret, index;
+
+	index = of_property_match_string(dev->of_node, "clock-names", parent);
+	if (index < 0)
+		return ERR_PTR(index);
+
+	parent_name = of_clk_get_parent_name(dev->of_node, index);
+
+	ret = of_property_read_string_index(dev->of_node,
+					    "clock-output-names", name_index,
+					    &clk_name);
+	if (ret)
+		return ERR_PTR(ret);
+
+	return clk_hw_register_gate(dev, clk_name, parent_name,
+				    CLK_SET_RATE_PARENT,
+				    regs + TCON_TOP_GATE_SRC_REG,
+				    bit, 0, lock);
+};
+
+static int sun8i_tcon_top_bind(struct device *dev, struct device *master,
+			       void *data)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct clk_hw_onecell_data *clk_data;
+	struct sun8i_tcon_top *tcon_top;
+	struct resource *res;
+	void __iomem *regs;
+	int ret, i;
+
+	tcon_top = devm_kzalloc(dev, sizeof(*tcon_top), GFP_KERNEL);
+	if (!tcon_top)
+		return -ENOMEM;
+
+	clk_data = devm_kzalloc(dev, sizeof(*clk_data) +
+				sizeof(*clk_data->hws) * CLK_NUM,
+				GFP_KERNEL);
+	if (!clk_data)
+		return -ENOMEM;
+	tcon_top->clk_data = clk_data;
+
+	spin_lock_init(&tcon_top->reg_lock);
+
+	tcon_top->rst = devm_reset_control_get(dev, NULL);
+	if (IS_ERR(tcon_top->rst)) {
+		dev_err(dev, "Couldn't get our reset line\n");
+		return PTR_ERR(tcon_top->rst);
+	}
+
+	tcon_top->bus = devm_clk_get(dev, "bus");
+	if (IS_ERR(tcon_top->bus)) {
+		dev_err(dev, "Couldn't get the bus clock\n");
+		return PTR_ERR(tcon_top->bus);
+	}
+
+	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	regs = devm_ioremap_resource(dev, res);
+	tcon_top->regs = regs;
+	if (IS_ERR(regs))
+		return PTR_ERR(regs);
+
+	ret = reset_control_deassert(tcon_top->rst);
+	if (ret) {
+		dev_err(dev, "Could not deassert ctrl reset control\n");
+		return ret;
+	}
+
+	ret = clk_prepare_enable(tcon_top->bus);
+	if (ret) {
+		dev_err(dev, "Could not enable bus clock\n");
+		goto err_assert_reset;
+	}
+
+	/*
+	 * TCON TOP has two muxes, which select parent clock for each TCON TV
+	 * channel clock. Parent could be either TCON TV or TVE clock. For now
+	 * we leave this fixed to TCON TV, since TVE driver for R40 is not yet
+	 * implemented. Once it is, graph needs to be traversed to determine
+	 * if TVE is active on each TCON TV. If it is, mux should be switched
+	 * to TVE clock parent.
+	 */
+	clk_data->hws[CLK_TCON_TOP_TV0] =
+		sun8i_tcon_top_register_gate(dev, "tcon-tv0", regs,
+					     &tcon_top->reg_lock,
+					     TCON_TOP_TCON_TV0_GATE, 0);
+
+	clk_data->hws[CLK_TCON_TOP_TV1] =
+		sun8i_tcon_top_register_gate(dev, "tcon-tv1", regs,
+					     &tcon_top->reg_lock,
+					     TCON_TOP_TCON_TV1_GATE, 1);
+
+	clk_data->hws[CLK_TCON_TOP_DSI] =
+		sun8i_tcon_top_register_gate(dev, "dsi", regs,
+					     &tcon_top->reg_lock,
+					     TCON_TOP_TCON_DSI_GATE, 2);
+
+	for (i = 0; i < CLK_NUM; i++)
+		if (IS_ERR(clk_data->hws[i])) {
+			ret = PTR_ERR(clk_data->hws[i]);
+			goto err_unregister_gates;
+		}
+
+	clk_data->num = CLK_NUM;
+
+	ret = of_clk_add_hw_provider(dev->of_node, of_clk_hw_onecell_get,
+				     clk_data);
+	if (ret)
+		goto err_unregister_gates;
+
+	dev_set_drvdata(dev, tcon_top);
+
+	return 0;
+
+err_unregister_gates:
+	for (i = 0; i < CLK_NUM; i++)
+		if (clk_data->hws[i])
+			clk_hw_unregister_gate(clk_data->hws[i]);
+	clk_disable_unprepare(tcon_top->bus);
+err_assert_reset:
+	reset_control_assert(tcon_top->rst);
+
+	return ret;
+}
+
+static void sun8i_tcon_top_unbind(struct device *dev, struct device *master,
+				  void *data)
+{
+	struct sun8i_tcon_top *tcon_top = dev_get_drvdata(dev);
+	struct clk_hw_onecell_data *clk_data = tcon_top->clk_data;
+	int i;
+
+	of_clk_del_provider(dev->of_node);
+	for (i = 0; i < CLK_NUM; i++)
+		clk_hw_unregister_gate(clk_data->hws[i]);
+
+	clk_disable_unprepare(tcon_top->bus);
+	reset_control_assert(tcon_top->rst);
+}
+
+static const struct component_ops sun8i_tcon_top_ops = {
+	.bind	= sun8i_tcon_top_bind,
+	.unbind	= sun8i_tcon_top_unbind,
+};
+
+static int sun8i_tcon_top_probe(struct platform_device *pdev)
+{
+	return component_add(&pdev->dev, &sun8i_tcon_top_ops);
+}
+
+static int sun8i_tcon_top_remove(struct platform_device *pdev)
+{
+	component_del(&pdev->dev, &sun8i_tcon_top_ops);
+
+	return 0;
+}
+
+/* sun4i_drv uses this list to check if a device node is a TCON TOP */
+const struct of_device_id sun8i_tcon_top_of_table[] = {
+	{ .compatible = "allwinner,sun8i-r40-tcon-top" },
+	{ /* sentinel */ }
+};
+MODULE_DEVICE_TABLE(of, sun8i_tcon_top_of_table);
+EXPORT_SYMBOL(sun8i_tcon_top_of_table);
+
+static struct platform_driver sun8i_tcon_top_platform_driver = {
+	.probe		= sun8i_tcon_top_probe,
+	.remove		= sun8i_tcon_top_remove,
+	.driver		= {
+		.name		= "sun8i-tcon-top",
+		.of_match_table	= sun8i_tcon_top_of_table,
+	},
+};
+module_platform_driver(sun8i_tcon_top_platform_driver);
+
+MODULE_AUTHOR("Jernej Skrabec <jernej.skrabec@siol.net>");
+MODULE_DESCRIPTION("Allwinner R40 TCON TOP driver");
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/sun4i/sun8i_tcon_top.h b/drivers/gpu/drm/sun4i/sun8i_tcon_top.h
new file mode 100644
index 000000000000..0390584a330e
--- /dev/null
+++ b/drivers/gpu/drm/sun4i/sun8i_tcon_top.h
@@ -0,0 +1,44 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+/* Copyright (c) 2018 Jernej Skrabec <jernej.skrabec@siol.net> */
+
+#ifndef _SUN8I_TCON_TOP_H_
+#define _SUN8I_TCON_TOP_H_
+
+#include <linux/clk.h>
+#include <linux/clk-provider.h>
+#include <linux/reset.h>
+#include <linux/spinlock.h>
+
+#define TCON_TOP_TCON_TV_SETUP_REG	0x00
+
+#define TCON_TOP_PORT_SEL_REG		0x1C
+#define TCON_TOP_PORT_DE0_MSK			GENMASK(1, 0)
+#define TCON_TOP_PORT_DE1_MSK			GENMASK(5, 4)
+
+#define TCON_TOP_GATE_SRC_REG		0x20
+#define TCON_TOP_HDMI_SRC_MSK			GENMASK(29, 28)
+#define TCON_TOP_TCON_TV1_GATE			24
+#define TCON_TOP_TCON_TV0_GATE			20
+#define TCON_TOP_TCON_DSI_GATE			16
+
+#define CLK_NUM					3
+
+struct sun8i_tcon_top {
+	struct clk			*bus;
+	struct clk_hw_onecell_data	*clk_data;
+	void __iomem			*regs;
+	struct reset_control		*rst;
+
+	/*
+	 * spinlock is used to synchronize access to same
+	 * register where multiple clock gates can be set.
+	 */
+	spinlock_t			reg_lock;
+};
+
+extern const struct of_device_id sun8i_tcon_top_of_table[];
+
+int sun8i_tcon_top_set_hdmi_src(struct device *dev, int tcon);
+int sun8i_tcon_top_de_config(struct device *dev, int mixer, int tcon);
+
+#endif /* _SUN8I_TCON_TOP_H_ */
diff --git a/drivers/gpu/drm/sun4i/sun8i_ui_layer.c b/drivers/gpu/drm/sun4i/sun8i_ui_layer.c
index 9a540330cb79..28c15c6ef1ef 100644
--- a/drivers/gpu/drm/sun4i/sun8i_ui_layer.c
+++ b/drivers/gpu/drm/sun4i/sun8i_ui_layer.c
@@ -27,7 +27,8 @@
 #include "sun8i_ui_scaler.h"
 
 static void sun8i_ui_layer_enable(struct sun8i_mixer *mixer, int channel,
-				  int overlay, bool enable)
+				  int overlay, bool enable, unsigned int zpos,
+				  unsigned int old_zpos)
 {
 	u32 val;
 
@@ -43,18 +44,36 @@ static void sun8i_ui_layer_enable(struct sun8i_mixer *mixer, int channel,
 			   SUN8I_MIXER_CHAN_UI_LAYER_ATTR(channel, overlay),
 			   SUN8I_MIXER_CHAN_UI_LAYER_ATTR_EN, val);
 
-	if (enable)
-		val = SUN8I_MIXER_BLEND_PIPE_CTL_EN(channel);
-	else
-		val = 0;
+	if (!enable || zpos != old_zpos) {
+		regmap_update_bits(mixer->engine.regs,
+				   SUN8I_MIXER_BLEND_PIPE_CTL,
+				   SUN8I_MIXER_BLEND_PIPE_CTL_EN(old_zpos),
+				   0);
 
-	regmap_update_bits(mixer->engine.regs,
-			   SUN8I_MIXER_BLEND_PIPE_CTL,
-			   SUN8I_MIXER_BLEND_PIPE_CTL_EN(channel), val);
+		regmap_update_bits(mixer->engine.regs,
+				   SUN8I_MIXER_BLEND_ROUTE,
+				   SUN8I_MIXER_BLEND_ROUTE_PIPE_MSK(old_zpos),
+				   0);
+	}
+
+	if (enable) {
+		val = SUN8I_MIXER_BLEND_PIPE_CTL_EN(zpos);
+
+		regmap_update_bits(mixer->engine.regs,
+				   SUN8I_MIXER_BLEND_PIPE_CTL, val, val);
+
+		val = channel << SUN8I_MIXER_BLEND_ROUTE_PIPE_SHIFT(zpos);
+
+		regmap_update_bits(mixer->engine.regs,
+				   SUN8I_MIXER_BLEND_ROUTE,
+				   SUN8I_MIXER_BLEND_ROUTE_PIPE_MSK(zpos),
+				   val);
+	}
 }
 
 static int sun8i_ui_layer_update_coord(struct sun8i_mixer *mixer, int channel,
-				       int overlay, struct drm_plane *plane)
+				       int overlay, struct drm_plane *plane,
+				       unsigned int zpos)
 {
 	struct drm_plane_state *state = plane->state;
 	u32 src_w, src_h, dst_w, dst_h;
@@ -137,10 +156,10 @@ static int sun8i_ui_layer_update_coord(struct sun8i_mixer *mixer, int channel,
 			 state->dst.x1, state->dst.y1);
 	DRM_DEBUG_DRIVER("Layer destination size W: %d H: %d\n", dst_w, dst_h);
 	regmap_write(mixer->engine.regs,
-		     SUN8I_MIXER_BLEND_ATTR_COORD(channel),
+		     SUN8I_MIXER_BLEND_ATTR_COORD(zpos),
 		     SUN8I_MIXER_COORD(state->dst.x1, state->dst.y1));
 	regmap_write(mixer->engine.regs,
-		     SUN8I_MIXER_BLEND_ATTR_INSIZE(channel),
+		     SUN8I_MIXER_BLEND_ATTR_INSIZE(zpos),
 		     outsize);
 
 	return 0;
@@ -236,30 +255,35 @@ static void sun8i_ui_layer_atomic_disable(struct drm_plane *plane,
 					  struct drm_plane_state *old_state)
 {
 	struct sun8i_ui_layer *layer = plane_to_sun8i_ui_layer(plane);
+	unsigned int old_zpos = old_state->normalized_zpos;
 	struct sun8i_mixer *mixer = layer->mixer;
 
-	sun8i_ui_layer_enable(mixer, layer->channel, layer->overlay, false);
+	sun8i_ui_layer_enable(mixer, layer->channel, layer->overlay, false, 0,
+			      old_zpos);
 }
 
 static void sun8i_ui_layer_atomic_update(struct drm_plane *plane,
 					 struct drm_plane_state *old_state)
 {
 	struct sun8i_ui_layer *layer = plane_to_sun8i_ui_layer(plane);
+	unsigned int zpos = plane->state->normalized_zpos;
+	unsigned int old_zpos = old_state->normalized_zpos;
 	struct sun8i_mixer *mixer = layer->mixer;
 
 	if (!plane->state->visible) {
 		sun8i_ui_layer_enable(mixer, layer->channel,
-				      layer->overlay, false);
+				      layer->overlay, false, 0, old_zpos);
 		return;
 	}
 
 	sun8i_ui_layer_update_coord(mixer, layer->channel,
-				    layer->overlay, plane);
+				    layer->overlay, plane, zpos);
 	sun8i_ui_layer_update_formats(mixer, layer->channel,
 				      layer->overlay, plane);
 	sun8i_ui_layer_update_buffer(mixer, layer->channel,
 				     layer->overlay, plane);
-	sun8i_ui_layer_enable(mixer, layer->channel, layer->overlay, true);
+	sun8i_ui_layer_enable(mixer, layer->channel, layer->overlay,
+			      true, zpos, old_zpos);
 }
 
 static struct drm_plane_helper_funcs sun8i_ui_layer_helper_funcs = {
@@ -307,6 +331,7 @@ struct sun8i_ui_layer *sun8i_ui_layer_init_one(struct drm_device *drm,
 	enum drm_plane_type type = DRM_PLANE_TYPE_OVERLAY;
 	int channel = mixer->cfg->vi_num + index;
 	struct sun8i_ui_layer *layer;
+	unsigned int plane_cnt;
 	int ret;
 
 	layer = devm_kzalloc(drm->dev, sizeof(*layer), GFP_KERNEL);
@@ -327,8 +352,10 @@ struct sun8i_ui_layer *sun8i_ui_layer_init_one(struct drm_device *drm,
 		return ERR_PTR(ret);
 	}
 
-	/* fixed zpos for now */
-	ret = drm_plane_create_zpos_immutable_property(&layer->plane, channel);
+	plane_cnt = mixer->cfg->ui_num + mixer->cfg->vi_num;
+
+	ret = drm_plane_create_zpos_property(&layer->plane, channel,
+					     0, plane_cnt - 1);
 	if (ret) {
 		dev_err(drm->dev, "Couldn't add zpos property\n");
 		return ERR_PTR(ret);
diff --git a/drivers/gpu/drm/sun4i/sun8i_vi_layer.c b/drivers/gpu/drm/sun4i/sun8i_vi_layer.c
index 5877f8ef5895..f4fe97813f94 100644
--- a/drivers/gpu/drm/sun4i/sun8i_vi_layer.c
+++ b/drivers/gpu/drm/sun4i/sun8i_vi_layer.c
@@ -21,7 +21,8 @@
 #include "sun8i_vi_scaler.h"
 
 static void sun8i_vi_layer_enable(struct sun8i_mixer *mixer, int channel,
-				  int overlay, bool enable)
+				  int overlay, bool enable, unsigned int zpos,
+				  unsigned int old_zpos)
 {
 	u32 val;
 
@@ -37,18 +38,36 @@ static void sun8i_vi_layer_enable(struct sun8i_mixer *mixer, int channel,
 			   SUN8I_MIXER_CHAN_VI_LAYER_ATTR(channel, overlay),
 			   SUN8I_MIXER_CHAN_VI_LAYER_ATTR_EN, val);
 
-	if (enable)
-		val = SUN8I_MIXER_BLEND_PIPE_CTL_EN(channel);
-	else
-		val = 0;
+	if (!enable || zpos != old_zpos) {
+		regmap_update_bits(mixer->engine.regs,
+				   SUN8I_MIXER_BLEND_PIPE_CTL,
+				   SUN8I_MIXER_BLEND_PIPE_CTL_EN(old_zpos),
+				   0);
 
-	regmap_update_bits(mixer->engine.regs,
-			   SUN8I_MIXER_BLEND_PIPE_CTL,
-			   SUN8I_MIXER_BLEND_PIPE_CTL_EN(channel), val);
+		regmap_update_bits(mixer->engine.regs,
+				   SUN8I_MIXER_BLEND_ROUTE,
+				   SUN8I_MIXER_BLEND_ROUTE_PIPE_MSK(old_zpos),
+				   0);
+	}
+
+	if (enable) {
+		val = SUN8I_MIXER_BLEND_PIPE_CTL_EN(zpos);
+
+		regmap_update_bits(mixer->engine.regs,
+				   SUN8I_MIXER_BLEND_PIPE_CTL, val, val);
+
+		val = channel << SUN8I_MIXER_BLEND_ROUTE_PIPE_SHIFT(zpos);
+
+		regmap_update_bits(mixer->engine.regs,
+				   SUN8I_MIXER_BLEND_ROUTE,
+				   SUN8I_MIXER_BLEND_ROUTE_PIPE_MSK(zpos),
+				   val);
+	}
 }
 
 static int sun8i_vi_layer_update_coord(struct sun8i_mixer *mixer, int channel,
-				       int overlay, struct drm_plane *plane)
+				       int overlay, struct drm_plane *plane,
+				       unsigned int zpos)
 {
 	struct drm_plane_state *state = plane->state;
 	const struct drm_format_info *format = state->fb->format;
@@ -130,10 +149,10 @@ static int sun8i_vi_layer_update_coord(struct sun8i_mixer *mixer, int channel,
 			 state->dst.x1, state->dst.y1);
 	DRM_DEBUG_DRIVER("Layer destination size W: %d H: %d\n", dst_w, dst_h);
 	regmap_write(mixer->engine.regs,
-		     SUN8I_MIXER_BLEND_ATTR_COORD(channel),
+		     SUN8I_MIXER_BLEND_ATTR_COORD(zpos),
 		     SUN8I_MIXER_COORD(state->dst.x1, state->dst.y1));
 	regmap_write(mixer->engine.regs,
-		     SUN8I_MIXER_BLEND_ATTR_INSIZE(channel),
+		     SUN8I_MIXER_BLEND_ATTR_INSIZE(zpos),
 		     outsize);
 
 	return 0;
@@ -264,30 +283,35 @@ static void sun8i_vi_layer_atomic_disable(struct drm_plane *plane,
 					  struct drm_plane_state *old_state)
 {
 	struct sun8i_vi_layer *layer = plane_to_sun8i_vi_layer(plane);
+	unsigned int old_zpos = old_state->normalized_zpos;
 	struct sun8i_mixer *mixer = layer->mixer;
 
-	sun8i_vi_layer_enable(mixer, layer->channel, layer->overlay, false);
+	sun8i_vi_layer_enable(mixer, layer->channel, layer->overlay, false, 0,
+			      old_zpos);
 }
 
 static void sun8i_vi_layer_atomic_update(struct drm_plane *plane,
 					 struct drm_plane_state *old_state)
 {
 	struct sun8i_vi_layer *layer = plane_to_sun8i_vi_layer(plane);
+	unsigned int zpos = plane->state->normalized_zpos;
+	unsigned int old_zpos = old_state->normalized_zpos;
 	struct sun8i_mixer *mixer = layer->mixer;
 
 	if (!plane->state->visible) {
 		sun8i_vi_layer_enable(mixer, layer->channel,
-				      layer->overlay, false);
+				      layer->overlay, false, 0, old_zpos);
 		return;
 	}
 
 	sun8i_vi_layer_update_coord(mixer, layer->channel,
-				    layer->overlay, plane);
+				    layer->overlay, plane, zpos);
 	sun8i_vi_layer_update_formats(mixer, layer->channel,
 				      layer->overlay, plane);
 	sun8i_vi_layer_update_buffer(mixer, layer->channel,
 				     layer->overlay, plane);
-	sun8i_vi_layer_enable(mixer, layer->channel, layer->overlay, true);
+	sun8i_vi_layer_enable(mixer, layer->channel, layer->overlay,
+			      true, zpos, old_zpos);
 }
 
 static struct drm_plane_helper_funcs sun8i_vi_layer_helper_funcs = {
@@ -351,6 +375,7 @@ struct sun8i_vi_layer *sun8i_vi_layer_init_one(struct drm_device *drm,
 					       int index)
 {
 	struct sun8i_vi_layer *layer;
+	unsigned int plane_cnt;
 	int ret;
 
 	layer = devm_kzalloc(drm->dev, sizeof(*layer), GFP_KERNEL);
@@ -368,8 +393,10 @@ struct sun8i_vi_layer *sun8i_vi_layer_init_one(struct drm_device *drm,
 		return ERR_PTR(ret);
 	}
 
-	/* fixed zpos for now */
-	ret = drm_plane_create_zpos_immutable_property(&layer->plane, index);
+	plane_cnt = mixer->cfg->ui_num + mixer->cfg->vi_num;
+
+	ret = drm_plane_create_zpos_property(&layer->plane, index,
+					     0, plane_cnt - 1);
 	if (ret) {
 		dev_err(drm->dev, "Couldn't add zpos property\n");
 		return ERR_PTR(ret);
diff --git a/drivers/gpu/drm/tegra/dc.c b/drivers/gpu/drm/tegra/dc.c
index 9f83a65b5ea9..c3afe7b2237e 100644
--- a/drivers/gpu/drm/tegra/dc.c
+++ b/drivers/gpu/drm/tegra/dc.c
@@ -163,28 +163,89 @@ static void tegra_plane_setup_blending_legacy(struct tegra_plane *plane)
 			 BLEND_COLOR_KEY_NONE;
 	u32 blendnokey = BLEND_WEIGHT1(255) | BLEND_WEIGHT0(255);
 	struct tegra_plane_state *state;
+	u32 blending[2];
 	unsigned int i;
 
+	/* disable blending for non-overlapping case */
+	tegra_plane_writel(plane, blendnokey, DC_WIN_BLEND_NOKEY);
+	tegra_plane_writel(plane, foreground, DC_WIN_BLEND_1WIN);
+
 	state = to_tegra_plane_state(plane->base.state);
 
-	/* alpha contribution is 1 minus sum of overlapping windows */
-	for (i = 0; i < 3; i++) {
-		if (state->dependent[i])
-			background[i] |= BLEND_CONTROL_DEPENDENT;
-	}
+	if (state->opaque) {
+		/*
+		 * Since custom fix-weight blending isn't utilized and weight
+		 * of top window is set to max, we can enforce dependent
+		 * blending which in this case results in transparent bottom
+		 * window if top window is opaque and if top window enables
+		 * alpha blending, then bottom window is getting alpha value
+		 * of 1 minus the sum of alpha components of the overlapping
+		 * plane.
+		 */
+		background[0] |= BLEND_CONTROL_DEPENDENT;
+		background[1] |= BLEND_CONTROL_DEPENDENT;
 
-	/* enable alpha blending if pixel format has an alpha component */
-	if (!state->opaque)
+		/*
+		 * The region where three windows overlap is the intersection
+		 * of the two regions where two windows overlap. It contributes
+		 * to the area if all of the windows on top of it have an alpha
+		 * component.
+		 */
+		switch (state->base.normalized_zpos) {
+		case 0:
+			if (state->blending[0].alpha &&
+			    state->blending[1].alpha)
+				background[2] |= BLEND_CONTROL_DEPENDENT;
+			break;
+
+		case 1:
+			background[2] |= BLEND_CONTROL_DEPENDENT;
+			break;
+		}
+	} else {
+		/*
+		 * Enable alpha blending if pixel format has an alpha
+		 * component.
+		 */
 		foreground |= BLEND_CONTROL_ALPHA;
 
-	/*
-	 * Disable blending and assume Window A is the bottom-most window,
-	 * Window C is the top-most window and Window B is in the middle.
-	 */
-	tegra_plane_writel(plane, blendnokey, DC_WIN_BLEND_NOKEY);
-	tegra_plane_writel(plane, foreground, DC_WIN_BLEND_1WIN);
+		/*
+		 * If any of the windows on top of this window is opaque, it
+		 * will completely conceal this window within that area. If
+		 * top window has an alpha component, it is blended over the
+		 * bottom window.
+		 */
+		for (i = 0; i < 2; i++) {
+			if (state->blending[i].alpha &&
+			    state->blending[i].top)
+				background[i] |= BLEND_CONTROL_DEPENDENT;
+		}
+
+		switch (state->base.normalized_zpos) {
+		case 0:
+			if (state->blending[0].alpha &&
+			    state->blending[1].alpha)
+				background[2] |= BLEND_CONTROL_DEPENDENT;
+			break;
 
-	switch (plane->index) {
+		case 1:
+			/*
+			 * When both middle and topmost windows have an alpha,
+			 * these windows a mixed together and then the result
+			 * is blended over the bottom window.
+			 */
+			if (state->blending[0].alpha &&
+			    state->blending[0].top)
+				background[2] |= BLEND_CONTROL_ALPHA;
+
+			if (state->blending[1].alpha &&
+			    state->blending[1].top)
+				background[2] |= BLEND_CONTROL_ALPHA;
+			break;
+		}
+	}
+
+	switch (state->base.normalized_zpos) {
 	case 0:
 		tegra_plane_writel(plane, background[0], DC_WIN_BLEND_2WIN_X);
 		tegra_plane_writel(plane, background[1], DC_WIN_BLEND_2WIN_Y);
@@ -192,8 +253,21 @@ static void tegra_plane_setup_blending_legacy(struct tegra_plane *plane)
 		break;
 
 	case 1:
-		tegra_plane_writel(plane, foreground, DC_WIN_BLEND_2WIN_X);
-		tegra_plane_writel(plane, background[1], DC_WIN_BLEND_2WIN_Y);
+		/*
+		 * If window B / C is topmost, then X / Y registers are
+		 * matching the order of blending[...] state indices,
+		 * otherwise a swap is required.
+		 */
+		if (!state->blending[0].top && state->blending[1].top) {
+			blending[0] = foreground;
+			blending[1] = background[1];
+		} else {
+			blending[0] = background[0];
+			blending[1] = foreground;
+		}
+
+		tegra_plane_writel(plane, blending[0], DC_WIN_BLEND_2WIN_X);
+		tegra_plane_writel(plane, blending[1], DC_WIN_BLEND_2WIN_Y);
 		tegra_plane_writel(plane, background[2], DC_WIN_BLEND_3WIN_XY);
 		break;
 
@@ -224,6 +298,39 @@ static void tegra_plane_setup_blending(struct tegra_plane *plane,
 	tegra_plane_writel(plane, value, DC_WIN_BLEND_LAYER_CONTROL);
 }
 
+static bool
+tegra_plane_use_horizontal_filtering(struct tegra_plane *plane,
+				     const struct tegra_dc_window *window)
+{
+	struct tegra_dc *dc = plane->dc;
+
+	if (window->src.w == window->dst.w)
+		return false;
+
+	if (plane->index == 0 && dc->soc->has_win_a_without_filters)
+		return false;
+
+	return true;
+}
+
+static bool
+tegra_plane_use_vertical_filtering(struct tegra_plane *plane,
+				   const struct tegra_dc_window *window)
+{
+	struct tegra_dc *dc = plane->dc;
+
+	if (window->src.h == window->dst.h)
+		return false;
+
+	if (plane->index == 0 && dc->soc->has_win_a_without_filters)
+		return false;
+
+	if (plane->index == 2 && dc->soc->has_win_c_without_vert_filter)
+		return false;
+
+	return true;
+}
+
 static void tegra_dc_setup_window(struct tegra_plane *plane,
 				  const struct tegra_dc_window *window)
 {
@@ -361,12 +468,50 @@ static void tegra_dc_setup_window(struct tegra_plane *plane,
 	if (window->bottom_up)
 		value |= V_DIRECTION;
 
+	if (tegra_plane_use_horizontal_filtering(plane, window)) {
+		/*
+		 * Enable horizontal 6-tap filter and set filtering
+		 * coefficients to the default values defined in TRM.
+		 */
+		tegra_plane_writel(plane, 0x00008000, DC_WIN_H_FILTER_P(0));
+		tegra_plane_writel(plane, 0x3e087ce1, DC_WIN_H_FILTER_P(1));
+		tegra_plane_writel(plane, 0x3b117ac1, DC_WIN_H_FILTER_P(2));
+		tegra_plane_writel(plane, 0x591b73aa, DC_WIN_H_FILTER_P(3));
+		tegra_plane_writel(plane, 0x57256d9a, DC_WIN_H_FILTER_P(4));
+		tegra_plane_writel(plane, 0x552f668b, DC_WIN_H_FILTER_P(5));
+		tegra_plane_writel(plane, 0x73385e8b, DC_WIN_H_FILTER_P(6));
+		tegra_plane_writel(plane, 0x72435583, DC_WIN_H_FILTER_P(7));
+		tegra_plane_writel(plane, 0x714c4c8b, DC_WIN_H_FILTER_P(8));
+		tegra_plane_writel(plane, 0x70554393, DC_WIN_H_FILTER_P(9));
+		tegra_plane_writel(plane, 0x715e389b, DC_WIN_H_FILTER_P(10));
+		tegra_plane_writel(plane, 0x71662faa, DC_WIN_H_FILTER_P(11));
+		tegra_plane_writel(plane, 0x536d25ba, DC_WIN_H_FILTER_P(12));
+		tegra_plane_writel(plane, 0x55731bca, DC_WIN_H_FILTER_P(13));
+		tegra_plane_writel(plane, 0x387a11d9, DC_WIN_H_FILTER_P(14));
+		tegra_plane_writel(plane, 0x3c7c08f1, DC_WIN_H_FILTER_P(15));
+
+		value |= H_FILTER;
+	}
+
+	if (tegra_plane_use_vertical_filtering(plane, window)) {
+		unsigned int i, k;
+
+		/*
+		 * Enable vertical 2-tap filter and set filtering
+		 * coefficients to the default values defined in TRM.
+		 */
+		for (i = 0, k = 128; i < 16; i++, k -= 8)
+			tegra_plane_writel(plane, k, DC_WIN_V_FILTER_P(i));
+
+		value |= V_FILTER;
+	}
+
 	tegra_plane_writel(plane, value, DC_WIN_WIN_OPTIONS);
 
-	if (dc->soc->supports_blending)
-		tegra_plane_setup_blending(plane, window);
-	else
+	if (dc->soc->has_legacy_blending)
 		tegra_plane_setup_blending_legacy(plane);
+	else
+		tegra_plane_setup_blending(plane, window);
 }
 
 static const u32 tegra20_primary_formats[] = {
@@ -451,17 +596,18 @@ static int tegra_plane_atomic_check(struct drm_plane *plane,
 				    struct drm_plane_state *state)
 {
 	struct tegra_plane_state *plane_state = to_tegra_plane_state(state);
+	unsigned int rotation = DRM_MODE_ROTATE_0 | DRM_MODE_REFLECT_Y;
 	struct tegra_bo_tiling *tiling = &plane_state->tiling;
 	struct tegra_plane *tegra = to_tegra_plane(plane);
 	struct tegra_dc *dc = to_tegra_dc(state->crtc);
-	unsigned int format;
 	int err;
 
 	/* no need for further checks if the plane is being disabled */
 	if (!state->crtc)
 		return 0;
 
-	err = tegra_plane_format(state->fb->format->format, &format,
+	err = tegra_plane_format(state->fb->format->format,
+				 &plane_state->format,
 				 &plane_state->swap);
 	if (err < 0)
 		return err;
@@ -472,22 +618,12 @@ static int tegra_plane_atomic_check(struct drm_plane *plane,
 	 * the corresponding opaque formats. However, the opaque formats can
 	 * be emulated by disabling alpha blending for the plane.
 	 */
-	if (!dc->soc->supports_blending) {
-		if (!tegra_plane_format_has_alpha(format)) {
-			err = tegra_plane_format_get_alpha(format, &format);
-			if (err < 0)
-				return err;
-
-			plane_state->opaque = true;
-		} else {
-			plane_state->opaque = false;
-		}
-
-		tegra_plane_check_dependent(tegra, plane_state);
+	if (dc->soc->has_legacy_blending) {
+		err = tegra_plane_setup_legacy_state(tegra, plane_state);
+		if (err < 0)
+			return err;
 	}
 
-	plane_state->format = format;
-
 	err = tegra_fb_get_tiling(state->fb, tiling);
 	if (err < 0)
 		return err;
@@ -498,6 +634,13 @@ static int tegra_plane_atomic_check(struct drm_plane *plane,
 		return -EINVAL;
 	}
 
+	rotation = drm_rotation_simplify(state->rotation, rotation);
+
+	if (rotation & DRM_MODE_REFLECT_Y)
+		plane_state->bottom_up = true;
+	else
+		plane_state->bottom_up = false;
+
 	/*
 	 * Tegra doesn't support different strides for U and V planes so we
 	 * error out if the user tries to display a framebuffer with such a
@@ -558,7 +701,7 @@ static void tegra_plane_atomic_update(struct drm_plane *plane,
 	window.dst.w = drm_rect_width(&plane->state->dst);
 	window.dst.h = drm_rect_height(&plane->state->dst);
 	window.bits_per_pixel = fb->format->cpp[0] * 8;
-	window.bottom_up = tegra_fb_is_bottom_up(fb);
+	window.bottom_up = tegra_fb_is_bottom_up(fb) || state->bottom_up;
 
 	/* copy from state */
 	window.zpos = plane->state->normalized_zpos;
@@ -639,9 +782,15 @@ static struct drm_plane *tegra_primary_plane_create(struct drm_device *drm,
 	}
 
 	drm_plane_helper_add(&plane->base, &tegra_plane_helper_funcs);
+	drm_plane_create_zpos_property(&plane->base, plane->index, 0, 255);
 
-	if (dc->soc->supports_blending)
-		drm_plane_create_zpos_property(&plane->base, 0, 0, 255);
+	err = drm_plane_create_rotation_property(&plane->base,
+						 DRM_MODE_ROTATE_0,
+						 DRM_MODE_ROTATE_0 |
+						 DRM_MODE_REFLECT_Y);
+	if (err < 0)
+		dev_err(dc->dev, "failed to create rotation property: %d\n",
+			err);
 
 	return &plane->base;
 }
@@ -918,9 +1067,15 @@ static struct drm_plane *tegra_dc_overlay_plane_create(struct drm_device *drm,
 	}
 
 	drm_plane_helper_add(&plane->base, &tegra_plane_helper_funcs);
+	drm_plane_create_zpos_property(&plane->base, plane->index, 0, 255);
 
-	if (dc->soc->supports_blending)
-		drm_plane_create_zpos_property(&plane->base, 0, 0, 255);
+	err = drm_plane_create_rotation_property(&plane->base,
+						 DRM_MODE_ROTATE_0,
+						 DRM_MODE_ROTATE_0 |
+						 DRM_MODE_REFLECT_Y);
+	if (err < 0)
+		dev_err(dc->dev, "failed to create rotation property: %d\n",
+			err);
 
 	return &plane->base;
 }
@@ -1826,7 +1981,6 @@ static irqreturn_t tegra_dc_irq(int irq, void *data)
 static int tegra_dc_init(struct host1x_client *client)
 {
 	struct drm_device *drm = dev_get_drvdata(client->parent);
-	struct iommu_group *group = iommu_group_get(client->dev);
 	unsigned long flags = HOST1X_SYNCPT_CLIENT_MANAGED;
 	struct tegra_dc *dc = host1x_client_to_dc(client);
 	struct tegra_drm *tegra = drm->dev_private;
@@ -1838,20 +1992,11 @@ static int tegra_dc_init(struct host1x_client *client)
 	if (!dc->syncpt)
 		dev_warn(dc->dev, "failed to allocate syncpoint\n");
 
-	if (group && tegra->domain) {
-		if (group != tegra->group) {
-			err = iommu_attach_group(tegra->domain, group);
-			if (err < 0) {
-				dev_err(dc->dev,
-					"failed to attach to domain: %d\n",
-					err);
-				return err;
-			}
-
-			tegra->group = group;
-		}
-
-		dc->domain = tegra->domain;
+	dc->group = host1x_client_iommu_attach(client, true);
+	if (IS_ERR(dc->group)) {
+		err = PTR_ERR(dc->group);
+		dev_err(client->dev, "failed to attach to domain: %d\n", err);
+		return err;
 	}
 
 	if (dc->soc->wgrps)
@@ -1916,24 +2061,15 @@ cleanup:
 	if (!IS_ERR(primary))
 		drm_plane_cleanup(primary);
 
-	if (group && dc->domain) {
-		if (group == tegra->group) {
-			iommu_detach_group(dc->domain, group);
-			tegra->group = NULL;
-		}
-
-		dc->domain = NULL;
-	}
+	host1x_client_iommu_detach(client, dc->group);
+	host1x_syncpt_free(dc->syncpt);
 
 	return err;
 }
 
 static int tegra_dc_exit(struct host1x_client *client)
 {
-	struct drm_device *drm = dev_get_drvdata(client->parent);
-	struct iommu_group *group = iommu_group_get(client->dev);
 	struct tegra_dc *dc = host1x_client_to_dc(client);
-	struct tegra_drm *tegra = drm->dev_private;
 	int err;
 
 	devm_free_irq(dc->dev, dc->irq, dc);
@@ -1944,15 +2080,7 @@ static int tegra_dc_exit(struct host1x_client *client)
 		return err;
 	}
 
-	if (group && dc->domain) {
-		if (group == tegra->group) {
-			iommu_detach_group(dc->domain, group);
-			tegra->group = NULL;
-		}
-
-		dc->domain = NULL;
-	}
-
+	host1x_client_iommu_detach(client, dc->group);
 	host1x_syncpt_free(dc->syncpt);
 
 	return 0;
@@ -1968,7 +2096,7 @@ static const struct tegra_dc_soc_info tegra20_dc_soc_info = {
 	.supports_interlacing = false,
 	.supports_cursor = false,
 	.supports_block_linear = false,
-	.supports_blending = false,
+	.has_legacy_blending = true,
 	.pitch_align = 8,
 	.has_powergate = false,
 	.coupled_pm = true,
@@ -1978,6 +2106,8 @@ static const struct tegra_dc_soc_info tegra20_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra20_overlay_formats),
 	.overlay_formats = tegra20_overlay_formats,
 	.modifiers = tegra20_modifiers,
+	.has_win_a_without_filters = true,
+	.has_win_c_without_vert_filter = true,
 };
 
 static const struct tegra_dc_soc_info tegra30_dc_soc_info = {
@@ -1985,7 +2115,7 @@ static const struct tegra_dc_soc_info tegra30_dc_soc_info = {
 	.supports_interlacing = false,
 	.supports_cursor = false,
 	.supports_block_linear = false,
-	.supports_blending = false,
+	.has_legacy_blending = true,
 	.pitch_align = 8,
 	.has_powergate = false,
 	.coupled_pm = false,
@@ -1995,6 +2125,8 @@ static const struct tegra_dc_soc_info tegra30_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra20_overlay_formats),
 	.overlay_formats = tegra20_overlay_formats,
 	.modifiers = tegra20_modifiers,
+	.has_win_a_without_filters = false,
+	.has_win_c_without_vert_filter = false,
 };
 
 static const struct tegra_dc_soc_info tegra114_dc_soc_info = {
@@ -2002,7 +2134,7 @@ static const struct tegra_dc_soc_info tegra114_dc_soc_info = {
 	.supports_interlacing = false,
 	.supports_cursor = false,
 	.supports_block_linear = false,
-	.supports_blending = false,
+	.has_legacy_blending = true,
 	.pitch_align = 64,
 	.has_powergate = true,
 	.coupled_pm = false,
@@ -2012,6 +2144,8 @@ static const struct tegra_dc_soc_info tegra114_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra114_overlay_formats),
 	.overlay_formats = tegra114_overlay_formats,
 	.modifiers = tegra20_modifiers,
+	.has_win_a_without_filters = false,
+	.has_win_c_without_vert_filter = false,
 };
 
 static const struct tegra_dc_soc_info tegra124_dc_soc_info = {
@@ -2019,7 +2153,7 @@ static const struct tegra_dc_soc_info tegra124_dc_soc_info = {
 	.supports_interlacing = true,
 	.supports_cursor = true,
 	.supports_block_linear = true,
-	.supports_blending = true,
+	.has_legacy_blending = false,
 	.pitch_align = 64,
 	.has_powergate = true,
 	.coupled_pm = false,
@@ -2029,6 +2163,8 @@ static const struct tegra_dc_soc_info tegra124_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra124_overlay_formats),
 	.overlay_formats = tegra124_overlay_formats,
 	.modifiers = tegra124_modifiers,
+	.has_win_a_without_filters = false,
+	.has_win_c_without_vert_filter = false,
 };
 
 static const struct tegra_dc_soc_info tegra210_dc_soc_info = {
@@ -2036,7 +2172,7 @@ static const struct tegra_dc_soc_info tegra210_dc_soc_info = {
 	.supports_interlacing = true,
 	.supports_cursor = true,
 	.supports_block_linear = true,
-	.supports_blending = true,
+	.has_legacy_blending = false,
 	.pitch_align = 64,
 	.has_powergate = true,
 	.coupled_pm = false,
@@ -2046,6 +2182,8 @@ static const struct tegra_dc_soc_info tegra210_dc_soc_info = {
 	.num_overlay_formats = ARRAY_SIZE(tegra114_overlay_formats),
 	.overlay_formats = tegra114_overlay_formats,
 	.modifiers = tegra124_modifiers,
+	.has_win_a_without_filters = false,
+	.has_win_c_without_vert_filter = false,
 };
 
 static const struct tegra_windowgroup_soc tegra186_dc_wgrps[] = {
@@ -2087,7 +2225,7 @@ static const struct tegra_dc_soc_info tegra186_dc_soc_info = {
 	.supports_interlacing = true,
 	.supports_cursor = true,
 	.supports_block_linear = true,
-	.supports_blending = true,
+	.has_legacy_blending = false,
 	.pitch_align = 64,
 	.has_powergate = false,
 	.coupled_pm = false,
diff --git a/drivers/gpu/drm/tegra/dc.h b/drivers/gpu/drm/tegra/dc.h
index d2b50d32de4d..e96f582ca692 100644
--- a/drivers/gpu/drm/tegra/dc.h
+++ b/drivers/gpu/drm/tegra/dc.h
@@ -55,7 +55,7 @@ struct tegra_dc_soc_info {
 	bool supports_interlacing;
 	bool supports_cursor;
 	bool supports_block_linear;
-	bool supports_blending;
+	bool has_legacy_blending;
 	unsigned int pitch_align;
 	bool has_powergate;
 	bool coupled_pm;
@@ -67,6 +67,8 @@ struct tegra_dc_soc_info {
 	const u32 *overlay_formats;
 	unsigned int num_overlay_formats;
 	const u64 *modifiers;
+	bool has_win_a_without_filters;
+	bool has_win_c_without_vert_filter;
 };
 
 struct tegra_dc {
@@ -92,7 +94,7 @@ struct tegra_dc {
 
 	const struct tegra_dc_soc_info *soc;
 
-	struct iommu_domain *domain;
+	struct iommu_group *group;
 };
 
 static inline struct tegra_dc *
@@ -553,6 +555,9 @@ int tegra_dc_rgb_exit(struct tegra_dc *dc);
 #define  THREAD_NUM(x) (((x) & 0x1f) << 1)
 #define  THREAD_GROUP_ENABLE (1 << 0)
 
+#define DC_WIN_H_FILTER_P(p)			(0x601 + (p))
+#define DC_WIN_V_FILTER_P(p)			(0x619 + (p))
+
 #define DC_WIN_CSC_YOF				0x611
 #define DC_WIN_CSC_KYRGB			0x612
 #define DC_WIN_CSC_KUR				0x613
@@ -566,6 +571,8 @@ int tegra_dc_rgb_exit(struct tegra_dc *dc);
 #define H_DIRECTION  (1 <<  0)
 #define V_DIRECTION  (1 <<  2)
 #define COLOR_EXPAND (1 <<  6)
+#define H_FILTER     (1 <<  8)
+#define V_FILTER     (1 << 10)
 #define CSC_ENABLE   (1 << 18)
 #define WIN_ENABLE   (1 << 30)
 
diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index a0519612ae2c..776c1513e582 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -98,6 +98,10 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags)
 			goto free;
 		}
 
+		err = iova_cache_get();
+		if (err < 0)
+			goto domain;
+
 		geometry = &tegra->domain->geometry;
 		gem_start = geometry->aperture_start;
 		gem_end = geometry->aperture_end - CARVEOUT_SZ;
@@ -191,11 +195,14 @@ config:
 	drm_mode_config_cleanup(drm);
 
 	if (tegra->domain) {
-		iommu_domain_free(tegra->domain);
-		drm_mm_takedown(&tegra->mm);
 		mutex_destroy(&tegra->mm_lock);
+		drm_mm_takedown(&tegra->mm);
 		put_iova_domain(&tegra->carveout.domain);
+		iova_cache_put();
 	}
+domain:
+	if (tegra->domain)
+		iommu_domain_free(tegra->domain);
 free:
 	kfree(tegra);
 	return err;
@@ -217,10 +224,11 @@ static void tegra_drm_unload(struct drm_device *drm)
 		return;
 
 	if (tegra->domain) {
-		iommu_domain_free(tegra->domain);
-		drm_mm_takedown(&tegra->mm);
 		mutex_destroy(&tegra->mm_lock);
+		drm_mm_takedown(&tegra->mm);
 		put_iova_domain(&tegra->carveout.domain);
+		iova_cache_put();
+		iommu_domain_free(tegra->domain);
 	}
 
 	kfree(tegra);
@@ -300,46 +308,15 @@ static int host1x_reloc_copy_from_user(struct host1x_reloc *dest,
 	return 0;
 }
 
-static int host1x_waitchk_copy_from_user(struct host1x_waitchk *dest,
-					 struct drm_tegra_waitchk __user *src,
-					 struct drm_file *file)
-{
-	u32 cmdbuf;
-	int err;
-
-	err = get_user(cmdbuf, &src->handle);
-	if (err < 0)
-		return err;
-
-	err = get_user(dest->offset, &src->offset);
-	if (err < 0)
-		return err;
-
-	err = get_user(dest->syncpt_id, &src->syncpt);
-	if (err < 0)
-		return err;
-
-	err = get_user(dest->thresh, &src->thresh);
-	if (err < 0)
-		return err;
-
-	dest->bo = host1x_bo_lookup(file, cmdbuf);
-	if (!dest->bo)
-		return -ENOENT;
-
-	return 0;
-}
-
 int tegra_drm_submit(struct tegra_drm_context *context,
 		     struct drm_tegra_submit *args, struct drm_device *drm,
 		     struct drm_file *file)
 {
+	struct host1x_client *client = &context->client->base;
 	unsigned int num_cmdbufs = args->num_cmdbufs;
 	unsigned int num_relocs = args->num_relocs;
-	unsigned int num_waitchks = args->num_waitchks;
 	struct drm_tegra_cmdbuf __user *user_cmdbufs;
 	struct drm_tegra_reloc __user *user_relocs;
-	struct drm_tegra_waitchk __user *user_waitchks;
 	struct drm_tegra_syncpt __user *user_syncpt;
 	struct drm_tegra_syncpt syncpt;
 	struct host1x *host1x = dev_get_drvdata(drm->dev->parent);
@@ -351,7 +328,6 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 
 	user_cmdbufs = u64_to_user_ptr(args->cmdbufs);
 	user_relocs = u64_to_user_ptr(args->relocs);
-	user_waitchks = u64_to_user_ptr(args->waitchks);
 	user_syncpt = u64_to_user_ptr(args->syncpts);
 
 	/* We don't yet support other than one syncpt_incr struct per submit */
@@ -363,21 +339,20 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 		return -EINVAL;
 
 	job = host1x_job_alloc(context->channel, args->num_cmdbufs,
-			       args->num_relocs, args->num_waitchks);
+			       args->num_relocs);
 	if (!job)
 		return -ENOMEM;
 
 	job->num_relocs = args->num_relocs;
-	job->num_waitchk = args->num_waitchks;
-	job->client = (u32)args->context;
-	job->class = context->client->base.class;
+	job->client = client;
+	job->class = client->class;
 	job->serialize = true;
 
 	/*
 	 * Track referenced BOs so that they can be unreferenced after the
 	 * submission is complete.
 	 */
-	num_refs = num_cmdbufs + num_relocs * 2 + num_waitchks;
+	num_refs = num_cmdbufs + num_relocs * 2;
 
 	refs = kmalloc_array(num_refs, sizeof(*refs), GFP_KERNEL);
 	if (!refs) {
@@ -438,13 +413,13 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 		struct host1x_reloc *reloc;
 		struct tegra_bo *obj;
 
-		err = host1x_reloc_copy_from_user(&job->relocarray[num_relocs],
+		err = host1x_reloc_copy_from_user(&job->relocs[num_relocs],
 						  &user_relocs[num_relocs], drm,
 						  file);
 		if (err < 0)
 			goto fail;
 
-		reloc = &job->relocarray[num_relocs];
+		reloc = &job->relocs[num_relocs];
 		obj = host1x_to_tegra_bo(reloc->cmdbuf.bo);
 		refs[num_refs++] = &obj->gem;
 
@@ -468,30 +443,6 @@ int tegra_drm_submit(struct tegra_drm_context *context,
 		}
 	}
 
-	/* copy and resolve waitchks from submit */
-	while (num_waitchks--) {
-		struct host1x_waitchk *wait = &job->waitchk[num_waitchks];
-		struct tegra_bo *obj;
-
-		err = host1x_waitchk_copy_from_user(
-			wait, &user_waitchks[num_waitchks], file);
-		if (err < 0)
-			goto fail;
-
-		obj = host1x_to_tegra_bo(wait->bo);
-		refs[num_refs++] = &obj->gem;
-
-		/*
-		 * The unaligned offset will cause an unaligned write during
-		 * of the waitchks patching, corrupting the commands stream.
-		 */
-		if (wait->offset & 3 ||
-		    wait->offset >= obj->gem.size) {
-			err = -EINVAL;
-			goto fail;
-		}
-	}
-
 	if (copy_from_user(&syncpt, user_syncpt, sizeof(syncpt))) {
 		err = -EFAULT;
 		goto fail;
@@ -1101,6 +1052,52 @@ int tegra_drm_unregister_client(struct tegra_drm *tegra,
 	return 0;
 }
 
+struct iommu_group *host1x_client_iommu_attach(struct host1x_client *client,
+					       bool shared)
+{
+	struct drm_device *drm = dev_get_drvdata(client->parent);
+	struct tegra_drm *tegra = drm->dev_private;
+	struct iommu_group *group = NULL;
+	int err;
+
+	if (tegra->domain) {
+		group = iommu_group_get(client->dev);
+		if (!group) {
+			dev_err(client->dev, "failed to get IOMMU group\n");
+			return ERR_PTR(-ENODEV);
+		}
+
+		if (!shared || (shared && (group != tegra->group))) {
+			err = iommu_attach_group(tegra->domain, group);
+			if (err < 0) {
+				iommu_group_put(group);
+				return ERR_PTR(err);
+			}
+
+			if (shared && !tegra->group)
+				tegra->group = group;
+		}
+	}
+
+	return group;
+}
+
+void host1x_client_iommu_detach(struct host1x_client *client,
+				struct iommu_group *group)
+{
+	struct drm_device *drm = dev_get_drvdata(client->parent);
+	struct tegra_drm *tegra = drm->dev_private;
+
+	if (group) {
+		if (group == tegra->group) {
+			iommu_detach_group(tegra->domain, group);
+			tegra->group = NULL;
+		}
+
+		iommu_group_put(group);
+	}
+}
+
 void *tegra_drm_alloc(struct tegra_drm *tegra, size_t size, dma_addr_t *dma)
 {
 	struct iova *alloc;
diff --git a/drivers/gpu/drm/tegra/drm.h b/drivers/gpu/drm/tegra/drm.h
index 4f41aaec8530..92d248784396 100644
--- a/drivers/gpu/drm/tegra/drm.h
+++ b/drivers/gpu/drm/tegra/drm.h
@@ -29,16 +29,10 @@
 
 struct reset_control;
 
-struct tegra_fb {
-	struct drm_framebuffer base;
-	struct tegra_bo **planes;
-	unsigned int num_planes;
-};
-
 #ifdef CONFIG_DRM_FBDEV_EMULATION
 struct tegra_fbdev {
 	struct drm_fb_helper base;
-	struct tegra_fb *fb;
+	struct drm_framebuffer *fb;
 };
 #endif
 
@@ -97,6 +91,7 @@ struct tegra_drm_client {
 	struct host1x_client base;
 	struct list_head list;
 
+	unsigned int version;
 	const struct tegra_drm_client_ops *ops;
 };
 
@@ -110,6 +105,10 @@ int tegra_drm_register_client(struct tegra_drm *tegra,
 			      struct tegra_drm_client *client);
 int tegra_drm_unregister_client(struct tegra_drm *tegra,
 				struct tegra_drm_client *client);
+struct iommu_group *host1x_client_iommu_attach(struct host1x_client *client,
+					       bool shared);
+void host1x_client_iommu_detach(struct host1x_client *client,
+				struct iommu_group *group);
 
 int tegra_drm_init(struct tegra_drm *tegra, struct drm_device *drm);
 int tegra_drm_exit(struct tegra_drm *tegra);
diff --git a/drivers/gpu/drm/tegra/dsi.c b/drivers/gpu/drm/tegra/dsi.c
index 87c5d89bc9ba..ee6ca8fa1c65 100644
--- a/drivers/gpu/drm/tegra/dsi.c
+++ b/drivers/gpu/drm/tegra/dsi.c
@@ -1052,7 +1052,7 @@ static int tegra_dsi_init(struct host1x_client *client)
 		drm_encoder_helper_add(&dsi->output.encoder,
 				       &tegra_dsi_encoder_helper_funcs);
 
-		drm_mode_connector_attach_encoder(&dsi->output.connector,
+		drm_connector_attach_encoder(&dsi->output.connector,
 						  &dsi->output.encoder);
 		drm_connector_register(&dsi->output.connector);
 
@@ -1411,6 +1411,9 @@ static int tegra_dsi_host_attach(struct mipi_dsi_host *host,
 		struct tegra_output *output = &dsi->output;
 
 		output->panel = of_drm_find_panel(device->dev.of_node);
+		if (IS_ERR(output->panel))
+			output->panel = NULL;
+
 		if (output->panel && output->connector.dev) {
 			drm_panel_attach(output->panel, &output->connector);
 			drm_helper_hpd_irq_event(output->connector.dev);
diff --git a/drivers/gpu/drm/tegra/fb.c b/drivers/gpu/drm/tegra/fb.c
index e69434909a42..4c22cdded3c2 100644
--- a/drivers/gpu/drm/tegra/fb.c
+++ b/drivers/gpu/drm/tegra/fb.c
@@ -14,11 +14,7 @@
 
 #include "drm.h"
 #include "gem.h"
-
-static inline struct tegra_fb *to_tegra_fb(struct drm_framebuffer *fb)
-{
-	return container_of(fb, struct tegra_fb, base);
-}
+#include <drm/drm_gem_framebuffer_helper.h>
 
 #ifdef CONFIG_DRM_FBDEV_EMULATION
 static inline struct tegra_fbdev *to_tegra_fbdev(struct drm_fb_helper *helper)
@@ -30,19 +26,14 @@ static inline struct tegra_fbdev *to_tegra_fbdev(struct drm_fb_helper *helper)
 struct tegra_bo *tegra_fb_get_plane(struct drm_framebuffer *framebuffer,
 				    unsigned int index)
 {
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
-
-	if (index >= framebuffer->format->num_planes)
-		return NULL;
-
-	return fb->planes[index];
+	return to_tegra_bo(drm_gem_fb_get_obj(framebuffer, index));
 }
 
 bool tegra_fb_is_bottom_up(struct drm_framebuffer *framebuffer)
 {
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
+	struct tegra_bo *bo = tegra_fb_get_plane(framebuffer, 0);
 
-	if (fb->planes[0]->flags & TEGRA_BO_BOTTOM_UP)
+	if (bo->flags & TEGRA_BO_BOTTOM_UP)
 		return true;
 
 	return false;
@@ -51,8 +42,7 @@ bool tegra_fb_is_bottom_up(struct drm_framebuffer *framebuffer)
 int tegra_fb_get_tiling(struct drm_framebuffer *framebuffer,
 			struct tegra_bo_tiling *tiling)
 {
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
-	uint64_t modifier = fb->base.modifier;
+	uint64_t modifier = framebuffer->modifier;
 
 	switch (modifier) {
 	case DRM_FORMAT_MOD_LINEAR:
@@ -102,46 +92,17 @@ int tegra_fb_get_tiling(struct drm_framebuffer *framebuffer,
 	return 0;
 }
 
-static void tegra_fb_destroy(struct drm_framebuffer *framebuffer)
-{
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
-	unsigned int i;
-
-	for (i = 0; i < fb->num_planes; i++) {
-		struct tegra_bo *bo = fb->planes[i];
-
-		if (bo) {
-			if (bo->pages)
-				vunmap(bo->vaddr);
-
-			drm_gem_object_put_unlocked(&bo->gem);
-		}
-	}
-
-	drm_framebuffer_cleanup(framebuffer);
-	kfree(fb->planes);
-	kfree(fb);
-}
-
-static int tegra_fb_create_handle(struct drm_framebuffer *framebuffer,
-				  struct drm_file *file, unsigned int *handle)
-{
-	struct tegra_fb *fb = to_tegra_fb(framebuffer);
-
-	return drm_gem_handle_create(file, &fb->planes[0]->gem, handle);
-}
-
 static const struct drm_framebuffer_funcs tegra_fb_funcs = {
-	.destroy = tegra_fb_destroy,
-	.create_handle = tegra_fb_create_handle,
+	.destroy = drm_gem_fb_destroy,
+	.create_handle = drm_gem_fb_create_handle,
 };
 
-static struct tegra_fb *tegra_fb_alloc(struct drm_device *drm,
-				       const struct drm_mode_fb_cmd2 *mode_cmd,
-				       struct tegra_bo **planes,
-				       unsigned int num_planes)
+static struct drm_framebuffer *tegra_fb_alloc(struct drm_device *drm,
+					      const struct drm_mode_fb_cmd2 *mode_cmd,
+					      struct tegra_bo **planes,
+					      unsigned int num_planes)
 {
-	struct tegra_fb *fb;
+	struct drm_framebuffer *fb;
 	unsigned int i;
 	int err;
 
@@ -149,24 +110,15 @@ static struct tegra_fb *tegra_fb_alloc(struct drm_device *drm,
 	if (!fb)
 		return ERR_PTR(-ENOMEM);
 
-	fb->planes = kzalloc(num_planes * sizeof(*planes), GFP_KERNEL);
-	if (!fb->planes) {
-		kfree(fb);
-		return ERR_PTR(-ENOMEM);
-	}
-
-	fb->num_planes = num_planes;
+	drm_helper_mode_fill_fb_struct(drm, fb, mode_cmd);
 
-	drm_helper_mode_fill_fb_struct(drm, &fb->base, mode_cmd);
+	for (i = 0; i < fb->format->num_planes; i++)
+		fb->obj[i] = &planes[i]->gem;
 
-	for (i = 0; i < fb->num_planes; i++)
-		fb->planes[i] = planes[i];
-
-	err = drm_framebuffer_init(drm, &fb->base, &tegra_fb_funcs);
+	err = drm_framebuffer_init(drm, fb, &tegra_fb_funcs);
 	if (err < 0) {
 		dev_err(drm->dev, "failed to initialize framebuffer: %d\n",
 			err);
-		kfree(fb->planes);
 		kfree(fb);
 		return ERR_PTR(err);
 	}
@@ -181,7 +133,7 @@ struct drm_framebuffer *tegra_fb_create(struct drm_device *drm,
 	unsigned int hsub, vsub, i;
 	struct tegra_bo *planes[4];
 	struct drm_gem_object *gem;
-	struct tegra_fb *fb;
+	struct drm_framebuffer *fb;
 	int err;
 
 	hsub = drm_format_horz_chroma_subsampling(cmd->pixel_format);
@@ -217,7 +169,7 @@ struct drm_framebuffer *tegra_fb_create(struct drm_device *drm,
 		goto unreference;
 	}
 
-	return &fb->base;
+	return fb;
 
 unreference:
 	while (i--)
@@ -298,7 +250,7 @@ static int tegra_fbdev_probe(struct drm_fb_helper *helper,
 		return PTR_ERR(fbdev->fb);
 	}
 
-	fb = &fbdev->fb->base;
+	fb = fbdev->fb;
 	helper->fb = fb;
 	helper->fbdev = info;
 
@@ -398,8 +350,17 @@ static void tegra_fbdev_exit(struct tegra_fbdev *fbdev)
 {
 	drm_fb_helper_unregister_fbi(&fbdev->base);
 
-	if (fbdev->fb)
-		drm_framebuffer_remove(&fbdev->fb->base);
+	if (fbdev->fb) {
+		struct tegra_bo *bo = tegra_fb_get_plane(fbdev->fb, 0);
+
+		/* Undo the special mapping we made in fbdev probe. */
+		if (bo && bo->pages) {
+			vunmap(bo->vaddr);
+			bo->vaddr = 0;
+		}
+
+		drm_framebuffer_remove(fbdev->fb);
+	}
 
 	drm_fb_helper_fini(&fbdev->base);
 	tegra_fbdev_free(fbdev);
diff --git a/drivers/gpu/drm/tegra/gem.c b/drivers/gpu/drm/tegra/gem.c
index 8b0b4ff64bb4..4f80100ff5f3 100644
--- a/drivers/gpu/drm/tegra/gem.c
+++ b/drivers/gpu/drm/tegra/gem.c
@@ -422,14 +422,13 @@ int tegra_bo_dumb_create(struct drm_file *file, struct drm_device *drm,
 	return 0;
 }
 
-static int tegra_bo_fault(struct vm_fault *vmf)
+static vm_fault_t tegra_bo_fault(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 	struct drm_gem_object *gem = vma->vm_private_data;
 	struct tegra_bo *bo = to_tegra_bo(gem);
 	struct page *page;
 	pgoff_t offset;
-	int err;
 
 	if (!bo->pages)
 		return VM_FAULT_SIGBUS;
@@ -437,20 +436,7 @@ static int tegra_bo_fault(struct vm_fault *vmf)
 	offset = (vmf->address - vma->vm_start) >> PAGE_SHIFT;
 	page = bo->pages[offset];
 
-	err = vm_insert_page(vma, vmf->address, page);
-	switch (err) {
-	case -EAGAIN:
-	case 0:
-	case -ERESTARTSYS:
-	case -EINTR:
-	case -EBUSY:
-		return VM_FAULT_NOPAGE;
-
-	case -ENOMEM:
-		return VM_FAULT_OOM;
-	}
-
-	return VM_FAULT_SIGBUS;
+	return vmf_insert_page(vma, vmf->address, page);
 }
 
 const struct vm_operations_struct tegra_bo_vm_ops = {
@@ -596,18 +582,6 @@ static int tegra_gem_prime_end_cpu_access(struct dma_buf *buf,
 	return 0;
 }
 
-static void *tegra_gem_prime_kmap_atomic(struct dma_buf *buf,
-					 unsigned long page)
-{
-	return NULL;
-}
-
-static void tegra_gem_prime_kunmap_atomic(struct dma_buf *buf,
-					  unsigned long page,
-					  void *addr)
-{
-}
-
 static void *tegra_gem_prime_kmap(struct dma_buf *buf, unsigned long page)
 {
 	return NULL;
@@ -648,8 +622,6 @@ static const struct dma_buf_ops tegra_gem_prime_dmabuf_ops = {
 	.release = tegra_gem_prime_release,
 	.begin_cpu_access = tegra_gem_prime_begin_cpu_access,
 	.end_cpu_access = tegra_gem_prime_end_cpu_access,
-	.map_atomic = tegra_gem_prime_kmap_atomic,
-	.unmap_atomic = tegra_gem_prime_kunmap_atomic,
 	.map = tegra_gem_prime_kmap,
 	.unmap = tegra_gem_prime_kunmap,
 	.mmap = tegra_gem_prime_mmap,
@@ -663,6 +635,8 @@ struct dma_buf *tegra_gem_prime_export(struct drm_device *drm,
 {
 	DEFINE_DMA_BUF_EXPORT_INFO(exp_info);
 
+	exp_info.exp_name = KBUILD_MODNAME;
+	exp_info.owner = drm->driver->fops->owner;
 	exp_info.ops = &tegra_gem_prime_dmabuf_ops;
 	exp_info.size = gem->size;
 	exp_info.flags = flags;
diff --git a/drivers/gpu/drm/tegra/gr2d.c b/drivers/gpu/drm/tegra/gr2d.c
index 9a8ea93016a9..673059fd2fcb 100644
--- a/drivers/gpu/drm/tegra/gr2d.c
+++ b/drivers/gpu/drm/tegra/gr2d.c
@@ -7,16 +7,25 @@
  */
 
 #include <linux/clk.h>
+#include <linux/iommu.h>
+#include <linux/of_device.h>
 
 #include "drm.h"
 #include "gem.h"
 #include "gr2d.h"
 
+struct gr2d_soc {
+	unsigned int version;
+};
+
 struct gr2d {
+	struct iommu_group *group;
 	struct tegra_drm_client client;
 	struct host1x_channel *channel;
 	struct clk *clk;
 
+	const struct gr2d_soc *soc;
+
 	DECLARE_BITMAP(addr_regs, GR2D_NUM_REGS);
 };
 
@@ -31,6 +40,7 @@ static int gr2d_init(struct host1x_client *client)
 	struct drm_device *dev = dev_get_drvdata(client->parent);
 	unsigned long flags = HOST1X_SYNCPT_HAS_BASE;
 	struct gr2d *gr2d = to_gr2d(drm);
+	int err;
 
 	gr2d->channel = host1x_channel_request(client->dev);
 	if (!gr2d->channel)
@@ -38,24 +48,48 @@ static int gr2d_init(struct host1x_client *client)
 
 	client->syncpts[0] = host1x_syncpt_request(client, flags);
 	if (!client->syncpts[0]) {
-		host1x_channel_put(gr2d->channel);
-		return -ENOMEM;
+		err = -ENOMEM;
+		dev_err(client->dev, "failed to request syncpoint: %d\n", err);
+		goto put;
+	}
+
+	gr2d->group = host1x_client_iommu_attach(client, false);
+	if (IS_ERR(gr2d->group)) {
+		err = PTR_ERR(gr2d->group);
+		dev_err(client->dev, "failed to attach to domain: %d\n", err);
+		goto free;
+	}
+
+	err = tegra_drm_register_client(dev->dev_private, drm);
+	if (err < 0) {
+		dev_err(client->dev, "failed to register client: %d\n", err);
+		goto detach;
 	}
 
-	return tegra_drm_register_client(dev->dev_private, drm);
+	return 0;
+
+detach:
+	host1x_client_iommu_detach(client, gr2d->group);
+free:
+	host1x_syncpt_free(client->syncpts[0]);
+put:
+	host1x_channel_put(gr2d->channel);
+	return err;
 }
 
 static int gr2d_exit(struct host1x_client *client)
 {
 	struct tegra_drm_client *drm = host1x_to_drm_client(client);
 	struct drm_device *dev = dev_get_drvdata(client->parent);
+	struct tegra_drm *tegra = dev->dev_private;
 	struct gr2d *gr2d = to_gr2d(drm);
 	int err;
 
-	err = tegra_drm_unregister_client(dev->dev_private, drm);
+	err = tegra_drm_unregister_client(tegra, drm);
 	if (err < 0)
 		return err;
 
+	host1x_client_iommu_detach(client, gr2d->group);
 	host1x_syncpt_free(client->syncpts[0]);
 	host1x_channel_put(gr2d->channel);
 
@@ -123,9 +157,17 @@ static const struct tegra_drm_client_ops gr2d_ops = {
 	.submit = tegra_drm_submit,
 };
 
+static const struct gr2d_soc tegra20_gr2d_soc = {
+	.version = 0x20,
+};
+
+static const struct gr2d_soc tegra30_gr2d_soc = {
+	.version = 0x30,
+};
+
 static const struct of_device_id gr2d_match[] = {
-	{ .compatible = "nvidia,tegra30-gr2d" },
-	{ .compatible = "nvidia,tegra20-gr2d" },
+	{ .compatible = "nvidia,tegra30-gr2d", .data = &tegra20_gr2d_soc },
+	{ .compatible = "nvidia,tegra20-gr2d", .data = &tegra30_gr2d_soc },
 	{ },
 };
 MODULE_DEVICE_TABLE(of, gr2d_match);
@@ -158,6 +200,8 @@ static int gr2d_probe(struct platform_device *pdev)
 	if (!gr2d)
 		return -ENOMEM;
 
+	gr2d->soc = of_device_get_match_data(dev);
+
 	syncpts = devm_kzalloc(dev, sizeof(*syncpts), GFP_KERNEL);
 	if (!syncpts)
 		return -ENOMEM;
@@ -182,6 +226,7 @@ static int gr2d_probe(struct platform_device *pdev)
 	gr2d->client.base.num_syncpts = 1;
 
 	INIT_LIST_HEAD(&gr2d->client.list);
+	gr2d->client.version = gr2d->soc->version;
 	gr2d->client.ops = &gr2d_ops;
 
 	err = host1x_client_register(&gr2d->client.base);
diff --git a/drivers/gpu/drm/tegra/gr3d.c b/drivers/gpu/drm/tegra/gr3d.c
index 28c4ef63065b..4778ae999668 100644
--- a/drivers/gpu/drm/tegra/gr3d.c
+++ b/drivers/gpu/drm/tegra/gr3d.c
@@ -9,7 +9,9 @@
 
 #include <linux/clk.h>
 #include <linux/host1x.h>
+#include <linux/iommu.h>
 #include <linux/module.h>
+#include <linux/of_device.h>
 #include <linux/platform_device.h>
 #include <linux/reset.h>
 
@@ -19,7 +21,12 @@
 #include "gem.h"
 #include "gr3d.h"
 
+struct gr3d_soc {
+	unsigned int version;
+};
+
 struct gr3d {
+	struct iommu_group *group;
 	struct tegra_drm_client client;
 	struct host1x_channel *channel;
 	struct clk *clk_secondary;
@@ -27,6 +34,8 @@ struct gr3d {
 	struct reset_control *rst_secondary;
 	struct reset_control *rst;
 
+	const struct gr3d_soc *soc;
+
 	DECLARE_BITMAP(addr_regs, GR3D_NUM_REGS);
 };
 
@@ -41,6 +50,7 @@ static int gr3d_init(struct host1x_client *client)
 	struct drm_device *dev = dev_get_drvdata(client->parent);
 	unsigned long flags = HOST1X_SYNCPT_HAS_BASE;
 	struct gr3d *gr3d = to_gr3d(drm);
+	int err;
 
 	gr3d->channel = host1x_channel_request(client->dev);
 	if (!gr3d->channel)
@@ -48,11 +58,33 @@ static int gr3d_init(struct host1x_client *client)
 
 	client->syncpts[0] = host1x_syncpt_request(client, flags);
 	if (!client->syncpts[0]) {
-		host1x_channel_put(gr3d->channel);
-		return -ENOMEM;
+		err = -ENOMEM;
+		dev_err(client->dev, "failed to request syncpoint: %d\n", err);
+		goto put;
 	}
 
-	return tegra_drm_register_client(dev->dev_private, drm);
+	gr3d->group = host1x_client_iommu_attach(client, false);
+	if (IS_ERR(gr3d->group)) {
+		err = PTR_ERR(gr3d->group);
+		dev_err(client->dev, "failed to attach to domain: %d\n", err);
+		goto free;
+	}
+
+	err = tegra_drm_register_client(dev->dev_private, drm);
+	if (err < 0) {
+		dev_err(client->dev, "failed to register client: %d\n", err);
+		goto detach;
+	}
+
+	return 0;
+
+detach:
+	host1x_client_iommu_detach(client, gr3d->group);
+free:
+	host1x_syncpt_free(client->syncpts[0]);
+put:
+	host1x_channel_put(gr3d->channel);
+	return err;
 }
 
 static int gr3d_exit(struct host1x_client *client)
@@ -66,6 +98,7 @@ static int gr3d_exit(struct host1x_client *client)
 	if (err < 0)
 		return err;
 
+	host1x_client_iommu_detach(client, gr3d->group);
 	host1x_syncpt_free(client->syncpts[0]);
 	host1x_channel_put(gr3d->channel);
 
@@ -125,10 +158,22 @@ static const struct tegra_drm_client_ops gr3d_ops = {
 	.submit = tegra_drm_submit,
 };
 
+static const struct gr3d_soc tegra20_gr3d_soc = {
+	.version = 0x20,
+};
+
+static const struct gr3d_soc tegra30_gr3d_soc = {
+	.version = 0x30,
+};
+
+static const struct gr3d_soc tegra114_gr3d_soc = {
+	.version = 0x35,
+};
+
 static const struct of_device_id tegra_gr3d_match[] = {
-	{ .compatible = "nvidia,tegra114-gr3d" },
-	{ .compatible = "nvidia,tegra30-gr3d" },
-	{ .compatible = "nvidia,tegra20-gr3d" },
+	{ .compatible = "nvidia,tegra114-gr3d", .data = &tegra114_gr3d_soc },
+	{ .compatible = "nvidia,tegra30-gr3d", .data = &tegra30_gr3d_soc },
+	{ .compatible = "nvidia,tegra20-gr3d", .data = &tegra20_gr3d_soc },
 	{ }
 };
 MODULE_DEVICE_TABLE(of, tegra_gr3d_match);
@@ -250,6 +295,8 @@ static int gr3d_probe(struct platform_device *pdev)
 	if (!gr3d)
 		return -ENOMEM;
 
+	gr3d->soc = of_device_get_match_data(&pdev->dev);
+
 	syncpts = devm_kzalloc(&pdev->dev, sizeof(*syncpts), GFP_KERNEL);
 	if (!syncpts)
 		return -ENOMEM;
@@ -307,6 +354,7 @@ static int gr3d_probe(struct platform_device *pdev)
 	gr3d->client.base.num_syncpts = 1;
 
 	INIT_LIST_HEAD(&gr3d->client.list);
+	gr3d->client.version = gr3d->soc->version;
 	gr3d->client.ops = &gr3d_ops;
 
 	err = host1x_client_register(&gr3d->client.base);
diff --git a/drivers/gpu/drm/tegra/hdmi.c b/drivers/gpu/drm/tegra/hdmi.c
index 784739a9f497..0082468f703c 100644
--- a/drivers/gpu/drm/tegra/hdmi.c
+++ b/drivers/gpu/drm/tegra/hdmi.c
@@ -1488,7 +1488,7 @@ static int tegra_hdmi_init(struct host1x_client *client)
 	drm_encoder_helper_add(&hdmi->output.encoder,
 			       &tegra_hdmi_encoder_helper_funcs);
 
-	drm_mode_connector_attach_encoder(&hdmi->output.connector,
+	drm_connector_attach_encoder(&hdmi->output.connector,
 					  &hdmi->output.encoder);
 	drm_connector_register(&hdmi->output.connector);
 
diff --git a/drivers/gpu/drm/tegra/hub.c b/drivers/gpu/drm/tegra/hub.c
index 9a3f23d4780f..8f4fcbb515fb 100644
--- a/drivers/gpu/drm/tegra/hub.c
+++ b/drivers/gpu/drm/tegra/hub.c
@@ -687,7 +687,7 @@ void tegra_display_hub_atomic_commit(struct drm_device *drm,
 	struct device *dev = hub->client.dev;
 	int err;
 
-	hub_state = tegra_display_hub_get_state(hub, state);
+	hub_state = to_tegra_display_hub_state(hub->base.state);
 
 	if (hub_state->clk) {
 		err = clk_set_rate(hub_state->clk, hub_state->rate);
diff --git a/drivers/gpu/drm/tegra/output.c b/drivers/gpu/drm/tegra/output.c
index ffe34bd0bb9d..c662efc7e413 100644
--- a/drivers/gpu/drm/tegra/output.c
+++ b/drivers/gpu/drm/tegra/output.c
@@ -37,7 +37,7 @@ int tegra_output_connector_get_modes(struct drm_connector *connector)
 		edid = drm_get_edid(connector, output->ddc);
 
 	cec_notifier_set_phys_addr_from_edid(output->notifier, edid);
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 
 	if (edid) {
 		err = drm_add_edid_modes(connector, edid);
@@ -110,8 +110,8 @@ int tegra_output_probe(struct tegra_output *output)
 	panel = of_parse_phandle(output->of_node, "nvidia,panel", 0);
 	if (panel) {
 		output->panel = of_drm_find_panel(panel);
-		if (!output->panel)
-			return -EPROBE_DEFER;
+		if (IS_ERR(output->panel))
+			return PTR_ERR(output->panel);
 
 		of_node_put(panel);
 	}
diff --git a/drivers/gpu/drm/tegra/plane.c b/drivers/gpu/drm/tegra/plane.c
index 176ef46c615c..d068e8aa3553 100644
--- a/drivers/gpu/drm/tegra/plane.c
+++ b/drivers/gpu/drm/tegra/plane.c
@@ -23,6 +23,7 @@ static void tegra_plane_destroy(struct drm_plane *plane)
 
 static void tegra_plane_reset(struct drm_plane *plane)
 {
+	struct tegra_plane *p = to_tegra_plane(plane);
 	struct tegra_plane_state *state;
 
 	if (plane->state)
@@ -35,6 +36,8 @@ static void tegra_plane_reset(struct drm_plane *plane)
 	if (state) {
 		plane->state = &state->base;
 		plane->state->plane = plane;
+		plane->state->zpos = p->index;
+		plane->state->normalized_zpos = p->index;
 	}
 }
 
@@ -53,10 +56,11 @@ tegra_plane_atomic_duplicate_state(struct drm_plane *plane)
 	copy->tiling = state->tiling;
 	copy->format = state->format;
 	copy->swap = state->swap;
+	copy->bottom_up = state->bottom_up;
 	copy->opaque = state->opaque;
 
-	for (i = 0; i < 3; i++)
-		copy->dependent[i] = state->dependent[i];
+	for (i = 0; i < 2; i++)
+		copy->blending[i] = state->blending[i];
 
 	return &copy->base;
 }
@@ -267,24 +271,8 @@ static bool __drm_format_has_alpha(u32 format)
 	return false;
 }
 
-/*
- * This is applicable to Tegra20 and Tegra30 only where the opaque formats can
- * be emulated using the alpha formats and alpha blending disabled.
- */
-bool tegra_plane_format_has_alpha(unsigned int format)
-{
-	switch (format) {
-	case WIN_COLOR_DEPTH_B5G5R5A1:
-	case WIN_COLOR_DEPTH_A1B5G5R5:
-	case WIN_COLOR_DEPTH_R8G8B8A8:
-	case WIN_COLOR_DEPTH_B8G8R8A8:
-		return true;
-	}
-
-	return false;
-}
-
-int tegra_plane_format_get_alpha(unsigned int opaque, unsigned int *alpha)
+static int tegra_plane_format_get_alpha(unsigned int opaque,
+					unsigned int *alpha)
 {
 	if (tegra_plane_format_is_yuv(opaque, NULL)) {
 		*alpha = opaque;
@@ -316,6 +304,67 @@ int tegra_plane_format_get_alpha(unsigned int opaque, unsigned int *alpha)
 	return -EINVAL;
 }
 
+/*
+ * This is applicable to Tegra20 and Tegra30 only where the opaque formats can
+ * be emulated using the alpha formats and alpha blending disabled.
+ */
+static int tegra_plane_setup_opacity(struct tegra_plane *tegra,
+				     struct tegra_plane_state *state)
+{
+	unsigned int format;
+	int err;
+
+	switch (state->format) {
+	case WIN_COLOR_DEPTH_B5G5R5A1:
+	case WIN_COLOR_DEPTH_A1B5G5R5:
+	case WIN_COLOR_DEPTH_R8G8B8A8:
+	case WIN_COLOR_DEPTH_B8G8R8A8:
+		state->opaque = false;
+		break;
+
+	default:
+		err = tegra_plane_format_get_alpha(state->format, &format);
+		if (err < 0)
+			return err;
+
+		state->format = format;
+		state->opaque = true;
+		break;
+	}
+
+	return 0;
+}
+
+static int tegra_plane_check_transparency(struct tegra_plane *tegra,
+					  struct tegra_plane_state *state)
+{
+	struct drm_plane_state *old, *plane_state;
+	struct drm_plane *plane;
+
+	old = drm_atomic_get_old_plane_state(state->base.state, &tegra->base);
+
+	/* check if zpos / transparency changed */
+	if (old->normalized_zpos == state->base.normalized_zpos &&
+	    to_tegra_plane_state(old)->opaque == state->opaque)
+		return 0;
+
+	/* include all sibling planes into this commit */
+	drm_for_each_plane(plane, tegra->base.dev) {
+		struct tegra_plane *p = to_tegra_plane(plane);
+
+		/* skip this plane and planes on different CRTCs */
+		if (p == tegra || p->dc != tegra->dc)
+			continue;
+
+		plane_state = drm_atomic_get_plane_state(state->base.state,
+							 plane);
+		if (IS_ERR(plane_state))
+			return PTR_ERR(plane_state);
+	}
+
+	return 1;
+}
+
 static unsigned int tegra_plane_get_overlap_index(struct tegra_plane *plane,
 						  struct tegra_plane *other)
 {
@@ -336,61 +385,98 @@ static unsigned int tegra_plane_get_overlap_index(struct tegra_plane *plane,
 	return index;
 }
 
-void tegra_plane_check_dependent(struct tegra_plane *tegra,
-				 struct tegra_plane_state *state)
+static void tegra_plane_update_transparency(struct tegra_plane *tegra,
+					    struct tegra_plane_state *state)
 {
-	struct drm_plane_state *old, *new;
+	struct drm_plane_state *new;
 	struct drm_plane *plane;
-	unsigned int zpos[2];
 	unsigned int i;
 
-	for (i = 0; i < 2; i++)
-		zpos[i] = 0;
-
-	for_each_oldnew_plane_in_state(state->base.state, plane, old, new, i) {
+	for_each_new_plane_in_state(state->base.state, plane, new, i) {
 		struct tegra_plane *p = to_tegra_plane(plane);
 		unsigned index;
 
 		/* skip this plane and planes on different CRTCs */
-		if (p == tegra || new->crtc != state->base.crtc)
+		if (p == tegra || p->dc != tegra->dc)
 			continue;
 
 		index = tegra_plane_get_overlap_index(tegra, p);
 
-		state->dependent[index] = false;
+		if (new->fb && __drm_format_has_alpha(new->fb->format->format))
+			state->blending[index].alpha = true;
+		else
+			state->blending[index].alpha = false;
+
+		if (new->normalized_zpos > state->base.normalized_zpos)
+			state->blending[index].top = true;
+		else
+			state->blending[index].top = false;
 
 		/*
-		 * If any of the other planes is on top of this plane and uses
-		 * a format with an alpha component, mark this plane as being
-		 * dependent, meaning it's alpha value will be 1 minus the sum
-		 * of alpha components of the overlapping planes.
+		 * Missing framebuffer means that plane is disabled, in this
+		 * case mark B / C window as top to be able to differentiate
+		 * windows indices order in regards to zPos for the middle
+		 * window X / Y registers programming.
 		 */
-		if (p->index > tegra->index) {
-			if (__drm_format_has_alpha(new->fb->format->format))
-				state->dependent[index] = true;
-
-			/* keep track of the Z position */
-			zpos[index] = p->index;
-		}
+		if (!new->fb)
+			state->blending[index].top = (index == 1);
 	}
+}
+
+static int tegra_plane_setup_transparency(struct tegra_plane *tegra,
+					  struct tegra_plane_state *state)
+{
+	struct tegra_plane_state *tegra_state;
+	struct drm_plane_state *new;
+	struct drm_plane *plane;
+	int err;
 
 	/*
-	 * The region where three windows overlap is the intersection of the
-	 * two regions where two windows overlap. It contributes to the area
-	 * if any of the windows on top of it have an alpha component.
+	 * If planes zpos / transparency changed, sibling planes blending
+	 * state may require adjustment and in this case they will be included
+	 * into this atom commit, otherwise blending state is unchanged.
 	 */
-	for (i = 0; i < 2; i++)
-		state->dependent[2] = state->dependent[2] ||
-				      state->dependent[i];
+	err = tegra_plane_check_transparency(tegra, state);
+	if (err <= 0)
+		return err;
 
 	/*
-	 * However, if any of the windows on top of this window is opaque, it
-	 * will completely conceal this window within that area, so avoid the
-	 * window from contributing to the area.
+	 * All planes are now in the atomic state, walk them up and update
+	 * transparency state for each plane.
 	 */
-	for (i = 0; i < 2; i++) {
-		if (zpos[i] > tegra->index)
-			state->dependent[2] = state->dependent[2] &&
-					      state->dependent[i];
+	drm_for_each_plane(plane, tegra->base.dev) {
+		struct tegra_plane *p = to_tegra_plane(plane);
+
+		/* skip planes on different CRTCs */
+		if (p->dc != tegra->dc)
+			continue;
+
+		new = drm_atomic_get_new_plane_state(state->base.state, plane);
+		tegra_state = to_tegra_plane_state(new);
+
+		/*
+		 * There is no need to update blending state for the disabled
+		 * plane.
+		 */
+		if (new->fb)
+			tegra_plane_update_transparency(p, tegra_state);
 	}
+
+	return 0;
+}
+
+int tegra_plane_setup_legacy_state(struct tegra_plane *tegra,
+				   struct tegra_plane_state *state)
+{
+	int err;
+
+	err = tegra_plane_setup_opacity(tegra, state);
+	if (err < 0)
+		return err;
+
+	err = tegra_plane_setup_transparency(tegra, state);
+	if (err < 0)
+		return err;
+
+	return 0;
 }
diff --git a/drivers/gpu/drm/tegra/plane.h b/drivers/gpu/drm/tegra/plane.h
index 6938719e7e5d..e79e6b4a8e0a 100644
--- a/drivers/gpu/drm/tegra/plane.h
+++ b/drivers/gpu/drm/tegra/plane.h
@@ -34,6 +34,11 @@ static inline struct tegra_plane *to_tegra_plane(struct drm_plane *plane)
 	return container_of(plane, struct tegra_plane, base);
 }
 
+struct tegra_plane_legacy_blending_state {
+	bool alpha;
+	bool top;
+};
+
 struct tegra_plane_state {
 	struct drm_plane_state base;
 
@@ -41,9 +46,11 @@ struct tegra_plane_state {
 	u32 format;
 	u32 swap;
 
+	bool bottom_up;
+
 	/* used for legacy blending support only */
+	struct tegra_plane_legacy_blending_state blending[2];
 	bool opaque;
-	bool dependent[3];
 };
 
 static inline struct tegra_plane_state *
@@ -62,9 +69,7 @@ int tegra_plane_state_add(struct tegra_plane *plane,
 
 int tegra_plane_format(u32 fourcc, u32 *format, u32 *swap);
 bool tegra_plane_format_is_yuv(unsigned int format, bool *planar);
-bool tegra_plane_format_has_alpha(unsigned int format);
-int tegra_plane_format_get_alpha(unsigned int opaque, unsigned int *alpha);
-void tegra_plane_check_dependent(struct tegra_plane *tegra,
-				 struct tegra_plane_state *state);
+int tegra_plane_setup_legacy_state(struct tegra_plane *tegra,
+				   struct tegra_plane_state *state);
 
 #endif /* TEGRA_PLANE_H */
diff --git a/drivers/gpu/drm/tegra/rgb.c b/drivers/gpu/drm/tegra/rgb.c
index 78ec5193741d..28a78d3120bc 100644
--- a/drivers/gpu/drm/tegra/rgb.c
+++ b/drivers/gpu/drm/tegra/rgb.c
@@ -289,7 +289,7 @@ int tegra_dc_rgb_init(struct drm_device *drm, struct tegra_dc *dc)
 	drm_encoder_helper_add(&output->encoder,
 			       &tegra_rgb_encoder_helper_funcs);
 
-	drm_mode_connector_attach_encoder(&output->connector,
+	drm_connector_attach_encoder(&output->connector,
 					  &output->encoder);
 	drm_connector_register(&output->connector);
 
diff --git a/drivers/gpu/drm/tegra/sor.c b/drivers/gpu/drm/tegra/sor.c
index 7d2a955fc515..d7fe9f15def1 100644
--- a/drivers/gpu/drm/tegra/sor.c
+++ b/drivers/gpu/drm/tegra/sor.c
@@ -2622,7 +2622,7 @@ static int tegra_sor_init(struct host1x_client *client)
 			 encoder, NULL);
 	drm_encoder_helper_add(&sor->output.encoder, helpers);
 
-	drm_mode_connector_attach_encoder(&sor->output.connector,
+	drm_connector_attach_encoder(&sor->output.connector,
 					  &sor->output.encoder);
 	drm_connector_register(&sor->output.connector);
 
diff --git a/drivers/gpu/drm/tegra/vic.c b/drivers/gpu/drm/tegra/vic.c
index f5794dd49f3b..9f657a63b0bb 100644
--- a/drivers/gpu/drm/tegra/vic.c
+++ b/drivers/gpu/drm/tegra/vic.c
@@ -25,6 +25,7 @@
 
 struct vic_config {
 	const char *firmware;
+	unsigned int version;
 };
 
 struct vic {
@@ -264,18 +265,21 @@ static const struct tegra_drm_client_ops vic_ops = {
 
 static const struct vic_config vic_t124_config = {
 	.firmware = NVIDIA_TEGRA_124_VIC_FIRMWARE,
+	.version = 0x40,
 };
 
 #define NVIDIA_TEGRA_210_VIC_FIRMWARE "nvidia/tegra210/vic04_ucode.bin"
 
 static const struct vic_config vic_t210_config = {
 	.firmware = NVIDIA_TEGRA_210_VIC_FIRMWARE,
+	.version = 0x21,
 };
 
 #define NVIDIA_TEGRA_186_VIC_FIRMWARE "nvidia/tegra186/vic04_ucode.bin"
 
 static const struct vic_config vic_t186_config = {
 	.firmware = NVIDIA_TEGRA_186_VIC_FIRMWARE,
+	.version = 0x18,
 };
 
 static const struct of_device_id vic_match[] = {
@@ -342,6 +346,7 @@ static int vic_probe(struct platform_device *pdev)
 	vic->dev = dev;
 
 	INIT_LIST_HEAD(&vic->client.list);
+	vic->client.version = vic->config->version;
 	vic->client.ops = &vic_ops;
 
 	err = host1x_client_register(&vic->client.base);
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_crtc.c b/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
index 1b278a22c8b7..1067e702c22c 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_crtc.c
@@ -224,7 +224,7 @@ static void tilcdc_crtc_set_clk(struct drm_crtc *crtc)
 
 	ret = clk_set_rate(priv->clk, req_rate * clkdiv);
 	clk_rate = clk_get_rate(priv->clk);
-	if (ret < 0) {
+	if (ret < 0 || tilcdc_pclk_diff(req_rate, clk_rate) > 5) {
 		/*
 		 * If we fail to set the clock rate (some architectures don't
 		 * use the common clock framework yet and may not implement
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_drv.c b/drivers/gpu/drm/tilcdc/tilcdc_drv.c
index b8a5e4ed22e6..0fb300d41a09 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_drv.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_drv.c
@@ -378,7 +378,7 @@ static int tilcdc_init(struct drm_driver *ddrv, struct device *dev)
 	if (!priv->external_connector &&
 	    ((priv->num_encoders == 0) || (priv->num_connectors == 0))) {
 		dev_err(dev, "no encoders/connectors found\n");
-		ret = -ENXIO;
+		ret = -EPROBE_DEFER;
 		goto init_failed;
 	}
 
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_external.c b/drivers/gpu/drm/tilcdc/tilcdc_external.c
index d651bdd6597e..b4eaf9bc87f8 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_external.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_external.c
@@ -103,12 +103,11 @@ struct drm_connector *tilcdc_encoder_find_connector(struct drm_device *ddev,
 						    struct drm_encoder *encoder)
 {
 	struct drm_connector *connector;
-	int i;
 
-	list_for_each_entry(connector, &ddev->mode_config.connector_list, head)
-		for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++)
-			if (connector->encoder_ids[i] == encoder->base.id)
-				return connector;
+	list_for_each_entry(connector, &ddev->mode_config.connector_list, head) {
+		if (drm_connector_has_possible_encoder(connector, encoder))
+			return connector;
+	}
 
 	dev_err(ddev->dev, "No connector found for %s encoder (id %d)\n",
 		encoder->name, encoder->base.id);
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_panel.c b/drivers/gpu/drm/tilcdc/tilcdc_panel.c
index d616d64a6725..a1acab39d87f 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_panel.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_panel.c
@@ -223,7 +223,7 @@ static struct drm_connector *panel_connector_create(struct drm_device *dev,
 	connector->interlace_allowed = 0;
 	connector->doublescan_allowed = 0;
 
-	ret = drm_mode_connector_attach_encoder(connector, encoder);
+	ret = drm_connector_attach_encoder(connector, encoder);
 	if (ret)
 		goto fail;
 
diff --git a/drivers/gpu/drm/tilcdc/tilcdc_tfp410.c b/drivers/gpu/drm/tilcdc/tilcdc_tfp410.c
index c45cabb38db0..daebf1aa6b0a 100644
--- a/drivers/gpu/drm/tilcdc/tilcdc_tfp410.c
+++ b/drivers/gpu/drm/tilcdc/tilcdc_tfp410.c
@@ -173,7 +173,7 @@ static int tfp410_connector_get_modes(struct drm_connector *connector)
 
 	edid = drm_get_edid(connector, tfp410_connector->mod->i2c);
 
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 
 	if (edid) {
 		ret = drm_add_edid_modes(connector, edid);
@@ -240,7 +240,7 @@ static struct drm_connector *tfp410_connector_create(struct drm_device *dev,
 	connector->interlace_allowed = 0;
 	connector->doublescan_allowed = 0;
 
-	ret = drm_mode_connector_attach_encoder(connector, encoder);
+	ret = drm_connector_attach_encoder(connector, encoder);
 	if (ret)
 		goto fail;
 
diff --git a/drivers/gpu/drm/tinydrm/Kconfig b/drivers/gpu/drm/tinydrm/Kconfig
index 4592a5e3f20b..16f4b5c91f1b 100644
--- a/drivers/gpu/drm/tinydrm/Kconfig
+++ b/drivers/gpu/drm/tinydrm/Kconfig
@@ -20,6 +20,17 @@ config TINYDRM_ILI9225
 
 	  If M is selected the module will be called ili9225.
 
+config TINYDRM_ILI9341
+	tristate "DRM support for ILI9341 display panels"
+	depends on DRM_TINYDRM && SPI
+	depends on BACKLIGHT_CLASS_DEVICE
+	select TINYDRM_MIPI_DBI
+	help
+	  DRM driver for the following Ilitek ILI9341 panels:
+	  * YX240QV29-T 2.4" 240x320 TFT (Adafruit 2.4")
+
+	  If M is selected the module will be called ili9341.
+
 config TINYDRM_MI0283QT
 	tristate "DRM support for MI0283QT"
 	depends on DRM_TINYDRM && SPI
diff --git a/drivers/gpu/drm/tinydrm/Makefile b/drivers/gpu/drm/tinydrm/Makefile
index 49a111929724..14d99080665a 100644
--- a/drivers/gpu/drm/tinydrm/Makefile
+++ b/drivers/gpu/drm/tinydrm/Makefile
@@ -5,6 +5,7 @@ obj-$(CONFIG_TINYDRM_MIPI_DBI)		+= mipi-dbi.o
 
 # Displays
 obj-$(CONFIG_TINYDRM_ILI9225)		+= ili9225.o
+obj-$(CONFIG_TINYDRM_ILI9341)		+= ili9341.o
 obj-$(CONFIG_TINYDRM_MI0283QT)		+= mi0283qt.o
 obj-$(CONFIG_TINYDRM_REPAPER)		+= repaper.o
 obj-$(CONFIG_TINYDRM_ST7586)		+= st7586.o
diff --git a/drivers/gpu/drm/tinydrm/core/tinydrm-core.c b/drivers/gpu/drm/tinydrm/core/tinydrm-core.c
index 24a33bf862fa..19c7f70adfa5 100644
--- a/drivers/gpu/drm/tinydrm/core/tinydrm-core.c
+++ b/drivers/gpu/drm/tinydrm/core/tinydrm-core.c
@@ -204,7 +204,7 @@ static int tinydrm_register(struct tinydrm_device *tdev)
 	if (ret)
 		return ret;
 
-	ret = drm_fb_cma_fbdev_init_with_funcs(drm, 0, 0, tdev->fb_funcs);
+	ret = drm_fbdev_generic_setup(drm, 0);
 	if (ret)
 		DRM_ERROR("Failed to initialize fbdev: %d\n", ret);
 
@@ -214,7 +214,6 @@ static int tinydrm_register(struct tinydrm_device *tdev)
 static void tinydrm_unregister(struct tinydrm_device *tdev)
 {
 	drm_atomic_helper_shutdown(tdev->drm);
-	drm_fb_cma_fbdev_fini(tdev->drm);
 	drm_dev_unregister(tdev->drm);
 }
 
diff --git a/drivers/gpu/drm/tinydrm/ili9225.c b/drivers/gpu/drm/tinydrm/ili9225.c
index 841c69aba059..455fefe012f5 100644
--- a/drivers/gpu/drm/tinydrm/ili9225.c
+++ b/drivers/gpu/drm/tinydrm/ili9225.c
@@ -368,7 +368,6 @@ static struct drm_driver ili9225_driver = {
 				  DRIVER_ATOMIC,
 	.fops			= &ili9225_fops,
 	TINYDRM_GEM_DRIVER_OPS,
-	.lastclose		= drm_fb_helper_lastclose,
 	.name			= "ili9225",
 	.desc			= "Ilitek ILI9225",
 	.date			= "20171106",
diff --git a/drivers/gpu/drm/tinydrm/ili9341.c b/drivers/gpu/drm/tinydrm/ili9341.c
new file mode 100644
index 000000000000..6701037749a7
--- /dev/null
+++ b/drivers/gpu/drm/tinydrm/ili9341.c
@@ -0,0 +1,232 @@
+// SPDX-License-Identifier: GPL-2.0+
+/*
+ * DRM driver for Ilitek ILI9341 panels
+ *
+ * Copyright 2018 David Lechner <david@lechnology.com>
+ *
+ * Based on mi0283qt.c:
+ * Copyright 2016 Noralf Trønnes
+ */
+
+#include <linux/backlight.h>
+#include <linux/delay.h>
+#include <linux/gpio/consumer.h>
+#include <linux/module.h>
+#include <linux/property.h>
+#include <linux/spi/spi.h>
+
+#include <drm/drm_fb_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
+#include <drm/drm_modeset_helper.h>
+#include <drm/tinydrm/mipi-dbi.h>
+#include <drm/tinydrm/tinydrm-helpers.h>
+#include <video/mipi_display.h>
+
+#define ILI9341_FRMCTR1		0xb1
+#define ILI9341_DISCTRL		0xb6
+#define ILI9341_ETMOD		0xb7
+
+#define ILI9341_PWCTRL1		0xc0
+#define ILI9341_PWCTRL2		0xc1
+#define ILI9341_VMCTRL1		0xc5
+#define ILI9341_VMCTRL2		0xc7
+#define ILI9341_PWCTRLA		0xcb
+#define ILI9341_PWCTRLB		0xcf
+
+#define ILI9341_PGAMCTRL	0xe0
+#define ILI9341_NGAMCTRL	0xe1
+#define ILI9341_DTCTRLA		0xe8
+#define ILI9341_DTCTRLB		0xea
+#define ILI9341_PWRSEQ		0xed
+
+#define ILI9341_EN3GAM		0xf2
+#define ILI9341_PUMPCTRL	0xf7
+
+#define ILI9341_MADCTL_BGR	BIT(3)
+#define ILI9341_MADCTL_MV	BIT(5)
+#define ILI9341_MADCTL_MX	BIT(6)
+#define ILI9341_MADCTL_MY	BIT(7)
+
+static void yx240qv29_enable(struct drm_simple_display_pipe *pipe,
+			     struct drm_crtc_state *crtc_state,
+			     struct drm_plane_state *plane_state)
+{
+	struct tinydrm_device *tdev = pipe_to_tinydrm(pipe);
+	struct mipi_dbi *mipi = mipi_dbi_from_tinydrm(tdev);
+	u8 addr_mode;
+	int ret;
+
+	DRM_DEBUG_KMS("\n");
+
+	ret = mipi_dbi_poweron_conditional_reset(mipi);
+	if (ret < 0)
+		return;
+	if (ret == 1)
+		goto out_enable;
+
+	mipi_dbi_command(mipi, MIPI_DCS_SET_DISPLAY_OFF);
+
+	mipi_dbi_command(mipi, ILI9341_PWCTRLB, 0x00, 0xc1, 0x30);
+	mipi_dbi_command(mipi, ILI9341_PWRSEQ, 0x64, 0x03, 0x12, 0x81);
+	mipi_dbi_command(mipi, ILI9341_DTCTRLA, 0x85, 0x00, 0x78);
+	mipi_dbi_command(mipi, ILI9341_PWCTRLA, 0x39, 0x2c, 0x00, 0x34, 0x02);
+	mipi_dbi_command(mipi, ILI9341_PUMPCTRL, 0x20);
+	mipi_dbi_command(mipi, ILI9341_DTCTRLB, 0x00, 0x00);
+
+	/* Power Control */
+	mipi_dbi_command(mipi, ILI9341_PWCTRL1, 0x23);
+	mipi_dbi_command(mipi, ILI9341_PWCTRL2, 0x10);
+	/* VCOM */
+	mipi_dbi_command(mipi, ILI9341_VMCTRL1, 0x3e, 0x28);
+	mipi_dbi_command(mipi, ILI9341_VMCTRL2, 0x86);
+
+	/* Memory Access Control */
+	mipi_dbi_command(mipi, MIPI_DCS_SET_PIXEL_FORMAT, MIPI_DCS_PIXEL_FMT_16BIT);
+
+	/* Frame Rate */
+	mipi_dbi_command(mipi, ILI9341_FRMCTR1, 0x00, 0x1b);
+
+	/* Gamma */
+	mipi_dbi_command(mipi, ILI9341_EN3GAM, 0x00);
+	mipi_dbi_command(mipi, MIPI_DCS_SET_GAMMA_CURVE, 0x01);
+	mipi_dbi_command(mipi, ILI9341_PGAMCTRL,
+			 0x0f, 0x31, 0x2b, 0x0c, 0x0e, 0x08, 0x4e, 0xf1,
+			 0x37, 0x07, 0x10, 0x03, 0x0e, 0x09, 0x00);
+	mipi_dbi_command(mipi, ILI9341_NGAMCTRL,
+			 0x00, 0x0e, 0x14, 0x03, 0x11, 0x07, 0x31, 0xc1,
+			 0x48, 0x08, 0x0f, 0x0c, 0x31, 0x36, 0x0f);
+
+	/* DDRAM */
+	mipi_dbi_command(mipi, ILI9341_ETMOD, 0x07);
+
+	/* Display */
+	mipi_dbi_command(mipi, ILI9341_DISCTRL, 0x08, 0x82, 0x27, 0x00);
+	mipi_dbi_command(mipi, MIPI_DCS_EXIT_SLEEP_MODE);
+	msleep(100);
+
+	mipi_dbi_command(mipi, MIPI_DCS_SET_DISPLAY_ON);
+	msleep(100);
+
+out_enable:
+	switch (mipi->rotation) {
+	default:
+		addr_mode = ILI9341_MADCTL_MX;
+		break;
+	case 90:
+		addr_mode = ILI9341_MADCTL_MV;
+		break;
+	case 180:
+		addr_mode = ILI9341_MADCTL_MY;
+		break;
+	case 270:
+		addr_mode = ILI9341_MADCTL_MV | ILI9341_MADCTL_MY |
+			    ILI9341_MADCTL_MX;
+		break;
+	}
+	addr_mode |= ILI9341_MADCTL_BGR;
+	mipi_dbi_command(mipi, MIPI_DCS_SET_ADDRESS_MODE, addr_mode);
+	mipi_dbi_enable_flush(mipi, crtc_state, plane_state);
+}
+
+static const struct drm_simple_display_pipe_funcs ili9341_pipe_funcs = {
+	.enable = yx240qv29_enable,
+	.disable = mipi_dbi_pipe_disable,
+	.update = tinydrm_display_pipe_update,
+	.prepare_fb = drm_gem_fb_simple_display_pipe_prepare_fb,
+};
+
+static const struct drm_display_mode yx240qv29_mode = {
+	TINYDRM_MODE(240, 320, 37, 49),
+};
+
+DEFINE_DRM_GEM_CMA_FOPS(ili9341_fops);
+
+static struct drm_driver ili9341_driver = {
+	.driver_features	= DRIVER_GEM | DRIVER_MODESET | DRIVER_PRIME | DRIVER_ATOMIC,
+	.fops			= &ili9341_fops,
+	TINYDRM_GEM_DRIVER_OPS,
+	.debugfs_init		= mipi_dbi_debugfs_init,
+	.name			= "ili9341",
+	.desc			= "Ilitek ILI9341",
+	.date			= "20180514",
+	.major			= 1,
+	.minor			= 0,
+};
+
+static const struct of_device_id ili9341_of_match[] = {
+	{ .compatible = "adafruit,yx240qv29" },
+	{ }
+};
+MODULE_DEVICE_TABLE(of, ili9341_of_match);
+
+static const struct spi_device_id ili9341_id[] = {
+	{ "yx240qv29", 0 },
+	{ }
+};
+MODULE_DEVICE_TABLE(spi, ili9341_id);
+
+static int ili9341_probe(struct spi_device *spi)
+{
+	struct device *dev = &spi->dev;
+	struct mipi_dbi *mipi;
+	struct gpio_desc *dc;
+	u32 rotation = 0;
+	int ret;
+
+	mipi = devm_kzalloc(dev, sizeof(*mipi), GFP_KERNEL);
+	if (!mipi)
+		return -ENOMEM;
+
+	mipi->reset = devm_gpiod_get_optional(dev, "reset", GPIOD_OUT_HIGH);
+	if (IS_ERR(mipi->reset)) {
+		DRM_DEV_ERROR(dev, "Failed to get gpio 'reset'\n");
+		return PTR_ERR(mipi->reset);
+	}
+
+	dc = devm_gpiod_get_optional(dev, "dc", GPIOD_OUT_LOW);
+	if (IS_ERR(dc)) {
+		DRM_DEV_ERROR(dev, "Failed to get gpio 'dc'\n");
+		return PTR_ERR(dc);
+	}
+
+	mipi->backlight = devm_of_find_backlight(dev);
+	if (IS_ERR(mipi->backlight))
+		return PTR_ERR(mipi->backlight);
+
+	device_property_read_u32(dev, "rotation", &rotation);
+
+	ret = mipi_dbi_spi_init(spi, mipi, dc);
+	if (ret)
+		return ret;
+
+	ret = mipi_dbi_init(&spi->dev, mipi, &ili9341_pipe_funcs,
+			    &ili9341_driver, &yx240qv29_mode, rotation);
+	if (ret)
+		return ret;
+
+	spi_set_drvdata(spi, mipi);
+
+	return devm_tinydrm_register(&mipi->tinydrm);
+}
+
+static void ili9341_shutdown(struct spi_device *spi)
+{
+	struct mipi_dbi *mipi = spi_get_drvdata(spi);
+
+	tinydrm_shutdown(&mipi->tinydrm);
+}
+
+static struct spi_driver ili9341_spi_driver = {
+	.driver = {
+		.name = "ili9341",
+		.of_match_table = ili9341_of_match,
+	},
+	.id_table = ili9341_id,
+	.probe = ili9341_probe,
+	.shutdown = ili9341_shutdown,
+};
+module_spi_driver(ili9341_spi_driver);
+
+MODULE_DESCRIPTION("Ilitek ILI9341 DRM driver");
+MODULE_AUTHOR("David Lechner <david@lechnology.com>");
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/tinydrm/mi0283qt.c b/drivers/gpu/drm/tinydrm/mi0283qt.c
index d5ef65179c16..d7bb4c5e6657 100644
--- a/drivers/gpu/drm/tinydrm/mi0283qt.c
+++ b/drivers/gpu/drm/tinydrm/mi0283qt.c
@@ -85,24 +85,6 @@ static void mi0283qt_enable(struct drm_simple_display_pipe *pipe,
 	/* Memory Access Control */
 	mipi_dbi_command(mipi, MIPI_DCS_SET_PIXEL_FORMAT, MIPI_DCS_PIXEL_FMT_16BIT);
 
-	switch (mipi->rotation) {
-	default:
-		addr_mode = ILI9341_MADCTL_MV | ILI9341_MADCTL_MY |
-			    ILI9341_MADCTL_MX;
-		break;
-	case 90:
-		addr_mode = ILI9341_MADCTL_MY;
-		break;
-	case 180:
-		addr_mode = ILI9341_MADCTL_MV;
-		break;
-	case 270:
-		addr_mode = ILI9341_MADCTL_MX;
-		break;
-	}
-	addr_mode |= ILI9341_MADCTL_BGR;
-	mipi_dbi_command(mipi, MIPI_DCS_SET_ADDRESS_MODE, addr_mode);
-
 	/* Frame Rate */
 	mipi_dbi_command(mipi, ILI9341_FRMCTR1, 0x00, 0x1b);
 
@@ -128,6 +110,29 @@ static void mi0283qt_enable(struct drm_simple_display_pipe *pipe,
 	msleep(100);
 
 out_enable:
+	/* The PiTFT (ili9340) has a hardware reset circuit that
+	 * resets only on power-on and not on each reboot through
+	 * a gpio like the rpi-display does.
+	 * As a result, we need to always apply the rotation value
+	 * regardless of the display "on/off" state.
+	 */
+	switch (mipi->rotation) {
+	default:
+		addr_mode = ILI9341_MADCTL_MV | ILI9341_MADCTL_MY |
+			    ILI9341_MADCTL_MX;
+		break;
+	case 90:
+		addr_mode = ILI9341_MADCTL_MY;
+		break;
+	case 180:
+		addr_mode = ILI9341_MADCTL_MV;
+		break;
+	case 270:
+		addr_mode = ILI9341_MADCTL_MX;
+		break;
+	}
+	addr_mode |= ILI9341_MADCTL_BGR;
+	mipi_dbi_command(mipi, MIPI_DCS_SET_ADDRESS_MODE, addr_mode);
 	mipi_dbi_enable_flush(mipi, crtc_state, plane_state);
 }
 
@@ -149,7 +154,6 @@ static struct drm_driver mi0283qt_driver = {
 				  DRIVER_ATOMIC,
 	.fops			= &mi0283qt_fops,
 	TINYDRM_GEM_DRIVER_OPS,
-	.lastclose		= drm_fb_helper_lastclose,
 	.debugfs_init		= mipi_dbi_debugfs_init,
 	.name			= "mi0283qt",
 	.desc			= "Multi-Inno MI0283QT",
diff --git a/drivers/gpu/drm/tinydrm/mipi-dbi.c b/drivers/gpu/drm/tinydrm/mipi-dbi.c
index 4d1fb31a781f..cb3441e51d5f 100644
--- a/drivers/gpu/drm/tinydrm/mipi-dbi.c
+++ b/drivers/gpu/drm/tinydrm/mipi-dbi.c
@@ -260,6 +260,8 @@ static const struct drm_framebuffer_funcs mipi_dbi_fb_funcs = {
 /**
  * mipi_dbi_enable_flush - MIPI DBI enable helper
  * @mipi: MIPI DBI structure
+ * @crtc_state: CRTC state
+ * @plane_state: Plane state
  *
  * This function sets &mipi_dbi->enabled, flushes the whole framebuffer and
  * enables the backlight. Drivers can use this in their
diff --git a/drivers/gpu/drm/tinydrm/repaper.c b/drivers/gpu/drm/tinydrm/repaper.c
index 1ee6855212a0..50a1d4216ce7 100644
--- a/drivers/gpu/drm/tinydrm/repaper.c
+++ b/drivers/gpu/drm/tinydrm/repaper.c
@@ -548,7 +548,7 @@ static int repaper_fb_dirty(struct drm_framebuffer *fb,
 	DRM_DEBUG("Flushing [FB:%d] st=%ums\n", fb->base.id,
 		  epd->factored_stage_time);
 
-	buf = kmalloc(fb->width * fb->height, GFP_KERNEL);
+	buf = kmalloc_array(fb->width, fb->height, GFP_KERNEL);
 	if (!buf)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/tinydrm/st7586.c b/drivers/gpu/drm/tinydrm/st7586.c
index 5c29e3803ecb..2fcbc3067d71 100644
--- a/drivers/gpu/drm/tinydrm/st7586.c
+++ b/drivers/gpu/drm/tinydrm/st7586.c
@@ -304,7 +304,6 @@ static struct drm_driver st7586_driver = {
 				  DRIVER_ATOMIC,
 	.fops			= &st7586_fops,
 	TINYDRM_GEM_DRIVER_OPS,
-	.lastclose		= drm_fb_helper_lastclose,
 	.debugfs_init		= mipi_dbi_debugfs_init,
 	.name			= "st7586",
 	.desc			= "Sitronix ST7586",
diff --git a/drivers/gpu/drm/tinydrm/st7735r.c b/drivers/gpu/drm/tinydrm/st7735r.c
index 6c7b15c9da4f..3081bc57c116 100644
--- a/drivers/gpu/drm/tinydrm/st7735r.c
+++ b/drivers/gpu/drm/tinydrm/st7735r.c
@@ -120,7 +120,6 @@ static struct drm_driver st7735r_driver = {
 				  DRIVER_ATOMIC,
 	.fops			= &st7735r_fops,
 	TINYDRM_GEM_DRIVER_OPS,
-	.lastclose		= drm_fb_helper_lastclose,
 	.debugfs_init		= mipi_dbi_debugfs_init,
 	.name			= "st7735r",
 	.desc			= "Sitronix ST7735R",
diff --git a/drivers/gpu/drm/ttm/ttm_agp_backend.c b/drivers/gpu/drm/ttm/ttm_agp_backend.c
index 7c2485fe88d8..ea4d59eb8966 100644
--- a/drivers/gpu/drm/ttm/ttm_agp_backend.c
+++ b/drivers/gpu/drm/ttm/ttm_agp_backend.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index 98e06f8bf23b..7c484729f9b2 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
@@ -286,12 +287,9 @@ static int ttm_bo_handle_move_mem(struct ttm_buffer_object *bo,
 
 	if (ret) {
 		if (bdev->driver->move_notify) {
-			struct ttm_mem_reg tmp_mem = *mem;
-			*mem = bo->mem;
-			bo->mem = tmp_mem;
+			swap(*mem, bo->mem);
 			bdev->driver->move_notify(bo, false, mem);
-			bo->mem = *mem;
-			*mem = tmp_mem;
+			swap(*mem, bo->mem);
 		}
 
 		goto out_err;
@@ -589,12 +587,18 @@ static void ttm_bo_release(struct kref *kref)
 	kref_put(&bo->list_kref, ttm_bo_release_list);
 }
 
+void ttm_bo_put(struct ttm_buffer_object *bo)
+{
+	kref_put(&bo->kref, ttm_bo_release);
+}
+EXPORT_SYMBOL(ttm_bo_put);
+
 void ttm_bo_unref(struct ttm_buffer_object **p_bo)
 {
 	struct ttm_buffer_object *bo = *p_bo;
 
 	*p_bo = NULL;
-	kref_put(&bo->kref, ttm_bo_release);
+	ttm_bo_put(bo);
 }
 EXPORT_SYMBOL(ttm_bo_unref);
 
@@ -1175,7 +1179,6 @@ int ttm_bo_init_reserved(struct ttm_bo_device *bdev,
 	reservation_object_init(&bo->ttm_resv);
 	atomic_inc(&bo->bdev->glob->bo_count);
 	drm_vma_node_reset(&bo->vma_node);
-	bo->priority = 0;
 
 	/*
 	 * For ttm_bo_type_device buffers, allocate
@@ -1201,7 +1204,7 @@ int ttm_bo_init_reserved(struct ttm_bo_device *bdev,
 		if (!resv)
 			ttm_bo_unreserve(bo);
 
-		ttm_bo_unref(&bo);
+		ttm_bo_put(bo);
 		return ret;
 	}
 
diff --git a/drivers/gpu/drm/ttm/ttm_bo_manager.c b/drivers/gpu/drm/ttm/ttm_bo_manager.c
index a7c232dc39cb..18d3debcc949 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_manager.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_manager.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2007-2010 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_bo_util.c b/drivers/gpu/drm/ttm/ttm_bo_util.c
index 2ebbae6067ab..046a6dda690a 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_util.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_util.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2007-2009 VMware, Inc., Palo Alto, CA., USA
@@ -39,6 +40,11 @@
 #include <linux/module.h>
 #include <linux/reservation.h>
 
+struct ttm_transfer_obj {
+	struct ttm_buffer_object base;
+	struct ttm_buffer_object *bo;
+};
+
 void ttm_bo_free_old_node(struct ttm_buffer_object *bo)
 {
 	ttm_bo_mem_put(bo, &bo->mem);
@@ -454,7 +460,11 @@ EXPORT_SYMBOL(ttm_bo_move_memcpy);
 
 static void ttm_transfered_destroy(struct ttm_buffer_object *bo)
 {
-	kfree(bo);
+	struct ttm_transfer_obj *fbo;
+
+	fbo = container_of(bo, struct ttm_transfer_obj, base);
+	ttm_bo_put(fbo->bo);
+	kfree(fbo);
 }
 
 /**
@@ -475,14 +485,16 @@ static void ttm_transfered_destroy(struct ttm_buffer_object *bo)
 static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo,
 				      struct ttm_buffer_object **new_obj)
 {
-	struct ttm_buffer_object *fbo;
+	struct ttm_transfer_obj *fbo;
 	int ret;
 
 	fbo = kmalloc(sizeof(*fbo), GFP_KERNEL);
 	if (!fbo)
 		return -ENOMEM;
 
-	*fbo = *bo;
+	ttm_bo_get(bo);
+	fbo->base = *bo;
+	fbo->bo = bo;
 
 	/**
 	 * Fix up members that we shouldn't copy directly:
@@ -490,25 +502,25 @@ static int ttm_buffer_object_transfer(struct ttm_buffer_object *bo,
 	 */
 
 	atomic_inc(&bo->bdev->glob->bo_count);
-	INIT_LIST_HEAD(&fbo->ddestroy);
-	INIT_LIST_HEAD(&fbo->lru);
-	INIT_LIST_HEAD(&fbo->swap);
-	INIT_LIST_HEAD(&fbo->io_reserve_lru);
-	mutex_init(&fbo->wu_mutex);
-	fbo->moving = NULL;
-	drm_vma_node_reset(&fbo->vma_node);
-	atomic_set(&fbo->cpu_writers, 0);
-
-	kref_init(&fbo->list_kref);
-	kref_init(&fbo->kref);
-	fbo->destroy = &ttm_transfered_destroy;
-	fbo->acc_size = 0;
-	fbo->resv = &fbo->ttm_resv;
-	reservation_object_init(fbo->resv);
-	ret = reservation_object_trylock(fbo->resv);
+	INIT_LIST_HEAD(&fbo->base.ddestroy);
+	INIT_LIST_HEAD(&fbo->base.lru);
+	INIT_LIST_HEAD(&fbo->base.swap);
+	INIT_LIST_HEAD(&fbo->base.io_reserve_lru);
+	mutex_init(&fbo->base.wu_mutex);
+	fbo->base.moving = NULL;
+	drm_vma_node_reset(&fbo->base.vma_node);
+	atomic_set(&fbo->base.cpu_writers, 0);
+
+	kref_init(&fbo->base.list_kref);
+	kref_init(&fbo->base.kref);
+	fbo->base.destroy = &ttm_transfered_destroy;
+	fbo->base.acc_size = 0;
+	fbo->base.resv = &fbo->base.ttm_resv;
+	reservation_object_init(fbo->base.resv);
+	ret = reservation_object_trylock(fbo->base.resv);
 	WARN_ON(!ret);
 
-	*new_obj = fbo;
+	*new_obj = &fbo->base;
 	return 0;
 }
 
@@ -719,7 +731,7 @@ int ttm_bo_move_accel_cleanup(struct ttm_buffer_object *bo,
 			bo->ttm = NULL;
 
 		ttm_bo_unreserve(ghost_obj);
-		ttm_bo_unref(&ghost_obj);
+		ttm_bo_put(ghost_obj);
 	}
 
 	*old_mem = *new_mem;
@@ -775,7 +787,7 @@ int ttm_bo_pipeline_move(struct ttm_buffer_object *bo,
 			bo->ttm = NULL;
 
 		ttm_bo_unreserve(ghost_obj);
-		ttm_bo_unref(&ghost_obj);
+		ttm_bo_put(ghost_obj);
 
 	} else if (from->flags & TTM_MEMTYPE_FLAG_FIXED) {
 
@@ -840,7 +852,7 @@ int ttm_bo_pipeline_gutting(struct ttm_buffer_object *bo)
 	bo->ttm = NULL;
 
 	ttm_bo_unreserve(ghost);
-	ttm_bo_unref(&ghost);
+	ttm_bo_put(ghost);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/ttm/ttm_bo_vm.c b/drivers/gpu/drm/ttm/ttm_bo_vm.c
index 8eba95b3c737..6fe91c1b692d 100644
--- a/drivers/gpu/drm/ttm/ttm_bo_vm.c
+++ b/drivers/gpu/drm/ttm/ttm_bo_vm.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
@@ -43,10 +44,11 @@
 
 #define TTM_BO_VM_NUM_PREFAULT 16
 
-static int ttm_bo_vm_fault_idle(struct ttm_buffer_object *bo,
+static vm_fault_t ttm_bo_vm_fault_idle(struct ttm_buffer_object *bo,
 				struct vm_fault *vmf)
 {
-	int ret = 0;
+	vm_fault_t ret = 0;
+	int err = 0;
 
 	if (likely(!bo->moving))
 		goto out_unlock;
@@ -66,20 +68,20 @@ static int ttm_bo_vm_fault_idle(struct ttm_buffer_object *bo,
 		if (vmf->flags & FAULT_FLAG_RETRY_NOWAIT)
 			goto out_unlock;
 
-		ttm_bo_reference(bo);
+		ttm_bo_get(bo);
 		up_read(&vmf->vma->vm_mm->mmap_sem);
 		(void) dma_fence_wait(bo->moving, true);
 		ttm_bo_unreserve(bo);
-		ttm_bo_unref(&bo);
+		ttm_bo_put(bo);
 		goto out_unlock;
 	}
 
 	/*
 	 * Ordinary wait.
 	 */
-	ret = dma_fence_wait(bo->moving, true);
-	if (unlikely(ret != 0)) {
-		ret = (ret != -ERESTARTSYS) ? VM_FAULT_SIGBUS :
+	err = dma_fence_wait(bo->moving, true);
+	if (unlikely(err != 0)) {
+		ret = (err != -ERESTARTSYS) ? VM_FAULT_SIGBUS :
 			VM_FAULT_NOPAGE;
 		goto out_unlock;
 	}
@@ -104,7 +106,7 @@ static unsigned long ttm_bo_io_mem_pfn(struct ttm_buffer_object *bo,
 		+ page_offset;
 }
 
-static int ttm_bo_vm_fault(struct vm_fault *vmf)
+static vm_fault_t ttm_bo_vm_fault(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 	struct ttm_buffer_object *bo = (struct ttm_buffer_object *)
@@ -115,8 +117,9 @@ static int ttm_bo_vm_fault(struct vm_fault *vmf)
 	unsigned long pfn;
 	struct ttm_tt *ttm = NULL;
 	struct page *page;
-	int ret;
+	int err;
 	int i;
+	vm_fault_t ret = VM_FAULT_NOPAGE;
 	unsigned long address = vmf->address;
 	struct ttm_mem_type_manager *man =
 		&bdev->man[bo->mem.mem_type];
@@ -128,17 +131,17 @@ static int ttm_bo_vm_fault(struct vm_fault *vmf)
 	 * for reserve, and if it fails, retry the fault after waiting
 	 * for the buffer to become unreserved.
 	 */
-	ret = ttm_bo_reserve(bo, true, true, NULL);
-	if (unlikely(ret != 0)) {
-		if (ret != -EBUSY)
+	err = ttm_bo_reserve(bo, true, true, NULL);
+	if (unlikely(err != 0)) {
+		if (err != -EBUSY)
 			return VM_FAULT_NOPAGE;
 
 		if (vmf->flags & FAULT_FLAG_ALLOW_RETRY) {
 			if (!(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) {
-				ttm_bo_reference(bo);
+				ttm_bo_get(bo);
 				up_read(&vmf->vma->vm_mm->mmap_sem);
 				(void) ttm_bo_wait_unreserved(bo);
-				ttm_bo_unref(&bo);
+				ttm_bo_put(bo);
 			}
 
 			return VM_FAULT_RETRY;
@@ -162,8 +165,8 @@ static int ttm_bo_vm_fault(struct vm_fault *vmf)
 	}
 
 	if (bdev->driver->fault_reserve_notify) {
-		ret = bdev->driver->fault_reserve_notify(bo);
-		switch (ret) {
+		err = bdev->driver->fault_reserve_notify(bo);
+		switch (err) {
 		case 0:
 			break;
 		case -EBUSY:
@@ -191,13 +194,13 @@ static int ttm_bo_vm_fault(struct vm_fault *vmf)
 		goto out_unlock;
 	}
 
-	ret = ttm_mem_io_lock(man, true);
-	if (unlikely(ret != 0)) {
+	err = ttm_mem_io_lock(man, true);
+	if (unlikely(err != 0)) {
 		ret = VM_FAULT_NOPAGE;
 		goto out_unlock;
 	}
-	ret = ttm_mem_io_reserve_vm(bo);
-	if (unlikely(ret != 0)) {
+	err = ttm_mem_io_reserve_vm(bo);
+	if (unlikely(err != 0)) {
 		ret = VM_FAULT_SIGBUS;
 		goto out_io_unlock;
 	}
@@ -265,23 +268,20 @@ static int ttm_bo_vm_fault(struct vm_fault *vmf)
 		}
 
 		if (vma->vm_flags & VM_MIXEDMAP)
-			ret = vm_insert_mixed(&cvma, address,
+			ret = vmf_insert_mixed(&cvma, address,
 					__pfn_to_pfn_t(pfn, PFN_DEV));
 		else
-			ret = vm_insert_pfn(&cvma, address, pfn);
+			ret = vmf_insert_pfn(&cvma, address, pfn);
 
 		/*
 		 * Somebody beat us to this PTE or prefaulting to
 		 * an already populated PTE, or prefaulting error.
 		 */
 
-		if (unlikely((ret == -EBUSY) || (ret != 0 && i > 0)))
+		if (unlikely((ret == VM_FAULT_NOPAGE && i > 0)))
 			break;
-		else if (unlikely(ret != 0)) {
-			ret =
-			    (ret == -ENOMEM) ? VM_FAULT_OOM : VM_FAULT_SIGBUS;
+		else if (unlikely(ret & VM_FAULT_ERROR))
 			goto out_io_unlock;
-		}
 
 		address += PAGE_SIZE;
 		if (unlikely(++page_offset >= page_last))
@@ -302,14 +302,14 @@ static void ttm_bo_vm_open(struct vm_area_struct *vma)
 
 	WARN_ON(bo->bdev->dev_mapping != vma->vm_file->f_mapping);
 
-	(void)ttm_bo_reference(bo);
+	ttm_bo_get(bo);
 }
 
 static void ttm_bo_vm_close(struct vm_area_struct *vma)
 {
 	struct ttm_buffer_object *bo = (struct ttm_buffer_object *)vma->vm_private_data;
 
-	ttm_bo_unref(&bo);
+	ttm_bo_put(bo);
 	vma->vm_private_data = NULL;
 }
 
@@ -461,7 +461,7 @@ int ttm_bo_mmap(struct file *filp, struct vm_area_struct *vma,
 	vma->vm_flags |= VM_IO | VM_DONTEXPAND | VM_DONTDUMP;
 	return 0;
 out_unref:
-	ttm_bo_unref(&bo);
+	ttm_bo_put(bo);
 	return ret;
 }
 EXPORT_SYMBOL(ttm_bo_mmap);
@@ -471,8 +471,10 @@ int ttm_fbdev_mmap(struct vm_area_struct *vma, struct ttm_buffer_object *bo)
 	if (vma->vm_pgoff != 0)
 		return -EACCES;
 
+	ttm_bo_get(bo);
+
 	vma->vm_ops = &ttm_bo_vm_ops;
-	vma->vm_private_data = ttm_bo_reference(bo);
+	vma->vm_private_data = bo;
 	vma->vm_flags |= VM_MIXEDMAP;
 	vma->vm_flags |= VM_IO | VM_DONTEXPAND;
 	return 0;
diff --git a/drivers/gpu/drm/ttm/ttm_execbuf_util.c b/drivers/gpu/drm/ttm/ttm_execbuf_util.c
index 3dca206e85f7..e73ae0d22897 100644
--- a/drivers/gpu/drm/ttm/ttm_execbuf_util.c
+++ b/drivers/gpu/drm/ttm/ttm_execbuf_util.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_lock.c b/drivers/gpu/drm/ttm/ttm_lock.c
index 913f4318cdc0..20694b8a01ca 100644
--- a/drivers/gpu/drm/ttm/ttm_lock.c
+++ b/drivers/gpu/drm/ttm/ttm_lock.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2007-2009 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_memory.c b/drivers/gpu/drm/ttm/ttm_memory.c
index 27856c55dc84..450387c92b63 100644
--- a/drivers/gpu/drm/ttm/ttm_memory.c
+++ b/drivers/gpu/drm/ttm/ttm_memory.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_module.c b/drivers/gpu/drm/ttm/ttm_module.c
index 66fc6395eb54..6ff40c041d79 100644
--- a/drivers/gpu/drm/ttm/ttm_module.c
+++ b/drivers/gpu/drm/ttm/ttm_module.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_object.c b/drivers/gpu/drm/ttm/ttm_object.c
index 1aa2baa83959..74f1b1eb1f8e 100644
--- a/drivers/gpu/drm/ttm/ttm_object.c
+++ b/drivers/gpu/drm/ttm/ttm_object.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2009-2013 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc.c b/drivers/gpu/drm/ttm/ttm_page_alloc.c
index f0481b7b60c5..6e2d1300b457 100644
--- a/drivers/gpu/drm/ttm/ttm_page_alloc.c
+++ b/drivers/gpu/drm/ttm/ttm_page_alloc.c
@@ -348,8 +348,9 @@ static int ttm_page_pool_free(struct ttm_page_pool *pool, unsigned nr_free,
 	if (use_static)
 		pages_to_free = static_buf;
 	else
-		pages_to_free = kmalloc(npages_to_free * sizeof(struct page *),
-					GFP_KERNEL);
+		pages_to_free = kmalloc_array(npages_to_free,
+					      sizeof(struct page *),
+					      GFP_KERNEL);
 	if (!pages_to_free) {
 		pr_debug("Failed to allocate memory for pool free operation\n");
 		return 0;
@@ -547,7 +548,8 @@ static int ttm_alloc_new_pages(struct list_head *pages, gfp_t gfp_flags,
 	unsigned max_cpages = min(count << order, (unsigned)NUM_PAGES_TO_ALLOC);
 
 	/* allocate array for page caching change */
-	caching_array = kmalloc(max_cpages*sizeof(struct page *), GFP_KERNEL);
+	caching_array = kmalloc_array(max_cpages, sizeof(struct page *),
+				      GFP_KERNEL);
 
 	if (!caching_array) {
 		pr_debug("Unable to allocate table for new pages\n");
@@ -910,7 +912,8 @@ static int ttm_get_pages(struct page **pages, unsigned npages, int flags,
 			while (npages >= HPAGE_PMD_NR) {
 				gfp_t huge_flags = gfp_flags;
 
-				huge_flags |= GFP_TRANSHUGE;
+				huge_flags |= GFP_TRANSHUGE_LIGHT | __GFP_NORETRY |
+					__GFP_KSWAPD_RECLAIM;
 				huge_flags &= ~__GFP_MOVABLE;
 				huge_flags &= ~__GFP_COMP;
 				p = alloc_pages(huge_flags, HPAGE_PMD_ORDER);
@@ -1027,11 +1030,15 @@ int ttm_page_alloc_init(struct ttm_mem_global *glob, unsigned max_pages)
 				  GFP_USER | GFP_DMA32, "uc dma", 0);
 
 	ttm_page_pool_init_locked(&_manager->wc_pool_huge,
-				  GFP_TRANSHUGE	& ~(__GFP_MOVABLE | __GFP_COMP),
+				  (GFP_TRANSHUGE_LIGHT | __GFP_NORETRY |
+				   __GFP_KSWAPD_RECLAIM) &
+				  ~(__GFP_MOVABLE | __GFP_COMP),
 				  "wc huge", order);
 
 	ttm_page_pool_init_locked(&_manager->uc_pool_huge,
-				  GFP_TRANSHUGE	& ~(__GFP_MOVABLE | __GFP_COMP)
+				  (GFP_TRANSHUGE_LIGHT | __GFP_NORETRY |
+				   __GFP_KSWAPD_RECLAIM) &
+				  ~(__GFP_MOVABLE | __GFP_COMP)
 				  , "uc huge", order);
 
 	_manager->options.max_size = max_pages;
diff --git a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
index 8a25d1974385..3f14c1cc0789 100644
--- a/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
+++ b/drivers/gpu/drm/ttm/ttm_page_alloc_dma.c
@@ -463,8 +463,9 @@ static unsigned ttm_dma_page_pool_free(struct dma_pool *pool, unsigned nr_free,
 	if (use_static)
 		pages_to_free = static_buf;
 	else
-		pages_to_free = kmalloc(npages_to_free * sizeof(struct page *),
-					GFP_KERNEL);
+		pages_to_free = kmalloc_array(npages_to_free,
+					      sizeof(struct page *),
+					      GFP_KERNEL);
 
 	if (!pages_to_free) {
 		pr_debug("%s: Failed to allocate memory for pool free operation\n",
@@ -753,7 +754,8 @@ static int ttm_dma_pool_alloc_new_pages(struct dma_pool *pool,
 			(unsigned)(PAGE_SIZE/sizeof(struct page *)));
 
 	/* allocate array for page caching change */
-	caching_array = kmalloc(max_cpages*sizeof(struct page *), GFP_KERNEL);
+	caching_array = kmalloc_array(max_cpages, sizeof(struct page *),
+				      GFP_KERNEL);
 
 	if (!caching_array) {
 		pr_debug("%s: Unable to allocate table for new pages\n",
@@ -910,7 +912,8 @@ static gfp_t ttm_dma_pool_gfp_flags(struct ttm_dma_tt *ttm_dma, bool huge)
 		gfp_flags |= __GFP_ZERO;
 
 	if (huge) {
-		gfp_flags |= GFP_TRANSHUGE;
+		gfp_flags |= GFP_TRANSHUGE_LIGHT | __GFP_NORETRY |
+			__GFP_KSWAPD_RECLAIM;
 		gfp_flags &= ~__GFP_MOVABLE;
 		gfp_flags &= ~__GFP_COMP;
 	}
diff --git a/drivers/gpu/drm/ttm/ttm_tt.c b/drivers/gpu/drm/ttm/ttm_tt.c
index 7e672be987b5..a1e543972ca7 100644
--- a/drivers/gpu/drm/ttm/ttm_tt.c
+++ b/drivers/gpu/drm/ttm/ttm_tt.c
@@ -1,3 +1,4 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
  * Copyright (c) 2006-2009 VMware, Inc., Palo Alto, CA., USA
diff --git a/drivers/gpu/drm/udl/udl_connector.c b/drivers/gpu/drm/udl/udl_connector.c
index 09dc585aa46f..68e88bed77ca 100644
--- a/drivers/gpu/drm/udl/udl_connector.c
+++ b/drivers/gpu/drm/udl/udl_connector.c
@@ -99,7 +99,7 @@ static int udl_get_modes(struct drm_connector *connector)
 					struct udl_drm_connector,
 					connector);
 
-	drm_mode_connector_update_edid_property(connector, udl_connector->edid);
+	drm_connector_update_edid_property(connector, udl_connector->edid);
 	if (udl_connector->edid)
 		return drm_add_edid_modes(connector, udl_connector->edid);
 	return 0;
@@ -200,7 +200,7 @@ int udl_connector_init(struct drm_device *dev, struct drm_encoder *encoder)
 	drm_connector_helper_add(connector, &udl_connector_helper_funcs);
 
 	drm_connector_register(connector);
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 	connector->polled = DRM_CONNECTOR_POLL_HPD |
 		DRM_CONNECTOR_POLL_CONNECT | DRM_CONNECTOR_POLL_DISCONNECT;
 
diff --git a/drivers/gpu/drm/udl/udl_dmabuf.c b/drivers/gpu/drm/udl/udl_dmabuf.c
index 0a20695eb120..556f62662aa9 100644
--- a/drivers/gpu/drm/udl/udl_dmabuf.c
+++ b/drivers/gpu/drm/udl/udl_dmabuf.c
@@ -29,7 +29,6 @@ struct udl_drm_dmabuf_attachment {
 };
 
 static int udl_attach_dma_buf(struct dma_buf *dmabuf,
-			      struct device *dev,
 			      struct dma_buf_attachment *attach)
 {
 	struct udl_drm_dmabuf_attachment *udl_attach;
@@ -158,27 +157,12 @@ static void *udl_dmabuf_kmap(struct dma_buf *dma_buf, unsigned long page_num)
 	return NULL;
 }
 
-static void *udl_dmabuf_kmap_atomic(struct dma_buf *dma_buf,
-				    unsigned long page_num)
-{
-	/* TODO */
-
-	return NULL;
-}
-
 static void udl_dmabuf_kunmap(struct dma_buf *dma_buf,
 			      unsigned long page_num, void *addr)
 {
 	/* TODO */
 }
 
-static void udl_dmabuf_kunmap_atomic(struct dma_buf *dma_buf,
-				     unsigned long page_num,
-				     void *addr)
-{
-	/* TODO */
-}
-
 static int udl_dmabuf_mmap(struct dma_buf *dma_buf,
 			   struct vm_area_struct *vma)
 {
@@ -193,9 +177,7 @@ static const struct dma_buf_ops udl_dmabuf_ops = {
 	.map_dma_buf		= udl_map_dma_buf,
 	.unmap_dma_buf		= udl_unmap_dma_buf,
 	.map			= udl_dmabuf_kmap,
-	.map_atomic		= udl_dmabuf_kmap_atomic,
 	.unmap			= udl_dmabuf_kunmap,
-	.unmap_atomic		= udl_dmabuf_kunmap_atomic,
 	.mmap			= udl_dmabuf_mmap,
 	.release		= drm_gem_dmabuf_release,
 };
diff --git a/drivers/gpu/drm/udl/udl_drv.h b/drivers/gpu/drm/udl/udl_drv.h
index 55c0cc309198..072582570a4f 100644
--- a/drivers/gpu/drm/udl/udl_drv.h
+++ b/drivers/gpu/drm/udl/udl_drv.h
@@ -16,6 +16,7 @@
 
 #include <linux/usb.h>
 #include <drm/drm_gem.h>
+#include <linux/mm_types.h>
 
 #define DRIVER_NAME		"udl"
 #define DRIVER_DESC		"DisplayLink"
@@ -136,7 +137,7 @@ void udl_gem_put_pages(struct udl_gem_object *obj);
 int udl_gem_vmap(struct udl_gem_object *obj);
 void udl_gem_vunmap(struct udl_gem_object *obj);
 int udl_drm_gem_mmap(struct file *filp, struct vm_area_struct *vma);
-int udl_gem_fault(struct vm_fault *vmf);
+vm_fault_t udl_gem_fault(struct vm_fault *vmf);
 
 int udl_handle_damage(struct udl_framebuffer *fb, int x, int y,
 		      int width, int height);
diff --git a/drivers/gpu/drm/udl/udl_gem.c b/drivers/gpu/drm/udl/udl_gem.c
index 9a15cce22cce..d5a23295dd80 100644
--- a/drivers/gpu/drm/udl/udl_gem.c
+++ b/drivers/gpu/drm/udl/udl_gem.c
@@ -100,13 +100,12 @@ int udl_drm_gem_mmap(struct file *filp, struct vm_area_struct *vma)
 	return ret;
 }
 
-int udl_gem_fault(struct vm_fault *vmf)
+vm_fault_t udl_gem_fault(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 	struct udl_gem_object *obj = to_udl_bo(vma->vm_private_data);
 	struct page *page;
 	unsigned int page_offset;
-	int ret = 0;
 
 	page_offset = (vmf->address - vma->vm_start) >> PAGE_SHIFT;
 
@@ -114,17 +113,7 @@ int udl_gem_fault(struct vm_fault *vmf)
 		return VM_FAULT_SIGBUS;
 
 	page = obj->pages[page_offset];
-	ret = vm_insert_page(vma, vmf->address, page);
-	switch (ret) {
-	case -EAGAIN:
-	case 0:
-	case -ERESTARTSYS:
-		return VM_FAULT_NOPAGE;
-	case -ENOMEM:
-		return VM_FAULT_OOM;
-	default:
-		return VM_FAULT_SIGBUS;
-	}
+	return vmf_insert_page(vma, vmf->address, page);
 }
 
 int udl_gem_get_pages(struct udl_gem_object *obj)
diff --git a/drivers/gpu/drm/v3d/Kconfig b/drivers/gpu/drm/v3d/Kconfig
new file mode 100644
index 000000000000..1552bf552c94
--- /dev/null
+++ b/drivers/gpu/drm/v3d/Kconfig
@@ -0,0 +1,10 @@
+config DRM_V3D
+	tristate "Broadcom V3D 3.x and newer"
+	depends on ARCH_BCM || ARCH_BCMSTB || COMPILE_TEST
+	depends on DRM
+	depends on COMMON_CLK
+	depends on MMU
+	select DRM_SCHED
+	help
+	  Choose this option if you have a system that has a Broadcom
+	  V3D 3.x or newer GPU, such as BCM7268.
diff --git a/drivers/gpu/drm/v3d/Makefile b/drivers/gpu/drm/v3d/Makefile
new file mode 100644
index 000000000000..34446e1de64f
--- /dev/null
+++ b/drivers/gpu/drm/v3d/Makefile
@@ -0,0 +1,18 @@
+# Please keep these build lists sorted!
+
+# core driver code
+v3d-y := \
+	v3d_bo.o \
+	v3d_drv.o \
+	v3d_fence.o \
+	v3d_gem.o \
+	v3d_irq.o \
+	v3d_mmu.o \
+	v3d_trace_points.o \
+	v3d_sched.o
+
+v3d-$(CONFIG_DEBUG_FS) += v3d_debugfs.o
+
+obj-$(CONFIG_DRM_V3D)  += v3d.o
+
+CFLAGS_v3d_trace_points.o := -I$(src)
diff --git a/drivers/gpu/drm/v3d/v3d_bo.c b/drivers/gpu/drm/v3d/v3d_bo.c
new file mode 100644
index 000000000000..54d96518a131
--- /dev/null
+++ b/drivers/gpu/drm/v3d/v3d_bo.c
@@ -0,0 +1,371 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* Copyright (C) 2015-2018 Broadcom */
+
+/**
+ * DOC: V3D GEM BO management support
+ *
+ * Compared to VC4 (V3D 2.x), V3D 3.3 introduces an MMU between the
+ * GPU and the bus, allowing us to use shmem objects for our storage
+ * instead of CMA.
+ *
+ * Physically contiguous objects may still be imported to V3D, but the
+ * driver doesn't allocate physically contiguous objects on its own.
+ * Display engines requiring physically contiguous allocations should
+ * look into Mesa's "renderonly" support (as used by the Mesa pl111
+ * driver) for an example of how to integrate with V3D.
+ *
+ * Long term, we should support evicting pages from the MMU when under
+ * memory pressure (thus the v3d_bo_get_pages() refcounting), but
+ * that's not a high priority since our systems tend to not have swap.
+ */
+
+#include <linux/dma-buf.h>
+#include <linux/pfn_t.h>
+
+#include "v3d_drv.h"
+#include "uapi/drm/v3d_drm.h"
+
+/* Pins the shmem pages, fills in the .pages and .sgt fields of the BO, and maps
+ * it for DMA.
+ */
+static int
+v3d_bo_get_pages(struct v3d_bo *bo)
+{
+	struct drm_gem_object *obj = &bo->base;
+	struct drm_device *dev = obj->dev;
+	int npages = obj->size >> PAGE_SHIFT;
+	int ret = 0;
+
+	mutex_lock(&bo->lock);
+	if (bo->pages_refcount++ != 0)
+		goto unlock;
+
+	if (!obj->import_attach) {
+		bo->pages = drm_gem_get_pages(obj);
+		if (IS_ERR(bo->pages)) {
+			ret = PTR_ERR(bo->pages);
+			goto unlock;
+		}
+
+		bo->sgt = drm_prime_pages_to_sg(bo->pages, npages);
+		if (IS_ERR(bo->sgt)) {
+			ret = PTR_ERR(bo->sgt);
+			goto put_pages;
+		}
+
+		/* Map the pages for use by the GPU. */
+		dma_map_sg(dev->dev, bo->sgt->sgl,
+			   bo->sgt->nents, DMA_BIDIRECTIONAL);
+	} else {
+		bo->pages = kcalloc(npages, sizeof(*bo->pages), GFP_KERNEL);
+		if (!bo->pages)
+			goto put_pages;
+
+		drm_prime_sg_to_page_addr_arrays(bo->sgt, bo->pages,
+						 NULL, npages);
+
+		/* Note that dma-bufs come in mapped. */
+	}
+
+	mutex_unlock(&bo->lock);
+
+	return 0;
+
+put_pages:
+	drm_gem_put_pages(obj, bo->pages, true, true);
+	bo->pages = NULL;
+unlock:
+	bo->pages_refcount--;
+	mutex_unlock(&bo->lock);
+	return ret;
+}
+
+static void
+v3d_bo_put_pages(struct v3d_bo *bo)
+{
+	struct drm_gem_object *obj = &bo->base;
+
+	mutex_lock(&bo->lock);
+	if (--bo->pages_refcount == 0) {
+		if (!obj->import_attach) {
+			dma_unmap_sg(obj->dev->dev, bo->sgt->sgl,
+				     bo->sgt->nents, DMA_BIDIRECTIONAL);
+			sg_free_table(bo->sgt);
+			kfree(bo->sgt);
+			drm_gem_put_pages(obj, bo->pages, true, true);
+		} else {
+			kfree(bo->pages);
+		}
+	}
+	mutex_unlock(&bo->lock);
+}
+
+static struct v3d_bo *v3d_bo_create_struct(struct drm_device *dev,
+					   size_t unaligned_size)
+{
+	struct v3d_dev *v3d = to_v3d_dev(dev);
+	struct drm_gem_object *obj;
+	struct v3d_bo *bo;
+	size_t size = roundup(unaligned_size, PAGE_SIZE);
+	int ret;
+
+	if (size == 0)
+		return ERR_PTR(-EINVAL);
+
+	bo = kzalloc(sizeof(*bo), GFP_KERNEL);
+	if (!bo)
+		return ERR_PTR(-ENOMEM);
+	obj = &bo->base;
+
+	INIT_LIST_HEAD(&bo->vmas);
+	INIT_LIST_HEAD(&bo->unref_head);
+	mutex_init(&bo->lock);
+
+	ret = drm_gem_object_init(dev, obj, size);
+	if (ret)
+		goto free_bo;
+
+	spin_lock(&v3d->mm_lock);
+	ret = drm_mm_insert_node_generic(&v3d->mm, &bo->node,
+					 obj->size >> PAGE_SHIFT,
+					 GMP_GRANULARITY >> PAGE_SHIFT, 0, 0);
+	spin_unlock(&v3d->mm_lock);
+	if (ret)
+		goto free_obj;
+
+	return bo;
+
+free_obj:
+	drm_gem_object_release(obj);
+free_bo:
+	kfree(bo);
+	return ERR_PTR(ret);
+}
+
+struct v3d_bo *v3d_bo_create(struct drm_device *dev, struct drm_file *file_priv,
+			     size_t unaligned_size)
+{
+	struct v3d_dev *v3d = to_v3d_dev(dev);
+	struct drm_gem_object *obj;
+	struct v3d_bo *bo;
+	int ret;
+
+	bo = v3d_bo_create_struct(dev, unaligned_size);
+	if (IS_ERR(bo))
+		return bo;
+	obj = &bo->base;
+
+	bo->resv = &bo->_resv;
+	reservation_object_init(bo->resv);
+
+	ret = v3d_bo_get_pages(bo);
+	if (ret)
+		goto free_mm;
+
+	v3d_mmu_insert_ptes(bo);
+
+	mutex_lock(&v3d->bo_lock);
+	v3d->bo_stats.num_allocated++;
+	v3d->bo_stats.pages_allocated += obj->size >> PAGE_SHIFT;
+	mutex_unlock(&v3d->bo_lock);
+
+	return bo;
+
+free_mm:
+	spin_lock(&v3d->mm_lock);
+	drm_mm_remove_node(&bo->node);
+	spin_unlock(&v3d->mm_lock);
+
+	drm_gem_object_release(obj);
+	kfree(bo);
+	return ERR_PTR(ret);
+}
+
+/* Called DRM core on the last userspace/kernel unreference of the
+ * BO.
+ */
+void v3d_free_object(struct drm_gem_object *obj)
+{
+	struct v3d_dev *v3d = to_v3d_dev(obj->dev);
+	struct v3d_bo *bo = to_v3d_bo(obj);
+
+	mutex_lock(&v3d->bo_lock);
+	v3d->bo_stats.num_allocated--;
+	v3d->bo_stats.pages_allocated -= obj->size >> PAGE_SHIFT;
+	mutex_unlock(&v3d->bo_lock);
+
+	reservation_object_fini(&bo->_resv);
+
+	v3d_bo_put_pages(bo);
+
+	if (obj->import_attach)
+		drm_prime_gem_destroy(obj, bo->sgt);
+
+	v3d_mmu_remove_ptes(bo);
+	spin_lock(&v3d->mm_lock);
+	drm_mm_remove_node(&bo->node);
+	spin_unlock(&v3d->mm_lock);
+
+	mutex_destroy(&bo->lock);
+
+	drm_gem_object_release(obj);
+	kfree(bo);
+}
+
+struct reservation_object *v3d_prime_res_obj(struct drm_gem_object *obj)
+{
+	struct v3d_bo *bo = to_v3d_bo(obj);
+
+	return bo->resv;
+}
+
+static void
+v3d_set_mmap_vma_flags(struct vm_area_struct *vma)
+{
+	vma->vm_flags &= ~VM_PFNMAP;
+	vma->vm_flags |= VM_MIXEDMAP;
+	vma->vm_page_prot = pgprot_writecombine(vm_get_page_prot(vma->vm_flags));
+}
+
+vm_fault_t v3d_gem_fault(struct vm_fault *vmf)
+{
+	struct vm_area_struct *vma = vmf->vma;
+	struct drm_gem_object *obj = vma->vm_private_data;
+	struct v3d_bo *bo = to_v3d_bo(obj);
+	pfn_t pfn;
+	pgoff_t pgoff;
+
+	/* We don't use vmf->pgoff since that has the fake offset: */
+	pgoff = (vmf->address - vma->vm_start) >> PAGE_SHIFT;
+	pfn = __pfn_to_pfn_t(page_to_pfn(bo->pages[pgoff]), PFN_DEV);
+
+	return vmf_insert_mixed(vma, vmf->address, pfn);
+}
+
+int v3d_mmap(struct file *filp, struct vm_area_struct *vma)
+{
+	int ret;
+
+	ret = drm_gem_mmap(filp, vma);
+	if (ret)
+		return ret;
+
+	v3d_set_mmap_vma_flags(vma);
+
+	return ret;
+}
+
+int v3d_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma)
+{
+	int ret;
+
+	ret = drm_gem_mmap_obj(obj, obj->size, vma);
+	if (ret < 0)
+		return ret;
+
+	v3d_set_mmap_vma_flags(vma);
+
+	return 0;
+}
+
+struct sg_table *
+v3d_prime_get_sg_table(struct drm_gem_object *obj)
+{
+	struct v3d_bo *bo = to_v3d_bo(obj);
+	int npages = obj->size >> PAGE_SHIFT;
+
+	return drm_prime_pages_to_sg(bo->pages, npages);
+}
+
+struct drm_gem_object *
+v3d_prime_import_sg_table(struct drm_device *dev,
+			  struct dma_buf_attachment *attach,
+			  struct sg_table *sgt)
+{
+	struct drm_gem_object *obj;
+	struct v3d_bo *bo;
+
+	bo = v3d_bo_create_struct(dev, attach->dmabuf->size);
+	if (IS_ERR(bo))
+		return ERR_CAST(bo);
+	obj = &bo->base;
+
+	bo->resv = attach->dmabuf->resv;
+
+	bo->sgt = sgt;
+	v3d_bo_get_pages(bo);
+
+	v3d_mmu_insert_ptes(bo);
+
+	return obj;
+}
+
+int v3d_create_bo_ioctl(struct drm_device *dev, void *data,
+			struct drm_file *file_priv)
+{
+	struct drm_v3d_create_bo *args = data;
+	struct v3d_bo *bo = NULL;
+	int ret;
+
+	if (args->flags != 0) {
+		DRM_INFO("unknown create_bo flags: %d\n", args->flags);
+		return -EINVAL;
+	}
+
+	bo = v3d_bo_create(dev, file_priv, PAGE_ALIGN(args->size));
+	if (IS_ERR(bo))
+		return PTR_ERR(bo);
+
+	args->offset = bo->node.start << PAGE_SHIFT;
+
+	ret = drm_gem_handle_create(file_priv, &bo->base, &args->handle);
+	drm_gem_object_put_unlocked(&bo->base);
+
+	return ret;
+}
+
+int v3d_mmap_bo_ioctl(struct drm_device *dev, void *data,
+		      struct drm_file *file_priv)
+{
+	struct drm_v3d_mmap_bo *args = data;
+	struct drm_gem_object *gem_obj;
+	int ret;
+
+	if (args->flags != 0) {
+		DRM_INFO("unknown mmap_bo flags: %d\n", args->flags);
+		return -EINVAL;
+	}
+
+	gem_obj = drm_gem_object_lookup(file_priv, args->handle);
+	if (!gem_obj) {
+		DRM_DEBUG("Failed to look up GEM BO %d\n", args->handle);
+		return -ENOENT;
+	}
+
+	ret = drm_gem_create_mmap_offset(gem_obj);
+	if (ret == 0)
+		args->offset = drm_vma_node_offset_addr(&gem_obj->vma_node);
+	drm_gem_object_put_unlocked(gem_obj);
+
+	return ret;
+}
+
+int v3d_get_bo_offset_ioctl(struct drm_device *dev, void *data,
+			    struct drm_file *file_priv)
+{
+	struct drm_v3d_get_bo_offset *args = data;
+	struct drm_gem_object *gem_obj;
+	struct v3d_bo *bo;
+
+	gem_obj = drm_gem_object_lookup(file_priv, args->handle);
+	if (!gem_obj) {
+		DRM_DEBUG("Failed to look up GEM BO %d\n", args->handle);
+		return -ENOENT;
+	}
+	bo = to_v3d_bo(gem_obj);
+
+	args->offset = bo->node.start << PAGE_SHIFT;
+
+	drm_gem_object_put_unlocked(gem_obj);
+	return 0;
+}
diff --git a/drivers/gpu/drm/v3d/v3d_debugfs.c b/drivers/gpu/drm/v3d/v3d_debugfs.c
new file mode 100644
index 000000000000..4db62c545748
--- /dev/null
+++ b/drivers/gpu/drm/v3d/v3d_debugfs.c
@@ -0,0 +1,191 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* Copyright (C) 2014-2018 Broadcom */
+
+#include <linux/circ_buf.h>
+#include <linux/ctype.h>
+#include <linux/debugfs.h>
+#include <linux/pm_runtime.h>
+#include <linux/seq_file.h>
+#include <drm/drmP.h>
+
+#include "v3d_drv.h"
+#include "v3d_regs.h"
+
+#define REGDEF(reg) { reg, #reg }
+struct v3d_reg_def {
+	u32 reg;
+	const char *name;
+};
+
+static const struct v3d_reg_def v3d_hub_reg_defs[] = {
+	REGDEF(V3D_HUB_AXICFG),
+	REGDEF(V3D_HUB_UIFCFG),
+	REGDEF(V3D_HUB_IDENT0),
+	REGDEF(V3D_HUB_IDENT1),
+	REGDEF(V3D_HUB_IDENT2),
+	REGDEF(V3D_HUB_IDENT3),
+	REGDEF(V3D_HUB_INT_STS),
+	REGDEF(V3D_HUB_INT_MSK_STS),
+};
+
+static const struct v3d_reg_def v3d_gca_reg_defs[] = {
+	REGDEF(V3D_GCA_SAFE_SHUTDOWN),
+	REGDEF(V3D_GCA_SAFE_SHUTDOWN_ACK),
+};
+
+static const struct v3d_reg_def v3d_core_reg_defs[] = {
+	REGDEF(V3D_CTL_IDENT0),
+	REGDEF(V3D_CTL_IDENT1),
+	REGDEF(V3D_CTL_IDENT2),
+	REGDEF(V3D_CTL_MISCCFG),
+	REGDEF(V3D_CTL_INT_STS),
+	REGDEF(V3D_CTL_INT_MSK_STS),
+	REGDEF(V3D_CLE_CT0CS),
+	REGDEF(V3D_CLE_CT0CA),
+	REGDEF(V3D_CLE_CT0EA),
+	REGDEF(V3D_CLE_CT1CS),
+	REGDEF(V3D_CLE_CT1CA),
+	REGDEF(V3D_CLE_CT1EA),
+
+	REGDEF(V3D_PTB_BPCA),
+	REGDEF(V3D_PTB_BPCS),
+
+	REGDEF(V3D_MMU_CTL),
+	REGDEF(V3D_MMU_VIO_ADDR),
+
+	REGDEF(V3D_GMP_STATUS),
+	REGDEF(V3D_GMP_CFG),
+	REGDEF(V3D_GMP_VIO_ADDR),
+};
+
+static int v3d_v3d_debugfs_regs(struct seq_file *m, void *unused)
+{
+	struct drm_info_node *node = (struct drm_info_node *)m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct v3d_dev *v3d = to_v3d_dev(dev);
+	int i, core;
+
+	for (i = 0; i < ARRAY_SIZE(v3d_hub_reg_defs); i++) {
+		seq_printf(m, "%s (0x%04x): 0x%08x\n",
+			   v3d_hub_reg_defs[i].name, v3d_hub_reg_defs[i].reg,
+			   V3D_READ(v3d_hub_reg_defs[i].reg));
+	}
+
+	for (i = 0; i < ARRAY_SIZE(v3d_gca_reg_defs); i++) {
+		seq_printf(m, "%s (0x%04x): 0x%08x\n",
+			   v3d_gca_reg_defs[i].name, v3d_gca_reg_defs[i].reg,
+			   V3D_GCA_READ(v3d_gca_reg_defs[i].reg));
+	}
+
+	for (core = 0; core < v3d->cores; core++) {
+		for (i = 0; i < ARRAY_SIZE(v3d_core_reg_defs); i++) {
+			seq_printf(m, "core %d %s (0x%04x): 0x%08x\n",
+				   core,
+				   v3d_core_reg_defs[i].name,
+				   v3d_core_reg_defs[i].reg,
+				   V3D_CORE_READ(core,
+						 v3d_core_reg_defs[i].reg));
+		}
+	}
+
+	return 0;
+}
+
+static int v3d_v3d_debugfs_ident(struct seq_file *m, void *unused)
+{
+	struct drm_info_node *node = (struct drm_info_node *)m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct v3d_dev *v3d = to_v3d_dev(dev);
+	u32 ident0, ident1, ident2, ident3, cores;
+	int ret, core;
+
+	ret = pm_runtime_get_sync(v3d->dev);
+	if (ret < 0)
+		return ret;
+
+	ident0 = V3D_READ(V3D_HUB_IDENT0);
+	ident1 = V3D_READ(V3D_HUB_IDENT1);
+	ident2 = V3D_READ(V3D_HUB_IDENT2);
+	ident3 = V3D_READ(V3D_HUB_IDENT3);
+	cores = V3D_GET_FIELD(ident1, V3D_HUB_IDENT1_NCORES);
+
+	seq_printf(m, "Revision:   %d.%d.%d.%d\n",
+		   V3D_GET_FIELD(ident1, V3D_HUB_IDENT1_TVER),
+		   V3D_GET_FIELD(ident1, V3D_HUB_IDENT1_REV),
+		   V3D_GET_FIELD(ident3, V3D_HUB_IDENT3_IPREV),
+		   V3D_GET_FIELD(ident3, V3D_HUB_IDENT3_IPIDX));
+	seq_printf(m, "MMU:        %s\n",
+		   (ident2 & V3D_HUB_IDENT2_WITH_MMU) ? "yes" : "no");
+	seq_printf(m, "TFU:        %s\n",
+		   (ident1 & V3D_HUB_IDENT1_WITH_TFU) ? "yes" : "no");
+	seq_printf(m, "TSY:        %s\n",
+		   (ident1 & V3D_HUB_IDENT1_WITH_TSY) ? "yes" : "no");
+	seq_printf(m, "MSO:        %s\n",
+		   (ident1 & V3D_HUB_IDENT1_WITH_MSO) ? "yes" : "no");
+	seq_printf(m, "L3C:        %s (%dkb)\n",
+		   (ident1 & V3D_HUB_IDENT1_WITH_L3C) ? "yes" : "no",
+		   V3D_GET_FIELD(ident2, V3D_HUB_IDENT2_L3C_NKB));
+
+	for (core = 0; core < cores; core++) {
+		u32 misccfg;
+		u32 nslc, ntmu, qups;
+
+		ident0 = V3D_CORE_READ(core, V3D_CTL_IDENT0);
+		ident1 = V3D_CORE_READ(core, V3D_CTL_IDENT1);
+		ident2 = V3D_CORE_READ(core, V3D_CTL_IDENT2);
+		misccfg = V3D_CORE_READ(core, V3D_CTL_MISCCFG);
+
+		nslc = V3D_GET_FIELD(ident1, V3D_IDENT1_NSLC);
+		ntmu = V3D_GET_FIELD(ident1, V3D_IDENT1_NTMU);
+		qups = V3D_GET_FIELD(ident1, V3D_IDENT1_QUPS);
+
+		seq_printf(m, "Core %d:\n", core);
+		seq_printf(m, "  Revision:     %d.%d\n",
+			   V3D_GET_FIELD(ident0, V3D_IDENT0_VER),
+			   V3D_GET_FIELD(ident1, V3D_IDENT1_REV));
+		seq_printf(m, "  Slices:       %d\n", nslc);
+		seq_printf(m, "  TMUs:         %d\n", nslc * ntmu);
+		seq_printf(m, "  QPUs:         %d\n", nslc * qups);
+		seq_printf(m, "  Semaphores:   %d\n",
+			   V3D_GET_FIELD(ident1, V3D_IDENT1_NSEM));
+		seq_printf(m, "  BCG int:      %d\n",
+			   (ident2 & V3D_IDENT2_BCG_INT) != 0);
+		seq_printf(m, "  Override TMU: %d\n",
+			   (misccfg & V3D_MISCCFG_OVRTMUOUT) != 0);
+	}
+
+	pm_runtime_mark_last_busy(v3d->dev);
+	pm_runtime_put_autosuspend(v3d->dev);
+
+	return 0;
+}
+
+static int v3d_debugfs_bo_stats(struct seq_file *m, void *unused)
+{
+	struct drm_info_node *node = (struct drm_info_node *)m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct v3d_dev *v3d = to_v3d_dev(dev);
+
+	mutex_lock(&v3d->bo_lock);
+	seq_printf(m, "allocated bos:          %d\n",
+		   v3d->bo_stats.num_allocated);
+	seq_printf(m, "allocated bo size (kb): %ld\n",
+		   (long)v3d->bo_stats.pages_allocated << (PAGE_SHIFT - 10));
+	mutex_unlock(&v3d->bo_lock);
+
+	return 0;
+}
+
+static const struct drm_info_list v3d_debugfs_list[] = {
+	{"v3d_ident", v3d_v3d_debugfs_ident, 0},
+	{"v3d_regs", v3d_v3d_debugfs_regs, 0},
+	{"bo_stats", v3d_debugfs_bo_stats, 0},
+};
+
+int
+v3d_debugfs_init(struct drm_minor *minor)
+{
+	return drm_debugfs_create_files(v3d_debugfs_list,
+					ARRAY_SIZE(v3d_debugfs_list),
+					minor->debugfs_root, minor);
+}
diff --git a/drivers/gpu/drm/v3d/v3d_drv.c b/drivers/gpu/drm/v3d/v3d_drv.c
new file mode 100644
index 000000000000..1dceba2b42fd
--- /dev/null
+++ b/drivers/gpu/drm/v3d/v3d_drv.c
@@ -0,0 +1,370 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* Copyright (C) 2014-2018 Broadcom */
+
+/**
+ * DOC: Broadcom V3D Graphics Driver
+ *
+ * This driver supports the Broadcom V3D 3.3 and 4.1 OpenGL ES GPUs.
+ * For V3D 2.x support, see the VC4 driver.
+ *
+ * Currently only single-core rendering using the binner and renderer
+ * is supported.  The TFU (texture formatting unit) and V3D 4.x's CSD
+ * (compute shader dispatch) are not yet supported.
+ */
+
+#include <linux/clk.h>
+#include <linux/device.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of_platform.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <drm/drm_fb_cma_helper.h>
+#include <drm/drm_fb_helper.h>
+
+#include "uapi/drm/v3d_drm.h"
+#include "v3d_drv.h"
+#include "v3d_regs.h"
+
+#define DRIVER_NAME "v3d"
+#define DRIVER_DESC "Broadcom V3D graphics"
+#define DRIVER_DATE "20180419"
+#define DRIVER_MAJOR 1
+#define DRIVER_MINOR 0
+#define DRIVER_PATCHLEVEL 0
+
+#ifdef CONFIG_PM
+static int v3d_runtime_suspend(struct device *dev)
+{
+	struct drm_device *drm = dev_get_drvdata(dev);
+	struct v3d_dev *v3d = to_v3d_dev(drm);
+
+	v3d_irq_disable(v3d);
+
+	clk_disable_unprepare(v3d->clk);
+
+	return 0;
+}
+
+static int v3d_runtime_resume(struct device *dev)
+{
+	struct drm_device *drm = dev_get_drvdata(dev);
+	struct v3d_dev *v3d = to_v3d_dev(drm);
+	int ret;
+
+	ret = clk_prepare_enable(v3d->clk);
+	if (ret != 0)
+		return ret;
+
+	/* XXX: VPM base */
+
+	v3d_mmu_set_page_table(v3d);
+	v3d_irq_enable(v3d);
+
+	return 0;
+}
+#endif
+
+static const struct dev_pm_ops v3d_v3d_pm_ops = {
+	SET_RUNTIME_PM_OPS(v3d_runtime_suspend, v3d_runtime_resume, NULL)
+};
+
+static int v3d_get_param_ioctl(struct drm_device *dev, void *data,
+			       struct drm_file *file_priv)
+{
+	struct v3d_dev *v3d = to_v3d_dev(dev);
+	struct drm_v3d_get_param *args = data;
+	int ret;
+	static const u32 reg_map[] = {
+		[DRM_V3D_PARAM_V3D_UIFCFG] = V3D_HUB_UIFCFG,
+		[DRM_V3D_PARAM_V3D_HUB_IDENT1] = V3D_HUB_IDENT1,
+		[DRM_V3D_PARAM_V3D_HUB_IDENT2] = V3D_HUB_IDENT2,
+		[DRM_V3D_PARAM_V3D_HUB_IDENT3] = V3D_HUB_IDENT3,
+		[DRM_V3D_PARAM_V3D_CORE0_IDENT0] = V3D_CTL_IDENT0,
+		[DRM_V3D_PARAM_V3D_CORE0_IDENT1] = V3D_CTL_IDENT1,
+		[DRM_V3D_PARAM_V3D_CORE0_IDENT2] = V3D_CTL_IDENT2,
+	};
+
+	if (args->pad != 0)
+		return -EINVAL;
+
+	/* Note that DRM_V3D_PARAM_V3D_CORE0_IDENT0 is 0, so we need
+	 * to explicitly allow it in the "the register in our
+	 * parameter map" check.
+	 */
+	if (args->param < ARRAY_SIZE(reg_map) &&
+	    (reg_map[args->param] ||
+	     args->param == DRM_V3D_PARAM_V3D_CORE0_IDENT0)) {
+		u32 offset = reg_map[args->param];
+
+		if (args->value != 0)
+			return -EINVAL;
+
+		ret = pm_runtime_get_sync(v3d->dev);
+		if (args->param >= DRM_V3D_PARAM_V3D_CORE0_IDENT0 &&
+		    args->param <= DRM_V3D_PARAM_V3D_CORE0_IDENT2) {
+			args->value = V3D_CORE_READ(0, offset);
+		} else {
+			args->value = V3D_READ(offset);
+		}
+		pm_runtime_mark_last_busy(v3d->dev);
+		pm_runtime_put_autosuspend(v3d->dev);
+		return 0;
+	}
+
+	/* Any params that aren't just register reads would go here. */
+
+	DRM_DEBUG("Unknown parameter %d\n", args->param);
+	return -EINVAL;
+}
+
+static int
+v3d_open(struct drm_device *dev, struct drm_file *file)
+{
+	struct v3d_dev *v3d = to_v3d_dev(dev);
+	struct v3d_file_priv *v3d_priv;
+	struct drm_sched_rq *rq;
+	int i;
+
+	v3d_priv = kzalloc(sizeof(*v3d_priv), GFP_KERNEL);
+	if (!v3d_priv)
+		return -ENOMEM;
+
+	v3d_priv->v3d = v3d;
+
+	for (i = 0; i < V3D_MAX_QUEUES; i++) {
+		rq = &v3d->queue[i].sched.sched_rq[DRM_SCHED_PRIORITY_NORMAL];
+		drm_sched_entity_init(&v3d_priv->sched_entity[i], &rq, 1, NULL);
+	}
+
+	file->driver_priv = v3d_priv;
+
+	return 0;
+}
+
+static void
+v3d_postclose(struct drm_device *dev, struct drm_file *file)
+{
+	struct v3d_dev *v3d = to_v3d_dev(dev);
+	struct v3d_file_priv *v3d_priv = file->driver_priv;
+	enum v3d_queue q;
+
+	for (q = 0; q < V3D_MAX_QUEUES; q++) {
+		drm_sched_entity_destroy(&v3d->queue[q].sched,
+				      &v3d_priv->sched_entity[q]);
+	}
+
+	kfree(v3d_priv);
+}
+
+static const struct file_operations v3d_drm_fops = {
+	.owner = THIS_MODULE,
+	.open = drm_open,
+	.release = drm_release,
+	.unlocked_ioctl = drm_ioctl,
+	.mmap = v3d_mmap,
+	.poll = drm_poll,
+	.read = drm_read,
+	.compat_ioctl = drm_compat_ioctl,
+	.llseek = noop_llseek,
+};
+
+/* DRM_AUTH is required on SUBMIT_CL for now, while we don't have GMP
+ * protection between clients.  Note that render nodes would be be
+ * able to submit CLs that could access BOs from clients authenticated
+ * with the master node.
+ */
+static const struct drm_ioctl_desc v3d_drm_ioctls[] = {
+	DRM_IOCTL_DEF_DRV(V3D_SUBMIT_CL, v3d_submit_cl_ioctl, DRM_RENDER_ALLOW | DRM_AUTH),
+	DRM_IOCTL_DEF_DRV(V3D_WAIT_BO, v3d_wait_bo_ioctl, DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(V3D_CREATE_BO, v3d_create_bo_ioctl, DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(V3D_MMAP_BO, v3d_mmap_bo_ioctl, DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(V3D_GET_PARAM, v3d_get_param_ioctl, DRM_RENDER_ALLOW),
+	DRM_IOCTL_DEF_DRV(V3D_GET_BO_OFFSET, v3d_get_bo_offset_ioctl, DRM_RENDER_ALLOW),
+};
+
+static const struct vm_operations_struct v3d_vm_ops = {
+	.fault = v3d_gem_fault,
+	.open = drm_gem_vm_open,
+	.close = drm_gem_vm_close,
+};
+
+static struct drm_driver v3d_drm_driver = {
+	.driver_features = (DRIVER_GEM |
+			    DRIVER_RENDER |
+			    DRIVER_PRIME |
+			    DRIVER_SYNCOBJ),
+
+	.open = v3d_open,
+	.postclose = v3d_postclose,
+
+#if defined(CONFIG_DEBUG_FS)
+	.debugfs_init = v3d_debugfs_init,
+#endif
+
+	.gem_free_object_unlocked = v3d_free_object,
+	.gem_vm_ops = &v3d_vm_ops,
+
+	.prime_handle_to_fd = drm_gem_prime_handle_to_fd,
+	.prime_fd_to_handle = drm_gem_prime_fd_to_handle,
+	.gem_prime_import = drm_gem_prime_import,
+	.gem_prime_export = drm_gem_prime_export,
+	.gem_prime_res_obj = v3d_prime_res_obj,
+	.gem_prime_get_sg_table	= v3d_prime_get_sg_table,
+	.gem_prime_import_sg_table = v3d_prime_import_sg_table,
+	.gem_prime_mmap = v3d_prime_mmap,
+
+	.ioctls = v3d_drm_ioctls,
+	.num_ioctls = ARRAY_SIZE(v3d_drm_ioctls),
+	.fops = &v3d_drm_fops,
+
+	.name = DRIVER_NAME,
+	.desc = DRIVER_DESC,
+	.date = DRIVER_DATE,
+	.major = DRIVER_MAJOR,
+	.minor = DRIVER_MINOR,
+	.patchlevel = DRIVER_PATCHLEVEL,
+};
+
+static const struct of_device_id v3d_of_match[] = {
+	{ .compatible = "brcm,7268-v3d" },
+	{ .compatible = "brcm,7278-v3d" },
+	{},
+};
+MODULE_DEVICE_TABLE(of, v3d_of_match);
+
+static int
+map_regs(struct v3d_dev *v3d, void __iomem **regs, const char *name)
+{
+	struct resource *res =
+		platform_get_resource_byname(v3d->pdev, IORESOURCE_MEM, name);
+
+	*regs = devm_ioremap_resource(v3d->dev, res);
+	return PTR_ERR_OR_ZERO(*regs);
+}
+
+static int v3d_platform_drm_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct drm_device *drm;
+	struct v3d_dev *v3d;
+	int ret;
+	u32 ident1;
+
+	dev->coherent_dma_mask = DMA_BIT_MASK(36);
+
+	v3d = kzalloc(sizeof(*v3d), GFP_KERNEL);
+	if (!v3d)
+		return -ENOMEM;
+	v3d->dev = dev;
+	v3d->pdev = pdev;
+	drm = &v3d->drm;
+
+	ret = map_regs(v3d, &v3d->bridge_regs, "bridge");
+	if (ret)
+		goto dev_free;
+
+	ret = map_regs(v3d, &v3d->hub_regs, "hub");
+	if (ret)
+		goto dev_free;
+
+	ret = map_regs(v3d, &v3d->core_regs[0], "core0");
+	if (ret)
+		goto dev_free;
+
+	ident1 = V3D_READ(V3D_HUB_IDENT1);
+	v3d->ver = (V3D_GET_FIELD(ident1, V3D_HUB_IDENT1_TVER) * 10 +
+		    V3D_GET_FIELD(ident1, V3D_HUB_IDENT1_REV));
+	v3d->cores = V3D_GET_FIELD(ident1, V3D_HUB_IDENT1_NCORES);
+	WARN_ON(v3d->cores > 1); /* multicore not yet implemented */
+
+	if (v3d->ver < 41) {
+		ret = map_regs(v3d, &v3d->gca_regs, "gca");
+		if (ret)
+			goto dev_free;
+	}
+
+	v3d->mmu_scratch = dma_alloc_wc(dev, 4096, &v3d->mmu_scratch_paddr,
+					GFP_KERNEL | __GFP_NOWARN | __GFP_ZERO);
+	if (!v3d->mmu_scratch) {
+		dev_err(dev, "Failed to allocate MMU scratch page\n");
+		ret = -ENOMEM;
+		goto dev_free;
+	}
+
+	pm_runtime_use_autosuspend(dev);
+	pm_runtime_set_autosuspend_delay(dev, 50);
+	pm_runtime_enable(dev);
+
+	ret = drm_dev_init(&v3d->drm, &v3d_drm_driver, dev);
+	if (ret)
+		goto dma_free;
+
+	platform_set_drvdata(pdev, drm);
+	drm->dev_private = v3d;
+
+	ret = v3d_gem_init(drm);
+	if (ret)
+		goto dev_destroy;
+
+	v3d_irq_init(v3d);
+
+	ret = drm_dev_register(drm, 0);
+	if (ret)
+		goto gem_destroy;
+
+	return 0;
+
+gem_destroy:
+	v3d_gem_destroy(drm);
+dev_destroy:
+	drm_dev_put(drm);
+dma_free:
+	dma_free_wc(dev, 4096, v3d->mmu_scratch, v3d->mmu_scratch_paddr);
+dev_free:
+	kfree(v3d);
+	return ret;
+}
+
+static int v3d_platform_drm_remove(struct platform_device *pdev)
+{
+	struct drm_device *drm = platform_get_drvdata(pdev);
+	struct v3d_dev *v3d = to_v3d_dev(drm);
+
+	drm_dev_unregister(drm);
+
+	v3d_gem_destroy(drm);
+
+	drm_dev_put(drm);
+
+	dma_free_wc(v3d->dev, 4096, v3d->mmu_scratch, v3d->mmu_scratch_paddr);
+
+	return 0;
+}
+
+static struct platform_driver v3d_platform_driver = {
+	.probe		= v3d_platform_drm_probe,
+	.remove		= v3d_platform_drm_remove,
+	.driver		= {
+		.name	= "v3d",
+		.of_match_table = v3d_of_match,
+	},
+};
+
+static int __init v3d_drm_register(void)
+{
+	return platform_driver_register(&v3d_platform_driver);
+}
+
+static void __exit v3d_drm_unregister(void)
+{
+	platform_driver_unregister(&v3d_platform_driver);
+}
+
+module_init(v3d_drm_register);
+module_exit(v3d_drm_unregister);
+
+MODULE_ALIAS("platform:v3d-drm");
+MODULE_DESCRIPTION("Broadcom V3D DRM Driver");
+MODULE_AUTHOR("Eric Anholt <eric@anholt.net>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/gpu/drm/v3d/v3d_drv.h b/drivers/gpu/drm/v3d/v3d_drv.h
new file mode 100644
index 000000000000..e6fed696ad86
--- /dev/null
+++ b/drivers/gpu/drm/v3d/v3d_drv.h
@@ -0,0 +1,301 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* Copyright (C) 2015-2018 Broadcom */
+
+#include <linux/reservation.h>
+#include <linux/mm_types.h>
+#include <drm/drmP.h>
+#include <drm/drm_encoder.h>
+#include <drm/drm_gem.h>
+#include <drm/gpu_scheduler.h>
+
+#define GMP_GRANULARITY (128 * 1024)
+
+/* Enum for each of the V3D queues.  We maintain various queue
+ * tracking as an array because at some point we'll want to support
+ * the TFU (texture formatting unit) as another queue.
+ */
+enum v3d_queue {
+	V3D_BIN,
+	V3D_RENDER,
+};
+
+#define V3D_MAX_QUEUES (V3D_RENDER + 1)
+
+struct v3d_queue_state {
+	struct drm_gpu_scheduler sched;
+
+	u64 fence_context;
+	u64 emit_seqno;
+};
+
+struct v3d_dev {
+	struct drm_device drm;
+
+	/* Short representation (e.g. 33, 41) of the V3D tech version
+	 * and revision.
+	 */
+	int ver;
+
+	struct device *dev;
+	struct platform_device *pdev;
+	void __iomem *hub_regs;
+	void __iomem *core_regs[3];
+	void __iomem *bridge_regs;
+	void __iomem *gca_regs;
+	struct clk *clk;
+
+	/* Virtual and DMA addresses of the single shared page table. */
+	volatile u32 *pt;
+	dma_addr_t pt_paddr;
+
+	/* Virtual and DMA addresses of the MMU's scratch page.  When
+	 * a read or write is invalid in the MMU, it will be
+	 * redirected here.
+	 */
+	void *mmu_scratch;
+	dma_addr_t mmu_scratch_paddr;
+
+	/* Number of V3D cores. */
+	u32 cores;
+
+	/* Allocator managing the address space.  All units are in
+	 * number of pages.
+	 */
+	struct drm_mm mm;
+	spinlock_t mm_lock;
+
+	struct work_struct overflow_mem_work;
+
+	struct v3d_exec_info *bin_job;
+	struct v3d_exec_info *render_job;
+
+	struct v3d_queue_state queue[V3D_MAX_QUEUES];
+
+	/* Spinlock used to synchronize the overflow memory
+	 * management against bin job submission.
+	 */
+	spinlock_t job_lock;
+
+	/* Protects bo_stats */
+	struct mutex bo_lock;
+
+	/* Lock taken when resetting the GPU, to keep multiple
+	 * processes from trying to park the scheduler threads and
+	 * reset at once.
+	 */
+	struct mutex reset_lock;
+
+	/* Lock taken when creating and pushing the GPU scheduler
+	 * jobs, to keep the sched-fence seqnos in order.
+	 */
+	struct mutex sched_lock;
+
+	struct {
+		u32 num_allocated;
+		u32 pages_allocated;
+	} bo_stats;
+};
+
+static inline struct v3d_dev *
+to_v3d_dev(struct drm_device *dev)
+{
+	return (struct v3d_dev *)dev->dev_private;
+}
+
+/* The per-fd struct, which tracks the MMU mappings. */
+struct v3d_file_priv {
+	struct v3d_dev *v3d;
+
+	struct drm_sched_entity sched_entity[V3D_MAX_QUEUES];
+};
+
+/* Tracks a mapping of a BO into a per-fd address space */
+struct v3d_vma {
+	struct v3d_page_table *pt;
+	struct list_head list; /* entry in v3d_bo.vmas */
+};
+
+struct v3d_bo {
+	struct drm_gem_object base;
+
+	struct mutex lock;
+
+	struct drm_mm_node node;
+
+	u32 pages_refcount;
+	struct page **pages;
+	struct sg_table *sgt;
+	void *vaddr;
+
+	struct list_head vmas;    /* list of v3d_vma */
+
+	/* List entry for the BO's position in
+	 * v3d_exec_info->unref_list
+	 */
+	struct list_head unref_head;
+
+	/* normally (resv == &_resv) except for imported bo's */
+	struct reservation_object *resv;
+	struct reservation_object _resv;
+};
+
+static inline struct v3d_bo *
+to_v3d_bo(struct drm_gem_object *bo)
+{
+	return (struct v3d_bo *)bo;
+}
+
+struct v3d_fence {
+	struct dma_fence base;
+	struct drm_device *dev;
+	/* v3d seqno for signaled() test */
+	u64 seqno;
+	enum v3d_queue queue;
+};
+
+static inline struct v3d_fence *
+to_v3d_fence(struct dma_fence *fence)
+{
+	return (struct v3d_fence *)fence;
+}
+
+#define V3D_READ(offset) readl(v3d->hub_regs + offset)
+#define V3D_WRITE(offset, val) writel(val, v3d->hub_regs + offset)
+
+#define V3D_BRIDGE_READ(offset) readl(v3d->bridge_regs + offset)
+#define V3D_BRIDGE_WRITE(offset, val) writel(val, v3d->bridge_regs + offset)
+
+#define V3D_GCA_READ(offset) readl(v3d->gca_regs + offset)
+#define V3D_GCA_WRITE(offset, val) writel(val, v3d->gca_regs + offset)
+
+#define V3D_CORE_READ(core, offset) readl(v3d->core_regs[core] + offset)
+#define V3D_CORE_WRITE(core, offset, val) writel(val, v3d->core_regs[core] + offset)
+
+struct v3d_job {
+	struct drm_sched_job base;
+
+	struct v3d_exec_info *exec;
+
+	/* An optional fence userspace can pass in for the job to depend on. */
+	struct dma_fence *in_fence;
+
+	/* v3d fence to be signaled by IRQ handler when the job is complete. */
+	struct dma_fence *done_fence;
+
+	/* GPU virtual addresses of the start/end of the CL job. */
+	u32 start, end;
+
+	u32 timedout_ctca, timedout_ctra;
+};
+
+struct v3d_exec_info {
+	struct v3d_dev *v3d;
+
+	struct v3d_job bin, render;
+
+	/* Fence for when the scheduler considers the binner to be
+	 * done, for render to depend on.
+	 */
+	struct dma_fence *bin_done_fence;
+
+	struct kref refcount;
+
+	/* This is the array of BOs that were looked up at the start of exec. */
+	struct v3d_bo **bo;
+	u32 bo_count;
+
+	/* List of overflow BOs used in the job that need to be
+	 * released once the job is complete.
+	 */
+	struct list_head unref_list;
+
+	/* Submitted tile memory allocation start/size, tile state. */
+	u32 qma, qms, qts;
+};
+
+/**
+ * _wait_for - magic (register) wait macro
+ *
+ * Does the right thing for modeset paths when run under kdgb or similar atomic
+ * contexts. Note that it's important that we check the condition again after
+ * having timed out, since the timeout could be due to preemption or similar and
+ * we've never had a chance to check the condition before the timeout.
+ */
+#define wait_for(COND, MS) ({ \
+	unsigned long timeout__ = jiffies + msecs_to_jiffies(MS) + 1;	\
+	int ret__ = 0;							\
+	while (!(COND)) {						\
+		if (time_after(jiffies, timeout__)) {			\
+			if (!(COND))					\
+				ret__ = -ETIMEDOUT;			\
+			break;						\
+		}							\
+		msleep(1);					\
+	}								\
+	ret__;								\
+})
+
+static inline unsigned long nsecs_to_jiffies_timeout(const u64 n)
+{
+	/* nsecs_to_jiffies64() does not guard against overflow */
+	if (NSEC_PER_SEC % HZ &&
+	    div_u64(n, NSEC_PER_SEC) >= MAX_JIFFY_OFFSET / HZ)
+		return MAX_JIFFY_OFFSET;
+
+	return min_t(u64, MAX_JIFFY_OFFSET, nsecs_to_jiffies64(n) + 1);
+}
+
+/* v3d_bo.c */
+void v3d_free_object(struct drm_gem_object *gem_obj);
+struct v3d_bo *v3d_bo_create(struct drm_device *dev, struct drm_file *file_priv,
+			     size_t size);
+int v3d_create_bo_ioctl(struct drm_device *dev, void *data,
+			struct drm_file *file_priv);
+int v3d_mmap_bo_ioctl(struct drm_device *dev, void *data,
+		      struct drm_file *file_priv);
+int v3d_get_bo_offset_ioctl(struct drm_device *dev, void *data,
+			    struct drm_file *file_priv);
+vm_fault_t v3d_gem_fault(struct vm_fault *vmf);
+int v3d_mmap(struct file *filp, struct vm_area_struct *vma);
+struct reservation_object *v3d_prime_res_obj(struct drm_gem_object *obj);
+int v3d_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma);
+struct sg_table *v3d_prime_get_sg_table(struct drm_gem_object *obj);
+struct drm_gem_object *v3d_prime_import_sg_table(struct drm_device *dev,
+						 struct dma_buf_attachment *attach,
+						 struct sg_table *sgt);
+
+/* v3d_debugfs.c */
+int v3d_debugfs_init(struct drm_minor *minor);
+
+/* v3d_fence.c */
+extern const struct dma_fence_ops v3d_fence_ops;
+struct dma_fence *v3d_fence_create(struct v3d_dev *v3d, enum v3d_queue queue);
+
+/* v3d_gem.c */
+int v3d_gem_init(struct drm_device *dev);
+void v3d_gem_destroy(struct drm_device *dev);
+int v3d_submit_cl_ioctl(struct drm_device *dev, void *data,
+			struct drm_file *file_priv);
+int v3d_wait_bo_ioctl(struct drm_device *dev, void *data,
+		      struct drm_file *file_priv);
+void v3d_exec_put(struct v3d_exec_info *exec);
+void v3d_reset(struct v3d_dev *v3d);
+void v3d_invalidate_caches(struct v3d_dev *v3d);
+void v3d_flush_caches(struct v3d_dev *v3d);
+
+/* v3d_irq.c */
+void v3d_irq_init(struct v3d_dev *v3d);
+void v3d_irq_enable(struct v3d_dev *v3d);
+void v3d_irq_disable(struct v3d_dev *v3d);
+void v3d_irq_reset(struct v3d_dev *v3d);
+
+/* v3d_mmu.c */
+int v3d_mmu_get_offset(struct drm_file *file_priv, struct v3d_bo *bo,
+		       u32 *offset);
+int v3d_mmu_set_page_table(struct v3d_dev *v3d);
+void v3d_mmu_insert_ptes(struct v3d_bo *bo);
+void v3d_mmu_remove_ptes(struct v3d_bo *bo);
+
+/* v3d_sched.c */
+int v3d_sched_init(struct v3d_dev *v3d);
+void v3d_sched_fini(struct v3d_dev *v3d);
diff --git a/drivers/gpu/drm/v3d/v3d_fence.c b/drivers/gpu/drm/v3d/v3d_fence.c
new file mode 100644
index 000000000000..50bfcf9a8a1a
--- /dev/null
+++ b/drivers/gpu/drm/v3d/v3d_fence.c
@@ -0,0 +1,41 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* Copyright (C) 2017-2018 Broadcom */
+
+#include "v3d_drv.h"
+
+struct dma_fence *v3d_fence_create(struct v3d_dev *v3d, enum v3d_queue queue)
+{
+	struct v3d_fence *fence;
+
+	fence = kzalloc(sizeof(*fence), GFP_KERNEL);
+	if (!fence)
+		return ERR_PTR(-ENOMEM);
+
+	fence->dev = &v3d->drm;
+	fence->queue = queue;
+	fence->seqno = ++v3d->queue[queue].emit_seqno;
+	dma_fence_init(&fence->base, &v3d_fence_ops, &v3d->job_lock,
+		       v3d->queue[queue].fence_context, fence->seqno);
+
+	return &fence->base;
+}
+
+static const char *v3d_fence_get_driver_name(struct dma_fence *fence)
+{
+	return "v3d";
+}
+
+static const char *v3d_fence_get_timeline_name(struct dma_fence *fence)
+{
+	struct v3d_fence *f = to_v3d_fence(fence);
+
+	if (f->queue == V3D_BIN)
+		return "v3d-bin";
+	else
+		return "v3d-render";
+}
+
+const struct dma_fence_ops v3d_fence_ops = {
+	.get_driver_name = v3d_fence_get_driver_name,
+	.get_timeline_name = v3d_fence_get_timeline_name,
+};
diff --git a/drivers/gpu/drm/v3d/v3d_gem.c b/drivers/gpu/drm/v3d/v3d_gem.c
new file mode 100644
index 000000000000..e1fcbb4cd0ae
--- /dev/null
+++ b/drivers/gpu/drm/v3d/v3d_gem.c
@@ -0,0 +1,669 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* Copyright (C) 2014-2018 Broadcom */
+
+#include <drm/drmP.h>
+#include <drm/drm_syncobj.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/pm_runtime.h>
+#include <linux/device.h>
+#include <linux/io.h>
+#include <linux/sched/signal.h>
+
+#include "uapi/drm/v3d_drm.h"
+#include "v3d_drv.h"
+#include "v3d_regs.h"
+#include "v3d_trace.h"
+
+static void
+v3d_init_core(struct v3d_dev *v3d, int core)
+{
+	/* Set OVRTMUOUT, which means that the texture sampler uniform
+	 * configuration's tmu output type field is used, instead of
+	 * using the hardware default behavior based on the texture
+	 * type.  If you want the default behavior, you can still put
+	 * "2" in the indirect texture state's output_type field.
+	 */
+	V3D_CORE_WRITE(core, V3D_CTL_MISCCFG, V3D_MISCCFG_OVRTMUOUT);
+
+	/* Whenever we flush the L2T cache, we always want to flush
+	 * the whole thing.
+	 */
+	V3D_CORE_WRITE(core, V3D_CTL_L2TFLSTA, 0);
+	V3D_CORE_WRITE(core, V3D_CTL_L2TFLEND, ~0);
+}
+
+/* Sets invariant state for the HW. */
+static void
+v3d_init_hw_state(struct v3d_dev *v3d)
+{
+	v3d_init_core(v3d, 0);
+}
+
+static void
+v3d_idle_axi(struct v3d_dev *v3d, int core)
+{
+	V3D_CORE_WRITE(core, V3D_GMP_CFG, V3D_GMP_CFG_STOP_REQ);
+
+	if (wait_for((V3D_CORE_READ(core, V3D_GMP_STATUS) &
+		      (V3D_GMP_STATUS_RD_COUNT_MASK |
+		       V3D_GMP_STATUS_WR_COUNT_MASK |
+		       V3D_GMP_STATUS_CFG_BUSY)) == 0, 100)) {
+		DRM_ERROR("Failed to wait for safe GMP shutdown\n");
+	}
+}
+
+static void
+v3d_idle_gca(struct v3d_dev *v3d)
+{
+	if (v3d->ver >= 41)
+		return;
+
+	V3D_GCA_WRITE(V3D_GCA_SAFE_SHUTDOWN, V3D_GCA_SAFE_SHUTDOWN_EN);
+
+	if (wait_for((V3D_GCA_READ(V3D_GCA_SAFE_SHUTDOWN_ACK) &
+		      V3D_GCA_SAFE_SHUTDOWN_ACK_ACKED) ==
+		     V3D_GCA_SAFE_SHUTDOWN_ACK_ACKED, 100)) {
+		DRM_ERROR("Failed to wait for safe GCA shutdown\n");
+	}
+}
+
+static void
+v3d_reset_v3d(struct v3d_dev *v3d)
+{
+	int version = V3D_BRIDGE_READ(V3D_TOP_GR_BRIDGE_REVISION);
+
+	if (V3D_GET_FIELD(version, V3D_TOP_GR_BRIDGE_MAJOR) == 2) {
+		V3D_BRIDGE_WRITE(V3D_TOP_GR_BRIDGE_SW_INIT_0,
+				 V3D_TOP_GR_BRIDGE_SW_INIT_0_V3D_CLK_108_SW_INIT);
+		V3D_BRIDGE_WRITE(V3D_TOP_GR_BRIDGE_SW_INIT_0, 0);
+
+		/* GFXH-1383: The SW_INIT may cause a stray write to address 0
+		 * of the unit, so reset it to its power-on value here.
+		 */
+		V3D_WRITE(V3D_HUB_AXICFG, V3D_HUB_AXICFG_MAX_LEN_MASK);
+	} else {
+		WARN_ON_ONCE(V3D_GET_FIELD(version,
+					   V3D_TOP_GR_BRIDGE_MAJOR) != 7);
+		V3D_BRIDGE_WRITE(V3D_TOP_GR_BRIDGE_SW_INIT_1,
+				 V3D_TOP_GR_BRIDGE_SW_INIT_1_V3D_CLK_108_SW_INIT);
+		V3D_BRIDGE_WRITE(V3D_TOP_GR_BRIDGE_SW_INIT_1, 0);
+	}
+
+	v3d_init_hw_state(v3d);
+}
+
+void
+v3d_reset(struct v3d_dev *v3d)
+{
+	struct drm_device *dev = &v3d->drm;
+
+	DRM_ERROR("Resetting GPU.\n");
+	trace_v3d_reset_begin(dev);
+
+	/* XXX: only needed for safe powerdown, not reset. */
+	if (false)
+		v3d_idle_axi(v3d, 0);
+
+	v3d_idle_gca(v3d);
+	v3d_reset_v3d(v3d);
+
+	v3d_mmu_set_page_table(v3d);
+	v3d_irq_reset(v3d);
+
+	trace_v3d_reset_end(dev);
+}
+
+static void
+v3d_flush_l3(struct v3d_dev *v3d)
+{
+	if (v3d->ver < 41) {
+		u32 gca_ctrl = V3D_GCA_READ(V3D_GCA_CACHE_CTRL);
+
+		V3D_GCA_WRITE(V3D_GCA_CACHE_CTRL,
+			      gca_ctrl | V3D_GCA_CACHE_CTRL_FLUSH);
+
+		if (v3d->ver < 33) {
+			V3D_GCA_WRITE(V3D_GCA_CACHE_CTRL,
+				      gca_ctrl & ~V3D_GCA_CACHE_CTRL_FLUSH);
+		}
+	}
+}
+
+/* Invalidates the (read-only) L2 cache. */
+static void
+v3d_invalidate_l2(struct v3d_dev *v3d, int core)
+{
+	V3D_CORE_WRITE(core, V3D_CTL_L2CACTL,
+		       V3D_L2CACTL_L2CCLR |
+		       V3D_L2CACTL_L2CENA);
+}
+
+static void
+v3d_invalidate_l1td(struct v3d_dev *v3d, int core)
+{
+	V3D_CORE_WRITE(core, V3D_CTL_L2TCACTL, V3D_L2TCACTL_TMUWCF);
+	if (wait_for(!(V3D_CORE_READ(core, V3D_CTL_L2TCACTL) &
+		       V3D_L2TCACTL_L2TFLS), 100)) {
+		DRM_ERROR("Timeout waiting for L1T write combiner flush\n");
+	}
+}
+
+/* Invalidates texture L2 cachelines */
+static void
+v3d_flush_l2t(struct v3d_dev *v3d, int core)
+{
+	v3d_invalidate_l1td(v3d, core);
+
+	V3D_CORE_WRITE(core, V3D_CTL_L2TCACTL,
+		       V3D_L2TCACTL_L2TFLS |
+		       V3D_SET_FIELD(V3D_L2TCACTL_FLM_FLUSH, V3D_L2TCACTL_FLM));
+	if (wait_for(!(V3D_CORE_READ(core, V3D_CTL_L2TCACTL) &
+		       V3D_L2TCACTL_L2TFLS), 100)) {
+		DRM_ERROR("Timeout waiting for L2T flush\n");
+	}
+}
+
+/* Invalidates the slice caches.  These are read-only caches. */
+static void
+v3d_invalidate_slices(struct v3d_dev *v3d, int core)
+{
+	V3D_CORE_WRITE(core, V3D_CTL_SLCACTL,
+		       V3D_SET_FIELD(0xf, V3D_SLCACTL_TVCCS) |
+		       V3D_SET_FIELD(0xf, V3D_SLCACTL_TDCCS) |
+		       V3D_SET_FIELD(0xf, V3D_SLCACTL_UCC) |
+		       V3D_SET_FIELD(0xf, V3D_SLCACTL_ICC));
+}
+
+/* Invalidates texture L2 cachelines */
+static void
+v3d_invalidate_l2t(struct v3d_dev *v3d, int core)
+{
+	V3D_CORE_WRITE(core,
+		       V3D_CTL_L2TCACTL,
+		       V3D_L2TCACTL_L2TFLS |
+		       V3D_SET_FIELD(V3D_L2TCACTL_FLM_CLEAR, V3D_L2TCACTL_FLM));
+	if (wait_for(!(V3D_CORE_READ(core, V3D_CTL_L2TCACTL) &
+		       V3D_L2TCACTL_L2TFLS), 100)) {
+		DRM_ERROR("Timeout waiting for L2T invalidate\n");
+	}
+}
+
+void
+v3d_invalidate_caches(struct v3d_dev *v3d)
+{
+	v3d_flush_l3(v3d);
+
+	v3d_invalidate_l2(v3d, 0);
+	v3d_invalidate_slices(v3d, 0);
+	v3d_flush_l2t(v3d, 0);
+}
+
+void
+v3d_flush_caches(struct v3d_dev *v3d)
+{
+	v3d_invalidate_l1td(v3d, 0);
+	v3d_invalidate_l2t(v3d, 0);
+}
+
+static void
+v3d_attach_object_fences(struct v3d_exec_info *exec)
+{
+	struct dma_fence *out_fence = &exec->render.base.s_fence->finished;
+	struct v3d_bo *bo;
+	int i;
+
+	for (i = 0; i < exec->bo_count; i++) {
+		bo = to_v3d_bo(&exec->bo[i]->base);
+
+		/* XXX: Use shared fences for read-only objects. */
+		reservation_object_add_excl_fence(bo->resv, out_fence);
+	}
+}
+
+static void
+v3d_unlock_bo_reservations(struct drm_device *dev,
+			   struct v3d_exec_info *exec,
+			   struct ww_acquire_ctx *acquire_ctx)
+{
+	int i;
+
+	for (i = 0; i < exec->bo_count; i++) {
+		struct v3d_bo *bo = to_v3d_bo(&exec->bo[i]->base);
+
+		ww_mutex_unlock(&bo->resv->lock);
+	}
+
+	ww_acquire_fini(acquire_ctx);
+}
+
+/* Takes the reservation lock on all the BOs being referenced, so that
+ * at queue submit time we can update the reservations.
+ *
+ * We don't lock the RCL the tile alloc/state BOs, or overflow memory
+ * (all of which are on exec->unref_list).  They're entirely private
+ * to v3d, so we don't attach dma-buf fences to them.
+ */
+static int
+v3d_lock_bo_reservations(struct drm_device *dev,
+			 struct v3d_exec_info *exec,
+			 struct ww_acquire_ctx *acquire_ctx)
+{
+	int contended_lock = -1;
+	int i, ret;
+	struct v3d_bo *bo;
+
+	ww_acquire_init(acquire_ctx, &reservation_ww_class);
+
+retry:
+	if (contended_lock != -1) {
+		bo = to_v3d_bo(&exec->bo[contended_lock]->base);
+		ret = ww_mutex_lock_slow_interruptible(&bo->resv->lock,
+						       acquire_ctx);
+		if (ret) {
+			ww_acquire_done(acquire_ctx);
+			return ret;
+		}
+	}
+
+	for (i = 0; i < exec->bo_count; i++) {
+		if (i == contended_lock)
+			continue;
+
+		bo = to_v3d_bo(&exec->bo[i]->base);
+
+		ret = ww_mutex_lock_interruptible(&bo->resv->lock, acquire_ctx);
+		if (ret) {
+			int j;
+
+			for (j = 0; j < i; j++) {
+				bo = to_v3d_bo(&exec->bo[j]->base);
+				ww_mutex_unlock(&bo->resv->lock);
+			}
+
+			if (contended_lock != -1 && contended_lock >= i) {
+				bo = to_v3d_bo(&exec->bo[contended_lock]->base);
+
+				ww_mutex_unlock(&bo->resv->lock);
+			}
+
+			if (ret == -EDEADLK) {
+				contended_lock = i;
+				goto retry;
+			}
+
+			ww_acquire_done(acquire_ctx);
+			return ret;
+		}
+	}
+
+	ww_acquire_done(acquire_ctx);
+
+	/* Reserve space for our shared (read-only) fence references,
+	 * before we commit the CL to the hardware.
+	 */
+	for (i = 0; i < exec->bo_count; i++) {
+		bo = to_v3d_bo(&exec->bo[i]->base);
+
+		ret = reservation_object_reserve_shared(bo->resv);
+		if (ret) {
+			v3d_unlock_bo_reservations(dev, exec, acquire_ctx);
+			return ret;
+		}
+	}
+
+	return 0;
+}
+
+/**
+ * v3d_cl_lookup_bos() - Sets up exec->bo[] with the GEM objects
+ * referenced by the job.
+ * @dev: DRM device
+ * @file_priv: DRM file for this fd
+ * @exec: V3D job being set up
+ *
+ * The command validator needs to reference BOs by their index within
+ * the submitted job's BO list.  This does the validation of the job's
+ * BO list and reference counting for the lifetime of the job.
+ *
+ * Note that this function doesn't need to unreference the BOs on
+ * failure, because that will happen at v3d_exec_cleanup() time.
+ */
+static int
+v3d_cl_lookup_bos(struct drm_device *dev,
+		  struct drm_file *file_priv,
+		  struct drm_v3d_submit_cl *args,
+		  struct v3d_exec_info *exec)
+{
+	u32 *handles;
+	int ret = 0;
+	int i;
+
+	exec->bo_count = args->bo_handle_count;
+
+	if (!exec->bo_count) {
+		/* See comment on bo_index for why we have to check
+		 * this.
+		 */
+		DRM_DEBUG("Rendering requires BOs\n");
+		return -EINVAL;
+	}
+
+	exec->bo = kvmalloc_array(exec->bo_count,
+				  sizeof(struct drm_gem_cma_object *),
+				  GFP_KERNEL | __GFP_ZERO);
+	if (!exec->bo) {
+		DRM_DEBUG("Failed to allocate validated BO pointers\n");
+		return -ENOMEM;
+	}
+
+	handles = kvmalloc_array(exec->bo_count, sizeof(u32), GFP_KERNEL);
+	if (!handles) {
+		ret = -ENOMEM;
+		DRM_DEBUG("Failed to allocate incoming GEM handles\n");
+		goto fail;
+	}
+
+	if (copy_from_user(handles,
+			   (void __user *)(uintptr_t)args->bo_handles,
+			   exec->bo_count * sizeof(u32))) {
+		ret = -EFAULT;
+		DRM_DEBUG("Failed to copy in GEM handles\n");
+		goto fail;
+	}
+
+	spin_lock(&file_priv->table_lock);
+	for (i = 0; i < exec->bo_count; i++) {
+		struct drm_gem_object *bo = idr_find(&file_priv->object_idr,
+						     handles[i]);
+		if (!bo) {
+			DRM_DEBUG("Failed to look up GEM BO %d: %d\n",
+				  i, handles[i]);
+			ret = -ENOENT;
+			spin_unlock(&file_priv->table_lock);
+			goto fail;
+		}
+		drm_gem_object_get(bo);
+		exec->bo[i] = to_v3d_bo(bo);
+	}
+	spin_unlock(&file_priv->table_lock);
+
+fail:
+	kvfree(handles);
+	return ret;
+}
+
+static void
+v3d_exec_cleanup(struct kref *ref)
+{
+	struct v3d_exec_info *exec = container_of(ref, struct v3d_exec_info,
+						  refcount);
+	struct v3d_dev *v3d = exec->v3d;
+	unsigned int i;
+	struct v3d_bo *bo, *save;
+
+	dma_fence_put(exec->bin.in_fence);
+	dma_fence_put(exec->render.in_fence);
+
+	dma_fence_put(exec->bin.done_fence);
+	dma_fence_put(exec->render.done_fence);
+
+	dma_fence_put(exec->bin_done_fence);
+
+	for (i = 0; i < exec->bo_count; i++)
+		drm_gem_object_put_unlocked(&exec->bo[i]->base);
+	kvfree(exec->bo);
+
+	list_for_each_entry_safe(bo, save, &exec->unref_list, unref_head) {
+		drm_gem_object_put_unlocked(&bo->base);
+	}
+
+	pm_runtime_mark_last_busy(v3d->dev);
+	pm_runtime_put_autosuspend(v3d->dev);
+
+	kfree(exec);
+}
+
+void v3d_exec_put(struct v3d_exec_info *exec)
+{
+	kref_put(&exec->refcount, v3d_exec_cleanup);
+}
+
+int
+v3d_wait_bo_ioctl(struct drm_device *dev, void *data,
+		  struct drm_file *file_priv)
+{
+	int ret;
+	struct drm_v3d_wait_bo *args = data;
+	struct drm_gem_object *gem_obj;
+	struct v3d_bo *bo;
+	ktime_t start = ktime_get();
+	u64 delta_ns;
+	unsigned long timeout_jiffies =
+		nsecs_to_jiffies_timeout(args->timeout_ns);
+
+	if (args->pad != 0)
+		return -EINVAL;
+
+	gem_obj = drm_gem_object_lookup(file_priv, args->handle);
+	if (!gem_obj) {
+		DRM_DEBUG("Failed to look up GEM BO %d\n", args->handle);
+		return -EINVAL;
+	}
+	bo = to_v3d_bo(gem_obj);
+
+	ret = reservation_object_wait_timeout_rcu(bo->resv,
+						  true, true,
+						  timeout_jiffies);
+
+	if (ret == 0)
+		ret = -ETIME;
+	else if (ret > 0)
+		ret = 0;
+
+	/* Decrement the user's timeout, in case we got interrupted
+	 * such that the ioctl will be restarted.
+	 */
+	delta_ns = ktime_to_ns(ktime_sub(ktime_get(), start));
+	if (delta_ns < args->timeout_ns)
+		args->timeout_ns -= delta_ns;
+	else
+		args->timeout_ns = 0;
+
+	/* Asked to wait beyond the jiffie/scheduler precision? */
+	if (ret == -ETIME && args->timeout_ns)
+		ret = -EAGAIN;
+
+	drm_gem_object_put_unlocked(gem_obj);
+
+	return ret;
+}
+
+/**
+ * v3d_submit_cl_ioctl() - Submits a job (frame) to the V3D.
+ * @dev: DRM device
+ * @data: ioctl argument
+ * @file_priv: DRM file for this fd
+ *
+ * This is the main entrypoint for userspace to submit a 3D frame to
+ * the GPU.  Userspace provides the binner command list (if
+ * applicable), and the kernel sets up the render command list to draw
+ * to the framebuffer described in the ioctl, using the command lists
+ * that the 3D engine's binner will produce.
+ */
+int
+v3d_submit_cl_ioctl(struct drm_device *dev, void *data,
+		    struct drm_file *file_priv)
+{
+	struct v3d_dev *v3d = to_v3d_dev(dev);
+	struct v3d_file_priv *v3d_priv = file_priv->driver_priv;
+	struct drm_v3d_submit_cl *args = data;
+	struct v3d_exec_info *exec;
+	struct ww_acquire_ctx acquire_ctx;
+	struct drm_syncobj *sync_out;
+	int ret = 0;
+
+	if (args->pad != 0) {
+		DRM_INFO("pad must be zero: %d\n", args->pad);
+		return -EINVAL;
+	}
+
+	exec = kcalloc(1, sizeof(*exec), GFP_KERNEL);
+	if (!exec)
+		return -ENOMEM;
+
+	ret = pm_runtime_get_sync(v3d->dev);
+	if (ret < 0) {
+		kfree(exec);
+		return ret;
+	}
+
+	kref_init(&exec->refcount);
+
+	ret = drm_syncobj_find_fence(file_priv, args->in_sync_bcl,
+				     &exec->bin.in_fence);
+	if (ret == -EINVAL)
+		goto fail;
+
+	ret = drm_syncobj_find_fence(file_priv, args->in_sync_rcl,
+				     &exec->render.in_fence);
+	if (ret == -EINVAL)
+		goto fail;
+
+	exec->qma = args->qma;
+	exec->qms = args->qms;
+	exec->qts = args->qts;
+	exec->bin.exec = exec;
+	exec->bin.start = args->bcl_start;
+	exec->bin.end = args->bcl_end;
+	exec->render.exec = exec;
+	exec->render.start = args->rcl_start;
+	exec->render.end = args->rcl_end;
+	exec->v3d = v3d;
+	INIT_LIST_HEAD(&exec->unref_list);
+
+	ret = v3d_cl_lookup_bos(dev, file_priv, args, exec);
+	if (ret)
+		goto fail;
+
+	ret = v3d_lock_bo_reservations(dev, exec, &acquire_ctx);
+	if (ret)
+		goto fail;
+
+	mutex_lock(&v3d->sched_lock);
+	if (exec->bin.start != exec->bin.end) {
+		ret = drm_sched_job_init(&exec->bin.base,
+					 &v3d->queue[V3D_BIN].sched,
+					 &v3d_priv->sched_entity[V3D_BIN],
+					 v3d_priv);
+		if (ret)
+			goto fail_unreserve;
+
+		exec->bin_done_fence =
+			dma_fence_get(&exec->bin.base.s_fence->finished);
+
+		kref_get(&exec->refcount); /* put by scheduler job completion */
+		drm_sched_entity_push_job(&exec->bin.base,
+					  &v3d_priv->sched_entity[V3D_BIN]);
+	}
+
+	ret = drm_sched_job_init(&exec->render.base,
+				 &v3d->queue[V3D_RENDER].sched,
+				 &v3d_priv->sched_entity[V3D_RENDER],
+				 v3d_priv);
+	if (ret)
+		goto fail_unreserve;
+
+	kref_get(&exec->refcount); /* put by scheduler job completion */
+	drm_sched_entity_push_job(&exec->render.base,
+				  &v3d_priv->sched_entity[V3D_RENDER]);
+	mutex_unlock(&v3d->sched_lock);
+
+	v3d_attach_object_fences(exec);
+
+	v3d_unlock_bo_reservations(dev, exec, &acquire_ctx);
+
+	/* Update the return sync object for the */
+	sync_out = drm_syncobj_find(file_priv, args->out_sync);
+	if (sync_out) {
+		drm_syncobj_replace_fence(sync_out,
+					  &exec->render.base.s_fence->finished);
+		drm_syncobj_put(sync_out);
+	}
+
+	v3d_exec_put(exec);
+
+	return 0;
+
+fail_unreserve:
+	mutex_unlock(&v3d->sched_lock);
+	v3d_unlock_bo_reservations(dev, exec, &acquire_ctx);
+fail:
+	v3d_exec_put(exec);
+
+	return ret;
+}
+
+int
+v3d_gem_init(struct drm_device *dev)
+{
+	struct v3d_dev *v3d = to_v3d_dev(dev);
+	u32 pt_size = 4096 * 1024;
+	int ret, i;
+
+	for (i = 0; i < V3D_MAX_QUEUES; i++)
+		v3d->queue[i].fence_context = dma_fence_context_alloc(1);
+
+	spin_lock_init(&v3d->mm_lock);
+	spin_lock_init(&v3d->job_lock);
+	mutex_init(&v3d->bo_lock);
+	mutex_init(&v3d->reset_lock);
+	mutex_init(&v3d->sched_lock);
+
+	/* Note: We don't allocate address 0.  Various bits of HW
+	 * treat 0 as special, such as the occlusion query counters
+	 * where 0 means "disabled".
+	 */
+	drm_mm_init(&v3d->mm, 1, pt_size / sizeof(u32) - 1);
+
+	v3d->pt = dma_alloc_wc(v3d->dev, pt_size,
+			       &v3d->pt_paddr,
+			       GFP_KERNEL | __GFP_NOWARN | __GFP_ZERO);
+	if (!v3d->pt) {
+		drm_mm_takedown(&v3d->mm);
+		dev_err(v3d->dev,
+			"Failed to allocate page tables. "
+			"Please ensure you have CMA enabled.\n");
+		return -ENOMEM;
+	}
+
+	v3d_init_hw_state(v3d);
+	v3d_mmu_set_page_table(v3d);
+
+	ret = v3d_sched_init(v3d);
+	if (ret) {
+		drm_mm_takedown(&v3d->mm);
+		dma_free_coherent(v3d->dev, 4096 * 1024, (void *)v3d->pt,
+				  v3d->pt_paddr);
+	}
+
+	return 0;
+}
+
+void
+v3d_gem_destroy(struct drm_device *dev)
+{
+	struct v3d_dev *v3d = to_v3d_dev(dev);
+
+	v3d_sched_fini(v3d);
+
+	/* Waiting for exec to finish would need to be done before
+	 * unregistering V3D.
+	 */
+	WARN_ON(v3d->bin_job);
+	WARN_ON(v3d->render_job);
+
+	drm_mm_takedown(&v3d->mm);
+
+	dma_free_coherent(v3d->dev, 4096 * 1024, (void *)v3d->pt, v3d->pt_paddr);
+}
diff --git a/drivers/gpu/drm/v3d/v3d_irq.c b/drivers/gpu/drm/v3d/v3d_irq.c
new file mode 100644
index 000000000000..e07514eb11b5
--- /dev/null
+++ b/drivers/gpu/drm/v3d/v3d_irq.c
@@ -0,0 +1,203 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* Copyright (C) 2014-2018 Broadcom */
+
+/**
+ * DOC: Interrupt management for the V3D engine
+ *
+ * When we take a binning or rendering flush done interrupt, we need
+ * to signal the fence for that job so that the scheduler can queue up
+ * the next one and unblock any waiters.
+ *
+ * When we take the binner out of memory interrupt, we need to
+ * allocate some new memory and pass it to the binner so that the
+ * current job can make progress.
+ */
+
+#include "v3d_drv.h"
+#include "v3d_regs.h"
+
+#define V3D_CORE_IRQS ((u32)(V3D_INT_OUTOMEM |	\
+			     V3D_INT_FLDONE |	\
+			     V3D_INT_FRDONE |	\
+			     V3D_INT_GMPV))
+
+#define V3D_HUB_IRQS ((u32)(V3D_HUB_INT_MMU_WRV |	\
+			    V3D_HUB_INT_MMU_PTI |	\
+			    V3D_HUB_INT_MMU_CAP))
+
+static void
+v3d_overflow_mem_work(struct work_struct *work)
+{
+	struct v3d_dev *v3d =
+		container_of(work, struct v3d_dev, overflow_mem_work);
+	struct drm_device *dev = &v3d->drm;
+	struct v3d_bo *bo = v3d_bo_create(dev, NULL /* XXX: GMP */, 256 * 1024);
+	unsigned long irqflags;
+
+	if (IS_ERR(bo)) {
+		DRM_ERROR("Couldn't allocate binner overflow mem\n");
+		return;
+	}
+
+	/* We lost a race, and our work task came in after the bin job
+	 * completed and exited.  This can happen because the HW
+	 * signals OOM before it's fully OOM, so the binner might just
+	 * barely complete.
+	 *
+	 * If we lose the race and our work task comes in after a new
+	 * bin job got scheduled, that's fine.  We'll just give them
+	 * some binner pool anyway.
+	 */
+	spin_lock_irqsave(&v3d->job_lock, irqflags);
+	if (!v3d->bin_job) {
+		spin_unlock_irqrestore(&v3d->job_lock, irqflags);
+		goto out;
+	}
+
+	drm_gem_object_get(&bo->base);
+	list_add_tail(&bo->unref_head, &v3d->bin_job->unref_list);
+	spin_unlock_irqrestore(&v3d->job_lock, irqflags);
+
+	V3D_CORE_WRITE(0, V3D_PTB_BPOA, bo->node.start << PAGE_SHIFT);
+	V3D_CORE_WRITE(0, V3D_PTB_BPOS, bo->base.size);
+
+out:
+	drm_gem_object_put_unlocked(&bo->base);
+}
+
+static irqreturn_t
+v3d_irq(int irq, void *arg)
+{
+	struct v3d_dev *v3d = arg;
+	u32 intsts;
+	irqreturn_t status = IRQ_NONE;
+
+	intsts = V3D_CORE_READ(0, V3D_CTL_INT_STS);
+
+	/* Acknowledge the interrupts we're handling here. */
+	V3D_CORE_WRITE(0, V3D_CTL_INT_CLR, intsts);
+
+	if (intsts & V3D_INT_OUTOMEM) {
+		/* Note that the OOM status is edge signaled, so the
+		 * interrupt won't happen again until the we actually
+		 * add more memory.
+		 */
+		schedule_work(&v3d->overflow_mem_work);
+		status = IRQ_HANDLED;
+	}
+
+	if (intsts & V3D_INT_FLDONE) {
+		dma_fence_signal(v3d->bin_job->bin.done_fence);
+		status = IRQ_HANDLED;
+	}
+
+	if (intsts & V3D_INT_FRDONE) {
+		dma_fence_signal(v3d->render_job->render.done_fence);
+		status = IRQ_HANDLED;
+	}
+
+	/* We shouldn't be triggering these if we have GMP in
+	 * always-allowed mode.
+	 */
+	if (intsts & V3D_INT_GMPV)
+		dev_err(v3d->dev, "GMP violation\n");
+
+	return status;
+}
+
+static irqreturn_t
+v3d_hub_irq(int irq, void *arg)
+{
+	struct v3d_dev *v3d = arg;
+	u32 intsts;
+	irqreturn_t status = IRQ_NONE;
+
+	intsts = V3D_READ(V3D_HUB_INT_STS);
+
+	/* Acknowledge the interrupts we're handling here. */
+	V3D_WRITE(V3D_HUB_INT_CLR, intsts);
+
+	if (intsts & (V3D_HUB_INT_MMU_WRV |
+		      V3D_HUB_INT_MMU_PTI |
+		      V3D_HUB_INT_MMU_CAP)) {
+		u32 axi_id = V3D_READ(V3D_MMU_VIO_ID);
+		u64 vio_addr = (u64)V3D_READ(V3D_MMU_VIO_ADDR) << 8;
+
+		dev_err(v3d->dev, "MMU error from client %d at 0x%08llx%s%s%s\n",
+			axi_id, (long long)vio_addr,
+			((intsts & V3D_HUB_INT_MMU_WRV) ?
+			 ", write violation" : ""),
+			((intsts & V3D_HUB_INT_MMU_PTI) ?
+			 ", pte invalid" : ""),
+			((intsts & V3D_HUB_INT_MMU_CAP) ?
+			 ", cap exceeded" : ""));
+		status = IRQ_HANDLED;
+	}
+
+	return status;
+}
+
+void
+v3d_irq_init(struct v3d_dev *v3d)
+{
+	int ret, core;
+
+	INIT_WORK(&v3d->overflow_mem_work, v3d_overflow_mem_work);
+
+	/* Clear any pending interrupts someone might have left around
+	 * for us.
+	 */
+	for (core = 0; core < v3d->cores; core++)
+		V3D_CORE_WRITE(core, V3D_CTL_INT_CLR, V3D_CORE_IRQS);
+	V3D_WRITE(V3D_HUB_INT_CLR, V3D_HUB_IRQS);
+
+	ret = devm_request_irq(v3d->dev, platform_get_irq(v3d->pdev, 0),
+			       v3d_hub_irq, IRQF_SHARED,
+			       "v3d_hub", v3d);
+	ret = devm_request_irq(v3d->dev, platform_get_irq(v3d->pdev, 1),
+			       v3d_irq, IRQF_SHARED,
+			       "v3d_core0", v3d);
+	if (ret)
+		dev_err(v3d->dev, "IRQ setup failed: %d\n", ret);
+
+	v3d_irq_enable(v3d);
+}
+
+void
+v3d_irq_enable(struct v3d_dev *v3d)
+{
+	int core;
+
+	/* Enable our set of interrupts, masking out any others. */
+	for (core = 0; core < v3d->cores; core++) {
+		V3D_CORE_WRITE(core, V3D_CTL_INT_MSK_SET, ~V3D_CORE_IRQS);
+		V3D_CORE_WRITE(core, V3D_CTL_INT_MSK_CLR, V3D_CORE_IRQS);
+	}
+
+	V3D_WRITE(V3D_HUB_INT_MSK_SET, ~V3D_HUB_IRQS);
+	V3D_WRITE(V3D_HUB_INT_MSK_CLR, V3D_HUB_IRQS);
+}
+
+void
+v3d_irq_disable(struct v3d_dev *v3d)
+{
+	int core;
+
+	/* Disable all interrupts. */
+	for (core = 0; core < v3d->cores; core++)
+		V3D_CORE_WRITE(core, V3D_CTL_INT_MSK_SET, ~0);
+	V3D_WRITE(V3D_HUB_INT_MSK_SET, ~0);
+
+	/* Clear any pending interrupts we might have left. */
+	for (core = 0; core < v3d->cores; core++)
+		V3D_CORE_WRITE(core, V3D_CTL_INT_CLR, V3D_CORE_IRQS);
+	V3D_WRITE(V3D_HUB_INT_CLR, V3D_HUB_IRQS);
+
+	cancel_work_sync(&v3d->overflow_mem_work);
+}
+
+/** Reinitializes interrupt registers when a GPU reset is performed. */
+void v3d_irq_reset(struct v3d_dev *v3d)
+{
+	v3d_irq_enable(v3d);
+}
diff --git a/drivers/gpu/drm/v3d/v3d_mmu.c b/drivers/gpu/drm/v3d/v3d_mmu.c
new file mode 100644
index 000000000000..b00f97c31b70
--- /dev/null
+++ b/drivers/gpu/drm/v3d/v3d_mmu.c
@@ -0,0 +1,122 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* Copyright (C) 2017-2018 Broadcom */
+
+/**
+ * DOC: Broadcom V3D MMU
+ *
+ * The V3D 3.x hardware (compared to VC4) now includes an MMU.  It has
+ * a single level of page tables for the V3D's 4GB address space to
+ * map to AXI bus addresses, thus it could need up to 4MB of
+ * physically contiguous memory to store the PTEs.
+ *
+ * Because the 4MB of contiguous memory for page tables is precious,
+ * and switching between them is expensive, we load all BOs into the
+ * same 4GB address space.
+ *
+ * To protect clients from each other, we should use the GMP to
+ * quickly mask out (at 128kb granularity) what pages are available to
+ * each client.  This is not yet implemented.
+ */
+
+#include "v3d_drv.h"
+#include "v3d_regs.h"
+
+#define V3D_MMU_PAGE_SHIFT 12
+
+/* Note: All PTEs for the 1MB superpage must be filled with the
+ * superpage bit set.
+ */
+#define V3D_PTE_SUPERPAGE BIT(31)
+#define V3D_PTE_WRITEABLE BIT(29)
+#define V3D_PTE_VALID BIT(28)
+
+static int v3d_mmu_flush_all(struct v3d_dev *v3d)
+{
+	int ret;
+
+	/* Make sure that another flush isn't already running when we
+	 * start this one.
+	 */
+	ret = wait_for(!(V3D_READ(V3D_MMU_CTL) &
+			 V3D_MMU_CTL_TLB_CLEARING), 100);
+	if (ret)
+		dev_err(v3d->dev, "TLB clear wait idle pre-wait failed\n");
+
+	V3D_WRITE(V3D_MMU_CTL, V3D_READ(V3D_MMU_CTL) |
+		  V3D_MMU_CTL_TLB_CLEAR);
+
+	V3D_WRITE(V3D_MMUC_CONTROL,
+		  V3D_MMUC_CONTROL_FLUSH |
+		  V3D_MMUC_CONTROL_ENABLE);
+
+	ret = wait_for(!(V3D_READ(V3D_MMU_CTL) &
+			 V3D_MMU_CTL_TLB_CLEARING), 100);
+	if (ret) {
+		dev_err(v3d->dev, "TLB clear wait idle failed\n");
+		return ret;
+	}
+
+	ret = wait_for(!(V3D_READ(V3D_MMUC_CONTROL) &
+			 V3D_MMUC_CONTROL_FLUSHING), 100);
+	if (ret)
+		dev_err(v3d->dev, "MMUC flush wait idle failed\n");
+
+	return ret;
+}
+
+int v3d_mmu_set_page_table(struct v3d_dev *v3d)
+{
+	V3D_WRITE(V3D_MMU_PT_PA_BASE, v3d->pt_paddr >> V3D_MMU_PAGE_SHIFT);
+	V3D_WRITE(V3D_MMU_CTL,
+		  V3D_MMU_CTL_ENABLE |
+		  V3D_MMU_CTL_PT_INVALID |
+		  V3D_MMU_CTL_PT_INVALID_ABORT |
+		  V3D_MMU_CTL_WRITE_VIOLATION_ABORT |
+		  V3D_MMU_CTL_CAP_EXCEEDED_ABORT);
+	V3D_WRITE(V3D_MMU_ILLEGAL_ADDR,
+		  (v3d->mmu_scratch_paddr >> V3D_MMU_PAGE_SHIFT) |
+		  V3D_MMU_ILLEGAL_ADDR_ENABLE);
+	V3D_WRITE(V3D_MMUC_CONTROL, V3D_MMUC_CONTROL_ENABLE);
+
+	return v3d_mmu_flush_all(v3d);
+}
+
+void v3d_mmu_insert_ptes(struct v3d_bo *bo)
+{
+	struct v3d_dev *v3d = to_v3d_dev(bo->base.dev);
+	u32 page = bo->node.start;
+	u32 page_prot = V3D_PTE_WRITEABLE | V3D_PTE_VALID;
+	unsigned int count;
+	struct scatterlist *sgl;
+
+	for_each_sg(bo->sgt->sgl, sgl, bo->sgt->nents, count) {
+		u32 page_address = sg_dma_address(sgl) >> V3D_MMU_PAGE_SHIFT;
+		u32 pte = page_prot | page_address;
+		u32 i;
+
+		BUG_ON(page_address + (sg_dma_len(sgl) >> V3D_MMU_PAGE_SHIFT) >=
+		       BIT(24));
+
+		for (i = 0; i < sg_dma_len(sgl) >> V3D_MMU_PAGE_SHIFT; i++)
+			v3d->pt[page++] = pte + i;
+	}
+
+	WARN_ON_ONCE(page - bo->node.start !=
+		     bo->base.size >> V3D_MMU_PAGE_SHIFT);
+
+	if (v3d_mmu_flush_all(v3d))
+		dev_err(v3d->dev, "MMU flush timeout\n");
+}
+
+void v3d_mmu_remove_ptes(struct v3d_bo *bo)
+{
+	struct v3d_dev *v3d = to_v3d_dev(bo->base.dev);
+	u32 npages = bo->base.size >> V3D_MMU_PAGE_SHIFT;
+	u32 page;
+
+	for (page = bo->node.start; page < bo->node.start + npages; page++)
+		v3d->pt[page] = 0;
+
+	if (v3d_mmu_flush_all(v3d))
+		dev_err(v3d->dev, "MMU flush timeout\n");
+}
diff --git a/drivers/gpu/drm/v3d/v3d_regs.h b/drivers/gpu/drm/v3d/v3d_regs.h
new file mode 100644
index 000000000000..854046565989
--- /dev/null
+++ b/drivers/gpu/drm/v3d/v3d_regs.h
@@ -0,0 +1,296 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* Copyright (C) 2017-2018 Broadcom */
+
+#ifndef V3D_REGS_H
+#define V3D_REGS_H
+
+#include <linux/bitops.h>
+
+#define V3D_MASK(high, low) ((u32)GENMASK(high, low))
+/* Using the GNU statement expression extension */
+#define V3D_SET_FIELD(value, field)					\
+	({								\
+		u32 fieldval = (value) << field##_SHIFT;		\
+		WARN_ON((fieldval & ~field##_MASK) != 0);		\
+		fieldval & field##_MASK;				\
+	 })
+
+#define V3D_GET_FIELD(word, field) (((word) & field##_MASK) >>		\
+				    field##_SHIFT)
+
+/* Hub registers for shared hardware between V3D cores. */
+
+#define V3D_HUB_AXICFG                                 0x00000
+# define V3D_HUB_AXICFG_MAX_LEN_MASK                   V3D_MASK(3, 0)
+# define V3D_HUB_AXICFG_MAX_LEN_SHIFT                  0
+#define V3D_HUB_UIFCFG                                 0x00004
+#define V3D_HUB_IDENT0                                 0x00008
+
+#define V3D_HUB_IDENT1                                 0x0000c
+# define V3D_HUB_IDENT1_WITH_MSO                       BIT(19)
+# define V3D_HUB_IDENT1_WITH_TSY                       BIT(18)
+# define V3D_HUB_IDENT1_WITH_TFU                       BIT(17)
+# define V3D_HUB_IDENT1_WITH_L3C                       BIT(16)
+# define V3D_HUB_IDENT1_NHOSTS_MASK                    V3D_MASK(15, 12)
+# define V3D_HUB_IDENT1_NHOSTS_SHIFT                   12
+# define V3D_HUB_IDENT1_NCORES_MASK                    V3D_MASK(11, 8)
+# define V3D_HUB_IDENT1_NCORES_SHIFT                   8
+# define V3D_HUB_IDENT1_REV_MASK                       V3D_MASK(7, 4)
+# define V3D_HUB_IDENT1_REV_SHIFT                      4
+# define V3D_HUB_IDENT1_TVER_MASK                      V3D_MASK(3, 0)
+# define V3D_HUB_IDENT1_TVER_SHIFT                     0
+
+#define V3D_HUB_IDENT2                                 0x00010
+# define V3D_HUB_IDENT2_WITH_MMU                       BIT(8)
+# define V3D_HUB_IDENT2_L3C_NKB_MASK                   V3D_MASK(7, 0)
+# define V3D_HUB_IDENT2_L3C_NKB_SHIFT                  0
+
+#define V3D_HUB_IDENT3                                 0x00014
+# define V3D_HUB_IDENT3_IPREV_MASK                     V3D_MASK(15, 8)
+# define V3D_HUB_IDENT3_IPREV_SHIFT                    8
+# define V3D_HUB_IDENT3_IPIDX_MASK                     V3D_MASK(7, 0)
+# define V3D_HUB_IDENT3_IPIDX_SHIFT                    0
+
+#define V3D_HUB_INT_STS                                0x00050
+#define V3D_HUB_INT_SET                                0x00054
+#define V3D_HUB_INT_CLR                                0x00058
+#define V3D_HUB_INT_MSK_STS                            0x0005c
+#define V3D_HUB_INT_MSK_SET                            0x00060
+#define V3D_HUB_INT_MSK_CLR                            0x00064
+# define V3D_HUB_INT_MMU_WRV                           BIT(5)
+# define V3D_HUB_INT_MMU_PTI                           BIT(4)
+# define V3D_HUB_INT_MMU_CAP                           BIT(3)
+# define V3D_HUB_INT_MSO                               BIT(2)
+# define V3D_HUB_INT_TFUC                              BIT(1)
+# define V3D_HUB_INT_TFUF                              BIT(0)
+
+#define V3D_GCA_CACHE_CTRL                             0x0000c
+# define V3D_GCA_CACHE_CTRL_FLUSH                      BIT(0)
+
+#define V3D_GCA_SAFE_SHUTDOWN                          0x000b0
+# define V3D_GCA_SAFE_SHUTDOWN_EN                      BIT(0)
+
+#define V3D_GCA_SAFE_SHUTDOWN_ACK                      0x000b4
+# define V3D_GCA_SAFE_SHUTDOWN_ACK_ACKED               3
+
+# define V3D_TOP_GR_BRIDGE_REVISION                    0x00000
+# define V3D_TOP_GR_BRIDGE_MAJOR_MASK                  V3D_MASK(15, 8)
+# define V3D_TOP_GR_BRIDGE_MAJOR_SHIFT                 8
+# define V3D_TOP_GR_BRIDGE_MINOR_MASK                  V3D_MASK(7, 0)
+# define V3D_TOP_GR_BRIDGE_MINOR_SHIFT                 0
+
+/* 7268 reset reg */
+# define V3D_TOP_GR_BRIDGE_SW_INIT_0                   0x00008
+# define V3D_TOP_GR_BRIDGE_SW_INIT_0_V3D_CLK_108_SW_INIT BIT(0)
+/* 7278 reset reg */
+# define V3D_TOP_GR_BRIDGE_SW_INIT_1                   0x0000c
+# define V3D_TOP_GR_BRIDGE_SW_INIT_1_V3D_CLK_108_SW_INIT BIT(0)
+
+/* Per-MMU registers. */
+
+#define V3D_MMUC_CONTROL                               0x01000
+# define V3D_MMUC_CONTROL_CLEAR                        BIT(3)
+# define V3D_MMUC_CONTROL_FLUSHING                     BIT(2)
+# define V3D_MMUC_CONTROL_FLUSH                        BIT(1)
+# define V3D_MMUC_CONTROL_ENABLE                       BIT(0)
+
+#define V3D_MMU_CTL                                    0x01200
+# define V3D_MMU_CTL_CAP_EXCEEDED                      BIT(27)
+# define V3D_MMU_CTL_CAP_EXCEEDED_ABORT                BIT(26)
+# define V3D_MMU_CTL_CAP_EXCEEDED_INT                  BIT(25)
+# define V3D_MMU_CTL_CAP_EXCEEDED_EXCEPTION            BIT(24)
+# define V3D_MMU_CTL_PT_INVALID                        BIT(20)
+# define V3D_MMU_CTL_PT_INVALID_ABORT                  BIT(19)
+# define V3D_MMU_CTL_PT_INVALID_INT                    BIT(18)
+# define V3D_MMU_CTL_PT_INVALID_EXCEPTION              BIT(17)
+# define V3D_MMU_CTL_WRITE_VIOLATION                   BIT(16)
+# define V3D_MMU_CTL_WRITE_VIOLATION_ABORT             BIT(11)
+# define V3D_MMU_CTL_WRITE_VIOLATION_INT               BIT(10)
+# define V3D_MMU_CTL_WRITE_VIOLATION_EXCEPTION         BIT(9)
+# define V3D_MMU_CTL_TLB_CLEARING                      BIT(7)
+# define V3D_MMU_CTL_TLB_STATS_CLEAR                   BIT(3)
+# define V3D_MMU_CTL_TLB_CLEAR                         BIT(2)
+# define V3D_MMU_CTL_TLB_STATS_ENABLE                  BIT(1)
+# define V3D_MMU_CTL_ENABLE                            BIT(0)
+
+#define V3D_MMU_PT_PA_BASE                             0x01204
+#define V3D_MMU_HIT                                    0x01208
+#define V3D_MMU_MISSES                                 0x0120c
+#define V3D_MMU_STALLS                                 0x01210
+
+#define V3D_MMU_ADDR_CAP                               0x01214
+# define V3D_MMU_ADDR_CAP_ENABLE                       BIT(31)
+# define V3D_MMU_ADDR_CAP_MPAGE_MASK                   V3D_MASK(11, 0)
+# define V3D_MMU_ADDR_CAP_MPAGE_SHIFT                  0
+
+#define V3D_MMU_SHOOT_DOWN                             0x01218
+# define V3D_MMU_SHOOT_DOWN_SHOOTING                   BIT(29)
+# define V3D_MMU_SHOOT_DOWN_SHOOT                      BIT(28)
+# define V3D_MMU_SHOOT_DOWN_PAGE_MASK                  V3D_MASK(27, 0)
+# define V3D_MMU_SHOOT_DOWN_PAGE_SHIFT                 0
+
+#define V3D_MMU_BYPASS_START                           0x0121c
+#define V3D_MMU_BYPASS_END                             0x01220
+
+/* AXI ID of the access that faulted */
+#define V3D_MMU_VIO_ID                                 0x0122c
+
+/* Address for illegal PTEs to return */
+#define V3D_MMU_ILLEGAL_ADDR                           0x01230
+# define V3D_MMU_ILLEGAL_ADDR_ENABLE                   BIT(31)
+
+/* Address that faulted */
+#define V3D_MMU_VIO_ADDR                               0x01234
+
+/* Per-V3D-core registers */
+
+#define V3D_CTL_IDENT0                                 0x00000
+# define V3D_IDENT0_VER_MASK                           V3D_MASK(31, 24)
+# define V3D_IDENT0_VER_SHIFT                          24
+
+#define V3D_CTL_IDENT1                                 0x00004
+/* Multiples of 1kb */
+# define V3D_IDENT1_VPM_SIZE_MASK                      V3D_MASK(31, 28)
+# define V3D_IDENT1_VPM_SIZE_SHIFT                     28
+# define V3D_IDENT1_NSEM_MASK                          V3D_MASK(23, 16)
+# define V3D_IDENT1_NSEM_SHIFT                         16
+# define V3D_IDENT1_NTMU_MASK                          V3D_MASK(15, 12)
+# define V3D_IDENT1_NTMU_SHIFT                         12
+# define V3D_IDENT1_QUPS_MASK                          V3D_MASK(11, 8)
+# define V3D_IDENT1_QUPS_SHIFT                         8
+# define V3D_IDENT1_NSLC_MASK                          V3D_MASK(7, 4)
+# define V3D_IDENT1_NSLC_SHIFT                         4
+# define V3D_IDENT1_REV_MASK                           V3D_MASK(3, 0)
+# define V3D_IDENT1_REV_SHIFT                          0
+
+#define V3D_CTL_IDENT2                                 0x00008
+# define V3D_IDENT2_BCG_INT                            BIT(28)
+
+#define V3D_CTL_MISCCFG                                0x00018
+# define V3D_MISCCFG_OVRTMUOUT                         BIT(0)
+
+#define V3D_CTL_L2CACTL                                0x00020
+# define V3D_L2CACTL_L2CCLR                            BIT(2)
+# define V3D_L2CACTL_L2CDIS                            BIT(1)
+# define V3D_L2CACTL_L2CENA                            BIT(0)
+
+#define V3D_CTL_SLCACTL                                0x00024
+# define V3D_SLCACTL_TVCCS_MASK                        V3D_MASK(27, 24)
+# define V3D_SLCACTL_TVCCS_SHIFT                       24
+# define V3D_SLCACTL_TDCCS_MASK                        V3D_MASK(19, 16)
+# define V3D_SLCACTL_TDCCS_SHIFT                       16
+# define V3D_SLCACTL_UCC_MASK                          V3D_MASK(11, 8)
+# define V3D_SLCACTL_UCC_SHIFT                         8
+# define V3D_SLCACTL_ICC_MASK                          V3D_MASK(3, 0)
+# define V3D_SLCACTL_ICC_SHIFT                         0
+
+#define V3D_CTL_L2TCACTL                               0x00030
+# define V3D_L2TCACTL_TMUWCF                           BIT(8)
+# define V3D_L2TCACTL_L2T_NO_WM                        BIT(4)
+# define V3D_L2TCACTL_FLM_FLUSH                        0
+# define V3D_L2TCACTL_FLM_CLEAR                        1
+# define V3D_L2TCACTL_FLM_CLEAN                        2
+# define V3D_L2TCACTL_FLM_MASK                         V3D_MASK(2, 1)
+# define V3D_L2TCACTL_FLM_SHIFT                        1
+# define V3D_L2TCACTL_L2TFLS                           BIT(0)
+#define V3D_CTL_L2TFLSTA                               0x00034
+#define V3D_CTL_L2TFLEND                               0x00038
+
+#define V3D_CTL_INT_STS                                0x00050
+#define V3D_CTL_INT_SET                                0x00054
+#define V3D_CTL_INT_CLR                                0x00058
+#define V3D_CTL_INT_MSK_STS                            0x0005c
+#define V3D_CTL_INT_MSK_SET                            0x00060
+#define V3D_CTL_INT_MSK_CLR                            0x00064
+# define V3D_INT_QPU_MASK                              V3D_MASK(27, 16)
+# define V3D_INT_QPU_SHIFT                             16
+# define V3D_INT_GMPV                                  BIT(5)
+# define V3D_INT_TRFB                                  BIT(4)
+# define V3D_INT_SPILLUSE                              BIT(3)
+# define V3D_INT_OUTOMEM                               BIT(2)
+# define V3D_INT_FLDONE                                BIT(1)
+# define V3D_INT_FRDONE                                BIT(0)
+
+#define V3D_CLE_CT0CS                                  0x00100
+#define V3D_CLE_CT1CS                                  0x00104
+#define V3D_CLE_CTNCS(n) (V3D_CLE_CT0CS + 4 * n)
+#define V3D_CLE_CT0EA                                  0x00108
+#define V3D_CLE_CT1EA                                  0x0010c
+#define V3D_CLE_CTNEA(n) (V3D_CLE_CT0EA + 4 * n)
+#define V3D_CLE_CT0CA                                  0x00110
+#define V3D_CLE_CT1CA                                  0x00114
+#define V3D_CLE_CTNCA(n) (V3D_CLE_CT0CA + 4 * n)
+#define V3D_CLE_CT0RA                                  0x00118
+#define V3D_CLE_CT1RA                                  0x0011c
+#define V3D_CLE_CTNRA(n) (V3D_CLE_CT0RA + 4 * n)
+#define V3D_CLE_CT0LC                                  0x00120
+#define V3D_CLE_CT1LC                                  0x00124
+#define V3D_CLE_CT0PC                                  0x00128
+#define V3D_CLE_CT1PC                                  0x0012c
+#define V3D_CLE_PCS                                    0x00130
+#define V3D_CLE_BFC                                    0x00134
+#define V3D_CLE_RFC                                    0x00138
+#define V3D_CLE_TFBC                                   0x0013c
+#define V3D_CLE_TFIT                                   0x00140
+#define V3D_CLE_CT1CFG                                 0x00144
+#define V3D_CLE_CT1TILECT                              0x00148
+#define V3D_CLE_CT1TSKIP                               0x0014c
+#define V3D_CLE_CT1PTCT                                0x00150
+#define V3D_CLE_CT0SYNC                                0x00154
+#define V3D_CLE_CT1SYNC                                0x00158
+#define V3D_CLE_CT0QTS                                 0x0015c
+# define V3D_CLE_CT0QTS_ENABLE                         BIT(1)
+#define V3D_CLE_CT0QBA                                 0x00160
+#define V3D_CLE_CT1QBA                                 0x00164
+#define V3D_CLE_CTNQBA(n) (V3D_CLE_CT0QBA + 4 * n)
+#define V3D_CLE_CT0QEA                                 0x00168
+#define V3D_CLE_CT1QEA                                 0x0016c
+#define V3D_CLE_CTNQEA(n) (V3D_CLE_CT0QEA + 4 * n)
+#define V3D_CLE_CT0QMA                                 0x00170
+#define V3D_CLE_CT0QMS                                 0x00174
+#define V3D_CLE_CT1QCFG                                0x00178
+/* If set without ETPROC, entirely skip tiles with no primitives. */
+# define V3D_CLE_QCFG_ETFILT                           BIT(7)
+/* If set with ETFILT, just write the clear color to tiles with no
+ * primitives.
+ */
+# define V3D_CLE_QCFG_ETPROC                           BIT(6)
+# define V3D_CLE_QCFG_ETSFLUSH                         BIT(1)
+# define V3D_CLE_QCFG_MCDIS                            BIT(0)
+
+#define V3D_PTB_BPCA                                   0x00300
+#define V3D_PTB_BPCS                                   0x00304
+#define V3D_PTB_BPOA                                   0x00308
+#define V3D_PTB_BPOS                                   0x0030c
+
+#define V3D_PTB_BXCF                                   0x00310
+# define V3D_PTB_BXCF_RWORDERDISA                      BIT(1)
+# define V3D_PTB_BXCF_CLIPDISA                         BIT(0)
+
+#define V3D_GMP_STATUS                                 0x00800
+# define V3D_GMP_STATUS_GMPRST                         BIT(31)
+# define V3D_GMP_STATUS_WR_COUNT_MASK                  V3D_MASK(30, 24)
+# define V3D_GMP_STATUS_WR_COUNT_SHIFT                 24
+# define V3D_GMP_STATUS_RD_COUNT_MASK                  V3D_MASK(22, 16)
+# define V3D_GMP_STATUS_RD_COUNT_SHIFT                 16
+# define V3D_GMP_STATUS_WR_ACTIVE                      BIT(5)
+# define V3D_GMP_STATUS_RD_ACTIVE                      BIT(4)
+# define V3D_GMP_STATUS_CFG_BUSY                       BIT(3)
+# define V3D_GMP_STATUS_CNTOVF                         BIT(2)
+# define V3D_GMP_STATUS_INVPROT                        BIT(1)
+# define V3D_GMP_STATUS_VIO                            BIT(0)
+
+#define V3D_GMP_CFG                                    0x00804
+# define V3D_GMP_CFG_LBURSTEN                          BIT(3)
+# define V3D_GMP_CFG_PGCRSEN                           BIT()
+# define V3D_GMP_CFG_STOP_REQ                          BIT(1)
+# define V3D_GMP_CFG_PROT_ENABLE                       BIT(0)
+
+#define V3D_GMP_VIO_ADDR                               0x00808
+#define V3D_GMP_VIO_TYPE                               0x0080c
+#define V3D_GMP_TABLE_ADDR                             0x00810
+#define V3D_GMP_CLEAR_LOAD                             0x00814
+#define V3D_GMP_PRESERVE_LOAD                          0x00818
+#define V3D_GMP_VALID_LINES                            0x00820
+
+#endif /* V3D_REGS_H */
diff --git a/drivers/gpu/drm/v3d/v3d_sched.c b/drivers/gpu/drm/v3d/v3d_sched.c
new file mode 100644
index 000000000000..a5501581d96b
--- /dev/null
+++ b/drivers/gpu/drm/v3d/v3d_sched.c
@@ -0,0 +1,246 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* Copyright (C) 2018 Broadcom */
+
+/**
+ * DOC: Broadcom V3D scheduling
+ *
+ * The shared DRM GPU scheduler is used to coordinate submitting jobs
+ * to the hardware.  Each DRM fd (roughly a client process) gets its
+ * own scheduler entity, which will process jobs in order.  The GPU
+ * scheduler will round-robin between clients to submit the next job.
+ *
+ * For simplicity, and in order to keep latency low for interactive
+ * jobs when bulk background jobs are queued up, we submit a new job
+ * to the HW only when it has completed the last one, instead of
+ * filling up the CT[01]Q FIFOs with jobs.  Similarly, we use
+ * v3d_job_dependency() to manage the dependency between bin and
+ * render, instead of having the clients submit jobs using the HW's
+ * semaphores to interlock between them.
+ */
+
+#include <linux/kthread.h>
+
+#include "v3d_drv.h"
+#include "v3d_regs.h"
+#include "v3d_trace.h"
+
+static struct v3d_job *
+to_v3d_job(struct drm_sched_job *sched_job)
+{
+	return container_of(sched_job, struct v3d_job, base);
+}
+
+static void
+v3d_job_free(struct drm_sched_job *sched_job)
+{
+	struct v3d_job *job = to_v3d_job(sched_job);
+
+	v3d_exec_put(job->exec);
+}
+
+/**
+ * Returns the fences that the bin job depends on, one by one.
+ * v3d_job_run() won't be called until all of them have been signaled.
+ */
+static struct dma_fence *
+v3d_job_dependency(struct drm_sched_job *sched_job,
+		   struct drm_sched_entity *s_entity)
+{
+	struct v3d_job *job = to_v3d_job(sched_job);
+	struct v3d_exec_info *exec = job->exec;
+	enum v3d_queue q = job == &exec->bin ? V3D_BIN : V3D_RENDER;
+	struct dma_fence *fence;
+
+	fence = job->in_fence;
+	if (fence) {
+		job->in_fence = NULL;
+		return fence;
+	}
+
+	if (q == V3D_RENDER) {
+		/* If we had a bin job, the render job definitely depends on
+		 * it. We first have to wait for bin to be scheduled, so that
+		 * its done_fence is created.
+		 */
+		fence = exec->bin_done_fence;
+		if (fence) {
+			exec->bin_done_fence = NULL;
+			return fence;
+		}
+	}
+
+	/* XXX: Wait on a fence for switching the GMP if necessary,
+	 * and then do so.
+	 */
+
+	return fence;
+}
+
+static struct dma_fence *v3d_job_run(struct drm_sched_job *sched_job)
+{
+	struct v3d_job *job = to_v3d_job(sched_job);
+	struct v3d_exec_info *exec = job->exec;
+	enum v3d_queue q = job == &exec->bin ? V3D_BIN : V3D_RENDER;
+	struct v3d_dev *v3d = exec->v3d;
+	struct drm_device *dev = &v3d->drm;
+	struct dma_fence *fence;
+	unsigned long irqflags;
+
+	if (unlikely(job->base.s_fence->finished.error))
+		return NULL;
+
+	/* Lock required around bin_job update vs
+	 * v3d_overflow_mem_work().
+	 */
+	spin_lock_irqsave(&v3d->job_lock, irqflags);
+	if (q == V3D_BIN) {
+		v3d->bin_job = job->exec;
+
+		/* Clear out the overflow allocation, so we don't
+		 * reuse the overflow attached to a previous job.
+		 */
+		V3D_CORE_WRITE(0, V3D_PTB_BPOS, 0);
+	} else {
+		v3d->render_job = job->exec;
+	}
+	spin_unlock_irqrestore(&v3d->job_lock, irqflags);
+
+	/* Can we avoid this flush when q==RENDER?  We need to be
+	 * careful of scheduling, though -- imagine job0 rendering to
+	 * texture and job1 reading, and them being executed as bin0,
+	 * bin1, render0, render1, so that render1's flush at bin time
+	 * wasn't enough.
+	 */
+	v3d_invalidate_caches(v3d);
+
+	fence = v3d_fence_create(v3d, q);
+	if (IS_ERR(fence))
+		return NULL;
+
+	if (job->done_fence)
+		dma_fence_put(job->done_fence);
+	job->done_fence = dma_fence_get(fence);
+
+	trace_v3d_submit_cl(dev, q == V3D_RENDER, to_v3d_fence(fence)->seqno,
+			    job->start, job->end);
+
+	if (q == V3D_BIN) {
+		if (exec->qma) {
+			V3D_CORE_WRITE(0, V3D_CLE_CT0QMA, exec->qma);
+			V3D_CORE_WRITE(0, V3D_CLE_CT0QMS, exec->qms);
+		}
+		if (exec->qts) {
+			V3D_CORE_WRITE(0, V3D_CLE_CT0QTS,
+				       V3D_CLE_CT0QTS_ENABLE |
+				       exec->qts);
+		}
+	} else {
+		/* XXX: Set the QCFG */
+	}
+
+	/* Set the current and end address of the control list.
+	 * Writing the end register is what starts the job.
+	 */
+	V3D_CORE_WRITE(0, V3D_CLE_CTNQBA(q), job->start);
+	V3D_CORE_WRITE(0, V3D_CLE_CTNQEA(q), job->end);
+
+	return fence;
+}
+
+static void
+v3d_job_timedout(struct drm_sched_job *sched_job)
+{
+	struct v3d_job *job = to_v3d_job(sched_job);
+	struct v3d_exec_info *exec = job->exec;
+	struct v3d_dev *v3d = exec->v3d;
+	enum v3d_queue job_q = job == &exec->bin ? V3D_BIN : V3D_RENDER;
+	enum v3d_queue q;
+	u32 ctca = V3D_CORE_READ(0, V3D_CLE_CTNCA(job_q));
+	u32 ctra = V3D_CORE_READ(0, V3D_CLE_CTNRA(job_q));
+
+	/* If the current address or return address have changed, then
+	 * the GPU has probably made progress and we should delay the
+	 * reset.  This could fail if the GPU got in an infinite loop
+	 * in the CL, but that is pretty unlikely outside of an i-g-t
+	 * testcase.
+	 */
+	if (job->timedout_ctca != ctca || job->timedout_ctra != ctra) {
+		job->timedout_ctca = ctca;
+		job->timedout_ctra = ctra;
+
+		schedule_delayed_work(&job->base.work_tdr,
+				      job->base.sched->timeout);
+		return;
+	}
+
+	mutex_lock(&v3d->reset_lock);
+
+	/* block scheduler */
+	for (q = 0; q < V3D_MAX_QUEUES; q++) {
+		struct drm_gpu_scheduler *sched = &v3d->queue[q].sched;
+
+		kthread_park(sched->thread);
+		drm_sched_hw_job_reset(sched, (sched_job->sched == sched ?
+					       sched_job : NULL));
+	}
+
+	/* get the GPU back into the init state */
+	v3d_reset(v3d);
+
+	/* Unblock schedulers and restart their jobs. */
+	for (q = 0; q < V3D_MAX_QUEUES; q++) {
+		drm_sched_job_recovery(&v3d->queue[q].sched);
+		kthread_unpark(v3d->queue[q].sched.thread);
+	}
+
+	mutex_unlock(&v3d->reset_lock);
+}
+
+static const struct drm_sched_backend_ops v3d_sched_ops = {
+	.dependency = v3d_job_dependency,
+	.run_job = v3d_job_run,
+	.timedout_job = v3d_job_timedout,
+	.free_job = v3d_job_free
+};
+
+int
+v3d_sched_init(struct v3d_dev *v3d)
+{
+	int hw_jobs_limit = 1;
+	int job_hang_limit = 0;
+	int hang_limit_ms = 500;
+	int ret;
+
+	ret = drm_sched_init(&v3d->queue[V3D_BIN].sched,
+			     &v3d_sched_ops,
+			     hw_jobs_limit, job_hang_limit,
+			     msecs_to_jiffies(hang_limit_ms),
+			     "v3d_bin");
+	if (ret) {
+		dev_err(v3d->dev, "Failed to create bin scheduler: %d.", ret);
+		return ret;
+	}
+
+	ret = drm_sched_init(&v3d->queue[V3D_RENDER].sched,
+			     &v3d_sched_ops,
+			     hw_jobs_limit, job_hang_limit,
+			     msecs_to_jiffies(hang_limit_ms),
+			     "v3d_render");
+	if (ret) {
+		dev_err(v3d->dev, "Failed to create render scheduler: %d.",
+			ret);
+		drm_sched_fini(&v3d->queue[V3D_BIN].sched);
+		return ret;
+	}
+
+	return 0;
+}
+
+void
+v3d_sched_fini(struct v3d_dev *v3d)
+{
+	enum v3d_queue q;
+
+	for (q = 0; q < V3D_MAX_QUEUES; q++)
+		drm_sched_fini(&v3d->queue[q].sched);
+}
diff --git a/drivers/gpu/drm/v3d/v3d_trace.h b/drivers/gpu/drm/v3d/v3d_trace.h
new file mode 100644
index 000000000000..85dd351e1e09
--- /dev/null
+++ b/drivers/gpu/drm/v3d/v3d_trace.h
@@ -0,0 +1,82 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* Copyright (C) 2015-2018 Broadcom */
+
+#if !defined(_V3D_TRACE_H_) || defined(TRACE_HEADER_MULTI_READ)
+#define _V3D_TRACE_H_
+
+#include <linux/stringify.h>
+#include <linux/types.h>
+#include <linux/tracepoint.h>
+
+#undef TRACE_SYSTEM
+#define TRACE_SYSTEM v3d
+#define TRACE_INCLUDE_FILE v3d_trace
+
+TRACE_EVENT(v3d_submit_cl,
+	    TP_PROTO(struct drm_device *dev, bool is_render,
+		     uint64_t seqno,
+		     u32 ctnqba, u32 ctnqea),
+	    TP_ARGS(dev, is_render, seqno, ctnqba, ctnqea),
+
+	    TP_STRUCT__entry(
+			     __field(u32, dev)
+			     __field(bool, is_render)
+			     __field(u64, seqno)
+			     __field(u32, ctnqba)
+			     __field(u32, ctnqea)
+			     ),
+
+	    TP_fast_assign(
+			   __entry->dev = dev->primary->index;
+			   __entry->is_render = is_render;
+			   __entry->seqno = seqno;
+			   __entry->ctnqba = ctnqba;
+			   __entry->ctnqea = ctnqea;
+			   ),
+
+	    TP_printk("dev=%u, %s, seqno=%llu, 0x%08x..0x%08x",
+		      __entry->dev,
+		      __entry->is_render ? "RCL" : "BCL",
+		      __entry->seqno,
+		      __entry->ctnqba,
+		      __entry->ctnqea)
+);
+
+TRACE_EVENT(v3d_reset_begin,
+	    TP_PROTO(struct drm_device *dev),
+	    TP_ARGS(dev),
+
+	    TP_STRUCT__entry(
+			     __field(u32, dev)
+			     ),
+
+	    TP_fast_assign(
+			   __entry->dev = dev->primary->index;
+			   ),
+
+	    TP_printk("dev=%u",
+		      __entry->dev)
+);
+
+TRACE_EVENT(v3d_reset_end,
+	    TP_PROTO(struct drm_device *dev),
+	    TP_ARGS(dev),
+
+	    TP_STRUCT__entry(
+			     __field(u32, dev)
+			     ),
+
+	    TP_fast_assign(
+			   __entry->dev = dev->primary->index;
+			   ),
+
+	    TP_printk("dev=%u",
+		      __entry->dev)
+);
+
+#endif /* _V3D_TRACE_H_ */
+
+/* This part must be outside protection */
+#undef TRACE_INCLUDE_PATH
+#define TRACE_INCLUDE_PATH .
+#include <trace/define_trace.h>
diff --git a/drivers/gpu/drm/v3d/v3d_trace_points.c b/drivers/gpu/drm/v3d/v3d_trace_points.c
new file mode 100644
index 000000000000..482922d7c7e1
--- /dev/null
+++ b/drivers/gpu/drm/v3d/v3d_trace_points.c
@@ -0,0 +1,9 @@
+// SPDX-License-Identifier: GPL-2.0+
+/* Copyright (C) 2015 Broadcom */
+
+#include "v3d_drv.h"
+
+#ifndef __CHECKER__
+#define CREATE_TRACE_POINTS
+#include "v3d_trace.h"
+#endif
diff --git a/drivers/gpu/drm/vc4/Makefile b/drivers/gpu/drm/vc4/Makefile
index 4a3a868235f8..b303703bc7f3 100644
--- a/drivers/gpu/drm/vc4/Makefile
+++ b/drivers/gpu/drm/vc4/Makefile
@@ -19,6 +19,7 @@ vc4-y := \
 	vc4_plane.o \
 	vc4_render_cl.o \
 	vc4_trace_points.o \
+	vc4_txp.o \
 	vc4_v3d.o \
 	vc4_validate.o \
 	vc4_validate_shaders.o
diff --git a/drivers/gpu/drm/vc4/vc4_bo.c b/drivers/gpu/drm/vc4/vc4_bo.c
index add9cc97a3b6..8dcce7182bb7 100644
--- a/drivers/gpu/drm/vc4/vc4_bo.c
+++ b/drivers/gpu/drm/vc4/vc4_bo.c
@@ -721,7 +721,7 @@ vc4_prime_export(struct drm_device *dev, struct drm_gem_object *obj, int flags)
 	return dmabuf;
 }
 
-int vc4_fault(struct vm_fault *vmf)
+vm_fault_t vc4_fault(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 	struct drm_gem_object *obj = vma->vm_private_data;
diff --git a/drivers/gpu/drm/vc4/vc4_crtc.c b/drivers/gpu/drm/vc4/vc4_crtc.c
index 83d3b7912fc2..0e6a121858d1 100644
--- a/drivers/gpu/drm/vc4/vc4_crtc.c
+++ b/drivers/gpu/drm/vc4/vc4_crtc.c
@@ -46,6 +46,8 @@ struct vc4_crtc_state {
 	struct drm_crtc_state base;
 	/* Dlist area for this CRTC configuration. */
 	struct drm_mm_node mm;
+	bool feed_txp;
+	bool txp_armed;
 };
 
 static inline struct vc4_crtc_state *
@@ -324,10 +326,8 @@ static struct drm_encoder *vc4_get_crtc_encoder(struct drm_crtc *crtc)
 	return NULL;
 }
 
-static void vc4_crtc_mode_set_nofb(struct drm_crtc *crtc)
+static void vc4_crtc_config_pv(struct drm_crtc *crtc)
 {
-	struct drm_device *dev = crtc->dev;
-	struct vc4_dev *vc4 = to_vc4_dev(dev);
 	struct drm_encoder *encoder = vc4_get_crtc_encoder(crtc);
 	struct vc4_encoder *vc4_encoder = to_vc4_encoder(encoder);
 	struct vc4_crtc *vc4_crtc = to_vc4_crtc(crtc);
@@ -338,12 +338,6 @@ static void vc4_crtc_mode_set_nofb(struct drm_crtc *crtc)
 	bool is_dsi = (vc4_encoder->type == VC4_ENCODER_TYPE_DSI0 ||
 		       vc4_encoder->type == VC4_ENCODER_TYPE_DSI1);
 	u32 format = is_dsi ? PV_CONTROL_FORMAT_DSIV_24 : PV_CONTROL_FORMAT_24;
-	bool debug_dump_regs = false;
-
-	if (debug_dump_regs) {
-		DRM_INFO("CRTC %d regs before:\n", drm_crtc_index(crtc));
-		vc4_crtc_dump_regs(vc4_crtc);
-	}
 
 	/* Reset the PV fifo. */
 	CRTC_WRITE(PV_CONTROL, 0);
@@ -419,6 +413,49 @@ static void vc4_crtc_mode_set_nofb(struct drm_crtc *crtc)
 				 PV_CONTROL_CLK_SELECT) |
 		   PV_CONTROL_FIFO_CLR |
 		   PV_CONTROL_EN);
+}
+
+static void vc4_crtc_mode_set_nofb(struct drm_crtc *crtc)
+{
+	struct drm_device *dev = crtc->dev;
+	struct vc4_dev *vc4 = to_vc4_dev(dev);
+	struct vc4_crtc *vc4_crtc = to_vc4_crtc(crtc);
+	struct vc4_crtc_state *vc4_state = to_vc4_crtc_state(crtc->state);
+	struct drm_display_mode *mode = &crtc->state->adjusted_mode;
+	bool interlace = mode->flags & DRM_MODE_FLAG_INTERLACE;
+	bool debug_dump_regs = false;
+
+	if (debug_dump_regs) {
+		DRM_INFO("CRTC %d regs before:\n", drm_crtc_index(crtc));
+		vc4_crtc_dump_regs(vc4_crtc);
+	}
+
+	if (vc4_crtc->channel == 2) {
+		u32 dispctrl;
+		u32 dsp3_mux;
+
+		/*
+		 * SCALER_DISPCTRL_DSP3 = X, where X < 2 means 'connect DSP3 to
+		 * FIFO X'.
+		 * SCALER_DISPCTRL_DSP3 = 3 means 'disable DSP 3'.
+		 *
+		 * DSP3 is connected to FIFO2 unless the transposer is
+		 * enabled. In this case, FIFO 2 is directly accessed by the
+		 * TXP IP, and we need to disable the FIFO2 -> pixelvalve1
+		 * route.
+		 */
+		if (vc4_state->feed_txp)
+			dsp3_mux = VC4_SET_FIELD(3, SCALER_DISPCTRL_DSP3_MUX);
+		else
+			dsp3_mux = VC4_SET_FIELD(2, SCALER_DISPCTRL_DSP3_MUX);
+
+		dispctrl = HVS_READ(SCALER_DISPCTRL) &
+			   ~SCALER_DISPCTRL_DSP3_MUX_MASK;
+		HVS_WRITE(SCALER_DISPCTRL, dispctrl | dsp3_mux);
+	}
+
+	if (!vc4_state->feed_txp)
+		vc4_crtc_config_pv(crtc);
 
 	HVS_WRITE(SCALER_DISPBKGNDX(vc4_crtc->channel),
 		  SCALER_DISPBKGND_AUTOHS |
@@ -499,6 +536,13 @@ static void vc4_crtc_atomic_disable(struct drm_crtc *crtc,
 	}
 }
 
+void vc4_crtc_txp_armed(struct drm_crtc_state *state)
+{
+	struct vc4_crtc_state *vc4_state = to_vc4_crtc_state(state);
+
+	vc4_state->txp_armed = true;
+}
+
 static void vc4_crtc_update_dlist(struct drm_crtc *crtc)
 {
 	struct drm_device *dev = crtc->dev;
@@ -514,8 +558,11 @@ static void vc4_crtc_update_dlist(struct drm_crtc *crtc)
 		WARN_ON(drm_crtc_vblank_get(crtc) != 0);
 
 		spin_lock_irqsave(&dev->event_lock, flags);
-		vc4_crtc->event = crtc->state->event;
-		crtc->state->event = NULL;
+
+		if (!vc4_state->feed_txp || vc4_state->txp_armed) {
+			vc4_crtc->event = crtc->state->event;
+			crtc->state->event = NULL;
+		}
 
 		HVS_WRITE(SCALER_DISPLISTX(vc4_crtc->channel),
 			  vc4_state->mm.start);
@@ -533,8 +580,8 @@ static void vc4_crtc_atomic_enable(struct drm_crtc *crtc,
 	struct drm_device *dev = crtc->dev;
 	struct vc4_dev *vc4 = to_vc4_dev(dev);
 	struct vc4_crtc *vc4_crtc = to_vc4_crtc(crtc);
-	struct drm_crtc_state *state = crtc->state;
-	struct drm_display_mode *mode = &state->adjusted_mode;
+	struct vc4_crtc_state *vc4_state = to_vc4_crtc_state(crtc->state);
+	struct drm_display_mode *mode = &crtc->state->adjusted_mode;
 
 	require_hvs_enabled(dev);
 
@@ -546,15 +593,21 @@ static void vc4_crtc_atomic_enable(struct drm_crtc *crtc,
 
 	/* Turn on the scaler, which will wait for vstart to start
 	 * compositing.
+	 * When feeding the transposer, we should operate in oneshot
+	 * mode.
 	 */
 	HVS_WRITE(SCALER_DISPCTRLX(vc4_crtc->channel),
 		  VC4_SET_FIELD(mode->hdisplay, SCALER_DISPCTRLX_WIDTH) |
 		  VC4_SET_FIELD(mode->vdisplay, SCALER_DISPCTRLX_HEIGHT) |
-		  SCALER_DISPCTRLX_ENABLE);
+		  SCALER_DISPCTRLX_ENABLE |
+		  (vc4_state->feed_txp ? SCALER_DISPCTRLX_ONESHOT : 0));
 
-	/* Turn on the pixel valve, which will emit the vstart signal. */
-	CRTC_WRITE(PV_V_CONTROL,
-		   CRTC_READ(PV_V_CONTROL) | PV_VCONTROL_VIDEN);
+	/* When feeding the transposer block the pixelvalve is unneeded and
+	 * should not be enabled.
+	 */
+	if (!vc4_state->feed_txp)
+		CRTC_WRITE(PV_V_CONTROL,
+			   CRTC_READ(PV_V_CONTROL) | PV_VCONTROL_VIDEN);
 }
 
 static enum drm_mode_status vc4_crtc_mode_valid(struct drm_crtc *crtc,
@@ -579,8 +632,10 @@ static int vc4_crtc_atomic_check(struct drm_crtc *crtc,
 	struct drm_plane *plane;
 	unsigned long flags;
 	const struct drm_plane_state *plane_state;
+	struct drm_connector *conn;
+	struct drm_connector_state *conn_state;
 	u32 dlist_count = 0;
-	int ret;
+	int ret, i;
 
 	/* The pixelvalve can only feed one encoder (and encoders are
 	 * 1:1 with connectors.)
@@ -600,6 +655,24 @@ static int vc4_crtc_atomic_check(struct drm_crtc *crtc,
 	if (ret)
 		return ret;
 
+	for_each_new_connector_in_state(state->state, conn, conn_state, i) {
+		if (conn_state->crtc != crtc)
+			continue;
+
+		/* The writeback connector is implemented using the transposer
+		 * block which is directly taking its data from the HVS FIFO.
+		 */
+		if (conn->connector_type == DRM_MODE_CONNECTOR_WRITEBACK) {
+			state->no_vblank = true;
+			vc4_state->feed_txp = true;
+		} else {
+			state->no_vblank = false;
+			vc4_state->feed_txp = false;
+		}
+
+		break;
+	}
+
 	return 0;
 }
 
@@ -713,7 +786,8 @@ static void vc4_crtc_handle_page_flip(struct vc4_crtc *vc4_crtc)
 
 	spin_lock_irqsave(&dev->event_lock, flags);
 	if (vc4_crtc->event &&
-	    (vc4_state->mm.start == HVS_READ(SCALER_DISPLACTX(chan)))) {
+	    (vc4_state->mm.start == HVS_READ(SCALER_DISPLACTX(chan)) ||
+	     vc4_state->feed_txp)) {
 		drm_crtc_send_vblank_event(crtc, vc4_crtc->event);
 		vc4_crtc->event = NULL;
 		drm_crtc_vblank_put(crtc);
@@ -721,6 +795,13 @@ static void vc4_crtc_handle_page_flip(struct vc4_crtc *vc4_crtc)
 	spin_unlock_irqrestore(&dev->event_lock, flags);
 }
 
+void vc4_crtc_handle_vblank(struct vc4_crtc *crtc)
+{
+	crtc->t_vblank = ktime_get();
+	drm_crtc_handle_vblank(&crtc->base);
+	vc4_crtc_handle_page_flip(crtc);
+}
+
 static irqreturn_t vc4_crtc_irq_handler(int irq, void *data)
 {
 	struct vc4_crtc *vc4_crtc = data;
@@ -728,10 +809,8 @@ static irqreturn_t vc4_crtc_irq_handler(int irq, void *data)
 	irqreturn_t ret = IRQ_NONE;
 
 	if (stat & PV_INT_VFP_START) {
-		vc4_crtc->t_vblank = ktime_get();
 		CRTC_WRITE(PV_INTSTAT, PV_INT_VFP_START);
-		drm_crtc_handle_vblank(&vc4_crtc->base);
-		vc4_crtc_handle_page_flip(vc4_crtc);
+		vc4_crtc_handle_vblank(vc4_crtc);
 		ret = IRQ_HANDLED;
 	}
 
@@ -741,6 +820,7 @@ static irqreturn_t vc4_crtc_irq_handler(int irq, void *data)
 struct vc4_async_flip_state {
 	struct drm_crtc *crtc;
 	struct drm_framebuffer *fb;
+	struct drm_framebuffer *old_fb;
 	struct drm_pending_vblank_event *event;
 
 	struct vc4_seqno_cb cb;
@@ -770,6 +850,23 @@ vc4_async_page_flip_complete(struct vc4_seqno_cb *cb)
 
 	drm_crtc_vblank_put(crtc);
 	drm_framebuffer_put(flip_state->fb);
+
+	/* Decrement the BO usecnt in order to keep the inc/dec calls balanced
+	 * when the planes are updated through the async update path.
+	 * FIXME: we should move to generic async-page-flip when it's
+	 * available, so that we can get rid of this hand-made cleanup_fb()
+	 * logic.
+	 */
+	if (flip_state->old_fb) {
+		struct drm_gem_cma_object *cma_bo;
+		struct vc4_bo *bo;
+
+		cma_bo = drm_fb_cma_get_gem_obj(flip_state->old_fb, 0);
+		bo = to_vc4_bo(&cma_bo->base);
+		vc4_bo_dec_usecnt(bo);
+		drm_framebuffer_put(flip_state->old_fb);
+	}
+
 	kfree(flip_state);
 
 	up(&vc4->async_modeset);
@@ -794,9 +891,22 @@ static int vc4_async_page_flip(struct drm_crtc *crtc,
 	struct drm_gem_cma_object *cma_bo = drm_fb_cma_get_gem_obj(fb, 0);
 	struct vc4_bo *bo = to_vc4_bo(&cma_bo->base);
 
+	/* Increment the BO usecnt here, so that we never end up with an
+	 * unbalanced number of vc4_bo_{dec,inc}_usecnt() calls when the
+	 * plane is later updated through the non-async path.
+	 * FIXME: we should move to generic async-page-flip when it's
+	 * available, so that we can get rid of this hand-made prepare_fb()
+	 * logic.
+	 */
+	ret = vc4_bo_inc_usecnt(bo);
+	if (ret)
+		return ret;
+
 	flip_state = kzalloc(sizeof(*flip_state), GFP_KERNEL);
-	if (!flip_state)
+	if (!flip_state) {
+		vc4_bo_dec_usecnt(bo);
 		return -ENOMEM;
+	}
 
 	drm_framebuffer_get(fb);
 	flip_state->fb = fb;
@@ -807,10 +917,23 @@ static int vc4_async_page_flip(struct drm_crtc *crtc,
 	ret = down_interruptible(&vc4->async_modeset);
 	if (ret) {
 		drm_framebuffer_put(fb);
+		vc4_bo_dec_usecnt(bo);
 		kfree(flip_state);
 		return ret;
 	}
 
+	/* Save the current FB before it's replaced by the new one in
+	 * drm_atomic_set_fb_for_plane(). We'll need the old FB in
+	 * vc4_async_page_flip_complete() to decrement the BO usecnt and keep
+	 * it consistent.
+	 * FIXME: we should move to generic async-page-flip when it's
+	 * available, so that we can get rid of this hand-made cleanup_fb()
+	 * logic.
+	 */
+	flip_state->old_fb = plane->state->fb;
+	if (flip_state->old_fb)
+		drm_framebuffer_get(flip_state->old_fb);
+
 	WARN_ON(drm_crtc_vblank_get(crtc) != 0);
 
 	/* Immediately update the plane's legacy fb pointer, so that later
@@ -818,7 +941,6 @@ static int vc4_async_page_flip(struct drm_crtc *crtc,
 	 * is released.
 	 */
 	drm_atomic_set_fb_for_plane(plane->state, fb);
-	plane->fb = fb;
 
 	vc4_queue_seqno_cb(dev, &flip_state->cb, bo->seqno,
 			   vc4_async_page_flip_complete);
@@ -841,12 +963,15 @@ static int vc4_page_flip(struct drm_crtc *crtc,
 
 static struct drm_crtc_state *vc4_crtc_duplicate_state(struct drm_crtc *crtc)
 {
-	struct vc4_crtc_state *vc4_state;
+	struct vc4_crtc_state *vc4_state, *old_vc4_state;
 
 	vc4_state = kzalloc(sizeof(*vc4_state), GFP_KERNEL);
 	if (!vc4_state)
 		return NULL;
 
+	old_vc4_state = to_vc4_crtc_state(crtc->state);
+	vc4_state->feed_txp = old_vc4_state->feed_txp;
+
 	__drm_atomic_helper_crtc_duplicate_state(crtc, &vc4_state->base);
 	return &vc4_state->base;
 }
@@ -944,9 +1069,17 @@ static void vc4_set_crtc_possible_masks(struct drm_device *drm,
 	struct drm_encoder *encoder;
 
 	drm_for_each_encoder(encoder, drm) {
-		struct vc4_encoder *vc4_encoder = to_vc4_encoder(encoder);
+		struct vc4_encoder *vc4_encoder;
 		int i;
 
+		/* HVS FIFO2 can feed the TXP IP. */
+		if (crtc_data->hvs_channel == 2 &&
+		    encoder->encoder_type == DRM_MODE_ENCODER_VIRTUAL) {
+			encoder->possible_crtcs |= drm_crtc_mask(crtc);
+			continue;
+		}
+
+		vc4_encoder = to_vc4_encoder(encoder);
 		for (i = 0; i < ARRAY_SIZE(crtc_data->encoder_types); i++) {
 			if (vc4_encoder->type == encoder_types[i]) {
 				vc4_encoder->clock_select = i;
@@ -1013,7 +1146,6 @@ static int vc4_crtc_bind(struct device *dev, struct device *master, void *data)
 	drm_crtc_init_with_planes(drm, crtc, primary_plane, NULL,
 				  &vc4_crtc_funcs, NULL);
 	drm_crtc_helper_add(crtc, &vc4_crtc_helper_funcs);
-	primary_plane->crtc = crtc;
 	vc4_crtc->channel = vc4_crtc->data->hvs_channel;
 	drm_mode_crtc_set_gamma_size(crtc, ARRAY_SIZE(vc4_crtc->lut_r));
 	drm_crtc_enable_color_mgmt(crtc, 0, false, crtc->gamma_size);
@@ -1039,7 +1171,7 @@ static int vc4_crtc_bind(struct device *dev, struct device *master, void *data)
 		if (IS_ERR(plane))
 			continue;
 
-		plane->possible_crtcs = 1 << drm_crtc_index(crtc);
+		plane->possible_crtcs = drm_crtc_mask(crtc);
 	}
 
 	/* Set up the legacy cursor after overlay initialization,
@@ -1048,8 +1180,7 @@ static int vc4_crtc_bind(struct device *dev, struct device *master, void *data)
 	 */
 	cursor_plane = vc4_plane_init(drm, DRM_PLANE_TYPE_CURSOR);
 	if (!IS_ERR(cursor_plane)) {
-		cursor_plane->possible_crtcs = 1 << drm_crtc_index(crtc);
-		cursor_plane->crtc = crtc;
+		cursor_plane->possible_crtcs = drm_crtc_mask(crtc);
 		crtc->cursor = cursor_plane;
 	}
 
@@ -1077,7 +1208,7 @@ static int vc4_crtc_bind(struct device *dev, struct device *master, void *data)
 err_destroy_planes:
 	list_for_each_entry_safe(destroy_plane, temp,
 				 &drm->mode_config.plane_list, head) {
-		if (destroy_plane->possible_crtcs == 1 << drm_crtc_index(crtc))
+		if (destroy_plane->possible_crtcs == drm_crtc_mask(crtc))
 		    destroy_plane->funcs->destroy(destroy_plane);
 	}
 err:
diff --git a/drivers/gpu/drm/vc4/vc4_debugfs.c b/drivers/gpu/drm/vc4/vc4_debugfs.c
index 5db06bdb5f27..7a0003de71ab 100644
--- a/drivers/gpu/drm/vc4/vc4_debugfs.c
+++ b/drivers/gpu/drm/vc4/vc4_debugfs.c
@@ -21,6 +21,7 @@ static const struct drm_info_list vc4_debugfs_list[] = {
 	{"dsi1_regs", vc4_dsi_debugfs_regs, 0, (void *)(uintptr_t)1},
 	{"hdmi_regs", vc4_hdmi_debugfs_regs, 0},
 	{"vec_regs", vc4_vec_debugfs_regs, 0},
+	{"txp_regs", vc4_txp_debugfs_regs, 0},
 	{"hvs_regs", vc4_hvs_debugfs_regs, 0},
 	{"crtc0_regs", vc4_crtc_debugfs_regs, 0, (void *)(uintptr_t)0},
 	{"crtc1_regs", vc4_crtc_debugfs_regs, 0, (void *)(uintptr_t)1},
diff --git a/drivers/gpu/drm/vc4/vc4_dpi.c b/drivers/gpu/drm/vc4/vc4_dpi.c
index 72c9dbd81d7f..f185812970da 100644
--- a/drivers/gpu/drm/vc4/vc4_dpi.c
+++ b/drivers/gpu/drm/vc4/vc4_dpi.c
@@ -96,7 +96,6 @@ struct vc4_dpi {
 	struct platform_device *pdev;
 
 	struct drm_encoder *encoder;
-	struct drm_connector *connector;
 
 	void __iomem *regs;
 
@@ -164,14 +163,31 @@ static void vc4_dpi_encoder_disable(struct drm_encoder *encoder)
 
 static void vc4_dpi_encoder_enable(struct drm_encoder *encoder)
 {
+	struct drm_device *dev = encoder->dev;
 	struct drm_display_mode *mode = &encoder->crtc->mode;
 	struct vc4_dpi_encoder *vc4_encoder = to_vc4_dpi_encoder(encoder);
 	struct vc4_dpi *dpi = vc4_encoder->dpi;
+	struct drm_connector_list_iter conn_iter;
+	struct drm_connector *connector = NULL, *connector_scan;
 	u32 dpi_c = DPI_ENABLE | DPI_OUTPUT_ENABLE_MODE;
 	int ret;
 
-	if (dpi->connector->display_info.num_bus_formats) {
-		u32 bus_format = dpi->connector->display_info.bus_formats[0];
+	/* Look up the connector attached to DPI so we can get the
+	 * bus_format.  Ideally the bridge would tell us the
+	 * bus_format we want, but it doesn't yet, so assume that it's
+	 * uniform throughout the bridge chain.
+	 */
+	drm_connector_list_iter_begin(dev, &conn_iter);
+	drm_for_each_connector_iter(connector_scan, &conn_iter) {
+		if (connector_scan->encoder == encoder) {
+			connector = connector_scan;
+			break;
+		}
+	}
+	drm_connector_list_iter_end(&conn_iter);
+
+	if (connector && connector->display_info.num_bus_formats) {
+		u32 bus_format = connector->display_info.bus_formats[0];
 
 		switch (bus_format) {
 		case MEDIA_BUS_FMT_RGB888_1X24:
@@ -199,6 +215,9 @@ static void vc4_dpi_encoder_enable(struct drm_encoder *encoder)
 			DRM_ERROR("Unknown media bus format %d\n", bus_format);
 			break;
 		}
+	} else {
+		/* Default to 24bit if no connector found. */
+		dpi_c |= VC4_SET_FIELD(DPI_FORMAT_24BIT_888_RGB, DPI_FORMAT);
 	}
 
 	if (mode->flags & DRM_MODE_FLAG_NHSYNC)
diff --git a/drivers/gpu/drm/vc4/vc4_drv.c b/drivers/gpu/drm/vc4/vc4_drv.c
index 40ddeaafd65f..04270a14fcaa 100644
--- a/drivers/gpu/drm/vc4/vc4_drv.c
+++ b/drivers/gpu/drm/vc4/vc4_drv.c
@@ -130,6 +130,7 @@ static void vc4_close(struct drm_device *dev, struct drm_file *file)
 	struct vc4_file *vc4file = file->driver_priv;
 
 	vc4_perfmon_close_file(vc4file);
+	kfree(vc4file);
 }
 
 static const struct vm_operations_struct vc4_vm_ops = {
@@ -175,7 +176,8 @@ static struct drm_driver vc4_drm_driver = {
 			    DRIVER_GEM |
 			    DRIVER_HAVE_IRQ |
 			    DRIVER_RENDER |
-			    DRIVER_PRIME),
+			    DRIVER_PRIME |
+			    DRIVER_SYNCOBJ),
 	.lastclose = drm_fb_helper_lastclose,
 	.open = vc4_open,
 	.postclose = vc4_close,
@@ -286,7 +288,7 @@ static int vc4_drm_bind(struct device *dev)
 
 	ret = vc4_bo_cache_init(drm);
 	if (ret)
-		goto dev_unref;
+		goto dev_put;
 
 	drm_mode_config_init(drm);
 
@@ -311,8 +313,8 @@ unbind_all:
 gem_destroy:
 	vc4_gem_destroy(drm);
 	vc4_bo_cache_destroy(drm);
-dev_unref:
-	drm_dev_unref(drm);
+dev_put:
+	drm_dev_put(drm);
 	return ret;
 }
 
@@ -329,7 +331,7 @@ static void vc4_drm_unbind(struct device *dev)
 
 	drm_atomic_private_obj_fini(&vc4->ctm_manager);
 
-	drm_dev_unref(drm);
+	drm_dev_put(drm);
 }
 
 static const struct component_master_ops vc4_drm_ops = {
@@ -342,6 +344,7 @@ static struct platform_driver *const component_drivers[] = {
 	&vc4_vec_driver,
 	&vc4_dpi_driver,
 	&vc4_dsi_driver,
+	&vc4_txp_driver,
 	&vc4_hvs_driver,
 	&vc4_crtc_driver,
 	&vc4_v3d_driver,
diff --git a/drivers/gpu/drm/vc4/vc4_drv.h b/drivers/gpu/drm/vc4/vc4_drv.h
index 22589d39083c..bd6ef1f31822 100644
--- a/drivers/gpu/drm/vc4/vc4_drv.h
+++ b/drivers/gpu/drm/vc4/vc4_drv.h
@@ -6,11 +6,13 @@
  * published by the Free Software Foundation.
  */
 
+#include <linux/mm_types.h>
 #include <linux/reservation.h>
 #include <drm/drmP.h>
 #include <drm/drm_encoder.h>
 #include <drm/drm_gem_cma_helper.h>
 #include <drm/drm_atomic.h>
+#include <drm/drm_syncobj.h>
 
 #include "uapi/drm/vc4_drm.h"
 
@@ -71,6 +73,7 @@ struct vc4_dev {
 	struct vc4_dpi *dpi;
 	struct vc4_dsi *dsi1;
 	struct vc4_vec *vec;
+	struct vc4_txp *txp;
 
 	struct vc4_hang_state *hang_state;
 
@@ -673,7 +676,7 @@ int vc4_get_hang_state_ioctl(struct drm_device *dev, void *data,
 			     struct drm_file *file_priv);
 int vc4_label_bo_ioctl(struct drm_device *dev, void *data,
 		       struct drm_file *file_priv);
-int vc4_fault(struct vm_fault *vmf);
+vm_fault_t vc4_fault(struct vm_fault *vmf);
 int vc4_mmap(struct file *filp, struct vm_area_struct *vma);
 struct reservation_object *vc4_prime_res_obj(struct drm_gem_object *obj);
 int vc4_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma);
@@ -696,6 +699,8 @@ bool vc4_crtc_get_scanoutpos(struct drm_device *dev, unsigned int crtc_id,
 			     bool in_vblank_irq, int *vpos, int *hpos,
 			     ktime_t *stime, ktime_t *etime,
 			     const struct drm_display_mode *mode);
+void vc4_crtc_handle_vblank(struct vc4_crtc *crtc);
+void vc4_crtc_txp_armed(struct drm_crtc_state *state);
 
 /* vc4_debugfs.c */
 int vc4_debugfs_init(struct drm_minor *minor);
@@ -743,6 +748,10 @@ int vc4_hdmi_debugfs_regs(struct seq_file *m, void *unused);
 extern struct platform_driver vc4_vec_driver;
 int vc4_vec_debugfs_regs(struct seq_file *m, void *unused);
 
+/* vc4_txp.c */
+extern struct platform_driver vc4_txp_driver;
+int vc4_txp_debugfs_regs(struct seq_file *m, void *unused);
+
 /* vc4_irq.c */
 irqreturn_t vc4_irq(int irq, void *arg);
 void vc4_irq_preinstall(struct drm_device *dev);
diff --git a/drivers/gpu/drm/vc4/vc4_dsi.c b/drivers/gpu/drm/vc4/vc4_dsi.c
index 94085f8bcd68..0c607eb33d7e 100644
--- a/drivers/gpu/drm/vc4/vc4_dsi.c
+++ b/drivers/gpu/drm/vc4/vc4_dsi.c
@@ -753,6 +753,11 @@ static void vc4_dsi_ulps(struct vc4_dsi *dsi, bool ulps)
 			 (dsi->lanes > 2 ? DSI1_STAT_PHY_D2_STOP : 0) |
 			 (dsi->lanes > 3 ? DSI1_STAT_PHY_D3_STOP : 0));
 	int ret;
+	bool ulps_currently_enabled = (DSI_PORT_READ(PHY_AFEC0) &
+				       DSI_PORT_BIT(PHY_AFEC0_LATCH_ULPS));
+
+	if (ulps == ulps_currently_enabled)
+		return;
 
 	DSI_PORT_WRITE(STAT, stat_ulps);
 	DSI_PORT_WRITE(PHYC, DSI_PORT_READ(PHYC) | phyc_ulps);
@@ -809,7 +814,9 @@ static void vc4_dsi_encoder_disable(struct drm_encoder *encoder)
 	struct vc4_dsi *dsi = vc4_encoder->dsi;
 	struct device *dev = &dsi->pdev->dev;
 
+	drm_bridge_disable(dsi->bridge);
 	vc4_dsi_ulps(dsi, true);
+	drm_bridge_post_disable(dsi->bridge);
 
 	clk_disable_unprepare(dsi->pll_phy_clock);
 	clk_disable_unprepare(dsi->escape_clock);
@@ -1084,21 +1091,6 @@ static void vc4_dsi_encoder_enable(struct drm_encoder *encoder)
 	/* Display reset sequence timeout */
 	DSI_PORT_WRITE(PR_TO_CNT, 100000);
 
-	if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO) {
-		DSI_PORT_WRITE(DISP0_CTRL,
-			       VC4_SET_FIELD(dsi->divider,
-					     DSI_DISP0_PIX_CLK_DIV) |
-			       VC4_SET_FIELD(dsi->format, DSI_DISP0_PFORMAT) |
-			       VC4_SET_FIELD(DSI_DISP0_LP_STOP_PERFRAME,
-					     DSI_DISP0_LP_STOP_CTRL) |
-			       DSI_DISP0_ST_END |
-			       DSI_DISP0_ENABLE);
-	} else {
-		DSI_PORT_WRITE(DISP0_CTRL,
-			       DSI_DISP0_COMMAND_MODE |
-			       DSI_DISP0_ENABLE);
-	}
-
 	/* Set up DISP1 for transferring long command payloads through
 	 * the pixfifo.
 	 */
@@ -1123,6 +1115,25 @@ static void vc4_dsi_encoder_enable(struct drm_encoder *encoder)
 
 	vc4_dsi_ulps(dsi, false);
 
+	drm_bridge_pre_enable(dsi->bridge);
+
+	if (dsi->mode_flags & MIPI_DSI_MODE_VIDEO) {
+		DSI_PORT_WRITE(DISP0_CTRL,
+			       VC4_SET_FIELD(dsi->divider,
+					     DSI_DISP0_PIX_CLK_DIV) |
+			       VC4_SET_FIELD(dsi->format, DSI_DISP0_PFORMAT) |
+			       VC4_SET_FIELD(DSI_DISP0_LP_STOP_PERFRAME,
+					     DSI_DISP0_LP_STOP_CTRL) |
+			       DSI_DISP0_ST_END |
+			       DSI_DISP0_ENABLE);
+	} else {
+		DSI_PORT_WRITE(DISP0_CTRL,
+			       DSI_DISP0_COMMAND_MODE |
+			       DSI_DISP0_ENABLE);
+	}
+
+	drm_bridge_enable(dsi->bridge);
+
 	if (debug_dump_regs) {
 		DRM_INFO("DSI regs after:\n");
 		vc4_dsi_dump_regs(dsi);
@@ -1601,8 +1612,18 @@ static int vc4_dsi_bind(struct device *dev, struct device *master, void *data)
 
 	ret = drm_of_find_panel_or_bridge(dev->of_node, 0, 0,
 					  &panel, &dsi->bridge);
-	if (ret)
+	if (ret) {
+		/* If the bridge or panel pointed by dev->of_node is not
+		 * enabled, just return 0 here so that we don't prevent the DRM
+		 * dev from being registered. Of course that means the DSI
+		 * encoder won't be exposed, but that's not a problem since
+		 * nothing is connected to it.
+		 */
+		if (ret == -ENODEV)
+			return 0;
+
 		return ret;
+	}
 
 	if (panel) {
 		dsi->bridge = devm_drm_panel_bridge_add(dev, panel,
@@ -1634,6 +1655,12 @@ static int vc4_dsi_bind(struct device *dev, struct device *master, void *data)
 		dev_err(dev, "bridge attach failed: %d\n", ret);
 		return ret;
 	}
+	/* Disable the atomic helper calls into the bridge.  We
+	 * manually call the bridge pre_enable / enable / etc. calls
+	 * from our driver, since we need to sequence them within the
+	 * encoder's enable/disable paths.
+	 */
+	dsi->encoder->bridge = NULL;
 
 	pm_runtime_enable(dev);
 
@@ -1647,7 +1674,8 @@ static void vc4_dsi_unbind(struct device *dev, struct device *master,
 	struct vc4_dev *vc4 = to_vc4_dev(drm);
 	struct vc4_dsi *dsi = dev_get_drvdata(dev);
 
-	pm_runtime_disable(dev);
+	if (dsi->bridge)
+		pm_runtime_disable(dev);
 
 	vc4_dsi_encoder_destroy(dsi->encoder);
 
diff --git a/drivers/gpu/drm/vc4/vc4_fence.c b/drivers/gpu/drm/vc4/vc4_fence.c
index dbf5a5a5d5f5..580214e2158c 100644
--- a/drivers/gpu/drm/vc4/vc4_fence.c
+++ b/drivers/gpu/drm/vc4/vc4_fence.c
@@ -33,11 +33,6 @@ static const char *vc4_fence_get_timeline_name(struct dma_fence *fence)
 	return "vc4-v3d";
 }
 
-static bool vc4_fence_enable_signaling(struct dma_fence *fence)
-{
-	return true;
-}
-
 static bool vc4_fence_signaled(struct dma_fence *fence)
 {
 	struct vc4_fence *f = to_vc4_fence(fence);
@@ -49,8 +44,5 @@ static bool vc4_fence_signaled(struct dma_fence *fence)
 const struct dma_fence_ops vc4_fence_ops = {
 	.get_driver_name = vc4_fence_get_driver_name,
 	.get_timeline_name = vc4_fence_get_timeline_name,
-	.enable_signaling = vc4_fence_enable_signaling,
 	.signaled = vc4_fence_signaled,
-	.wait = dma_fence_default_wait,
-	.release = dma_fence_free,
 };
diff --git a/drivers/gpu/drm/vc4/vc4_gem.c b/drivers/gpu/drm/vc4/vc4_gem.c
index 2107b0daf8ef..7910b9acedd6 100644
--- a/drivers/gpu/drm/vc4/vc4_gem.c
+++ b/drivers/gpu/drm/vc4/vc4_gem.c
@@ -27,6 +27,7 @@
 #include <linux/device.h>
 #include <linux/io.h>
 #include <linux/sched/signal.h>
+#include <linux/dma-fence-array.h>
 
 #include "uapi/drm/vc4_drm.h"
 #include "vc4_drv.h"
@@ -655,7 +656,8 @@ retry:
  */
 static int
 vc4_queue_submit(struct drm_device *dev, struct vc4_exec_info *exec,
-		 struct ww_acquire_ctx *acquire_ctx)
+		 struct ww_acquire_ctx *acquire_ctx,
+		 struct drm_syncobj *out_sync)
 {
 	struct vc4_dev *vc4 = to_vc4_dev(dev);
 	struct vc4_exec_info *renderjob;
@@ -678,6 +680,9 @@ vc4_queue_submit(struct drm_device *dev, struct vc4_exec_info *exec,
 	fence->seqno = exec->seqno;
 	exec->fence = &fence->base;
 
+	if (out_sync)
+		drm_syncobj_replace_fence(out_sync, exec->fence);
+
 	vc4_update_bo_seqnos(exec, seqno);
 
 	vc4_unlock_bo_reservations(dev, exec, acquire_ctx);
@@ -1113,8 +1118,10 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data,
 	struct vc4_dev *vc4 = to_vc4_dev(dev);
 	struct vc4_file *vc4file = file_priv->driver_priv;
 	struct drm_vc4_submit_cl *args = data;
+	struct drm_syncobj *out_sync = NULL;
 	struct vc4_exec_info *exec;
 	struct ww_acquire_ctx acquire_ctx;
+	struct dma_fence *in_fence;
 	int ret = 0;
 
 	if ((args->flags & ~(VC4_SUBMIT_CL_USE_CLEAR_COLOR |
@@ -1126,7 +1133,7 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data,
 	}
 
 	if (args->pad2 != 0) {
-		DRM_DEBUG("->pad2 must be set to zero\n");
+		DRM_DEBUG("Invalid pad: 0x%08x\n", args->pad2);
 		return -EINVAL;
 	}
 
@@ -1164,6 +1171,29 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data,
 		}
 	}
 
+	if (args->in_sync) {
+		ret = drm_syncobj_find_fence(file_priv, args->in_sync,
+					     &in_fence);
+		if (ret)
+			goto fail;
+
+		/* When the fence (or fence array) is exclusively from our
+		 * context we can skip the wait since jobs are executed in
+		 * order of their submission through this ioctl and this can
+		 * only have fences from a prior job.
+		 */
+		if (!dma_fence_match_context(in_fence,
+					     vc4->dma_fence_context)) {
+			ret = dma_fence_wait(in_fence, true);
+			if (ret) {
+				dma_fence_put(in_fence);
+				goto fail;
+			}
+		}
+
+		dma_fence_put(in_fence);
+	}
+
 	if (exec->args->bin_cl_size != 0) {
 		ret = vc4_get_bcl(dev, exec);
 		if (ret)
@@ -1181,12 +1211,33 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data,
 	if (ret)
 		goto fail;
 
+	if (args->out_sync) {
+		out_sync = drm_syncobj_find(file_priv, args->out_sync);
+		if (!out_sync) {
+			ret = -EINVAL;
+			goto fail;
+		}
+
+		/* We replace the fence in out_sync in vc4_queue_submit since
+		 * the render job could execute immediately after that call.
+		 * If it finishes before our ioctl processing resumes the
+		 * render job fence could already have been freed.
+		 */
+	}
+
 	/* Clear this out of the struct we'll be putting in the queue,
 	 * since it's part of our stack.
 	 */
 	exec->args = NULL;
 
-	ret = vc4_queue_submit(dev, exec, &acquire_ctx);
+	ret = vc4_queue_submit(dev, exec, &acquire_ctx, out_sync);
+
+	/* The syncobj isn't part of the exec data and we need to free our
+	 * reference even if job submission failed.
+	 */
+	if (out_sync)
+		drm_syncobj_put(out_sync);
+
 	if (ret)
 		goto fail;
 
diff --git a/drivers/gpu/drm/vc4/vc4_hdmi.c b/drivers/gpu/drm/vc4/vc4_hdmi.c
index b8d50533e2bb..fd5522fd179e 100644
--- a/drivers/gpu/drm/vc4/vc4_hdmi.c
+++ b/drivers/gpu/drm/vc4/vc4_hdmi.c
@@ -285,7 +285,7 @@ static int vc4_hdmi_connector_get_modes(struct drm_connector *connector)
 			drm_rgb_quant_range_selectable(edid);
 	}
 
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 	ret = drm_add_edid_modes(connector, edid);
 	kfree(edid);
 
@@ -329,7 +329,7 @@ static struct drm_connector *vc4_hdmi_connector_init(struct drm_device *dev,
 	connector->interlace_allowed = 1;
 	connector->doublescan_allowed = 0;
 
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	return connector;
 }
diff --git a/drivers/gpu/drm/vc4/vc4_kms.c b/drivers/gpu/drm/vc4/vc4_kms.c
index 8a411e5f8776..ca5aa7fba769 100644
--- a/drivers/gpu/drm/vc4/vc4_kms.c
+++ b/drivers/gpu/drm/vc4/vc4_kms.c
@@ -153,18 +153,11 @@ vc4_atomic_complete_commit(struct drm_atomic_state *state)
 
 	drm_atomic_helper_commit_modeset_enables(dev, state);
 
-	/* Make sure that drm_atomic_helper_wait_for_vblanks()
-	 * actually waits for vblank.  If we're doing a full atomic
-	 * modeset (as opposed to a vc4_update_plane() short circuit),
-	 * then we need to wait for scanout to be done with our
-	 * display lists before we free it and potentially reallocate
-	 * and overwrite the dlist memory with a new modeset.
-	 */
-	state->legacy_cursor_update = false;
+	drm_atomic_helper_fake_vblank(state);
 
 	drm_atomic_helper_commit_hw_done(state);
 
-	drm_atomic_helper_wait_for_vblanks(dev, state);
+	drm_atomic_helper_wait_for_flip_done(dev, state);
 
 	drm_atomic_helper_cleanup_planes(dev, state);
 
diff --git a/drivers/gpu/drm/vc4/vc4_plane.c b/drivers/gpu/drm/vc4/vc4_plane.c
index 3483c05cc3d6..9d7a36f148cf 100644
--- a/drivers/gpu/drm/vc4/vc4_plane.c
+++ b/drivers/gpu/drm/vc4/vc4_plane.c
@@ -209,7 +209,7 @@ static void vc4_dlist_write(struct vc4_plane_state *vc4_state, u32 val)
 {
 	if (vc4_state->dlist_count == vc4_state->dlist_size) {
 		u32 new_size = max(4u, vc4_state->dlist_count * 2);
-		u32 *new_dlist = kmalloc(new_size * 4, GFP_KERNEL);
+		u32 *new_dlist = kmalloc_array(new_size, 4, GFP_KERNEL);
 
 		if (!new_dlist)
 			return;
@@ -467,12 +467,14 @@ static int vc4_plane_mode_set(struct drm_plane *plane,
 	struct drm_framebuffer *fb = state->fb;
 	u32 ctl0_offset = vc4_state->dlist_count;
 	const struct hvs_format *format = vc4_get_hvs_format(fb->format->format);
+	u64 base_format_mod = fourcc_mod_broadcom_mod(fb->modifier);
 	int num_planes = drm_format_num_planes(format->drm);
 	bool mix_plane_alpha;
 	bool covers_screen;
 	u32 scl0, scl1, pitch0;
 	u32 lbm_size, tiling;
 	unsigned long irqflags;
+	u32 hvs_format = format->hvs;
 	int ret, i;
 
 	ret = vc4_plane_setup_clipping_and_scaling(state);
@@ -505,14 +507,14 @@ static int vc4_plane_mode_set(struct drm_plane *plane,
 	 * the scl fields here.
 	 */
 	if (num_planes == 1) {
-		scl0 = vc4_get_scl_field(state, 1);
+		scl0 = vc4_get_scl_field(state, 0);
 		scl1 = scl0;
 	} else {
 		scl0 = vc4_get_scl_field(state, 1);
 		scl1 = vc4_get_scl_field(state, 0);
 	}
 
-	switch (fb->modifier) {
+	switch (base_format_mod) {
 	case DRM_FORMAT_MOD_LINEAR:
 		tiling = SCALER_CTL0_TILING_LINEAR;
 		pitch0 = VC4_SET_FIELD(fb->pitches[0], SCALER_SRC_PITCH);
@@ -535,6 +537,49 @@ static int vc4_plane_mode_set(struct drm_plane *plane,
 		break;
 	}
 
+	case DRM_FORMAT_MOD_BROADCOM_SAND64:
+	case DRM_FORMAT_MOD_BROADCOM_SAND128:
+	case DRM_FORMAT_MOD_BROADCOM_SAND256: {
+		uint32_t param = fourcc_mod_broadcom_param(fb->modifier);
+
+		/* Column-based NV12 or RGBA.
+		 */
+		if (fb->format->num_planes > 1) {
+			if (hvs_format != HVS_PIXEL_FORMAT_YCBCR_YUV420_2PLANE) {
+				DRM_DEBUG_KMS("SAND format only valid for NV12/21");
+				return -EINVAL;
+			}
+			hvs_format = HVS_PIXEL_FORMAT_H264;
+		} else {
+			if (base_format_mod == DRM_FORMAT_MOD_BROADCOM_SAND256) {
+				DRM_DEBUG_KMS("SAND256 format only valid for H.264");
+				return -EINVAL;
+			}
+		}
+
+		switch (base_format_mod) {
+		case DRM_FORMAT_MOD_BROADCOM_SAND64:
+			tiling = SCALER_CTL0_TILING_64B;
+			break;
+		case DRM_FORMAT_MOD_BROADCOM_SAND128:
+			tiling = SCALER_CTL0_TILING_128B;
+			break;
+		case DRM_FORMAT_MOD_BROADCOM_SAND256:
+			tiling = SCALER_CTL0_TILING_256B_OR_T;
+			break;
+		default:
+			break;
+		}
+
+		if (param > SCALER_TILE_HEIGHT_MASK) {
+			DRM_DEBUG_KMS("SAND height too large (%d)\n", param);
+			return -EINVAL;
+		}
+
+		pitch0 = VC4_SET_FIELD(param, SCALER_TILE_HEIGHT);
+		break;
+	}
+
 	default:
 		DRM_DEBUG_KMS("Unsupported FB tiling flag 0x%16llx",
 			      (long long)fb->modifier);
@@ -544,8 +589,9 @@ static int vc4_plane_mode_set(struct drm_plane *plane,
 	/* Control word */
 	vc4_dlist_write(vc4_state,
 			SCALER_CTL0_VALID |
+			VC4_SET_FIELD(SCALER_CTL0_RGBA_EXPAND_ROUND, SCALER_CTL0_RGBA_EXPAND) |
 			(format->pixel_order << SCALER_CTL0_ORDER_SHIFT) |
-			(format->hvs << SCALER_CTL0_PIXEL_FORMAT_SHIFT) |
+			(hvs_format << SCALER_CTL0_PIXEL_FORMAT_SHIFT) |
 			VC4_SET_FIELD(tiling, SCALER_CTL0_TILING) |
 			(vc4_state->is_unity ? SCALER_CTL0_UNITY : 0) |
 			VC4_SET_FIELD(scl0, SCALER_CTL0_SCL0) |
@@ -607,8 +653,13 @@ static int vc4_plane_mode_set(struct drm_plane *plane,
 
 	/* Pitch word 1/2 */
 	for (i = 1; i < num_planes; i++) {
-		vc4_dlist_write(vc4_state,
-				VC4_SET_FIELD(fb->pitches[i], SCALER_SRC_PITCH));
+		if (hvs_format != HVS_PIXEL_FORMAT_H264) {
+			vc4_dlist_write(vc4_state,
+					VC4_SET_FIELD(fb->pitches[i],
+						      SCALER_SRC_PITCH));
+		} else {
+			vc4_dlist_write(vc4_state, pitch0);
+		}
 	}
 
 	/* Colorspace conversion words */
@@ -810,18 +861,21 @@ static int vc4_prepare_fb(struct drm_plane *plane,
 	struct dma_fence *fence;
 	int ret;
 
-	if ((plane->state->fb == state->fb) || !state->fb)
+	if (!state->fb)
 		return 0;
 
 	bo = to_vc4_bo(&drm_fb_cma_get_gem_obj(state->fb, 0)->base);
 
+	fence = reservation_object_get_excl_rcu(bo->resv);
+	drm_atomic_set_fence_for_plane(state, fence);
+
+	if (plane->state->fb == state->fb)
+		return 0;
+
 	ret = vc4_bo_inc_usecnt(bo);
 	if (ret)
 		return ret;
 
-	fence = reservation_object_get_excl_rcu(bo->resv);
-	drm_atomic_set_fence_for_plane(state, fence);
-
 	return 0;
 }
 
@@ -848,7 +902,7 @@ static const struct drm_plane_helper_funcs vc4_plane_helper_funcs = {
 
 static void vc4_plane_destroy(struct drm_plane *plane)
 {
-	drm_plane_helper_disable(plane);
+	drm_plane_helper_disable(plane, NULL);
 	drm_plane_cleanup(plane);
 }
 
@@ -866,13 +920,32 @@ static bool vc4_format_mod_supported(struct drm_plane *plane,
 	case DRM_FORMAT_BGR565:
 	case DRM_FORMAT_ARGB1555:
 	case DRM_FORMAT_XRGB1555:
-		return true;
+		switch (fourcc_mod_broadcom_mod(modifier)) {
+		case DRM_FORMAT_MOD_LINEAR:
+		case DRM_FORMAT_MOD_BROADCOM_VC4_T_TILED:
+		case DRM_FORMAT_MOD_BROADCOM_SAND64:
+		case DRM_FORMAT_MOD_BROADCOM_SAND128:
+			return true;
+		default:
+			return false;
+		}
+	case DRM_FORMAT_NV12:
+	case DRM_FORMAT_NV21:
+		switch (fourcc_mod_broadcom_mod(modifier)) {
+		case DRM_FORMAT_MOD_LINEAR:
+		case DRM_FORMAT_MOD_BROADCOM_SAND64:
+		case DRM_FORMAT_MOD_BROADCOM_SAND128:
+		case DRM_FORMAT_MOD_BROADCOM_SAND256:
+			return true;
+		default:
+			return false;
+		}
 	case DRM_FORMAT_YUV422:
 	case DRM_FORMAT_YVU422:
 	case DRM_FORMAT_YUV420:
 	case DRM_FORMAT_YVU420:
-	case DRM_FORMAT_NV12:
 	case DRM_FORMAT_NV16:
+	case DRM_FORMAT_NV61:
 	default:
 		return (modifier == DRM_FORMAT_MOD_LINEAR);
 	}
@@ -900,6 +973,9 @@ struct drm_plane *vc4_plane_init(struct drm_device *dev,
 	unsigned i;
 	static const uint64_t modifiers[] = {
 		DRM_FORMAT_MOD_BROADCOM_VC4_T_TILED,
+		DRM_FORMAT_MOD_BROADCOM_SAND128,
+		DRM_FORMAT_MOD_BROADCOM_SAND64,
+		DRM_FORMAT_MOD_BROADCOM_SAND256,
 		DRM_FORMAT_MOD_LINEAR,
 		DRM_FORMAT_MOD_INVALID
 	};
diff --git a/drivers/gpu/drm/vc4/vc4_regs.h b/drivers/gpu/drm/vc4/vc4_regs.h
index d1fb6fec46eb..d6864fa4bd14 100644
--- a/drivers/gpu/drm/vc4/vc4_regs.h
+++ b/drivers/gpu/drm/vc4/vc4_regs.h
@@ -1031,6 +1031,12 @@ enum hvs_pixel_format {
 #define SCALER_SRC_PITCH_MASK			VC4_MASK(15, 0)
 #define SCALER_SRC_PITCH_SHIFT			0
 
+/* PITCH0/1/2 fields for tiled (SAND). */
+#define SCALER_TILE_SKIP_0_MASK			VC4_MASK(18, 16)
+#define SCALER_TILE_SKIP_0_SHIFT		16
+#define SCALER_TILE_HEIGHT_MASK			VC4_MASK(15, 0)
+#define SCALER_TILE_HEIGHT_SHIFT		0
+
 /* PITCH0 fields for T-tiled. */
 #define SCALER_PITCH0_TILE_WIDTH_L_MASK		VC4_MASK(22, 16)
 #define SCALER_PITCH0_TILE_WIDTH_L_SHIFT	16
diff --git a/drivers/gpu/drm/vc4/vc4_txp.c b/drivers/gpu/drm/vc4/vc4_txp.c
new file mode 100644
index 000000000000..6e23c50168f9
--- /dev/null
+++ b/drivers/gpu/drm/vc4/vc4_txp.c
@@ -0,0 +1,477 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright © 2018 Broadcom
+ *
+ * Authors:
+ *	Eric Anholt <eric@anholt.net>
+ *	Boris Brezillon <boris.brezillon@bootlin.com>
+ */
+
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_fb_cma_helper.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_edid.h>
+#include <drm/drm_panel.h>
+#include <drm/drm_writeback.h>
+#include <linux/clk.h>
+#include <linux/component.h>
+#include <linux/of_graph.h>
+#include <linux/of_platform.h>
+#include <linux/pm_runtime.h>
+
+#include "vc4_drv.h"
+#include "vc4_regs.h"
+
+/* Base address of the output.  Raster formats must be 4-byte aligned,
+ * T and LT must be 16-byte aligned or maybe utile-aligned (docs are
+ * inconsistent, but probably utile).
+ */
+#define TXP_DST_PTR		0x00
+
+/* Pitch in bytes for raster images, 16-byte aligned.  For tiled, it's
+ * the width in tiles.
+ */
+#define TXP_DST_PITCH		0x04
+/* For T-tiled imgaes, DST_PITCH should be the number of tiles wide,
+ * shifted up.
+ */
+# define TXP_T_TILE_WIDTH_SHIFT		7
+/* For LT-tiled images, DST_PITCH should be the number of utiles wide,
+ * shifted up.
+ */
+# define TXP_LT_TILE_WIDTH_SHIFT	4
+
+/* Pre-rotation width/height of the image.  Must match HVS config.
+ *
+ * If TFORMAT and 32-bit, limit is 1920 for 32-bit and 3840 to 16-bit
+ * and width/height must be tile or utile-aligned as appropriate.  If
+ * transposing (rotating), width is limited to 1920.
+ *
+ * Height is limited to various numbers between 4088 and 4095.  I'd
+ * just use 4088 to be safe.
+ */
+#define TXP_DIM			0x08
+# define TXP_HEIGHT_SHIFT		16
+# define TXP_HEIGHT_MASK		GENMASK(31, 16)
+# define TXP_WIDTH_SHIFT		0
+# define TXP_WIDTH_MASK			GENMASK(15, 0)
+
+#define TXP_DST_CTRL		0x0c
+/* These bits are set to 0x54 */
+#define TXP_PILOT_SHIFT			24
+#define TXP_PILOT_MASK			GENMASK(31, 24)
+/* Bits 22-23 are set to 0x01 */
+#define TXP_VERSION_SHIFT		22
+#define TXP_VERSION_MASK		GENMASK(23, 22)
+
+/* Powers down the internal memory. */
+# define TXP_POWERDOWN			BIT(21)
+
+/* Enables storing the alpha component in 8888/4444, instead of
+ * filling with ~ALPHA_INVERT.
+ */
+# define TXP_ALPHA_ENABLE		BIT(20)
+
+/* 4 bits, each enables stores for a channel in each set of 4 bytes.
+ * Set to 0xf for normal operation.
+ */
+# define TXP_BYTE_ENABLE_SHIFT		16
+# define TXP_BYTE_ENABLE_MASK		GENMASK(19, 16)
+
+/* Debug: Generate VSTART again at EOF. */
+# define TXP_VSTART_AT_EOF		BIT(15)
+
+/* Debug: Terminate the current frame immediately.  Stops AXI
+ * writes.
+ */
+# define TXP_ABORT			BIT(14)
+
+# define TXP_DITHER			BIT(13)
+
+/* Inverts alpha if TXP_ALPHA_ENABLE, chooses fill value for
+ * !TXP_ALPHA_ENABLE.
+ */
+# define TXP_ALPHA_INVERT		BIT(12)
+
+/* Note: I've listed the channels here in high bit (in byte 3/2/1) to
+ * low bit (in byte 0) order.
+ */
+# define TXP_FORMAT_SHIFT		8
+# define TXP_FORMAT_MASK		GENMASK(11, 8)
+# define TXP_FORMAT_ABGR4444		0
+# define TXP_FORMAT_ARGB4444		1
+# define TXP_FORMAT_BGRA4444		2
+# define TXP_FORMAT_RGBA4444		3
+# define TXP_FORMAT_BGR565		6
+# define TXP_FORMAT_RGB565		7
+/* 888s are non-rotated, raster-only */
+# define TXP_FORMAT_BGR888		8
+# define TXP_FORMAT_RGB888		9
+# define TXP_FORMAT_ABGR8888		12
+# define TXP_FORMAT_ARGB8888		13
+# define TXP_FORMAT_BGRA8888		14
+# define TXP_FORMAT_RGBA8888		15
+
+/* If TFORMAT is set, generates LT instead of T format. */
+# define TXP_LINEAR_UTILE		BIT(7)
+
+/* Rotate output by 90 degrees. */
+# define TXP_TRANSPOSE			BIT(6)
+
+/* Generate a tiled format for V3D. */
+# define TXP_TFORMAT			BIT(5)
+
+/* Generates some undefined test mode output. */
+# define TXP_TEST_MODE			BIT(4)
+
+/* Request odd field from HVS. */
+# define TXP_FIELD			BIT(3)
+
+/* Raise interrupt when idle. */
+# define TXP_EI				BIT(2)
+
+/* Set when generating a frame, clears when idle. */
+# define TXP_BUSY			BIT(1)
+
+/* Starts a frame.  Self-clearing. */
+# define TXP_GO				BIT(0)
+
+/* Number of lines received and committed to memory. */
+#define TXP_PROGRESS		0x10
+
+#define TXP_READ(offset) readl(txp->regs + (offset))
+#define TXP_WRITE(offset, val) writel(val, txp->regs + (offset))
+
+struct vc4_txp {
+	struct platform_device *pdev;
+
+	struct drm_writeback_connector connector;
+
+	void __iomem *regs;
+};
+
+static inline struct vc4_txp *encoder_to_vc4_txp(struct drm_encoder *encoder)
+{
+	return container_of(encoder, struct vc4_txp, connector.encoder);
+}
+
+static inline struct vc4_txp *connector_to_vc4_txp(struct drm_connector *conn)
+{
+	return container_of(conn, struct vc4_txp, connector.base);
+}
+
+#define TXP_REG(reg) { reg, #reg }
+static const struct {
+	u32 reg;
+	const char *name;
+} txp_regs[] = {
+	TXP_REG(TXP_DST_PTR),
+	TXP_REG(TXP_DST_PITCH),
+	TXP_REG(TXP_DIM),
+	TXP_REG(TXP_DST_CTRL),
+	TXP_REG(TXP_PROGRESS),
+};
+
+#ifdef CONFIG_DEBUG_FS
+int vc4_txp_debugfs_regs(struct seq_file *m, void *unused)
+{
+	struct drm_info_node *node = (struct drm_info_node *)m->private;
+	struct drm_device *dev = node->minor->dev;
+	struct vc4_dev *vc4 = to_vc4_dev(dev);
+	struct vc4_txp *txp = vc4->txp;
+	int i;
+
+	if (!txp)
+		return 0;
+
+	for (i = 0; i < ARRAY_SIZE(txp_regs); i++) {
+		seq_printf(m, "%s (0x%04x): 0x%08x\n",
+			   txp_regs[i].name, txp_regs[i].reg,
+			   TXP_READ(txp_regs[i].reg));
+	}
+
+	return 0;
+}
+#endif
+
+static int vc4_txp_connector_get_modes(struct drm_connector *connector)
+{
+	struct drm_device *dev = connector->dev;
+
+	return drm_add_modes_noedid(connector, dev->mode_config.max_width,
+				    dev->mode_config.max_height);
+}
+
+static enum drm_mode_status
+vc4_txp_connector_mode_valid(struct drm_connector *connector,
+			     struct drm_display_mode *mode)
+{
+	struct drm_device *dev = connector->dev;
+	struct drm_mode_config *mode_config = &dev->mode_config;
+	int w = mode->hdisplay, h = mode->vdisplay;
+
+	if (w < mode_config->min_width || w > mode_config->max_width)
+		return MODE_BAD_HVALUE;
+
+	if (h < mode_config->min_height || h > mode_config->max_height)
+		return MODE_BAD_VVALUE;
+
+	return MODE_OK;
+}
+
+static const u32 drm_fmts[] = {
+	DRM_FORMAT_RGB888,
+	DRM_FORMAT_BGR888,
+	DRM_FORMAT_XRGB8888,
+	DRM_FORMAT_XBGR8888,
+	DRM_FORMAT_ARGB8888,
+	DRM_FORMAT_ABGR8888,
+	DRM_FORMAT_RGBX8888,
+	DRM_FORMAT_BGRX8888,
+	DRM_FORMAT_RGBA8888,
+	DRM_FORMAT_BGRA8888,
+};
+
+static const u32 txp_fmts[] = {
+	TXP_FORMAT_RGB888,
+	TXP_FORMAT_BGR888,
+	TXP_FORMAT_ARGB8888,
+	TXP_FORMAT_ABGR8888,
+	TXP_FORMAT_ARGB8888,
+	TXP_FORMAT_ABGR8888,
+	TXP_FORMAT_RGBA8888,
+	TXP_FORMAT_BGRA8888,
+	TXP_FORMAT_RGBA8888,
+	TXP_FORMAT_BGRA8888,
+};
+
+static int vc4_txp_connector_atomic_check(struct drm_connector *conn,
+					struct drm_connector_state *conn_state)
+{
+	struct drm_crtc_state *crtc_state;
+	struct drm_gem_cma_object *gem;
+	struct drm_framebuffer *fb;
+	int i;
+
+	if (!conn_state->writeback_job || !conn_state->writeback_job->fb)
+		return 0;
+
+	crtc_state = drm_atomic_get_new_crtc_state(conn_state->state,
+						   conn_state->crtc);
+
+	fb = conn_state->writeback_job->fb;
+	if (fb->width != crtc_state->mode.hdisplay ||
+	    fb->height != crtc_state->mode.vdisplay) {
+		DRM_DEBUG_KMS("Invalid framebuffer size %ux%u\n",
+			      fb->width, fb->height);
+		return -EINVAL;
+	}
+
+	for (i = 0; i < ARRAY_SIZE(drm_fmts); i++) {
+		if (fb->format->format == drm_fmts[i])
+			break;
+	}
+
+	if (i == ARRAY_SIZE(drm_fmts))
+		return -EINVAL;
+
+	gem = drm_fb_cma_get_gem_obj(fb, 0);
+
+	/* Pitch must be aligned on 16 bytes. */
+	if (fb->pitches[0] & GENMASK(3, 0))
+		return -EINVAL;
+
+	vc4_crtc_txp_armed(crtc_state);
+
+	return 0;
+}
+
+static void vc4_txp_connector_atomic_commit(struct drm_connector *conn,
+					struct drm_connector_state *conn_state)
+{
+	struct vc4_txp *txp = connector_to_vc4_txp(conn);
+	struct drm_gem_cma_object *gem;
+	struct drm_display_mode *mode;
+	struct drm_framebuffer *fb;
+	u32 ctrl;
+	int i;
+
+	if (WARN_ON(!conn_state->writeback_job ||
+		    !conn_state->writeback_job->fb))
+		return;
+
+	mode = &conn_state->crtc->state->adjusted_mode;
+	fb = conn_state->writeback_job->fb;
+
+	for (i = 0; i < ARRAY_SIZE(drm_fmts); i++) {
+		if (fb->format->format == drm_fmts[i])
+			break;
+	}
+
+	if (WARN_ON(i == ARRAY_SIZE(drm_fmts)))
+		return;
+
+	ctrl = TXP_GO | TXP_VSTART_AT_EOF | TXP_EI |
+	       VC4_SET_FIELD(0xf, TXP_BYTE_ENABLE) |
+	       VC4_SET_FIELD(txp_fmts[i], TXP_FORMAT);
+
+	if (fb->format->has_alpha)
+		ctrl |= TXP_ALPHA_ENABLE;
+
+	gem = drm_fb_cma_get_gem_obj(fb, 0);
+	TXP_WRITE(TXP_DST_PTR, gem->paddr + fb->offsets[0]);
+	TXP_WRITE(TXP_DST_PITCH, fb->pitches[0]);
+	TXP_WRITE(TXP_DIM,
+		  VC4_SET_FIELD(mode->hdisplay, TXP_WIDTH) |
+		  VC4_SET_FIELD(mode->vdisplay, TXP_HEIGHT));
+
+	TXP_WRITE(TXP_DST_CTRL, ctrl);
+
+	drm_writeback_queue_job(&txp->connector, conn_state->writeback_job);
+}
+
+static const struct drm_connector_helper_funcs vc4_txp_connector_helper_funcs = {
+	.get_modes = vc4_txp_connector_get_modes,
+	.mode_valid = vc4_txp_connector_mode_valid,
+	.atomic_check = vc4_txp_connector_atomic_check,
+	.atomic_commit = vc4_txp_connector_atomic_commit,
+};
+
+static enum drm_connector_status
+vc4_txp_connector_detect(struct drm_connector *connector, bool force)
+{
+	return connector_status_connected;
+}
+
+static void vc4_txp_connector_destroy(struct drm_connector *connector)
+{
+	drm_connector_unregister(connector);
+	drm_connector_cleanup(connector);
+}
+
+static const struct drm_connector_funcs vc4_txp_connector_funcs = {
+	.detect = vc4_txp_connector_detect,
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = vc4_txp_connector_destroy,
+	.reset = drm_atomic_helper_connector_reset,
+	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+static void vc4_txp_encoder_disable(struct drm_encoder *encoder)
+{
+	struct vc4_txp *txp = encoder_to_vc4_txp(encoder);
+
+	if (TXP_READ(TXP_DST_CTRL) & TXP_BUSY) {
+		unsigned long timeout = jiffies + msecs_to_jiffies(1000);
+
+		TXP_WRITE(TXP_DST_CTRL, TXP_ABORT);
+
+		while (TXP_READ(TXP_DST_CTRL) & TXP_BUSY &&
+		       time_before(jiffies, timeout))
+			;
+
+		WARN_ON(TXP_READ(TXP_DST_CTRL) & TXP_BUSY);
+	}
+
+	TXP_WRITE(TXP_DST_CTRL, TXP_POWERDOWN);
+}
+
+static const struct drm_encoder_helper_funcs vc4_txp_encoder_helper_funcs = {
+	.disable = vc4_txp_encoder_disable,
+};
+
+static irqreturn_t vc4_txp_interrupt(int irq, void *data)
+{
+	struct vc4_txp *txp = data;
+
+	TXP_WRITE(TXP_DST_CTRL, TXP_READ(TXP_DST_CTRL) & ~TXP_EI);
+	vc4_crtc_handle_vblank(to_vc4_crtc(txp->connector.base.state->crtc));
+	drm_writeback_signal_completion(&txp->connector, 0);
+
+	return IRQ_HANDLED;
+}
+
+static int vc4_txp_bind(struct device *dev, struct device *master, void *data)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct drm_device *drm = dev_get_drvdata(master);
+	struct vc4_dev *vc4 = to_vc4_dev(drm);
+	struct vc4_txp *txp;
+	int ret, irq;
+
+	irq = platform_get_irq(pdev, 0);
+	if (irq < 0)
+		return irq;
+
+	txp = devm_kzalloc(dev, sizeof(*txp), GFP_KERNEL);
+	if (!txp)
+		return -ENOMEM;
+
+	txp->pdev = pdev;
+
+	txp->regs = vc4_ioremap_regs(pdev, 0);
+	if (IS_ERR(txp->regs))
+		return PTR_ERR(txp->regs);
+
+	drm_connector_helper_add(&txp->connector.base,
+				 &vc4_txp_connector_helper_funcs);
+	ret = drm_writeback_connector_init(drm, &txp->connector,
+					   &vc4_txp_connector_funcs,
+					   &vc4_txp_encoder_helper_funcs,
+					   drm_fmts, ARRAY_SIZE(drm_fmts));
+	if (ret)
+		return ret;
+
+	ret = devm_request_irq(dev, irq, vc4_txp_interrupt, 0,
+			       dev_name(dev), txp);
+	if (ret)
+		return ret;
+
+	dev_set_drvdata(dev, txp);
+	vc4->txp = txp;
+
+	return 0;
+}
+
+static void vc4_txp_unbind(struct device *dev, struct device *master,
+			   void *data)
+{
+	struct drm_device *drm = dev_get_drvdata(master);
+	struct vc4_dev *vc4 = to_vc4_dev(drm);
+	struct vc4_txp *txp = dev_get_drvdata(dev);
+
+	vc4_txp_connector_destroy(&txp->connector.base);
+
+	vc4->txp = NULL;
+}
+
+static const struct component_ops vc4_txp_ops = {
+	.bind   = vc4_txp_bind,
+	.unbind = vc4_txp_unbind,
+};
+
+static int vc4_txp_probe(struct platform_device *pdev)
+{
+	return component_add(&pdev->dev, &vc4_txp_ops);
+}
+
+static int vc4_txp_remove(struct platform_device *pdev)
+{
+	component_del(&pdev->dev, &vc4_txp_ops);
+	return 0;
+}
+
+static const struct of_device_id vc4_txp_dt_match[] = {
+	{ .compatible = "brcm,bcm2835-txp" },
+	{ /* sentinel */ },
+};
+
+struct platform_driver vc4_txp_driver = {
+	.probe = vc4_txp_probe,
+	.remove = vc4_txp_remove,
+	.driver = {
+		.name = "vc4_txp",
+		.of_match_table = vc4_txp_dt_match,
+	},
+};
diff --git a/drivers/gpu/drm/vc4/vc4_v3d.c b/drivers/gpu/drm/vc4/vc4_v3d.c
index bfc2fa73d2ae..e47e29426078 100644
--- a/drivers/gpu/drm/vc4/vc4_v3d.c
+++ b/drivers/gpu/drm/vc4/vc4_v3d.c
@@ -218,8 +218,7 @@ try_again:
  * overall CMA pool before they make scenes complicated enough to run
  * out of bin space.
  */
-int
-vc4_allocate_bin_bo(struct drm_device *drm)
+static int vc4_allocate_bin_bo(struct drm_device *drm)
 {
 	struct vc4_dev *vc4 = to_vc4_dev(drm);
 	struct vc4_v3d *v3d = vc4->v3d;
diff --git a/drivers/gpu/drm/vc4/vc4_vec.c b/drivers/gpu/drm/vc4/vc4_vec.c
index 3a9a302247a2..8e7facb6514e 100644
--- a/drivers/gpu/drm/vc4/vc4_vec.c
+++ b/drivers/gpu/drm/vc4/vc4_vec.c
@@ -404,7 +404,7 @@ static struct drm_connector *vc4_vec_connector_init(struct drm_device *dev,
 				   VC4_VEC_TV_MODE_NTSC);
 	vec->tv_mode = &vc4_vec_tv_modes[VC4_VEC_TV_MODE_NTSC];
 
-	drm_mode_connector_attach_encoder(connector, vec->encoder);
+	drm_connector_attach_encoder(connector, vec->encoder);
 
 	return connector;
 }
diff --git a/drivers/gpu/drm/vgem/vgem_drv.c b/drivers/gpu/drm/vgem/vgem_drv.c
index 2524ff116f00..0e5620f76ee0 100644
--- a/drivers/gpu/drm/vgem/vgem_drv.c
+++ b/drivers/gpu/drm/vgem/vgem_drv.c
@@ -61,23 +61,22 @@ static void vgem_gem_free_object(struct drm_gem_object *obj)
 	kfree(vgem_obj);
 }
 
-static int vgem_gem_fault(struct vm_fault *vmf)
+static vm_fault_t vgem_gem_fault(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 	struct drm_vgem_gem_object *obj = vma->vm_private_data;
 	/* We don't use vmf->pgoff since that has the fake offset */
 	unsigned long vaddr = vmf->address;
-	int ret;
+	vm_fault_t ret = VM_FAULT_SIGBUS;
 	loff_t num_pages;
 	pgoff_t page_offset;
 	page_offset = (vaddr - vma->vm_start) >> PAGE_SHIFT;
 
 	num_pages = DIV_ROUND_UP(obj->base.size, PAGE_SIZE);
 
-	if (page_offset > num_pages)
+	if (page_offset >= num_pages)
 		return VM_FAULT_SIGBUS;
 
-	ret = -ENOENT;
 	mutex_lock(&obj->pages_lock);
 	if (obj->pages) {
 		get_page(obj->pages[page_offset]);
diff --git a/drivers/gpu/drm/via/via_dmablit.c b/drivers/gpu/drm/via/via_dmablit.c
index d6e84a589ef1..345bda4494e1 100644
--- a/drivers/gpu/drm/via/via_dmablit.c
+++ b/drivers/gpu/drm/via/via_dmablit.c
@@ -235,7 +235,7 @@ via_lock_all_dma_pages(drm_via_sg_info_t *vsg,  drm_via_dmablit_t *xfer)
 	vsg->num_pages = VIA_PFN(xfer->mem_addr + (xfer->num_lines * xfer->mem_stride - 1)) -
 		first_pfn + 1;
 
-	vsg->pages = vzalloc(sizeof(struct page *) * vsg->num_pages);
+	vsg->pages = vzalloc(array_size(sizeof(struct page *), vsg->num_pages));
 	if (NULL == vsg->pages)
 		return -ENOMEM;
 	ret = get_user_pages_fast((unsigned long)xfer->mem_addr,
diff --git a/drivers/gpu/drm/virtio/virtgpu_display.c b/drivers/gpu/drm/virtio/virtgpu_display.c
index a5edd86603d9..25503b933599 100644
--- a/drivers/gpu/drm/virtio/virtgpu_display.c
+++ b/drivers/gpu/drm/virtio/virtgpu_display.c
@@ -28,6 +28,7 @@
 #include "virtgpu_drv.h"
 #include <drm/drm_crtc_helper.h>
 #include <drm/drm_atomic_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
 
 #define XRES_MIN    32
 #define YRES_MIN    32
@@ -48,16 +49,6 @@ static const struct drm_crtc_funcs virtio_gpu_crtc_funcs = {
 	.atomic_destroy_state   = drm_atomic_helper_crtc_destroy_state,
 };
 
-static void virtio_gpu_user_framebuffer_destroy(struct drm_framebuffer *fb)
-{
-	struct virtio_gpu_framebuffer *virtio_gpu_fb
-		= to_virtio_gpu_framebuffer(fb);
-
-	drm_gem_object_put_unlocked(virtio_gpu_fb->obj);
-	drm_framebuffer_cleanup(fb);
-	kfree(virtio_gpu_fb);
-}
-
 static int
 virtio_gpu_framebuffer_surface_dirty(struct drm_framebuffer *fb,
 				     struct drm_file *file_priv,
@@ -71,20 +62,9 @@ virtio_gpu_framebuffer_surface_dirty(struct drm_framebuffer *fb,
 	return virtio_gpu_surface_dirty(virtio_gpu_fb, clips, num_clips);
 }
 
-static int
-virtio_gpu_framebuffer_create_handle(struct drm_framebuffer *fb,
-				     struct drm_file *file_priv,
-				     unsigned int *handle)
-{
-	struct virtio_gpu_framebuffer *virtio_gpu_fb =
-		to_virtio_gpu_framebuffer(fb);
-
-	return drm_gem_handle_create(file_priv, virtio_gpu_fb->obj, handle);
-}
-
 static const struct drm_framebuffer_funcs virtio_gpu_fb_funcs = {
-	.create_handle = virtio_gpu_framebuffer_create_handle,
-	.destroy = virtio_gpu_user_framebuffer_destroy,
+	.create_handle = drm_gem_fb_create_handle,
+	.destroy = drm_gem_fb_destroy,
 	.dirty = virtio_gpu_framebuffer_surface_dirty,
 };
 
@@ -97,7 +77,7 @@ virtio_gpu_framebuffer_init(struct drm_device *dev,
 	int ret;
 	struct virtio_gpu_object *bo;
 
-	vgfb->obj = obj;
+	vgfb->base.obj[0] = obj;
 
 	bo = gem_to_virtio_gpu_obj(obj);
 
@@ -105,7 +85,7 @@ virtio_gpu_framebuffer_init(struct drm_device *dev,
 
 	ret = drm_framebuffer_init(dev, &vgfb->base, &virtio_gpu_fb_funcs);
 	if (ret) {
-		vgfb->obj = NULL;
+		vgfb->base.obj[0] = NULL;
 		return ret;
 	}
 
@@ -302,8 +282,6 @@ static int vgdev_output_init(struct virtio_gpu_device *vgdev, int index)
 	drm_crtc_init_with_planes(dev, crtc, primary, cursor,
 				  &virtio_gpu_crtc_funcs, NULL);
 	drm_crtc_helper_add(crtc, &virtio_gpu_crtc_helper_funcs);
-	primary->crtc = crtc;
-	cursor->crtc = crtc;
 
 	drm_connector_init(dev, connector, &virtio_gpu_connector_funcs,
 			   DRM_MODE_CONNECTOR_VIRTUAL);
@@ -314,7 +292,7 @@ static int vgdev_output_init(struct virtio_gpu_device *vgdev, int index)
 	drm_encoder_helper_add(encoder, &virtio_gpu_enc_helper_funcs);
 	encoder->possible_crtcs = 1 << index;
 
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 	drm_connector_register(connector);
 	return 0;
 }
diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h
index d25c8ca224aa..65605e207bbe 100644
--- a/drivers/gpu/drm/virtio/virtgpu_drv.h
+++ b/drivers/gpu/drm/virtio/virtgpu_drv.h
@@ -124,7 +124,6 @@ struct virtio_gpu_output {
 
 struct virtio_gpu_framebuffer {
 	struct drm_framebuffer base;
-	struct drm_gem_object *obj;
 	int x1, y1, x2, y2; /* dirty rect */
 	spinlock_t dirty_lock;
 	uint32_t hw_res_handle;
diff --git a/drivers/gpu/drm/virtio/virtgpu_fb.c b/drivers/gpu/drm/virtio/virtgpu_fb.c
index 8af69ab58b89..a121b1c79522 100644
--- a/drivers/gpu/drm/virtio/virtgpu_fb.c
+++ b/drivers/gpu/drm/virtio/virtgpu_fb.c
@@ -46,7 +46,7 @@ static int virtio_gpu_dirty_update(struct virtio_gpu_framebuffer *fb,
 	int bpp = fb->base.format->cpp[0];
 	int x2, y2;
 	unsigned long flags;
-	struct virtio_gpu_object *obj = gem_to_virtio_gpu_obj(fb->obj);
+	struct virtio_gpu_object *obj = gem_to_virtio_gpu_obj(fb->base.obj[0]);
 
 	if ((width <= 0) ||
 	    (x + width > fb->base.width) ||
@@ -121,7 +121,7 @@ int virtio_gpu_surface_dirty(struct virtio_gpu_framebuffer *vgfb,
 			     unsigned int num_clips)
 {
 	struct virtio_gpu_device *vgdev = vgfb->base.dev->dev_private;
-	struct virtio_gpu_object *obj = gem_to_virtio_gpu_obj(vgfb->obj);
+	struct virtio_gpu_object *obj = gem_to_virtio_gpu_obj(vgfb->base.obj[0]);
 	struct drm_clip_rect norect;
 	struct drm_clip_rect *clips_ptr;
 	int left, right, top, bottom;
@@ -305,8 +305,8 @@ static int virtio_gpu_fbdev_destroy(struct drm_device *dev,
 
 	drm_fb_helper_unregister_fbi(&vgfbdev->helper);
 
-	if (vgfb->obj)
-		vgfb->obj = NULL;
+	if (vgfb->base.obj[0])
+		vgfb->base.obj[0] = NULL;
 	drm_fb_helper_fini(&vgfbdev->helper);
 	drm_framebuffer_cleanup(&vgfb->base);
 
diff --git a/drivers/gpu/drm/virtio/virtgpu_fence.c b/drivers/gpu/drm/virtio/virtgpu_fence.c
index 23353521f903..00c742a441bf 100644
--- a/drivers/gpu/drm/virtio/virtgpu_fence.c
+++ b/drivers/gpu/drm/virtio/virtgpu_fence.c
@@ -36,11 +36,6 @@ static const char *virtio_get_timeline_name(struct dma_fence *f)
 	return "controlq";
 }
 
-static bool virtio_enable_signaling(struct dma_fence *f)
-{
-	return true;
-}
-
 static bool virtio_signaled(struct dma_fence *f)
 {
 	struct virtio_gpu_fence *fence = to_virtio_fence(f);
@@ -67,9 +62,7 @@ static void virtio_timeline_value_str(struct dma_fence *f, char *str, int size)
 static const struct dma_fence_ops virtio_fence_ops = {
 	.get_driver_name     = virtio_get_driver_name,
 	.get_timeline_name   = virtio_get_timeline_name,
-	.enable_signaling    = virtio_enable_signaling,
 	.signaled            = virtio_signaled,
-	.wait                = dma_fence_default_wait,
 	.fence_value_str     = virtio_fence_value_str,
 	.timeline_value_str  = virtio_timeline_value_str,
 };
diff --git a/drivers/gpu/drm/virtio/virtgpu_plane.c b/drivers/gpu/drm/virtio/virtgpu_plane.c
index 71ba455af915..dc5b5b2b7aab 100644
--- a/drivers/gpu/drm/virtio/virtgpu_plane.c
+++ b/drivers/gpu/drm/virtio/virtgpu_plane.c
@@ -154,7 +154,7 @@ static void virtio_gpu_primary_plane_update(struct drm_plane *plane,
 
 	if (plane->state->fb) {
 		vgfb = to_virtio_gpu_framebuffer(plane->state->fb);
-		bo = gem_to_virtio_gpu_obj(vgfb->obj);
+		bo = gem_to_virtio_gpu_obj(vgfb->base.obj[0]);
 		handle = bo->hw_res_handle;
 		if (bo->dumb) {
 			virtio_gpu_cmd_transfer_to_host_2d
@@ -208,7 +208,7 @@ static void virtio_gpu_cursor_plane_update(struct drm_plane *plane,
 
 	if (plane->state->fb) {
 		vgfb = to_virtio_gpu_framebuffer(plane->state->fb);
-		bo = gem_to_virtio_gpu_obj(vgfb->obj);
+		bo = gem_to_virtio_gpu_obj(vgfb->base.obj[0]);
 		handle = bo->hw_res_handle;
 	} else {
 		handle = 0;
diff --git a/drivers/gpu/drm/vkms/Makefile b/drivers/gpu/drm/vkms/Makefile
new file mode 100644
index 000000000000..986297da51bf
--- /dev/null
+++ b/drivers/gpu/drm/vkms/Makefile
@@ -0,0 +1,3 @@
+vkms-y := vkms_drv.o vkms_plane.o vkms_output.o vkms_crtc.o vkms_gem.o
+
+obj-$(CONFIG_DRM_VKMS) += vkms.o
diff --git a/drivers/gpu/drm/vkms/vkms_crtc.c b/drivers/gpu/drm/vkms/vkms_crtc.c
new file mode 100644
index 000000000000..875fca662ac0
--- /dev/null
+++ b/drivers/gpu/drm/vkms/vkms_crtc.c
@@ -0,0 +1,130 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include "vkms_drv.h"
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_crtc_helper.h>
+
+static enum hrtimer_restart vkms_vblank_simulate(struct hrtimer *timer)
+{
+	struct vkms_output *output = container_of(timer, struct vkms_output,
+						  vblank_hrtimer);
+	struct drm_crtc *crtc = &output->crtc;
+	int ret_overrun;
+	bool ret;
+
+	ret = drm_crtc_handle_vblank(crtc);
+	if (!ret)
+		DRM_ERROR("vkms failure on handling vblank");
+
+	ret_overrun = hrtimer_forward_now(&output->vblank_hrtimer,
+					  output->period_ns);
+
+	return HRTIMER_RESTART;
+}
+
+static int vkms_enable_vblank(struct drm_crtc *crtc)
+{
+	struct drm_device *dev = crtc->dev;
+	unsigned int pipe = drm_crtc_index(crtc);
+	struct drm_vblank_crtc *vblank = &dev->vblank[pipe];
+	struct vkms_output *out = drm_crtc_to_vkms_output(crtc);
+
+	drm_calc_timestamping_constants(crtc, &crtc->mode);
+
+	hrtimer_init(&out->vblank_hrtimer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
+	out->vblank_hrtimer.function = &vkms_vblank_simulate;
+	out->period_ns = ktime_set(0, vblank->framedur_ns);
+	hrtimer_start(&out->vblank_hrtimer, out->period_ns, HRTIMER_MODE_REL);
+
+	return 0;
+}
+
+static void vkms_disable_vblank(struct drm_crtc *crtc)
+{
+	struct vkms_output *out = drm_crtc_to_vkms_output(crtc);
+
+	hrtimer_cancel(&out->vblank_hrtimer);
+}
+
+bool vkms_get_vblank_timestamp(struct drm_device *dev, unsigned int pipe,
+			       int *max_error, ktime_t *vblank_time,
+			       bool in_vblank_irq)
+{
+	struct vkms_device *vkmsdev = drm_device_to_vkms_device(dev);
+	struct vkms_output *output = &vkmsdev->output;
+
+	*vblank_time = output->vblank_hrtimer.node.expires;
+
+	return true;
+}
+
+static const struct drm_crtc_funcs vkms_crtc_funcs = {
+	.set_config             = drm_atomic_helper_set_config,
+	.destroy                = drm_crtc_cleanup,
+	.page_flip              = drm_atomic_helper_page_flip,
+	.reset                  = drm_atomic_helper_crtc_reset,
+	.atomic_duplicate_state = drm_atomic_helper_crtc_duplicate_state,
+	.atomic_destroy_state   = drm_atomic_helper_crtc_destroy_state,
+	.enable_vblank		= vkms_enable_vblank,
+	.disable_vblank		= vkms_disable_vblank,
+};
+
+static void vkms_crtc_atomic_enable(struct drm_crtc *crtc,
+				    struct drm_crtc_state *old_state)
+{
+	drm_crtc_vblank_on(crtc);
+}
+
+static void vkms_crtc_atomic_disable(struct drm_crtc *crtc,
+				     struct drm_crtc_state *old_state)
+{
+	drm_crtc_vblank_off(crtc);
+}
+
+static void vkms_crtc_atomic_flush(struct drm_crtc *crtc,
+				   struct drm_crtc_state *old_crtc_state)
+{
+	unsigned long flags;
+
+	if (crtc->state->event) {
+		spin_lock_irqsave(&crtc->dev->event_lock, flags);
+
+		if (drm_crtc_vblank_get(crtc) != 0)
+			drm_crtc_send_vblank_event(crtc, crtc->state->event);
+		else
+			drm_crtc_arm_vblank_event(crtc, crtc->state->event);
+
+		spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
+
+		crtc->state->event = NULL;
+	}
+}
+
+static const struct drm_crtc_helper_funcs vkms_crtc_helper_funcs = {
+	.atomic_flush	= vkms_crtc_atomic_flush,
+	.atomic_enable	= vkms_crtc_atomic_enable,
+	.atomic_disable	= vkms_crtc_atomic_disable,
+};
+
+int vkms_crtc_init(struct drm_device *dev, struct drm_crtc *crtc,
+		   struct drm_plane *primary, struct drm_plane *cursor)
+{
+	int ret;
+
+	ret = drm_crtc_init_with_planes(dev, crtc, primary, cursor,
+					&vkms_crtc_funcs, NULL);
+	if (ret) {
+		DRM_ERROR("Failed to init CRTC\n");
+		return ret;
+	}
+
+	drm_crtc_helper_add(crtc, &vkms_crtc_helper_funcs);
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/vkms/vkms_drv.c b/drivers/gpu/drm/vkms/vkms_drv.c
new file mode 100644
index 000000000000..37aa2ef33b21
--- /dev/null
+++ b/drivers/gpu/drm/vkms/vkms_drv.c
@@ -0,0 +1,155 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/module.h>
+#include <drm/drm_gem.h>
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_atomic_helper.h>
+#include <drm/drm_gem_framebuffer_helper.h>
+#include <drm/drm_fb_helper.h>
+#include "vkms_drv.h"
+
+#define DRIVER_NAME	"vkms"
+#define DRIVER_DESC	"Virtual Kernel Mode Setting"
+#define DRIVER_DATE	"20180514"
+#define DRIVER_MAJOR	1
+#define DRIVER_MINOR	0
+
+static struct vkms_device *vkms_device;
+
+static const struct file_operations vkms_driver_fops = {
+	.owner		= THIS_MODULE,
+	.open		= drm_open,
+	.mmap		= drm_gem_mmap,
+	.unlocked_ioctl	= drm_ioctl,
+	.compat_ioctl	= drm_compat_ioctl,
+	.poll		= drm_poll,
+	.read		= drm_read,
+	.llseek		= no_llseek,
+	.release	= drm_release,
+};
+
+static const struct vm_operations_struct vkms_gem_vm_ops = {
+	.fault = vkms_gem_fault,
+	.open = drm_gem_vm_open,
+	.close = drm_gem_vm_close,
+};
+
+static void vkms_release(struct drm_device *dev)
+{
+	struct vkms_device *vkms = container_of(dev, struct vkms_device, drm);
+
+	platform_device_unregister(vkms->platform);
+	drm_mode_config_cleanup(&vkms->drm);
+	drm_dev_fini(&vkms->drm);
+}
+
+static struct drm_driver vkms_driver = {
+	.driver_features	= DRIVER_MODESET | DRIVER_ATOMIC | DRIVER_GEM,
+	.release		= vkms_release,
+	.fops			= &vkms_driver_fops,
+	.dumb_create		= vkms_dumb_create,
+	.dumb_map_offset	= vkms_dumb_map,
+	.gem_vm_ops		= &vkms_gem_vm_ops,
+	.gem_free_object_unlocked = vkms_gem_free_object,
+	.get_vblank_timestamp	= vkms_get_vblank_timestamp,
+
+	.name			= DRIVER_NAME,
+	.desc			= DRIVER_DESC,
+	.date			= DRIVER_DATE,
+	.major			= DRIVER_MAJOR,
+	.minor			= DRIVER_MINOR,
+};
+
+static const struct drm_mode_config_funcs vkms_mode_funcs = {
+	.fb_create = drm_gem_fb_create,
+	.atomic_check = drm_atomic_helper_check,
+	.atomic_commit = drm_atomic_helper_commit,
+};
+
+static int vkms_modeset_init(struct vkms_device *vkmsdev)
+{
+	struct drm_device *dev = &vkmsdev->drm;
+
+	drm_mode_config_init(dev);
+	dev->mode_config.funcs = &vkms_mode_funcs;
+	dev->mode_config.min_width = XRES_MIN;
+	dev->mode_config.min_height = YRES_MIN;
+	dev->mode_config.max_width = XRES_MAX;
+	dev->mode_config.max_height = YRES_MAX;
+
+	return vkms_output_init(vkmsdev);
+}
+
+static int __init vkms_init(void)
+{
+	int ret;
+
+	vkms_device = kzalloc(sizeof(*vkms_device), GFP_KERNEL);
+	if (!vkms_device)
+		return -ENOMEM;
+
+	ret = drm_dev_init(&vkms_device->drm, &vkms_driver, NULL);
+	if (ret)
+		goto out_free;
+
+	vkms_device->platform =
+		platform_device_register_simple(DRIVER_NAME, -1, NULL, 0);
+	if (IS_ERR(vkms_device->platform)) {
+		ret = PTR_ERR(vkms_device->platform);
+		goto out_fini;
+	}
+
+	vkms_device->drm.irq_enabled = true;
+
+	ret = drm_vblank_init(&vkms_device->drm, 1);
+	if (ret) {
+		DRM_ERROR("Failed to vblank\n");
+		goto out_fini;
+	}
+
+	ret = vkms_modeset_init(vkms_device);
+	if (ret)
+		goto out_unregister;
+
+	ret = drm_dev_register(&vkms_device->drm, 0);
+	if (ret)
+		goto out_unregister;
+
+	return 0;
+
+out_unregister:
+	platform_device_unregister(vkms_device->platform);
+
+out_fini:
+	drm_dev_fini(&vkms_device->drm);
+
+out_free:
+	kfree(vkms_device);
+	return ret;
+}
+
+static void __exit vkms_exit(void)
+{
+	if (!vkms_device) {
+		DRM_INFO("vkms_device is NULL.\n");
+		return;
+	}
+
+	drm_dev_unregister(&vkms_device->drm);
+	drm_dev_put(&vkms_device->drm);
+
+	kfree(vkms_device);
+}
+
+module_init(vkms_init);
+module_exit(vkms_exit);
+
+MODULE_AUTHOR("Haneen Mohammed <hamohammed.sa@gmail.com>");
+MODULE_AUTHOR("Rodrigo Siqueira <rodrigosiqueiramelo@gmail.com>");
+MODULE_DESCRIPTION(DRIVER_DESC);
+MODULE_LICENSE("GPL");
diff --git a/drivers/gpu/drm/vkms/vkms_drv.h b/drivers/gpu/drm/vkms/vkms_drv.h
new file mode 100644
index 000000000000..07be29f2dc44
--- /dev/null
+++ b/drivers/gpu/drm/vkms/vkms_drv.h
@@ -0,0 +1,78 @@
+#ifndef _VKMS_DRV_H_
+#define _VKMS_DRV_H_
+
+#include <drm/drmP.h>
+#include <drm/drm.h>
+#include <drm/drm_gem.h>
+#include <drm/drm_encoder.h>
+#include <linux/hrtimer.h>
+
+#define XRES_MIN    32
+#define YRES_MIN    32
+
+#define XRES_DEF  1024
+#define YRES_DEF   768
+
+#define XRES_MAX  8192
+#define YRES_MAX  8192
+
+static const u32 vkms_formats[] = {
+	DRM_FORMAT_XRGB8888,
+};
+
+struct vkms_output {
+	struct drm_crtc crtc;
+	struct drm_encoder encoder;
+	struct drm_connector connector;
+	struct hrtimer vblank_hrtimer;
+	ktime_t period_ns;
+	struct drm_pending_vblank_event *event;
+};
+
+struct vkms_device {
+	struct drm_device drm;
+	struct platform_device *platform;
+	struct vkms_output output;
+};
+
+struct vkms_gem_object {
+	struct drm_gem_object gem;
+	struct mutex pages_lock; /* Page lock used in page fault handler */
+	struct page **pages;
+};
+
+#define drm_crtc_to_vkms_output(target) \
+	container_of(target, struct vkms_output, crtc)
+
+#define drm_device_to_vkms_device(target) \
+	container_of(target, struct vkms_device, drm)
+
+/* CRTC */
+int vkms_crtc_init(struct drm_device *dev, struct drm_crtc *crtc,
+		   struct drm_plane *primary, struct drm_plane *cursor);
+
+bool vkms_get_vblank_timestamp(struct drm_device *dev, unsigned int pipe,
+			       int *max_error, ktime_t *vblank_time,
+			       bool in_vblank_irq);
+
+int vkms_output_init(struct vkms_device *vkmsdev);
+
+struct drm_plane *vkms_plane_init(struct vkms_device *vkmsdev);
+
+/* Gem stuff */
+struct drm_gem_object *vkms_gem_create(struct drm_device *dev,
+				       struct drm_file *file,
+				       u32 *handle,
+				       u64 size);
+
+int vkms_gem_fault(struct vm_fault *vmf);
+
+int vkms_dumb_create(struct drm_file *file, struct drm_device *dev,
+		     struct drm_mode_create_dumb *args);
+
+int vkms_dumb_map(struct drm_file *file, struct drm_device *dev,
+		  u32 handle, u64 *offset);
+
+void vkms_gem_free_object(struct drm_gem_object *obj);
+
+#endif /* _VKMS_DRV_H_ */
diff --git a/drivers/gpu/drm/vkms/vkms_gem.c b/drivers/gpu/drm/vkms/vkms_gem.c
new file mode 100644
index 000000000000..c7e38368602b
--- /dev/null
+++ b/drivers/gpu/drm/vkms/vkms_gem.c
@@ -0,0 +1,179 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include <linux/shmem_fs.h>
+
+#include "vkms_drv.h"
+
+static struct vkms_gem_object *__vkms_gem_create(struct drm_device *dev,
+						 u64 size)
+{
+	struct vkms_gem_object *obj;
+	int ret;
+
+	obj = kzalloc(sizeof(*obj), GFP_KERNEL);
+	if (!obj)
+		return ERR_PTR(-ENOMEM);
+
+	size = roundup(size, PAGE_SIZE);
+	ret = drm_gem_object_init(dev, &obj->gem, size);
+	if (ret) {
+		kfree(obj);
+		return ERR_PTR(ret);
+	}
+
+	mutex_init(&obj->pages_lock);
+
+	return obj;
+}
+
+void vkms_gem_free_object(struct drm_gem_object *obj)
+{
+	struct vkms_gem_object *gem = container_of(obj, struct vkms_gem_object,
+						   gem);
+
+	kvfree(gem->pages);
+	mutex_destroy(&gem->pages_lock);
+	drm_gem_object_release(obj);
+	kfree(gem);
+}
+
+int vkms_gem_fault(struct vm_fault *vmf)
+{
+	struct vm_area_struct *vma = vmf->vma;
+	struct vkms_gem_object *obj = vma->vm_private_data;
+	unsigned long vaddr = vmf->address;
+	pgoff_t page_offset;
+	loff_t num_pages;
+	int ret;
+
+	page_offset = (vaddr - vma->vm_start) >> PAGE_SHIFT;
+	num_pages = DIV_ROUND_UP(obj->gem.size, PAGE_SIZE);
+
+	if (page_offset > num_pages)
+		return VM_FAULT_SIGBUS;
+
+	ret = -ENOENT;
+	mutex_lock(&obj->pages_lock);
+	if (obj->pages) {
+		get_page(obj->pages[page_offset]);
+		vmf->page = obj->pages[page_offset];
+		ret = 0;
+	}
+	mutex_unlock(&obj->pages_lock);
+	if (ret) {
+		struct page *page;
+		struct address_space *mapping;
+
+		mapping = file_inode(obj->gem.filp)->i_mapping;
+		page = shmem_read_mapping_page(mapping, page_offset);
+
+		if (!IS_ERR(page)) {
+			vmf->page = page;
+			ret = 0;
+		} else {
+			switch (PTR_ERR(page)) {
+			case -ENOSPC:
+			case -ENOMEM:
+				ret = VM_FAULT_OOM;
+				break;
+			case -EBUSY:
+				ret = VM_FAULT_RETRY;
+				break;
+			case -EFAULT:
+			case -EINVAL:
+				ret = VM_FAULT_SIGBUS;
+				break;
+			default:
+				WARN_ON(PTR_ERR(page));
+				ret = VM_FAULT_SIGBUS;
+				break;
+			}
+		}
+	}
+	return ret;
+}
+
+struct drm_gem_object *vkms_gem_create(struct drm_device *dev,
+				       struct drm_file *file,
+				       u32 *handle,
+				       u64 size)
+{
+	struct vkms_gem_object *obj;
+	int ret;
+
+	if (!file || !dev || !handle)
+		return ERR_PTR(-EINVAL);
+
+	obj = __vkms_gem_create(dev, size);
+	if (IS_ERR(obj))
+		return ERR_CAST(obj);
+
+	ret = drm_gem_handle_create(file, &obj->gem, handle);
+	drm_gem_object_put_unlocked(&obj->gem);
+	if (ret) {
+		drm_gem_object_release(&obj->gem);
+		kfree(obj);
+		return ERR_PTR(ret);
+	}
+
+	return &obj->gem;
+}
+
+int vkms_dumb_create(struct drm_file *file, struct drm_device *dev,
+		     struct drm_mode_create_dumb *args)
+{
+	struct drm_gem_object *gem_obj;
+	u64 pitch, size;
+
+	if (!args || !dev || !file)
+		return -EINVAL;
+
+	pitch = args->width * DIV_ROUND_UP(args->bpp, 8);
+	size = pitch * args->height;
+
+	if (!size)
+		return -EINVAL;
+
+	gem_obj = vkms_gem_create(dev, file, &args->handle, size);
+	if (IS_ERR(gem_obj))
+		return PTR_ERR(gem_obj);
+
+	args->size = gem_obj->size;
+	args->pitch = pitch;
+
+	DRM_DEBUG_DRIVER("Created object of size %lld\n", size);
+
+	return 0;
+}
+
+int vkms_dumb_map(struct drm_file *file, struct drm_device *dev,
+		  u32 handle, u64 *offset)
+{
+	struct drm_gem_object *obj;
+	int ret;
+
+	obj = drm_gem_object_lookup(file, handle);
+	if (!obj)
+		return -ENOENT;
+
+	if (!obj->filp) {
+		ret = -EINVAL;
+		goto unref;
+	}
+
+	ret = drm_gem_create_mmap_offset(obj);
+	if (ret)
+		goto unref;
+
+	*offset = drm_vma_node_offset_addr(&obj->vma_node);
+unref:
+	drm_gem_object_put_unlocked(obj);
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/vkms/vkms_output.c b/drivers/gpu/drm/vkms/vkms_output.c
new file mode 100644
index 000000000000..901012cb1af1
--- /dev/null
+++ b/drivers/gpu/drm/vkms/vkms_output.c
@@ -0,0 +1,111 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include "vkms_drv.h"
+#include <drm/drm_crtc_helper.h>
+#include <drm/drm_atomic_helper.h>
+
+static void vkms_connector_destroy(struct drm_connector *connector)
+{
+	drm_connector_unregister(connector);
+	drm_connector_cleanup(connector);
+}
+
+static const struct drm_connector_funcs vkms_connector_funcs = {
+	.fill_modes = drm_helper_probe_single_connector_modes,
+	.destroy = vkms_connector_destroy,
+	.reset = drm_atomic_helper_connector_reset,
+	.atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state,
+	.atomic_destroy_state = drm_atomic_helper_connector_destroy_state,
+};
+
+static const struct drm_encoder_funcs vkms_encoder_funcs = {
+	.destroy = drm_encoder_cleanup,
+};
+
+static int vkms_conn_get_modes(struct drm_connector *connector)
+{
+	int count;
+
+	count = drm_add_modes_noedid(connector, XRES_MAX, YRES_MAX);
+	drm_set_preferred_mode(connector, XRES_DEF, YRES_DEF);
+
+	return count;
+}
+
+static const struct drm_connector_helper_funcs vkms_conn_helper_funcs = {
+	.get_modes    = vkms_conn_get_modes,
+};
+
+int vkms_output_init(struct vkms_device *vkmsdev)
+{
+	struct vkms_output *output = &vkmsdev->output;
+	struct drm_device *dev = &vkmsdev->drm;
+	struct drm_connector *connector = &output->connector;
+	struct drm_encoder *encoder = &output->encoder;
+	struct drm_crtc *crtc = &output->crtc;
+	struct drm_plane *primary;
+	int ret;
+
+	primary = vkms_plane_init(vkmsdev);
+	if (IS_ERR(primary))
+		return PTR_ERR(primary);
+
+	ret = vkms_crtc_init(dev, crtc, primary, NULL);
+	if (ret)
+		goto err_crtc;
+
+	ret = drm_connector_init(dev, connector, &vkms_connector_funcs,
+				 DRM_MODE_CONNECTOR_VIRTUAL);
+	if (ret) {
+		DRM_ERROR("Failed to init connector\n");
+		goto err_connector;
+	}
+
+	drm_connector_helper_add(connector, &vkms_conn_helper_funcs);
+
+	ret = drm_connector_register(connector);
+	if (ret) {
+		DRM_ERROR("Failed to register connector\n");
+		goto err_connector_register;
+	}
+
+	ret = drm_encoder_init(dev, encoder, &vkms_encoder_funcs,
+			       DRM_MODE_ENCODER_VIRTUAL, NULL);
+	if (ret) {
+		DRM_ERROR("Failed to init encoder\n");
+		goto err_encoder;
+	}
+	encoder->possible_crtcs = 1;
+
+	ret = drm_connector_attach_encoder(connector, encoder);
+	if (ret) {
+		DRM_ERROR("Failed to attach connector to encoder\n");
+		goto err_attach;
+	}
+
+	drm_mode_config_reset(dev);
+
+	return 0;
+
+err_attach:
+	drm_encoder_cleanup(encoder);
+
+err_encoder:
+	drm_connector_unregister(connector);
+
+err_connector_register:
+	drm_connector_cleanup(connector);
+
+err_connector:
+	drm_crtc_cleanup(crtc);
+
+err_crtc:
+	drm_plane_cleanup(primary);
+	return ret;
+}
diff --git a/drivers/gpu/drm/vkms/vkms_plane.c b/drivers/gpu/drm/vkms/vkms_plane.c
new file mode 100644
index 000000000000..9f75b1e2c1c4
--- /dev/null
+++ b/drivers/gpu/drm/vkms/vkms_plane.c
@@ -0,0 +1,57 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ */
+
+#include "vkms_drv.h"
+#include <drm/drm_plane_helper.h>
+#include <drm/drm_atomic_helper.h>
+
+static const struct drm_plane_funcs vkms_plane_funcs = {
+	.update_plane		= drm_atomic_helper_update_plane,
+	.disable_plane		= drm_atomic_helper_disable_plane,
+	.destroy		= drm_plane_cleanup,
+	.reset			= drm_atomic_helper_plane_reset,
+	.atomic_duplicate_state = drm_atomic_helper_plane_duplicate_state,
+	.atomic_destroy_state	= drm_atomic_helper_plane_destroy_state,
+};
+
+static void vkms_primary_plane_update(struct drm_plane *plane,
+				      struct drm_plane_state *old_state)
+{
+}
+
+static const struct drm_plane_helper_funcs vkms_primary_helper_funcs = {
+	.atomic_update		= vkms_primary_plane_update,
+};
+
+struct drm_plane *vkms_plane_init(struct vkms_device *vkmsdev)
+{
+	struct drm_device *dev = &vkmsdev->drm;
+	struct drm_plane *plane;
+	const u32 *formats;
+	int ret, nformats;
+
+	plane = kzalloc(sizeof(*plane), GFP_KERNEL);
+	if (!plane)
+		return ERR_PTR(-ENOMEM);
+
+	formats = vkms_formats;
+	nformats = ARRAY_SIZE(vkms_formats);
+
+	ret = drm_universal_plane_init(dev, plane, 0,
+				       &vkms_plane_funcs,
+				       formats, nformats,
+				       NULL, DRM_PLANE_TYPE_PRIMARY, NULL);
+	if (ret) {
+		kfree(plane);
+		return ERR_PTR(ret);
+	}
+
+	drm_plane_helper_add(plane, &vkms_primary_helper_funcs);
+
+	return plane;
+}
diff --git a/drivers/gpu/drm/vmwgfx/Kconfig b/drivers/gpu/drm/vmwgfx/Kconfig
index 8c308dac99c5..6b28a326f8bb 100644
--- a/drivers/gpu/drm/vmwgfx/Kconfig
+++ b/drivers/gpu/drm/vmwgfx/Kconfig
@@ -1,3 +1,4 @@
+# SPDX-License-Identifier: GPL-2.0
 config DRM_VMWGFX
 	tristate "DRM driver for VMware Virtual GPU"
 	depends on DRM && PCI && X86 && MMU
diff --git a/drivers/gpu/drm/vmwgfx/Makefile b/drivers/gpu/drm/vmwgfx/Makefile
index 794cc9d5c9b0..09b2aa08363e 100644
--- a/drivers/gpu/drm/vmwgfx/Makefile
+++ b/drivers/gpu/drm/vmwgfx/Makefile
@@ -1,9 +1,9 @@
 # SPDX-License-Identifier: GPL-2.0
 vmwgfx-y := vmwgfx_execbuf.o vmwgfx_gmr.o vmwgfx_kms.o vmwgfx_drv.o \
-	    vmwgfx_fb.o vmwgfx_ioctl.o vmwgfx_resource.o vmwgfx_buffer.o \
+	    vmwgfx_fb.o vmwgfx_ioctl.o vmwgfx_resource.o vmwgfx_ttm_buffer.o \
 	    vmwgfx_fifo.o vmwgfx_irq.o vmwgfx_ldu.o vmwgfx_ttm_glue.o \
 	    vmwgfx_overlay.o vmwgfx_marker.o vmwgfx_gmrid_manager.o \
-	    vmwgfx_fence.o vmwgfx_dmabuf.o vmwgfx_scrn.o vmwgfx_context.o \
+	    vmwgfx_fence.o vmwgfx_bo.o vmwgfx_scrn.o vmwgfx_context.o \
 	    vmwgfx_surface.o vmwgfx_prime.o vmwgfx_mob.o vmwgfx_shader.o \
 	    vmwgfx_cmdbuf_res.o vmwgfx_cmdbuf.o vmwgfx_stdu.o \
 	    vmwgfx_cotable.o vmwgfx_so.o vmwgfx_binding.o vmwgfx_msg.o \
diff --git a/drivers/gpu/drm/vmwgfx/device_include/svga3d_caps.h b/drivers/gpu/drm/vmwgfx/device_include/svga3d_caps.h
index 9ce2466a5d00..69c4253fbfbb 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/svga3d_caps.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/svga3d_caps.h
@@ -1,5 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**********************************************************
- * Copyright 2007-2015 VMware, Inc.  All rights reserved.
+ * Copyright 2007-2015 VMware, Inc.
  *
  * Permission is hereby granted, free of charge, to any person
  * obtaining a copy of this software and associated documentation
diff --git a/drivers/gpu/drm/vmwgfx/device_include/svga3d_cmd.h b/drivers/gpu/drm/vmwgfx/device_include/svga3d_cmd.h
index 2dfd57c5f463..9cbba0e8ce6a 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/svga3d_cmd.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/svga3d_cmd.h
@@ -1,5 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**********************************************************
- * Copyright 1998-2015 VMware, Inc.  All rights reserved.
+ * Copyright 1998-2015 VMware, Inc.
  *
  * Permission is hereby granted, free of charge, to any person
  * obtaining a copy of this software and associated documentation
@@ -46,10 +47,10 @@
  * the SVGA3D protocol and remain reserved; they should not be used in the
  * future.
  *
- * IDs between 1040 and 1999 (inclusive) are available for use by the
+ * IDs between 1040 and 2999 (inclusive) are available for use by the
  * current SVGA3D protocol.
  *
- * FIFO clients other than SVGA3D should stay below 1000, or at 2000
+ * FIFO clients other than SVGA3D should stay below 1000, or at 3000
  * and up.
  */
 
@@ -89,19 +90,19 @@ typedef enum {
    SVGA_3D_CMD_BLIT_SURFACE_TO_SCREEN                     = 1069,
    SVGA_3D_CMD_SURFACE_DEFINE_V2                          = 1070,
    SVGA_3D_CMD_GENERATE_MIPMAPS                           = 1071,
-   SVGA_3D_CMD_VIDEO_CREATE_DECODER                       = 1072,
-   SVGA_3D_CMD_VIDEO_DESTROY_DECODER                      = 1073,
-   SVGA_3D_CMD_VIDEO_CREATE_PROCESSOR                     = 1074,
-   SVGA_3D_CMD_VIDEO_DESTROY_PROCESSOR                    = 1075,
-   SVGA_3D_CMD_VIDEO_DECODE_START_FRAME                   = 1076,
-   SVGA_3D_CMD_VIDEO_DECODE_RENDER                        = 1077,
-   SVGA_3D_CMD_VIDEO_DECODE_END_FRAME                     = 1078,
-   SVGA_3D_CMD_VIDEO_PROCESS_FRAME                        = 1079,
+   SVGA_3D_CMD_DEAD4                                      = 1072,
+   SVGA_3D_CMD_DEAD5                                      = 1073,
+   SVGA_3D_CMD_DEAD6                                      = 1074,
+   SVGA_3D_CMD_DEAD7                                      = 1075,
+   SVGA_3D_CMD_DEAD8                                      = 1076,
+   SVGA_3D_CMD_DEAD9                                      = 1077,
+   SVGA_3D_CMD_DEAD10                                     = 1078,
+   SVGA_3D_CMD_DEAD11                                     = 1079,
    SVGA_3D_CMD_ACTIVATE_SURFACE                           = 1080,
    SVGA_3D_CMD_DEACTIVATE_SURFACE                         = 1081,
    SVGA_3D_CMD_SCREEN_DMA                                 = 1082,
-   SVGA_3D_CMD_SET_UNITY_SURFACE_COOKIE                   = 1083,
-   SVGA_3D_CMD_OPEN_CONTEXT_SURFACE                       = 1084,
+   SVGA_3D_CMD_DEAD1                                      = 1083,
+   SVGA_3D_CMD_DEAD2                                      = 1084,
 
    SVGA_3D_CMD_LOGICOPS_BITBLT                            = 1085,
    SVGA_3D_CMD_LOGICOPS_TRANSBLT                          = 1086,
@@ -217,7 +218,7 @@ typedef enum {
    SVGA_3D_CMD_DX_CLEAR_DEPTHSTENCIL_VIEW                 = 1177,
    SVGA_3D_CMD_DX_PRED_COPY_REGION                        = 1178,
    SVGA_3D_CMD_DX_PRED_COPY                               = 1179,
-   SVGA_3D_CMD_DX_STRETCHBLT                              = 1180,
+   SVGA_3D_CMD_DX_PRESENTBLT                              = 1180,
    SVGA_3D_CMD_DX_GENMIPS                                 = 1181,
    SVGA_3D_CMD_DX_UPDATE_SUBRESOURCE                      = 1182,
    SVGA_3D_CMD_DX_READBACK_SUBRESOURCE                    = 1183,
@@ -254,7 +255,7 @@ typedef enum {
    SVGA_3D_CMD_DX_READBACK_ALL_QUERY                      = 1214,
    SVGA_3D_CMD_DX_PRED_TRANSFER_FROM_BUFFER               = 1215,
    SVGA_3D_CMD_DX_MOB_FENCE_64                            = 1216,
-   SVGA_3D_CMD_DX_BIND_SHADER_ON_CONTEXT                  = 1217,
+   SVGA_3D_CMD_DX_BIND_ALL_SHADER                         = 1217,
    SVGA_3D_CMD_DX_HINT                                    = 1218,
    SVGA_3D_CMD_DX_BUFFER_UPDATE                           = 1219,
    SVGA_3D_CMD_DX_SET_VS_CONSTANT_BUFFER_OFFSET           = 1220,
@@ -262,17 +263,47 @@ typedef enum {
    SVGA_3D_CMD_DX_SET_GS_CONSTANT_BUFFER_OFFSET           = 1222,
 
    /*
-    * Reserve some IDs to be used for the DX11 shader types.
+    * Reserve some IDs to be used for the SM5 shader types.
     */
    SVGA_3D_CMD_DX_RESERVED1                               = 1223,
    SVGA_3D_CMD_DX_RESERVED2                               = 1224,
    SVGA_3D_CMD_DX_RESERVED3                               = 1225,
 
-   SVGA_3D_CMD_DX_MAX                                     = 1226,
-   SVGA_3D_CMD_MAX                                        = 1226,
+   SVGA_3D_CMD_DX_COND_BIND_ALL_SHADER                    = 1226,
+   SVGA_3D_CMD_DX_MAX                                     = 1227,
+
+   SVGA_3D_CMD_SCREEN_COPY                                = 1227,
+
+   /*
+    * Reserve some IDs to be used for video.
+    */
+   SVGA_3D_CMD_VIDEO_RESERVED1                            = 1228,
+   SVGA_3D_CMD_VIDEO_RESERVED2                            = 1229,
+   SVGA_3D_CMD_VIDEO_RESERVED3                            = 1230,
+   SVGA_3D_CMD_VIDEO_RESERVED4                            = 1231,
+   SVGA_3D_CMD_VIDEO_RESERVED5                            = 1232,
+   SVGA_3D_CMD_VIDEO_RESERVED6                            = 1233,
+   SVGA_3D_CMD_VIDEO_RESERVED7                            = 1234,
+   SVGA_3D_CMD_VIDEO_RESERVED8                            = 1235,
+
+   SVGA_3D_CMD_GROW_OTABLE                                = 1236,
+   SVGA_3D_CMD_DX_GROW_COTABLE                            = 1237,
+   SVGA_3D_CMD_INTRA_SURFACE_COPY                         = 1238,
+
+   SVGA_3D_CMD_DEFINE_GB_SURFACE_V3                       = 1239,
+
+   SVGA_3D_CMD_DX_RESOLVE_COPY                            = 1240,
+   SVGA_3D_CMD_DX_PRED_RESOLVE_COPY                       = 1241,
+   SVGA_3D_CMD_DX_PRED_CONVERT_REGION                     = 1242,
+   SVGA_3D_CMD_DX_PRED_CONVERT                            = 1243,
+   SVGA_3D_CMD_WHOLE_SURFACE_COPY                         = 1244,
+
+   SVGA_3D_CMD_MAX                                        = 1245,
    SVGA_3D_CMD_FUTURE_MAX                                 = 3000
 } SVGAFifo3dCmdId;
 
+#define SVGA_NUM_3D_CMD (SVGA_3D_CMD_MAX - SVGA_3D_CMD_BASE)
+
 /*
  * FIFO command format definitions:
  */
@@ -301,7 +332,7 @@ typedef
 #include "vmware_pack_begin.h"
 struct {
    uint32                      sid;
-   SVGA3dSurfaceFlags          surfaceFlags;
+   SVGA3dSurface1Flags         surfaceFlags;
    SVGA3dSurfaceFormat         format;
    /*
     * If surfaceFlags has SVGA3D_SURFACE_CUBEMAP bit set, all SVGA3dSurfaceFace
@@ -327,7 +358,7 @@ typedef
 #include "vmware_pack_begin.h"
 struct {
    uint32                      sid;
-   SVGA3dSurfaceFlags          surfaceFlags;
+   SVGA3dSurface1Flags         surfaceFlags;
    SVGA3dSurfaceFormat         format;
    /*
     * If surfaceFlags has SVGA3D_SURFACE_CUBEMAP bit set, all SVGA3dSurfaceFace
@@ -459,6 +490,28 @@ struct {
 #include "vmware_pack_end.h"
 SVGA3dCmdSurfaceCopy;               /* SVGA_3D_CMD_SURFACE_COPY */
 
+/*
+ * Perform a surface copy within the same image.
+ * The src/dest boxes are allowed to overlap.
+ */
+typedef
+#include "vmware_pack_begin.h"
+struct {
+   SVGA3dSurfaceImageId  surface;
+   SVGA3dCopyBox box;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdIntraSurfaceCopy;               /* SVGA_3D_CMD_INTRA_SURFACE_COPY */
+
+typedef
+#include "vmware_pack_begin.h"
+struct {
+   uint32 srcSid;
+   uint32 destSid;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdWholeSurfaceCopy;               /* SVGA_3D_CMD_WHOLE_SURFACE_COPY */
+
 typedef
 #include "vmware_pack_begin.h"
 struct {
@@ -772,6 +825,17 @@ struct {
 #include "vmware_pack_end.h"
 SVGA3dVertexElement;
 
+/*
+ * Should the vertex element respect the stream value?  The high bit of the
+ * stream should be set to indicate that the stream should be respected.  If
+ * the high bit is not set, the stream will be ignored and replaced by the index
+ * of the position of the currently considered vertex element.
+ *
+ * All guests should set this bit and correctly specify the stream going
+ * forward.
+ */
+#define SVGA3D_VERTEX_ELEMENT_RESPECT_STREAM (1 << 7)
+
 typedef
 #include "vmware_pack_begin.h"
 struct {
@@ -1102,8 +1166,6 @@ struct {
 #include "vmware_pack_end.h"
 SVGA3dCmdGenerateMipmaps;             /* SVGA_3D_CMD_GENERATE_MIPMAPS */
 
-
-
 typedef
 #include "vmware_pack_begin.h"
 struct {
@@ -1147,38 +1209,6 @@ struct SVGA3dCmdScreenDMA {
 SVGA3dCmdScreenDMA;        /* SVGA_3D_CMD_SCREEN_DMA */
 
 /*
- * Set Unity Surface Cookie
- *
- * Associates the supplied cookie with the surface id for use with
- * Unity.  This cookie is a hint from guest to host, there is no way
- * for the guest to readback the cookie and the host is free to drop
- * the cookie association at will.  The default value for the cookie
- * on all surfaces is 0.
- */
-
-typedef
-#include "vmware_pack_begin.h"
-struct SVGA3dCmdSetUnitySurfaceCookie {
-   uint32 sid;
-   uint64 cookie;
-}
-#include "vmware_pack_end.h"
-SVGA3dCmdSetUnitySurfaceCookie;   /* SVGA_3D_CMD_SET_UNITY_SURFACE_COOKIE */
-
-/*
- * Open a context-specific surface in a non-context-specific manner.
- */
-
-typedef
-#include "vmware_pack_begin.h"
-struct SVGA3dCmdOpenContextSurface {
-   uint32 sid;
-}
-#include "vmware_pack_end.h"
-SVGA3dCmdOpenContextSurface;   /* SVGA_3D_CMD_OPEN_CONTEXT_SURFACE */
-
-
-/*
  * Logic ops
  */
 
@@ -1324,7 +1354,7 @@ typedef
 #include "vmware_pack_begin.h"
 struct {
    SVGA3dSurfaceFormat format;
-   SVGA3dSurfaceFlags surfaceFlags;
+   SVGA3dSurface1Flags surface1Flags;
    uint32 numMipLevels;
    uint32 multisampleCount;
    SVGA3dTextureFilter autogenFilter;
@@ -1332,7 +1362,11 @@ struct {
    SVGAMobId mobid;
    uint32 arraySize;
    uint32 mobPitch;
-   uint32 pad[5];
+   SVGA3dSurface2Flags surface2Flags;
+   uint8 multisamplePattern;
+   uint8 qualityLevel;
+   uint8  pad0[2];
+   uint32 pad1[3];
 }
 #include "vmware_pack_end.h"
 SVGAOTableSurfaceEntry;
@@ -1360,7 +1394,8 @@ struct {
 SVGAOTableShaderEntry;
 #define SVGA3D_OTABLE_SHADER_ENTRY_SIZE (sizeof(SVGAOTableShaderEntry))
 
-#define SVGA_STFLAG_PRIMARY (1 << 0)
+#define SVGA_STFLAG_PRIMARY  (1 << 0)
+#define SVGA_STFLAG_RESERVED (1 << 1) /* Added with cap SVGA_CAP_HP_CMD_QUEUE */
 typedef uint32 SVGAScreenTargetFlags;
 
 typedef
@@ -1528,6 +1563,25 @@ struct {
 #include "vmware_pack_end.h"
 SVGA3dCmdSetOTableBase64;  /* SVGA_3D_CMD_SET_OTABLE_BASE64 */
 
+/*
+ * Guests using SVGA_3D_CMD_GROW_OTABLE are promising that
+ * the new OTable contains the same contents as the old one, except possibly
+ * for some new invalid entries at the end.
+ *
+ * (Otherwise, guests should use one of the SetOTableBase commands.)
+ */
+typedef
+#include "vmware_pack_begin.h"
+struct {
+   SVGAOTableType type;
+   PPN64 baseAddress;
+   uint32 sizeInBytes;
+   uint32 validSizeInBytes;
+   SVGAMobFormat ptDepth;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdGrowOTable;  /* SVGA_3D_CMD_GROW_OTABLE */
+
 typedef
 #include "vmware_pack_begin.h"
 struct {
@@ -1615,7 +1669,7 @@ typedef
 #include "vmware_pack_begin.h"
 struct SVGA3dCmdDefineGBSurface {
    uint32 sid;
-   SVGA3dSurfaceFlags surfaceFlags;
+   SVGA3dSurface1Flags surfaceFlags;
    SVGA3dSurfaceFormat format;
    uint32 numMipLevels;
    uint32 multisampleCount;
@@ -1626,6 +1680,45 @@ struct SVGA3dCmdDefineGBSurface {
 SVGA3dCmdDefineGBSurface;   /* SVGA_3D_CMD_DEFINE_GB_SURFACE */
 
 /*
+ * Defines a guest-backed surface, adding the arraySize field.
+ */
+typedef
+#include "vmware_pack_begin.h"
+struct SVGA3dCmdDefineGBSurface_v2 {
+   uint32 sid;
+   SVGA3dSurface1Flags surfaceFlags;
+   SVGA3dSurfaceFormat format;
+   uint32 numMipLevels;
+   uint32 multisampleCount;
+   SVGA3dTextureFilter autogenFilter;
+   SVGA3dSize size;
+   uint32 arraySize;
+   uint32 pad;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdDefineGBSurface_v2;   /* SVGA_3D_CMD_DEFINE_GB_SURFACE_V2 */
+
+/*
+ * Defines a guest-backed surface, adding the larger flags.
+ */
+typedef
+#include "vmware_pack_begin.h"
+struct SVGA3dCmdDefineGBSurface_v3 {
+   uint32 sid;
+   SVGA3dSurfaceAllFlags surfaceFlags;
+   SVGA3dSurfaceFormat format;
+   uint32 numMipLevels;
+   uint32 multisampleCount;
+   SVGA3dMSPattern multisamplePattern;
+   SVGA3dMSQualityLevel qualityLevel;
+   SVGA3dTextureFilter autogenFilter;
+   SVGA3dSize size;
+   uint32 arraySize;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdDefineGBSurface_v3;   /* SVGA_3D_CMD_DEFINE_GB_SURFACE_V3 */
+
+/*
  * Destroy a guest-backed surface.
  */
 
@@ -1672,7 +1765,7 @@ SVGA3dCmdBindGBSurfaceWithPitch;   /* SVGA_3D_CMD_BIND_GB_SURFACE_WITH_PITCH */
 
 typedef
 #include "vmware_pack_begin.h"
-struct{
+struct SVGA3dCmdCondBindGBSurface {
    uint32 sid;
    SVGAMobId testMobid;
    SVGAMobId mobid;
@@ -2066,6 +2159,26 @@ struct {
    uint32 mobOffset;
 }
 #include "vmware_pack_end.h"
-SVGA3dCmdGBMobFence;  /* SVGA_3D_CMD_GB_MOB_FENCE*/
+SVGA3dCmdGBMobFence;  /* SVGA_3D_CMD_GB_MOB_FENCE */
+
+typedef
+#include "vmware_pack_begin.h"
+struct {
+   uint32 stid;
+   SVGA3dSurfaceImageId dest;
+
+   uint32 statusMobId;
+   uint32 statusMobOffset;
+
+   /* Reserved fields */
+   uint32 mustBeInvalidId;
+   uint32 mustBeZero;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdScreenCopy;  /* SVGA_3D_CMD_SCREEN_COPY */
+
+#define SVGA_SCREEN_COPY_STATUS_FAILURE 0x00
+#define SVGA_SCREEN_COPY_STATUS_SUCCESS 0x01
+#define SVGA_SCREEN_COPY_STATUS_INVALID 0xFFFFFFFF
 
 #endif /* _SVGA3D_CMD_H_ */
diff --git a/drivers/gpu/drm/vmwgfx/device_include/svga3d_devcaps.h b/drivers/gpu/drm/vmwgfx/device_include/svga3d_devcaps.h
index c18b663f360f..f256560049bf 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/svga3d_devcaps.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/svga3d_devcaps.h
@@ -1,5 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**********************************************************
- * Copyright 1998-2015 VMware, Inc.  All rights reserved.
+ * Copyright 1998-2015 VMware, Inc.
  *
  * Permission is hereby granted, free of charge, to any person
  * obtaining a copy of this software and associated documentation
@@ -229,9 +230,9 @@ typedef enum {
    SVGA3D_DEVCAP_DEAD2                             = 94,
 
    /*
-    * Does the device support the DX commands?
+    * Does the device support DXContexts?
     */
-   SVGA3D_DEVCAP_DX                                = 95,
+   SVGA3D_DEVCAP_DXCONTEXT                         = 95,
 
    /*
     * What is the maximum size of a texture array?
@@ -241,21 +242,47 @@ typedef enum {
    SVGA3D_DEVCAP_MAX_TEXTURE_ARRAY_SIZE            = 96,
 
    /*
-    * What is the maximum number of vertex buffers that can
-    * be used in the DXContext inputAssembly?
+    * What is the maximum number of vertex buffers or vertex input registers
+    * that can be expected to work correctly with a DXContext?
+    *
+    * The guest is allowed to set up to SVGA3D_DX_MAX_VERTEXBUFFERS, but
+    * anything in excess of this cap is not guaranteed to render correctly.
+    *
+    * Similarly, the guest can set up to SVGA3D_DX_MAX_VERTEXINPUTREGISTERS
+    * input registers without the SVGA3D_DEVCAP_SM4_1 cap, or
+    * SVGA3D_DX_SM41_MAX_VERTEXINPUTREGISTERS with the SVGA3D_DEVCAP_SM4_1,
+    * but only the registers up to this cap value are guaranteed to render
+    * correctly.
+    *
+    * If guest-drivers are able to expose a lower-limit, it's recommended
+    * that they clamp to this value.  Otherwise, the host will make a
+    * best-effort on case-by-case basis if guests exceed this.
     */
    SVGA3D_DEVCAP_DX_MAX_VERTEXBUFFERS              = 97,
 
    /*
-    * What is the maximum number of constant buffers
-    * that can be expected to work correctly with a
-    * DX context?
+    * What is the maximum number of constant buffers that can be expected to
+    * work correctly with a DX context?
+    *
+    * The guest is allowed to set up to SVGA3D_DX_MAX_CONSTBUFFERS, but
+    * anything in excess of this cap is not guaranteed to render correctly.
+    *
+    * If guest-drivers are able to expose a lower-limit, it's recommended
+    * that they clamp to this value.  Otherwise, the host will make a
+    * best-effort on case-by-case basis if guests exceed this.
     */
    SVGA3D_DEVCAP_DX_MAX_CONSTANT_BUFFERS           = 98,
 
    /*
     * Does the device support provoking vertex control?
-    * If zero, the first vertex will always be the provoking vertex.
+    *
+    * If this cap is present, the provokingVertexLast field in the
+    * rasterizer state is enabled.  (Guests can then set it to FALSE,
+    * meaning that the first vertex is the provoking vertex, or TRUE,
+    * meaning that the last verteix is the provoking vertex.)
+    *
+    * If this cap is FALSE, then guests should set the provokingVertexLast
+    * to FALSE, otherwise rendering behavior is undefined.
     */
    SVGA3D_DEVCAP_DX_PROVOKING_VERTEX               = 99,
 
@@ -281,7 +308,7 @@ typedef enum {
    SVGA3D_DEVCAP_DXFMT_BUMPU8V8                    = 119,
    SVGA3D_DEVCAP_DXFMT_BUMPL6V5U5                  = 120,
    SVGA3D_DEVCAP_DXFMT_BUMPX8L8V8U8                = 121,
-   SVGA3D_DEVCAP_DXFMT_BUMPL8V8U8                  = 122,
+   SVGA3D_DEVCAP_DXFMT_FORMAT_DEAD1                = 122,
    SVGA3D_DEVCAP_DXFMT_ARGB_S10E5                  = 123,
    SVGA3D_DEVCAP_DXFMT_ARGB_S23E8                  = 124,
    SVGA3D_DEVCAP_DXFMT_A2R10G10B10                 = 125,
@@ -320,8 +347,8 @@ typedef enum {
    SVGA3D_DEVCAP_DXFMT_R32G32_SINT                 = 158,
    SVGA3D_DEVCAP_DXFMT_R32G8X24_TYPELESS           = 159,
    SVGA3D_DEVCAP_DXFMT_D32_FLOAT_S8X24_UINT        = 160,
-   SVGA3D_DEVCAP_DXFMT_R32_FLOAT_X8X24_TYPELESS    = 161,
-   SVGA3D_DEVCAP_DXFMT_X32_TYPELESS_G8X24_UINT     = 162,
+   SVGA3D_DEVCAP_DXFMT_R32_FLOAT_X8X24             = 161,
+   SVGA3D_DEVCAP_DXFMT_X32_G8X24_UINT              = 162,
    SVGA3D_DEVCAP_DXFMT_R10G10B10A2_TYPELESS        = 163,
    SVGA3D_DEVCAP_DXFMT_R10G10B10A2_UINT            = 164,
    SVGA3D_DEVCAP_DXFMT_R11G11B10_FLOAT             = 165,
@@ -339,8 +366,8 @@ typedef enum {
    SVGA3D_DEVCAP_DXFMT_R32_SINT                    = 177,
    SVGA3D_DEVCAP_DXFMT_R24G8_TYPELESS              = 178,
    SVGA3D_DEVCAP_DXFMT_D24_UNORM_S8_UINT           = 179,
-   SVGA3D_DEVCAP_DXFMT_R24_UNORM_X8_TYPELESS       = 180,
-   SVGA3D_DEVCAP_DXFMT_X24_TYPELESS_G8_UINT        = 181,
+   SVGA3D_DEVCAP_DXFMT_R24_UNORM_X8                = 180,
+   SVGA3D_DEVCAP_DXFMT_X24_G8_UINT                 = 181,
    SVGA3D_DEVCAP_DXFMT_R8G8_TYPELESS               = 182,
    SVGA3D_DEVCAP_DXFMT_R8G8_UNORM                  = 183,
    SVGA3D_DEVCAP_DXFMT_R8G8_UINT                   = 184,
@@ -404,6 +431,17 @@ typedef enum {
    SVGA3D_DEVCAP_DXFMT_BC4_UNORM                   = 242,
    SVGA3D_DEVCAP_DXFMT_BC5_UNORM                   = 243,
 
+   /*
+    * Advertises shaderModel 4.1 support, independent blend-states,
+    * cube-map arrays, and a higher vertex input registers limit.
+    *
+    * (See documentation on SVGA3D_DEVCAP_DX_MAX_VERTEXBUFFERS.)
+    */
+   SVGA3D_DEVCAP_SM41                              = 244,
+
+   SVGA3D_DEVCAP_MULTISAMPLE_2X                    = 245,
+   SVGA3D_DEVCAP_MULTISAMPLE_4X                    = 246,
+
    SVGA3D_DEVCAP_MAX                       /* This must be the last index. */
 } SVGA3dDevCapIndex;
 
@@ -419,9 +457,7 @@ typedef enum {
  * MIPS: Does the format support mip levels?
  * ARRAY: Does the format support texture arrays?
  * VOLUME: Does the format support having volume?
- * MULTISAMPLE_2: Does the format support 2x multisample?
- * MULTISAMPLE_4: Does the format support 4x multisample?
- * MULTISAMPLE_8: Does the format support 8x multisample?
+ * MULTISAMPLE: Does the format support multisample?
  */
 #define SVGA3D_DXFMT_SUPPORTED                (1 <<  0)
 #define SVGA3D_DXFMT_SHADER_SAMPLE            (1 <<  1)
@@ -432,20 +468,8 @@ typedef enum {
 #define SVGA3D_DXFMT_ARRAY                    (1 <<  6)
 #define SVGA3D_DXFMT_VOLUME                   (1 <<  7)
 #define SVGA3D_DXFMT_DX_VERTEX_BUFFER         (1 <<  8)
-#define SVGADX_DXFMT_MULTISAMPLE_2            (1 <<  9)
-#define SVGADX_DXFMT_MULTISAMPLE_4            (1 << 10)
-#define SVGADX_DXFMT_MULTISAMPLE_8            (1 << 11)
-#define SVGADX_DXFMT_MAX                      (1 << 12)
-
-/*
- * Convenience mask for any multisample capability.
- *
- * The multisample bits imply both load and render capability.
- */
-#define SVGA3D_DXFMT_MULTISAMPLE ( \
-           SVGADX_DXFMT_MULTISAMPLE_2 | \
-           SVGADX_DXFMT_MULTISAMPLE_4 | \
-           SVGADX_DXFMT_MULTISAMPLE_8 )
+#define SVGA3D_DXFMT_MULTISAMPLE              (1 <<  9)
+#define SVGA3D_DXFMT_MAX                      (1 << 10)
 
 typedef union {
    Bool   b;
diff --git a/drivers/gpu/drm/vmwgfx/device_include/svga3d_dx.h b/drivers/gpu/drm/vmwgfx/device_include/svga3d_dx.h
index 8c5ae608cfb4..7a49c94df221 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/svga3d_dx.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/svga3d_dx.h
@@ -1,5 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**********************************************************
- * Copyright 2012-2015 VMware, Inc.  All rights reserved.
+ * Copyright 2012-2015 VMware, Inc.
  *
  * Permission is hereby granted, free of charge, to any person
  * obtaining a copy of this software and associated documentation
@@ -56,6 +57,16 @@ typedef uint32 SVGA3dInputClassification;
 #define SVGA3D_RESOURCE_TYPE_MAX      7
 typedef uint32 SVGA3dResourceType;
 
+#define SVGA3D_COLOR_WRITE_ENABLE_RED     (1 << 0)
+#define SVGA3D_COLOR_WRITE_ENABLE_GREEN   (1 << 1)
+#define SVGA3D_COLOR_WRITE_ENABLE_BLUE    (1 << 2)
+#define SVGA3D_COLOR_WRITE_ENABLE_ALPHA   (1 << 3)
+#define SVGA3D_COLOR_WRITE_ENABLE_ALL     (SVGA3D_COLOR_WRITE_ENABLE_RED |   \
+                                           SVGA3D_COLOR_WRITE_ENABLE_GREEN | \
+                                           SVGA3D_COLOR_WRITE_ENABLE_BLUE |  \
+                                           SVGA3D_COLOR_WRITE_ENABLE_ALPHA)
+typedef uint8 SVGA3dColorWriteEnable;
+
 #define SVGA3D_DEPTH_WRITE_MASK_ZERO   0
 #define SVGA3D_DEPTH_WRITE_MASK_ALL    1
 typedef uint8 SVGA3dDepthWriteMask;
@@ -88,17 +99,28 @@ typedef uint8 SVGA3dCullMode;
 #define SVGA3D_COMPARISON_MAX             9
 typedef uint8 SVGA3dComparisonFunc;
 
+/*
+ * SVGA3D_MULTISAMPLE_RAST_DISABLE disables MSAA for all primitives.
+ * SVGA3D_MULTISAMPLE_RAST_DISABLE_LINE, which is supported in SM41,
+ * disables MSAA for lines only.
+ */
+#define SVGA3D_MULTISAMPLE_RAST_DISABLE        0
+#define SVGA3D_MULTISAMPLE_RAST_ENABLE         1
+#define SVGA3D_MULTISAMPLE_RAST_DX_MAX         1
+#define SVGA3D_MULTISAMPLE_RAST_DISABLE_LINE   2
+#define SVGA3D_MULTISAMPLE_RAST_MAX            2
+typedef uint8 SVGA3dMultisampleRastEnable;
+
 #define SVGA3D_DX_MAX_VERTEXBUFFERS 32
+#define SVGA3D_DX_MAX_VERTEXINPUTREGISTERS 16
+#define SVGA3D_DX_SM41_MAX_VERTEXINPUTREGISTERS 32
 #define SVGA3D_DX_MAX_SOTARGETS 4
 #define SVGA3D_DX_MAX_SRVIEWS 128
 #define SVGA3D_DX_MAX_CONSTBUFFERS 16
 #define SVGA3D_DX_MAX_SAMPLERS 16
 
-/* Id limits */
-static const uint32 SVGA3dBlendObjectCountPerContext = 4096;
-static const uint32 SVGA3dDepthStencilObjectCountPerContext = 4096;
+#define SVGA3D_DX_MAX_CONSTBUF_BINDING_SIZE (4096 * 4 * (uint32)sizeof(uint32))
 
-typedef uint32 SVGA3dSurfaceId;
 typedef uint32 SVGA3dShaderResourceViewId;
 typedef uint32 SVGA3dRenderTargetViewId;
 typedef uint32 SVGA3dDepthStencilViewId;
@@ -194,20 +216,6 @@ SVGA3dCmdDXInvalidateContext;   /* SVGA_3D_CMD_DX_INVALIDATE_CONTEXT */
 
 typedef
 #include "vmware_pack_begin.h"
-struct SVGA3dReplyFormatData {
-   uint32 formatSupport;
-   uint32 msaa2xQualityLevels:5;
-   uint32 msaa4xQualityLevels:5;
-   uint32 msaa8xQualityLevels:5;
-   uint32 msaa16xQualityLevels:5;
-   uint32 msaa32xQualityLevels:5;
-   uint32 pad:7;
-}
-#include "vmware_pack_end.h"
-SVGA3dReplyFormatData;
-
-typedef
-#include "vmware_pack_begin.h"
 struct SVGA3dCmdDXSetSingleConstantBuffer {
    uint32 slot;
    SVGA3dShaderType type;
@@ -624,6 +632,28 @@ SVGA3dCmdDXPredCopy; /* SVGA_3D_CMD_DX_PRED_COPY */
 
 typedef
 #include "vmware_pack_begin.h"
+struct SVGA3dCmdDXPredConvertRegion {
+   SVGA3dSurfaceId dstSid;
+   uint32 dstSubResource;
+   SVGA3dBox destBox;
+   SVGA3dSurfaceId srcSid;
+   uint32 srcSubResource;
+   SVGA3dBox srcBox;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdDXPredConvertRegion; /* SVGA_3D_CMD_DX_PRED_CONVERT_REGION */
+
+typedef
+#include "vmware_pack_begin.h"
+struct SVGA3dCmdDXPredConvert {
+   SVGA3dSurfaceId dstSid;
+   SVGA3dSurfaceId srcSid;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdDXPredConvert; /* SVGA_3D_CMD_DX_PRED_CONVERT */
+
+typedef
+#include "vmware_pack_begin.h"
 struct SVGA3dCmdDXBufferCopy {
    SVGA3dSurfaceId dest;
    SVGA3dSurfaceId src;
@@ -635,23 +665,57 @@ struct SVGA3dCmdDXBufferCopy {
 SVGA3dCmdDXBufferCopy;
 /* SVGA_3D_CMD_DX_BUFFER_COPY */
 
-typedef uint32 SVGA3dDXStretchBltMode;
-#define SVGADX_STRETCHBLT_LINEAR         (1 << 0)
-#define SVGADX_STRETCHBLT_FORCE_SRC_SRGB (1 << 1)
+/*
+ * Perform a surface copy between a multisample, and a non-multisampled
+ * surface.
+ */
+typedef
+#include "vmware_pack_begin.h"
+struct {
+   SVGA3dSurfaceId dstSid;
+   uint32 dstSubResource;
+   SVGA3dSurfaceId srcSid;
+   uint32 srcSubResource;
+   SVGA3dSurfaceFormat copyFormat;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdDXResolveCopy;               /* SVGA_3D_CMD_DX_RESOLVE_COPY */
+
+/*
+ * Perform a predicated surface copy between a multisample, and a
+ * non-multisampled surface.
+ */
+typedef
+#include "vmware_pack_begin.h"
+struct {
+   SVGA3dSurfaceId dstSid;
+   uint32 dstSubResource;
+   SVGA3dSurfaceId srcSid;
+   uint32 srcSubResource;
+   SVGA3dSurfaceFormat copyFormat;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdDXPredResolveCopy;           /* SVGA_3D_CMD_DX_PRED_RESOLVE_COPY */
+
+typedef uint32 SVGA3dDXPresentBltMode;
+#define SVGADX_PRESENTBLT_LINEAR           (1 << 0)
+#define SVGADX_PRESENTBLT_FORCE_SRC_SRGB   (1 << 1)
+#define SVGADX_PRESENTBLT_FORCE_SRC_XRBIAS (1 << 2)
+#define SVGADX_PRESENTBLT_MODE_MAX         (1 << 3)
 
 typedef
 #include "vmware_pack_begin.h"
-struct SVGA3dCmdDXStretchBlt {
+struct SVGA3dCmdDXPresentBlt {
    SVGA3dSurfaceId srcSid;
    uint32 srcSubResource;
    SVGA3dSurfaceId dstSid;
    uint32 destSubResource;
    SVGA3dBox boxSrc;
    SVGA3dBox boxDest;
-   SVGA3dDXStretchBltMode mode;
+   SVGA3dDXPresentBltMode mode;
 }
 #include "vmware_pack_end.h"
-SVGA3dCmdDXStretchBlt; /* SVGA_3D_CMD_DX_STRETCHBLT */
+SVGA3dCmdDXPresentBlt; /* SVGA_3D_CMD_DX_PRESENTBLT*/
 
 typedef
 #include "vmware_pack_begin.h"
@@ -662,26 +726,6 @@ struct SVGA3dCmdDXGenMips {
 SVGA3dCmdDXGenMips; /* SVGA_3D_CMD_DX_GENMIPS */
 
 /*
- * Defines a resource/DX surface.  Resources share the surfaceId namespace.
- *
- */
-typedef
-#include "vmware_pack_begin.h"
-struct SVGA3dCmdDefineGBSurface_v2 {
-   uint32 sid;
-   SVGA3dSurfaceFlags surfaceFlags;
-   SVGA3dSurfaceFormat format;
-   uint32 numMipLevels;
-   uint32 multisampleCount;
-   SVGA3dTextureFilter autogenFilter;
-   SVGA3dSize size;
-   uint32 arraySize;
-   uint32 pad;
-}
-#include "vmware_pack_end.h"
-SVGA3dCmdDefineGBSurface_v2;   /* SVGA_3D_CMD_DEFINE_GB_SURFACE_V2 */
-
-/*
  * Update a sub-resource in a guest-backed resource.
  * (Inform the device that the guest-contents have been updated.)
  */
@@ -724,7 +768,8 @@ SVGA3dCmdDXInvalidateSubResource;   /* SVGA_3D_CMD_DX_INVALIDATE_SUBRESOURCE */
 
 /*
  * Raw byte wise transfer from a buffer surface into another surface
- * of the requested box.
+ * of the requested box.  Supported if 3d is enabled and SVGA_CAP_DX
+ * is set.  This command does not take a context.
  */
 typedef
 #include "vmware_pack_begin.h"
@@ -773,6 +818,93 @@ struct SVGA3dCmdDXSurfaceCopyAndReadback {
 SVGA3dCmdDXSurfaceCopyAndReadback;
 /* SVGA_3D_CMD_DX_SURFACE_COPY_AND_READBACK */
 
+/*
+ * SVGA_DX_HINT_NONE: Does nothing.
+ *
+ * SVGA_DX_HINT_PREFETCH_OBJECT:
+ * SVGA_DX_HINT_PREEVICT_OBJECT:
+ *      Consumes a SVGAObjectRef, and hints that the host should consider
+ *      fetching/evicting the specified object.
+ *
+ *      An id of SVGA3D_INVALID_ID can be used if the guest isn't sure
+ *      what object was affected.  (For instance, if the guest knows that
+ *      it is about to evict a DXShader, but doesn't know precisely which one,
+ *      the device can still use this to help limit it's search, or track
+ *      how many page-outs have happened.)
+ *
+ * SVGA_DX_HINT_PREFETCH_COBJECT:
+ * SVGA_DX_HINT_PREEVICT_COBJECT:
+ *      Same as the above, except they consume an SVGACObjectRef.
+ */
+typedef uint32 SVGADXHintId;
+#define SVGA_DX_HINT_NONE              0
+#define SVGA_DX_HINT_PREFETCH_OBJECT   1
+#define SVGA_DX_HINT_PREEVICT_OBJECT   2
+#define SVGA_DX_HINT_PREFETCH_COBJECT  3
+#define SVGA_DX_HINT_PREEVICT_COBJECT  4
+#define SVGA_DX_HINT_MAX               5
+
+typedef
+#include "vmware_pack_begin.h"
+struct SVGAObjectRef {
+   SVGAOTableType type;
+   uint32 id;
+}
+#include "vmware_pack_end.h"
+SVGAObjectRef;
+
+typedef
+#include "vmware_pack_begin.h"
+struct SVGACObjectRef {
+   SVGACOTableType type;
+   uint32 cid;
+   uint32 id;
+}
+#include "vmware_pack_end.h"
+SVGACObjectRef;
+
+typedef
+#include "vmware_pack_begin.h"
+struct SVGA3dCmdDXHint {
+   SVGADXHintId hintId;
+
+   /*
+    * Followed by variable sized data depending on the hintId.
+    */
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdDXHint;
+/* SVGA_3D_CMD_DX_HINT */
+
+typedef
+#include "vmware_pack_begin.h"
+struct SVGA3dCmdDXBufferUpdate {
+   SVGA3dSurfaceId sid;
+   uint32 x;
+   uint32 width;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdDXBufferUpdate;
+/* SVGA_3D_CMD_DX_BUFFER_UPDATE */
+
+typedef
+#include "vmware_pack_begin.h"
+struct SVGA3dCmdDXSetConstantBufferOffset {
+   uint32 slot;
+   uint32 offsetInBytes;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdDXSetConstantBufferOffset;
+
+typedef SVGA3dCmdDXSetConstantBufferOffset SVGA3dCmdDXSetVSConstantBufferOffset;
+/* SVGA_3D_CMD_DX_SET_VS_CONSTANT_BUFFER_OFFSET */
+
+typedef SVGA3dCmdDXSetConstantBufferOffset SVGA3dCmdDXSetPSConstantBufferOffset;
+/* SVGA_3D_CMD_DX_SET_PS_CONSTANT_BUFFER_OFFSET */
+
+typedef SVGA3dCmdDXSetConstantBufferOffset SVGA3dCmdDXSetGSConstantBufferOffset;
+/* SVGA_3D_CMD_DX_SET_GS_CONSTANT_BUFFER_OFFSET */
+
 
 typedef
 #include "vmware_pack_begin.h"
@@ -789,7 +921,7 @@ struct {
          uint32 firstArraySlice;
          uint32 mipLevels;
          uint32 arraySize;
-      } tex;
+      } tex; /* 1d, 2d, 3d, cube */
       struct {
          uint32 firstElement;
          uint32 numElements;
@@ -844,6 +976,7 @@ struct SVGA3dRenderTargetViewDesc {
       struct {
          uint32 firstElement;
          uint32 numElements;
+         uint32 padding0;
       } buffer;
       struct {
          uint32 mipSlice;
@@ -964,9 +1097,6 @@ SVGA3dInputElementDesc;
 typedef
 #include "vmware_pack_begin.h"
 struct {
-   /*
-    * XXX: How many of these can there be?
-    */
    uint32 elid;
    uint32 numDescs;
    SVGA3dInputElementDesc desc[32];
@@ -1007,7 +1137,7 @@ struct SVGA3dDXBlendStatePerRT {
       uint8 srcBlendAlpha;
       uint8 destBlendAlpha;
       uint8 blendOpAlpha;
-      uint8 renderTargetWriteMask;
+      SVGA3dColorWriteEnable renderTargetWriteMask;
       uint8 logicOpEnable;
       uint8 logicOp;
       uint16 pad0;
@@ -1125,7 +1255,7 @@ struct {
    float slopeScaledDepthBias;
    uint8 depthClipEnable;
    uint8 scissorEnable;
-   uint8 multisampleEnable;
+   SVGA3dMultisampleRastEnable multisampleEnable;
    uint8 antialiasedLineEnable;
    float lineWidth;
    uint8 lineStippleEnable;
@@ -1152,7 +1282,7 @@ struct SVGA3dCmdDXDefineRasterizerState {
    float slopeScaledDepthBias;
    uint8 depthClipEnable;
    uint8 scissorEnable;
-   uint8 multisampleEnable;
+   SVGA3dMultisampleRastEnable multisampleEnable;
    uint8 antialiasedLineEnable;
    float lineWidth;
    uint8 lineStippleEnable;
@@ -1222,21 +1352,6 @@ struct SVGA3dCmdDXDestroySamplerState {
 #include "vmware_pack_end.h"
 SVGA3dCmdDXDestroySamplerState; /* SVGA_3D_CMD_DX_DESTROY_SAMPLER_STATE */
 
-/*
- */
-typedef
-#include "vmware_pack_begin.h"
-struct SVGA3dSignatureEntry {
-   uint8 systemValue;
-   uint8 reg;                 /* register is a reserved word */
-   uint16 mask;
-   uint8 registerComponentType;
-   uint8 minPrecision;
-   uint16 pad0;
-}
-#include "vmware_pack_end.h"
-SVGA3dSignatureEntry;
-
 typedef
 #include "vmware_pack_begin.h"
 struct SVGA3dCmdDXDefineShader {
@@ -1254,12 +1369,7 @@ struct SVGACOTableDXShaderEntry {
    uint32 sizeInBytes;
    uint32 offsetInBytes;
    SVGAMobId mobid;
-   uint32 numInputSignatureEntries;
-   uint32 numOutputSignatureEntries;
-
-   uint32 numPatchConstantSignatureEntries;
-
-   uint32 pad;
+   uint32 pad[4];
 }
 #include "vmware_pack_end.h"
 SVGACOTableDXShaderEntry;
@@ -1283,6 +1393,25 @@ struct SVGA3dCmdDXBindShader {
 #include "vmware_pack_end.h"
 SVGA3dCmdDXBindShader;   /* SVGA_3D_CMD_DX_BIND_SHADER */
 
+typedef
+#include "vmware_pack_begin.h"
+struct SVGA3dCmdDXBindAllShader {
+   uint32 cid;
+   SVGAMobId mobid;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdDXBindAllShader;   /* SVGA_3D_CMD_DX_BIND_ALL_SHADER */
+
+typedef
+#include "vmware_pack_begin.h"
+struct SVGA3dCmdDXCondBindAllShader {
+   uint32 cid;
+   SVGAMobId testMobid;
+   SVGAMobId mobid;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdDXCondBindAllShader;   /* SVGA_3D_CMD_DX_COND_BIND_ALL_SHADER */
+
 /*
  * The maximum number of streamout decl's in each streamout entry.
  */
@@ -1356,7 +1485,6 @@ SVGA3dCmdDXMobFence64;  /* SVGA_3D_CMD_DX_MOB_FENCE_64 */
  *
  * This command allows the guest to bind a mob to a context-object table.
  */
-
 typedef
 #include "vmware_pack_begin.h"
 struct SVGA3dCmdDXSetCOTable {
@@ -1368,6 +1496,26 @@ struct SVGA3dCmdDXSetCOTable {
 #include "vmware_pack_end.h"
 SVGA3dCmdDXSetCOTable; /* SVGA_3D_CMD_DX_SET_COTABLE */
 
+/*
+ * Guests using SVGA_3D_CMD_DX_GROW_COTABLE are promising that
+ * the new COTable contains the same contents as the old one, except possibly
+ * for some new invalid entries at the end.
+ *
+ * If there is an old cotable mob bound, it also has to still be valid.
+ *
+ * (Otherwise, guests should use the DXSetCOTableBase command.)
+ */
+typedef
+#include "vmware_pack_begin.h"
+struct SVGA3dCmdDXGrowCOTable {
+   uint32 cid;
+   uint32 mobid;
+   SVGACOTableType type;
+   uint32 validSizeInBytes;
+}
+#include "vmware_pack_end.h"
+SVGA3dCmdDXGrowCOTable; /* SVGA_3D_CMD_DX_GROW_COTABLE */
+
 typedef
 #include "vmware_pack_begin.h"
 struct SVGA3dCmdDXReadbackCOTable {
@@ -1471,7 +1619,7 @@ struct SVGADXContextMobFormat {
    SVGA3dQueryId queryID[SVGA3D_MAX_QUERY];
 
    SVGA3dCOTableData cotables[SVGA_COTABLE_MAX];
-   uint32 pad7[381];
+   uint32 pad7[380];
 }
 #include "vmware_pack_end.h"
 SVGADXContextMobFormat;
diff --git a/drivers/gpu/drm/vmwgfx/device_include/svga3d_limits.h b/drivers/gpu/drm/vmwgfx/device_include/svga3d_limits.h
index a1c36877ad55..b22a67f15660 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/svga3d_limits.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/svga3d_limits.h
@@ -1,5 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**********************************************************
- * Copyright 2007-2015 VMware, Inc.  All rights reserved.
+ * Copyright 2007-2015 VMware, Inc.
  *
  * Permission is hereby granted, free of charge, to any person
  * obtaining a copy of this software and associated documentation
@@ -62,7 +63,9 @@
  * Maximum size in dwords of shader text the SVGA device will allow.
  * Currently 8 MB.
  */
-#define SVGA3D_MAX_SHADER_MEMORY  (8 * 1024 * 1024 / sizeof(uint32))
+#define SVGA3D_MAX_SHADER_MEMORY_BYTES (8 * 1024 * 1024)
+#define SVGA3D_MAX_SHADER_MEMORY  (SVGA3D_MAX_SHADER_MEMORY_BYTES / \
+                                   sizeof(uint32))
 
 #define SVGA3D_MAX_CLIP_PLANES    6
 
diff --git a/drivers/gpu/drm/vmwgfx/device_include/svga3d_reg.h b/drivers/gpu/drm/vmwgfx/device_include/svga3d_reg.h
index b44ce648f592..bdfc404c91e3 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/svga3d_reg.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/svga3d_reg.h
@@ -1,5 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**********************************************************
- * Copyright 1998-2015 VMware, Inc.  All rights reserved.
+ * Copyright 1998-2015 VMware, Inc.
  *
  * Permission is hereby granted, free of charge, to any person
  * obtaining a copy of this software and associated documentation
diff --git a/drivers/gpu/drm/vmwgfx/device_include/svga3d_surfacedefs.h b/drivers/gpu/drm/vmwgfx/device_include/svga3d_surfacedefs.h
index babe7cb84fc2..f2bfd3d80598 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/svga3d_surfacedefs.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/svga3d_surfacedefs.h
@@ -1,7 +1,7 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
- * Copyright © 2008-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2008-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -25,189 +25,355 @@
  *
  **************************************************************************/
 
-#include <linux/kernel.h>
-
-#ifdef __KERNEL__
-
-#include <drm/vmwgfx_drm.h>
-#define surf_size_struct struct drm_vmw_size
-
-#else /* __KERNEL__ */
+/*
+ * svga3d_surfacedefs.h --
+ *
+ *      Surface definitions and inlineable utilities for SVGA3d.
+ */
 
-#ifndef ARRAY_SIZE
-#define ARRAY_SIZE(_A) (sizeof(_A) / sizeof((_A)[0]))
-#endif /* ARRAY_SIZE */
+#ifndef _SVGA3D_SURFACEDEFS_H_
+#define _SVGA3D_SURFACEDEFS_H_
 
-#define max_t(type, x, y)  ((x) > (y) ? (x) : (y))
-#define surf_size_struct SVGA3dSize
-#define u32 uint32
+#define INCLUDE_ALLOW_USERLEVEL
+#define INCLUDE_ALLOW_MODULE
+#include "includeCheck.h"
 
-#endif /* __KERNEL__ */
+#include <linux/kernel.h>
+#include <drm/vmwgfx_drm.h>
 
 #include "svga3d_reg.h"
 
+#define surf_size_struct struct drm_vmw_size
+
 /*
- * enum svga3d_block_desc describes the active data channels in a block.
- *
- * There can be at-most four active channels in a block:
- *    1. Red, bump W, luminance and depth are stored in the first channel.
- *    2. Green, bump V and stencil are stored in the second channel.
- *    3. Blue and bump U are stored in the third channel.
- *    4. Alpha and bump Q are stored in the fourth channel.
- *
- * Block channels can be used to store compressed and buffer data:
- *    1. For compressed formats, only the data channel is used and its size
- *       is equal to that of a singular block in the compression scheme.
- *    2. For buffer formats, only the data channel is used and its size is
- *       exactly one byte in length.
- *    3. In each case the bit depth represent the size of a singular block.
- *
- * Note: Compressed and IEEE formats do not use the bitMask structure.
+ * enum svga3d_block_desc - describes generic properties about formats.
  */
-
 enum svga3d_block_desc {
-	SVGA3DBLOCKDESC_NONE        = 0,         /* No channels are active */
-	SVGA3DBLOCKDESC_BLUE        = 1 << 0,    /* Block with red channel
-						    data */
-	SVGA3DBLOCKDESC_U           = 1 << 0,    /* Block with bump U channel
-						    data */
-	SVGA3DBLOCKDESC_UV_VIDEO    = 1 << 7,    /* Block with alternating video
-						    U and V */
-	SVGA3DBLOCKDESC_GREEN       = 1 << 1,    /* Block with green channel
-						    data */
-	SVGA3DBLOCKDESC_V           = 1 << 1,    /* Block with bump V channel
-						    data */
-	SVGA3DBLOCKDESC_STENCIL     = 1 << 1,    /* Block with a stencil
-						    channel */
-	SVGA3DBLOCKDESC_RED         = 1 << 2,    /* Block with blue channel
-						    data */
-	SVGA3DBLOCKDESC_W           = 1 << 2,    /* Block with bump W channel
-						    data */
-	SVGA3DBLOCKDESC_LUMINANCE   = 1 << 2,    /* Block with luminance channel
-						    data */
-	SVGA3DBLOCKDESC_Y           = 1 << 2,    /* Block with video luminance
-						    data */
-	SVGA3DBLOCKDESC_DEPTH       = 1 << 2,    /* Block with depth channel */
-	SVGA3DBLOCKDESC_ALPHA       = 1 << 3,    /* Block with an alpha
-						    channel */
-	SVGA3DBLOCKDESC_Q           = 1 << 3,    /* Block with bump Q channel
-						    data */
-	SVGA3DBLOCKDESC_BUFFER      = 1 << 4,    /* Block stores 1 byte of
-						    data */
-	SVGA3DBLOCKDESC_COMPRESSED  = 1 << 5,    /* Block stores n bytes of
-						    data depending on the
-						    compression method used */
-	SVGA3DBLOCKDESC_IEEE_FP     = 1 << 6,    /* Block stores data in an IEEE
-						    floating point
-						    representation in
-						    all channels */
-	SVGA3DBLOCKDESC_PLANAR_YUV  = 1 << 8,    /* Three separate blocks store
-						    data. */
-	SVGA3DBLOCKDESC_U_VIDEO     = 1 << 9,    /* Block with U video data */
-	SVGA3DBLOCKDESC_V_VIDEO     = 1 << 10,   /* Block with V video data */
-	SVGA3DBLOCKDESC_EXP         = 1 << 11,   /* Shared exponent */
-	SVGA3DBLOCKDESC_SRGB        = 1 << 12,   /* Data is in sRGB format */
-	SVGA3DBLOCKDESC_2PLANAR_YUV = 1 << 13,   /* 2 planes of Y, UV,
-						    e.g., NV12. */
-	SVGA3DBLOCKDESC_3PLANAR_YUV = 1 << 14,   /* 3 planes of separate
-						    Y, U, V, e.g., YV12. */
-
-	SVGA3DBLOCKDESC_RG         = SVGA3DBLOCKDESC_RED |
-	SVGA3DBLOCKDESC_GREEN,
-	SVGA3DBLOCKDESC_RGB        = SVGA3DBLOCKDESC_RG |
-	SVGA3DBLOCKDESC_BLUE,
-	SVGA3DBLOCKDESC_RGB_SRGB   = SVGA3DBLOCKDESC_RGB |
-	SVGA3DBLOCKDESC_SRGB,
-	SVGA3DBLOCKDESC_RGBA       = SVGA3DBLOCKDESC_RGB |
-	SVGA3DBLOCKDESC_ALPHA,
-	SVGA3DBLOCKDESC_RGBA_SRGB  = SVGA3DBLOCKDESC_RGBA |
-	SVGA3DBLOCKDESC_SRGB,
+	/* Nothing special can be said about this format. */
+	SVGA3DBLOCKDESC_NONE        = 0,
+
+	/* Format contains Blue/U data */
+	SVGA3DBLOCKDESC_BLUE        = 1 << 0,
+	SVGA3DBLOCKDESC_W           = 1 << 0,
+	SVGA3DBLOCKDESC_BUMP_L      = 1 << 0,
+
+	/* Format contains Green/V data */
+	SVGA3DBLOCKDESC_GREEN       = 1 << 1,
+	SVGA3DBLOCKDESC_V           = 1 << 1,
+
+	/* Format contains Red/W/Luminance data */
+	SVGA3DBLOCKDESC_RED         = 1 << 2,
+	SVGA3DBLOCKDESC_U           = 1 << 2,
+	SVGA3DBLOCKDESC_LUMINANCE   = 1 << 2,
+
+	/* Format contains Alpha/Q data */
+	SVGA3DBLOCKDESC_ALPHA       = 1 << 3,
+	SVGA3DBLOCKDESC_Q           = 1 << 3,
+
+	/* Format is a buffer */
+	SVGA3DBLOCKDESC_BUFFER      = 1 << 4,
+
+	/* Format is compressed */
+	SVGA3DBLOCKDESC_COMPRESSED  = 1 << 5,
+
+	/* Format uses IEEE floating point */
+	SVGA3DBLOCKDESC_FP          = 1 << 6,
+
+	/* Three separate blocks store data. */
+	SVGA3DBLOCKDESC_PLANAR_YUV  = 1 << 7,
+
+	/* 2 planes of Y, UV, e.g., NV12. */
+	SVGA3DBLOCKDESC_2PLANAR_YUV = 1 << 8,
+
+	/* 3 planes of separate Y, U, V, e.g., YV12. */
+	SVGA3DBLOCKDESC_3PLANAR_YUV = 1 << 9,
+
+	/* Block with a stencil channel */
+	SVGA3DBLOCKDESC_STENCIL     = 1 << 11,
+
+	/* Typeless format */
+	SVGA3DBLOCKDESC_TYPELESS    = 1 << 12,
+
+	/* Channels are signed integers */
+	SVGA3DBLOCKDESC_SINT        = 1 << 13,
+
+	/* Channels are unsigned integers */
+	SVGA3DBLOCKDESC_UINT        = 1 << 14,
+
+	/* Channels are normalized (when sampling) */
+	SVGA3DBLOCKDESC_NORM        = 1 << 15,
+
+	/* Channels are in SRGB */
+	SVGA3DBLOCKDESC_SRGB        = 1 << 16,
+
+	/* Shared exponent */
+	SVGA3DBLOCKDESC_EXP         = 1 << 17,
+
+	/* Format contains color data. */
+	SVGA3DBLOCKDESC_COLOR       = 1 << 18,
+	/* Format contains depth data. */
+	SVGA3DBLOCKDESC_DEPTH       = 1 << 19,
+	/* Format contains bump data. */
+	SVGA3DBLOCKDESC_BUMP        = 1 << 20,
+
+	/* Format contains YUV video data. */
+	SVGA3DBLOCKDESC_YUV_VIDEO   = 1 << 21,
+
+	/* For mixed unsigned/signed formats. */
+	SVGA3DBLOCKDESC_MIXED       = 1 << 22,
+
+	/* For distingushing CxV8U8. */
+	SVGA3DBLOCKDESC_CX          = 1 << 23,
+
+	/* Different compressed format groups. */
+	SVGA3DBLOCKDESC_BC1         = 1 << 24,
+	SVGA3DBLOCKDESC_BC2         = 1 << 25,
+	SVGA3DBLOCKDESC_BC3         = 1 << 26,
+	SVGA3DBLOCKDESC_BC4         = 1 << 27,
+	SVGA3DBLOCKDESC_BC5         = 1 << 28,
+
+	SVGA3DBLOCKDESC_A_UINT    = SVGA3DBLOCKDESC_ALPHA |
+				    SVGA3DBLOCKDESC_UINT |
+				    SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_A_UNORM   = SVGA3DBLOCKDESC_A_UINT |
+				    SVGA3DBLOCKDESC_NORM,
+	SVGA3DBLOCKDESC_R_UINT    = SVGA3DBLOCKDESC_RED |
+				    SVGA3DBLOCKDESC_UINT |
+				    SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_R_UNORM   = SVGA3DBLOCKDESC_R_UINT |
+				    SVGA3DBLOCKDESC_NORM,
+	SVGA3DBLOCKDESC_R_SINT    = SVGA3DBLOCKDESC_RED |
+				    SVGA3DBLOCKDESC_SINT |
+				    SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_R_SNORM   = SVGA3DBLOCKDESC_R_SINT |
+				    SVGA3DBLOCKDESC_NORM,
+	SVGA3DBLOCKDESC_G_UINT    = SVGA3DBLOCKDESC_GREEN |
+				    SVGA3DBLOCKDESC_UINT |
+				    SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_RG_UINT    = SVGA3DBLOCKDESC_RED |
+				     SVGA3DBLOCKDESC_GREEN |
+				     SVGA3DBLOCKDESC_UINT |
+				     SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_RG_UNORM   = SVGA3DBLOCKDESC_RG_UINT |
+				     SVGA3DBLOCKDESC_NORM,
+	SVGA3DBLOCKDESC_RG_SINT    = SVGA3DBLOCKDESC_RED |
+				     SVGA3DBLOCKDESC_GREEN |
+				     SVGA3DBLOCKDESC_SINT |
+				     SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_RG_SNORM   = SVGA3DBLOCKDESC_RG_SINT |
+				     SVGA3DBLOCKDESC_NORM,
+	SVGA3DBLOCKDESC_RGB_UINT   = SVGA3DBLOCKDESC_RED |
+				     SVGA3DBLOCKDESC_GREEN |
+				     SVGA3DBLOCKDESC_BLUE |
+				     SVGA3DBLOCKDESC_UINT |
+				     SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_RGB_SINT   = SVGA3DBLOCKDESC_RED |
+				     SVGA3DBLOCKDESC_GREEN |
+				     SVGA3DBLOCKDESC_BLUE |
+				     SVGA3DBLOCKDESC_SINT |
+				     SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_RGB_UNORM   = SVGA3DBLOCKDESC_RGB_UINT |
+				      SVGA3DBLOCKDESC_NORM,
+	SVGA3DBLOCKDESC_RGB_UNORM_SRGB = SVGA3DBLOCKDESC_RGB_UNORM |
+					 SVGA3DBLOCKDESC_SRGB,
+	SVGA3DBLOCKDESC_RGBA_UINT  = SVGA3DBLOCKDESC_RED |
+				     SVGA3DBLOCKDESC_GREEN |
+				     SVGA3DBLOCKDESC_BLUE |
+				     SVGA3DBLOCKDESC_ALPHA |
+				     SVGA3DBLOCKDESC_UINT |
+				     SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_RGBA_UNORM = SVGA3DBLOCKDESC_RGBA_UINT |
+				     SVGA3DBLOCKDESC_NORM,
+	SVGA3DBLOCKDESC_RGBA_UNORM_SRGB = SVGA3DBLOCKDESC_RGBA_UNORM |
+					  SVGA3DBLOCKDESC_SRGB,
+	SVGA3DBLOCKDESC_RGBA_SINT  = SVGA3DBLOCKDESC_RED |
+				     SVGA3DBLOCKDESC_GREEN |
+				     SVGA3DBLOCKDESC_BLUE |
+				     SVGA3DBLOCKDESC_ALPHA |
+				     SVGA3DBLOCKDESC_SINT |
+				     SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_RGBA_SNORM = SVGA3DBLOCKDESC_RGBA_SINT |
+				     SVGA3DBLOCKDESC_NORM,
+	SVGA3DBLOCKDESC_RGBA_FP    = SVGA3DBLOCKDESC_RED |
+				     SVGA3DBLOCKDESC_GREEN |
+				     SVGA3DBLOCKDESC_BLUE |
+				     SVGA3DBLOCKDESC_ALPHA |
+				     SVGA3DBLOCKDESC_FP |
+				     SVGA3DBLOCKDESC_COLOR,
 	SVGA3DBLOCKDESC_UV         = SVGA3DBLOCKDESC_U |
-	SVGA3DBLOCKDESC_V,
+				     SVGA3DBLOCKDESC_V |
+				     SVGA3DBLOCKDESC_BUMP,
 	SVGA3DBLOCKDESC_UVL        = SVGA3DBLOCKDESC_UV |
-	SVGA3DBLOCKDESC_LUMINANCE,
+				     SVGA3DBLOCKDESC_BUMP_L |
+				     SVGA3DBLOCKDESC_MIXED |
+				     SVGA3DBLOCKDESC_BUMP,
 	SVGA3DBLOCKDESC_UVW        = SVGA3DBLOCKDESC_UV |
-	SVGA3DBLOCKDESC_W,
+				     SVGA3DBLOCKDESC_W |
+				     SVGA3DBLOCKDESC_BUMP,
 	SVGA3DBLOCKDESC_UVWA       = SVGA3DBLOCKDESC_UVW |
-	SVGA3DBLOCKDESC_ALPHA,
+				     SVGA3DBLOCKDESC_ALPHA |
+				     SVGA3DBLOCKDESC_MIXED |
+				     SVGA3DBLOCKDESC_BUMP,
 	SVGA3DBLOCKDESC_UVWQ       = SVGA3DBLOCKDESC_U |
-	SVGA3DBLOCKDESC_V |
-	SVGA3DBLOCKDESC_W |
-	SVGA3DBLOCKDESC_Q,
-	SVGA3DBLOCKDESC_LA         = SVGA3DBLOCKDESC_LUMINANCE |
-	SVGA3DBLOCKDESC_ALPHA,
+				     SVGA3DBLOCKDESC_V |
+				     SVGA3DBLOCKDESC_W |
+				     SVGA3DBLOCKDESC_Q |
+				     SVGA3DBLOCKDESC_BUMP,
+	SVGA3DBLOCKDESC_L_UNORM    = SVGA3DBLOCKDESC_LUMINANCE |
+				     SVGA3DBLOCKDESC_UINT |
+				     SVGA3DBLOCKDESC_NORM |
+				     SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_LA_UNORM   = SVGA3DBLOCKDESC_LUMINANCE |
+				     SVGA3DBLOCKDESC_ALPHA |
+				     SVGA3DBLOCKDESC_UINT |
+				     SVGA3DBLOCKDESC_NORM |
+				     SVGA3DBLOCKDESC_COLOR,
 	SVGA3DBLOCKDESC_R_FP       = SVGA3DBLOCKDESC_RED |
-	SVGA3DBLOCKDESC_IEEE_FP,
+				     SVGA3DBLOCKDESC_FP |
+				     SVGA3DBLOCKDESC_COLOR,
 	SVGA3DBLOCKDESC_RG_FP      = SVGA3DBLOCKDESC_R_FP |
-	SVGA3DBLOCKDESC_GREEN,
+				     SVGA3DBLOCKDESC_GREEN |
+				     SVGA3DBLOCKDESC_COLOR,
 	SVGA3DBLOCKDESC_RGB_FP     = SVGA3DBLOCKDESC_RG_FP |
-	SVGA3DBLOCKDESC_BLUE,
-	SVGA3DBLOCKDESC_RGBA_FP    = SVGA3DBLOCKDESC_RGB_FP |
-	SVGA3DBLOCKDESC_ALPHA,
-	SVGA3DBLOCKDESC_DS         = SVGA3DBLOCKDESC_DEPTH |
-	SVGA3DBLOCKDESC_STENCIL,
-	SVGA3DBLOCKDESC_YUV        = SVGA3DBLOCKDESC_UV_VIDEO |
-	SVGA3DBLOCKDESC_Y,
+				     SVGA3DBLOCKDESC_BLUE |
+				     SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_YUV        = SVGA3DBLOCKDESC_YUV_VIDEO |
+				     SVGA3DBLOCKDESC_COLOR,
 	SVGA3DBLOCKDESC_AYUV       = SVGA3DBLOCKDESC_ALPHA |
-	SVGA3DBLOCKDESC_Y |
-	SVGA3DBLOCKDESC_U_VIDEO |
-	SVGA3DBLOCKDESC_V_VIDEO,
-	SVGA3DBLOCKDESC_RGBE       = SVGA3DBLOCKDESC_RGB |
-	SVGA3DBLOCKDESC_EXP,
-	SVGA3DBLOCKDESC_COMPRESSED_SRGB = SVGA3DBLOCKDESC_COMPRESSED |
-	SVGA3DBLOCKDESC_SRGB,
-	SVGA3DBLOCKDESC_NV12       = SVGA3DBLOCKDESC_PLANAR_YUV |
-	SVGA3DBLOCKDESC_2PLANAR_YUV,
-	SVGA3DBLOCKDESC_YV12       = SVGA3DBLOCKDESC_PLANAR_YUV |
-	SVGA3DBLOCKDESC_3PLANAR_YUV,
+				     SVGA3DBLOCKDESC_YUV_VIDEO |
+				     SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_RGB_EXP       = SVGA3DBLOCKDESC_RED |
+					SVGA3DBLOCKDESC_GREEN |
+					SVGA3DBLOCKDESC_BLUE |
+					SVGA3DBLOCKDESC_EXP |
+					SVGA3DBLOCKDESC_COLOR,
+
+	SVGA3DBLOCKDESC_COMP_TYPELESS = SVGA3DBLOCKDESC_COMPRESSED |
+					SVGA3DBLOCKDESC_TYPELESS,
+	SVGA3DBLOCKDESC_COMP_UNORM = SVGA3DBLOCKDESC_COMPRESSED |
+				     SVGA3DBLOCKDESC_UINT |
+				     SVGA3DBLOCKDESC_NORM |
+				     SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_COMP_SNORM = SVGA3DBLOCKDESC_COMPRESSED |
+				     SVGA3DBLOCKDESC_SINT |
+				     SVGA3DBLOCKDESC_NORM |
+				     SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_COMP_UNORM_SRGB = SVGA3DBLOCKDESC_COMP_UNORM |
+					  SVGA3DBLOCKDESC_SRGB,
+	SVGA3DBLOCKDESC_BC1_COMP_TYPELESS = SVGA3DBLOCKDESC_BC1 |
+					    SVGA3DBLOCKDESC_COMP_TYPELESS,
+	SVGA3DBLOCKDESC_BC1_COMP_UNORM = SVGA3DBLOCKDESC_BC1 |
+					 SVGA3DBLOCKDESC_COMP_UNORM,
+	SVGA3DBLOCKDESC_BC1_COMP_UNORM_SRGB = SVGA3DBLOCKDESC_BC1_COMP_UNORM |
+					      SVGA3DBLOCKDESC_SRGB,
+	SVGA3DBLOCKDESC_BC2_COMP_TYPELESS = SVGA3DBLOCKDESC_BC2 |
+					    SVGA3DBLOCKDESC_COMP_TYPELESS,
+	SVGA3DBLOCKDESC_BC2_COMP_UNORM = SVGA3DBLOCKDESC_BC2 |
+					 SVGA3DBLOCKDESC_COMP_UNORM,
+	SVGA3DBLOCKDESC_BC2_COMP_UNORM_SRGB = SVGA3DBLOCKDESC_BC2_COMP_UNORM |
+					      SVGA3DBLOCKDESC_SRGB,
+	SVGA3DBLOCKDESC_BC3_COMP_TYPELESS = SVGA3DBLOCKDESC_BC3 |
+					    SVGA3DBLOCKDESC_COMP_TYPELESS,
+	SVGA3DBLOCKDESC_BC3_COMP_UNORM = SVGA3DBLOCKDESC_BC3 |
+					 SVGA3DBLOCKDESC_COMP_UNORM,
+	SVGA3DBLOCKDESC_BC3_COMP_UNORM_SRGB = SVGA3DBLOCKDESC_BC3_COMP_UNORM |
+					      SVGA3DBLOCKDESC_SRGB,
+	SVGA3DBLOCKDESC_BC4_COMP_TYPELESS = SVGA3DBLOCKDESC_BC4 |
+					    SVGA3DBLOCKDESC_COMP_TYPELESS,
+	SVGA3DBLOCKDESC_BC4_COMP_UNORM = SVGA3DBLOCKDESC_BC4 |
+					 SVGA3DBLOCKDESC_COMP_UNORM,
+	SVGA3DBLOCKDESC_BC4_COMP_SNORM = SVGA3DBLOCKDESC_BC4 |
+					 SVGA3DBLOCKDESC_COMP_SNORM,
+	SVGA3DBLOCKDESC_BC5_COMP_TYPELESS = SVGA3DBLOCKDESC_BC5 |
+					    SVGA3DBLOCKDESC_COMP_TYPELESS,
+	SVGA3DBLOCKDESC_BC5_COMP_UNORM = SVGA3DBLOCKDESC_BC5 |
+					 SVGA3DBLOCKDESC_COMP_UNORM,
+	SVGA3DBLOCKDESC_BC5_COMP_SNORM = SVGA3DBLOCKDESC_BC5 |
+					 SVGA3DBLOCKDESC_COMP_SNORM,
+
+	SVGA3DBLOCKDESC_NV12       = SVGA3DBLOCKDESC_YUV_VIDEO |
+				     SVGA3DBLOCKDESC_PLANAR_YUV |
+				     SVGA3DBLOCKDESC_2PLANAR_YUV |
+				     SVGA3DBLOCKDESC_COLOR,
+	SVGA3DBLOCKDESC_YV12       = SVGA3DBLOCKDESC_YUV_VIDEO |
+				     SVGA3DBLOCKDESC_PLANAR_YUV |
+				     SVGA3DBLOCKDESC_3PLANAR_YUV |
+				     SVGA3DBLOCKDESC_COLOR,
+
+	SVGA3DBLOCKDESC_DEPTH_UINT = SVGA3DBLOCKDESC_DEPTH |
+				     SVGA3DBLOCKDESC_UINT,
+	SVGA3DBLOCKDESC_DEPTH_UNORM = SVGA3DBLOCKDESC_DEPTH_UINT |
+				     SVGA3DBLOCKDESC_NORM,
+	SVGA3DBLOCKDESC_DS      =    SVGA3DBLOCKDESC_DEPTH |
+				     SVGA3DBLOCKDESC_STENCIL,
+	SVGA3DBLOCKDESC_DS_UINT =    SVGA3DBLOCKDESC_DEPTH |
+				     SVGA3DBLOCKDESC_STENCIL |
+				     SVGA3DBLOCKDESC_UINT,
+	SVGA3DBLOCKDESC_DS_UNORM =   SVGA3DBLOCKDESC_DS_UINT |
+				     SVGA3DBLOCKDESC_NORM,
+	SVGA3DBLOCKDESC_DEPTH_FP   = SVGA3DBLOCKDESC_DEPTH |
+				     SVGA3DBLOCKDESC_FP,
+
+	SVGA3DBLOCKDESC_UV_UINT    = SVGA3DBLOCKDESC_UV |
+				     SVGA3DBLOCKDESC_UINT,
+	SVGA3DBLOCKDESC_UV_SNORM   = SVGA3DBLOCKDESC_UV |
+				     SVGA3DBLOCKDESC_SINT |
+				     SVGA3DBLOCKDESC_NORM,
+	SVGA3DBLOCKDESC_UVCX_SNORM = SVGA3DBLOCKDESC_UV_SNORM |
+				     SVGA3DBLOCKDESC_CX,
+	SVGA3DBLOCKDESC_UVWQ_SNORM = SVGA3DBLOCKDESC_UVWQ |
+				     SVGA3DBLOCKDESC_SINT |
+				     SVGA3DBLOCKDESC_NORM,
 };
 
-/*
- * SVGA3dSurfaceDesc describes the actual pixel data.
- *
- * This structure provides the following information:
- *    1. Block description.
- *    2. Dimensions of a block in the surface.
- *    3. Size of block in bytes.
- *    4. Bit depth of the pixel data.
- *    5. Channel bit depths and masks (if applicable).
- */
 struct svga3d_channel_def {
 	union {
 		u8 blue;
-		u8 u;
+		u8 w_bump;
+		u8 l_bump;
 		u8 uv_video;
 		u8 u_video;
 	};
 	union {
 		u8 green;
-		u8 v;
 		u8 stencil;
+		u8 v_bump;
 		u8 v_video;
 	};
 	union {
 		u8 red;
-		u8 w;
+		u8 u_bump;
 		u8 luminance;
-		u8 y;
+		u8 y_video;
 		u8 depth;
 		u8 data;
 	};
 	union {
 		u8 alpha;
-		u8 q;
+		u8 q_bump;
 		u8 exp;
 	};
 };
 
+/*
+ * struct svga3d_surface_desc - describes the actual pixel data.
+ *
+ * @format: Format
+ * @block_desc: Block description
+ * @block_size: Dimensions in pixels of a block
+ * @bytes_per_block: Size of block in bytes
+ * @pitch_bytes_per_block: Size of a block in bytes for purposes of pitch
+ * @bit_depth: Channel bit depths
+ * @bit_offset: Channel bit masks (in bits offset from the start of the pointer)
+ */
 struct svga3d_surface_desc {
 	SVGA3dSurfaceFormat format;
 	enum svga3d_block_desc block_desc;
+
 	surf_size_struct block_size;
 	u32 bytes_per_block;
 	u32 pitch_bytes_per_block;
 
-	u32 total_bit_depth;
 	struct svga3d_channel_def bit_depth;
 	struct svga3d_channel_def bit_offset;
 };
@@ -215,729 +381,728 @@ struct svga3d_surface_desc {
 static const struct svga3d_surface_desc svga3d_surface_descs[] = {
    {SVGA3D_FORMAT_INVALID, SVGA3DBLOCKDESC_NONE,
       {1, 1, 1},  0, 0,
-      0, {{0}, {0}, {0}, {0}},
+      {{0}, {0}, {0}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_X8R8G8B8, SVGA3DBLOCKDESC_RGB,
+   {SVGA3D_X8R8G8B8, SVGA3DBLOCKDESC_RGB_UNORM,
       {1, 1, 1},  4, 4,
-      24, {{8}, {8}, {8}, {0}},
+      {{8}, {8}, {8}, {0}},
       {{0}, {8}, {16}, {24}}},
 
-   {SVGA3D_A8R8G8B8, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_A8R8G8B8, SVGA3DBLOCKDESC_RGBA_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{8}, {8}, {8}, {8}},
+      {{8}, {8}, {8}, {8}},
       {{0}, {8}, {16}, {24}}},
 
-   {SVGA3D_R5G6B5, SVGA3DBLOCKDESC_RGB,
+   {SVGA3D_R5G6B5, SVGA3DBLOCKDESC_RGB_UNORM,
       {1, 1, 1},  2, 2,
-      16, {{5}, {6}, {5}, {0}},
+      {{5}, {6}, {5}, {0}},
       {{0}, {5}, {11}, {0}}},
 
-   {SVGA3D_X1R5G5B5, SVGA3DBLOCKDESC_RGB,
+   {SVGA3D_X1R5G5B5, SVGA3DBLOCKDESC_RGB_UNORM,
       {1, 1, 1},  2, 2,
-      15, {{5}, {5}, {5}, {0}},
+      {{5}, {5}, {5}, {0}},
       {{0}, {5}, {10}, {0}}},
 
-   {SVGA3D_A1R5G5B5, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_A1R5G5B5, SVGA3DBLOCKDESC_RGBA_UNORM,
       {1, 1, 1},  2, 2,
-      16, {{5}, {5}, {5}, {1}},
+      {{5}, {5}, {5}, {1}},
       {{0}, {5}, {10}, {15}}},
 
-   {SVGA3D_A4R4G4B4, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_A4R4G4B4, SVGA3DBLOCKDESC_RGBA_UNORM,
       {1, 1, 1},  2, 2,
-      16, {{4}, {4}, {4}, {4}},
+      {{4}, {4}, {4}, {4}},
       {{0}, {4}, {8}, {12}}},
 
-   {SVGA3D_Z_D32, SVGA3DBLOCKDESC_DEPTH,
+   {SVGA3D_Z_D32, SVGA3DBLOCKDESC_DEPTH_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{0}, {0}, {32}, {0}},
+      {{0}, {0}, {32}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_Z_D16, SVGA3DBLOCKDESC_DEPTH,
+   {SVGA3D_Z_D16, SVGA3DBLOCKDESC_DEPTH_UNORM,
       {1, 1, 1},  2, 2,
-      16, {{0}, {0}, {16}, {0}},
+      {{0}, {0}, {16}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_Z_D24S8, SVGA3DBLOCKDESC_DS,
+   {SVGA3D_Z_D24S8, SVGA3DBLOCKDESC_DS_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{0}, {8}, {24}, {0}},
-      {{0}, {24}, {0}, {0}}},
+      {{0}, {8}, {24}, {0}},
+      {{0}, {0}, {8}, {0}}},
 
-   {SVGA3D_Z_D15S1, SVGA3DBLOCKDESC_DS,
+   {SVGA3D_Z_D15S1, SVGA3DBLOCKDESC_DS_UNORM,
       {1, 1, 1},  2, 2,
-      16, {{0}, {1}, {15}, {0}},
-      {{0}, {15}, {0}, {0}}},
+      {{0}, {1}, {15}, {0}},
+      {{0}, {0}, {1}, {0}}},
 
-   {SVGA3D_LUMINANCE8, SVGA3DBLOCKDESC_LUMINANCE,
+   {SVGA3D_LUMINANCE8, SVGA3DBLOCKDESC_L_UNORM,
       {1, 1, 1},  1, 1,
-      8, {{0}, {0}, {8}, {0}},
+      {{0}, {0}, {8}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_LUMINANCE4_ALPHA4, SVGA3DBLOCKDESC_LA,
-    {1  , 1, 1},  1, 1,
-      8, {{0}, {0}, {4}, {4}},
+   {SVGA3D_LUMINANCE4_ALPHA4, SVGA3DBLOCKDESC_LA_UNORM,
+      {1, 1, 1},  1, 1,
+      {{0}, {0}, {4}, {4}},
       {{0}, {0}, {0}, {4}}},
 
-   {SVGA3D_LUMINANCE16, SVGA3DBLOCKDESC_LUMINANCE,
+   {SVGA3D_LUMINANCE16, SVGA3DBLOCKDESC_L_UNORM,
       {1, 1, 1},  2, 2,
-      16, {{0}, {0}, {16}, {0}},
+      {{0}, {0}, {16}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_LUMINANCE8_ALPHA8, SVGA3DBLOCKDESC_LA,
+   {SVGA3D_LUMINANCE8_ALPHA8, SVGA3DBLOCKDESC_LA_UNORM,
       {1, 1, 1},  2, 2,
-      16, {{0}, {0}, {8}, {8}},
+      {{0}, {0}, {8}, {8}},
       {{0}, {0}, {0}, {8}}},
 
-   {SVGA3D_DXT1, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_DXT1, SVGA3DBLOCKDESC_BC1_COMP_UNORM,
       {4, 4, 1},  8, 8,
-      64, {{0}, {0}, {64}, {0}},
+      {{0}, {0}, {64}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_DXT2, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_DXT2, SVGA3DBLOCKDESC_BC2_COMP_UNORM,
       {4, 4, 1},  16, 16,
-      128, {{0}, {0}, {128}, {0}},
+      {{0}, {0}, {128}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_DXT3, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_DXT3, SVGA3DBLOCKDESC_BC2_COMP_UNORM,
       {4, 4, 1},  16, 16,
-      128, {{0}, {0}, {128}, {0}},
+      {{0}, {0}, {128}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_DXT4, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_DXT4, SVGA3DBLOCKDESC_BC3_COMP_UNORM,
       {4, 4, 1},  16, 16,
-      128, {{0}, {0}, {128}, {0}},
+      {{0}, {0}, {128}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_DXT5, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_DXT5, SVGA3DBLOCKDESC_BC3_COMP_UNORM,
       {4, 4, 1},  16, 16,
-      128, {{0}, {0}, {128}, {0}},
+      {{0}, {0}, {128}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_BUMPU8V8, SVGA3DBLOCKDESC_UV,
+   {SVGA3D_BUMPU8V8, SVGA3DBLOCKDESC_UV_SNORM,
       {1, 1, 1},  2, 2,
-      16, {{0}, {0}, {8}, {8}},
-      {{0}, {0}, {0}, {8}}},
+      {{0}, {8}, {8}, {0}},
+      {{0}, {8}, {0}, {0}}},
 
    {SVGA3D_BUMPL6V5U5, SVGA3DBLOCKDESC_UVL,
       {1, 1, 1},  2, 2,
-      16, {{5}, {5}, {6}, {0}},
-      {{11}, {6}, {0}, {0}}},
+      {{6}, {5}, {5}, {0}},
+      {{10}, {5}, {0}, {0}}},
 
    {SVGA3D_BUMPX8L8V8U8, SVGA3DBLOCKDESC_UVL,
       {1, 1, 1},  4, 4,
-      32, {{8}, {8}, {8}, {0}},
+      {{8}, {8}, {8}, {0}},
       {{16}, {8}, {0}, {0}}},
 
-   {SVGA3D_BUMPL8V8U8, SVGA3DBLOCKDESC_UVL,
+   {SVGA3D_FORMAT_DEAD1, SVGA3DBLOCKDESC_UVL,
       {1, 1, 1},  3, 3,
-      24, {{8}, {8}, {8}, {0}},
+      {{8}, {8}, {8}, {0}},
       {{16}, {8}, {0}, {0}}},
 
    {SVGA3D_ARGB_S10E5, SVGA3DBLOCKDESC_RGBA_FP,
       {1, 1, 1},  8, 8,
-      64, {{16}, {16}, {16}, {16}},
+      {{16}, {16}, {16}, {16}},
       {{32}, {16}, {0}, {48}}},
 
    {SVGA3D_ARGB_S23E8, SVGA3DBLOCKDESC_RGBA_FP,
       {1, 1, 1},  16, 16,
-      128, {{32}, {32}, {32}, {32}},
+      {{32}, {32}, {32}, {32}},
       {{64}, {32}, {0}, {96}}},
 
-   {SVGA3D_A2R10G10B10, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_A2R10G10B10, SVGA3DBLOCKDESC_RGBA_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{10}, {10}, {10}, {2}},
+      {{10}, {10}, {10}, {2}},
       {{0}, {10}, {20}, {30}}},
 
-   {SVGA3D_V8U8, SVGA3DBLOCKDESC_UV,
+   {SVGA3D_V8U8, SVGA3DBLOCKDESC_UV_SNORM,
       {1, 1, 1},  2, 2,
-      16, {{8}, {8}, {0}, {0}},
-      {{8}, {0}, {0}, {0}}},
+      {{0}, {8}, {8}, {0}},
+      {{0}, {8}, {0}, {0}}},
 
-   {SVGA3D_Q8W8V8U8, SVGA3DBLOCKDESC_UVWQ,
+   {SVGA3D_Q8W8V8U8, SVGA3DBLOCKDESC_UVWQ_SNORM,
       {1, 1, 1},  4, 4,
-      32, {{8}, {8}, {8}, {8}},
-      {{24}, {16}, {8}, {0}}},
+      {{8}, {8}, {8}, {8}},
+      {{16}, {8}, {0}, {24}}},
 
-   {SVGA3D_CxV8U8, SVGA3DBLOCKDESC_UV,
+   {SVGA3D_CxV8U8, SVGA3DBLOCKDESC_UVCX_SNORM,
       {1, 1, 1},  2, 2,
-      16, {{8}, {8}, {0}, {0}},
-      {{8}, {0}, {0}, {0}}},
+      {{0}, {8}, {8}, {0}},
+      {{0}, {8}, {0}, {0}}},
 
    {SVGA3D_X8L8V8U8, SVGA3DBLOCKDESC_UVL,
       {1, 1, 1},  4, 4,
-      24, {{8}, {8}, {8}, {0}},
+      {{8}, {8}, {8}, {0}},
       {{16}, {8}, {0}, {0}}},
 
    {SVGA3D_A2W10V10U10, SVGA3DBLOCKDESC_UVWA,
       {1, 1, 1},  4, 4,
-      32, {{10}, {10}, {10}, {2}},
-      {{0}, {10}, {20}, {30}}},
+      {{10}, {10}, {10}, {2}},
+      {{20}, {10}, {0}, {30}}},
 
-   {SVGA3D_ALPHA8, SVGA3DBLOCKDESC_ALPHA,
+   {SVGA3D_ALPHA8, SVGA3DBLOCKDESC_A_UNORM,
       {1, 1, 1},  1, 1,
-      8, {{0}, {0}, {0}, {8}},
+      {{0}, {0}, {0}, {8}},
       {{0}, {0}, {0}, {0}}},
 
    {SVGA3D_R_S10E5, SVGA3DBLOCKDESC_R_FP,
       {1, 1, 1},  2, 2,
-      16, {{0}, {0}, {16}, {0}},
+      {{0}, {0}, {16}, {0}},
       {{0}, {0}, {0}, {0}}},
 
    {SVGA3D_R_S23E8, SVGA3DBLOCKDESC_R_FP,
       {1, 1, 1},  4, 4,
-      32, {{0}, {0}, {32}, {0}},
+      {{0}, {0}, {32}, {0}},
       {{0}, {0}, {0}, {0}}},
 
    {SVGA3D_RG_S10E5, SVGA3DBLOCKDESC_RG_FP,
       {1, 1, 1},  4, 4,
-      32, {{0}, {16}, {16}, {0}},
+      {{0}, {16}, {16}, {0}},
       {{0}, {16}, {0}, {0}}},
 
    {SVGA3D_RG_S23E8, SVGA3DBLOCKDESC_RG_FP,
       {1, 1, 1},  8, 8,
-      64, {{0}, {32}, {32}, {0}},
+      {{0}, {32}, {32}, {0}},
       {{0}, {32}, {0}, {0}}},
 
    {SVGA3D_BUFFER, SVGA3DBLOCKDESC_BUFFER,
       {1, 1, 1},  1, 1,
-      8, {{0}, {0}, {8}, {0}},
+      {{0}, {0}, {8}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_Z_D24X8, SVGA3DBLOCKDESC_DEPTH,
+   {SVGA3D_Z_D24X8, SVGA3DBLOCKDESC_DEPTH_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{0}, {0}, {24}, {0}},
-      {{0}, {24}, {0}, {0}}},
+      {{0}, {0}, {24}, {0}},
+      {{0}, {0}, {8}, {0}}},
 
-   {SVGA3D_V16U16, SVGA3DBLOCKDESC_UV,
+   {SVGA3D_V16U16, SVGA3DBLOCKDESC_UV_SNORM,
       {1, 1, 1},  4, 4,
-      32, {{16}, {16}, {0}, {0}},
-      {{16}, {0}, {0}, {0}}},
+      {{0}, {16}, {16}, {0}},
+      {{0}, {16}, {0}, {0}}},
 
-   {SVGA3D_G16R16, SVGA3DBLOCKDESC_RG,
+   {SVGA3D_G16R16, SVGA3DBLOCKDESC_RG_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{0}, {16}, {16}, {0}},
-      {{0}, {0}, {16}, {0}}},
+      {{0}, {16}, {16}, {0}},
+      {{0}, {16}, {0}, {0}}},
 
-   {SVGA3D_A16B16G16R16, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_A16B16G16R16, SVGA3DBLOCKDESC_RGBA_UNORM,
       {1, 1, 1},  8, 8,
-      64, {{16}, {16}, {16}, {16}},
+      {{16}, {16}, {16}, {16}},
       {{32}, {16}, {0}, {48}}},
 
    {SVGA3D_UYVY, SVGA3DBLOCKDESC_YUV,
-      {1, 1, 1},  2, 2,
-      16, {{8}, {0}, {8}, {0}},
+      {2, 1, 1},  4, 4,
+      {{8}, {0}, {8}, {0}},
       {{0}, {0}, {8}, {0}}},
 
    {SVGA3D_YUY2, SVGA3DBLOCKDESC_YUV,
-      {1, 1, 1},  2, 2,
-      16, {{8}, {0}, {8}, {0}},
+      {2, 1, 1},  4, 4,
+      {{8}, {0}, {8}, {0}},
       {{8}, {0}, {0}, {0}}},
 
    {SVGA3D_NV12, SVGA3DBLOCKDESC_NV12,
       {2, 2, 1},  6, 2,
-      48, {{0}, {0}, {48}, {0}},
+      {{0}, {0}, {48}, {0}},
       {{0}, {0}, {0}, {0}}},
 
    {SVGA3D_AYUV, SVGA3DBLOCKDESC_AYUV,
       {1, 1, 1},  4, 4,
-      32, {{8}, {8}, {8}, {8}},
+      {{8}, {8}, {8}, {8}},
       {{0}, {8}, {16}, {24}}},
 
-   {SVGA3D_R32G32B32A32_TYPELESS, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_R32G32B32A32_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  16, 16,
-      128, {{32}, {32}, {32}, {32}},
+      {{32}, {32}, {32}, {32}},
       {{64}, {32}, {0}, {96}}},
 
-   {SVGA3D_R32G32B32A32_UINT, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_R32G32B32A32_UINT, SVGA3DBLOCKDESC_RGBA_UINT,
       {1, 1, 1},  16, 16,
-      128, {{32}, {32}, {32}, {32}},
+      {{32}, {32}, {32}, {32}},
       {{64}, {32}, {0}, {96}}},
 
-   {SVGA3D_R32G32B32A32_SINT, SVGA3DBLOCKDESC_UVWQ,
+   {SVGA3D_R32G32B32A32_SINT, SVGA3DBLOCKDESC_RGBA_SINT,
       {1, 1, 1},  16, 16,
-      128, {{32}, {32}, {32}, {32}},
+      {{32}, {32}, {32}, {32}},
       {{64}, {32}, {0}, {96}}},
 
-   {SVGA3D_R32G32B32_TYPELESS, SVGA3DBLOCKDESC_RGB,
+   {SVGA3D_R32G32B32_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  12, 12,
-      96, {{32}, {32}, {32}, {0}},
+      {{32}, {32}, {32}, {0}},
       {{64}, {32}, {0}, {0}}},
 
    {SVGA3D_R32G32B32_FLOAT, SVGA3DBLOCKDESC_RGB_FP,
       {1, 1, 1},  12, 12,
-      96, {{32}, {32}, {32}, {0}},
+      {{32}, {32}, {32}, {0}},
       {{64}, {32}, {0}, {0}}},
 
-   {SVGA3D_R32G32B32_UINT, SVGA3DBLOCKDESC_RGB,
+   {SVGA3D_R32G32B32_UINT, SVGA3DBLOCKDESC_RGB_UINT,
       {1, 1, 1},  12, 12,
-      96, {{32}, {32}, {32}, {0}},
+      {{32}, {32}, {32}, {0}},
       {{64}, {32}, {0}, {0}}},
 
-   {SVGA3D_R32G32B32_SINT, SVGA3DBLOCKDESC_UVW,
+   {SVGA3D_R32G32B32_SINT, SVGA3DBLOCKDESC_RGB_SINT,
       {1, 1, 1},  12, 12,
-      96, {{32}, {32}, {32}, {0}},
+      {{32}, {32}, {32}, {0}},
       {{64}, {32}, {0}, {0}}},
 
-   {SVGA3D_R16G16B16A16_TYPELESS, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_R16G16B16A16_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  8, 8,
-      64, {{16}, {16}, {16}, {16}},
+      {{16}, {16}, {16}, {16}},
       {{32}, {16}, {0}, {48}}},
 
-   {SVGA3D_R16G16B16A16_UINT, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_R16G16B16A16_UINT, SVGA3DBLOCKDESC_RGBA_UINT,
       {1, 1, 1},  8, 8,
-      64, {{16}, {16}, {16}, {16}},
+      {{16}, {16}, {16}, {16}},
       {{32}, {16}, {0}, {48}}},
 
-   {SVGA3D_R16G16B16A16_SNORM, SVGA3DBLOCKDESC_UVWQ,
+   {SVGA3D_R16G16B16A16_SNORM, SVGA3DBLOCKDESC_RGBA_SNORM,
       {1, 1, 1},  8, 8,
-      64, {{16}, {16}, {16}, {16}},
+      {{16}, {16}, {16}, {16}},
       {{32}, {16}, {0}, {48}}},
 
-   {SVGA3D_R16G16B16A16_SINT, SVGA3DBLOCKDESC_UVWQ,
+   {SVGA3D_R16G16B16A16_SINT, SVGA3DBLOCKDESC_RGBA_SINT,
       {1, 1, 1},  8, 8,
-      64, {{16}, {16}, {16}, {16}},
+      {{16}, {16}, {16}, {16}},
       {{32}, {16}, {0}, {48}}},
 
-   {SVGA3D_R32G32_TYPELESS, SVGA3DBLOCKDESC_RG,
+   {SVGA3D_R32G32_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  8, 8,
-      64, {{0}, {32}, {32}, {0}},
+      {{0}, {32}, {32}, {0}},
       {{0}, {32}, {0}, {0}}},
 
-   {SVGA3D_R32G32_UINT, SVGA3DBLOCKDESC_RG,
+   {SVGA3D_R32G32_UINT, SVGA3DBLOCKDESC_RG_UINT,
       {1, 1, 1},  8, 8,
-      64, {{0}, {32}, {32}, {0}},
+      {{0}, {32}, {32}, {0}},
       {{0}, {32}, {0}, {0}}},
 
-   {SVGA3D_R32G32_SINT, SVGA3DBLOCKDESC_UV,
+   {SVGA3D_R32G32_SINT, SVGA3DBLOCKDESC_RG_SINT,
       {1, 1, 1},  8, 8,
-      64, {{0}, {32}, {32}, {0}},
+      {{0}, {32}, {32}, {0}},
       {{0}, {32}, {0}, {0}}},
 
-   {SVGA3D_R32G8X24_TYPELESS, SVGA3DBLOCKDESC_RG,
+   {SVGA3D_R32G8X24_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  8, 8,
-      64, {{0}, {8}, {32}, {0}},
+      {{0}, {8}, {32}, {0}},
       {{0}, {32}, {0}, {0}}},
 
    {SVGA3D_D32_FLOAT_S8X24_UINT, SVGA3DBLOCKDESC_DS,
       {1, 1, 1},  8, 8,
-      64, {{0}, {8}, {32}, {0}},
+      {{0}, {8}, {32}, {0}},
       {{0}, {32}, {0}, {0}}},
 
-   {SVGA3D_R32_FLOAT_X8X24_TYPELESS, SVGA3DBLOCKDESC_R_FP,
+   {SVGA3D_R32_FLOAT_X8X24, SVGA3DBLOCKDESC_R_FP,
       {1, 1, 1},  8, 8,
-      64, {{0}, {0}, {32}, {0}},
+      {{0}, {0}, {32}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_X32_TYPELESS_G8X24_UINT, SVGA3DBLOCKDESC_GREEN,
+   {SVGA3D_X32_G8X24_UINT, SVGA3DBLOCKDESC_G_UINT,
       {1, 1, 1},  8, 8,
-      64, {{0}, {8}, {0}, {0}},
+      {{0}, {8}, {0}, {0}},
       {{0}, {32}, {0}, {0}}},
 
-   {SVGA3D_R10G10B10A2_TYPELESS, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_R10G10B10A2_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  4, 4,
-      32, {{10}, {10}, {10}, {2}},
-      {{0}, {10}, {20}, {30}}},
+      {{10}, {10}, {10}, {2}},
+      {{20}, {10}, {0}, {30}}},
 
-   {SVGA3D_R10G10B10A2_UINT, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_R10G10B10A2_UINT, SVGA3DBLOCKDESC_RGBA_UINT,
       {1, 1, 1},  4, 4,
-      32, {{10}, {10}, {10}, {2}},
-      {{0}, {10}, {20}, {30}}},
+      {{10}, {10}, {10}, {2}},
+      {{20}, {10}, {0}, {30}}},
 
    {SVGA3D_R11G11B10_FLOAT, SVGA3DBLOCKDESC_RGB_FP,
       {1, 1, 1},  4, 4,
-      32, {{10}, {11}, {11}, {0}},
-      {{0}, {10}, {21}, {0}}},
+      {{10}, {11}, {11}, {0}},
+      {{22}, {11}, {0}, {0}}},
 
-   {SVGA3D_R8G8B8A8_TYPELESS, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_R8G8B8A8_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  4, 4,
-      32, {{8}, {8}, {8}, {8}},
+      {{8}, {8}, {8}, {8}},
       {{16}, {8}, {0}, {24}}},
 
-   {SVGA3D_R8G8B8A8_UNORM, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_R8G8B8A8_UNORM, SVGA3DBLOCKDESC_RGBA_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{8}, {8}, {8}, {8}},
+      {{8}, {8}, {8}, {8}},
       {{16}, {8}, {0}, {24}}},
 
-   {SVGA3D_R8G8B8A8_UNORM_SRGB, SVGA3DBLOCKDESC_RGBA_SRGB,
+   {SVGA3D_R8G8B8A8_UNORM_SRGB, SVGA3DBLOCKDESC_RGBA_UNORM_SRGB,
       {1, 1, 1},  4, 4,
-      32, {{8}, {8}, {8}, {8}},
+      {{8}, {8}, {8}, {8}},
       {{16}, {8}, {0}, {24}}},
 
-   {SVGA3D_R8G8B8A8_UINT, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_R8G8B8A8_UINT, SVGA3DBLOCKDESC_RGBA_UINT,
       {1, 1, 1},  4, 4,
-      32, {{8}, {8}, {8}, {8}},
+      {{8}, {8}, {8}, {8}},
       {{16}, {8}, {0}, {24}}},
 
-   {SVGA3D_R8G8B8A8_SINT, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_R8G8B8A8_SINT, SVGA3DBLOCKDESC_RGBA_SINT,
       {1, 1, 1},  4, 4,
-      32, {{8}, {8}, {8}, {8}},
+      {{8}, {8}, {8}, {8}},
       {{16}, {8}, {0}, {24}}},
 
-   {SVGA3D_R16G16_TYPELESS, SVGA3DBLOCKDESC_RG,
+   {SVGA3D_R16G16_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  4, 4,
-      32, {{0}, {16}, {16}, {0}},
+      {{0}, {16}, {16}, {0}},
       {{0}, {16}, {0}, {0}}},
 
-   {SVGA3D_R16G16_UINT, SVGA3DBLOCKDESC_RG_FP,
+   {SVGA3D_R16G16_UINT, SVGA3DBLOCKDESC_RG_UINT,
       {1, 1, 1},  4, 4,
-      32, {{0}, {16}, {16}, {0}},
+      {{0}, {16}, {16}, {0}},
       {{0}, {16}, {0}, {0}}},
 
-   {SVGA3D_R16G16_SINT, SVGA3DBLOCKDESC_UV,
+   {SVGA3D_R16G16_SINT, SVGA3DBLOCKDESC_RG_SINT,
       {1, 1, 1},  4, 4,
-      32, {{0}, {16}, {16}, {0}},
+      {{0}, {16}, {16}, {0}},
       {{0}, {16}, {0}, {0}}},
 
-   {SVGA3D_R32_TYPELESS, SVGA3DBLOCKDESC_RED,
+   {SVGA3D_R32_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  4, 4,
-      32, {{0}, {0}, {32}, {0}},
+      {{0}, {0}, {32}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_D32_FLOAT, SVGA3DBLOCKDESC_DEPTH,
+   {SVGA3D_D32_FLOAT, SVGA3DBLOCKDESC_DEPTH_FP,
       {1, 1, 1},  4, 4,
-      32, {{0}, {0}, {32}, {0}},
+      {{0}, {0}, {32}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R32_UINT, SVGA3DBLOCKDESC_RED,
+   {SVGA3D_R32_UINT, SVGA3DBLOCKDESC_R_UINT,
       {1, 1, 1},  4, 4,
-      32, {{0}, {0}, {32}, {0}},
+      {{0}, {0}, {32}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R32_SINT, SVGA3DBLOCKDESC_RED,
+   {SVGA3D_R32_SINT, SVGA3DBLOCKDESC_R_SINT,
       {1, 1, 1},  4, 4,
-      32, {{0}, {0}, {32}, {0}},
+      {{0}, {0}, {32}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R24G8_TYPELESS, SVGA3DBLOCKDESC_RG,
+   {SVGA3D_R24G8_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  4, 4,
-      32, {{0}, {8}, {24}, {0}},
+      {{0}, {8}, {24}, {0}},
       {{0}, {24}, {0}, {0}}},
 
-   {SVGA3D_D24_UNORM_S8_UINT, SVGA3DBLOCKDESC_DS,
+   {SVGA3D_D24_UNORM_S8_UINT, SVGA3DBLOCKDESC_DS_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{0}, {8}, {24}, {0}},
+      {{0}, {8}, {24}, {0}},
       {{0}, {24}, {0}, {0}}},
 
-   {SVGA3D_R24_UNORM_X8_TYPELESS, SVGA3DBLOCKDESC_RED,
+   {SVGA3D_R24_UNORM_X8, SVGA3DBLOCKDESC_R_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{0}, {0}, {24}, {0}},
+      {{0}, {0}, {24}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_X24_TYPELESS_G8_UINT, SVGA3DBLOCKDESC_GREEN,
+   {SVGA3D_X24_G8_UINT, SVGA3DBLOCKDESC_G_UINT,
       {1, 1, 1},  4, 4,
-      32, {{0}, {8}, {0}, {0}},
+      {{0}, {8}, {0}, {0}},
       {{0}, {24}, {0}, {0}}},
 
-   {SVGA3D_R8G8_TYPELESS, SVGA3DBLOCKDESC_RG,
+   {SVGA3D_R8G8_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  2, 2,
-      16, {{0}, {8}, {8}, {0}},
+      {{0}, {8}, {8}, {0}},
       {{0}, {8}, {0}, {0}}},
 
-   {SVGA3D_R8G8_UNORM, SVGA3DBLOCKDESC_RG,
+   {SVGA3D_R8G8_UNORM, SVGA3DBLOCKDESC_RG_UNORM,
       {1, 1, 1},  2, 2,
-      16, {{0}, {8}, {8}, {0}},
+      {{0}, {8}, {8}, {0}},
       {{0}, {8}, {0}, {0}}},
 
-   {SVGA3D_R8G8_UINT, SVGA3DBLOCKDESC_RG,
+   {SVGA3D_R8G8_UINT, SVGA3DBLOCKDESC_RG_UINT,
       {1, 1, 1},  2, 2,
-      16, {{0}, {8}, {8}, {0}},
+      {{0}, {8}, {8}, {0}},
       {{0}, {8}, {0}, {0}}},
 
-   {SVGA3D_R8G8_SINT, SVGA3DBLOCKDESC_UV,
+   {SVGA3D_R8G8_SINT, SVGA3DBLOCKDESC_RG_SINT,
       {1, 1, 1},  2, 2,
-      16, {{0}, {8}, {8}, {0}},
+      {{0}, {8}, {8}, {0}},
       {{0}, {8}, {0}, {0}}},
 
-   {SVGA3D_R16_TYPELESS, SVGA3DBLOCKDESC_RED,
+   {SVGA3D_R16_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  2, 2,
-      16, {{0}, {0}, {16}, {0}},
+      {{0}, {0}, {16}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R16_UNORM, SVGA3DBLOCKDESC_RED,
+   {SVGA3D_R16_UNORM, SVGA3DBLOCKDESC_R_UNORM,
       {1, 1, 1},  2, 2,
-      16, {{0}, {0}, {16}, {0}},
+      {{0}, {0}, {16}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R16_UINT, SVGA3DBLOCKDESC_RED,
+   {SVGA3D_R16_UINT, SVGA3DBLOCKDESC_R_UINT,
       {1, 1, 1},  2, 2,
-      16, {{0}, {0}, {16}, {0}},
+      {{0}, {0}, {16}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R16_SNORM, SVGA3DBLOCKDESC_U,
+   {SVGA3D_R16_SNORM, SVGA3DBLOCKDESC_R_SNORM,
       {1, 1, 1},  2, 2,
-      16, {{0}, {0}, {16}, {0}},
+      {{0}, {0}, {16}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R16_SINT, SVGA3DBLOCKDESC_U,
+   {SVGA3D_R16_SINT, SVGA3DBLOCKDESC_R_SINT,
       {1, 1, 1},  2, 2,
-      16, {{0}, {0}, {16}, {0}},
+      {{0}, {0}, {16}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R8_TYPELESS, SVGA3DBLOCKDESC_RED,
+   {SVGA3D_R8_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  1, 1,
-      8, {{0}, {0}, {8}, {0}},
+      {{0}, {0}, {8}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R8_UNORM, SVGA3DBLOCKDESC_RED,
+   {SVGA3D_R8_UNORM, SVGA3DBLOCKDESC_R_UNORM,
       {1, 1, 1},  1, 1,
-      8, {{0}, {0}, {8}, {0}},
+      {{0}, {0}, {8}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R8_UINT, SVGA3DBLOCKDESC_RED,
+   {SVGA3D_R8_UINT, SVGA3DBLOCKDESC_R_UINT,
       {1, 1, 1},  1, 1,
-      8, {{0}, {0}, {8}, {0}},
+      {{0}, {0}, {8}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R8_SNORM, SVGA3DBLOCKDESC_U,
+   {SVGA3D_R8_SNORM, SVGA3DBLOCKDESC_R_SNORM,
       {1, 1, 1},  1, 1,
-      8, {{0}, {0}, {8}, {0}},
+      {{0}, {0}, {8}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R8_SINT, SVGA3DBLOCKDESC_U,
+   {SVGA3D_R8_SINT, SVGA3DBLOCKDESC_R_SINT,
       {1, 1, 1},  1, 1,
-      8, {{0}, {0}, {8}, {0}},
+      {{0}, {0}, {8}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_P8, SVGA3DBLOCKDESC_RED,
+   {SVGA3D_P8, SVGA3DBLOCKDESC_NONE,
       {1, 1, 1},  1, 1,
-      8, {{0}, {0}, {8}, {0}},
+      {{0}, {0}, {8}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R9G9B9E5_SHAREDEXP, SVGA3DBLOCKDESC_RGBE,
+   {SVGA3D_R9G9B9E5_SHAREDEXP, SVGA3DBLOCKDESC_RGB_EXP,
       {1, 1, 1},  4, 4,
-      32, {{9}, {9}, {9}, {5}},
+      {{9}, {9}, {9}, {5}},
       {{18}, {9}, {0}, {27}}},
 
-   {SVGA3D_R8G8_B8G8_UNORM, SVGA3DBLOCKDESC_RG,
-      {1, 1, 1},  2, 2,
-      16, {{0}, {8}, {8}, {0}},
-      {{0}, {8}, {0}, {0}}},
+   {SVGA3D_R8G8_B8G8_UNORM, SVGA3DBLOCKDESC_NONE,
+      {2, 1, 1},  4, 4,
+      {{0}, {8}, {8}, {0}},
+      {{0}, {0}, {8}, {0}}},
 
-   {SVGA3D_G8R8_G8B8_UNORM, SVGA3DBLOCKDESC_RG,
-      {1, 1, 1},  2, 2,
-      16, {{0}, {8}, {8}, {0}},
+   {SVGA3D_G8R8_G8B8_UNORM, SVGA3DBLOCKDESC_NONE,
+      {2, 1, 1},  4, 4,
+      {{0}, {8}, {8}, {0}},
       {{0}, {8}, {0}, {0}}},
 
-   {SVGA3D_BC1_TYPELESS, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_BC1_TYPELESS, SVGA3DBLOCKDESC_BC1_COMP_TYPELESS,
       {4, 4, 1},  8, 8,
-      64, {{0}, {0}, {64}, {0}},
+      {{0}, {0}, {64}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_BC1_UNORM_SRGB, SVGA3DBLOCKDESC_COMPRESSED_SRGB,
+   {SVGA3D_BC1_UNORM_SRGB, SVGA3DBLOCKDESC_BC1_COMP_UNORM_SRGB,
       {4, 4, 1},  8, 8,
-      64, {{0}, {0}, {64}, {0}},
+      {{0}, {0}, {64}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_BC2_TYPELESS, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_BC2_TYPELESS, SVGA3DBLOCKDESC_BC2_COMP_TYPELESS,
       {4, 4, 1},  16, 16,
-      128, {{0}, {0}, {128}, {0}},
+      {{0}, {0}, {128}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_BC2_UNORM_SRGB, SVGA3DBLOCKDESC_COMPRESSED_SRGB,
+   {SVGA3D_BC2_UNORM_SRGB, SVGA3DBLOCKDESC_BC2_COMP_UNORM_SRGB,
       {4, 4, 1},  16, 16,
-      128, {{0}, {0}, {128}, {0}},
+      {{0}, {0}, {128}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_BC3_TYPELESS, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_BC3_TYPELESS, SVGA3DBLOCKDESC_BC3_COMP_TYPELESS,
       {4, 4, 1},  16, 16,
-      128, {{0}, {0}, {128}, {0}},
+      {{0}, {0}, {128}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_BC3_UNORM_SRGB, SVGA3DBLOCKDESC_COMPRESSED_SRGB,
+   {SVGA3D_BC3_UNORM_SRGB, SVGA3DBLOCKDESC_BC3_COMP_UNORM_SRGB,
       {4, 4, 1},  16, 16,
-      128, {{0}, {0}, {128}, {0}},
+      {{0}, {0}, {128}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_BC4_TYPELESS, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_BC4_TYPELESS, SVGA3DBLOCKDESC_BC4_COMP_TYPELESS,
       {4, 4, 1},  8, 8,
-      64, {{0}, {0}, {64}, {0}},
+      {{0}, {0}, {64}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_ATI1, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_ATI1, SVGA3DBLOCKDESC_BC4_COMP_UNORM,
       {4, 4, 1},  8, 8,
-      64, {{0}, {0}, {64}, {0}},
+      {{0}, {0}, {64}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_BC4_SNORM, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_BC4_SNORM, SVGA3DBLOCKDESC_BC4_COMP_SNORM,
       {4, 4, 1},  8, 8,
-      64, {{0}, {0}, {64}, {0}},
+      {{0}, {0}, {64}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_BC5_TYPELESS, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_BC5_TYPELESS, SVGA3DBLOCKDESC_BC5_COMP_TYPELESS,
       {4, 4, 1},  16, 16,
-      128, {{0}, {0}, {128}, {0}},
+      {{0}, {0}, {128}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_ATI2, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_ATI2, SVGA3DBLOCKDESC_BC5_COMP_UNORM,
       {4, 4, 1},  16, 16,
-      128, {{0}, {0}, {128}, {0}},
+      {{0}, {0}, {128}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_BC5_SNORM, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_BC5_SNORM, SVGA3DBLOCKDESC_BC5_COMP_SNORM,
       {4, 4, 1},  16, 16,
-      128, {{0}, {0}, {128}, {0}},
+      {{0}, {0}, {128}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R10G10B10_XR_BIAS_A2_UNORM, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_R10G10B10_XR_BIAS_A2_UNORM, SVGA3DBLOCKDESC_RGBA_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{10}, {10}, {10}, {2}},
-      {{0}, {10}, {20}, {30}}},
+      {{10}, {10}, {10}, {2}},
+     {{20}, {10}, {0}, {30}}},
 
-   {SVGA3D_B8G8R8A8_TYPELESS, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_B8G8R8A8_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  4, 4,
-      32, {{8}, {8}, {8}, {8}},
+      {{8}, {8}, {8}, {8}},
       {{0}, {8}, {16}, {24}}},
 
-   {SVGA3D_B8G8R8A8_UNORM_SRGB, SVGA3DBLOCKDESC_RGBA_SRGB,
+   {SVGA3D_B8G8R8A8_UNORM_SRGB, SVGA3DBLOCKDESC_RGBA_UNORM_SRGB,
       {1, 1, 1},  4, 4,
-      32, {{8}, {8}, {8}, {8}},
+      {{8}, {8}, {8}, {8}},
       {{0}, {8}, {16}, {24}}},
 
-   {SVGA3D_B8G8R8X8_TYPELESS, SVGA3DBLOCKDESC_RGB,
+   {SVGA3D_B8G8R8X8_TYPELESS, SVGA3DBLOCKDESC_TYPELESS,
       {1, 1, 1},  4, 4,
-      24, {{8}, {8}, {8}, {0}},
+      {{8}, {8}, {8}, {0}},
       {{0}, {8}, {16}, {24}}},
 
-   {SVGA3D_B8G8R8X8_UNORM_SRGB, SVGA3DBLOCKDESC_RGB_SRGB,
+   {SVGA3D_B8G8R8X8_UNORM_SRGB, SVGA3DBLOCKDESC_RGB_UNORM_SRGB,
       {1, 1, 1},  4, 4,
-      24, {{8}, {8}, {8}, {0}},
+      {{8}, {8}, {8}, {0}},
       {{0}, {8}, {16}, {24}}},
 
-   {SVGA3D_Z_DF16, SVGA3DBLOCKDESC_DEPTH,
+   {SVGA3D_Z_DF16, SVGA3DBLOCKDESC_DEPTH_UNORM,
       {1, 1, 1},  2, 2,
-      16, {{0}, {0}, {16}, {0}},
+      {{0}, {0}, {16}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_Z_DF24, SVGA3DBLOCKDESC_DEPTH,
+   {SVGA3D_Z_DF24, SVGA3DBLOCKDESC_DEPTH_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{0}, {8}, {24}, {0}},
-      {{0}, {24}, {0}, {0}}},
+      {{0}, {0}, {24}, {0}},
+      {{0}, {0}, {8}, {0}}},
 
-   {SVGA3D_Z_D24S8_INT, SVGA3DBLOCKDESC_DS,
+   {SVGA3D_Z_D24S8_INT, SVGA3DBLOCKDESC_DS_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{0}, {8}, {24}, {0}},
-      {{0}, {24}, {0}, {0}}},
+      {{0}, {8}, {24}, {0}},
+      {{0}, {0}, {8}, {0}}},
 
    {SVGA3D_YV12, SVGA3DBLOCKDESC_YV12,
       {2, 2, 1},  6, 2,
-      48, {{0}, {0}, {48}, {0}},
+      {{0}, {0}, {48}, {0}},
       {{0}, {0}, {0}, {0}}},
 
    {SVGA3D_R32G32B32A32_FLOAT, SVGA3DBLOCKDESC_RGBA_FP,
       {1, 1, 1},  16, 16,
-      128, {{32}, {32}, {32}, {32}},
+      {{32}, {32}, {32}, {32}},
       {{64}, {32}, {0}, {96}}},
 
    {SVGA3D_R16G16B16A16_FLOAT, SVGA3DBLOCKDESC_RGBA_FP,
       {1, 1, 1},  8, 8,
-      64, {{16}, {16}, {16}, {16}},
+      {{16}, {16}, {16}, {16}},
       {{32}, {16}, {0}, {48}}},
 
-   {SVGA3D_R16G16B16A16_UNORM, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_R16G16B16A16_UNORM, SVGA3DBLOCKDESC_RGBA_UNORM,
       {1, 1, 1},  8, 8,
-      64, {{16}, {16}, {16}, {16}},
+      {{16}, {16}, {16}, {16}},
       {{32}, {16}, {0}, {48}}},
 
    {SVGA3D_R32G32_FLOAT, SVGA3DBLOCKDESC_RG_FP,
       {1, 1, 1},  8, 8,
-      64, {{0}, {32}, {32}, {0}},
+      {{0}, {32}, {32}, {0}},
       {{0}, {32}, {0}, {0}}},
 
-   {SVGA3D_R10G10B10A2_UNORM, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_R10G10B10A2_UNORM, SVGA3DBLOCKDESC_RGBA_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{10}, {10}, {10}, {2}},
-      {{0}, {10}, {20}, {30}}},
+      {{10}, {10}, {10}, {2}},
+      {{20}, {10}, {0}, {30}}},
 
-   {SVGA3D_R8G8B8A8_SNORM, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_R8G8B8A8_SNORM, SVGA3DBLOCKDESC_RGBA_SNORM,
       {1, 1, 1},  4, 4,
-      32, {{8}, {8}, {8}, {8}},
-      {{24}, {16}, {8}, {0}}},
+      {{8}, {8}, {8}, {8}},
+      {{16}, {8}, {0}, {24}}},
 
    {SVGA3D_R16G16_FLOAT, SVGA3DBLOCKDESC_RG_FP,
       {1, 1, 1},  4, 4,
-      32, {{0}, {16}, {16}, {0}},
+      {{0}, {16}, {16}, {0}},
       {{0}, {16}, {0}, {0}}},
 
-   {SVGA3D_R16G16_UNORM, SVGA3DBLOCKDESC_RG,
+   {SVGA3D_R16G16_UNORM, SVGA3DBLOCKDESC_RG_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{0}, {16}, {16}, {0}},
-      {{0}, {0}, {16}, {0}}},
+      {{0}, {16}, {16}, {0}},
+      {{0}, {16}, {0}, {0}}},
 
-   {SVGA3D_R16G16_SNORM, SVGA3DBLOCKDESC_RG,
+   {SVGA3D_R16G16_SNORM, SVGA3DBLOCKDESC_RG_SNORM,
       {1, 1, 1},  4, 4,
-      32, {{16}, {16}, {0}, {0}},
-      {{16}, {0}, {0}, {0}}},
+      {{0}, {16}, {16}, {0}},
+      {{0}, {16}, {0}, {0}}},
 
    {SVGA3D_R32_FLOAT, SVGA3DBLOCKDESC_R_FP,
       {1, 1, 1},  4, 4,
-      32, {{0}, {0}, {32}, {0}},
+      {{0}, {0}, {32}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_R8G8_SNORM, SVGA3DBLOCKDESC_RG,
+   {SVGA3D_R8G8_SNORM, SVGA3DBLOCKDESC_RG_SNORM,
       {1, 1, 1},  2, 2,
-      16, {{8}, {8}, {0}, {0}},
-      {{8}, {0}, {0}, {0}}},
+      {{0}, {8}, {8}, {0}},
+      {{0}, {8}, {0}, {0}}},
 
    {SVGA3D_R16_FLOAT, SVGA3DBLOCKDESC_R_FP,
       {1, 1, 1},  2, 2,
-      16, {{0}, {0}, {16}, {0}},
+      {{0}, {0}, {16}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_D16_UNORM, SVGA3DBLOCKDESC_DEPTH,
+   {SVGA3D_D16_UNORM, SVGA3DBLOCKDESC_DEPTH_UNORM,
       {1, 1, 1},  2, 2,
-      16, {{0}, {0}, {16}, {0}},
+      {{0}, {0}, {16}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_A8_UNORM, SVGA3DBLOCKDESC_ALPHA,
+   {SVGA3D_A8_UNORM, SVGA3DBLOCKDESC_A_UNORM,
       {1, 1, 1},  1, 1,
-      8, {{0}, {0}, {0}, {8}},
+      {{0}, {0}, {0}, {8}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_BC1_UNORM, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_BC1_UNORM, SVGA3DBLOCKDESC_BC1_COMP_UNORM,
       {4, 4, 1},  8, 8,
-      64, {{0}, {0}, {64}, {0}},
+      {{0}, {0}, {64}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_BC2_UNORM, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_BC2_UNORM, SVGA3DBLOCKDESC_BC2_COMP_UNORM,
       {4, 4, 1},  16, 16,
-      128, {{0}, {0}, {128}, {0}},
+      {{0}, {0}, {128}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_BC3_UNORM, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_BC3_UNORM, SVGA3DBLOCKDESC_BC3_COMP_UNORM,
       {4, 4, 1},  16, 16,
-      128, {{0}, {0}, {128}, {0}},
+      {{0}, {0}, {128}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_B5G6R5_UNORM, SVGA3DBLOCKDESC_RGB,
+   {SVGA3D_B5G6R5_UNORM, SVGA3DBLOCKDESC_RGB_UNORM,
       {1, 1, 1},  2, 2,
-      16, {{5}, {6}, {5}, {0}},
+      {{5}, {6}, {5}, {0}},
       {{0}, {5}, {11}, {0}}},
 
-   {SVGA3D_B5G5R5A1_UNORM, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_B5G5R5A1_UNORM, SVGA3DBLOCKDESC_RGBA_UNORM,
       {1, 1, 1},  2, 2,
-      16, {{5}, {5}, {5}, {1}},
+      {{5}, {5}, {5}, {1}},
       {{0}, {5}, {10}, {15}}},
 
-   {SVGA3D_B8G8R8A8_UNORM, SVGA3DBLOCKDESC_RGBA,
+   {SVGA3D_B8G8R8A8_UNORM, SVGA3DBLOCKDESC_RGBA_UNORM,
       {1, 1, 1},  4, 4,
-      32, {{8}, {8}, {8}, {8}},
+      {{8}, {8}, {8}, {8}},
       {{0}, {8}, {16}, {24}}},
 
-   {SVGA3D_B8G8R8X8_UNORM, SVGA3DBLOCKDESC_RGB,
+   {SVGA3D_B8G8R8X8_UNORM, SVGA3DBLOCKDESC_RGB_UNORM,
       {1, 1, 1},  4, 4,
-      24, {{8}, {8}, {8}, {0}},
+      {{8}, {8}, {8}, {0}},
       {{0}, {8}, {16}, {24}}},
 
-   {SVGA3D_BC4_UNORM, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_BC4_UNORM, SVGA3DBLOCKDESC_BC4_COMP_UNORM,
       {4, 4, 1},  8, 8,
-      64, {{0}, {0}, {64}, {0}},
+      {{0}, {0}, {64}, {0}},
       {{0}, {0}, {0}, {0}}},
 
-   {SVGA3D_BC5_UNORM, SVGA3DBLOCKDESC_COMPRESSED,
+   {SVGA3D_BC5_UNORM, SVGA3DBLOCKDESC_BC5_COMP_UNORM,
       {4, 4, 1},  16, 16,
-      128, {{0}, {0}, {128}, {0}},
+      {{0}, {0}, {128}, {0}},
       {{0}, {0}, {0}, {0}}},
-
 };
 
 static inline u32 clamped_umul32(u32 a, u32 b)
@@ -946,6 +1111,10 @@ static inline u32 clamped_umul32(u32 a, u32 b)
 	return (tmp > (uint64_t) ((u32) -1)) ? (u32) -1 : tmp;
 }
 
+/**
+ * svga3dsurface_get_desc - Look up the appropriate SVGA3dSurfaceDesc for the
+ * given format.
+ */
 static inline const struct svga3d_surface_desc *
 svga3dsurface_get_desc(SVGA3dSurfaceFormat format)
 {
@@ -955,23 +1124,10 @@ svga3dsurface_get_desc(SVGA3dSurfaceFormat format)
 	return &svga3d_surface_descs[SVGA3D_FORMAT_INVALID];
 }
 
-/*
- *----------------------------------------------------------------------
- *
- * svga3dsurface_get_mip_size --
- *
- *      Given a base level size and the mip level, compute the size of
- *      the mip level.
- *
- * Results:
- *      See above.
- *
- * Side effects:
- *      None.
- *
- *----------------------------------------------------------------------
+/**
+ * svga3dsurface_get_mip_size -  Given a base level size and the mip level,
+ * compute the size of the mip level.
  */
-
 static inline surf_size_struct
 svga3dsurface_get_mip_size(surf_size_struct base_level, u32 mip_level)
 {
@@ -1018,28 +1174,17 @@ svga3dsurface_calculate_pitch(const struct svga3d_surface_desc *desc,
 	return pitch;
 }
 
-/*
- *-----------------------------------------------------------------------------
- *
- * svga3dsurface_get_image_buffer_size --
- *
- *      Return the number of bytes of buffer space required to store
- *      one image of a surface, optionally using the specified pitch.
- *
- *      If pitch is zero, it is assumed that rows are tightly packed.
+/**
+ * svga3dsurface_get_image_buffer_size - Calculates image buffer size.
  *
- *      This function is overflow-safe. If the result would have
- *      overflowed, instead we return MAX_UINT32.
+ * Return the number of bytes of buffer space required to store one image of a
+ * surface, optionally using the specified pitch.
  *
- * Results:
- *      Byte count.
+ * If pitch is zero, it is assumed that rows are tightly packed.
  *
- * Side effects:
- *      None.
- *
- *-----------------------------------------------------------------------------
+ * This function is overflow-safe. If the result would have overflowed, instead
+ * we return MAX_UINT32.
  */
-
 static inline u32
 svga3dsurface_get_image_buffer_size(const struct svga3d_surface_desc *desc,
 				    const surf_size_struct *size,
@@ -1067,6 +1212,9 @@ svga3dsurface_get_image_buffer_size(const struct svga3d_surface_desc *desc,
 	return total_size;
 }
 
+/**
+ * svga3dsurface_get_serialized_size - Get the serialized size for the image.
+ */
 static inline u32
 svga3dsurface_get_serialized_size(SVGA3dSurfaceFormat format,
 				  surf_size_struct base_level_size,
@@ -1087,6 +1235,26 @@ svga3dsurface_get_serialized_size(SVGA3dSurfaceFormat format,
 	return total_size * num_layers;
 }
 
+/**
+ * svga3dsurface_get_serialized_size_extended - Returns the number of bytes
+ * required for a surface with given parameters. Support for sample count.
+ */
+static inline u32
+svga3dsurface_get_serialized_size_extended(SVGA3dSurfaceFormat format,
+					   surf_size_struct base_level_size,
+					   u32 num_mip_levels,
+					   u32 num_layers,
+					   u32 num_samples)
+{
+	uint64_t total_size =
+		svga3dsurface_get_serialized_size(format,
+						  base_level_size,
+						  num_mip_levels,
+						  num_layers);
+	total_size *= max_t(u32, 1, num_samples);
+
+	return min_t(uint64_t, total_size, (uint64_t)U32_MAX);
+}
 
 /**
  * svga3dsurface_get_pixel_offset - Compute the offset (in bytes) to a pixel
@@ -1206,3 +1374,5 @@ svga3dsurface_is_screen_target_format(SVGA3dSurfaceFormat format)
 	}
 	return svga3dsurface_is_dx_screen_target_format(format);
 }
+
+#endif /* _SVGA3D_SURFACEDEFS_H_ */
diff --git a/drivers/gpu/drm/vmwgfx/device_include/svga3d_types.h b/drivers/gpu/drm/vmwgfx/device_include/svga3d_types.h
index 27b33ba88430..308370665a8e 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/svga3d_types.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/svga3d_types.h
@@ -1,5 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**********************************************************
- * Copyright 2012-2015 VMware, Inc.  All rights reserved.
+ * Copyright 2012-2015 VMware, Inc.
  *
  * Permission is hereby granted, free of charge, to any person
  * obtaining a copy of this software and associated documentation
@@ -44,9 +45,21 @@
 
 #define SVGA3D_INVALID_ID         ((uint32)-1)
 
+typedef uint8 SVGABool8;   /* 8-bit Bool definition */
 typedef uint32 SVGA3dBool; /* 32-bit Bool definition */
 typedef uint32 SVGA3dColor; /* a, r, g, b */
 
+typedef uint32 SVGA3dSurfaceId;
+
+typedef
+#include "vmware_pack_begin.h"
+struct {
+   uint32 numerator;
+   uint32 denominator;
+}
+#include "vmware_pack_end.h"
+SVGA3dFraction64;
+
 typedef
 #include "vmware_pack_begin.h"
 struct SVGA3dCopyRect {
@@ -145,7 +158,7 @@ typedef enum SVGA3dSurfaceFormat {
    SVGA3D_BUMPU8V8                     = 20,
    SVGA3D_BUMPL6V5U5                   = 21,
    SVGA3D_BUMPX8L8V8U8                 = 22,
-   SVGA3D_BUMPL8V8U8                   = 23,
+   SVGA3D_FORMAT_DEAD1                 = 23,
 
    SVGA3D_ARGB_S10E5                   = 24,   /* 16-bit floating-point ARGB */
    SVGA3D_ARGB_S23E8                   = 25,   /* 32-bit floating-point ARGB */
@@ -204,8 +217,8 @@ typedef enum SVGA3dSurfaceFormat {
    SVGA3D_R32G32_SINT                  = 59,
    SVGA3D_R32G8X24_TYPELESS            = 60,
    SVGA3D_D32_FLOAT_S8X24_UINT         = 61,
-   SVGA3D_R32_FLOAT_X8X24_TYPELESS     = 62,
-   SVGA3D_X32_TYPELESS_G8X24_UINT      = 63,
+   SVGA3D_R32_FLOAT_X8X24              = 62,
+   SVGA3D_X32_G8X24_UINT               = 63,
    SVGA3D_R10G10B10A2_TYPELESS         = 64,
    SVGA3D_R10G10B10A2_UINT             = 65,
    SVGA3D_R11G11B10_FLOAT              = 66,
@@ -223,8 +236,8 @@ typedef enum SVGA3dSurfaceFormat {
    SVGA3D_R32_SINT                     = 78,
    SVGA3D_R24G8_TYPELESS               = 79,
    SVGA3D_D24_UNORM_S8_UINT            = 80,
-   SVGA3D_R24_UNORM_X8_TYPELESS        = 81,
-   SVGA3D_X24_TYPELESS_G8_UINT         = 82,
+   SVGA3D_R24_UNORM_X8                 = 81,
+   SVGA3D_X24_G8_UINT                  = 82,
    SVGA3D_R8G8_TYPELESS                = 83,
    SVGA3D_R8G8_UNORM                   = 84,
    SVGA3D_R8G8_UINT                    = 85,
@@ -296,92 +309,114 @@ typedef enum SVGA3dSurfaceFormat {
    SVGA3D_FORMAT_MAX
 } SVGA3dSurfaceFormat;
 
-typedef enum SVGA3dSurfaceFlags {
-   SVGA3D_SURFACE_CUBEMAP               = (1 << 0),
+/*
+ * SVGA3d Surface Flags --
+ */
+#define SVGA3D_SURFACE_CUBEMAP                (1 << 0)
 
-   /*
-    * HINT flags are not enforced by the device but are useful for
-    * performance.
-    */
-   SVGA3D_SURFACE_HINT_STATIC           = (1 << 1),
-   SVGA3D_SURFACE_HINT_DYNAMIC          = (1 << 2),
-   SVGA3D_SURFACE_HINT_INDEXBUFFER      = (1 << 3),
-   SVGA3D_SURFACE_HINT_VERTEXBUFFER     = (1 << 4),
-   SVGA3D_SURFACE_HINT_TEXTURE          = (1 << 5),
-   SVGA3D_SURFACE_HINT_RENDERTARGET     = (1 << 6),
-   SVGA3D_SURFACE_HINT_DEPTHSTENCIL     = (1 << 7),
-   SVGA3D_SURFACE_HINT_WRITEONLY        = (1 << 8),
-   SVGA3D_SURFACE_MASKABLE_ANTIALIAS    = (1 << 9),
-   SVGA3D_SURFACE_AUTOGENMIPMAPS        = (1 << 10),
-   SVGA3D_SURFACE_DECODE_RENDERTARGET   = (1 << 11),
+/*
+ * HINT flags are not enforced by the device but are useful for
+ * performance.
+ */
+#define SVGA3D_SURFACE_HINT_STATIC            (CONST64U(1) << 1)
+#define SVGA3D_SURFACE_HINT_DYNAMIC           (CONST64U(1) << 2)
+#define SVGA3D_SURFACE_HINT_INDEXBUFFER       (CONST64U(1) << 3)
+#define SVGA3D_SURFACE_HINT_VERTEXBUFFER      (CONST64U(1) << 4)
+#define SVGA3D_SURFACE_HINT_TEXTURE           (CONST64U(1) << 5)
+#define SVGA3D_SURFACE_HINT_RENDERTARGET      (CONST64U(1) << 6)
+#define SVGA3D_SURFACE_HINT_DEPTHSTENCIL      (CONST64U(1) << 7)
+#define SVGA3D_SURFACE_HINT_WRITEONLY         (CONST64U(1) << 8)
+#define SVGA3D_SURFACE_MASKABLE_ANTIALIAS     (CONST64U(1) << 9)
+#define SVGA3D_SURFACE_AUTOGENMIPMAPS         (CONST64U(1) << 10)
+
+#define SVGA3D_SURFACE_DECODE_RENDERTARGET    (CONST64U(1) << 11)
 
-   /*
-    * Is this surface using a base-level pitch for it's mob backing?
-    *
-    * This flag is not intended to be set by guest-drivers, but is instead
-    * set by the device when the surface is bound to a mob with a specified
-    * pitch.
-    */
-   SVGA3D_SURFACE_MOB_PITCH             = (1 << 12),
+/*
+ * Is this surface using a base-level pitch for it's mob backing?
+ *
+ * This flag is not intended to be set by guest-drivers, but is instead
+ * set by the device when the surface is bound to a mob with a specified
+ * pitch.
+ */
+#define SVGA3D_SURFACE_MOB_PITCH              (CONST64U(1) << 12)
 
-   SVGA3D_SURFACE_INACTIVE              = (1 << 13),
-   SVGA3D_SURFACE_HINT_RT_LOCKABLE      = (1 << 14),
-   SVGA3D_SURFACE_VOLUME                = (1 << 15),
+#define SVGA3D_SURFACE_INACTIVE               (CONST64U(1) << 13)
+#define SVGA3D_SURFACE_HINT_RT_LOCKABLE       (CONST64U(1) << 14)
+#define SVGA3D_SURFACE_VOLUME                 (CONST64U(1) << 15)
 
-   /*
-    * Required to be set on a surface to bind it to a screen target.
-    */
-   SVGA3D_SURFACE_SCREENTARGET          = (1 << 16),
+/*
+ * Required to be set on a surface to bind it to a screen target.
+ */
+#define SVGA3D_SURFACE_SCREENTARGET           (CONST64U(1) << 16)
 
-   /*
-    * Align images in the guest-backing mob to 16-bytes.
-    */
-   SVGA3D_SURFACE_ALIGN16               = (1 << 17),
+/*
+ * Align images in the guest-backing mob to 16-bytes.
+ */
+#define SVGA3D_SURFACE_ALIGN16                (CONST64U(1) << 17)
 
-   SVGA3D_SURFACE_1D                    = (1 << 18),
-   SVGA3D_SURFACE_ARRAY                 = (1 << 19),
+#define SVGA3D_SURFACE_1D                     (CONST64U(1) << 18)
+#define SVGA3D_SURFACE_ARRAY                  (CONST64U(1) << 19)
 
-   /*
-    * Bind flags.
-    * These are enforced for any surface defined with DefineGBSurface_v2.
-    */
-   SVGA3D_SURFACE_BIND_VERTEX_BUFFER    = (1 << 20),
-   SVGA3D_SURFACE_BIND_INDEX_BUFFER     = (1 << 21),
-   SVGA3D_SURFACE_BIND_CONSTANT_BUFFER  = (1 << 22),
-   SVGA3D_SURFACE_BIND_SHADER_RESOURCE  = (1 << 23),
-   SVGA3D_SURFACE_BIND_RENDER_TARGET    = (1 << 24),
-   SVGA3D_SURFACE_BIND_DEPTH_STENCIL    = (1 << 25),
-   SVGA3D_SURFACE_BIND_STREAM_OUTPUT    = (1 << 26),
+/*
+ * Bind flags.
+ * These are enforced for any surface defined with DefineGBSurface_v2.
+ */
+#define SVGA3D_SURFACE_BIND_VERTEX_BUFFER     (CONST64U(1) << 20)
+#define SVGA3D_SURFACE_BIND_INDEX_BUFFER      (CONST64U(1) << 21)
+#define SVGA3D_SURFACE_BIND_CONSTANT_BUFFER   (CONST64U(1) << 22)
+#define SVGA3D_SURFACE_BIND_SHADER_RESOURCE   (CONST64U(1) << 23)
+#define SVGA3D_SURFACE_BIND_RENDER_TARGET     (CONST64U(1) << 24)
+#define SVGA3D_SURFACE_BIND_DEPTH_STENCIL     (CONST64U(1) << 25)
+#define SVGA3D_SURFACE_BIND_STREAM_OUTPUT     (CONST64U(1) << 26)
 
-   /*
-    * A note on staging flags:
-    *
-    * The STAGING flags notes that the surface will not be used directly by the
-    * drawing pipeline, i.e. that it will not be bound to any bind point.
-    * Staging surfaces may be used by copy operations to move data in and out
-    * of other surfaces.
-    *
-    * The HINT_INDIRECT_UPDATE flag suggests that the surface will receive
-    * updates indirectly, i.e. the surface will not be updated directly, but
-    * will receive copies from staging surfaces.
-    */
-   SVGA3D_SURFACE_STAGING_UPLOAD        = (1 << 27),
-   SVGA3D_SURFACE_STAGING_DOWNLOAD      = (1 << 28),
-   SVGA3D_SURFACE_HINT_INDIRECT_UPDATE  = (1 << 29),
+/*
+ * The STAGING flags notes that the surface will not be used directly by the
+ * drawing pipeline, i.e. that it will not be bound to any bind point.
+ * Staging surfaces may be used by copy operations to move data in and out
+ * of other surfaces.  No bind flags may be set on surfaces with this flag.
+ *
+ * The HINT_INDIRECT_UPDATE flag suggests that the surface will receive
+ * updates indirectly, i.e. the surface will not be updated directly, but
+ * will receive copies from staging surfaces.
+ */
+#define SVGA3D_SURFACE_STAGING_UPLOAD         (CONST64U(1) << 27)
+#define SVGA3D_SURFACE_STAGING_DOWNLOAD       (CONST64U(1) << 28)
+#define SVGA3D_SURFACE_HINT_INDIRECT_UPDATE   (CONST64U(1) << 29)
 
-   /*
-    * Setting this flag allow this surface to be used with the
-    * SVGA_3D_CMD_DX_TRANSFER_FROM_BUFFER command.  It is only valid for
-    * buffer surfaces, an no bind flags are allowed to be set on surfaces
-    * with this flag.
-    */
-   SVGA3D_SURFACE_TRANSFER_FROM_BUFFER  = (1 << 30),
+/*
+ * Setting this flag allow this surface to be used with the
+ * SVGA_3D_CMD_DX_TRANSFER_FROM_BUFFER command.  It is only valid for
+ * buffer surfaces, and no bind flags are allowed to be set on surfaces
+ * with this flag.
+ */
+#define SVGA3D_SURFACE_TRANSFER_FROM_BUFFER   (CONST64U(1) << 30)
 
-   /*
-    * Marker for the last defined bit.
-    */
-   SVGA3D_SURFACE_FLAG_MAX              = (1 << 31),
-} SVGA3dSurfaceFlags;
+/*
+ * Reserved for video operations.
+ */
+#define SVGA3D_SURFACE_RESERVED1              (CONST64U(1) << 31)
+
+/*
+ * Specifies that a surface is multisample, and therefore requires the full
+ * mob-backing to store all the samples.
+ */
+#define SVGA3D_SURFACE_MULTISAMPLE            (CONST64U(1) << 32)
+
+#define SVGA3D_SURFACE_FLAG_MAX               (CONST64U(1) << 33)
+
+/*
+ * Surface flags types:
+ *
+ * SVGA3dSurface1Flags:  Lower 32-bits of flags.
+ * SVGA3dSurface2Flags:  Upper 32-bits of flags.
+ * SVGA3dSurfaceAllFlags: Full 64-bits of flags.
+ */
+typedef uint32 SVGA3dSurface1Flags;
+typedef uint32 SVGA3dSurface2Flags;
+typedef uint64 SVGA3dSurfaceAllFlags;
+
+#define SVGA3D_SURFACE_FLAGS1_MASK ((uint64_t)MAX_UINT32)
+#define SVGA3D_SURFACE_FLAGS2_MASK (MAX_UINT64 & ~SVGA3D_SURFACE_FLAGS1_MASK)
 
 #define SVGA3D_SURFACE_HB_DISALLOWED_MASK        \
         (  SVGA3D_SURFACE_MOB_PITCH    |         \
@@ -392,29 +427,41 @@ typedef enum SVGA3dSurfaceFlags {
            SVGA3D_SURFACE_STAGING_UPLOAD |       \
            SVGA3D_SURFACE_STAGING_DOWNLOAD |     \
            SVGA3D_SURFACE_HINT_INDIRECT_UPDATE | \
-           SVGA3D_SURFACE_TRANSFER_FROM_BUFFER   \
+           SVGA3D_SURFACE_TRANSFER_FROM_BUFFER | \
+           SVGA3D_SURFACE_MULTISAMPLE            \
+        )
+
+#define SVGA3D_SURFACE_HB_PRESENT_DISALLOWED_MASK   \
+       (   SVGA3D_SURFACE_1D |                      \
+           SVGA3D_SURFACE_MULTISAMPLE               \
         )
 
 #define SVGA3D_SURFACE_2D_DISALLOWED_MASK           \
         (  SVGA3D_SURFACE_CUBEMAP |                 \
            SVGA3D_SURFACE_MASKABLE_ANTIALIAS |      \
            SVGA3D_SURFACE_AUTOGENMIPMAPS |          \
-           SVGA3D_SURFACE_DECODE_RENDERTARGET |     \
            SVGA3D_SURFACE_VOLUME |                  \
            SVGA3D_SURFACE_1D |                      \
-           SVGA3D_SURFACE_ARRAY |                   \
            SVGA3D_SURFACE_BIND_VERTEX_BUFFER |      \
            SVGA3D_SURFACE_BIND_INDEX_BUFFER |       \
            SVGA3D_SURFACE_BIND_CONSTANT_BUFFER |    \
            SVGA3D_SURFACE_BIND_DEPTH_STENCIL |      \
            SVGA3D_SURFACE_BIND_STREAM_OUTPUT |      \
-           SVGA3D_SURFACE_TRANSFER_FROM_BUFFER      \
+           SVGA3D_SURFACE_TRANSFER_FROM_BUFFER |    \
+           SVGA3D_SURFACE_MULTISAMPLE               \
+        )
+
+#define SVGA3D_SURFACE_BASICOPS_DISALLOWED_MASK     \
+        (  SVGA3D_SURFACE_CUBEMAP |                 \
+           SVGA3D_SURFACE_AUTOGENMIPMAPS |          \
+           SVGA3D_SURFACE_VOLUME |                  \
+           SVGA3D_SURFACE_1D |                      \
+           SVGA3D_SURFACE_MULTISAMPLE               \
         )
 
 #define SVGA3D_SURFACE_SCREENTARGET_DISALLOWED_MASK \
         (  SVGA3D_SURFACE_CUBEMAP |                 \
            SVGA3D_SURFACE_AUTOGENMIPMAPS |          \
-           SVGA3D_SURFACE_DECODE_RENDERTARGET |     \
            SVGA3D_SURFACE_VOLUME |                  \
            SVGA3D_SURFACE_1D |                      \
            SVGA3D_SURFACE_BIND_VERTEX_BUFFER |      \
@@ -426,12 +473,36 @@ typedef enum SVGA3dSurfaceFlags {
            SVGA3D_SURFACE_STAGING_UPLOAD |          \
            SVGA3D_SURFACE_STAGING_DOWNLOAD |        \
            SVGA3D_SURFACE_HINT_INDIRECT_UPDATE |    \
-           SVGA3D_SURFACE_TRANSFER_FROM_BUFFER      \
+           SVGA3D_SURFACE_TRANSFER_FROM_BUFFER |    \
+           SVGA3D_SURFACE_MULTISAMPLE               \
+        )
+
+#define SVGA3D_SURFACE_BUFFER_DISALLOWED_MASK       \
+        (  SVGA3D_SURFACE_CUBEMAP |                 \
+           SVGA3D_SURFACE_AUTOGENMIPMAPS |          \
+           SVGA3D_SURFACE_VOLUME |                  \
+           SVGA3D_SURFACE_1D |                      \
+           SVGA3D_SURFACE_MASKABLE_ANTIALIAS |      \
+           SVGA3D_SURFACE_ARRAY |                   \
+           SVGA3D_SURFACE_MULTISAMPLE |             \
+           SVGA3D_SURFACE_MOB_PITCH                 \
+        )
+
+#define SVGA3D_SURFACE_MULTISAMPLE_DISALLOWED_MASK  \
+        (  SVGA3D_SURFACE_CUBEMAP |                 \
+           SVGA3D_SURFACE_AUTOGENMIPMAPS |          \
+           SVGA3D_SURFACE_VOLUME |                  \
+           SVGA3D_SURFACE_1D |                      \
+           SVGA3D_SURFACE_SCREENTARGET |            \
+           SVGA3D_SURFACE_MOB_PITCH                 \
         )
 
 #define SVGA3D_SURFACE_DX_ONLY_MASK             \
         (  SVGA3D_SURFACE_BIND_STREAM_OUTPUT |  \
+           SVGA3D_SURFACE_STAGING_UPLOAD |      \
+           SVGA3D_SURFACE_STAGING_DOWNLOAD |    \
            SVGA3D_SURFACE_TRANSFER_FROM_BUFFER  \
+        )
 
 #define SVGA3D_SURFACE_STAGING_MASK             \
         (  SVGA3D_SURFACE_STAGING_UPLOAD |      \
@@ -487,7 +558,7 @@ typedef enum {
 
 /*
  * Indicates that this format can be converted to any RGB format for which
- * SVGA3DFORMAT_OP_MEMBEROFGROUP_ARGB is specified
+ * SVGA3DFORMAT_OP_MEMBEROFGROUP_ARGB is specified.
  */
    SVGA3DFORMAT_OP_CONVERT_TO_ARGB                       = 0x00002000,
 
@@ -498,22 +569,22 @@ typedef enum {
 
 /*
  * Indicated that this format can be read as an SRGB texture (meaning that the
- * sampler will linearize the looked up data)
+ * sampler will linearize the looked up data).
  */
    SVGA3DFORMAT_OP_SRGBREAD                              = 0x00008000,
 
 /*
- * Indicates that this format can be used in the bumpmap instructions
+ * Indicates that this format can be used in the bumpmap instructions.
  */
    SVGA3DFORMAT_OP_BUMPMAP                               = 0x00010000,
 
 /*
- * Indicates that this format can be sampled by the displacement map sampler
+ * Indicates that this format can be sampled by the displacement map sampler.
  */
    SVGA3DFORMAT_OP_DMAP                                  = 0x00020000,
 
 /*
- * Indicates that this format cannot be used with texture filtering
+ * Indicates that this format cannot be used with texture filtering.
  */
    SVGA3DFORMAT_OP_NOFILTER                              = 0x00040000,
 
@@ -530,18 +601,18 @@ typedef enum {
    SVGA3DFORMAT_OP_SRGBWRITE                             = 0x00100000,
 
 /*
- * Indicates that this format cannot be used with alpha blending
+ * Indicates that this format cannot be used with alpha blending.
  */
    SVGA3DFORMAT_OP_NOALPHABLEND                          = 0x00200000,
 
 /*
  * Indicates that the device can auto-generated sublevels for resources
- * of this format
+ * of this format.
  */
    SVGA3DFORMAT_OP_AUTOGENMIPMAP                         = 0x00400000,
 
 /*
- * Indicates that this format can be used by vertex texture sampler
+ * Indicates that this format can be used by vertex texture sampler.
  */
    SVGA3DFORMAT_OP_VERTEXTEXTURE                         = 0x00800000,
 
@@ -1501,7 +1572,6 @@ union SVGADXQueryResultUnion {
 #include "vmware_pack_end.h"
 SVGADXQueryResultUnion;
 
-
 typedef enum {
    SVGA3D_QUERYSTATE_PENDING     = 0,      /* Query is not finished yet */
    SVGA3D_QUERYSTATE_SUCCEEDED   = 1,      /* Completed successfully */
@@ -1533,9 +1603,9 @@ typedef
 struct {
    union {
       struct {
-	 uint16  function;       /* SVGA3dFogFunction */
-	 uint8   type;           /* SVGA3dFogType */
-	 uint8   base;           /* SVGA3dFogBase */
+         uint16  function;       /* SVGA3dFogFunction */
+         uint8   type;           /* SVGA3dFogType */
+         uint8   base;           /* SVGA3dFogBase */
       };
       uint32     uintValue;
    };
@@ -1547,19 +1617,27 @@ SVGA3dFogMode;
  * Uniquely identify one image (a 1D/2D/3D array) from a surface. This
  * is a surface ID as well as face/mipmap indices.
  */
-
 typedef
 #include "vmware_pack_begin.h"
 struct SVGA3dSurfaceImageId {
-   uint32               sid;
-   uint32               face;
-   uint32               mipmap;
+   uint32 sid;
+   uint32 face;
+   uint32 mipmap;
 }
 #include "vmware_pack_end.h"
 SVGA3dSurfaceImageId;
 
 typedef
 #include "vmware_pack_begin.h"
+struct SVGA3dSubSurfaceId {
+   uint32 sid;
+   uint32 subResourceId;
+}
+#include "vmware_pack_end.h"
+SVGA3dSubSurfaceId;
+
+typedef
+#include "vmware_pack_begin.h"
 struct {
    uint32               width;
    uint32               height;
@@ -1582,13 +1660,18 @@ typedef enum {
    SVGA_OTABLE_DX9_MAX         = 5,
 
    SVGA_OTABLE_DXCONTEXT       = 5,
-   SVGA_OTABLE_MAX             = 6
-} SVGAOTableType;
+   SVGA_OTABLE_DX_MAX          = 6,
 
-/*
- * Deprecated.
- */
-#define SVGA_OTABLE_COUNT 4
+   SVGA_OTABLE_RESERVED1       = 6,
+   SVGA_OTABLE_RESERVED2       = 7,
+
+   /*
+    * Additions to this table need to be tied to HW-version features and
+    * checkpointed accordingly.
+    */
+   SVGA_OTABLE_DEVEL_MAX       = 8,
+   SVGA_OTABLE_MAX             = 8
+} SVGAOTableType;
 
 typedef enum {
    SVGA_COTABLE_MIN             = 0,
@@ -1605,7 +1688,7 @@ typedef enum {
    SVGA_COTABLE_DXSHADER        = 10,
    SVGA_COTABLE_DX10_MAX        = 11,
    SVGA_COTABLE_UAVIEW          = 11,
-   SVGA_COTABLE_MAX
+   SVGA_COTABLE_MAX             = 12,
 } SVGACOTableType;
 
 /*
@@ -1626,8 +1709,37 @@ typedef enum SVGAMobFormat {
    SVGA3D_MOBFMT_PREDX_MAX   = 7,
    SVGA3D_MOBFMT_EMPTY       = 7,
    SVGA3D_MOBFMT_MAX,
+
+   /*
+    * This isn't actually used by the guest, but is a mob-format used
+    * internally by the SVGA device (and is therefore not binary compatible).
+    */
+   SVGA3D_MOBFMT_HB,
 } SVGAMobFormat;
 
 #define SVGA3D_MOB_EMPTY_BASE 1
 
+/*
+ * Multisample pattern types.
+ */
+
+typedef enum SVGA3dMSPattern {
+   SVGA3D_MS_PATTERN_NONE     = 0,
+   SVGA3D_MS_PATTERN_MIN      = 0,
+   SVGA3D_MS_PATTERN_STANDARD = 1,
+   SVGA3D_MS_PATTERN_CENTER   = 2,
+   SVGA3D_MS_PATTERN_MAX      = 3,
+} SVGA3dMSPattern;
+
+/*
+ * Precision settings for each sample.
+ */
+
+typedef enum SVGA3dMSQualityLevel {
+   SVGA3D_MS_QUALITY_NONE = 0,
+   SVGA3D_MS_QUALITY_MIN  = 0,
+   SVGA3D_MS_QUALITY_FULL = 1,
+   SVGA3D_MS_QUALITY_MAX  = 2,
+} SVGA3dMSQualityLevel;
+
 #endif /* _SVGA3D_TYPES_H_ */
diff --git a/drivers/gpu/drm/vmwgfx/device_include/svga_escape.h b/drivers/gpu/drm/vmwgfx/device_include/svga_escape.h
index 884b1d1fb85f..acb41e28e46f 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/svga_escape.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/svga_escape.h
@@ -1,5 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**********************************************************
- * Copyright 2007-2015 VMware, Inc.  All rights reserved.
+ * Copyright 2007-2015 VMware, Inc.
  *
  * Permission is hereby granted, free of charge, to any person
  * obtaining a copy of this software and associated documentation
diff --git a/drivers/gpu/drm/vmwgfx/device_include/svga_overlay.h b/drivers/gpu/drm/vmwgfx/device_include/svga_overlay.h
index faf6d9b2b891..e5385146e7fc 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/svga_overlay.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/svga_overlay.h
@@ -1,5 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**********************************************************
- * Copyright 2007-2015 VMware, Inc.  All rights reserved.
+ * Copyright 2007-2015 VMware, Inc.
  *
  * Permission is hereby granted, free of charge, to any person
  * obtaining a copy of this software and associated documentation
diff --git a/drivers/gpu/drm/vmwgfx/device_include/svga_reg.h b/drivers/gpu/drm/vmwgfx/device_include/svga_reg.h
index 88e72bf9a534..056f54b35d73 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/svga_reg.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/svga_reg.h
@@ -1,5 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**********************************************************
- * Copyright 1998-2015 VMware, Inc.  All rights reserved.
+ * Copyright 1998-2015 VMware, Inc.
  *
  * Permission is hereby granted, free of charge, to any person
  * obtaining a copy of this software and associated documentation
@@ -63,16 +64,26 @@ typedef uint32 SVGAMobId;
 #define SVGA_MAX_BITS_PER_PIXEL         32
 #define SVGA_MAX_DEPTH                  24
 #define SVGA_MAX_DISPLAYS               10
+#define SVGA_MAX_SCREEN_SIZE            8192
+#define SVGA_SCREEN_ROOT_LIMIT (SVGA_MAX_SCREEN_SIZE * SVGA_MAX_DISPLAYS)
+
 
 /*
  * Legal values for the SVGA_REG_CURSOR_ON register in old-fashioned
  * cursor bypass mode. This is still supported, but no new guest
  * drivers should use it.
  */
-#define SVGA_CURSOR_ON_HIDE            0x0   /* Must be 0 to maintain backward compatibility */
-#define SVGA_CURSOR_ON_SHOW            0x1   /* Must be 1 to maintain backward compatibility */
-#define SVGA_CURSOR_ON_REMOVE_FROM_FB  0x2   /* Remove the cursor from the framebuffer because we need to see what's under it */
-#define SVGA_CURSOR_ON_RESTORE_TO_FB   0x3   /* Put the cursor back in the framebuffer so the user can see it */
+#define SVGA_CURSOR_ON_HIDE            0x0
+#define SVGA_CURSOR_ON_SHOW            0x1
+
+/*
+ * Remove the cursor from the framebuffer
+ * because we need to see what's under it
+ */
+#define SVGA_CURSOR_ON_REMOVE_FROM_FB  0x2
+
+/* Put the cursor back in the framebuffer so the user can see it */
+#define SVGA_CURSOR_ON_RESTORE_TO_FB   0x3
 
 /*
  * The maximum framebuffer size that can traced for guests unless the
@@ -101,7 +112,10 @@ typedef uint32 SVGAMobId;
 #define SVGA_VERSION_0     0
 #define SVGA_ID_0          SVGA_MAKE_ID(SVGA_VERSION_0)
 
-/* "Invalid" value for all SVGA IDs. (Version ID, screen object ID, surface ID...) */
+/*
+ * "Invalid" value for all SVGA IDs.
+ * (Version ID, screen object ID, surface ID...)
+ */
 #define SVGA_ID_INVALID    0xFFFFFFFF
 
 /* Port offsets, relative to BAR0 */
@@ -154,7 +168,7 @@ enum {
    SVGA_REG_CONFIG_DONE = 20,         /* Set when memory area configured */
    SVGA_REG_SYNC = 21,                /* See "FIFO Synchronization Registers" */
    SVGA_REG_BUSY = 22,                /* See "FIFO Synchronization Registers" */
-   SVGA_REG_GUEST_ID = 23,            /* Set guest OS identifier */
+   SVGA_REG_GUEST_ID = 23,            /* (Deprecated) */
    SVGA_REG_CURSOR_ID = 24,           /* (Deprecated) */
    SVGA_REG_CURSOR_X = 25,            /* (Deprecated) */
    SVGA_REG_CURSOR_Y = 26,            /* (Deprecated) */
@@ -186,7 +200,14 @@ enum {
    SVGA_REG_MEMORY_SIZE = 47,       /* Total dedicated device memory excluding FIFO */
    SVGA_REG_COMMAND_LOW = 48,       /* Lower 32 bits and submits commands */
    SVGA_REG_COMMAND_HIGH = 49,      /* Upper 32 bits of command buffer PA */
-   SVGA_REG_MAX_PRIMARY_BOUNDING_BOX_MEM = 50,   /* Max primary memory */
+
+   /*
+    * Max primary memory.
+    * See SVGA_CAP_NO_BB_RESTRICTION.
+    */
+   SVGA_REG_MAX_PRIMARY_MEM = 50,
+   SVGA_REG_MAX_PRIMARY_BOUNDING_BOX_MEM = 50,
+
    SVGA_REG_SUGGESTED_GBOBJECT_MEM_SIZE_KB = 51, /* Sugested limit on mob mem */
    SVGA_REG_DEV_CAP = 52,           /* Write dev cap index, read value */
    SVGA_REG_CMD_PREPEND_LOW = 53,
@@ -194,7 +215,10 @@ enum {
    SVGA_REG_SCREENTARGET_MAX_WIDTH = 55,
    SVGA_REG_SCREENTARGET_MAX_HEIGHT = 56,
    SVGA_REG_MOB_MAX_SIZE = 57,
-   SVGA_REG_TOP = 58,               /* Must be 1 more than the last register */
+   SVGA_REG_BLANK_SCREEN_TARGETS = 58,
+   SVGA_REG_CAP2 = 59,
+   SVGA_REG_DEVEL_CAP = 60,
+   SVGA_REG_TOP = 61,               /* Must be 1 more than the last register */
 
    SVGA_PALETTE_BASE = 1024,        /* Base of SVGA color map */
    /* Next 768 (== 256*3) registers exist for colormap */
@@ -392,6 +416,7 @@ typedef enum {
    SVGA_CB_CONTEXT_0      = 0x0,
    SVGA_CB_CONTEXT_1      = 0x1, /* Supported with SVGA_CAP_HP_CMD_QUEUE */
    SVGA_CB_CONTEXT_MAX    = 0x2,
+   SVGA_CB_CONTEXT_HP_MAX = 0x2,
 } SVGACBContext;
 
 
@@ -448,6 +473,18 @@ typedef enum {
     * due to an error.  No IRQ is raised.
     */
    SVGA_CB_STATUS_SUBMISSION_ERROR = 6,
+
+   /*
+    * Written by the host when the host finished a
+    * SVGA_DC_CMD_ASYNC_STOP_QUEUE request for this command buffer
+    * queue.  The offset of the first byte not processed is stored in
+    * the errorOffset field of the command buffer header.  All guest
+    * visible side effects of commands till that point are guaranteed
+    * to be finished before this is written.  The
+    * SVGA_IRQFLAG_COMMAND_BUFFER IRQ is raised as long as the
+    * SVGA_CB_FLAG_NO_IRQ is not set.
+    */
+   SVGA_CB_STATUS_PARTIAL_COMPLETE = 7,
 } SVGACBStatus;
 
 typedef enum {
@@ -460,8 +497,8 @@ typedef enum {
 typedef
 #include "vmware_pack_begin.h"
 struct {
-   volatile SVGACBStatus status;
-   volatile uint32 errorOffset;
+   volatile SVGACBStatus status; /* Modified by device. */
+   volatile uint32 errorOffset;  /* Modified by device. */
    uint64 id;
    SVGACBFlags flags;
    uint32 length;
@@ -472,7 +509,9 @@ struct {
          uint32 mobOffset;
       } mob;
    } ptr;
-   uint32 offset; /* Valid if CMD_BUFFERS_2 cap set, must be zero otherwise */
+   uint32 offset; /* Valid if CMD_BUFFERS_2 cap set, must be zero otherwise,
+                   * modified by device.
+                   */
    uint32 dxContext; /* Valid if DX_CONTEXT flag set, must be zero otherwise */
    uint32 mustBeZero[6];
 }
@@ -483,20 +522,26 @@ typedef enum {
    SVGA_DC_CMD_NOP                   = 0,
    SVGA_DC_CMD_START_STOP_CONTEXT    = 1,
    SVGA_DC_CMD_PREEMPT               = 2,
-   SVGA_DC_CMD_MAX                   = 3,
-   SVGA_DC_CMD_FORCE_UINT            = MAX_UINT32,
+   SVGA_DC_CMD_START_QUEUE           = 3, /* Requires SVGA_CAP_HP_CMD_QUEUE */
+   SVGA_DC_CMD_ASYNC_STOP_QUEUE      = 4, /* Requires SVGA_CAP_HP_CMD_QUEUE */
+   SVGA_DC_CMD_EMPTY_CONTEXT_QUEUE   = 5, /* Requires SVGA_CAP_HP_CMD_QUEUE */
+   SVGA_DC_CMD_MAX                   = 6,
 } SVGADeviceContextCmdId;
 
-typedef struct {
+/*
+ * Starts or stops both SVGA_CB_CONTEXT_0 and SVGA_CB_CONTEXT_1.
+ */
+
+typedef struct SVGADCCmdStartStop {
    uint32 enable;
-   SVGACBContext context;
+   SVGACBContext context; /* Must be zero */
 } SVGADCCmdStartStop;
 
 /*
  * SVGADCCmdPreempt --
  *
  * This command allows the guest to request that all command buffers
- * on the specified context be preempted that can be.  After execution
+ * on SVGA_CB_CONTEXT_0 be preempted that can be.  After execution
  * of this command all command buffers that were preempted will
  * already have SVGA_CB_STATUS_PREEMPTED written into the status
  * field.  The device might still be processing a command buffer,
@@ -506,12 +551,69 @@ typedef struct {
  * command buffer header set to zero.
  */
 
-typedef struct {
-   SVGACBContext context;
+typedef struct SVGADCCmdPreempt {
+   SVGACBContext context; /* Must be zero */
    uint32 ignoreIDZero;
 } SVGADCCmdPreempt;
 
 /*
+ * Starts the requested command buffer processing queue.  Valid only
+ * if the SVGA_CAP_HP_CMD_QUEUE cap is set.
+ *
+ * For a command queue to be considered runnable it must be enabled
+ * and any corresponding higher priority queues must also be enabled.
+ * For example in order for command buffers to be processed on
+ * SVGA_CB_CONTEXT_0 both SVGA_CB_CONTEXT_0 and SVGA_CB_CONTEXT_1 must
+ * be enabled.  But for commands to be runnable on SVGA_CB_CONTEXT_1
+ * only that queue must be enabled.
+ */
+
+typedef struct SVGADCCmdStartQueue {
+   SVGACBContext context;
+} SVGADCCmdStartQueue;
+
+/*
+ * Requests the SVGA device to stop processing the requested command
+ * buffer queue as soon as possible.  The guest knows the stop has
+ * completed when one of the following happens.
+ *
+ * 1) A command buffer status of SVGA_CB_STATUS_PARTIAL_COMPLETE is returned
+ * 2) A command buffer error is encountered with would stop the queue
+ *    regardless of the async stop request.
+ * 3) All command buffers that have been submitted complete successfully.
+ * 4) The stop completes synchronously if no command buffers are
+ *    active on the queue when it is issued.
+ *
+ * If the command queue is not in a runnable state there is no
+ * guarentee this async stop will finish.  For instance if the high
+ * priority queue is not enabled and a stop is requested on the low
+ * priority queue, the high priority queue must be reenabled to
+ * guarantee that the async stop will finish.
+ *
+ * This command along with SVGA_DC_CMD_EMPTY_CONTEXT_QUEUE can be used
+ * to implement mid command buffer preemption.
+ *
+ * Valid only if the SVGA_CAP_HP_CMD_QUEUE cap is set.
+ */
+
+typedef struct SVGADCCmdAsyncStopQueue {
+   SVGACBContext context;
+} SVGADCCmdAsyncStopQueue;
+
+/*
+ * Requests the SVGA device to throw away any full command buffers on
+ * the requested command queue that have not been started.  For a
+ * driver to know which command buffers were thrown away a driver
+ * should only issue this command when the queue is stopped, for
+ * whatever reason.
+ */
+
+typedef struct SVGADCCmdEmptyQueue {
+   SVGACBContext context;
+} SVGADCCmdEmptyQueue;
+
+
+/*
  * SVGAGMRImageFormat --
  *
  *    This is a packed representation of the source 2D image format
@@ -536,7 +638,7 @@ typedef struct SVGAGMRImageFormat {
       struct {
          uint32 bitsPerPixel : 8;
          uint32 colorDepth   : 8;
-	 uint32 reserved     : 16;  /* Must be zero */
+         uint32 reserved     : 16;  /* Must be zero */
       };
 
       uint32 value;
@@ -672,8 +774,36 @@ SVGASignedPoint;
  * SVGA_CAP_GBOBJECTS --
  *    Enable guest-backed objects and surfaces.
  *
- * SVGA_CAP_CMD_BUFFERS_3 --
- *    Enable support for command buffers in a mob.
+ * SVGA_CAP_DX --
+ *    Enable support for DX commands, and command buffers in a mob.
+ *
+ * SVGA_CAP_HP_CMD_QUEUE --
+ *    Enable support for the high priority command queue, and the
+ *    ScreenCopy command.
+ *
+ * SVGA_CAP_NO_BB_RESTRICTION --
+ *    Allow ScreenTargets to be defined without regard to the 32-bpp
+ *    bounding-box memory restrictions. ie:
+ *
+ *    The summed memory usage of all screens (assuming they were defined as
+ *    32-bpp) must always be less than the value of the
+ *    SVGA_REG_MAX_PRIMARY_MEM register.
+ *
+ *    If this cap is not present, the 32-bpp bounding box around all screens
+ *    must additionally be under the value of the SVGA_REG_MAX_PRIMARY_MEM
+ *    register.
+ *
+ *    If the cap is present, the bounding box restriction is lifted (and only
+ *    the screen-sum limit applies).
+ *
+ *    (Note that this is a slight lie... there is still a sanity limit on any
+ *     dimension of the topology to be less than SVGA_SCREEN_ROOT_LIMIT, even
+ *     when SVGA_CAP_NO_BB_RESTRICTION is present, but that should be
+ *     large enough to express any possible topology without holes between
+ *     monitors.)
+ *
+ * SVGA_CAP_CAP2_REGISTER --
+ *    If this cap is present, the SVGA_REG_CAP2 register is supported.
  */
 
 #define SVGA_CAP_NONE               0x00000000
@@ -699,8 +829,30 @@ SVGASignedPoint;
 #define SVGA_CAP_GBOBJECTS          0x08000000
 #define SVGA_CAP_DX                 0x10000000
 #define SVGA_CAP_HP_CMD_QUEUE       0x20000000
+#define SVGA_CAP_NO_BB_RESTRICTION  0x40000000
+#define SVGA_CAP_CAP2_REGISTER      0x80000000
 
-#define SVGA_CAP_CMD_RESERVED       0x80000000
+/*
+ * The SVGA_REG_CAP2 register is an additional set of SVGA capability bits.
+ *
+ * SVGA_CAP2_GROW_OTABLE --
+ *      Allow the GrowOTable/DXGrowCOTable commands.
+ *
+ * SVGA_CAP2_INTRA_SURFACE_COPY --
+ *      Allow the IntraSurfaceCopy command.
+ *
+ * SVGA_CAP2_DX2 --
+ *      Allow the DefineGBSurface_v3, WholeSurfaceCopy.
+ *
+ * SVGA_CAP2_RESERVED --
+ *      Reserve the last bit for extending the SVGA capabilities to some
+ *      future mechanisms.
+ */
+#define SVGA_CAP2_NONE               0x00000000
+#define SVGA_CAP2_GROW_OTABLE        0x00000001
+#define SVGA_CAP2_INTRA_SURFACE_COPY 0x00000002
+#define SVGA_CAP2_DX2                0x00000004
+#define SVGA_CAP2_RESERVED           0x80000000
 
 
 /*
@@ -722,7 +874,8 @@ typedef enum {
    SVGABackdoorCapDeviceCaps = 0,
    SVGABackdoorCapFifoCaps = 1,
    SVGABackdoorCap3dHWVersion = 2,
-   SVGABackdoorCapMax = 3,
+   SVGABackdoorCapDeviceCaps2 = 3,
+   SVGABackdoorCapMax = 4,
 } SVGABackdoorCapType;
 
 
@@ -1914,16 +2067,6 @@ SVGAFifoCmdRemapGMR2;
 
 #define SVGA_VRAM_SIZE_W2K          (64 * 1024 * 1024) /* 64 MB */
 
-/*
- * To simplify autoDetect display configuration, support a minimum of
- * two 1920x1200 monitors, 32bpp, side-by-side, optionally rotated:
- *   numDisplays = 2
- *   maxWidth = numDisplay * 1920 = 3840
- *   maxHeight = rotated width of single monitor = 1920
- *   vramSize = maxWidth * maxHeight * 4 = 29491200
- */
-#define SVGA_VRAM_SIZE_AUTODETECT   (32 * 1024 * 1024)
-
 #if defined(VMX86_SERVER)
 #define SVGA_VRAM_SIZE               (4 * 1024 * 1024)
 #define SVGA_VRAM_SIZE_3D           (64 * 1024 * 1024)
diff --git a/drivers/gpu/drm/vmwgfx/device_include/svga_types.h b/drivers/gpu/drm/vmwgfx/device_include/svga_types.h
index 2e8ba4df8de9..350bbc6fab02 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/svga_types.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/svga_types.h
@@ -1,5 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**********************************************************
- * Copyright 2015 VMware, Inc.  All rights reserved.
+ * Copyright 2015 VMware, Inc.
  *
  * Permission is hereby granted, free of charge, to any person
  * obtaining a copy of this software and associated documentation
@@ -40,7 +41,10 @@ typedef uint64 PPN64;
 
 typedef bool Bool;
 
+#define MAX_UINT64 U64_MAX
 #define MAX_UINT32 U32_MAX
 #define MAX_UINT16 U16_MAX
 
+#define CONST64U(x) x##ULL
+
 #endif
diff --git a/drivers/gpu/drm/vmwgfx/device_include/vmware_pack_begin.h b/drivers/gpu/drm/vmwgfx/device_include/vmware_pack_begin.h
index 7e7b0ce34aa2..75308bd0d970 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/vmware_pack_begin.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/vmware_pack_begin.h
@@ -1,25 +1,2 @@
-/**********************************************************
- * Copyright 2015 VMware, Inc.  All rights reserved.
- *
- * Permission is hereby granted, free of charge, to any person
- * obtaining a copy of this software and associated documentation
- * files (the "Software"), to deal in the Software without
- * restriction, including without limitation the rights to use, copy,
- * modify, merge, publish, distribute, sublicense, and/or sell copies
- * of the Software, and to permit persons to whom the Software is
- * furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be
- * included in all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
- * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
- * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
- * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
- * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
- * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
- * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- *
- **********************************************************/
+/* SPDX-License-Identifier: GPL-2.0 */
 #include <linux/compiler.h>
diff --git a/drivers/gpu/drm/vmwgfx/device_include/vmware_pack_end.h b/drivers/gpu/drm/vmwgfx/device_include/vmware_pack_end.h
index e2e440ed3d44..e93d6f28b68c 100644
--- a/drivers/gpu/drm/vmwgfx/device_include/vmware_pack_end.h
+++ b/drivers/gpu/drm/vmwgfx/device_include/vmware_pack_end.h
@@ -1,25 +1,2 @@
-/**********************************************************
- * Copyright 2015 VMware, Inc.  All rights reserved.
- *
- * Permission is hereby granted, free of charge, to any person
- * obtaining a copy of this software and associated documentation
- * files (the "Software"), to deal in the Software without
- * restriction, including without limitation the rights to use, copy,
- * modify, merge, publish, distribute, sublicense, and/or sell copies
- * of the Software, and to permit persons to whom the Software is
- * furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be
- * included in all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
- * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
- * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
- * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
- * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
- * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
- * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
- * SOFTWARE.
- *
- **********************************************************/
+/* SPDX-License-Identifier: GPL-2.0 */
 __packed
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_binding.c b/drivers/gpu/drm/vmwgfx/vmwgfx_binding.c
index 55d32ae43aa4..0b9ee7fb45d6 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_binding.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_binding.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_binding.h b/drivers/gpu/drm/vmwgfx/vmwgfx_binding.h
index bf2e77ad5a20..6a2a9d69043b 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_binding.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_binding.h
@@ -1,7 +1,7 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
- * Copyright © 2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_blit.c b/drivers/gpu/drm/vmwgfx/vmwgfx_blit.c
index e8c94b19db7b..fc6673cde289 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_blit.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_blit.c
@@ -1,6 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2017 VMware, Inc., Palo Alto, CA., USA
+ * Copyright 2017 VMware, Inc., Palo Alto, CA., USA
  * All Rights Reserved.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c
new file mode 100644
index 000000000000..2dda03345761
--- /dev/null
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c
@@ -0,0 +1,1123 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
+/**************************************************************************
+ *
+ * Copyright © 2011-2018 VMware, Inc., Palo Alto, CA., USA
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ **************************************************************************/
+
+#include <drm/ttm/ttm_placement.h>
+
+#include <drm/drmP.h>
+#include "vmwgfx_drv.h"
+#include "drm/ttm/ttm_object.h"
+
+
+/**
+ * struct vmw_user_buffer_object - User-space-visible buffer object
+ *
+ * @prime: The prime object providing user visibility.
+ * @vbo: The struct vmw_buffer_object
+ */
+struct vmw_user_buffer_object {
+	struct ttm_prime_object prime;
+	struct vmw_buffer_object vbo;
+};
+
+
+/**
+ * vmw_buffer_object - Convert a struct ttm_buffer_object to a struct
+ * vmw_buffer_object.
+ *
+ * @bo: Pointer to the TTM buffer object.
+ * Return: Pointer to the struct vmw_buffer_object embedding the
+ * TTM buffer object.
+ */
+static struct vmw_buffer_object *
+vmw_buffer_object(struct ttm_buffer_object *bo)
+{
+	return container_of(bo, struct vmw_buffer_object, base);
+}
+
+
+/**
+ * vmw_user_buffer_object - Convert a struct ttm_buffer_object to a struct
+ * vmw_user_buffer_object.
+ *
+ * @bo: Pointer to the TTM buffer object.
+ * Return: Pointer to the struct vmw_buffer_object embedding the TTM buffer
+ * object.
+ */
+static struct vmw_user_buffer_object *
+vmw_user_buffer_object(struct ttm_buffer_object *bo)
+{
+	struct vmw_buffer_object *vmw_bo = vmw_buffer_object(bo);
+
+	return container_of(vmw_bo, struct vmw_user_buffer_object, vbo);
+}
+
+
+/**
+ * vmw_bo_pin_in_placement - Validate a buffer to placement.
+ *
+ * @dev_priv:  Driver private.
+ * @buf:  DMA buffer to move.
+ * @placement:  The placement to pin it.
+ * @interruptible:  Use interruptible wait.
+ * Return: Zero on success, Negative error code on failure. In particular
+ * -ERESTARTSYS if interrupted by a signal
+ */
+int vmw_bo_pin_in_placement(struct vmw_private *dev_priv,
+			    struct vmw_buffer_object *buf,
+			    struct ttm_placement *placement,
+			    bool interruptible)
+{
+	struct ttm_operation_ctx ctx = {interruptible, false };
+	struct ttm_buffer_object *bo = &buf->base;
+	int ret;
+	uint32_t new_flags;
+
+	ret = ttm_write_lock(&dev_priv->reservation_sem, interruptible);
+	if (unlikely(ret != 0))
+		return ret;
+
+	vmw_execbuf_release_pinned_bo(dev_priv);
+
+	ret = ttm_bo_reserve(bo, interruptible, false, NULL);
+	if (unlikely(ret != 0))
+		goto err;
+
+	if (buf->pin_count > 0)
+		ret = ttm_bo_mem_compat(placement, &bo->mem,
+					&new_flags) == true ? 0 : -EINVAL;
+	else
+		ret = ttm_bo_validate(bo, placement, &ctx);
+
+	if (!ret)
+		vmw_bo_pin_reserved(buf, true);
+
+	ttm_bo_unreserve(bo);
+
+err:
+	ttm_write_unlock(&dev_priv->reservation_sem);
+	return ret;
+}
+
+
+/**
+ * vmw_bo_pin_in_vram_or_gmr - Move a buffer to vram or gmr.
+ *
+ * This function takes the reservation_sem in write mode.
+ * Flushes and unpins the query bo to avoid failures.
+ *
+ * @dev_priv:  Driver private.
+ * @buf:  DMA buffer to move.
+ * @pin:  Pin buffer if true.
+ * @interruptible:  Use interruptible wait.
+ * Return: Zero on success, Negative error code on failure. In particular
+ * -ERESTARTSYS if interrupted by a signal
+ */
+int vmw_bo_pin_in_vram_or_gmr(struct vmw_private *dev_priv,
+			      struct vmw_buffer_object *buf,
+			      bool interruptible)
+{
+	struct ttm_operation_ctx ctx = {interruptible, false };
+	struct ttm_buffer_object *bo = &buf->base;
+	int ret;
+	uint32_t new_flags;
+
+	ret = ttm_write_lock(&dev_priv->reservation_sem, interruptible);
+	if (unlikely(ret != 0))
+		return ret;
+
+	vmw_execbuf_release_pinned_bo(dev_priv);
+
+	ret = ttm_bo_reserve(bo, interruptible, false, NULL);
+	if (unlikely(ret != 0))
+		goto err;
+
+	if (buf->pin_count > 0) {
+		ret = ttm_bo_mem_compat(&vmw_vram_gmr_placement, &bo->mem,
+					&new_flags) == true ? 0 : -EINVAL;
+		goto out_unreserve;
+	}
+
+	ret = ttm_bo_validate(bo, &vmw_vram_gmr_placement, &ctx);
+	if (likely(ret == 0) || ret == -ERESTARTSYS)
+		goto out_unreserve;
+
+	ret = ttm_bo_validate(bo, &vmw_vram_placement, &ctx);
+
+out_unreserve:
+	if (!ret)
+		vmw_bo_pin_reserved(buf, true);
+
+	ttm_bo_unreserve(bo);
+err:
+	ttm_write_unlock(&dev_priv->reservation_sem);
+	return ret;
+}
+
+
+/**
+ * vmw_bo_pin_in_vram - Move a buffer to vram.
+ *
+ * This function takes the reservation_sem in write mode.
+ * Flushes and unpins the query bo to avoid failures.
+ *
+ * @dev_priv:  Driver private.
+ * @buf:  DMA buffer to move.
+ * @interruptible:  Use interruptible wait.
+ * Return: Zero on success, Negative error code on failure. In particular
+ * -ERESTARTSYS if interrupted by a signal
+ */
+int vmw_bo_pin_in_vram(struct vmw_private *dev_priv,
+		       struct vmw_buffer_object *buf,
+		       bool interruptible)
+{
+	return vmw_bo_pin_in_placement(dev_priv, buf, &vmw_vram_placement,
+				       interruptible);
+}
+
+
+/**
+ * vmw_bo_pin_in_start_of_vram - Move a buffer to start of vram.
+ *
+ * This function takes the reservation_sem in write mode.
+ * Flushes and unpins the query bo to avoid failures.
+ *
+ * @dev_priv:  Driver private.
+ * @buf:  DMA buffer to pin.
+ * @interruptible:  Use interruptible wait.
+ * Return: Zero on success, Negative error code on failure. In particular
+ * -ERESTARTSYS if interrupted by a signal
+ */
+int vmw_bo_pin_in_start_of_vram(struct vmw_private *dev_priv,
+				struct vmw_buffer_object *buf,
+				bool interruptible)
+{
+	struct ttm_operation_ctx ctx = {interruptible, false };
+	struct ttm_buffer_object *bo = &buf->base;
+	struct ttm_placement placement;
+	struct ttm_place place;
+	int ret = 0;
+	uint32_t new_flags;
+
+	place = vmw_vram_placement.placement[0];
+	place.lpfn = bo->num_pages;
+	placement.num_placement = 1;
+	placement.placement = &place;
+	placement.num_busy_placement = 1;
+	placement.busy_placement = &place;
+
+	ret = ttm_write_lock(&dev_priv->reservation_sem, interruptible);
+	if (unlikely(ret != 0))
+		return ret;
+
+	vmw_execbuf_release_pinned_bo(dev_priv);
+	ret = ttm_bo_reserve(bo, interruptible, false, NULL);
+	if (unlikely(ret != 0))
+		goto err_unlock;
+
+	/*
+	 * Is this buffer already in vram but not at the start of it?
+	 * In that case, evict it first because TTM isn't good at handling
+	 * that situation.
+	 */
+	if (bo->mem.mem_type == TTM_PL_VRAM &&
+	    bo->mem.start < bo->num_pages &&
+	    bo->mem.start > 0 &&
+	    buf->pin_count == 0) {
+		ctx.interruptible = false;
+		(void) ttm_bo_validate(bo, &vmw_sys_placement, &ctx);
+	}
+
+	if (buf->pin_count > 0)
+		ret = ttm_bo_mem_compat(&placement, &bo->mem,
+					&new_flags) == true ? 0 : -EINVAL;
+	else
+		ret = ttm_bo_validate(bo, &placement, &ctx);
+
+	/* For some reason we didn't end up at the start of vram */
+	WARN_ON(ret == 0 && bo->offset != 0);
+	if (!ret)
+		vmw_bo_pin_reserved(buf, true);
+
+	ttm_bo_unreserve(bo);
+err_unlock:
+	ttm_write_unlock(&dev_priv->reservation_sem);
+
+	return ret;
+}
+
+
+/**
+ * vmw_bo_unpin - Unpin the buffer given buffer, does not move the buffer.
+ *
+ * This function takes the reservation_sem in write mode.
+ *
+ * @dev_priv:  Driver private.
+ * @buf:  DMA buffer to unpin.
+ * @interruptible:  Use interruptible wait.
+ * Return: Zero on success, Negative error code on failure. In particular
+ * -ERESTARTSYS if interrupted by a signal
+ */
+int vmw_bo_unpin(struct vmw_private *dev_priv,
+		 struct vmw_buffer_object *buf,
+		 bool interruptible)
+{
+	struct ttm_buffer_object *bo = &buf->base;
+	int ret;
+
+	ret = ttm_read_lock(&dev_priv->reservation_sem, interruptible);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = ttm_bo_reserve(bo, interruptible, false, NULL);
+	if (unlikely(ret != 0))
+		goto err;
+
+	vmw_bo_pin_reserved(buf, false);
+
+	ttm_bo_unreserve(bo);
+
+err:
+	ttm_read_unlock(&dev_priv->reservation_sem);
+	return ret;
+}
+
+/**
+ * vmw_bo_get_guest_ptr - Get the guest ptr representing the current placement
+ * of a buffer.
+ *
+ * @bo: Pointer to a struct ttm_buffer_object. Must be pinned or reserved.
+ * @ptr: SVGAGuestPtr returning the result.
+ */
+void vmw_bo_get_guest_ptr(const struct ttm_buffer_object *bo,
+			  SVGAGuestPtr *ptr)
+{
+	if (bo->mem.mem_type == TTM_PL_VRAM) {
+		ptr->gmrId = SVGA_GMR_FRAMEBUFFER;
+		ptr->offset = bo->offset;
+	} else {
+		ptr->gmrId = bo->mem.start;
+		ptr->offset = 0;
+	}
+}
+
+
+/**
+ * vmw_bo_pin_reserved - Pin or unpin a buffer object without moving it.
+ *
+ * @vbo: The buffer object. Must be reserved.
+ * @pin: Whether to pin or unpin.
+ *
+ */
+void vmw_bo_pin_reserved(struct vmw_buffer_object *vbo, bool pin)
+{
+	struct ttm_operation_ctx ctx = { false, true };
+	struct ttm_place pl;
+	struct ttm_placement placement;
+	struct ttm_buffer_object *bo = &vbo->base;
+	uint32_t old_mem_type = bo->mem.mem_type;
+	int ret;
+
+	lockdep_assert_held(&bo->resv->lock.base);
+
+	if (pin) {
+		if (vbo->pin_count++ > 0)
+			return;
+	} else {
+		WARN_ON(vbo->pin_count <= 0);
+		if (--vbo->pin_count > 0)
+			return;
+	}
+
+	pl.fpfn = 0;
+	pl.lpfn = 0;
+	pl.flags = TTM_PL_FLAG_VRAM | VMW_PL_FLAG_GMR | VMW_PL_FLAG_MOB
+		| TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED;
+	if (pin)
+		pl.flags |= TTM_PL_FLAG_NO_EVICT;
+
+	memset(&placement, 0, sizeof(placement));
+	placement.num_placement = 1;
+	placement.placement = &pl;
+
+	ret = ttm_bo_validate(bo, &placement, &ctx);
+
+	BUG_ON(ret != 0 || bo->mem.mem_type != old_mem_type);
+}
+
+
+/**
+ * vmw_bo_map_and_cache - Map a buffer object and cache the map
+ *
+ * @vbo: The buffer object to map
+ * Return: A kernel virtual address or NULL if mapping failed.
+ *
+ * This function maps a buffer object into the kernel address space, or
+ * returns the virtual kernel address of an already existing map. The virtual
+ * address remains valid as long as the buffer object is pinned or reserved.
+ * The cached map is torn down on either
+ * 1) Buffer object move
+ * 2) Buffer object swapout
+ * 3) Buffer object destruction
+ *
+ */
+void *vmw_bo_map_and_cache(struct vmw_buffer_object *vbo)
+{
+	struct ttm_buffer_object *bo = &vbo->base;
+	bool not_used;
+	void *virtual;
+	int ret;
+
+	virtual = ttm_kmap_obj_virtual(&vbo->map, &not_used);
+	if (virtual)
+		return virtual;
+
+	ret = ttm_bo_kmap(bo, 0, bo->num_pages, &vbo->map);
+	if (ret)
+		DRM_ERROR("Buffer object map failed: %d.\n", ret);
+
+	return ttm_kmap_obj_virtual(&vbo->map, &not_used);
+}
+
+
+/**
+ * vmw_bo_unmap - Tear down a cached buffer object map.
+ *
+ * @vbo: The buffer object whose map we are tearing down.
+ *
+ * This function tears down a cached map set up using
+ * vmw_buffer_object_map_and_cache().
+ */
+void vmw_bo_unmap(struct vmw_buffer_object *vbo)
+{
+	if (vbo->map.bo == NULL)
+		return;
+
+	ttm_bo_kunmap(&vbo->map);
+}
+
+
+/**
+ * vmw_bo_acc_size - Calculate the pinned memory usage of buffers
+ *
+ * @dev_priv: Pointer to a struct vmw_private identifying the device.
+ * @size: The requested buffer size.
+ * @user: Whether this is an ordinary dma buffer or a user dma buffer.
+ */
+static size_t vmw_bo_acc_size(struct vmw_private *dev_priv, size_t size,
+			      bool user)
+{
+	static size_t struct_size, user_struct_size;
+	size_t num_pages = PAGE_ALIGN(size) >> PAGE_SHIFT;
+	size_t page_array_size = ttm_round_pot(num_pages * sizeof(void *));
+
+	if (unlikely(struct_size == 0)) {
+		size_t backend_size = ttm_round_pot(vmw_tt_size);
+
+		struct_size = backend_size +
+			ttm_round_pot(sizeof(struct vmw_buffer_object));
+		user_struct_size = backend_size +
+			ttm_round_pot(sizeof(struct vmw_user_buffer_object));
+	}
+
+	if (dev_priv->map_mode == vmw_dma_alloc_coherent)
+		page_array_size +=
+			ttm_round_pot(num_pages * sizeof(dma_addr_t));
+
+	return ((user) ? user_struct_size : struct_size) +
+		page_array_size;
+}
+
+
+/**
+ * vmw_bo_bo_free - vmw buffer object destructor
+ *
+ * @bo: Pointer to the embedded struct ttm_buffer_object
+ */
+void vmw_bo_bo_free(struct ttm_buffer_object *bo)
+{
+	struct vmw_buffer_object *vmw_bo = vmw_buffer_object(bo);
+
+	vmw_bo_unmap(vmw_bo);
+	kfree(vmw_bo);
+}
+
+
+/**
+ * vmw_user_bo_destroy - vmw buffer object destructor
+ *
+ * @bo: Pointer to the embedded struct ttm_buffer_object
+ */
+static void vmw_user_bo_destroy(struct ttm_buffer_object *bo)
+{
+	struct vmw_user_buffer_object *vmw_user_bo = vmw_user_buffer_object(bo);
+
+	vmw_bo_unmap(&vmw_user_bo->vbo);
+	ttm_prime_object_kfree(vmw_user_bo, prime);
+}
+
+
+/**
+ * vmw_bo_init - Initialize a vmw buffer object
+ *
+ * @dev_priv: Pointer to the device private struct
+ * @vmw_bo: Pointer to the struct vmw_buffer_object to initialize.
+ * @size: Buffer object size in bytes.
+ * @placement: Initial placement.
+ * @interruptible: Whether waits should be performed interruptible.
+ * @bo_free: The buffer object destructor.
+ * Returns: Zero on success, negative error code on error.
+ *
+ * Note that on error, the code will free the buffer object.
+ */
+int vmw_bo_init(struct vmw_private *dev_priv,
+		struct vmw_buffer_object *vmw_bo,
+		size_t size, struct ttm_placement *placement,
+		bool interruptible,
+		void (*bo_free)(struct ttm_buffer_object *bo))
+{
+	struct ttm_bo_device *bdev = &dev_priv->bdev;
+	size_t acc_size;
+	int ret;
+	bool user = (bo_free == &vmw_user_bo_destroy);
+
+	WARN_ON_ONCE(!bo_free && (!user && (bo_free != vmw_bo_bo_free)));
+
+	acc_size = vmw_bo_acc_size(dev_priv, size, user);
+	memset(vmw_bo, 0, sizeof(*vmw_bo));
+
+	INIT_LIST_HEAD(&vmw_bo->res_list);
+
+	ret = ttm_bo_init(bdev, &vmw_bo->base, size,
+			  ttm_bo_type_device, placement,
+			  0, interruptible, acc_size,
+			  NULL, NULL, bo_free);
+	return ret;
+}
+
+
+/**
+ * vmw_user_bo_release - TTM reference base object release callback for
+ * vmw user buffer objects
+ *
+ * @p_base: The TTM base object pointer about to be unreferenced.
+ *
+ * Clears the TTM base object pointer and drops the reference the
+ * base object has on the underlying struct vmw_buffer_object.
+ */
+static void vmw_user_bo_release(struct ttm_base_object **p_base)
+{
+	struct vmw_user_buffer_object *vmw_user_bo;
+	struct ttm_base_object *base = *p_base;
+	struct ttm_buffer_object *bo;
+
+	*p_base = NULL;
+
+	if (unlikely(base == NULL))
+		return;
+
+	vmw_user_bo = container_of(base, struct vmw_user_buffer_object,
+				   prime.base);
+	bo = &vmw_user_bo->vbo.base;
+	ttm_bo_unref(&bo);
+}
+
+
+/**
+ * vmw_user_bo_ref_obj-release - TTM synccpu reference object release callback
+ * for vmw user buffer objects
+ *
+ * @base: Pointer to the TTM base object
+ * @ref_type: Reference type of the reference reaching zero.
+ *
+ * Called when user-space drops its last synccpu reference on the buffer
+ * object, Either explicitly or as part of a cleanup file close.
+ */
+static void vmw_user_bo_ref_obj_release(struct ttm_base_object *base,
+					enum ttm_ref_type ref_type)
+{
+	struct vmw_user_buffer_object *user_bo;
+
+	user_bo = container_of(base, struct vmw_user_buffer_object, prime.base);
+
+	switch (ref_type) {
+	case TTM_REF_SYNCCPU_WRITE:
+		ttm_bo_synccpu_write_release(&user_bo->vbo.base);
+		break;
+	default:
+		WARN_ONCE(true, "Undefined buffer object reference release.\n");
+	}
+}
+
+
+/**
+ * vmw_user_bo_alloc - Allocate a user buffer object
+ *
+ * @dev_priv: Pointer to a struct device private.
+ * @tfile: Pointer to a struct ttm_object_file on which to register the user
+ * object.
+ * @size: Size of the buffer object.
+ * @shareable: Boolean whether the buffer is shareable with other open files.
+ * @handle: Pointer to where the handle value should be assigned.
+ * @p_vbo: Pointer to where the refcounted struct vmw_buffer_object pointer
+ * should be assigned.
+ * Return: Zero on success, negative error code on error.
+ */
+int vmw_user_bo_alloc(struct vmw_private *dev_priv,
+		      struct ttm_object_file *tfile,
+		      uint32_t size,
+		      bool shareable,
+		      uint32_t *handle,
+		      struct vmw_buffer_object **p_vbo,
+		      struct ttm_base_object **p_base)
+{
+	struct vmw_user_buffer_object *user_bo;
+	struct ttm_buffer_object *tmp;
+	int ret;
+
+	user_bo = kzalloc(sizeof(*user_bo), GFP_KERNEL);
+	if (unlikely(!user_bo)) {
+		DRM_ERROR("Failed to allocate a buffer.\n");
+		return -ENOMEM;
+	}
+
+	ret = vmw_bo_init(dev_priv, &user_bo->vbo, size,
+			  (dev_priv->has_mob) ?
+			  &vmw_sys_placement :
+			  &vmw_vram_sys_placement, true,
+			  &vmw_user_bo_destroy);
+	if (unlikely(ret != 0))
+		return ret;
+
+	tmp = ttm_bo_reference(&user_bo->vbo.base);
+	ret = ttm_prime_object_init(tfile,
+				    size,
+				    &user_bo->prime,
+				    shareable,
+				    ttm_buffer_type,
+				    &vmw_user_bo_release,
+				    &vmw_user_bo_ref_obj_release);
+	if (unlikely(ret != 0)) {
+		ttm_bo_unref(&tmp);
+		goto out_no_base_object;
+	}
+
+	*p_vbo = &user_bo->vbo;
+	if (p_base) {
+		*p_base = &user_bo->prime.base;
+		kref_get(&(*p_base)->refcount);
+	}
+	*handle = user_bo->prime.base.hash.key;
+
+out_no_base_object:
+	return ret;
+}
+
+
+/**
+ * vmw_user_bo_verify_access - verify access permissions on this
+ * buffer object.
+ *
+ * @bo: Pointer to the buffer object being accessed
+ * @tfile: Identifying the caller.
+ */
+int vmw_user_bo_verify_access(struct ttm_buffer_object *bo,
+			      struct ttm_object_file *tfile)
+{
+	struct vmw_user_buffer_object *vmw_user_bo;
+
+	if (unlikely(bo->destroy != vmw_user_bo_destroy))
+		return -EPERM;
+
+	vmw_user_bo = vmw_user_buffer_object(bo);
+
+	/* Check that the caller has opened the object. */
+	if (likely(ttm_ref_object_exists(tfile, &vmw_user_bo->prime.base)))
+		return 0;
+
+	DRM_ERROR("Could not grant buffer access.\n");
+	return -EPERM;
+}
+
+
+/**
+ * vmw_user_bo_synccpu_grab - Grab a struct vmw_user_buffer_object for cpu
+ * access, idling previous GPU operations on the buffer and optionally
+ * blocking it for further command submissions.
+ *
+ * @user_bo: Pointer to the buffer object being grabbed for CPU access
+ * @tfile: Identifying the caller.
+ * @flags: Flags indicating how the grab should be performed.
+ * Return: Zero on success, Negative error code on error. In particular,
+ * -EBUSY will be returned if a dontblock operation is requested and the
+ * buffer object is busy, and -ERESTARTSYS will be returned if a wait is
+ * interrupted by a signal.
+ *
+ * A blocking grab will be automatically released when @tfile is closed.
+ */
+static int vmw_user_bo_synccpu_grab(struct vmw_user_buffer_object *user_bo,
+				    struct ttm_object_file *tfile,
+				    uint32_t flags)
+{
+	struct ttm_buffer_object *bo = &user_bo->vbo.base;
+	bool existed;
+	int ret;
+
+	if (flags & drm_vmw_synccpu_allow_cs) {
+		bool nonblock = !!(flags & drm_vmw_synccpu_dontblock);
+		long lret;
+
+		lret = reservation_object_wait_timeout_rcu
+			(bo->resv, true, true,
+			 nonblock ? 0 : MAX_SCHEDULE_TIMEOUT);
+		if (!lret)
+			return -EBUSY;
+		else if (lret < 0)
+			return lret;
+		return 0;
+	}
+
+	ret = ttm_bo_synccpu_write_grab
+		(bo, !!(flags & drm_vmw_synccpu_dontblock));
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = ttm_ref_object_add(tfile, &user_bo->prime.base,
+				 TTM_REF_SYNCCPU_WRITE, &existed, false);
+	if (ret != 0 || existed)
+		ttm_bo_synccpu_write_release(&user_bo->vbo.base);
+
+	return ret;
+}
+
+/**
+ * vmw_user_bo_synccpu_release - Release a previous grab for CPU access,
+ * and unblock command submission on the buffer if blocked.
+ *
+ * @handle: Handle identifying the buffer object.
+ * @tfile: Identifying the caller.
+ * @flags: Flags indicating the type of release.
+ */
+static int vmw_user_bo_synccpu_release(uint32_t handle,
+					   struct ttm_object_file *tfile,
+					   uint32_t flags)
+{
+	if (!(flags & drm_vmw_synccpu_allow_cs))
+		return ttm_ref_object_base_unref(tfile, handle,
+						 TTM_REF_SYNCCPU_WRITE);
+
+	return 0;
+}
+
+
+/**
+ * vmw_user_bo_synccpu_ioctl - ioctl function implementing the synccpu
+ * functionality.
+ *
+ * @dev: Identifies the drm device.
+ * @data: Pointer to the ioctl argument.
+ * @file_priv: Identifies the caller.
+ * Return: Zero on success, negative error code on error.
+ *
+ * This function checks the ioctl arguments for validity and calls the
+ * relevant synccpu functions.
+ */
+int vmw_user_bo_synccpu_ioctl(struct drm_device *dev, void *data,
+			      struct drm_file *file_priv)
+{
+	struct drm_vmw_synccpu_arg *arg =
+		(struct drm_vmw_synccpu_arg *) data;
+	struct vmw_buffer_object *vbo;
+	struct vmw_user_buffer_object *user_bo;
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+	struct ttm_base_object *buffer_base;
+	int ret;
+
+	if ((arg->flags & (drm_vmw_synccpu_read | drm_vmw_synccpu_write)) == 0
+	    || (arg->flags & ~(drm_vmw_synccpu_read | drm_vmw_synccpu_write |
+			       drm_vmw_synccpu_dontblock |
+			       drm_vmw_synccpu_allow_cs)) != 0) {
+		DRM_ERROR("Illegal synccpu flags.\n");
+		return -EINVAL;
+	}
+
+	switch (arg->op) {
+	case drm_vmw_synccpu_grab:
+		ret = vmw_user_bo_lookup(tfile, arg->handle, &vbo,
+					     &buffer_base);
+		if (unlikely(ret != 0))
+			return ret;
+
+		user_bo = container_of(vbo, struct vmw_user_buffer_object,
+				       vbo);
+		ret = vmw_user_bo_synccpu_grab(user_bo, tfile, arg->flags);
+		vmw_bo_unreference(&vbo);
+		ttm_base_object_unref(&buffer_base);
+		if (unlikely(ret != 0 && ret != -ERESTARTSYS &&
+			     ret != -EBUSY)) {
+			DRM_ERROR("Failed synccpu grab on handle 0x%08x.\n",
+				  (unsigned int) arg->handle);
+			return ret;
+		}
+		break;
+	case drm_vmw_synccpu_release:
+		ret = vmw_user_bo_synccpu_release(arg->handle, tfile,
+						  arg->flags);
+		if (unlikely(ret != 0)) {
+			DRM_ERROR("Failed synccpu release on handle 0x%08x.\n",
+				  (unsigned int) arg->handle);
+			return ret;
+		}
+		break;
+	default:
+		DRM_ERROR("Invalid synccpu operation.\n");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+
+
+/**
+ * vmw_bo_alloc_ioctl - ioctl function implementing the buffer object
+ * allocation functionality.
+ *
+ * @dev: Identifies the drm device.
+ * @data: Pointer to the ioctl argument.
+ * @file_priv: Identifies the caller.
+ * Return: Zero on success, negative error code on error.
+ *
+ * This function checks the ioctl arguments for validity and allocates a
+ * struct vmw_user_buffer_object bo.
+ */
+int vmw_bo_alloc_ioctl(struct drm_device *dev, void *data,
+		       struct drm_file *file_priv)
+{
+	struct vmw_private *dev_priv = vmw_priv(dev);
+	union drm_vmw_alloc_dmabuf_arg *arg =
+	    (union drm_vmw_alloc_dmabuf_arg *)data;
+	struct drm_vmw_alloc_dmabuf_req *req = &arg->req;
+	struct drm_vmw_dmabuf_rep *rep = &arg->rep;
+	struct vmw_buffer_object *vbo;
+	uint32_t handle;
+	int ret;
+
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = vmw_user_bo_alloc(dev_priv, vmw_fpriv(file_priv)->tfile,
+				req->size, false, &handle, &vbo,
+				NULL);
+	if (unlikely(ret != 0))
+		goto out_no_bo;
+
+	rep->handle = handle;
+	rep->map_handle = drm_vma_node_offset_addr(&vbo->base.vma_node);
+	rep->cur_gmr_id = handle;
+	rep->cur_gmr_offset = 0;
+
+	vmw_bo_unreference(&vbo);
+
+out_no_bo:
+	ttm_read_unlock(&dev_priv->reservation_sem);
+
+	return ret;
+}
+
+
+/**
+ * vmw_bo_unref_ioctl - Generic handle close ioctl.
+ *
+ * @dev: Identifies the drm device.
+ * @data: Pointer to the ioctl argument.
+ * @file_priv: Identifies the caller.
+ * Return: Zero on success, negative error code on error.
+ *
+ * This function checks the ioctl arguments for validity and closes a
+ * handle to a TTM base object, optionally freeing the object.
+ */
+int vmw_bo_unref_ioctl(struct drm_device *dev, void *data,
+		       struct drm_file *file_priv)
+{
+	struct drm_vmw_unref_dmabuf_arg *arg =
+	    (struct drm_vmw_unref_dmabuf_arg *)data;
+
+	return ttm_ref_object_base_unref(vmw_fpriv(file_priv)->tfile,
+					 arg->handle,
+					 TTM_REF_USAGE);
+}
+
+
+/**
+ * vmw_user_bo_lookup - Look up a vmw user buffer object from a handle.
+ *
+ * @tfile: The TTM object file the handle is registered with.
+ * @handle: The user buffer object handle
+ * @out: Pointer to a where a pointer to the embedded
+ * struct vmw_buffer_object should be placed.
+ * @p_base: Pointer to where a pointer to the TTM base object should be
+ * placed, or NULL if no such pointer is required.
+ * Return: Zero on success, Negative error code on error.
+ *
+ * Both the output base object pointer and the vmw buffer object pointer
+ * will be refcounted.
+ */
+int vmw_user_bo_lookup(struct ttm_object_file *tfile,
+		       uint32_t handle, struct vmw_buffer_object **out,
+		       struct ttm_base_object **p_base)
+{
+	struct vmw_user_buffer_object *vmw_user_bo;
+	struct ttm_base_object *base;
+
+	base = ttm_base_object_lookup(tfile, handle);
+	if (unlikely(base == NULL)) {
+		DRM_ERROR("Invalid buffer object handle 0x%08lx.\n",
+			  (unsigned long)handle);
+		return -ESRCH;
+	}
+
+	if (unlikely(ttm_base_object_type(base) != ttm_buffer_type)) {
+		ttm_base_object_unref(&base);
+		DRM_ERROR("Invalid buffer object handle 0x%08lx.\n",
+			  (unsigned long)handle);
+		return -EINVAL;
+	}
+
+	vmw_user_bo = container_of(base, struct vmw_user_buffer_object,
+				   prime.base);
+	(void)ttm_bo_reference(&vmw_user_bo->vbo.base);
+	if (p_base)
+		*p_base = base;
+	else
+		ttm_base_object_unref(&base);
+	*out = &vmw_user_bo->vbo;
+
+	return 0;
+}
+
+
+/**
+ * vmw_user_bo_reference - Open a handle to a vmw user buffer object.
+ *
+ * @tfile: The TTM object file to register the handle with.
+ * @vbo: The embedded vmw buffer object.
+ * @handle: Pointer to where the new handle should be placed.
+ * Return: Zero on success, Negative error code on error.
+ */
+int vmw_user_bo_reference(struct ttm_object_file *tfile,
+			  struct vmw_buffer_object *vbo,
+			  uint32_t *handle)
+{
+	struct vmw_user_buffer_object *user_bo;
+
+	if (vbo->base.destroy != vmw_user_bo_destroy)
+		return -EINVAL;
+
+	user_bo = container_of(vbo, struct vmw_user_buffer_object, vbo);
+
+	*handle = user_bo->prime.base.hash.key;
+	return ttm_ref_object_add(tfile, &user_bo->prime.base,
+				  TTM_REF_USAGE, NULL, false);
+}
+
+
+/**
+ * vmw_bo_fence_single - Utility function to fence a single TTM buffer
+ *                       object without unreserving it.
+ *
+ * @bo:             Pointer to the struct ttm_buffer_object to fence.
+ * @fence:          Pointer to the fence. If NULL, this function will
+ *                  insert a fence into the command stream..
+ *
+ * Contrary to the ttm_eu version of this function, it takes only
+ * a single buffer object instead of a list, and it also doesn't
+ * unreserve the buffer object, which needs to be done separately.
+ */
+void vmw_bo_fence_single(struct ttm_buffer_object *bo,
+			 struct vmw_fence_obj *fence)
+{
+	struct ttm_bo_device *bdev = bo->bdev;
+
+	struct vmw_private *dev_priv =
+		container_of(bdev, struct vmw_private, bdev);
+
+	if (fence == NULL) {
+		vmw_execbuf_fence_commands(NULL, dev_priv, &fence, NULL);
+		reservation_object_add_excl_fence(bo->resv, &fence->base);
+		dma_fence_put(&fence->base);
+	} else
+		reservation_object_add_excl_fence(bo->resv, &fence->base);
+}
+
+
+/**
+ * vmw_dumb_create - Create a dumb kms buffer
+ *
+ * @file_priv: Pointer to a struct drm_file identifying the caller.
+ * @dev: Pointer to the drm device.
+ * @args: Pointer to a struct drm_mode_create_dumb structure
+ * Return: Zero on success, negative error code on failure.
+ *
+ * This is a driver callback for the core drm create_dumb functionality.
+ * Note that this is very similar to the vmw_bo_alloc ioctl, except
+ * that the arguments have a different format.
+ */
+int vmw_dumb_create(struct drm_file *file_priv,
+		    struct drm_device *dev,
+		    struct drm_mode_create_dumb *args)
+{
+	struct vmw_private *dev_priv = vmw_priv(dev);
+	struct vmw_buffer_object *vbo;
+	int ret;
+
+	args->pitch = args->width * ((args->bpp + 7) / 8);
+	args->size = args->pitch * args->height;
+
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
+	if (unlikely(ret != 0))
+		return ret;
+
+	ret = vmw_user_bo_alloc(dev_priv, vmw_fpriv(file_priv)->tfile,
+				    args->size, false, &args->handle,
+				    &vbo, NULL);
+	if (unlikely(ret != 0))
+		goto out_no_bo;
+
+	vmw_bo_unreference(&vbo);
+out_no_bo:
+	ttm_read_unlock(&dev_priv->reservation_sem);
+	return ret;
+}
+
+
+/**
+ * vmw_dumb_map_offset - Return the address space offset of a dumb buffer
+ *
+ * @file_priv: Pointer to a struct drm_file identifying the caller.
+ * @dev: Pointer to the drm device.
+ * @handle: Handle identifying the dumb buffer.
+ * @offset: The address space offset returned.
+ * Return: Zero on success, negative error code on failure.
+ *
+ * This is a driver callback for the core drm dumb_map_offset functionality.
+ */
+int vmw_dumb_map_offset(struct drm_file *file_priv,
+			struct drm_device *dev, uint32_t handle,
+			uint64_t *offset)
+{
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+	struct vmw_buffer_object *out_buf;
+	int ret;
+
+	ret = vmw_user_bo_lookup(tfile, handle, &out_buf, NULL);
+	if (ret != 0)
+		return -EINVAL;
+
+	*offset = drm_vma_node_offset_addr(&out_buf->base.vma_node);
+	vmw_bo_unreference(&out_buf);
+	return 0;
+}
+
+
+/**
+ * vmw_dumb_destroy - Destroy a dumb boffer
+ *
+ * @file_priv: Pointer to a struct drm_file identifying the caller.
+ * @dev: Pointer to the drm device.
+ * @handle: Handle identifying the dumb buffer.
+ * Return: Zero on success, negative error code on failure.
+ *
+ * This is a driver callback for the core drm dumb_destroy functionality.
+ */
+int vmw_dumb_destroy(struct drm_file *file_priv,
+		     struct drm_device *dev,
+		     uint32_t handle)
+{
+	return ttm_ref_object_base_unref(vmw_fpriv(file_priv)->tfile,
+					 handle, TTM_REF_USAGE);
+}
+
+
+/**
+ * vmw_bo_swap_notify - swapout notify callback.
+ *
+ * @bo: The buffer object to be swapped out.
+ */
+void vmw_bo_swap_notify(struct ttm_buffer_object *bo)
+{
+	/* Is @bo embedded in a struct vmw_buffer_object? */
+	if (bo->destroy != vmw_bo_bo_free &&
+	    bo->destroy != vmw_user_bo_destroy)
+		return;
+
+	/* Kill any cached kernel maps before swapout */
+	vmw_bo_unmap(vmw_buffer_object(bo));
+}
+
+
+/**
+ * vmw_bo_move_notify - TTM move_notify_callback
+ *
+ * @bo: The TTM buffer object about to move.
+ * @mem: The struct ttm_mem_reg indicating to what memory
+ *       region the move is taking place.
+ *
+ * Detaches cached maps and device bindings that require that the
+ * buffer doesn't move.
+ */
+void vmw_bo_move_notify(struct ttm_buffer_object *bo,
+			struct ttm_mem_reg *mem)
+{
+	struct vmw_buffer_object *vbo;
+
+	if (mem == NULL)
+		return;
+
+	/* Make sure @bo is embedded in a struct vmw_buffer_object? */
+	if (bo->destroy != vmw_bo_bo_free &&
+	    bo->destroy != vmw_user_bo_destroy)
+		return;
+
+	vbo = container_of(bo, struct vmw_buffer_object, base);
+
+	/*
+	 * Kill any cached kernel maps before move to or from VRAM.
+	 * With other types of moves, the underlying pages stay the same,
+	 * and the map can be kept.
+	 */
+	if (mem->mem_type == TTM_PL_VRAM || bo->mem.mem_type == TTM_PL_VRAM)
+		vmw_bo_unmap(vbo);
+
+	/*
+	 * If we're moving a backup MOB out of MOB placement, then make sure we
+	 * read back all resource content first, and unbind the MOB from
+	 * the resource.
+	 */
+	if (mem->mem_type != VMW_PL_MOB && bo->mem.mem_type == VMW_PL_MOB)
+		vmw_resource_unbind_list(vbo);
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c
index 9f45d5004cae..e7e4655d3f36 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf_res.c b/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf_res.c
index 36c7b6c839c0..3b75af9bf85f 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf_res.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf_res.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2014-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2014-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_context.c b/drivers/gpu/drm/vmwgfx/vmwgfx_context.c
index 3767ac335aca..7c3cb8efd11a 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_context.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_context.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -38,7 +38,7 @@ struct vmw_user_context {
 	struct vmw_cmdbuf_res_manager *man;
 	struct vmw_resource *cotables[SVGA_COTABLE_DX10_MAX];
 	spinlock_t cotable_lock;
-	struct vmw_dma_buffer *dx_query_mob;
+	struct vmw_buffer_object *dx_query_mob;
 };
 
 static void vmw_user_context_free(struct vmw_resource *res);
@@ -424,7 +424,7 @@ static int vmw_gb_context_unbind(struct vmw_resource *res,
 	(void) vmw_execbuf_fence_commands(NULL, dev_priv,
 					  &fence, NULL);
 
-	vmw_fence_single_bo(bo, fence);
+	vmw_bo_fence_single(bo, fence);
 
 	if (likely(fence != NULL))
 		vmw_fence_obj_unreference(&fence);
@@ -648,7 +648,7 @@ static int vmw_dx_context_unbind(struct vmw_resource *res,
 	(void) vmw_execbuf_fence_commands(NULL, dev_priv,
 					  &fence, NULL);
 
-	vmw_fence_single_bo(bo, fence);
+	vmw_bo_fence_single(bo, fence);
 
 	if (likely(fence != NULL))
 		vmw_fence_obj_unreference(&fence);
@@ -900,7 +900,7 @@ vmw_context_binding_state(struct vmw_resource *ctx)
  * specified in the parameter.  0 otherwise.
  */
 int vmw_context_bind_dx_query(struct vmw_resource *ctx_res,
-			      struct vmw_dma_buffer *mob)
+			      struct vmw_buffer_object *mob)
 {
 	struct vmw_user_context *uctx =
 		container_of(ctx_res, struct vmw_user_context, res);
@@ -908,7 +908,7 @@ int vmw_context_bind_dx_query(struct vmw_resource *ctx_res,
 	if (mob == NULL) {
 		if (uctx->dx_query_mob) {
 			uctx->dx_query_mob->dx_query_ctx = NULL;
-			vmw_dmabuf_unreference(&uctx->dx_query_mob);
+			vmw_bo_unreference(&uctx->dx_query_mob);
 			uctx->dx_query_mob = NULL;
 		}
 
@@ -922,7 +922,7 @@ int vmw_context_bind_dx_query(struct vmw_resource *ctx_res,
 	mob->dx_query_ctx  = ctx_res;
 
 	if (!uctx->dx_query_mob)
-		uctx->dx_query_mob = vmw_dmabuf_reference(mob);
+		uctx->dx_query_mob = vmw_bo_reference(mob);
 
 	return 0;
 }
@@ -932,7 +932,7 @@ int vmw_context_bind_dx_query(struct vmw_resource *ctx_res,
  *
  * @ctx_res: The context resource
  */
-struct vmw_dma_buffer *
+struct vmw_buffer_object *
 vmw_context_get_dx_query_mob(struct vmw_resource *ctx_res)
 {
 	struct vmw_user_context *uctx =
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_cotable.c b/drivers/gpu/drm/vmwgfx/vmwgfx_cotable.c
index cbf54ea7b4c0..1d45714e1d5a 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_cotable.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_cotable.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2014-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2014-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -324,7 +324,7 @@ static int vmw_cotable_unbind(struct vmw_resource *res,
 		vmw_dx_context_scrub_cotables(vcotbl->ctx, readback);
 	mutex_unlock(&dev_priv->binding_mutex);
 	(void) vmw_execbuf_fence_commands(NULL, dev_priv, &fence, NULL);
-	vmw_fence_single_bo(bo, fence);
+	vmw_bo_fence_single(bo, fence);
 	if (likely(fence != NULL))
 		vmw_fence_obj_unreference(&fence);
 
@@ -367,7 +367,7 @@ static int vmw_cotable_readback(struct vmw_resource *res)
 	}
 
 	(void) vmw_execbuf_fence_commands(NULL, dev_priv, &fence, NULL);
-	vmw_fence_single_bo(&res->backup->base, fence);
+	vmw_bo_fence_single(&res->backup->base, fence);
 	vmw_fence_obj_unreference(&fence);
 
 	return 0;
@@ -390,7 +390,7 @@ static int vmw_cotable_resize(struct vmw_resource *res, size_t new_size)
 	struct ttm_operation_ctx ctx = { false, false };
 	struct vmw_private *dev_priv = res->dev_priv;
 	struct vmw_cotable *vcotbl = vmw_cotable(res);
-	struct vmw_dma_buffer *buf, *old_buf = res->backup;
+	struct vmw_buffer_object *buf, *old_buf = res->backup;
 	struct ttm_buffer_object *bo, *old_bo = &res->backup->base;
 	size_t old_size = res->backup_size;
 	size_t old_size_read_back = vcotbl->size_read_back;
@@ -415,8 +415,8 @@ static int vmw_cotable_resize(struct vmw_resource *res, size_t new_size)
 	if (!buf)
 		return -ENOMEM;
 
-	ret = vmw_dmabuf_init(dev_priv, buf, new_size, &vmw_mob_ne_placement,
-			      true, vmw_dmabuf_bo_free);
+	ret = vmw_bo_init(dev_priv, buf, new_size, &vmw_mob_ne_placement,
+			  true, vmw_bo_bo_free);
 	if (ret) {
 		DRM_ERROR("Failed initializing new cotable MOB.\n");
 		return ret;
@@ -482,7 +482,7 @@ static int vmw_cotable_resize(struct vmw_resource *res, size_t new_size)
 	/* Let go of the old mob. */
 	list_del(&res->mob_head);
 	list_add_tail(&res->mob_head, &buf->res_list);
-	vmw_dmabuf_unreference(&old_buf);
+	vmw_bo_unreference(&old_buf);
 	res->id = vcotbl->type;
 
 	return 0;
@@ -491,7 +491,7 @@ out_map_new:
 	ttm_bo_kunmap(&old_map);
 out_wait:
 	ttm_bo_unreserve(bo);
-	vmw_dmabuf_unreference(&buf);
+	vmw_bo_unreference(&buf);
 
 	return ret;
 }
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
deleted file mode 100644
index d59d9dd16ebc..000000000000
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_dmabuf.c
+++ /dev/null
@@ -1,376 +0,0 @@
-/**************************************************************************
- *
- * Copyright © 2011-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the
- * "Software"), to deal in the Software without restriction, including
- * without limitation the rights to use, copy, modify, merge, publish,
- * distribute, sub license, and/or sell copies of the Software, and to
- * permit persons to whom the Software is furnished to do so, subject to
- * the following conditions:
- *
- * The above copyright notice and this permission notice (including the
- * next paragraph) shall be included in all copies or substantial portions
- * of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
- * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
- * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
- * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
- * USE OR OTHER DEALINGS IN THE SOFTWARE.
- *
- **************************************************************************/
-
-#include <drm/ttm/ttm_placement.h>
-
-#include <drm/drmP.h>
-#include "vmwgfx_drv.h"
-
-
-/**
- * vmw_dmabuf_pin_in_placement - Validate a buffer to placement.
- *
- * @dev_priv:  Driver private.
- * @buf:  DMA buffer to move.
- * @placement:  The placement to pin it.
- * @interruptible:  Use interruptible wait.
- *
- * Returns
- *  -ERESTARTSYS if interrupted by a signal.
- */
-int vmw_dmabuf_pin_in_placement(struct vmw_private *dev_priv,
-				struct vmw_dma_buffer *buf,
-				struct ttm_placement *placement,
-				bool interruptible)
-{
-	struct ttm_operation_ctx ctx = {interruptible, false };
-	struct ttm_buffer_object *bo = &buf->base;
-	int ret;
-	uint32_t new_flags;
-
-	ret = ttm_write_lock(&dev_priv->reservation_sem, interruptible);
-	if (unlikely(ret != 0))
-		return ret;
-
-	vmw_execbuf_release_pinned_bo(dev_priv);
-
-	ret = ttm_bo_reserve(bo, interruptible, false, NULL);
-	if (unlikely(ret != 0))
-		goto err;
-
-	if (buf->pin_count > 0)
-		ret = ttm_bo_mem_compat(placement, &bo->mem,
-					&new_flags) == true ? 0 : -EINVAL;
-	else
-		ret = ttm_bo_validate(bo, placement, &ctx);
-
-	if (!ret)
-		vmw_bo_pin_reserved(buf, true);
-
-	ttm_bo_unreserve(bo);
-
-err:
-	ttm_write_unlock(&dev_priv->reservation_sem);
-	return ret;
-}
-
-/**
- * vmw_dmabuf_pin_in_vram_or_gmr - Move a buffer to vram or gmr.
- *
- * This function takes the reservation_sem in write mode.
- * Flushes and unpins the query bo to avoid failures.
- *
- * @dev_priv:  Driver private.
- * @buf:  DMA buffer to move.
- * @pin:  Pin buffer if true.
- * @interruptible:  Use interruptible wait.
- *
- * Returns
- * -ERESTARTSYS if interrupted by a signal.
- */
-int vmw_dmabuf_pin_in_vram_or_gmr(struct vmw_private *dev_priv,
-				  struct vmw_dma_buffer *buf,
-				  bool interruptible)
-{
-	struct ttm_operation_ctx ctx = {interruptible, false };
-	struct ttm_buffer_object *bo = &buf->base;
-	int ret;
-	uint32_t new_flags;
-
-	ret = ttm_write_lock(&dev_priv->reservation_sem, interruptible);
-	if (unlikely(ret != 0))
-		return ret;
-
-	vmw_execbuf_release_pinned_bo(dev_priv);
-
-	ret = ttm_bo_reserve(bo, interruptible, false, NULL);
-	if (unlikely(ret != 0))
-		goto err;
-
-	if (buf->pin_count > 0) {
-		ret = ttm_bo_mem_compat(&vmw_vram_gmr_placement, &bo->mem,
-					&new_flags) == true ? 0 : -EINVAL;
-		goto out_unreserve;
-	}
-
-	ret = ttm_bo_validate(bo, &vmw_vram_gmr_placement, &ctx);
-	if (likely(ret == 0) || ret == -ERESTARTSYS)
-		goto out_unreserve;
-
-	ret = ttm_bo_validate(bo, &vmw_vram_placement, &ctx);
-
-out_unreserve:
-	if (!ret)
-		vmw_bo_pin_reserved(buf, true);
-
-	ttm_bo_unreserve(bo);
-err:
-	ttm_write_unlock(&dev_priv->reservation_sem);
-	return ret;
-}
-
-/**
- * vmw_dmabuf_pin_in_vram - Move a buffer to vram.
- *
- * This function takes the reservation_sem in write mode.
- * Flushes and unpins the query bo to avoid failures.
- *
- * @dev_priv:  Driver private.
- * @buf:  DMA buffer to move.
- * @interruptible:  Use interruptible wait.
- *
- * Returns
- * -ERESTARTSYS if interrupted by a signal.
- */
-int vmw_dmabuf_pin_in_vram(struct vmw_private *dev_priv,
-			   struct vmw_dma_buffer *buf,
-			   bool interruptible)
-{
-	return vmw_dmabuf_pin_in_placement(dev_priv, buf, &vmw_vram_placement,
-					   interruptible);
-}
-
-/**
- * vmw_dmabuf_pin_in_start_of_vram - Move a buffer to start of vram.
- *
- * This function takes the reservation_sem in write mode.
- * Flushes and unpins the query bo to avoid failures.
- *
- * @dev_priv:  Driver private.
- * @buf:  DMA buffer to pin.
- * @interruptible:  Use interruptible wait.
- *
- * Returns
- * -ERESTARTSYS if interrupted by a signal.
- */
-int vmw_dmabuf_pin_in_start_of_vram(struct vmw_private *dev_priv,
-				    struct vmw_dma_buffer *buf,
-				    bool interruptible)
-{
-	struct ttm_operation_ctx ctx = {interruptible, false };
-	struct ttm_buffer_object *bo = &buf->base;
-	struct ttm_placement placement;
-	struct ttm_place place;
-	int ret = 0;
-	uint32_t new_flags;
-
-	place = vmw_vram_placement.placement[0];
-	place.lpfn = bo->num_pages;
-	placement.num_placement = 1;
-	placement.placement = &place;
-	placement.num_busy_placement = 1;
-	placement.busy_placement = &place;
-
-	ret = ttm_write_lock(&dev_priv->reservation_sem, interruptible);
-	if (unlikely(ret != 0))
-		return ret;
-
-	vmw_execbuf_release_pinned_bo(dev_priv);
-	ret = ttm_bo_reserve(bo, interruptible, false, NULL);
-	if (unlikely(ret != 0))
-		goto err_unlock;
-
-	/*
-	 * Is this buffer already in vram but not at the start of it?
-	 * In that case, evict it first because TTM isn't good at handling
-	 * that situation.
-	 */
-	if (bo->mem.mem_type == TTM_PL_VRAM &&
-	    bo->mem.start < bo->num_pages &&
-	    bo->mem.start > 0 &&
-	    buf->pin_count == 0) {
-		ctx.interruptible = false;
-		(void) ttm_bo_validate(bo, &vmw_sys_placement, &ctx);
-	}
-
-	if (buf->pin_count > 0)
-		ret = ttm_bo_mem_compat(&placement, &bo->mem,
-					&new_flags) == true ? 0 : -EINVAL;
-	else
-		ret = ttm_bo_validate(bo, &placement, &ctx);
-
-	/* For some reason we didn't end up at the start of vram */
-	WARN_ON(ret == 0 && bo->offset != 0);
-	if (!ret)
-		vmw_bo_pin_reserved(buf, true);
-
-	ttm_bo_unreserve(bo);
-err_unlock:
-	ttm_write_unlock(&dev_priv->reservation_sem);
-
-	return ret;
-}
-
-/**
- * vmw_dmabuf_unpin - Unpin the buffer given buffer, does not move the buffer.
- *
- * This function takes the reservation_sem in write mode.
- *
- * @dev_priv:  Driver private.
- * @buf:  DMA buffer to unpin.
- * @interruptible:  Use interruptible wait.
- *
- * Returns
- * -ERESTARTSYS if interrupted by a signal.
- */
-int vmw_dmabuf_unpin(struct vmw_private *dev_priv,
-		     struct vmw_dma_buffer *buf,
-		     bool interruptible)
-{
-	struct ttm_buffer_object *bo = &buf->base;
-	int ret;
-
-	ret = ttm_read_lock(&dev_priv->reservation_sem, interruptible);
-	if (unlikely(ret != 0))
-		return ret;
-
-	ret = ttm_bo_reserve(bo, interruptible, false, NULL);
-	if (unlikely(ret != 0))
-		goto err;
-
-	vmw_bo_pin_reserved(buf, false);
-
-	ttm_bo_unreserve(bo);
-
-err:
-	ttm_read_unlock(&dev_priv->reservation_sem);
-	return ret;
-}
-
-/**
- * vmw_bo_get_guest_ptr - Get the guest ptr representing the current placement
- * of a buffer.
- *
- * @bo: Pointer to a struct ttm_buffer_object. Must be pinned or reserved.
- * @ptr: SVGAGuestPtr returning the result.
- */
-void vmw_bo_get_guest_ptr(const struct ttm_buffer_object *bo,
-			  SVGAGuestPtr *ptr)
-{
-	if (bo->mem.mem_type == TTM_PL_VRAM) {
-		ptr->gmrId = SVGA_GMR_FRAMEBUFFER;
-		ptr->offset = bo->offset;
-	} else {
-		ptr->gmrId = bo->mem.start;
-		ptr->offset = 0;
-	}
-}
-
-
-/**
- * vmw_bo_pin_reserved - Pin or unpin a buffer object without moving it.
- *
- * @vbo: The buffer object. Must be reserved.
- * @pin: Whether to pin or unpin.
- *
- */
-void vmw_bo_pin_reserved(struct vmw_dma_buffer *vbo, bool pin)
-{
-	struct ttm_operation_ctx ctx = { false, true };
-	struct ttm_place pl;
-	struct ttm_placement placement;
-	struct ttm_buffer_object *bo = &vbo->base;
-	uint32_t old_mem_type = bo->mem.mem_type;
-	int ret;
-
-	lockdep_assert_held(&bo->resv->lock.base);
-
-	if (pin) {
-		if (vbo->pin_count++ > 0)
-			return;
-	} else {
-		WARN_ON(vbo->pin_count <= 0);
-		if (--vbo->pin_count > 0)
-			return;
-	}
-
-	pl.fpfn = 0;
-	pl.lpfn = 0;
-	pl.flags = TTM_PL_FLAG_VRAM | VMW_PL_FLAG_GMR | VMW_PL_FLAG_MOB
-		| TTM_PL_FLAG_SYSTEM | TTM_PL_FLAG_CACHED;
-	if (pin)
-		pl.flags |= TTM_PL_FLAG_NO_EVICT;
-
-	memset(&placement, 0, sizeof(placement));
-	placement.num_placement = 1;
-	placement.placement = &pl;
-
-	ret = ttm_bo_validate(bo, &placement, &ctx);
-
-	BUG_ON(ret != 0 || bo->mem.mem_type != old_mem_type);
-}
-
-
-/*
- * vmw_dma_buffer_unmap - Tear down a cached buffer object map.
- *
- * @vbo: The buffer object whose map we are tearing down.
- *
- * This function tears down a cached map set up using
- * vmw_dma_buffer_map_and_cache().
- */
-void vmw_dma_buffer_unmap(struct vmw_dma_buffer *vbo)
-{
-	if (vbo->map.bo == NULL)
-		return;
-
-	ttm_bo_kunmap(&vbo->map);
-}
-
-
-/*
- * vmw_dma_buffer_map_and_cache - Map a buffer object and cache the map
- *
- * @vbo: The buffer object to map
- * Return: A kernel virtual address or NULL if mapping failed.
- *
- * This function maps a buffer object into the kernel address space, or
- * returns the virtual kernel address of an already existing map. The virtual
- * address remains valid as long as the buffer object is pinned or reserved.
- * The cached map is torn down on either
- * 1) Buffer object move
- * 2) Buffer object swapout
- * 3) Buffer object destruction
- *
- */
-void *vmw_dma_buffer_map_and_cache(struct vmw_dma_buffer *vbo)
-{
-	struct ttm_buffer_object *bo = &vbo->base;
-	bool not_used;
-	void *virtual;
-	int ret;
-
-	virtual = ttm_kmap_obj_virtual(&vbo->map, &not_used);
-	if (virtual)
-		return virtual;
-
-	ret = ttm_bo_kmap(bo, 0, bo->num_pages, &vbo->map);
-	if (ret)
-		DRM_ERROR("Buffer object map failed: %d.\n", ret);
-
-	return ttm_kmap_obj_virtual(&vbo->map, &not_used);
-}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
index 70e1a8820a7c..bb6dbbe18835 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009-2016 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2016 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -137,6 +137,12 @@
 #define DRM_IOCTL_VMW_CREATE_EXTENDED_CONTEXT			\
 	DRM_IOWR(DRM_COMMAND_BASE + DRM_VMW_CREATE_EXTENDED_CONTEXT,	\
 		struct drm_vmw_context_arg)
+#define DRM_IOCTL_VMW_GB_SURFACE_CREATE_EXT				\
+	DRM_IOWR(DRM_COMMAND_BASE + DRM_VMW_GB_SURFACE_CREATE_EXT,	\
+		union drm_vmw_gb_surface_create_ext_arg)
+#define DRM_IOCTL_VMW_GB_SURFACE_REF_EXT				\
+	DRM_IOWR(DRM_COMMAND_BASE + DRM_VMW_GB_SURFACE_REF_EXT,		\
+		union drm_vmw_gb_surface_reference_ext_arg)
 
 /**
  * The core DRM version of this macro doesn't account for
@@ -153,20 +159,20 @@
 static const struct drm_ioctl_desc vmw_ioctls[] = {
 	VMW_IOCTL_DEF(VMW_GET_PARAM, vmw_getparam_ioctl,
 		      DRM_AUTH | DRM_RENDER_ALLOW),
-	VMW_IOCTL_DEF(VMW_ALLOC_DMABUF, vmw_dmabuf_alloc_ioctl,
+	VMW_IOCTL_DEF(VMW_ALLOC_DMABUF, vmw_bo_alloc_ioctl,
 		      DRM_AUTH | DRM_RENDER_ALLOW),
-	VMW_IOCTL_DEF(VMW_UNREF_DMABUF, vmw_dmabuf_unref_ioctl,
+	VMW_IOCTL_DEF(VMW_UNREF_DMABUF, vmw_bo_unref_ioctl,
 		      DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_CURSOR_BYPASS,
 		      vmw_kms_cursor_bypass_ioctl,
-		      DRM_MASTER | DRM_CONTROL_ALLOW),
+		      DRM_MASTER),
 
 	VMW_IOCTL_DEF(VMW_CONTROL_STREAM, vmw_overlay_ioctl,
-		      DRM_MASTER | DRM_CONTROL_ALLOW),
+		      DRM_MASTER),
 	VMW_IOCTL_DEF(VMW_CLAIM_STREAM, vmw_stream_claim_ioctl,
-		      DRM_MASTER | DRM_CONTROL_ALLOW),
+		      DRM_MASTER),
 	VMW_IOCTL_DEF(VMW_UNREF_STREAM, vmw_stream_unref_ioctl,
-		      DRM_MASTER | DRM_CONTROL_ALLOW),
+		      DRM_MASTER),
 
 	VMW_IOCTL_DEF(VMW_CREATE_CONTEXT, vmw_context_define_ioctl,
 		      DRM_AUTH | DRM_RENDER_ALLOW),
@@ -219,11 +225,17 @@ static const struct drm_ioctl_desc vmw_ioctls[] = {
 		      vmw_gb_surface_reference_ioctl,
 		      DRM_AUTH | DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_SYNCCPU,
-		      vmw_user_dmabuf_synccpu_ioctl,
+		      vmw_user_bo_synccpu_ioctl,
 		      DRM_RENDER_ALLOW),
 	VMW_IOCTL_DEF(VMW_CREATE_EXTENDED_CONTEXT,
 		      vmw_extended_context_define_ioctl,
 		      DRM_AUTH | DRM_RENDER_ALLOW),
+	VMW_IOCTL_DEF(VMW_GB_SURFACE_CREATE_EXT,
+		      vmw_gb_surface_define_ext_ioctl,
+		      DRM_AUTH | DRM_RENDER_ALLOW),
+	VMW_IOCTL_DEF(VMW_GB_SURFACE_REF_EXT,
+		      vmw_gb_surface_reference_ext_ioctl,
+		      DRM_AUTH | DRM_RENDER_ALLOW),
 };
 
 static const struct pci_device_id vmw_pci_id_list[] = {
@@ -258,6 +270,15 @@ MODULE_PARM_DESC(assume_16bpp, "Assume 16-bpp when filtering modes");
 module_param_named(assume_16bpp, vmw_assume_16bpp, int, 0600);
 
 
+static void vmw_print_capabilities2(uint32_t capabilities2)
+{
+	DRM_INFO("Capabilities2:\n");
+	if (capabilities2 & SVGA_CAP2_GROW_OTABLE)
+		DRM_INFO("  Grow oTable.\n");
+	if (capabilities2 & SVGA_CAP2_INTRA_SURFACE_COPY)
+		DRM_INFO("  IntraSurface copy.\n");
+}
+
 static void vmw_print_capabilities(uint32_t capabilities)
 {
 	DRM_INFO("Capabilities:\n");
@@ -321,7 +342,7 @@ static void vmw_print_capabilities(uint32_t capabilities)
 static int vmw_dummy_query_bo_create(struct vmw_private *dev_priv)
 {
 	int ret;
-	struct vmw_dma_buffer *vbo;
+	struct vmw_buffer_object *vbo;
 	struct ttm_bo_kmap_obj map;
 	volatile SVGA3dQueryResult *result;
 	bool dummy;
@@ -335,9 +356,9 @@ static int vmw_dummy_query_bo_create(struct vmw_private *dev_priv)
 	if (!vbo)
 		return -ENOMEM;
 
-	ret = vmw_dmabuf_init(dev_priv, vbo, PAGE_SIZE,
-			      &vmw_sys_ne_placement, false,
-			      &vmw_dmabuf_bo_free);
+	ret = vmw_bo_init(dev_priv, vbo, PAGE_SIZE,
+			  &vmw_sys_ne_placement, false,
+			  &vmw_bo_bo_free);
 	if (unlikely(ret != 0))
 		return ret;
 
@@ -358,7 +379,7 @@ static int vmw_dummy_query_bo_create(struct vmw_private *dev_priv)
 
 	if (unlikely(ret != 0)) {
 		DRM_ERROR("Dummy query buffer map failed.\n");
-		vmw_dmabuf_unreference(&vbo);
+		vmw_bo_unreference(&vbo);
 	} else
 		dev_priv->dummy_query_bo = vbo;
 
@@ -460,7 +481,7 @@ static void vmw_release_device_early(struct vmw_private *dev_priv)
 
 	BUG_ON(dev_priv->pinned_bo != NULL);
 
-	vmw_dmabuf_unreference(&dev_priv->dummy_query_bo);
+	vmw_bo_unreference(&dev_priv->dummy_query_bo);
 	if (dev_priv->cman)
 		vmw_cmdbuf_remove_pool(dev_priv->cman);
 
@@ -644,6 +665,7 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 	mutex_init(&dev_priv->cmdbuf_mutex);
 	mutex_init(&dev_priv->release_mutex);
 	mutex_init(&dev_priv->binding_mutex);
+	mutex_init(&dev_priv->requested_layout_mutex);
 	mutex_init(&dev_priv->global_kms_state_mutex);
 	rwlock_init(&dev_priv->resource_lock);
 	ttm_lock_init(&dev_priv->reservation_sem);
@@ -683,6 +705,12 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 	}
 
 	dev_priv->capabilities = vmw_read(dev_priv, SVGA_REG_CAPABILITIES);
+
+	if (dev_priv->capabilities & SVGA_CAP_CAP2_REGISTER) {
+		dev_priv->capabilities2 = vmw_read(dev_priv, SVGA_REG_CAP2);
+	}
+
+
 	ret = vmw_dma_select_mode(dev_priv);
 	if (unlikely(ret != 0)) {
 		DRM_INFO("Restricting capabilities due to IOMMU setup.\n");
@@ -751,6 +779,8 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 	}
 
 	vmw_print_capabilities(dev_priv->capabilities);
+	if (dev_priv->capabilities & SVGA_CAP_CAP2_REGISTER)
+		vmw_print_capabilities2(dev_priv->capabilities2);
 
 	ret = vmw_dma_masks(dev_priv);
 	if (unlikely(ret != 0))
@@ -883,7 +913,7 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 
 	if (dev_priv->has_mob) {
 		spin_lock(&dev_priv->cap_lock);
-		vmw_write(dev_priv, SVGA_REG_DEV_CAP, SVGA3D_DEVCAP_DX);
+		vmw_write(dev_priv, SVGA_REG_DEV_CAP, SVGA3D_DEVCAP_DXCONTEXT);
 		dev_priv->has_dx = !!vmw_read(dev_priv, SVGA_REG_DEV_CAP);
 		spin_unlock(&dev_priv->cap_lock);
 	}
@@ -898,9 +928,23 @@ static int vmw_driver_load(struct drm_device *dev, unsigned long chipset)
 	if (ret)
 		goto out_no_fifo;
 
+	if (dev_priv->has_dx) {
+		/*
+		 * SVGA_CAP2_DX2 (DefineGBSurface_v3) is needed for SM4_1
+		 * support
+		 */
+		if ((dev_priv->capabilities2 & SVGA_CAP2_DX2) != 0) {
+			vmw_write(dev_priv, SVGA_REG_DEV_CAP,
+					SVGA3D_DEVCAP_SM41);
+			dev_priv->has_sm4_1 = vmw_read(dev_priv,
+							SVGA_REG_DEV_CAP);
+		}
+	}
+
 	DRM_INFO("DX: %s\n", dev_priv->has_dx ? "yes." : "no.");
-	DRM_INFO("Atomic: %s\n",
-		 (dev->driver->driver_features & DRIVER_ATOMIC) ? "yes" : "no");
+	DRM_INFO("Atomic: %s\n", (dev->driver->driver_features & DRIVER_ATOMIC)
+		 ? "yes." : "no.");
+	DRM_INFO("SM4_1: %s\n", dev_priv->has_sm4_1 ? "yes." : "no.");
 
 	snprintf(host_log, sizeof(host_log), "vmwgfx: %s-%s",
 		VMWGFX_REPO, VMWGFX_GIT_VERSION);
@@ -1278,8 +1322,6 @@ static void vmw_master_drop(struct drm_device *dev,
 	dev_priv->active_master = &dev_priv->fbdev_master;
 	ttm_lock_set_kill(&dev_priv->fbdev_master.lock, false, SIGTERM);
 	ttm_vt_unlock(&dev_priv->fbdev_master.lock);
-
-	vmw_fb_refresh(dev_priv);
 }
 
 /**
@@ -1483,7 +1525,6 @@ static int vmw_pm_freeze(struct device *kdev)
 			vmw_kms_resume(dev);
 		if (dev_priv->enable_fb)
 			vmw_fb_on(dev_priv);
-		vmw_fb_refresh(dev_priv);
 		return -EBUSY;
 	}
 
@@ -1523,8 +1564,6 @@ static int vmw_pm_restore(struct device *kdev)
 	if (dev_priv->enable_fb)
 		vmw_fb_on(dev_priv);
 
-	vmw_fb_refresh(dev_priv);
-
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
index f34f368c1a2e..1abe21758b0d 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
@@ -1,7 +1,7 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
- * Copyright © 2009-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -43,10 +43,10 @@
 #include <linux/sync_file.h>
 
 #define VMWGFX_DRIVER_NAME "vmwgfx"
-#define VMWGFX_DRIVER_DATE "20180322"
+#define VMWGFX_DRIVER_DATE "20180704"
 #define VMWGFX_DRIVER_MAJOR 2
-#define VMWGFX_DRIVER_MINOR 14
-#define VMWGFX_DRIVER_PATCHLEVEL 1
+#define VMWGFX_DRIVER_MINOR 15
+#define VMWGFX_DRIVER_PATCHLEVEL 0
 #define VMWGFX_FILE_PAGE_OFFSET 0x00100000
 #define VMWGFX_FIFO_STATIC_SIZE (1024*1024)
 #define VMWGFX_MAX_RELOCATIONS 2048
@@ -83,10 +83,10 @@
 struct vmw_fpriv {
 	struct drm_master *locked_master;
 	struct ttm_object_file *tfile;
-	bool gb_aware;
+	bool gb_aware; /* user-space is guest-backed aware */
 };
 
-struct vmw_dma_buffer {
+struct vmw_buffer_object {
 	struct ttm_buffer_object base;
 	struct list_head res_list;
 	s32 pin_count;
@@ -120,7 +120,7 @@ struct vmw_resource {
 	unsigned long backup_size;
 	bool res_dirty; /* Protected by backup buffer reserved */
 	bool backup_dirty; /* Protected by backup buffer reserved */
-	struct vmw_dma_buffer *backup;
+	struct vmw_buffer_object *backup;
 	unsigned long backup_offset;
 	unsigned long pin_count; /* Protected by resource reserved */
 	const struct vmw_res_func *func;
@@ -166,7 +166,7 @@ struct vmw_surface_offset;
 
 struct vmw_surface {
 	struct vmw_resource res;
-	uint32_t flags;
+	SVGA3dSurfaceAllFlags flags;
 	uint32_t format;
 	uint32_t mip_levels[DRM_VMW_MAX_SURFACE_FACES];
 	struct drm_vmw_size base_size;
@@ -180,6 +180,8 @@ struct vmw_surface {
 	SVGA3dTextureFilter autogen_filter;
 	uint32_t multisample_count;
 	struct list_head view_list;
+	SVGA3dMSPattern multisample_pattern;
+	SVGA3dMSQualityLevel quality_level;
 };
 
 struct vmw_marker_queue {
@@ -304,7 +306,7 @@ struct vmw_sw_context{
 	uint32_t cmd_bounce_size;
 	struct list_head resource_list;
 	struct list_head ctx_resource_list; /* For contexts and cotables */
-	struct vmw_dma_buffer *cur_query_bo;
+	struct vmw_buffer_object *cur_query_bo;
 	struct list_head res_relocations;
 	uint32_t *buf_start;
 	struct vmw_res_cache_entry res_cache[vmw_res_max];
@@ -315,7 +317,7 @@ struct vmw_sw_context{
 	bool staged_bindings_inuse;
 	struct list_head staged_cmd_res;
 	struct vmw_resource_val_node *dx_ctx_node;
-	struct vmw_dma_buffer *dx_query_mob;
+	struct vmw_buffer_object *dx_query_mob;
 	struct vmw_resource *dx_query_ctx;
 	struct vmw_cmdbuf_res_manager *man;
 };
@@ -386,6 +388,7 @@ struct vmw_private {
 	uint32_t initial_height;
 	u32 *mmio_virt;
 	uint32_t capabilities;
+	uint32_t capabilities2;
 	uint32_t max_gmr_ids;
 	uint32_t max_gmr_pages;
 	uint32_t max_mob_pages;
@@ -397,6 +400,7 @@ struct vmw_private {
 	spinlock_t cap_lock;
 	bool has_dx;
 	bool assume_16bpp;
+	bool has_sm4_1;
 
 	/*
 	 * VGA registers.
@@ -412,6 +416,15 @@ struct vmw_private {
 	uint32_t num_displays;
 
 	/*
+	 * Currently requested_layout_mutex is used to protect the gui
+	 * positionig state in display unit. With that use case currently this
+	 * mutex is only taken during layout ioctl and atomic check_modeset.
+	 * Other display unit state can be protected with this mutex but that
+	 * needs careful consideration.
+	 */
+	struct mutex requested_layout_mutex;
+
+	/*
 	 * Framebuffer info.
 	 */
 
@@ -513,8 +526,8 @@ struct vmw_private {
 	 * are protected by the cmdbuf mutex.
 	 */
 
-	struct vmw_dma_buffer *dummy_query_bo;
-	struct vmw_dma_buffer *pinned_bo;
+	struct vmw_buffer_object *dummy_query_bo;
+	struct vmw_buffer_object *pinned_bo;
 	uint32_t query_cid;
 	uint32_t query_cid_valid;
 	bool dummy_query_bo_pinned;
@@ -623,43 +636,13 @@ extern int vmw_user_lookup_handle(struct vmw_private *dev_priv,
 				  struct ttm_object_file *tfile,
 				  uint32_t handle,
 				  struct vmw_surface **out_surf,
-				  struct vmw_dma_buffer **out_buf);
+				  struct vmw_buffer_object **out_buf);
 extern int vmw_user_resource_lookup_handle(
 	struct vmw_private *dev_priv,
 	struct ttm_object_file *tfile,
 	uint32_t handle,
 	const struct vmw_user_resource_conv *converter,
 	struct vmw_resource **p_res);
-extern void vmw_dmabuf_bo_free(struct ttm_buffer_object *bo);
-extern int vmw_dmabuf_init(struct vmw_private *dev_priv,
-			   struct vmw_dma_buffer *vmw_bo,
-			   size_t size, struct ttm_placement *placement,
-			   bool interuptable,
-			   void (*bo_free) (struct ttm_buffer_object *bo));
-extern int vmw_user_dmabuf_verify_access(struct ttm_buffer_object *bo,
-				  struct ttm_object_file *tfile);
-extern int vmw_user_dmabuf_alloc(struct vmw_private *dev_priv,
-				 struct ttm_object_file *tfile,
-				 uint32_t size,
-				 bool shareable,
-				 uint32_t *handle,
-				 struct vmw_dma_buffer **p_dma_buf,
-				 struct ttm_base_object **p_base);
-extern int vmw_user_dmabuf_reference(struct ttm_object_file *tfile,
-				     struct vmw_dma_buffer *dma_buf,
-				     uint32_t *handle);
-extern int vmw_dmabuf_alloc_ioctl(struct drm_device *dev, void *data,
-				  struct drm_file *file_priv);
-extern int vmw_dmabuf_unref_ioctl(struct drm_device *dev, void *data,
-				  struct drm_file *file_priv);
-extern int vmw_user_dmabuf_synccpu_ioctl(struct drm_device *dev, void *data,
-					 struct drm_file *file_priv);
-extern uint32_t vmw_dmabuf_validate_node(struct ttm_buffer_object *bo,
-					 uint32_t cur_validate_node);
-extern void vmw_dmabuf_validate_clear(struct ttm_buffer_object *bo);
-extern int vmw_user_dmabuf_lookup(struct ttm_object_file *tfile,
-				  uint32_t id, struct vmw_dma_buffer **out,
-				  struct ttm_base_object **base);
 extern int vmw_stream_claim_ioctl(struct drm_device *dev, void *data,
 				  struct drm_file *file_priv);
 extern int vmw_stream_unref_ioctl(struct drm_device *dev, void *data,
@@ -670,43 +653,70 @@ extern int vmw_user_stream_lookup(struct vmw_private *dev_priv,
 				  struct vmw_resource **out);
 extern void vmw_resource_unreserve(struct vmw_resource *res,
 				   bool switch_backup,
-				   struct vmw_dma_buffer *new_backup,
+				   struct vmw_buffer_object *new_backup,
 				   unsigned long new_backup_offset);
-extern void vmw_resource_move_notify(struct ttm_buffer_object *bo,
-				     struct ttm_mem_reg *mem);
 extern void vmw_query_move_notify(struct ttm_buffer_object *bo,
 				  struct ttm_mem_reg *mem);
-extern void vmw_resource_swap_notify(struct ttm_buffer_object *bo);
-extern int vmw_query_readback_all(struct vmw_dma_buffer *dx_query_mob);
-extern void vmw_fence_single_bo(struct ttm_buffer_object *bo,
-				struct vmw_fence_obj *fence);
+extern int vmw_query_readback_all(struct vmw_buffer_object *dx_query_mob);
 extern void vmw_resource_evict_all(struct vmw_private *dev_priv);
-
+extern void vmw_resource_unbind_list(struct vmw_buffer_object *vbo);
 
 /**
- * DMA buffer helper routines - vmwgfx_dmabuf.c
+ * Buffer object helper functions - vmwgfx_bo.c
  */
-extern int vmw_dmabuf_pin_in_placement(struct vmw_private *vmw_priv,
-				       struct vmw_dma_buffer *bo,
-				       struct ttm_placement *placement,
+extern int vmw_bo_pin_in_placement(struct vmw_private *vmw_priv,
+				   struct vmw_buffer_object *bo,
+				   struct ttm_placement *placement,
+				   bool interruptible);
+extern int vmw_bo_pin_in_vram(struct vmw_private *dev_priv,
+			      struct vmw_buffer_object *buf,
+			      bool interruptible);
+extern int vmw_bo_pin_in_vram_or_gmr(struct vmw_private *dev_priv,
+				     struct vmw_buffer_object *buf,
+				     bool interruptible);
+extern int vmw_bo_pin_in_start_of_vram(struct vmw_private *vmw_priv,
+				       struct vmw_buffer_object *bo,
 				       bool interruptible);
-extern int vmw_dmabuf_pin_in_vram(struct vmw_private *dev_priv,
-				  struct vmw_dma_buffer *buf,
-				  bool interruptible);
-extern int vmw_dmabuf_pin_in_vram_or_gmr(struct vmw_private *dev_priv,
-					 struct vmw_dma_buffer *buf,
-					 bool interruptible);
-extern int vmw_dmabuf_pin_in_start_of_vram(struct vmw_private *vmw_priv,
-					   struct vmw_dma_buffer *bo,
-					   bool interruptible);
-extern int vmw_dmabuf_unpin(struct vmw_private *vmw_priv,
-			    struct vmw_dma_buffer *bo,
-			    bool interruptible);
+extern int vmw_bo_unpin(struct vmw_private *vmw_priv,
+			struct vmw_buffer_object *bo,
+			bool interruptible);
 extern void vmw_bo_get_guest_ptr(const struct ttm_buffer_object *buf,
 				 SVGAGuestPtr *ptr);
-extern void vmw_bo_pin_reserved(struct vmw_dma_buffer *bo, bool pin);
-extern void *vmw_dma_buffer_map_and_cache(struct vmw_dma_buffer *vbo);
-extern void vmw_dma_buffer_unmap(struct vmw_dma_buffer *vbo);
+extern void vmw_bo_pin_reserved(struct vmw_buffer_object *bo, bool pin);
+extern void vmw_bo_bo_free(struct ttm_buffer_object *bo);
+extern int vmw_bo_init(struct vmw_private *dev_priv,
+		       struct vmw_buffer_object *vmw_bo,
+		       size_t size, struct ttm_placement *placement,
+		       bool interuptable,
+		       void (*bo_free)(struct ttm_buffer_object *bo));
+extern int vmw_user_bo_verify_access(struct ttm_buffer_object *bo,
+				     struct ttm_object_file *tfile);
+extern int vmw_user_bo_alloc(struct vmw_private *dev_priv,
+			     struct ttm_object_file *tfile,
+			     uint32_t size,
+			     bool shareable,
+			     uint32_t *handle,
+			     struct vmw_buffer_object **p_dma_buf,
+			     struct ttm_base_object **p_base);
+extern int vmw_user_bo_reference(struct ttm_object_file *tfile,
+				 struct vmw_buffer_object *dma_buf,
+				 uint32_t *handle);
+extern int vmw_bo_alloc_ioctl(struct drm_device *dev, void *data,
+			      struct drm_file *file_priv);
+extern int vmw_bo_unref_ioctl(struct drm_device *dev, void *data,
+			      struct drm_file *file_priv);
+extern int vmw_user_bo_synccpu_ioctl(struct drm_device *dev, void *data,
+				     struct drm_file *file_priv);
+extern int vmw_user_bo_lookup(struct ttm_object_file *tfile,
+			      uint32_t id, struct vmw_buffer_object **out,
+			      struct ttm_base_object **base);
+extern void vmw_bo_fence_single(struct ttm_buffer_object *bo,
+				struct vmw_fence_obj *fence);
+extern void *vmw_bo_map_and_cache(struct vmw_buffer_object *vbo);
+extern void vmw_bo_unmap(struct vmw_buffer_object *vbo);
+extern void vmw_bo_move_notify(struct ttm_buffer_object *bo,
+			       struct ttm_mem_reg *mem);
+extern void vmw_bo_swap_notify(struct ttm_buffer_object *bo);
 
 /**
  * Misc Ioctl functionality - vmwgfx_ioctl.c
@@ -758,7 +768,7 @@ extern void vmw_ttm_global_release(struct vmw_private *dev_priv);
 extern int vmw_mmap(struct file *filp, struct vm_area_struct *vma);
 
 /**
- * TTM buffer object driver - vmwgfx_buffer.c
+ * TTM buffer object driver - vmwgfx_ttm_buffer.c
  */
 
 extern const size_t vmw_tt_size;
@@ -910,7 +920,6 @@ int vmw_fb_init(struct vmw_private *vmw_priv);
 int vmw_fb_close(struct vmw_private *dev_priv);
 int vmw_fb_off(struct vmw_private *vmw_priv);
 int vmw_fb_on(struct vmw_private *vmw_priv);
-void vmw_fb_refresh(struct vmw_private *vmw_priv);
 
 /**
  * Kernel modesetting - vmwgfx_kms.c
@@ -1042,8 +1051,8 @@ vmw_context_binding_state(struct vmw_resource *ctx);
 extern void vmw_dx_context_scrub_cotables(struct vmw_resource *ctx,
 					  bool readback);
 extern int vmw_context_bind_dx_query(struct vmw_resource *ctx_res,
-				     struct vmw_dma_buffer *mob);
-extern struct vmw_dma_buffer *
+				     struct vmw_buffer_object *mob);
+extern struct vmw_buffer_object *
 vmw_context_get_dx_query_mob(struct vmw_resource *ctx_res);
 
 
@@ -1071,14 +1080,22 @@ extern int vmw_surface_validate(struct vmw_private *dev_priv,
 				struct vmw_surface *srf);
 int vmw_surface_gb_priv_define(struct drm_device *dev,
 			       uint32_t user_accounting_size,
-			       uint32_t svga3d_flags,
+			       SVGA3dSurfaceAllFlags svga3d_flags,
 			       SVGA3dSurfaceFormat format,
 			       bool for_scanout,
 			       uint32_t num_mip_levels,
 			       uint32_t multisample_count,
 			       uint32_t array_size,
 			       struct drm_vmw_size size,
+			       SVGA3dMSPattern multisample_pattern,
+			       SVGA3dMSQualityLevel quality_level,
 			       struct vmw_surface **srf_out);
+extern int vmw_gb_surface_define_ext_ioctl(struct drm_device *dev,
+					   void *data,
+					   struct drm_file *file_priv);
+extern int vmw_gb_surface_reference_ext_ioctl(struct drm_device *dev,
+					      void *data,
+					      struct drm_file *file_priv);
 
 /*
  * Shader management - vmwgfx_shader.c
@@ -1225,6 +1242,11 @@ int vmw_bo_cpu_blit(struct ttm_buffer_object *dst,
 		    u32 w, u32 h,
 		    struct vmw_diff_cpy *diff);
 
+/* Host messaging -vmwgfx_msg.c: */
+int vmw_host_get_guestinfo(const char *guest_info_param,
+			   char *buffer, size_t *length);
+int vmw_host_log(const char *log);
+
 /**
  * Inline helper functions
  */
@@ -1244,9 +1266,9 @@ static inline struct vmw_surface *vmw_surface_reference(struct vmw_surface *srf)
 	return srf;
 }
 
-static inline void vmw_dmabuf_unreference(struct vmw_dma_buffer **buf)
+static inline void vmw_bo_unreference(struct vmw_buffer_object **buf)
 {
-	struct vmw_dma_buffer *tmp_buf = *buf;
+	struct vmw_buffer_object *tmp_buf = *buf;
 
 	*buf = NULL;
 	if (tmp_buf != NULL) {
@@ -1256,7 +1278,8 @@ static inline void vmw_dmabuf_unreference(struct vmw_dma_buffer **buf)
 	}
 }
 
-static inline struct vmw_dma_buffer *vmw_dmabuf_reference(struct vmw_dma_buffer *buf)
+static inline struct vmw_buffer_object *
+vmw_bo_reference(struct vmw_buffer_object *buf)
 {
 	if (ttm_bo_reference(&buf->base))
 		return buf;
@@ -1303,10 +1326,4 @@ static inline void vmw_mmio_write(u32 value, u32 *addr)
 {
 	WRITE_ONCE(*addr, value);
 }
-
-/**
- * Add vmw_msg module function
- */
-extern int vmw_host_log(const char *log);
-
 #endif
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
index c9d5cc237124..1f134570b759 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_execbuf.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009 - 2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009 - 2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -92,7 +92,7 @@ struct vmw_resource_val_node {
 	struct list_head head;
 	struct drm_hash_item hash;
 	struct vmw_resource *res;
-	struct vmw_dma_buffer *new_backup;
+	struct vmw_buffer_object *new_backup;
 	struct vmw_ctx_binding_state *staged_bindings;
 	unsigned long new_backup_offset;
 	u32 first_usage : 1;
@@ -126,9 +126,9 @@ static int vmw_resource_context_res_add(struct vmw_private *dev_priv,
 static int vmw_translate_mob_ptr(struct vmw_private *dev_priv,
 				 struct vmw_sw_context *sw_context,
 				 SVGAMobId *id,
-				 struct vmw_dma_buffer **vmw_bo_p);
+				 struct vmw_buffer_object **vmw_bo_p);
 static int vmw_bo_to_validate_list(struct vmw_sw_context *sw_context,
-				   struct vmw_dma_buffer *vbo,
+				   struct vmw_buffer_object *vbo,
 				   bool validate_as_mob,
 				   uint32_t *p_val_node);
 /**
@@ -185,7 +185,7 @@ static void vmw_resources_unreserve(struct vmw_sw_context *sw_context,
 		}
 		vmw_resource_unreserve(res, switch_backup, val->new_backup,
 				       val->new_backup_offset);
-		vmw_dmabuf_unreference(&val->new_backup);
+		vmw_bo_unreference(&val->new_backup);
 	}
 }
 
@@ -423,7 +423,7 @@ static int vmw_resource_context_res_add(struct vmw_private *dev_priv,
 	}
 
 	if (dev_priv->has_dx && vmw_res_type(ctx) == vmw_res_dx_context) {
-		struct vmw_dma_buffer *dx_query_mob;
+		struct vmw_buffer_object *dx_query_mob;
 
 		dx_query_mob = vmw_context_get_dx_query_mob(ctx);
 		if (dx_query_mob)
@@ -544,7 +544,7 @@ static int vmw_cmd_ok(struct vmw_private *dev_priv,
  * submission is reached.
  */
 static int vmw_bo_to_validate_list(struct vmw_sw_context *sw_context,
-				   struct vmw_dma_buffer *vbo,
+				   struct vmw_buffer_object *vbo,
 				   bool validate_as_mob,
 				   uint32_t *p_val_node)
 {
@@ -616,7 +616,7 @@ static int vmw_resources_reserve(struct vmw_sw_context *sw_context)
 			return ret;
 
 		if (res->backup) {
-			struct vmw_dma_buffer *vbo = res->backup;
+			struct vmw_buffer_object *vbo = res->backup;
 
 			ret = vmw_bo_to_validate_list
 				(sw_context, vbo,
@@ -628,7 +628,7 @@ static int vmw_resources_reserve(struct vmw_sw_context *sw_context)
 	}
 
 	if (sw_context->dx_query_mob) {
-		struct vmw_dma_buffer *expected_dx_query_mob;
+		struct vmw_buffer_object *expected_dx_query_mob;
 
 		expected_dx_query_mob =
 			vmw_context_get_dx_query_mob(sw_context->dx_query_ctx);
@@ -657,7 +657,7 @@ static int vmw_resources_validate(struct vmw_sw_context *sw_context)
 
 	list_for_each_entry(val, &sw_context->resource_list, head) {
 		struct vmw_resource *res = val->res;
-		struct vmw_dma_buffer *backup = res->backup;
+		struct vmw_buffer_object *backup = res->backup;
 
 		ret = vmw_resource_validate(res);
 		if (unlikely(ret != 0)) {
@@ -668,7 +668,7 @@ static int vmw_resources_validate(struct vmw_sw_context *sw_context)
 
 		/* Check if the resource switched backup buffer */
 		if (backup && res->backup && (backup != res->backup)) {
-			struct vmw_dma_buffer *vbo = res->backup;
+			struct vmw_buffer_object *vbo = res->backup;
 
 			ret = vmw_bo_to_validate_list
 				(sw_context, vbo,
@@ -821,7 +821,7 @@ out_no_reloc:
 static int vmw_rebind_all_dx_query(struct vmw_resource *ctx_res)
 {
 	struct vmw_private *dev_priv = ctx_res->dev_priv;
-	struct vmw_dma_buffer *dx_query_mob;
+	struct vmw_buffer_object *dx_query_mob;
 	struct {
 		SVGA3dCmdHeader header;
 		SVGA3dCmdDXBindAllQuery body;
@@ -1152,7 +1152,7 @@ static int vmw_cmd_present_check(struct vmw_private *dev_priv,
  * command batch.
  */
 static int vmw_query_bo_switch_prepare(struct vmw_private *dev_priv,
-				       struct vmw_dma_buffer *new_query_bo,
+				       struct vmw_buffer_object *new_query_bo,
 				       struct vmw_sw_context *sw_context)
 {
 	struct vmw_res_cache_entry *ctx_entry =
@@ -1234,7 +1234,7 @@ static void vmw_query_bo_switch_commit(struct vmw_private *dev_priv,
 	if (dev_priv->pinned_bo != sw_context->cur_query_bo) {
 		if (dev_priv->pinned_bo) {
 			vmw_bo_pin_reserved(dev_priv->pinned_bo, false);
-			vmw_dmabuf_unreference(&dev_priv->pinned_bo);
+			vmw_bo_unreference(&dev_priv->pinned_bo);
 		}
 
 		if (!sw_context->needs_post_query_barrier) {
@@ -1256,7 +1256,7 @@ static void vmw_query_bo_switch_commit(struct vmw_private *dev_priv,
 			dev_priv->query_cid = sw_context->last_query_ctx->id;
 			dev_priv->query_cid_valid = true;
 			dev_priv->pinned_bo =
-				vmw_dmabuf_reference(sw_context->cur_query_bo);
+				vmw_bo_reference(sw_context->cur_query_bo);
 		}
 	}
 }
@@ -1282,15 +1282,14 @@ static void vmw_query_bo_switch_commit(struct vmw_private *dev_priv,
 static int vmw_translate_mob_ptr(struct vmw_private *dev_priv,
 				 struct vmw_sw_context *sw_context,
 				 SVGAMobId *id,
-				 struct vmw_dma_buffer **vmw_bo_p)
+				 struct vmw_buffer_object **vmw_bo_p)
 {
-	struct vmw_dma_buffer *vmw_bo = NULL;
+	struct vmw_buffer_object *vmw_bo = NULL;
 	uint32_t handle = *id;
 	struct vmw_relocation *reloc;
 	int ret;
 
-	ret = vmw_user_dmabuf_lookup(sw_context->fp->tfile, handle, &vmw_bo,
-				     NULL);
+	ret = vmw_user_bo_lookup(sw_context->fp->tfile, handle, &vmw_bo, NULL);
 	if (unlikely(ret != 0)) {
 		DRM_ERROR("Could not find or use MOB buffer.\n");
 		ret = -EINVAL;
@@ -1316,7 +1315,7 @@ static int vmw_translate_mob_ptr(struct vmw_private *dev_priv,
 	return 0;
 
 out_no_reloc:
-	vmw_dmabuf_unreference(&vmw_bo);
+	vmw_bo_unreference(&vmw_bo);
 	*vmw_bo_p = NULL;
 	return ret;
 }
@@ -1343,15 +1342,14 @@ out_no_reloc:
 static int vmw_translate_guest_ptr(struct vmw_private *dev_priv,
 				   struct vmw_sw_context *sw_context,
 				   SVGAGuestPtr *ptr,
-				   struct vmw_dma_buffer **vmw_bo_p)
+				   struct vmw_buffer_object **vmw_bo_p)
 {
-	struct vmw_dma_buffer *vmw_bo = NULL;
+	struct vmw_buffer_object *vmw_bo = NULL;
 	uint32_t handle = ptr->gmrId;
 	struct vmw_relocation *reloc;
 	int ret;
 
-	ret = vmw_user_dmabuf_lookup(sw_context->fp->tfile, handle, &vmw_bo,
-				     NULL);
+	ret = vmw_user_bo_lookup(sw_context->fp->tfile, handle, &vmw_bo, NULL);
 	if (unlikely(ret != 0)) {
 		DRM_ERROR("Could not find or use GMR region.\n");
 		ret = -EINVAL;
@@ -1376,7 +1374,7 @@ static int vmw_translate_guest_ptr(struct vmw_private *dev_priv,
 	return 0;
 
 out_no_reloc:
-	vmw_dmabuf_unreference(&vmw_bo);
+	vmw_bo_unreference(&vmw_bo);
 	*vmw_bo_p = NULL;
 	return ret;
 }
@@ -1447,7 +1445,7 @@ static int vmw_cmd_dx_bind_query(struct vmw_private *dev_priv,
 		SVGA3dCmdDXBindQuery q;
 	} *cmd;
 
-	struct vmw_dma_buffer *vmw_bo;
+	struct vmw_buffer_object *vmw_bo;
 	int    ret;
 
 
@@ -1466,7 +1464,7 @@ static int vmw_cmd_dx_bind_query(struct vmw_private *dev_priv,
 	sw_context->dx_query_mob = vmw_bo;
 	sw_context->dx_query_ctx = sw_context->dx_ctx_node->res;
 
-	vmw_dmabuf_unreference(&vmw_bo);
+	vmw_bo_unreference(&vmw_bo);
 
 	return ret;
 }
@@ -1549,7 +1547,7 @@ static int vmw_cmd_end_gb_query(struct vmw_private *dev_priv,
 				struct vmw_sw_context *sw_context,
 				SVGA3dCmdHeader *header)
 {
-	struct vmw_dma_buffer *vmw_bo;
+	struct vmw_buffer_object *vmw_bo;
 	struct vmw_query_cmd {
 		SVGA3dCmdHeader header;
 		SVGA3dCmdEndGBQuery q;
@@ -1569,7 +1567,7 @@ static int vmw_cmd_end_gb_query(struct vmw_private *dev_priv,
 
 	ret = vmw_query_bo_switch_prepare(dev_priv, vmw_bo, sw_context);
 
-	vmw_dmabuf_unreference(&vmw_bo);
+	vmw_bo_unreference(&vmw_bo);
 	return ret;
 }
 
@@ -1584,7 +1582,7 @@ static int vmw_cmd_end_query(struct vmw_private *dev_priv,
 			     struct vmw_sw_context *sw_context,
 			     SVGA3dCmdHeader *header)
 {
-	struct vmw_dma_buffer *vmw_bo;
+	struct vmw_buffer_object *vmw_bo;
 	struct vmw_query_cmd {
 		SVGA3dCmdHeader header;
 		SVGA3dCmdEndQuery q;
@@ -1623,7 +1621,7 @@ static int vmw_cmd_end_query(struct vmw_private *dev_priv,
 
 	ret = vmw_query_bo_switch_prepare(dev_priv, vmw_bo, sw_context);
 
-	vmw_dmabuf_unreference(&vmw_bo);
+	vmw_bo_unreference(&vmw_bo);
 	return ret;
 }
 
@@ -1638,7 +1636,7 @@ static int vmw_cmd_wait_gb_query(struct vmw_private *dev_priv,
 				 struct vmw_sw_context *sw_context,
 				 SVGA3dCmdHeader *header)
 {
-	struct vmw_dma_buffer *vmw_bo;
+	struct vmw_buffer_object *vmw_bo;
 	struct vmw_query_cmd {
 		SVGA3dCmdHeader header;
 		SVGA3dCmdWaitForGBQuery q;
@@ -1656,7 +1654,7 @@ static int vmw_cmd_wait_gb_query(struct vmw_private *dev_priv,
 	if (unlikely(ret != 0))
 		return ret;
 
-	vmw_dmabuf_unreference(&vmw_bo);
+	vmw_bo_unreference(&vmw_bo);
 	return 0;
 }
 
@@ -1671,7 +1669,7 @@ static int vmw_cmd_wait_query(struct vmw_private *dev_priv,
 			      struct vmw_sw_context *sw_context,
 			      SVGA3dCmdHeader *header)
 {
-	struct vmw_dma_buffer *vmw_bo;
+	struct vmw_buffer_object *vmw_bo;
 	struct vmw_query_cmd {
 		SVGA3dCmdHeader header;
 		SVGA3dCmdWaitForQuery q;
@@ -1708,7 +1706,7 @@ static int vmw_cmd_wait_query(struct vmw_private *dev_priv,
 	if (unlikely(ret != 0))
 		return ret;
 
-	vmw_dmabuf_unreference(&vmw_bo);
+	vmw_bo_unreference(&vmw_bo);
 	return 0;
 }
 
@@ -1716,7 +1714,7 @@ static int vmw_cmd_dma(struct vmw_private *dev_priv,
 		       struct vmw_sw_context *sw_context,
 		       SVGA3dCmdHeader *header)
 {
-	struct vmw_dma_buffer *vmw_bo = NULL;
+	struct vmw_buffer_object *vmw_bo = NULL;
 	struct vmw_surface *srf = NULL;
 	struct vmw_dma_cmd {
 		SVGA3dCmdHeader header;
@@ -1768,7 +1766,7 @@ static int vmw_cmd_dma(struct vmw_private *dev_priv,
 			     header);
 
 out_no_surface:
-	vmw_dmabuf_unreference(&vmw_bo);
+	vmw_bo_unreference(&vmw_bo);
 	return ret;
 }
 
@@ -1887,7 +1885,7 @@ static int vmw_cmd_check_define_gmrfb(struct vmw_private *dev_priv,
 				      struct vmw_sw_context *sw_context,
 				      void *buf)
 {
-	struct vmw_dma_buffer *vmw_bo;
+	struct vmw_buffer_object *vmw_bo;
 	int ret;
 
 	struct {
@@ -1901,7 +1899,7 @@ static int vmw_cmd_check_define_gmrfb(struct vmw_private *dev_priv,
 	if (unlikely(ret != 0))
 		return ret;
 
-	vmw_dmabuf_unreference(&vmw_bo);
+	vmw_bo_unreference(&vmw_bo);
 
 	return ret;
 }
@@ -1928,7 +1926,7 @@ static int vmw_cmd_res_switch_backup(struct vmw_private *dev_priv,
 				     uint32_t *buf_id,
 				     unsigned long backup_offset)
 {
-	struct vmw_dma_buffer *dma_buf;
+	struct vmw_buffer_object *dma_buf;
 	int ret;
 
 	ret = vmw_translate_mob_ptr(dev_priv, sw_context, buf_id, &dma_buf);
@@ -1939,7 +1937,7 @@ static int vmw_cmd_res_switch_backup(struct vmw_private *dev_priv,
 	if (val_node->first_usage)
 		val_node->no_buffer_needed = true;
 
-	vmw_dmabuf_unreference(&val_node->new_backup);
+	vmw_bo_unreference(&val_node->new_backup);
 	val_node->new_backup = dma_buf;
 	val_node->new_backup_offset = backup_offset;
 
@@ -3118,6 +3116,32 @@ static int vmw_cmd_dx_transfer_from_buffer(struct vmw_private *dev_priv,
 				 &cmd->body.destSid, NULL);
 }
 
+/**
+ * vmw_cmd_intra_surface_copy -
+ * Validate an SVGA_3D_CMD_INTRA_SURFACE_COPY command
+ *
+ * @dev_priv: Pointer to a device private struct.
+ * @sw_context: The software context being used for this batch.
+ * @header: Pointer to the command header in the command stream.
+ */
+static int vmw_cmd_intra_surface_copy(struct vmw_private *dev_priv,
+					   struct vmw_sw_context *sw_context,
+					   SVGA3dCmdHeader *header)
+{
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdIntraSurfaceCopy body;
+	} *cmd = container_of(header, typeof(*cmd), header);
+
+	if (!(dev_priv->capabilities2 & SVGA_CAP2_INTRA_SURFACE_COPY))
+		return -EINVAL;
+
+	return vmw_cmd_res_check(dev_priv, sw_context, vmw_res_surface,
+				user_surface_converter,
+				&cmd->body.surface.sid, NULL);
+}
+
+
 static int vmw_cmd_check_not_3d(struct vmw_private *dev_priv,
 				struct vmw_sw_context *sw_context,
 				void *buf, uint32_t *size)
@@ -3232,9 +3256,9 @@ static const struct vmw_cmd_entry vmw_cmd_entries[SVGA_3D_CMD_MAX] = {
 		    false, false, false),
 	VMW_CMD_DEF(SVGA_3D_CMD_SCREEN_DMA, &vmw_cmd_invalid,
 		    false, false, false),
-	VMW_CMD_DEF(SVGA_3D_CMD_SET_UNITY_SURFACE_COOKIE, &vmw_cmd_invalid,
+	VMW_CMD_DEF(SVGA_3D_CMD_DEAD1, &vmw_cmd_invalid,
 		    false, false, false),
-	VMW_CMD_DEF(SVGA_3D_CMD_OPEN_CONTEXT_SURFACE, &vmw_cmd_invalid,
+	VMW_CMD_DEF(SVGA_3D_CMD_DEAD2, &vmw_cmd_invalid,
 		    false, false, false),
 	VMW_CMD_DEF(SVGA_3D_CMD_LOGICOPS_BITBLT, &vmw_cmd_invalid,
 		    false, false, false),
@@ -3473,6 +3497,8 @@ static const struct vmw_cmd_entry vmw_cmd_entries[SVGA_3D_CMD_MAX] = {
 	VMW_CMD_DEF(SVGA_3D_CMD_DX_TRANSFER_FROM_BUFFER,
 		    &vmw_cmd_dx_transfer_from_buffer,
 		    true, false, true),
+	VMW_CMD_DEF(SVGA_3D_CMD_INTRA_SURFACE_COPY, &vmw_cmd_intra_surface_copy,
+		    true, false, true),
 };
 
 bool vmw_cmd_describe(const void *buf, u32 *size, char const **cmd)
@@ -3701,8 +3727,8 @@ int vmw_validate_single_buffer(struct vmw_private *dev_priv,
 			       bool interruptible,
 			       bool validate_as_mob)
 {
-	struct vmw_dma_buffer *vbo = container_of(bo, struct vmw_dma_buffer,
-						  base);
+	struct vmw_buffer_object *vbo =
+		container_of(bo, struct vmw_buffer_object, base);
 	struct ttm_operation_ctx ctx = { interruptible, true };
 	int ret;
 
@@ -4423,7 +4449,7 @@ void __vmw_execbuf_release_pinned_bo(struct vmw_private *dev_priv,
 
 	ttm_bo_unref(&query_val.bo);
 	ttm_bo_unref(&pinned_val.bo);
-	vmw_dmabuf_unreference(&dev_priv->pinned_bo);
+	vmw_bo_unreference(&dev_priv->pinned_bo);
 out_unlock:
 	return;
 
@@ -4432,7 +4458,7 @@ out_no_emit:
 out_no_reserve:
 	ttm_bo_unref(&query_val.bo);
 	ttm_bo_unref(&pinned_val.bo);
-	vmw_dmabuf_unreference(&dev_priv->pinned_bo);
+	vmw_bo_unreference(&dev_priv->pinned_bo);
 }
 
 /**
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
index 2582ffd36bb5..b913a56f3426 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
@@ -42,7 +42,7 @@ struct vmw_fb_par {
 	void *vmalloc;
 
 	struct mutex bo_mutex;
-	struct vmw_dma_buffer *vmw_bo;
+	struct vmw_buffer_object *vmw_bo;
 	unsigned bo_size;
 	struct drm_framebuffer *set_fb;
 	struct drm_display_mode *set_mode;
@@ -184,7 +184,7 @@ static void vmw_fb_dirty_flush(struct work_struct *work)
 	struct drm_clip_rect clip;
 	struct drm_framebuffer *cur_fb;
 	u8 *src_ptr, *dst_ptr;
-	struct vmw_dma_buffer *vbo = par->vmw_bo;
+	struct vmw_buffer_object *vbo = par->vmw_bo;
 	void *virtual;
 
 	if (!READ_ONCE(par->dirty.active))
@@ -197,7 +197,7 @@ static void vmw_fb_dirty_flush(struct work_struct *work)
 
 	(void) ttm_read_lock(&vmw_priv->reservation_sem, false);
 	(void) ttm_bo_reserve(&vbo->base, false, false, NULL);
-	virtual = vmw_dma_buffer_map_and_cache(vbo);
+	virtual = vmw_bo_map_and_cache(vbo);
 	if (!virtual)
 		goto out_unreserve;
 
@@ -391,9 +391,9 @@ static void vmw_fb_imageblit(struct fb_info *info, const struct fb_image *image)
  */
 
 static int vmw_fb_create_bo(struct vmw_private *vmw_priv,
-			    size_t size, struct vmw_dma_buffer **out)
+			    size_t size, struct vmw_buffer_object **out)
 {
-	struct vmw_dma_buffer *vmw_bo;
+	struct vmw_buffer_object *vmw_bo;
 	int ret;
 
 	(void) ttm_write_lock(&vmw_priv->reservation_sem, false);
@@ -404,10 +404,10 @@ static int vmw_fb_create_bo(struct vmw_private *vmw_priv,
 		goto err_unlock;
 	}
 
-	ret = vmw_dmabuf_init(vmw_priv, vmw_bo, size,
+	ret = vmw_bo_init(vmw_priv, vmw_bo, size,
 			      &vmw_sys_placement,
 			      false,
-			      &vmw_dmabuf_bo_free);
+			      &vmw_bo_bo_free);
 	if (unlikely(ret != 0))
 		goto err_unlock; /* init frees the buffer on failure */
 
@@ -439,54 +439,22 @@ static int vmw_fb_compute_depth(struct fb_var_screeninfo *var,
 static int vmwgfx_set_config_internal(struct drm_mode_set *set)
 {
 	struct drm_crtc *crtc = set->crtc;
-	struct drm_framebuffer *fb;
-	struct drm_crtc *tmp;
-	struct drm_modeset_acquire_ctx *ctx;
-	struct drm_device *dev = set->crtc->dev;
+	struct drm_modeset_acquire_ctx ctx;
 	int ret;
 
-	ctx = dev->mode_config.acquire_ctx;
+	drm_modeset_acquire_init(&ctx, 0);
 
 restart:
-	/*
-	 * NOTE: ->set_config can also disable other crtcs (if we steal all
-	 * connectors from it), hence we need to refcount the fbs across all
-	 * crtcs. Atomic modeset will have saner semantics ...
-	 */
-	drm_for_each_crtc(tmp, dev)
-		tmp->primary->old_fb = tmp->primary->fb;
-
-	fb = set->fb;
-
-	ret = crtc->funcs->set_config(set, ctx);
-	if (ret == 0) {
-		crtc->primary->crtc = crtc;
-		crtc->primary->fb = fb;
-	}
-
-	drm_for_each_crtc(tmp, dev) {
-		if (tmp->primary->fb)
-			drm_framebuffer_get(tmp->primary->fb);
-		if (tmp->primary->old_fb)
-			drm_framebuffer_put(tmp->primary->old_fb);
-		tmp->primary->old_fb = NULL;
-	}
+	ret = crtc->funcs->set_config(set, &ctx);
 
 	if (ret == -EDEADLK) {
-		dev->mode_config.acquire_ctx = NULL;
-
-retry_locking:
-		drm_modeset_backoff(ctx);
-
-		ret = drm_modeset_lock_all_ctx(dev, ctx);
-		if (ret)
-			goto retry_locking;
-
-		dev->mode_config.acquire_ctx = ctx;
-
+		drm_modeset_backoff(&ctx);
 		goto restart;
 	}
 
+	drm_modeset_drop_locks(&ctx);
+	drm_modeset_acquire_fini(&ctx);
+
 	return ret;
 }
 
@@ -523,7 +491,7 @@ static int vmw_fb_kms_detach(struct vmw_fb_par *par,
 	}
 
 	if (par->vmw_bo && detach_bo && unref_bo)
-		vmw_dmabuf_unreference(&par->vmw_bo);
+		vmw_bo_unreference(&par->vmw_bo);
 
 	return 0;
 }
@@ -624,7 +592,6 @@ static int vmw_fb_set_par(struct fb_info *info)
 	}
 
 	mutex_lock(&par->bo_mutex);
-	drm_modeset_lock_all(vmw_priv->dev);
 	ret = vmw_fb_kms_framebuffer(info);
 	if (ret)
 		goto out_unlock;
@@ -657,7 +624,6 @@ out_unlock:
 		drm_mode_destroy(vmw_priv->dev, old_mode);
 	par->set_mode = mode;
 
-	drm_modeset_unlock_all(vmw_priv->dev);
 	mutex_unlock(&par->bo_mutex);
 
 	return ret;
@@ -713,18 +679,14 @@ int vmw_fb_init(struct vmw_private *vmw_priv)
 	par->max_width = fb_width;
 	par->max_height = fb_height;
 
-	drm_modeset_lock_all(vmw_priv->dev);
 	ret = vmw_kms_fbdev_init_data(vmw_priv, 0, par->max_width,
 				      par->max_height, &par->con,
 				      &par->crtc, &init_mode);
-	if (ret) {
-		drm_modeset_unlock_all(vmw_priv->dev);
+	if (ret)
 		goto err_kms;
-	}
 
 	info->var.xres = init_mode->hdisplay;
 	info->var.yres = init_mode->vdisplay;
-	drm_modeset_unlock_all(vmw_priv->dev);
 
 	/*
 	 * Create buffers and alloc memory
@@ -832,7 +794,9 @@ int vmw_fb_close(struct vmw_private *vmw_priv)
 	cancel_delayed_work_sync(&par->local_work);
 	unregister_framebuffer(info);
 
+	mutex_lock(&par->bo_mutex);
 	(void) vmw_fb_kms_detach(par, true, true);
+	mutex_unlock(&par->bo_mutex);
 
 	vfree(par->vmalloc);
 	framebuffer_release(info);
@@ -877,21 +841,13 @@ int vmw_fb_on(struct vmw_private *vmw_priv)
 	spin_lock_irqsave(&par->dirty.lock, flags);
 	par->dirty.active = true;
 	spin_unlock_irqrestore(&par->dirty.lock, flags);
- 
-	return 0;
-}
 
-/**
- * vmw_fb_refresh - Refresh fb display
- *
- * @vmw_priv: Pointer to device private
- *
- * Call into kms to show the fbdev display(s).
- */
-void vmw_fb_refresh(struct vmw_private *vmw_priv)
-{
-	if (!vmw_priv->fb_info)
-		return;
+	/*
+	 * Need to reschedule a dirty update, because otherwise that's
+	 * only done in dirty_mark() if the previous coalesced
+	 * dirty region was empty.
+	 */
+	schedule_delayed_work(&par->local_work, 0);
 
-	vmw_fb_set_par(vmw_priv->fb_info);
+	return 0;
 }
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
index 9ed544f8958f..3d546d409334 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2011-2014 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2011-2014 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -175,7 +175,6 @@ static long vmw_fence_wait(struct dma_fence *f, bool intr, signed long timeout)
 	struct vmw_private *dev_priv = fman->dev_priv;
 	struct vmwgfx_wait_cb cb;
 	long ret = timeout;
-	unsigned long irq_flags;
 
 	if (likely(vmw_fence_obj_signaled(fence)))
 		return timeout;
@@ -183,7 +182,7 @@ static long vmw_fence_wait(struct dma_fence *f, bool intr, signed long timeout)
 	vmw_fifo_ping_host(dev_priv, SVGA_SYNC_GENERIC);
 	vmw_seqno_waiter_add(dev_priv);
 
-	spin_lock_irqsave(f->lock, irq_flags);
+	spin_lock(f->lock);
 
 	if (intr && signal_pending(current)) {
 		ret = -ERESTARTSYS;
@@ -194,30 +193,45 @@ static long vmw_fence_wait(struct dma_fence *f, bool intr, signed long timeout)
 	cb.task = current;
 	list_add(&cb.base.node, &f->cb_list);
 
-	while (ret > 0) {
+	for (;;) {
 		__vmw_fences_update(fman);
-		if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &f->flags))
-			break;
 
+		/*
+		 * We can use the barrier free __set_current_state() since
+		 * DMA_FENCE_FLAG_SIGNALED_BIT + wakeup is protected by the
+		 * fence spinlock.
+		 */
 		if (intr)
 			__set_current_state(TASK_INTERRUPTIBLE);
 		else
 			__set_current_state(TASK_UNINTERRUPTIBLE);
-		spin_unlock_irqrestore(f->lock, irq_flags);
 
-		ret = schedule_timeout(ret);
+		if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &f->flags)) {
+			if (ret == 0 && timeout > 0)
+				ret = 1;
+			break;
+		}
 
-		spin_lock_irqsave(f->lock, irq_flags);
-		if (ret > 0 && intr && signal_pending(current))
+		if (intr && signal_pending(current)) {
 			ret = -ERESTARTSYS;
-	}
+			break;
+		}
 
+		if (ret == 0)
+			break;
+
+		spin_unlock(f->lock);
+
+		ret = schedule_timeout(ret);
+
+		spin_lock(f->lock);
+	}
+	__set_current_state(TASK_RUNNING);
 	if (!list_empty(&cb.base.node))
 		list_del(&cb.base.node);
-	__set_current_state(TASK_RUNNING);
 
 out:
-	spin_unlock_irqrestore(f->lock, irq_flags);
+	spin_unlock(f->lock);
 
 	vmw_seqno_waiter_remove(dev_priv);
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.h b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.h
index 20224dba9d8e..c9382933c2b9 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fence.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fence.h
@@ -1,7 +1,7 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
- * Copyright © 2011-2012 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2011-2012 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c b/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
index a1c68e6a689e..d0fd147ef75f 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c b/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c
index 66ffa1d4759c..007a0cc7f232 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_gmr.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c b/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c
index f2f9d88131f2..ddb1e9365a3e 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_gmrid_manager.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright (c) 2007-2010 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2007-2010 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c
index c5e8eae0dbe2..172a6ba6539c 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ioctl.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -56,6 +56,9 @@ int vmw_getparam_ioctl(struct drm_device *dev, void *data,
 	case DRM_VMW_PARAM_HW_CAPS:
 		param->value = dev_priv->capabilities;
 		break;
+	case DRM_VMW_PARAM_HW_CAPS2:
+		param->value = dev_priv->capabilities2;
+		break;
 	case DRM_VMW_PARAM_FIFO_CAPS:
 		param->value = dev_priv->fifo.capabilities;
 		break;
@@ -113,6 +116,9 @@ int vmw_getparam_ioctl(struct drm_device *dev, void *data,
 	case DRM_VMW_PARAM_DX:
 		param->value = dev_priv->has_dx;
 		break;
+	case DRM_VMW_PARAM_SM4_1:
+		param->value = dev_priv->has_sm4_1;
+		break;
 	default:
 		return -EINVAL;
 	}
@@ -122,15 +128,12 @@ int vmw_getparam_ioctl(struct drm_device *dev, void *data,
 
 static u32 vmw_mask_multisample(unsigned int cap, u32 fmt_value)
 {
-	/* If the header is updated, update the format test as well! */
-	BUILD_BUG_ON(SVGA3D_DEVCAP_DXFMT_BC5_UNORM + 1 != SVGA3D_DEVCAP_MAX);
-
-	if (cap >= SVGA3D_DEVCAP_DXFMT_X8R8G8B8 &&
-	    cap <= SVGA3D_DEVCAP_DXFMT_BC5_UNORM)
-		fmt_value &= ~(SVGADX_DXFMT_MULTISAMPLE_2 |
-			       SVGADX_DXFMT_MULTISAMPLE_4 |
-			       SVGADX_DXFMT_MULTISAMPLE_8);
-	else if (cap == SVGA3D_DEVCAP_MULTISAMPLE_MASKABLESAMPLES)
+	/*
+	 * A version of user-space exists which use MULTISAMPLE_MASKABLESAMPLES
+	 * to check the sample count supported by virtual device. Since there
+	 * never was support for multisample count for backing MOB return 0.
+	 */
+	if (cap == SVGA3D_DEVCAP_MULTISAMPLE_MASKABLESAMPLES)
 		return 0;
 
 	return fmt_value;
@@ -377,8 +380,8 @@ int vmw_present_readback_ioctl(struct drm_device *dev, void *data,
 	}
 
 	vfb = vmw_framebuffer_to_vfb(fb);
-	if (!vfb->dmabuf) {
-		DRM_ERROR("Framebuffer not dmabuf backed.\n");
+	if (!vfb->bo) {
+		DRM_ERROR("Framebuffer not buffer backed.\n");
 		ret = -EINVAL;
 		goto out_no_ttm_lock;
 	}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c b/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c
index b9239ba067c4..c3ad4478266b 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_irq.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
index 6728c6247b4b..23beff5d8e3c 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -85,10 +85,10 @@ static int vmw_cursor_update_image(struct vmw_private *dev_priv,
 	return 0;
 }
 
-static int vmw_cursor_update_dmabuf(struct vmw_private *dev_priv,
-				    struct vmw_dma_buffer *dmabuf,
-				    u32 width, u32 height,
-				    u32 hotspotX, u32 hotspotY)
+static int vmw_cursor_update_bo(struct vmw_private *dev_priv,
+				struct vmw_buffer_object *bo,
+				u32 width, u32 height,
+				u32 hotspotX, u32 hotspotY)
 {
 	struct ttm_bo_kmap_obj map;
 	unsigned long kmap_offset;
@@ -100,13 +100,13 @@ static int vmw_cursor_update_dmabuf(struct vmw_private *dev_priv,
 	kmap_offset = 0;
 	kmap_num = (width*height*4 + PAGE_SIZE - 1) >> PAGE_SHIFT;
 
-	ret = ttm_bo_reserve(&dmabuf->base, true, false, NULL);
+	ret = ttm_bo_reserve(&bo->base, true, false, NULL);
 	if (unlikely(ret != 0)) {
 		DRM_ERROR("reserve failed\n");
 		return -EINVAL;
 	}
 
-	ret = ttm_bo_kmap(&dmabuf->base, kmap_offset, kmap_num, &map);
+	ret = ttm_bo_kmap(&bo->base, kmap_offset, kmap_num, &map);
 	if (unlikely(ret != 0))
 		goto err_unreserve;
 
@@ -116,7 +116,7 @@ static int vmw_cursor_update_dmabuf(struct vmw_private *dev_priv,
 
 	ttm_bo_kunmap(&map);
 err_unreserve:
-	ttm_bo_unreserve(&dmabuf->base);
+	ttm_bo_unreserve(&bo->base);
 
 	return ret;
 }
@@ -352,13 +352,13 @@ vmw_du_cursor_plane_prepare_fb(struct drm_plane *plane,
 	if (vps->surf)
 		vmw_surface_unreference(&vps->surf);
 
-	if (vps->dmabuf)
-		vmw_dmabuf_unreference(&vps->dmabuf);
+	if (vps->bo)
+		vmw_bo_unreference(&vps->bo);
 
 	if (fb) {
-		if (vmw_framebuffer_to_vfb(fb)->dmabuf) {
-			vps->dmabuf = vmw_framebuffer_to_vfbd(fb)->buffer;
-			vmw_dmabuf_reference(vps->dmabuf);
+		if (vmw_framebuffer_to_vfb(fb)->bo) {
+			vps->bo = vmw_framebuffer_to_vfbd(fb)->buffer;
+			vmw_bo_reference(vps->bo);
 		} else {
 			vps->surf = vmw_framebuffer_to_vfbs(fb)->surface;
 			vmw_surface_reference(vps->surf);
@@ -390,7 +390,7 @@ vmw_du_cursor_plane_atomic_update(struct drm_plane *plane,
 	}
 
 	du->cursor_surface = vps->surf;
-	du->cursor_dmabuf = vps->dmabuf;
+	du->cursor_bo = vps->bo;
 
 	if (vps->surf) {
 		du->cursor_age = du->cursor_surface->snooper.age;
@@ -399,11 +399,11 @@ vmw_du_cursor_plane_atomic_update(struct drm_plane *plane,
 					      vps->surf->snooper.image,
 					      64, 64, hotspot_x,
 					      hotspot_y);
-	} else if (vps->dmabuf) {
-		ret = vmw_cursor_update_dmabuf(dev_priv, vps->dmabuf,
-					       plane->state->crtc_w,
-					       plane->state->crtc_h,
-					       hotspot_x, hotspot_y);
+	} else if (vps->bo) {
+		ret = vmw_cursor_update_bo(dev_priv, vps->bo,
+					   plane->state->crtc_w,
+					   plane->state->crtc_h,
+					   hotspot_x, hotspot_y);
 	} else {
 		vmw_cursor_update_position(dev_priv, false, 0, 0);
 		return;
@@ -519,7 +519,7 @@ int vmw_du_cursor_plane_atomic_check(struct drm_plane *plane,
 		ret = -EINVAL;
 	}
 
-	if (!vmw_framebuffer_to_vfb(fb)->dmabuf)
+	if (!vmw_framebuffer_to_vfb(fb)->bo)
 		surface = vmw_framebuffer_to_vfbs(fb)->surface;
 
 	if (surface && !surface->snooper.image) {
@@ -535,9 +535,9 @@ int vmw_du_crtc_atomic_check(struct drm_crtc *crtc,
 			     struct drm_crtc_state *new_state)
 {
 	struct vmw_display_unit *du = vmw_crtc_to_du(new_state->crtc);
-	int connector_mask = 1 << drm_connector_index(&du->connector);
+	int connector_mask = drm_connector_mask(&du->connector);
 	bool has_primary = new_state->plane_mask &
-			   BIT(drm_plane_index(crtc->primary));
+			   drm_plane_mask(crtc->primary);
 
 	/* We always want to have an active plane with an active CRTC */
 	if (has_primary != new_state->enable)
@@ -687,8 +687,8 @@ vmw_du_plane_duplicate_state(struct drm_plane *plane)
 	if (vps->surf)
 		(void) vmw_surface_reference(vps->surf);
 
-	if (vps->dmabuf)
-		(void) vmw_dmabuf_reference(vps->dmabuf);
+	if (vps->bo)
+		(void) vmw_bo_reference(vps->bo);
 
 	state = &vps->base;
 
@@ -745,8 +745,8 @@ vmw_du_plane_destroy_state(struct drm_plane *plane,
 	if (vps->surf)
 		vmw_surface_unreference(&vps->surf);
 
-	if (vps->dmabuf)
-		vmw_dmabuf_unreference(&vps->dmabuf);
+	if (vps->bo)
+		vmw_bo_unreference(&vps->bo);
 
 	drm_atomic_helper_plane_destroy_state(plane, state);
 }
@@ -902,12 +902,12 @@ static int vmw_framebuffer_surface_dirty(struct drm_framebuffer *framebuffer,
 
 /**
  * vmw_kms_readback - Perform a readback from the screen system to
- * a dma-buffer backed framebuffer.
+ * a buffer-object backed framebuffer.
  *
  * @dev_priv: Pointer to the device private structure.
  * @file_priv: Pointer to a struct drm_file identifying the caller.
  * Must be set to NULL if @user_fence_rep is NULL.
- * @vfb: Pointer to the dma-buffer backed framebuffer.
+ * @vfb: Pointer to the buffer-object backed framebuffer.
  * @user_fence_rep: User-space provided structure for fence information.
  * Must be set to non-NULL if @file_priv is non-NULL.
  * @vclips: Array of clip rects.
@@ -951,7 +951,7 @@ static int vmw_kms_new_framebuffer_surface(struct vmw_private *dev_priv,
 					   struct vmw_framebuffer **out,
 					   const struct drm_mode_fb_cmd2
 					   *mode_cmd,
-					   bool is_dmabuf_proxy)
+					   bool is_bo_proxy)
 
 {
 	struct drm_device *dev = dev_priv->dev;
@@ -1019,7 +1019,7 @@ static int vmw_kms_new_framebuffer_surface(struct vmw_private *dev_priv,
 	drm_helper_mode_fill_fb_struct(dev, &vfbs->base.base, mode_cmd);
 	vfbs->surface = vmw_surface_reference(surface);
 	vfbs->base.user_handle = mode_cmd->handles[0];
-	vfbs->is_dmabuf_proxy = is_dmabuf_proxy;
+	vfbs->is_bo_proxy = is_bo_proxy;
 
 	*out = &vfbs->base;
 
@@ -1038,30 +1038,30 @@ out_err1:
 }
 
 /*
- * Dmabuf framebuffer code
+ * Buffer-object framebuffer code
  */
 
-static void vmw_framebuffer_dmabuf_destroy(struct drm_framebuffer *framebuffer)
+static void vmw_framebuffer_bo_destroy(struct drm_framebuffer *framebuffer)
 {
-	struct vmw_framebuffer_dmabuf *vfbd =
+	struct vmw_framebuffer_bo *vfbd =
 		vmw_framebuffer_to_vfbd(framebuffer);
 
 	drm_framebuffer_cleanup(framebuffer);
-	vmw_dmabuf_unreference(&vfbd->buffer);
+	vmw_bo_unreference(&vfbd->buffer);
 	if (vfbd->base.user_obj)
 		ttm_base_object_unref(&vfbd->base.user_obj);
 
 	kfree(vfbd);
 }
 
-static int vmw_framebuffer_dmabuf_dirty(struct drm_framebuffer *framebuffer,
-				 struct drm_file *file_priv,
-				 unsigned flags, unsigned color,
-				 struct drm_clip_rect *clips,
-				 unsigned num_clips)
+static int vmw_framebuffer_bo_dirty(struct drm_framebuffer *framebuffer,
+				    struct drm_file *file_priv,
+				    unsigned int flags, unsigned int color,
+				    struct drm_clip_rect *clips,
+				    unsigned int num_clips)
 {
 	struct vmw_private *dev_priv = vmw_priv(framebuffer->dev);
-	struct vmw_framebuffer_dmabuf *vfbd =
+	struct vmw_framebuffer_bo *vfbd =
 		vmw_framebuffer_to_vfbd(framebuffer);
 	struct drm_clip_rect norect;
 	int ret, increment = 1;
@@ -1092,13 +1092,13 @@ static int vmw_framebuffer_dmabuf_dirty(struct drm_framebuffer *framebuffer,
 				       true, true, NULL);
 		break;
 	case vmw_du_screen_object:
-		ret = vmw_kms_sou_do_dmabuf_dirty(dev_priv, &vfbd->base,
-						  clips, NULL, num_clips,
-						  increment, true, NULL, NULL);
+		ret = vmw_kms_sou_do_bo_dirty(dev_priv, &vfbd->base,
+					      clips, NULL, num_clips,
+					      increment, true, NULL, NULL);
 		break;
 	case vmw_du_legacy:
-		ret = vmw_kms_ldu_do_dmabuf_dirty(dev_priv, &vfbd->base, 0, 0,
-						  clips, num_clips, increment);
+		ret = vmw_kms_ldu_do_bo_dirty(dev_priv, &vfbd->base, 0, 0,
+					      clips, num_clips, increment);
 		break;
 	default:
 		ret = -EINVAL;
@@ -1114,23 +1114,23 @@ static int vmw_framebuffer_dmabuf_dirty(struct drm_framebuffer *framebuffer,
 	return ret;
 }
 
-static const struct drm_framebuffer_funcs vmw_framebuffer_dmabuf_funcs = {
-	.destroy = vmw_framebuffer_dmabuf_destroy,
-	.dirty = vmw_framebuffer_dmabuf_dirty,
+static const struct drm_framebuffer_funcs vmw_framebuffer_bo_funcs = {
+	.destroy = vmw_framebuffer_bo_destroy,
+	.dirty = vmw_framebuffer_bo_dirty,
 };
 
 /**
- * Pin the dmabuffer in a location suitable for access by the
+ * Pin the bofer in a location suitable for access by the
  * display system.
  */
 static int vmw_framebuffer_pin(struct vmw_framebuffer *vfb)
 {
 	struct vmw_private *dev_priv = vmw_priv(vfb->base.dev);
-	struct vmw_dma_buffer *buf;
+	struct vmw_buffer_object *buf;
 	struct ttm_placement *placement;
 	int ret;
 
-	buf = vfb->dmabuf ?  vmw_framebuffer_to_vfbd(&vfb->base)->buffer :
+	buf = vfb->bo ?  vmw_framebuffer_to_vfbd(&vfb->base)->buffer :
 		vmw_framebuffer_to_vfbs(&vfb->base)->surface->res.backup;
 
 	if (!buf)
@@ -1139,12 +1139,12 @@ static int vmw_framebuffer_pin(struct vmw_framebuffer *vfb)
 	switch (dev_priv->active_display_unit) {
 	case vmw_du_legacy:
 		vmw_overlay_pause_all(dev_priv);
-		ret = vmw_dmabuf_pin_in_start_of_vram(dev_priv, buf, false);
+		ret = vmw_bo_pin_in_start_of_vram(dev_priv, buf, false);
 		vmw_overlay_resume_all(dev_priv);
 		break;
 	case vmw_du_screen_object:
 	case vmw_du_screen_target:
-		if (vfb->dmabuf) {
+		if (vfb->bo) {
 			if (dev_priv->capabilities & SVGA_CAP_3D) {
 				/*
 				 * Use surface DMA to get content to
@@ -1160,8 +1160,7 @@ static int vmw_framebuffer_pin(struct vmw_framebuffer *vfb)
 			placement = &vmw_mob_placement;
 		}
 
-		return vmw_dmabuf_pin_in_placement(dev_priv, buf, placement,
-						   false);
+		return vmw_bo_pin_in_placement(dev_priv, buf, placement, false);
 	default:
 		return -EINVAL;
 	}
@@ -1172,36 +1171,36 @@ static int vmw_framebuffer_pin(struct vmw_framebuffer *vfb)
 static int vmw_framebuffer_unpin(struct vmw_framebuffer *vfb)
 {
 	struct vmw_private *dev_priv = vmw_priv(vfb->base.dev);
-	struct vmw_dma_buffer *buf;
+	struct vmw_buffer_object *buf;
 
-	buf = vfb->dmabuf ?  vmw_framebuffer_to_vfbd(&vfb->base)->buffer :
+	buf = vfb->bo ?  vmw_framebuffer_to_vfbd(&vfb->base)->buffer :
 		vmw_framebuffer_to_vfbs(&vfb->base)->surface->res.backup;
 
 	if (WARN_ON(!buf))
 		return 0;
 
-	return vmw_dmabuf_unpin(dev_priv, buf, false);
+	return vmw_bo_unpin(dev_priv, buf, false);
 }
 
 /**
- * vmw_create_dmabuf_proxy - create a proxy surface for the DMA buf
+ * vmw_create_bo_proxy - create a proxy surface for the buffer object
  *
  * @dev: DRM device
  * @mode_cmd: parameters for the new surface
- * @dmabuf_mob: MOB backing the DMA buf
+ * @bo_mob: MOB backing the buffer object
  * @srf_out: newly created surface
  *
- * When the content FB is a DMA buf, we create a surface as a proxy to the
+ * When the content FB is a buffer object, we create a surface as a proxy to the
  * same buffer.  This way we can do a surface copy rather than a surface DMA.
  * This is a more efficient approach
  *
  * RETURNS:
  * 0 on success, error code otherwise
  */
-static int vmw_create_dmabuf_proxy(struct drm_device *dev,
-				   const struct drm_mode_fb_cmd2 *mode_cmd,
-				   struct vmw_dma_buffer *dmabuf_mob,
-				   struct vmw_surface **srf_out)
+static int vmw_create_bo_proxy(struct drm_device *dev,
+			       const struct drm_mode_fb_cmd2 *mode_cmd,
+			       struct vmw_buffer_object *bo_mob,
+			       struct vmw_surface **srf_out)
 {
 	uint32_t format;
 	struct drm_vmw_size content_base_size = {0};
@@ -1239,15 +1238,17 @@ static int vmw_create_dmabuf_proxy(struct drm_device *dev,
 	content_base_size.depth  = 1;
 
 	ret = vmw_surface_gb_priv_define(dev,
-			0, /* kernel visible only */
-			0, /* flags */
-			format,
-			true, /* can be a scanout buffer */
-			1, /* num of mip levels */
-			0,
-			0,
-			content_base_size,
-			srf_out);
+					 0, /* kernel visible only */
+					 0, /* flags */
+					 format,
+					 true, /* can be a scanout buffer */
+					 1, /* num of mip levels */
+					 0,
+					 0,
+					 content_base_size,
+					 SVGA3D_MS_PATTERN_NONE,
+					 SVGA3D_MS_QUALITY_NONE,
+					 srf_out);
 	if (ret) {
 		DRM_ERROR("Failed to allocate proxy content buffer\n");
 		return ret;
@@ -1258,8 +1259,8 @@ static int vmw_create_dmabuf_proxy(struct drm_device *dev,
 	/* Reserve and switch the backing mob. */
 	mutex_lock(&res->dev_priv->cmdbuf_mutex);
 	(void) vmw_resource_reserve(res, false, true);
-	vmw_dmabuf_unreference(&res->backup);
-	res->backup = vmw_dmabuf_reference(dmabuf_mob);
+	vmw_bo_unreference(&res->backup);
+	res->backup = vmw_bo_reference(bo_mob);
 	res->backup_offset = 0;
 	vmw_resource_unreserve(res, false, NULL, 0);
 	mutex_unlock(&res->dev_priv->cmdbuf_mutex);
@@ -1269,21 +1270,21 @@ static int vmw_create_dmabuf_proxy(struct drm_device *dev,
 
 
 
-static int vmw_kms_new_framebuffer_dmabuf(struct vmw_private *dev_priv,
-					  struct vmw_dma_buffer *dmabuf,
-					  struct vmw_framebuffer **out,
-					  const struct drm_mode_fb_cmd2
-					  *mode_cmd)
+static int vmw_kms_new_framebuffer_bo(struct vmw_private *dev_priv,
+				      struct vmw_buffer_object *bo,
+				      struct vmw_framebuffer **out,
+				      const struct drm_mode_fb_cmd2
+				      *mode_cmd)
 
 {
 	struct drm_device *dev = dev_priv->dev;
-	struct vmw_framebuffer_dmabuf *vfbd;
+	struct vmw_framebuffer_bo *vfbd;
 	unsigned int requested_size;
 	struct drm_format_name_buf format_name;
 	int ret;
 
 	requested_size = mode_cmd->height * mode_cmd->pitches[0];
-	if (unlikely(requested_size > dmabuf->base.num_pages * PAGE_SIZE)) {
+	if (unlikely(requested_size > bo->base.num_pages * PAGE_SIZE)) {
 		DRM_ERROR("Screen buffer object size is too small "
 			  "for requested mode.\n");
 		return -EINVAL;
@@ -1312,20 +1313,20 @@ static int vmw_kms_new_framebuffer_dmabuf(struct vmw_private *dev_priv,
 	}
 
 	drm_helper_mode_fill_fb_struct(dev, &vfbd->base.base, mode_cmd);
-	vfbd->base.dmabuf = true;
-	vfbd->buffer = vmw_dmabuf_reference(dmabuf);
+	vfbd->base.bo = true;
+	vfbd->buffer = vmw_bo_reference(bo);
 	vfbd->base.user_handle = mode_cmd->handles[0];
 	*out = &vfbd->base;
 
 	ret = drm_framebuffer_init(dev, &vfbd->base.base,
-				   &vmw_framebuffer_dmabuf_funcs);
+				   &vmw_framebuffer_bo_funcs);
 	if (ret)
 		goto out_err2;
 
 	return 0;
 
 out_err2:
-	vmw_dmabuf_unreference(&dmabuf);
+	vmw_bo_unreference(&bo);
 	kfree(vfbd);
 out_err1:
 	return ret;
@@ -1354,57 +1355,57 @@ vmw_kms_srf_ok(struct vmw_private *dev_priv, uint32_t width, uint32_t height)
  * vmw_kms_new_framebuffer - Create a new framebuffer.
  *
  * @dev_priv: Pointer to device private struct.
- * @dmabuf: Pointer to dma buffer to wrap the kms framebuffer around.
- * Either @dmabuf or @surface must be NULL.
+ * @bo: Pointer to buffer object to wrap the kms framebuffer around.
+ * Either @bo or @surface must be NULL.
  * @surface: Pointer to a surface to wrap the kms framebuffer around.
- * Either @dmabuf or @surface must be NULL.
- * @only_2d: No presents will occur to this dma buffer based framebuffer. This
- * Helps the code to do some important optimizations.
+ * Either @bo or @surface must be NULL.
+ * @only_2d: No presents will occur to this buffer object based framebuffer.
+ * This helps the code to do some important optimizations.
  * @mode_cmd: Frame-buffer metadata.
  */
 struct vmw_framebuffer *
 vmw_kms_new_framebuffer(struct vmw_private *dev_priv,
-			struct vmw_dma_buffer *dmabuf,
+			struct vmw_buffer_object *bo,
 			struct vmw_surface *surface,
 			bool only_2d,
 			const struct drm_mode_fb_cmd2 *mode_cmd)
 {
 	struct vmw_framebuffer *vfb = NULL;
-	bool is_dmabuf_proxy = false;
+	bool is_bo_proxy = false;
 	int ret;
 
 	/*
 	 * We cannot use the SurfaceDMA command in an non-accelerated VM,
-	 * therefore, wrap the DMA buf in a surface so we can use the
+	 * therefore, wrap the buffer object in a surface so we can use the
 	 * SurfaceCopy command.
 	 */
 	if (vmw_kms_srf_ok(dev_priv, mode_cmd->width, mode_cmd->height)  &&
-	    dmabuf && only_2d &&
+	    bo && only_2d &&
 	    mode_cmd->width > 64 &&  /* Don't create a proxy for cursor */
 	    dev_priv->active_display_unit == vmw_du_screen_target) {
-		ret = vmw_create_dmabuf_proxy(dev_priv->dev, mode_cmd,
-					      dmabuf, &surface);
+		ret = vmw_create_bo_proxy(dev_priv->dev, mode_cmd,
+					  bo, &surface);
 		if (ret)
 			return ERR_PTR(ret);
 
-		is_dmabuf_proxy = true;
+		is_bo_proxy = true;
 	}
 
 	/* Create the new framebuffer depending one what we have */
 	if (surface) {
 		ret = vmw_kms_new_framebuffer_surface(dev_priv, surface, &vfb,
 						      mode_cmd,
-						      is_dmabuf_proxy);
+						      is_bo_proxy);
 
 		/*
-		 * vmw_create_dmabuf_proxy() adds a reference that is no longer
+		 * vmw_create_bo_proxy() adds a reference that is no longer
 		 * needed
 		 */
-		if (is_dmabuf_proxy)
+		if (is_bo_proxy)
 			vmw_surface_unreference(&surface);
-	} else if (dmabuf) {
-		ret = vmw_kms_new_framebuffer_dmabuf(dev_priv, dmabuf, &vfb,
-						     mode_cmd);
+	} else if (bo) {
+		ret = vmw_kms_new_framebuffer_bo(dev_priv, bo, &vfb,
+						 mode_cmd);
 	} else {
 		BUG();
 	}
@@ -1430,23 +1431,10 @@ static struct drm_framebuffer *vmw_kms_fb_create(struct drm_device *dev,
 	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
 	struct vmw_framebuffer *vfb = NULL;
 	struct vmw_surface *surface = NULL;
-	struct vmw_dma_buffer *bo = NULL;
+	struct vmw_buffer_object *bo = NULL;
 	struct ttm_base_object *user_obj;
 	int ret;
 
-	/**
-	 * This code should be conditioned on Screen Objects not being used.
-	 * If screen objects are used, we can allocate a GMR to hold the
-	 * requested framebuffer.
-	 */
-
-	if (!vmw_kms_validate_mode_vram(dev_priv,
-					mode_cmd->pitches[0],
-					mode_cmd->height)) {
-		DRM_ERROR("Requested mode exceed bounding box limit.\n");
-		return ERR_PTR(-ENOMEM);
-	}
-
 	/*
 	 * Take a reference on the user object of the resource
 	 * backing the kms fb. This ensures that user-space handle
@@ -1466,7 +1454,7 @@ static struct drm_framebuffer *vmw_kms_fb_create(struct drm_device *dev,
 	 * End conditioned code.
 	 */
 
-	/* returns either a dmabuf or surface */
+	/* returns either a bo or surface */
 	ret = vmw_user_lookup_handle(dev_priv, tfile,
 				     mode_cmd->handles[0],
 				     &surface, &bo);
@@ -1494,7 +1482,7 @@ static struct drm_framebuffer *vmw_kms_fb_create(struct drm_device *dev,
 err_out:
 	/* vmw_user_lookup_handle takes one ref so does new_fb */
 	if (bo)
-		vmw_dmabuf_unreference(&bo);
+		vmw_bo_unreference(&bo);
 	if (surface)
 		vmw_surface_unreference(&surface);
 
@@ -1508,7 +1496,168 @@ err_out:
 	return &vfb->base;
 }
 
+/**
+ * vmw_kms_check_display_memory - Validates display memory required for a
+ * topology
+ * @dev: DRM device
+ * @num_rects: number of drm_rect in rects
+ * @rects: array of drm_rect representing the topology to validate indexed by
+ * crtc index.
+ *
+ * Returns:
+ * 0 on success otherwise negative error code
+ */
+static int vmw_kms_check_display_memory(struct drm_device *dev,
+					uint32_t num_rects,
+					struct drm_rect *rects)
+{
+	struct vmw_private *dev_priv = vmw_priv(dev);
+	struct drm_mode_config *mode_config = &dev->mode_config;
+	struct drm_rect bounding_box = {0};
+	u64 total_pixels = 0, pixel_mem, bb_mem;
+	int i;
+
+	for (i = 0; i < num_rects; i++) {
+		/*
+		 * Currently this check is limiting the topology within max
+		 * texture/screentarget size. This should change in future when
+		 * user-space support multiple fb with topology.
+		 */
+		if (rects[i].x1 < 0 ||  rects[i].y1 < 0 ||
+		    rects[i].x2 > mode_config->max_width ||
+		    rects[i].y2 > mode_config->max_height) {
+			DRM_ERROR("Invalid GUI layout.\n");
+			return -EINVAL;
+		}
+
+		/* Bounding box upper left is at (0,0). */
+		if (rects[i].x2 > bounding_box.x2)
+			bounding_box.x2 = rects[i].x2;
 
+		if (rects[i].y2 > bounding_box.y2)
+			bounding_box.y2 = rects[i].y2;
+
+		total_pixels += (u64) drm_rect_width(&rects[i]) *
+			(u64) drm_rect_height(&rects[i]);
+	}
+
+	/* Virtual svga device primary limits are always in 32-bpp. */
+	pixel_mem = total_pixels * 4;
+
+	/*
+	 * For HV10 and below prim_bb_mem is vram size. When
+	 * SVGA_REG_MAX_PRIMARY_BOUNDING_BOX_MEM is not present vram size is
+	 * limit on primary bounding box
+	 */
+	if (pixel_mem > dev_priv->prim_bb_mem) {
+		DRM_ERROR("Combined output size too large.\n");
+		return -EINVAL;
+	}
+
+	/* SVGA_CAP_NO_BB_RESTRICTION is available for STDU only. */
+	if (dev_priv->active_display_unit != vmw_du_screen_target ||
+	    !(dev_priv->capabilities & SVGA_CAP_NO_BB_RESTRICTION)) {
+		bb_mem = (u64) bounding_box.x2 * bounding_box.y2 * 4;
+
+		if (bb_mem > dev_priv->prim_bb_mem) {
+			DRM_ERROR("Topology is beyond supported limits.\n");
+			return -EINVAL;
+		}
+	}
+
+	return 0;
+}
+
+/**
+ * vmw_kms_check_topology - Validates topology in drm_atomic_state
+ * @dev: DRM device
+ * @state: the driver state object
+ *
+ * Returns:
+ * 0 on success otherwise negative error code
+ */
+static int vmw_kms_check_topology(struct drm_device *dev,
+				  struct drm_atomic_state *state)
+{
+	struct vmw_private *dev_priv = vmw_priv(dev);
+	struct drm_crtc_state *old_crtc_state, *new_crtc_state;
+	struct drm_rect *rects;
+	struct drm_crtc *crtc;
+	uint32_t i;
+	int ret = 0;
+
+	rects = kcalloc(dev->mode_config.num_crtc, sizeof(struct drm_rect),
+			GFP_KERNEL);
+	if (!rects)
+		return -ENOMEM;
+
+	mutex_lock(&dev_priv->requested_layout_mutex);
+
+	drm_for_each_crtc(crtc, dev) {
+		struct vmw_display_unit *du = vmw_crtc_to_du(crtc);
+		struct drm_crtc_state *crtc_state = crtc->state;
+
+		i = drm_crtc_index(crtc);
+
+		if (crtc_state && crtc_state->enable) {
+			rects[i].x1 = du->gui_x;
+			rects[i].y1 = du->gui_y;
+			rects[i].x2 = du->gui_x + crtc_state->mode.hdisplay;
+			rects[i].y2 = du->gui_y + crtc_state->mode.vdisplay;
+		}
+	}
+
+	/* Determine change to topology due to new atomic state */
+	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state,
+				      new_crtc_state, i) {
+		struct vmw_display_unit *du = vmw_crtc_to_du(crtc);
+		struct drm_connector *connector;
+		struct drm_connector_state *conn_state;
+		struct vmw_connector_state *vmw_conn_state;
+
+		if (!new_crtc_state->enable && old_crtc_state->enable) {
+			rects[i].x1 = 0;
+			rects[i].y1 = 0;
+			rects[i].x2 = 0;
+			rects[i].y2 = 0;
+			continue;
+		}
+
+		if (!du->pref_active) {
+			ret = -EINVAL;
+			goto clean;
+		}
+
+		/*
+		 * For vmwgfx each crtc has only one connector attached and it
+		 * is not changed so don't really need to check the
+		 * crtc->connector_mask and iterate over it.
+		 */
+		connector = &du->connector;
+		conn_state = drm_atomic_get_connector_state(state, connector);
+		if (IS_ERR(conn_state)) {
+			ret = PTR_ERR(conn_state);
+			goto clean;
+		}
+
+		vmw_conn_state = vmw_connector_state_to_vcs(conn_state);
+		vmw_conn_state->gui_x = du->gui_x;
+		vmw_conn_state->gui_y = du->gui_y;
+
+		rects[i].x1 = du->gui_x;
+		rects[i].y1 = du->gui_y;
+		rects[i].x2 = du->gui_x + new_crtc_state->mode.hdisplay;
+		rects[i].y2 = du->gui_y + new_crtc_state->mode.vdisplay;
+	}
+
+	ret = vmw_kms_check_display_memory(dev, dev->mode_config.num_crtc,
+					   rects);
+
+clean:
+	mutex_unlock(&dev_priv->requested_layout_mutex);
+	kfree(rects);
+	return ret;
+}
 
 /**
  * vmw_kms_atomic_check_modeset- validate state object for modeset changes
@@ -1520,36 +1669,39 @@ err_out:
  * us to assign a value to mode->crtc_clock so that
  * drm_calc_timestamping_constants() won't throw an error message
  *
- * RETURNS
+ * Returns:
  * Zero for success or -errno
  */
 static int
 vmw_kms_atomic_check_modeset(struct drm_device *dev,
 			     struct drm_atomic_state *state)
 {
-	struct drm_crtc_state *crtc_state;
 	struct drm_crtc *crtc;
-	struct vmw_private *dev_priv = vmw_priv(dev);
-	int i;
-
-	for_each_new_crtc_in_state(state, crtc, crtc_state, i) {
-		unsigned long requested_bb_mem = 0;
+	struct drm_crtc_state *crtc_state;
+	bool need_modeset = false;
+	int i, ret;
 
-		if (dev_priv->active_display_unit == vmw_du_screen_target) {
-			if (crtc->primary->fb) {
-				int cpp = crtc->primary->fb->pitches[0] /
-					  crtc->primary->fb->width;
+	ret = drm_atomic_helper_check(dev, state);
+	if (ret)
+		return ret;
 
-				requested_bb_mem += crtc->mode.hdisplay * cpp *
-						    crtc->mode.vdisplay;
-			}
+	if (!state->allow_modeset)
+		return ret;
 
-			if (requested_bb_mem > dev_priv->prim_bb_mem)
-				return -EINVAL;
-		}
+	/*
+	 * Legacy path do not set allow_modeset properly like
+	 * @drm_atomic_helper_update_plane, This will result in unnecessary call
+	 * to vmw_kms_check_topology. So extra set of check.
+	 */
+	for_each_new_crtc_in_state(state, crtc, crtc_state, i) {
+		if (drm_atomic_crtc_needs_modeset(crtc_state))
+			need_modeset = true;
 	}
 
-	return drm_atomic_helper_check(dev, state);
+	if (need_modeset)
+		return vmw_kms_check_topology(dev, state);
+
+	return ret;
 }
 
 static const struct drm_mode_config_funcs vmw_kms_funcs = {
@@ -1841,40 +1993,49 @@ void vmw_disable_vblank(struct drm_device *dev, unsigned int pipe)
 {
 }
 
-
-/*
- * Small shared kms functions.
+/**
+ * vmw_du_update_layout - Update the display unit with topology from resolution
+ * plugin and generate DRM uevent
+ * @dev_priv: device private
+ * @num_rects: number of drm_rect in rects
+ * @rects: toplogy to update
  */
-
-static int vmw_du_update_layout(struct vmw_private *dev_priv, unsigned num,
-			 struct drm_vmw_rect *rects)
+static int vmw_du_update_layout(struct vmw_private *dev_priv,
+				unsigned int num_rects, struct drm_rect *rects)
 {
 	struct drm_device *dev = dev_priv->dev;
 	struct vmw_display_unit *du;
 	struct drm_connector *con;
+	struct drm_connector_list_iter conn_iter;
 
-	mutex_lock(&dev->mode_config.mutex);
-
-#if 0
-	{
-		unsigned int i;
-
-		DRM_INFO("%s: new layout ", __func__);
-		for (i = 0; i < num; i++)
-			DRM_INFO("(%i, %i %ux%u) ", rects[i].x, rects[i].y,
-				 rects[i].w, rects[i].h);
-		DRM_INFO("\n");
+	/*
+	 * Currently only gui_x/y is protected with requested_layout_mutex.
+	 */
+	mutex_lock(&dev_priv->requested_layout_mutex);
+	drm_connector_list_iter_begin(dev, &conn_iter);
+	drm_for_each_connector_iter(con, &conn_iter) {
+		du = vmw_connector_to_du(con);
+		if (num_rects > du->unit) {
+			du->pref_width = drm_rect_width(&rects[du->unit]);
+			du->pref_height = drm_rect_height(&rects[du->unit]);
+			du->pref_active = true;
+			du->gui_x = rects[du->unit].x1;
+			du->gui_y = rects[du->unit].y1;
+		} else {
+			du->pref_width = 800;
+			du->pref_height = 600;
+			du->pref_active = false;
+			du->gui_x = 0;
+			du->gui_y = 0;
+		}
 	}
-#endif
+	drm_connector_list_iter_end(&conn_iter);
+	mutex_unlock(&dev_priv->requested_layout_mutex);
 
+	mutex_lock(&dev->mode_config.mutex);
 	list_for_each_entry(con, &dev->mode_config.connector_list, head) {
 		du = vmw_connector_to_du(con);
-		if (num > du->unit) {
-			du->pref_width = rects[du->unit].w;
-			du->pref_height = rects[du->unit].h;
-			du->pref_active = true;
-			du->gui_x = rects[du->unit].x;
-			du->gui_y = rects[du->unit].y;
+		if (num_rects > du->unit) {
 			drm_object_property_set_value
 			  (&con->base, dev->mode_config.suggested_x_property,
 			   du->gui_x);
@@ -1882,9 +2043,6 @@ static int vmw_du_update_layout(struct vmw_private *dev_priv, unsigned num,
 			  (&con->base, dev->mode_config.suggested_y_property,
 			   du->gui_y);
 		} else {
-			du->pref_width = 800;
-			du->pref_height = 600;
-			du->pref_active = false;
 			drm_object_property_set_value
 			  (&con->base, dev->mode_config.suggested_x_property,
 			   0);
@@ -1894,8 +2052,8 @@ static int vmw_du_update_layout(struct vmw_private *dev_priv, unsigned num,
 		}
 		con->status = vmw_du_connector_detect(con, true);
 	}
-
 	mutex_unlock(&dev->mode_config.mutex);
+
 	drm_sysfs_hotplug_event(dev);
 
 	return 0;
@@ -2110,7 +2268,7 @@ int vmw_du_connector_fill_modes(struct drm_connector *connector,
 		drm_mode_probed_add(connector, mode);
 	}
 
-	drm_mode_connector_list_update(connector);
+	drm_connector_list_update(connector);
 	/* Move the prefered mode first, help apps pick the right mode. */
 	drm_mode_sort(&connector->modes);
 
@@ -2195,7 +2353,25 @@ vmw_du_connector_atomic_get_property(struct drm_connector *connector,
 	return 0;
 }
 
-
+/**
+ * vmw_kms_update_layout_ioctl - Handler for DRM_VMW_UPDATE_LAYOUT ioctl
+ * @dev: drm device for the ioctl
+ * @data: data pointer for the ioctl
+ * @file_priv: drm file for the ioctl call
+ *
+ * Update preferred topology of display unit as per ioctl request. The topology
+ * is expressed as array of drm_vmw_rect.
+ * e.g.
+ * [0 0 640 480] [640 0 800 600] [0 480 640 480]
+ *
+ * NOTE:
+ * The x and y offset (upper left) in drm_vmw_rect cannot be less than 0. Beside
+ * device limit on topology, x + w and y + h (lower right) cannot be greater
+ * than INT_MAX. So topology beyond these limits will return with error.
+ *
+ * Returns:
+ * Zero on success, negative errno on failure.
+ */
 int vmw_kms_update_layout_ioctl(struct drm_device *dev, void *data,
 				struct drm_file *file_priv)
 {
@@ -2204,15 +2380,12 @@ int vmw_kms_update_layout_ioctl(struct drm_device *dev, void *data,
 		(struct drm_vmw_update_layout_arg *)data;
 	void __user *user_rects;
 	struct drm_vmw_rect *rects;
+	struct drm_rect *drm_rects;
 	unsigned rects_size;
-	int ret;
-	int i;
-	u64 total_pixels = 0;
-	struct drm_mode_config *mode_config = &dev->mode_config;
-	struct drm_vmw_rect bounding_box = {0};
+	int ret, i;
 
 	if (!arg->num_outputs) {
-		struct drm_vmw_rect def_rect = {0, 0, 800, 600};
+		struct drm_rect def_rect = {0, 0, 800, 600};
 		vmw_du_update_layout(dev_priv, 1, &def_rect);
 		return 0;
 	}
@@ -2231,52 +2404,29 @@ int vmw_kms_update_layout_ioctl(struct drm_device *dev, void *data,
 		goto out_free;
 	}
 
-	for (i = 0; i < arg->num_outputs; ++i) {
-		if (rects[i].x < 0 ||
-		    rects[i].y < 0 ||
-		    rects[i].x + rects[i].w > mode_config->max_width ||
-		    rects[i].y + rects[i].h > mode_config->max_height) {
-			DRM_ERROR("Invalid GUI layout.\n");
-			ret = -EINVAL;
-			goto out_free;
-		}
-
-		/*
-		 * bounding_box.w and bunding_box.h are used as
-		 * lower-right coordinates
-		 */
-		if (rects[i].x + rects[i].w > bounding_box.w)
-			bounding_box.w = rects[i].x + rects[i].w;
+	drm_rects = (struct drm_rect *)rects;
 
-		if (rects[i].y + rects[i].h > bounding_box.h)
-			bounding_box.h = rects[i].y + rects[i].h;
+	for (i = 0; i < arg->num_outputs; i++) {
+		struct drm_vmw_rect curr_rect;
 
-		total_pixels += (u64) rects[i].w * (u64) rects[i].h;
-	}
-
-	if (dev_priv->active_display_unit == vmw_du_screen_target) {
-		/*
-		 * For Screen Targets, the limits for a toplogy are:
-		 *	1. Bounding box (assuming 32bpp) must be < prim_bb_mem
-		 *      2. Total pixels (assuming 32bpp) must be < prim_bb_mem
-		 */
-		u64 bb_mem    = (u64) bounding_box.w * bounding_box.h * 4;
-		u64 pixel_mem = total_pixels * 4;
-
-		if (bb_mem > dev_priv->prim_bb_mem) {
-			DRM_ERROR("Topology is beyond supported limits.\n");
-			ret = -EINVAL;
+		/* Verify user-space for overflow as kernel use drm_rect */
+		if ((rects[i].x + rects[i].w > INT_MAX) ||
+		    (rects[i].y + rects[i].h > INT_MAX)) {
+			ret = -ERANGE;
 			goto out_free;
 		}
 
-		if (pixel_mem > dev_priv->prim_bb_mem) {
-			DRM_ERROR("Combined output size too large\n");
-			ret = -EINVAL;
-			goto out_free;
-		}
+		curr_rect = rects[i];
+		drm_rects[i].x1 = curr_rect.x;
+		drm_rects[i].y1 = curr_rect.y;
+		drm_rects[i].x2 = curr_rect.x + curr_rect.w;
+		drm_rects[i].y2 = curr_rect.y + curr_rect.h;
 	}
 
-	vmw_du_update_layout(dev_priv, arg->num_outputs, rects);
+	ret = vmw_kms_check_display_memory(dev, arg->num_outputs, drm_rects);
+
+	if (ret == 0)
+		vmw_du_update_layout(dev_priv, arg->num_outputs, drm_rects);
 
 out_free:
 	kfree(rects);
@@ -2322,9 +2472,10 @@ int vmw_kms_helper_dirty(struct vmw_private *dev_priv,
 	} else {
 		list_for_each_entry(crtc, &dev_priv->dev->mode_config.crtc_list,
 				    head) {
-			if (crtc->primary->fb != &framebuffer->base)
-				continue;
-			units[num_units++] = vmw_crtc_to_du(crtc);
+			struct drm_plane *plane = crtc->primary;
+
+			if (plane->state->fb == &framebuffer->base)
+				units[num_units++] = vmw_crtc_to_du(crtc);
 		}
 	}
 
@@ -2422,7 +2573,7 @@ int vmw_kms_helper_dirty(struct vmw_private *dev_priv,
  * interrupted by a signal.
  */
 int vmw_kms_helper_buffer_prepare(struct vmw_private *dev_priv,
-				  struct vmw_dma_buffer *buf,
+				  struct vmw_buffer_object *buf,
 				  bool interruptible,
 				  bool validate_as_mob,
 				  bool for_cpu_blit)
@@ -2454,7 +2605,7 @@ int vmw_kms_helper_buffer_prepare(struct vmw_private *dev_priv,
  * Helper to be used if an error forces the caller to undo the actions of
  * vmw_kms_helper_buffer_prepare.
  */
-void vmw_kms_helper_buffer_revert(struct vmw_dma_buffer *buf)
+void vmw_kms_helper_buffer_revert(struct vmw_buffer_object *buf)
 {
 	if (buf)
 		ttm_bo_unreserve(&buf->base);
@@ -2477,7 +2628,7 @@ void vmw_kms_helper_buffer_revert(struct vmw_dma_buffer *buf)
  */
 void vmw_kms_helper_buffer_finish(struct vmw_private *dev_priv,
 				  struct drm_file *file_priv,
-				  struct vmw_dma_buffer *buf,
+				  struct vmw_buffer_object *buf,
 				  struct vmw_fence_obj **out_fence,
 				  struct drm_vmw_fence_rep __user *
 				  user_fence_rep)
@@ -2489,7 +2640,7 @@ void vmw_kms_helper_buffer_finish(struct vmw_private *dev_priv,
 	ret = vmw_execbuf_fence_commands(file_priv, dev_priv, &fence,
 					 file_priv ? &handle : NULL);
 	if (buf)
-		vmw_fence_single_bo(&buf->base, fence);
+		vmw_bo_fence_single(&buf->base, fence);
 	if (file_priv)
 		vmw_execbuf_copy_fence_user(dev_priv, vmw_fpriv(file_priv),
 					    ret, user_fence_rep, fence,
@@ -2517,7 +2668,7 @@ void vmw_kms_helper_resource_revert(struct vmw_validation_ctx *ctx)
 	struct vmw_resource *res = ctx->res;
 
 	vmw_kms_helper_buffer_revert(ctx->buf);
-	vmw_dmabuf_unreference(&ctx->buf);
+	vmw_bo_unreference(&ctx->buf);
 	vmw_resource_unreserve(res, false, NULL, 0);
 	mutex_unlock(&res->dev_priv->cmdbuf_mutex);
 }
@@ -2562,7 +2713,7 @@ int vmw_kms_helper_resource_prepare(struct vmw_resource *res,
 		if (ret)
 			goto out_unreserve;
 
-		ctx->buf = vmw_dmabuf_reference(res->backup);
+		ctx->buf = vmw_bo_reference(res->backup);
 	}
 	ret = vmw_resource_validate(res);
 	if (ret)
@@ -2595,6 +2746,7 @@ void vmw_kms_helper_resource_finish(struct vmw_validation_ctx *ctx,
 		vmw_kms_helper_buffer_finish(res->dev_priv, NULL, ctx->buf,
 					     out_fence, NULL);
 
+	vmw_bo_unreference(&ctx->buf);
 	vmw_resource_unreserve(res, false, NULL, 0);
 	mutex_unlock(&res->dev_priv->cmdbuf_mutex);
 }
@@ -2680,7 +2832,9 @@ int vmw_kms_fbdev_init_data(struct vmw_private *dev_priv,
 	struct vmw_display_unit *du;
 	struct drm_display_mode *mode;
 	int i = 0;
+	int ret = 0;
 
+	mutex_lock(&dev_priv->dev->mode_config.mutex);
 	list_for_each_entry(con, &dev_priv->dev->mode_config.connector_list,
 			    head) {
 		if (i == unit)
@@ -2691,7 +2845,8 @@ int vmw_kms_fbdev_init_data(struct vmw_private *dev_priv,
 
 	if (i != unit) {
 		DRM_ERROR("Could not find initial display unit.\n");
-		return -EINVAL;
+		ret = -EINVAL;
+		goto out_unlock;
 	}
 
 	if (list_empty(&con->modes))
@@ -2699,7 +2854,8 @@ int vmw_kms_fbdev_init_data(struct vmw_private *dev_priv,
 
 	if (list_empty(&con->modes)) {
 		DRM_ERROR("Could not find initial display mode.\n");
-		return -EINVAL;
+		ret = -EINVAL;
+		goto out_unlock;
 	}
 
 	du = vmw_connector_to_du(con);
@@ -2720,7 +2876,10 @@ int vmw_kms_fbdev_init_data(struct vmw_private *dev_priv,
 					   head);
 	}
 
-	return 0;
+ out_unlock:
+	mutex_unlock(&dev_priv->dev->mode_config.mutex);
+
+	return ret;
 }
 
 /**
@@ -2798,6 +2957,7 @@ void vmw_kms_update_implicit_fb(struct vmw_private *dev_priv,
 				struct drm_crtc *crtc)
 {
 	struct vmw_display_unit *du = vmw_crtc_to_du(crtc);
+	struct drm_plane *plane = crtc->primary;
 	struct vmw_framebuffer *vfb;
 
 	mutex_lock(&dev_priv->global_kms_state_mutex);
@@ -2805,7 +2965,7 @@ void vmw_kms_update_implicit_fb(struct vmw_private *dev_priv,
 	if (!du->is_implicit)
 		goto out_unlock;
 
-	vfb = vmw_framebuffer_to_vfb(crtc->primary->fb);
+	vfb = vmw_framebuffer_to_vfb(plane->state->fb);
 	WARN_ON_ONCE(dev_priv->num_implicit != 1 &&
 		     dev_priv->implicit_fb != vfb);
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.h b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.h
index 6b7c012719f1..31311298ec0b 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_kms.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_kms.h
@@ -1,7 +1,7 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
- * Copyright © 2009-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -90,7 +90,7 @@ struct vmw_kms_dirty {
 #define vmw_framebuffer_to_vfbs(x) \
 	container_of(x, struct vmw_framebuffer_surface, base.base)
 #define vmw_framebuffer_to_vfbd(x) \
-	container_of(x, struct vmw_framebuffer_dmabuf, base.base)
+	container_of(x, struct vmw_framebuffer_bo, base.base)
 
 /**
  * Base class for framebuffers
@@ -102,7 +102,7 @@ struct vmw_framebuffer {
 	struct drm_framebuffer base;
 	int (*pin)(struct vmw_framebuffer *fb);
 	int (*unpin)(struct vmw_framebuffer *fb);
-	bool dmabuf;
+	bool bo;
 	struct ttm_base_object *user_obj;
 	uint32_t user_handle;
 };
@@ -117,15 +117,15 @@ struct vmw_clip_rect {
 struct vmw_framebuffer_surface {
 	struct vmw_framebuffer base;
 	struct vmw_surface *surface;
-	struct vmw_dma_buffer *buffer;
+	struct vmw_buffer_object *buffer;
 	struct list_head head;
-	bool is_dmabuf_proxy;  /* true if this is proxy surface for DMA buf */
+	bool is_bo_proxy;  /* true if this is proxy surface for DMA buf */
 };
 
 
-struct vmw_framebuffer_dmabuf {
+struct vmw_framebuffer_bo {
 	struct vmw_framebuffer base;
-	struct vmw_dma_buffer *buffer;
+	struct vmw_buffer_object *buffer;
 };
 
 
@@ -161,18 +161,18 @@ struct vmw_crtc_state {
  *
  * @base DRM plane object
  * @surf Display surface for STDU
- * @dmabuf display dmabuf for SOU
+ * @bo display bo for SOU
  * @content_fb_type Used by STDU.
- * @dmabuf_size Size of the dmabuf, used by Screen Object Display Unit
+ * @bo_size Size of the bo, used by Screen Object Display Unit
  * @pinned pin count for STDU display surface
  */
 struct vmw_plane_state {
 	struct drm_plane_state base;
 	struct vmw_surface *surf;
-	struct vmw_dma_buffer *dmabuf;
+	struct vmw_buffer_object *bo;
 
 	int content_fb_type;
-	unsigned long dmabuf_size;
+	unsigned long bo_size;
 
 	int pinned;
 
@@ -192,6 +192,24 @@ struct vmw_connector_state {
 	struct drm_connector_state base;
 
 	bool is_implicit;
+
+	/**
+	 * @gui_x:
+	 *
+	 * vmwgfx connector property representing the x position of this display
+	 * unit (connector is synonymous to display unit) in overall topology.
+	 * This is what the device expect as xRoot while creating screen.
+	 */
+	int gui_x;
+
+	/**
+	 * @gui_y:
+	 *
+	 * vmwgfx connector property representing the y position of this display
+	 * unit (connector is synonymous to display unit) in overall topology.
+	 * This is what the device expect as yRoot while creating screen.
+	 */
+	int gui_y;
 };
 
 /**
@@ -209,7 +227,7 @@ struct vmw_display_unit {
 	struct drm_plane cursor;
 
 	struct vmw_surface *cursor_surface;
-	struct vmw_dma_buffer *cursor_dmabuf;
+	struct vmw_buffer_object *cursor_bo;
 	size_t cursor_age;
 
 	int cursor_x;
@@ -243,7 +261,7 @@ struct vmw_display_unit {
 
 struct vmw_validation_ctx {
 	struct vmw_resource *res;
-	struct vmw_dma_buffer *buf;
+	struct vmw_buffer_object *buf;
 };
 
 #define vmw_crtc_to_du(x) \
@@ -291,14 +309,14 @@ int vmw_kms_helper_dirty(struct vmw_private *dev_priv,
 			 struct vmw_kms_dirty *dirty);
 
 int vmw_kms_helper_buffer_prepare(struct vmw_private *dev_priv,
-				  struct vmw_dma_buffer *buf,
+				  struct vmw_buffer_object *buf,
 				  bool interruptible,
 				  bool validate_as_mob,
 				  bool for_cpu_blit);
-void vmw_kms_helper_buffer_revert(struct vmw_dma_buffer *buf);
+void vmw_kms_helper_buffer_revert(struct vmw_buffer_object *buf);
 void vmw_kms_helper_buffer_finish(struct vmw_private *dev_priv,
 				  struct drm_file *file_priv,
-				  struct vmw_dma_buffer *buf,
+				  struct vmw_buffer_object *buf,
 				  struct vmw_fence_obj **out_fence,
 				  struct drm_vmw_fence_rep __user *
 				  user_fence_rep);
@@ -316,7 +334,7 @@ int vmw_kms_readback(struct vmw_private *dev_priv,
 		     uint32_t num_clips);
 struct vmw_framebuffer *
 vmw_kms_new_framebuffer(struct vmw_private *dev_priv,
-			struct vmw_dma_buffer *dmabuf,
+			struct vmw_buffer_object *bo,
 			struct vmw_surface *surface,
 			bool only_2d,
 			const struct drm_mode_fb_cmd2 *mode_cmd);
@@ -384,11 +402,11 @@ void vmw_du_connector_destroy_state(struct drm_connector *connector,
  */
 int vmw_kms_ldu_init_display(struct vmw_private *dev_priv);
 int vmw_kms_ldu_close_display(struct vmw_private *dev_priv);
-int vmw_kms_ldu_do_dmabuf_dirty(struct vmw_private *dev_priv,
-				struct vmw_framebuffer *framebuffer,
-				unsigned flags, unsigned color,
-				struct drm_clip_rect *clips,
-				unsigned num_clips, int increment);
+int vmw_kms_ldu_do_bo_dirty(struct vmw_private *dev_priv,
+			    struct vmw_framebuffer *framebuffer,
+			    unsigned int flags, unsigned int color,
+			    struct drm_clip_rect *clips,
+			    unsigned int num_clips, int increment);
 int vmw_kms_update_proxy(struct vmw_resource *res,
 			 const struct drm_clip_rect *clips,
 			 unsigned num_clips,
@@ -408,14 +426,14 @@ int vmw_kms_sou_do_surface_dirty(struct vmw_private *dev_priv,
 				 unsigned num_clips, int inc,
 				 struct vmw_fence_obj **out_fence,
 				 struct drm_crtc *crtc);
-int vmw_kms_sou_do_dmabuf_dirty(struct vmw_private *dev_priv,
-				struct vmw_framebuffer *framebuffer,
-				struct drm_clip_rect *clips,
-				struct drm_vmw_rect *vclips,
-				unsigned num_clips, int increment,
-				bool interruptible,
-				struct vmw_fence_obj **out_fence,
-				struct drm_crtc *crtc);
+int vmw_kms_sou_do_bo_dirty(struct vmw_private *dev_priv,
+			    struct vmw_framebuffer *framebuffer,
+			    struct drm_clip_rect *clips,
+			    struct drm_vmw_rect *vclips,
+			    unsigned int num_clips, int increment,
+			    bool interruptible,
+			    struct vmw_fence_obj **out_fence,
+			    struct drm_crtc *crtc);
 int vmw_kms_sou_readback(struct vmw_private *dev_priv,
 			 struct drm_file *file_priv,
 			 struct vmw_framebuffer *vfb,
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c
index 4a5907e3f560..723578117191 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -438,7 +438,7 @@ static int vmw_ldu_init(struct vmw_private *dev_priv, unsigned unit)
 		goto err_free_connector;
 	}
 
-	(void) drm_mode_connector_attach_encoder(connector, encoder);
+	(void) drm_connector_attach_encoder(connector, encoder);
 	encoder->possible_crtcs = (1 << unit);
 	encoder->possible_clones = 0;
 
@@ -547,11 +547,11 @@ int vmw_kms_ldu_close_display(struct vmw_private *dev_priv)
 }
 
 
-int vmw_kms_ldu_do_dmabuf_dirty(struct vmw_private *dev_priv,
-				struct vmw_framebuffer *framebuffer,
-				unsigned flags, unsigned color,
-				struct drm_clip_rect *clips,
-				unsigned num_clips, int increment)
+int vmw_kms_ldu_do_bo_dirty(struct vmw_private *dev_priv,
+			    struct vmw_framebuffer *framebuffer,
+			    unsigned int flags, unsigned int color,
+			    struct drm_clip_rect *clips,
+			    unsigned int num_clips, int increment)
 {
 	size_t fifo_size;
 	int i;
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_marker.c b/drivers/gpu/drm/vmwgfx/vmwgfx_marker.c
index efd1ffd68185..e53bc639a754 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_marker.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_marker.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright (C) 2010 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2010 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c b/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c
index d07c585e3c1d..7ed179d30ec5 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2012-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2012-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -225,7 +225,7 @@ static void vmw_takedown_otable_base(struct vmw_private *dev_priv,
 		ret = ttm_bo_reserve(bo, false, true, NULL);
 		BUG_ON(ret != 0);
 
-		vmw_fence_single_bo(bo, NULL);
+		vmw_bo_fence_single(bo, NULL);
 		ttm_bo_unreserve(bo);
 	}
 
@@ -362,7 +362,7 @@ static void vmw_otable_batch_takedown(struct vmw_private *dev_priv,
 	ret = ttm_bo_reserve(bo, false, true, NULL);
 	BUG_ON(ret != 0);
 
-	vmw_fence_single_bo(bo, NULL);
+	vmw_bo_fence_single(bo, NULL);
 	ttm_bo_unreserve(bo);
 
 	ttm_bo_unref(&batch->otable_bo);
@@ -620,7 +620,7 @@ void vmw_mob_unbind(struct vmw_private *dev_priv,
 		vmw_fifo_commit(dev_priv, sizeof(*cmd));
 	}
 	if (bo) {
-		vmw_fence_single_bo(bo, NULL);
+		vmw_bo_fence_single(bo, NULL);
 		ttm_bo_unreserve(bo);
 	}
 	vmw_fifo_resource_dec(dev_priv);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c
index cdff99211602..8b9270f31409 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c
@@ -1,6 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /*
- * Copyright © 2016 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2016 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -31,6 +31,7 @@
 #include <linux/frame.h>
 #include <asm/hypervisor.h>
 #include <drm/drmP.h>
+#include "vmwgfx_drv.h"
 #include "vmwgfx_msg.h"
 
 
@@ -234,7 +235,7 @@ static int vmw_recv_msg(struct rpc_channel *channel, void **msg,
 
 		if ((HIGH_WORD(ecx) & MESSAGE_STATUS_SUCCESS) == 0 ||
 		    (HIGH_WORD(ecx) & MESSAGE_STATUS_HB) == 0) {
-			DRM_ERROR("Failed to get reply size\n");
+			DRM_ERROR("Failed to get reply size for host message.\n");
 			return -EINVAL;
 		}
 
@@ -245,7 +246,7 @@ static int vmw_recv_msg(struct rpc_channel *channel, void **msg,
 		reply_len = ebx;
 		reply     = kzalloc(reply_len + 1, GFP_KERNEL);
 		if (!reply) {
-			DRM_ERROR("Cannot allocate memory for reply\n");
+			DRM_ERROR("Cannot allocate memory for host message reply.\n");
 			return -ENOMEM;
 		}
 
@@ -329,8 +330,6 @@ int vmw_host_get_guestinfo(const char *guest_info_param,
 	struct rpc_channel channel;
 	char *msg, *reply = NULL;
 	size_t reply_len = 0;
-	int ret = 0;
-
 
 	if (!vmw_msg_enabled)
 		return -ENODEV;
@@ -340,19 +339,19 @@ int vmw_host_get_guestinfo(const char *guest_info_param,
 
 	msg = kasprintf(GFP_KERNEL, "info-get %s", guest_info_param);
 	if (!msg) {
-		DRM_ERROR("Cannot allocate memory to get %s", guest_info_param);
+		DRM_ERROR("Cannot allocate memory to get guest info \"%s\".",
+			  guest_info_param);
 		return -ENOMEM;
 	}
 
-	if (vmw_open_channel(&channel, RPCI_PROTOCOL_NUM) ||
-	    vmw_send_msg(&channel, msg) ||
-	    vmw_recv_msg(&channel, (void *) &reply, &reply_len) ||
-	    vmw_close_channel(&channel)) {
-		DRM_ERROR("Failed to get %s", guest_info_param);
+	if (vmw_open_channel(&channel, RPCI_PROTOCOL_NUM))
+		goto out_open;
 
-		ret = -EINVAL;
-	}
+	if (vmw_send_msg(&channel, msg) ||
+	    vmw_recv_msg(&channel, (void *) &reply, &reply_len))
+		goto out_msg;
 
+	vmw_close_channel(&channel);
 	if (buffer && reply && reply_len > 0) {
 		/* Remove reply code, which are the first 2 characters of
 		 * the reply
@@ -369,7 +368,17 @@ int vmw_host_get_guestinfo(const char *guest_info_param,
 	kfree(reply);
 	kfree(msg);
 
-	return ret;
+	return 0;
+
+out_msg:
+	vmw_close_channel(&channel);
+	kfree(reply);
+out_open:
+	*length = 0;
+	kfree(msg);
+	DRM_ERROR("Failed to get guest info \"%s\".", guest_info_param);
+
+	return -EINVAL;
 }
 
 
@@ -396,19 +405,26 @@ int vmw_host_log(const char *log)
 
 	msg = kasprintf(GFP_KERNEL, "log %s", log);
 	if (!msg) {
-		DRM_ERROR("Cannot allocate memory for log message\n");
+		DRM_ERROR("Cannot allocate memory for host log message.\n");
 		return -ENOMEM;
 	}
 
-	if (vmw_open_channel(&channel, RPCI_PROTOCOL_NUM) ||
-	    vmw_send_msg(&channel, msg) ||
-	    vmw_close_channel(&channel)) {
-		DRM_ERROR("Failed to send log\n");
+	if (vmw_open_channel(&channel, RPCI_PROTOCOL_NUM))
+		goto out_open;
 
-		ret = -EINVAL;
-	}
+	if (vmw_send_msg(&channel, msg))
+		goto out_msg;
 
+	vmw_close_channel(&channel);
 	kfree(msg);
 
-	return ret;
+	return 0;
+
+out_msg:
+	vmw_close_channel(&channel);
+out_open:
+	kfree(msg);
+	DRM_ERROR("Failed to send host log message.\n");
+
+	return -EINVAL;
 }
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.h b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.h
index 557a033fb610..4907e50fb20a 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.h
@@ -1,16 +1,29 @@
-/*
- * Copyright (C) 2016, VMware, Inc.
+/* SPDX-License-Identifier: GPL-2.0+ OR MIT */
+/**************************************************************************
+ *
+ * Copyright 2016 VMware, Inc., Palo Alto, CA., USA
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
  *
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License as published by
- * the Free Software Foundation; either version 2 of the License, or
- * (at your option) any later version.
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
  *
- * This program is distributed in the hope that it will be useful, but
- * WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
- * NON INFRINGEMENT.  See the GNU General Public License for more
- * details.
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ **************************************************************************
  *
  * Based on code from vmware.c and vmmouse.c.
  * Author:
@@ -135,17 +148,24 @@
 
 #else
 
-/* In the 32-bit version of this macro, we use "m" because there is no
- * more register left for bp
+/*
+ * In the 32-bit version of this macro, we store bp in a memory location
+ * because we've ran out of registers.
+ * Now we can't reference that memory location while we've modified
+ * %esp or %ebp, so we first push it on the stack, just before we push
+ * %ebp, and then when we need it we read it from the stack where we
+ * just pushed it.
  */
 #define VMW_PORT_HB_OUT(cmd, in_ecx, in_si, in_di,	\
 			port_num, magic, bp,		\
 			eax, ebx, ecx, edx, si, di)	\
 ({							\
-	asm volatile ("push %%ebp;"			\
-		"mov %12, %%ebp;"			\
+	asm volatile ("push %12;"			\
+		"push %%ebp;"				\
+		"mov 0x04(%%esp), %%ebp;"		\
 		"rep outsb;"				\
-		"pop %%ebp;" :				\
+		"pop %%ebp;"				\
+		"add $0x04, %%esp;" :			\
 		"=a"(eax),				\
 		"=b"(ebx),				\
 		"=c"(ecx),				\
@@ -167,10 +187,12 @@
 		       port_num, magic, bp,		\
 		       eax, ebx, ecx, edx, si, di)	\
 ({							\
-	asm volatile ("push %%ebp;"			\
-		"mov %12, %%ebp;"			\
+	asm volatile ("push %12;"			\
+		"push %%ebp;"				\
+		"mov 0x04(%%esp), %%ebp;"		\
 		"rep insb;"				\
-		"pop %%ebp" :				\
+		"pop %%ebp;"				\
+		"add $0x04, %%esp;" :			\
 		"=a"(eax),				\
 		"=b"(ebx),				\
 		"=c"(ecx),				\
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_overlay.c b/drivers/gpu/drm/vmwgfx/vmwgfx_overlay.c
index 222c9c2123a1..9f1b9d289bec 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_overlay.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_overlay.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009-2014 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2014 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -38,7 +38,7 @@
 #define VMW_OVERLAY_CAP_MASK (SVGA_FIFO_CAP_VIDEO | SVGA_FIFO_CAP_ESCAPE)
 
 struct vmw_stream {
-	struct vmw_dma_buffer *buf;
+	struct vmw_buffer_object *buf;
 	bool claimed;
 	bool paused;
 	struct drm_vmw_control_stream_arg saved;
@@ -94,7 +94,7 @@ static inline void fill_flush(struct vmw_escape_video_flush *cmd,
  * -ERESTARTSYS if interrupted by a signal.
  */
 static int vmw_overlay_send_put(struct vmw_private *dev_priv,
-				struct vmw_dma_buffer *buf,
+				struct vmw_buffer_object *buf,
 				struct drm_vmw_control_stream_arg *arg,
 				bool interruptible)
 {
@@ -225,16 +225,16 @@ static int vmw_overlay_send_stop(struct vmw_private *dev_priv,
  * used with GMRs instead of being locked to vram.
  */
 static int vmw_overlay_move_buffer(struct vmw_private *dev_priv,
-				   struct vmw_dma_buffer *buf,
+				   struct vmw_buffer_object *buf,
 				   bool pin, bool inter)
 {
 	if (!pin)
-		return vmw_dmabuf_unpin(dev_priv, buf, inter);
+		return vmw_bo_unpin(dev_priv, buf, inter);
 
 	if (dev_priv->active_display_unit == vmw_du_legacy)
-		return vmw_dmabuf_pin_in_vram(dev_priv, buf, inter);
+		return vmw_bo_pin_in_vram(dev_priv, buf, inter);
 
-	return vmw_dmabuf_pin_in_vram_or_gmr(dev_priv, buf, inter);
+	return vmw_bo_pin_in_vram_or_gmr(dev_priv, buf, inter);
 }
 
 /**
@@ -278,7 +278,7 @@ static int vmw_overlay_stop(struct vmw_private *dev_priv,
 	}
 
 	if (!pause) {
-		vmw_dmabuf_unreference(&stream->buf);
+		vmw_bo_unreference(&stream->buf);
 		stream->paused = false;
 	} else {
 		stream->paused = true;
@@ -297,7 +297,7 @@ static int vmw_overlay_stop(struct vmw_private *dev_priv,
  * -ERESTARTSYS if interrupted.
  */
 static int vmw_overlay_update_stream(struct vmw_private *dev_priv,
-				     struct vmw_dma_buffer *buf,
+				     struct vmw_buffer_object *buf,
 				     struct drm_vmw_control_stream_arg *arg,
 				     bool interruptible)
 {
@@ -347,7 +347,7 @@ static int vmw_overlay_update_stream(struct vmw_private *dev_priv,
 	}
 
 	if (stream->buf != buf)
-		stream->buf = vmw_dmabuf_reference(buf);
+		stream->buf = vmw_bo_reference(buf);
 	stream->saved = *arg;
 	/* stream is no longer stopped/paused */
 	stream->paused = false;
@@ -466,7 +466,7 @@ int vmw_overlay_ioctl(struct drm_device *dev, void *data,
 	struct vmw_overlay *overlay = dev_priv->overlay_priv;
 	struct drm_vmw_control_stream_arg *arg =
 	    (struct drm_vmw_control_stream_arg *)data;
-	struct vmw_dma_buffer *buf;
+	struct vmw_buffer_object *buf;
 	struct vmw_resource *res;
 	int ret;
 
@@ -484,13 +484,13 @@ int vmw_overlay_ioctl(struct drm_device *dev, void *data,
 		goto out_unlock;
 	}
 
-	ret = vmw_user_dmabuf_lookup(tfile, arg->handle, &buf, NULL);
+	ret = vmw_user_bo_lookup(tfile, arg->handle, &buf, NULL);
 	if (ret)
 		goto out_unlock;
 
 	ret = vmw_overlay_update_stream(dev_priv, buf, arg, true);
 
-	vmw_dmabuf_unreference(&buf);
+	vmw_bo_unreference(&buf);
 
 out_unlock:
 	mutex_unlock(&overlay->mutex);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_prime.c b/drivers/gpu/drm/vmwgfx/vmwgfx_prime.c
index 0d42a46521fc..0861c821a7fe 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_prime.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_prime.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2013 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2013 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -40,7 +40,6 @@
  */
 
 static int vmw_prime_map_attach(struct dma_buf *dma_buf,
-				struct device *target_dev,
 				struct dma_buf_attachment *attach)
 {
 	return -ENOSYS;
@@ -72,17 +71,6 @@ static void vmw_prime_dmabuf_vunmap(struct dma_buf *dma_buf, void *vaddr)
 {
 }
 
-static void *vmw_prime_dmabuf_kmap_atomic(struct dma_buf *dma_buf,
-		unsigned long page_num)
-{
-	return NULL;
-}
-
-static void vmw_prime_dmabuf_kunmap_atomic(struct dma_buf *dma_buf,
-		unsigned long page_num, void *addr)
-{
-
-}
 static void *vmw_prime_dmabuf_kmap(struct dma_buf *dma_buf,
 		unsigned long page_num)
 {
@@ -109,9 +97,7 @@ const struct dma_buf_ops vmw_prime_dmabuf_ops =  {
 	.unmap_dma_buf = vmw_prime_unmap_dma_buf,
 	.release = NULL,
 	.map = vmw_prime_dmabuf_kmap,
-	.map_atomic = vmw_prime_dmabuf_kmap_atomic,
 	.unmap = vmw_prime_dmabuf_kunmap,
-	.unmap_atomic = vmw_prime_dmabuf_kunmap_atomic,
 	.mmap = vmw_prime_dmabuf_mmap,
 	.vmap = vmw_prime_dmabuf_vmap,
 	.vunmap = vmw_prime_dmabuf_vunmap,
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_reg.h b/drivers/gpu/drm/vmwgfx/vmwgfx_reg.h
index dce798053a96..e99f6cdbb091 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_reg.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_reg.h
@@ -1,7 +1,7 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
- * Copyright © 2009-2014 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2014 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
index 6b3a942b18df..92003ea5a219 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -27,7 +27,6 @@
 
 #include "vmwgfx_drv.h"
 #include <drm/vmwgfx_drm.h>
-#include <drm/ttm/ttm_object.h>
 #include <drm/ttm/ttm_placement.h>
 #include <drm/drmP.h>
 #include "vmwgfx_resource_priv.h"
@@ -35,29 +34,6 @@
 
 #define VMW_RES_EVICT_ERR_COUNT 10
 
-struct vmw_user_dma_buffer {
-	struct ttm_prime_object prime;
-	struct vmw_dma_buffer dma;
-};
-
-struct vmw_bo_user_rep {
-	uint32_t handle;
-	uint64_t map_handle;
-};
-
-static inline struct vmw_dma_buffer *
-vmw_dma_buffer(struct ttm_buffer_object *bo)
-{
-	return container_of(bo, struct vmw_dma_buffer, base);
-}
-
-static inline struct vmw_user_dma_buffer *
-vmw_user_dma_buffer(struct ttm_buffer_object *bo)
-{
-	struct vmw_dma_buffer *vmw_bo = vmw_dma_buffer(bo);
-	return container_of(vmw_bo, struct vmw_user_dma_buffer, dma);
-}
-
 struct vmw_resource *vmw_resource_reference(struct vmw_resource *res)
 {
 	kref_get(&res->kref);
@@ -116,7 +92,7 @@ static void vmw_resource_release(struct kref *kref)
 		res->backup_dirty = false;
 		list_del_init(&res->mob_head);
 		ttm_bo_unreserve(bo);
-		vmw_dmabuf_unreference(&res->backup);
+		vmw_bo_unreference(&res->backup);
 	}
 
 	if (likely(res->hw_destroy != NULL)) {
@@ -287,7 +263,7 @@ out_bad_resource:
 }
 
 /**
- * Helper function that looks either a surface or dmabuf.
+ * Helper function that looks either a surface or bo.
  *
  * The pointer this pointed at by out_surf and out_buf needs to be null.
  */
@@ -295,7 +271,7 @@ int vmw_user_lookup_handle(struct vmw_private *dev_priv,
 			   struct ttm_object_file *tfile,
 			   uint32_t handle,
 			   struct vmw_surface **out_surf,
-			   struct vmw_dma_buffer **out_buf)
+			   struct vmw_buffer_object **out_buf)
 {
 	struct vmw_resource *res;
 	int ret;
@@ -311,513 +287,11 @@ int vmw_user_lookup_handle(struct vmw_private *dev_priv,
 	}
 
 	*out_surf = NULL;
-	ret = vmw_user_dmabuf_lookup(tfile, handle, out_buf, NULL);
+	ret = vmw_user_bo_lookup(tfile, handle, out_buf, NULL);
 	return ret;
 }
 
 /**
- * Buffer management.
- */
-
-/**
- * vmw_dmabuf_acc_size - Calculate the pinned memory usage of buffers
- *
- * @dev_priv: Pointer to a struct vmw_private identifying the device.
- * @size: The requested buffer size.
- * @user: Whether this is an ordinary dma buffer or a user dma buffer.
- */
-static size_t vmw_dmabuf_acc_size(struct vmw_private *dev_priv, size_t size,
-				  bool user)
-{
-	static size_t struct_size, user_struct_size;
-	size_t num_pages = PAGE_ALIGN(size) >> PAGE_SHIFT;
-	size_t page_array_size = ttm_round_pot(num_pages * sizeof(void *));
-
-	if (unlikely(struct_size == 0)) {
-		size_t backend_size = ttm_round_pot(vmw_tt_size);
-
-		struct_size = backend_size +
-			ttm_round_pot(sizeof(struct vmw_dma_buffer));
-		user_struct_size = backend_size +
-			ttm_round_pot(sizeof(struct vmw_user_dma_buffer));
-	}
-
-	if (dev_priv->map_mode == vmw_dma_alloc_coherent)
-		page_array_size +=
-			ttm_round_pot(num_pages * sizeof(dma_addr_t));
-
-	return ((user) ? user_struct_size : struct_size) +
-		page_array_size;
-}
-
-void vmw_dmabuf_bo_free(struct ttm_buffer_object *bo)
-{
-	struct vmw_dma_buffer *vmw_bo = vmw_dma_buffer(bo);
-
-	vmw_dma_buffer_unmap(vmw_bo);
-	kfree(vmw_bo);
-}
-
-static void vmw_user_dmabuf_destroy(struct ttm_buffer_object *bo)
-{
-	struct vmw_user_dma_buffer *vmw_user_bo = vmw_user_dma_buffer(bo);
-
-	vmw_dma_buffer_unmap(&vmw_user_bo->dma);
-	ttm_prime_object_kfree(vmw_user_bo, prime);
-}
-
-int vmw_dmabuf_init(struct vmw_private *dev_priv,
-		    struct vmw_dma_buffer *vmw_bo,
-		    size_t size, struct ttm_placement *placement,
-		    bool interruptible,
-		    void (*bo_free) (struct ttm_buffer_object *bo))
-{
-	struct ttm_bo_device *bdev = &dev_priv->bdev;
-	size_t acc_size;
-	int ret;
-	bool user = (bo_free == &vmw_user_dmabuf_destroy);
-
-	BUG_ON(!bo_free && (!user && (bo_free != vmw_dmabuf_bo_free)));
-
-	acc_size = vmw_dmabuf_acc_size(dev_priv, size, user);
-	memset(vmw_bo, 0, sizeof(*vmw_bo));
-
-	INIT_LIST_HEAD(&vmw_bo->res_list);
-
-	ret = ttm_bo_init(bdev, &vmw_bo->base, size,
-			  ttm_bo_type_device, placement,
-			  0, interruptible, acc_size,
-			  NULL, NULL, bo_free);
-	return ret;
-}
-
-static void vmw_user_dmabuf_release(struct ttm_base_object **p_base)
-{
-	struct vmw_user_dma_buffer *vmw_user_bo;
-	struct ttm_base_object *base = *p_base;
-	struct ttm_buffer_object *bo;
-
-	*p_base = NULL;
-
-	if (unlikely(base == NULL))
-		return;
-
-	vmw_user_bo = container_of(base, struct vmw_user_dma_buffer,
-				   prime.base);
-	bo = &vmw_user_bo->dma.base;
-	ttm_bo_unref(&bo);
-}
-
-static void vmw_user_dmabuf_ref_obj_release(struct ttm_base_object *base,
-					    enum ttm_ref_type ref_type)
-{
-	struct vmw_user_dma_buffer *user_bo;
-	user_bo = container_of(base, struct vmw_user_dma_buffer, prime.base);
-
-	switch (ref_type) {
-	case TTM_REF_SYNCCPU_WRITE:
-		ttm_bo_synccpu_write_release(&user_bo->dma.base);
-		break;
-	default:
-		BUG();
-	}
-}
-
-/**
- * vmw_user_dmabuf_alloc - Allocate a user dma buffer
- *
- * @dev_priv: Pointer to a struct device private.
- * @tfile: Pointer to a struct ttm_object_file on which to register the user
- * object.
- * @size: Size of the dma buffer.
- * @shareable: Boolean whether the buffer is shareable with other open files.
- * @handle: Pointer to where the handle value should be assigned.
- * @p_dma_buf: Pointer to where the refcounted struct vmw_dma_buffer pointer
- * should be assigned.
- */
-int vmw_user_dmabuf_alloc(struct vmw_private *dev_priv,
-			  struct ttm_object_file *tfile,
-			  uint32_t size,
-			  bool shareable,
-			  uint32_t *handle,
-			  struct vmw_dma_buffer **p_dma_buf,
-			  struct ttm_base_object **p_base)
-{
-	struct vmw_user_dma_buffer *user_bo;
-	struct ttm_buffer_object *tmp;
-	int ret;
-
-	user_bo = kzalloc(sizeof(*user_bo), GFP_KERNEL);
-	if (unlikely(!user_bo)) {
-		DRM_ERROR("Failed to allocate a buffer.\n");
-		return -ENOMEM;
-	}
-
-	ret = vmw_dmabuf_init(dev_priv, &user_bo->dma, size,
-			      (dev_priv->has_mob) ?
-			      &vmw_sys_placement :
-			      &vmw_vram_sys_placement, true,
-			      &vmw_user_dmabuf_destroy);
-	if (unlikely(ret != 0))
-		return ret;
-
-	tmp = ttm_bo_reference(&user_bo->dma.base);
-	ret = ttm_prime_object_init(tfile,
-				    size,
-				    &user_bo->prime,
-				    shareable,
-				    ttm_buffer_type,
-				    &vmw_user_dmabuf_release,
-				    &vmw_user_dmabuf_ref_obj_release);
-	if (unlikely(ret != 0)) {
-		ttm_bo_unref(&tmp);
-		goto out_no_base_object;
-	}
-
-	*p_dma_buf = &user_bo->dma;
-	if (p_base) {
-		*p_base = &user_bo->prime.base;
-		kref_get(&(*p_base)->refcount);
-	}
-	*handle = user_bo->prime.base.hash.key;
-
-out_no_base_object:
-	return ret;
-}
-
-/**
- * vmw_user_dmabuf_verify_access - verify access permissions on this
- * buffer object.
- *
- * @bo: Pointer to the buffer object being accessed
- * @tfile: Identifying the caller.
- */
-int vmw_user_dmabuf_verify_access(struct ttm_buffer_object *bo,
-				  struct ttm_object_file *tfile)
-{
-	struct vmw_user_dma_buffer *vmw_user_bo;
-
-	if (unlikely(bo->destroy != vmw_user_dmabuf_destroy))
-		return -EPERM;
-
-	vmw_user_bo = vmw_user_dma_buffer(bo);
-
-	/* Check that the caller has opened the object. */
-	if (likely(ttm_ref_object_exists(tfile, &vmw_user_bo->prime.base)))
-		return 0;
-
-	DRM_ERROR("Could not grant buffer access.\n");
-	return -EPERM;
-}
-
-/**
- * vmw_user_dmabuf_synccpu_grab - Grab a struct vmw_user_dma_buffer for cpu
- * access, idling previous GPU operations on the buffer and optionally
- * blocking it for further command submissions.
- *
- * @user_bo: Pointer to the buffer object being grabbed for CPU access
- * @tfile: Identifying the caller.
- * @flags: Flags indicating how the grab should be performed.
- *
- * A blocking grab will be automatically released when @tfile is closed.
- */
-static int vmw_user_dmabuf_synccpu_grab(struct vmw_user_dma_buffer *user_bo,
-					struct ttm_object_file *tfile,
-					uint32_t flags)
-{
-	struct ttm_buffer_object *bo = &user_bo->dma.base;
-	bool existed;
-	int ret;
-
-	if (flags & drm_vmw_synccpu_allow_cs) {
-		bool nonblock = !!(flags & drm_vmw_synccpu_dontblock);
-		long lret;
-
-		lret = reservation_object_wait_timeout_rcu(bo->resv, true, true,
-							   nonblock ? 0 : MAX_SCHEDULE_TIMEOUT);
-		if (!lret)
-			return -EBUSY;
-		else if (lret < 0)
-			return lret;
-		return 0;
-	}
-
-	ret = ttm_bo_synccpu_write_grab
-		(bo, !!(flags & drm_vmw_synccpu_dontblock));
-	if (unlikely(ret != 0))
-		return ret;
-
-	ret = ttm_ref_object_add(tfile, &user_bo->prime.base,
-				 TTM_REF_SYNCCPU_WRITE, &existed, false);
-	if (ret != 0 || existed)
-		ttm_bo_synccpu_write_release(&user_bo->dma.base);
-
-	return ret;
-}
-
-/**
- * vmw_user_dmabuf_synccpu_release - Release a previous grab for CPU access,
- * and unblock command submission on the buffer if blocked.
- *
- * @handle: Handle identifying the buffer object.
- * @tfile: Identifying the caller.
- * @flags: Flags indicating the type of release.
- */
-static int vmw_user_dmabuf_synccpu_release(uint32_t handle,
-					   struct ttm_object_file *tfile,
-					   uint32_t flags)
-{
-	if (!(flags & drm_vmw_synccpu_allow_cs))
-		return ttm_ref_object_base_unref(tfile, handle,
-						 TTM_REF_SYNCCPU_WRITE);
-
-	return 0;
-}
-
-/**
- * vmw_user_dmabuf_synccpu_release - ioctl function implementing the synccpu
- * functionality.
- *
- * @dev: Identifies the drm device.
- * @data: Pointer to the ioctl argument.
- * @file_priv: Identifies the caller.
- *
- * This function checks the ioctl arguments for validity and calls the
- * relevant synccpu functions.
- */
-int vmw_user_dmabuf_synccpu_ioctl(struct drm_device *dev, void *data,
-				  struct drm_file *file_priv)
-{
-	struct drm_vmw_synccpu_arg *arg =
-		(struct drm_vmw_synccpu_arg *) data;
-	struct vmw_dma_buffer *dma_buf;
-	struct vmw_user_dma_buffer *user_bo;
-	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
-	struct ttm_base_object *buffer_base;
-	int ret;
-
-	if ((arg->flags & (drm_vmw_synccpu_read | drm_vmw_synccpu_write)) == 0
-	    || (arg->flags & ~(drm_vmw_synccpu_read | drm_vmw_synccpu_write |
-			       drm_vmw_synccpu_dontblock |
-			       drm_vmw_synccpu_allow_cs)) != 0) {
-		DRM_ERROR("Illegal synccpu flags.\n");
-		return -EINVAL;
-	}
-
-	switch (arg->op) {
-	case drm_vmw_synccpu_grab:
-		ret = vmw_user_dmabuf_lookup(tfile, arg->handle, &dma_buf,
-					     &buffer_base);
-		if (unlikely(ret != 0))
-			return ret;
-
-		user_bo = container_of(dma_buf, struct vmw_user_dma_buffer,
-				       dma);
-		ret = vmw_user_dmabuf_synccpu_grab(user_bo, tfile, arg->flags);
-		vmw_dmabuf_unreference(&dma_buf);
-		ttm_base_object_unref(&buffer_base);
-		if (unlikely(ret != 0 && ret != -ERESTARTSYS &&
-			     ret != -EBUSY)) {
-			DRM_ERROR("Failed synccpu grab on handle 0x%08x.\n",
-				  (unsigned int) arg->handle);
-			return ret;
-		}
-		break;
-	case drm_vmw_synccpu_release:
-		ret = vmw_user_dmabuf_synccpu_release(arg->handle, tfile,
-						      arg->flags);
-		if (unlikely(ret != 0)) {
-			DRM_ERROR("Failed synccpu release on handle 0x%08x.\n",
-				  (unsigned int) arg->handle);
-			return ret;
-		}
-		break;
-	default:
-		DRM_ERROR("Invalid synccpu operation.\n");
-		return -EINVAL;
-	}
-
-	return 0;
-}
-
-int vmw_dmabuf_alloc_ioctl(struct drm_device *dev, void *data,
-			   struct drm_file *file_priv)
-{
-	struct vmw_private *dev_priv = vmw_priv(dev);
-	union drm_vmw_alloc_dmabuf_arg *arg =
-	    (union drm_vmw_alloc_dmabuf_arg *)data;
-	struct drm_vmw_alloc_dmabuf_req *req = &arg->req;
-	struct drm_vmw_dmabuf_rep *rep = &arg->rep;
-	struct vmw_dma_buffer *dma_buf;
-	uint32_t handle;
-	int ret;
-
-	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
-	if (unlikely(ret != 0))
-		return ret;
-
-	ret = vmw_user_dmabuf_alloc(dev_priv, vmw_fpriv(file_priv)->tfile,
-				    req->size, false, &handle, &dma_buf,
-				    NULL);
-	if (unlikely(ret != 0))
-		goto out_no_dmabuf;
-
-	rep->handle = handle;
-	rep->map_handle = drm_vma_node_offset_addr(&dma_buf->base.vma_node);
-	rep->cur_gmr_id = handle;
-	rep->cur_gmr_offset = 0;
-
-	vmw_dmabuf_unreference(&dma_buf);
-
-out_no_dmabuf:
-	ttm_read_unlock(&dev_priv->reservation_sem);
-
-	return ret;
-}
-
-int vmw_dmabuf_unref_ioctl(struct drm_device *dev, void *data,
-			   struct drm_file *file_priv)
-{
-	struct drm_vmw_unref_dmabuf_arg *arg =
-	    (struct drm_vmw_unref_dmabuf_arg *)data;
-
-	return ttm_ref_object_base_unref(vmw_fpriv(file_priv)->tfile,
-					 arg->handle,
-					 TTM_REF_USAGE);
-}
-
-int vmw_user_dmabuf_lookup(struct ttm_object_file *tfile,
-			   uint32_t handle, struct vmw_dma_buffer **out,
-			   struct ttm_base_object **p_base)
-{
-	struct vmw_user_dma_buffer *vmw_user_bo;
-	struct ttm_base_object *base;
-
-	base = ttm_base_object_lookup(tfile, handle);
-	if (unlikely(base == NULL)) {
-		pr_err("Invalid buffer object handle 0x%08lx\n",
-		       (unsigned long)handle);
-		return -ESRCH;
-	}
-
-	if (unlikely(ttm_base_object_type(base) != ttm_buffer_type)) {
-		ttm_base_object_unref(&base);
-		pr_err("Invalid buffer object handle 0x%08lx\n",
-		       (unsigned long)handle);
-		return -EINVAL;
-	}
-
-	vmw_user_bo = container_of(base, struct vmw_user_dma_buffer,
-				   prime.base);
-	(void)ttm_bo_reference(&vmw_user_bo->dma.base);
-	if (p_base)
-		*p_base = base;
-	else
-		ttm_base_object_unref(&base);
-	*out = &vmw_user_bo->dma;
-
-	return 0;
-}
-
-int vmw_user_dmabuf_reference(struct ttm_object_file *tfile,
-			      struct vmw_dma_buffer *dma_buf,
-			      uint32_t *handle)
-{
-	struct vmw_user_dma_buffer *user_bo;
-
-	if (dma_buf->base.destroy != vmw_user_dmabuf_destroy)
-		return -EINVAL;
-
-	user_bo = container_of(dma_buf, struct vmw_user_dma_buffer, dma);
-
-	*handle = user_bo->prime.base.hash.key;
-	return ttm_ref_object_add(tfile, &user_bo->prime.base,
-				  TTM_REF_USAGE, NULL, false);
-}
-
-/**
- * vmw_dumb_create - Create a dumb kms buffer
- *
- * @file_priv: Pointer to a struct drm_file identifying the caller.
- * @dev: Pointer to the drm device.
- * @args: Pointer to a struct drm_mode_create_dumb structure
- *
- * This is a driver callback for the core drm create_dumb functionality.
- * Note that this is very similar to the vmw_dmabuf_alloc ioctl, except
- * that the arguments have a different format.
- */
-int vmw_dumb_create(struct drm_file *file_priv,
-		    struct drm_device *dev,
-		    struct drm_mode_create_dumb *args)
-{
-	struct vmw_private *dev_priv = vmw_priv(dev);
-	struct vmw_dma_buffer *dma_buf;
-	int ret;
-
-	args->pitch = args->width * ((args->bpp + 7) / 8);
-	args->size = args->pitch * args->height;
-
-	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
-	if (unlikely(ret != 0))
-		return ret;
-
-	ret = vmw_user_dmabuf_alloc(dev_priv, vmw_fpriv(file_priv)->tfile,
-				    args->size, false, &args->handle,
-				    &dma_buf, NULL);
-	if (unlikely(ret != 0))
-		goto out_no_dmabuf;
-
-	vmw_dmabuf_unreference(&dma_buf);
-out_no_dmabuf:
-	ttm_read_unlock(&dev_priv->reservation_sem);
-	return ret;
-}
-
-/**
- * vmw_dumb_map_offset - Return the address space offset of a dumb buffer
- *
- * @file_priv: Pointer to a struct drm_file identifying the caller.
- * @dev: Pointer to the drm device.
- * @handle: Handle identifying the dumb buffer.
- * @offset: The address space offset returned.
- *
- * This is a driver callback for the core drm dumb_map_offset functionality.
- */
-int vmw_dumb_map_offset(struct drm_file *file_priv,
-			struct drm_device *dev, uint32_t handle,
-			uint64_t *offset)
-{
-	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
-	struct vmw_dma_buffer *out_buf;
-	int ret;
-
-	ret = vmw_user_dmabuf_lookup(tfile, handle, &out_buf, NULL);
-	if (ret != 0)
-		return -EINVAL;
-
-	*offset = drm_vma_node_offset_addr(&out_buf->base.vma_node);
-	vmw_dmabuf_unreference(&out_buf);
-	return 0;
-}
-
-/**
- * vmw_dumb_destroy - Destroy a dumb boffer
- *
- * @file_priv: Pointer to a struct drm_file identifying the caller.
- * @dev: Pointer to the drm device.
- * @handle: Handle identifying the dumb buffer.
- *
- * This is a driver callback for the core drm dumb_destroy functionality.
- */
-int vmw_dumb_destroy(struct drm_file *file_priv,
-		     struct drm_device *dev,
-		     uint32_t handle)
-{
-	return ttm_ref_object_base_unref(vmw_fpriv(file_priv)->tfile,
-					 handle, TTM_REF_USAGE);
-}
-
-/**
  * vmw_resource_buf_alloc - Allocate a backup buffer for a resource.
  *
  * @res:            The resource for which to allocate a backup buffer.
@@ -829,7 +303,7 @@ static int vmw_resource_buf_alloc(struct vmw_resource *res,
 {
 	unsigned long size =
 		(res->backup_size + PAGE_SIZE - 1) & PAGE_MASK;
-	struct vmw_dma_buffer *backup;
+	struct vmw_buffer_object *backup;
 	int ret;
 
 	if (likely(res->backup)) {
@@ -841,16 +315,16 @@ static int vmw_resource_buf_alloc(struct vmw_resource *res,
 	if (unlikely(!backup))
 		return -ENOMEM;
 
-	ret = vmw_dmabuf_init(res->dev_priv, backup, res->backup_size,
+	ret = vmw_bo_init(res->dev_priv, backup, res->backup_size,
 			      res->func->backup_placement,
 			      interruptible,
-			      &vmw_dmabuf_bo_free);
+			      &vmw_bo_bo_free);
 	if (unlikely(ret != 0))
-		goto out_no_dmabuf;
+		goto out_no_bo;
 
 	res->backup = backup;
 
-out_no_dmabuf:
+out_no_bo:
 	return ret;
 }
 
@@ -919,7 +393,7 @@ out_bind_failed:
  */
 void vmw_resource_unreserve(struct vmw_resource *res,
 			    bool switch_backup,
-			    struct vmw_dma_buffer *new_backup,
+			    struct vmw_buffer_object *new_backup,
 			    unsigned long new_backup_offset)
 {
 	struct vmw_private *dev_priv = res->dev_priv;
@@ -931,11 +405,11 @@ void vmw_resource_unreserve(struct vmw_resource *res,
 		if (res->backup) {
 			lockdep_assert_held(&res->backup->base.resv->lock.base);
 			list_del_init(&res->mob_head);
-			vmw_dmabuf_unreference(&res->backup);
+			vmw_bo_unreference(&res->backup);
 		}
 
 		if (new_backup) {
-			res->backup = vmw_dmabuf_reference(new_backup);
+			res->backup = vmw_bo_reference(new_backup);
 			lockdep_assert_held(&new_backup->base.resv->lock.base);
 			list_add_tail(&res->mob_head, &new_backup->res_list);
 		} else {
@@ -959,6 +433,7 @@ void vmw_resource_unreserve(struct vmw_resource *res,
  *                             for a resource and in that case, allocate
  *                             one, reserve and validate it.
  *
+ * @ticket:         The ww aqcquire context to use, or NULL if trylocking.
  * @res:            The resource for which to allocate a backup buffer.
  * @interruptible:  Whether any sleeps during allocation should be
  *                  performed while interruptible.
@@ -966,7 +441,8 @@ void vmw_resource_unreserve(struct vmw_resource *res,
  *                  reserved and validated backup buffer.
  */
 static int
-vmw_resource_check_buffer(struct vmw_resource *res,
+vmw_resource_check_buffer(struct ww_acquire_ctx *ticket,
+			  struct vmw_resource *res,
 			  bool interruptible,
 			  struct ttm_validate_buffer *val_buf)
 {
@@ -985,7 +461,7 @@ vmw_resource_check_buffer(struct vmw_resource *res,
 	val_buf->bo = ttm_bo_reference(&res->backup->base);
 	val_buf->shared = false;
 	list_add_tail(&val_buf->head, &val_list);
-	ret = ttm_eu_reserve_buffers(NULL, &val_list, interruptible, NULL);
+	ret = ttm_eu_reserve_buffers(ticket, &val_list, interruptible, NULL);
 	if (unlikely(ret != 0))
 		goto out_no_reserve;
 
@@ -1003,11 +479,11 @@ vmw_resource_check_buffer(struct vmw_resource *res,
 	return 0;
 
 out_no_validate:
-	ttm_eu_backoff_reservation(NULL, &val_list);
+	ttm_eu_backoff_reservation(ticket, &val_list);
 out_no_reserve:
 	ttm_bo_unref(&val_buf->bo);
 	if (backup_dirty)
-		vmw_dmabuf_unreference(&res->backup);
+		vmw_bo_unreference(&res->backup);
 
 	return ret;
 }
@@ -1050,10 +526,12 @@ int vmw_resource_reserve(struct vmw_resource *res, bool interruptible,
  * vmw_resource_backoff_reservation - Unreserve and unreference a
  *                                    backup buffer
  *.
+ * @ticket:         The ww acquire ctx used for reservation.
  * @val_buf:        Backup buffer information.
  */
 static void
-vmw_resource_backoff_reservation(struct ttm_validate_buffer *val_buf)
+vmw_resource_backoff_reservation(struct ww_acquire_ctx *ticket,
+				 struct ttm_validate_buffer *val_buf)
 {
 	struct list_head val_list;
 
@@ -1062,7 +540,7 @@ vmw_resource_backoff_reservation(struct ttm_validate_buffer *val_buf)
 
 	INIT_LIST_HEAD(&val_list);
 	list_add_tail(&val_buf->head, &val_list);
-	ttm_eu_backoff_reservation(NULL, &val_list);
+	ttm_eu_backoff_reservation(ticket, &val_list);
 	ttm_bo_unref(&val_buf->bo);
 }
 
@@ -1070,10 +548,12 @@ vmw_resource_backoff_reservation(struct ttm_validate_buffer *val_buf)
  * vmw_resource_do_evict - Evict a resource, and transfer its data
  *                         to a backup buffer.
  *
+ * @ticket:         The ww acquire ticket to use, or NULL if trylocking.
  * @res:            The resource to evict.
  * @interruptible:  Whether to wait interruptible.
  */
-static int vmw_resource_do_evict(struct vmw_resource *res, bool interruptible)
+static int vmw_resource_do_evict(struct ww_acquire_ctx *ticket,
+				 struct vmw_resource *res, bool interruptible)
 {
 	struct ttm_validate_buffer val_buf;
 	const struct vmw_res_func *func = res->func;
@@ -1083,7 +563,7 @@ static int vmw_resource_do_evict(struct vmw_resource *res, bool interruptible)
 
 	val_buf.bo = NULL;
 	val_buf.shared = false;
-	ret = vmw_resource_check_buffer(res, interruptible, &val_buf);
+	ret = vmw_resource_check_buffer(ticket, res, interruptible, &val_buf);
 	if (unlikely(ret != 0))
 		return ret;
 
@@ -1098,7 +578,7 @@ static int vmw_resource_do_evict(struct vmw_resource *res, bool interruptible)
 	res->backup_dirty = true;
 	res->res_dirty = false;
 out_no_unbind:
-	vmw_resource_backoff_reservation(&val_buf);
+	vmw_resource_backoff_reservation(ticket, &val_buf);
 
 	return ret;
 }
@@ -1152,7 +632,8 @@ int vmw_resource_validate(struct vmw_resource *res)
 
 		write_unlock(&dev_priv->resource_lock);
 
-		ret = vmw_resource_do_evict(evict_res, true);
+		/* Trylock backup buffers with a NULL ticket. */
+		ret = vmw_resource_do_evict(NULL, evict_res, true);
 		if (unlikely(ret != 0)) {
 			write_lock(&dev_priv->resource_lock);
 			list_add_tail(&evict_res->lru_head, lru_list);
@@ -1171,7 +652,7 @@ int vmw_resource_validate(struct vmw_resource *res)
 		goto out_no_validate;
 	else if (!res->func->needs_backup && res->backup) {
 		list_del_init(&res->mob_head);
-		vmw_dmabuf_unreference(&res->backup);
+		vmw_bo_unreference(&res->backup);
 	}
 
 	return 0;
@@ -1180,109 +661,39 @@ out_no_validate:
 	return ret;
 }
 
-/**
- * vmw_fence_single_bo - Utility function to fence a single TTM buffer
- *                       object without unreserving it.
- *
- * @bo:             Pointer to the struct ttm_buffer_object to fence.
- * @fence:          Pointer to the fence. If NULL, this function will
- *                  insert a fence into the command stream..
- *
- * Contrary to the ttm_eu version of this function, it takes only
- * a single buffer object instead of a list, and it also doesn't
- * unreserve the buffer object, which needs to be done separately.
- */
-void vmw_fence_single_bo(struct ttm_buffer_object *bo,
-			 struct vmw_fence_obj *fence)
-{
-	struct ttm_bo_device *bdev = bo->bdev;
-
-	struct vmw_private *dev_priv =
-		container_of(bdev, struct vmw_private, bdev);
-
-	if (fence == NULL) {
-		vmw_execbuf_fence_commands(NULL, dev_priv, &fence, NULL);
-		reservation_object_add_excl_fence(bo->resv, &fence->base);
-		dma_fence_put(&fence->base);
-	} else
-		reservation_object_add_excl_fence(bo->resv, &fence->base);
-}
 
 /**
- * vmw_resource_move_notify - TTM move_notify_callback
+ * vmw_resource_unbind_list
  *
- * @bo: The TTM buffer object about to move.
- * @mem: The struct ttm_mem_reg indicating to what memory
- *       region the move is taking place.
+ * @vbo: Pointer to the current backing MOB.
  *
  * Evicts the Guest Backed hardware resource if the backup
  * buffer is being moved out of MOB memory.
- * Note that this function should not race with the resource
- * validation code as long as it accesses only members of struct
- * resource that remain static while bo::res is !NULL and
- * while we have @bo reserved. struct resource::backup is *not* a
- * static member. The resource validation code will take care
- * to set @bo::res to NULL, while having @bo reserved when the
- * buffer is no longer bound to the resource, so @bo:res can be
- * used to determine whether there is a need to unbind and whether
- * it is safe to unbind.
+ * Note that this function will not race with the resource
+ * validation code, since resource validation and eviction
+ * both require the backup buffer to be reserved.
  */
-void vmw_resource_move_notify(struct ttm_buffer_object *bo,
-			      struct ttm_mem_reg *mem)
+void vmw_resource_unbind_list(struct vmw_buffer_object *vbo)
 {
-	struct vmw_dma_buffer *dma_buf;
-
-	if (mem == NULL)
-		return;
-
-	if (bo->destroy != vmw_dmabuf_bo_free &&
-	    bo->destroy != vmw_user_dmabuf_destroy)
-		return;
-
-	dma_buf = container_of(bo, struct vmw_dma_buffer, base);
-
-	/*
-	 * Kill any cached kernel maps before move. An optimization could
-	 * be to do this iff source or destination memory type is VRAM.
-	 */
-	vmw_dma_buffer_unmap(dma_buf);
 
-	if (mem->mem_type != VMW_PL_MOB) {
-		struct vmw_resource *res, *n;
-		struct ttm_validate_buffer val_buf;
+	struct vmw_resource *res, *next;
+	struct ttm_validate_buffer val_buf = {
+		.bo = &vbo->base,
+		.shared = false
+	};
 
-		val_buf.bo = bo;
-		val_buf.shared = false;
+	lockdep_assert_held(&vbo->base.resv->lock.base);
+	list_for_each_entry_safe(res, next, &vbo->res_list, mob_head) {
+		if (!res->func->unbind)
+			continue;
 
-		list_for_each_entry_safe(res, n, &dma_buf->res_list, mob_head) {
-
-			if (unlikely(res->func->unbind == NULL))
-				continue;
-
-			(void) res->func->unbind(res, true, &val_buf);
-			res->backup_dirty = true;
-			res->res_dirty = false;
-			list_del_init(&res->mob_head);
-		}
-
-		(void) ttm_bo_wait(bo, false, false);
+		(void) res->func->unbind(res, true, &val_buf);
+		res->backup_dirty = true;
+		res->res_dirty = false;
+		list_del_init(&res->mob_head);
 	}
-}
-
-
-/**
- * vmw_resource_swap_notify - swapout notify callback.
- *
- * @bo: The buffer object to be swapped out.
- */
-void vmw_resource_swap_notify(struct ttm_buffer_object *bo)
-{
-	if (bo->destroy != vmw_dmabuf_bo_free &&
-	    bo->destroy != vmw_user_dmabuf_destroy)
-		return;
 
-	/* Kill any cached kernel maps before swapout */
-	vmw_dma_buffer_unmap(vmw_dma_buffer(bo));
+	(void) ttm_bo_wait(&vbo->base, false, false);
 }
 
 
@@ -1294,7 +705,7 @@ void vmw_resource_swap_notify(struct ttm_buffer_object *bo)
  * Read back cached states from the device if they exist.  This function
  * assumings binding_mutex is held.
  */
-int vmw_query_readback_all(struct vmw_dma_buffer *dx_query_mob)
+int vmw_query_readback_all(struct vmw_buffer_object *dx_query_mob)
 {
 	struct vmw_resource *dx_query_ctx;
 	struct vmw_private *dev_priv;
@@ -1344,7 +755,7 @@ int vmw_query_readback_all(struct vmw_dma_buffer *dx_query_mob)
 void vmw_query_move_notify(struct ttm_buffer_object *bo,
 			   struct ttm_mem_reg *mem)
 {
-	struct vmw_dma_buffer *dx_query_mob;
+	struct vmw_buffer_object *dx_query_mob;
 	struct ttm_bo_device *bdev = bo->bdev;
 	struct vmw_private *dev_priv;
 
@@ -1353,7 +764,7 @@ void vmw_query_move_notify(struct ttm_buffer_object *bo,
 
 	mutex_lock(&dev_priv->binding_mutex);
 
-	dx_query_mob = container_of(bo, struct vmw_dma_buffer, base);
+	dx_query_mob = container_of(bo, struct vmw_buffer_object, base);
 	if (mem == NULL || !dx_query_mob || !dx_query_mob->dx_query_ctx) {
 		mutex_unlock(&dev_priv->binding_mutex);
 		return;
@@ -1368,7 +779,7 @@ void vmw_query_move_notify(struct ttm_buffer_object *bo,
 
 		/* Create a fence and attach the BO to it */
 		(void) vmw_execbuf_fence_commands(NULL, dev_priv, &fence, NULL);
-		vmw_fence_single_bo(bo, fence);
+		vmw_bo_fence_single(bo, fence);
 
 		if (fence != NULL)
 			vmw_fence_obj_unreference(&fence);
@@ -1405,6 +816,7 @@ static void vmw_resource_evict_type(struct vmw_private *dev_priv,
 	struct vmw_resource *evict_res;
 	unsigned err_count = 0;
 	int ret;
+	struct ww_acquire_ctx ticket;
 
 	do {
 		write_lock(&dev_priv->resource_lock);
@@ -1418,7 +830,8 @@ static void vmw_resource_evict_type(struct vmw_private *dev_priv,
 		list_del_init(&evict_res->lru_head);
 		write_unlock(&dev_priv->resource_lock);
 
-		ret = vmw_resource_do_evict(evict_res, false);
+		/* Wait lock backup buffers with a ticket. */
+		ret = vmw_resource_do_evict(&ticket, evict_res, false);
 		if (unlikely(ret != 0)) {
 			write_lock(&dev_priv->resource_lock);
 			list_add_tail(&evict_res->lru_head, lru_list);
@@ -1481,7 +894,7 @@ int vmw_resource_pin(struct vmw_resource *res, bool interruptible)
 		goto out_no_reserve;
 
 	if (res->pin_count == 0) {
-		struct vmw_dma_buffer *vbo = NULL;
+		struct vmw_buffer_object *vbo = NULL;
 
 		if (res->backup) {
 			vbo = res->backup;
@@ -1539,7 +952,7 @@ void vmw_resource_unpin(struct vmw_resource *res)
 
 	WARN_ON(res->pin_count == 0);
 	if (--res->pin_count == 0 && res->backup) {
-		struct vmw_dma_buffer *vbo = res->backup;
+		struct vmw_buffer_object *vbo = res->backup;
 
 		(void) ttm_bo_reserve(&vbo->base, false, false, NULL);
 		vmw_bo_pin_reserved(vbo, false);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h
index ac05968a832b..a8c1c5ebd71d 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource_priv.h
@@ -1,7 +1,7 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
  *
- * Copyright © 2012-2014 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2012-2014 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c b/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c
index 648f8127f65a..ad0de7f0cd60 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_scrn.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2011-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2011-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -66,7 +66,7 @@ struct vmw_kms_sou_readback_blit {
 	SVGAFifoCmdBlitScreenToGMRFB body;
 };
 
-struct vmw_kms_sou_dmabuf_blit {
+struct vmw_kms_sou_bo_blit {
 	uint32 header;
 	SVGAFifoCmdBlitGMRFBToScreen body;
 };
@@ -83,7 +83,7 @@ struct vmw_screen_object_unit {
 	struct vmw_display_unit base;
 
 	unsigned long buffer_size; /**< Size of allocated buffer */
-	struct vmw_dma_buffer *buffer; /**< Backing store buffer */
+	struct vmw_buffer_object *buffer; /**< Backing store buffer */
 
 	bool defined;
 };
@@ -109,7 +109,7 @@ static void vmw_sou_crtc_destroy(struct drm_crtc *crtc)
  */
 static int vmw_sou_fifo_create(struct vmw_private *dev_priv,
 			       struct vmw_screen_object_unit *sou,
-			       uint32_t x, uint32_t y,
+			       int x, int y,
 			       struct drm_display_mode *mode)
 {
 	size_t fifo_size;
@@ -139,13 +139,8 @@ static int vmw_sou_fifo_create(struct vmw_private *dev_priv,
 		(sou->base.unit == 0 ? SVGA_SCREEN_IS_PRIMARY : 0);
 	cmd->obj.size.width = mode->hdisplay;
 	cmd->obj.size.height = mode->vdisplay;
-	if (sou->base.is_implicit) {
-		cmd->obj.root.x = x;
-		cmd->obj.root.y = y;
-	} else {
-		cmd->obj.root.x = sou->base.gui_x;
-		cmd->obj.root.y = sou->base.gui_y;
-	}
+	cmd->obj.root.x = x;
+	cmd->obj.root.y = y;
 	sou->base.set_gui_x = cmd->obj.root.x;
 	sou->base.set_gui_y = cmd->obj.root.y;
 
@@ -222,12 +217,11 @@ static void vmw_sou_crtc_mode_set_nofb(struct drm_crtc *crtc)
 	struct vmw_plane_state *vps;
 	int ret;
 
-
-	sou      = vmw_crtc_to_sou(crtc);
+	sou = vmw_crtc_to_sou(crtc);
 	dev_priv = vmw_priv(crtc->dev);
-	ps       = crtc->primary->state;
-	fb       = ps->fb;
-	vps      = vmw_plane_state_to_vps(ps);
+	ps = crtc->primary->state;
+	fb = ps->fb;
+	vps = vmw_plane_state_to_vps(ps);
 
 	vfb = (fb) ? vmw_framebuffer_to_vfb(fb) : NULL;
 
@@ -240,11 +234,25 @@ static void vmw_sou_crtc_mode_set_nofb(struct drm_crtc *crtc)
 	}
 
 	if (vfb) {
-		sou->buffer = vps->dmabuf;
-		sou->buffer_size = vps->dmabuf_size;
+		struct drm_connector_state *conn_state;
+		struct vmw_connector_state *vmw_conn_state;
+		int x, y;
+
+		sou->buffer = vps->bo;
+		sou->buffer_size = vps->bo_size;
+
+		if (sou->base.is_implicit) {
+			x = crtc->x;
+			y = crtc->y;
+		} else {
+			conn_state = sou->base.connector.state;
+			vmw_conn_state = vmw_connector_state_to_vcs(conn_state);
+
+			x = vmw_conn_state->gui_x;
+			y = vmw_conn_state->gui_y;
+		}
 
-		ret = vmw_sou_fifo_create(dev_priv, sou, crtc->x, crtc->y,
-					  &crtc->mode);
+		ret = vmw_sou_fifo_create(dev_priv, sou, x, y, &crtc->mode);
 		if (ret)
 			DRM_ERROR("Failed to define Screen Object %dx%d\n",
 				  crtc->x, crtc->y);
@@ -408,10 +416,10 @@ vmw_sou_primary_plane_cleanup_fb(struct drm_plane *plane,
 	struct drm_crtc *crtc = plane->state->crtc ?
 		plane->state->crtc : old_state->crtc;
 
-	if (vps->dmabuf)
-		vmw_dmabuf_unpin(vmw_priv(crtc->dev), vps->dmabuf, false);
-	vmw_dmabuf_unreference(&vps->dmabuf);
-	vps->dmabuf_size = 0;
+	if (vps->bo)
+		vmw_bo_unpin(vmw_priv(crtc->dev), vps->bo, false);
+	vmw_bo_unreference(&vps->bo);
+	vps->bo_size = 0;
 
 	vmw_du_plane_cleanup_fb(plane, old_state);
 }
@@ -440,8 +448,8 @@ vmw_sou_primary_plane_prepare_fb(struct drm_plane *plane,
 
 
 	if (!new_fb) {
-		vmw_dmabuf_unreference(&vps->dmabuf);
-		vps->dmabuf_size = 0;
+		vmw_bo_unreference(&vps->bo);
+		vps->bo_size = 0;
 
 		return 0;
 	}
@@ -449,22 +457,22 @@ vmw_sou_primary_plane_prepare_fb(struct drm_plane *plane,
 	size = new_state->crtc_w * new_state->crtc_h * 4;
 	dev_priv = vmw_priv(crtc->dev);
 
-	if (vps->dmabuf) {
-		if (vps->dmabuf_size == size) {
+	if (vps->bo) {
+		if (vps->bo_size == size) {
 			/*
 			 * Note that this might temporarily up the pin-count
 			 * to 2, until cleanup_fb() is called.
 			 */
-			return vmw_dmabuf_pin_in_vram(dev_priv, vps->dmabuf,
+			return vmw_bo_pin_in_vram(dev_priv, vps->bo,
 						      true);
 		}
 
-		vmw_dmabuf_unreference(&vps->dmabuf);
-		vps->dmabuf_size = 0;
+		vmw_bo_unreference(&vps->bo);
+		vps->bo_size = 0;
 	}
 
-	vps->dmabuf = kzalloc(sizeof(*vps->dmabuf), GFP_KERNEL);
-	if (!vps->dmabuf)
+	vps->bo = kzalloc(sizeof(*vps->bo), GFP_KERNEL);
+	if (!vps->bo)
 		return -ENOMEM;
 
 	vmw_svga_enable(dev_priv);
@@ -473,20 +481,22 @@ vmw_sou_primary_plane_prepare_fb(struct drm_plane *plane,
 	 * resume the overlays, this is preferred to failing to alloc.
 	 */
 	vmw_overlay_pause_all(dev_priv);
-	ret = vmw_dmabuf_init(dev_priv, vps->dmabuf, size,
+	ret = vmw_bo_init(dev_priv, vps->bo, size,
 			      &vmw_vram_ne_placement,
-			      false, &vmw_dmabuf_bo_free);
+			      false, &vmw_bo_bo_free);
 	vmw_overlay_resume_all(dev_priv);
 	if (ret) {
-		vps->dmabuf = NULL; /* vmw_dmabuf_init frees on error */
+		vps->bo = NULL; /* vmw_bo_init frees on error */
 		return ret;
 	}
 
+	vps->bo_size = size;
+
 	/*
 	 * TTM already thinks the buffer is pinned, but make sure the
 	 * pin_count is upped.
 	 */
-	return vmw_dmabuf_pin_in_vram(dev_priv, vps->dmabuf, true);
+	return vmw_bo_pin_in_vram(dev_priv, vps->bo, true);
 }
 
 
@@ -510,10 +520,10 @@ vmw_sou_primary_plane_atomic_update(struct drm_plane *plane,
 		vclips.w = crtc->mode.hdisplay;
 		vclips.h = crtc->mode.vdisplay;
 
-		if (vfb->dmabuf)
-			ret = vmw_kms_sou_do_dmabuf_dirty(dev_priv, vfb, NULL,
-							  &vclips, 1, 1, true,
-							  &fence, crtc);
+		if (vfb->bo)
+			ret = vmw_kms_sou_do_bo_dirty(dev_priv, vfb, NULL,
+						      &vclips, 1, 1, true,
+						      &fence, crtc);
 		else
 			ret = vmw_kms_sou_do_surface_dirty(dev_priv, vfb, NULL,
 							   &vclips, NULL, 0, 0,
@@ -525,8 +535,6 @@ vmw_sou_primary_plane_atomic_update(struct drm_plane *plane,
 		 */
 		if (ret != 0)
 			DRM_ERROR("Failed to update screen.\n");
-
-		crtc->primary->fb = plane->state->fb;
 	} else {
 		/*
 		 * When disabling a plane, CRTC and FB should always be NULL
@@ -695,7 +703,7 @@ static int vmw_sou_init(struct vmw_private *dev_priv, unsigned unit)
 		goto err_free_connector;
 	}
 
-	(void) drm_mode_connector_attach_encoder(connector, encoder);
+	(void) drm_connector_attach_encoder(connector, encoder);
 	encoder->possible_crtcs = (1 << unit);
 	encoder->possible_clones = 0;
 
@@ -775,11 +783,11 @@ int vmw_kms_sou_init_display(struct vmw_private *dev_priv)
 	return 0;
 }
 
-static int do_dmabuf_define_gmrfb(struct vmw_private *dev_priv,
+static int do_bo_define_gmrfb(struct vmw_private *dev_priv,
 				  struct vmw_framebuffer *framebuffer)
 {
-	struct vmw_dma_buffer *buf =
-		container_of(framebuffer, struct vmw_framebuffer_dmabuf,
+	struct vmw_buffer_object *buf =
+		container_of(framebuffer, struct vmw_framebuffer_bo,
 			     base)->buffer;
 	int depth = framebuffer->base.format->depth;
 	struct {
@@ -970,13 +978,13 @@ int vmw_kms_sou_do_surface_dirty(struct vmw_private *dev_priv,
 }
 
 /**
- * vmw_sou_dmabuf_fifo_commit - Callback to submit a set of readback clips.
+ * vmw_sou_bo_fifo_commit - Callback to submit a set of readback clips.
  *
  * @dirty: The closure structure.
  *
  * Commits a previously built command buffer of readback clips.
  */
-static void vmw_sou_dmabuf_fifo_commit(struct vmw_kms_dirty *dirty)
+static void vmw_sou_bo_fifo_commit(struct vmw_kms_dirty *dirty)
 {
 	if (!dirty->num_hits) {
 		vmw_fifo_commit(dirty->dev_priv, 0);
@@ -984,20 +992,20 @@ static void vmw_sou_dmabuf_fifo_commit(struct vmw_kms_dirty *dirty)
 	}
 
 	vmw_fifo_commit(dirty->dev_priv,
-			sizeof(struct vmw_kms_sou_dmabuf_blit) *
+			sizeof(struct vmw_kms_sou_bo_blit) *
 			dirty->num_hits);
 }
 
 /**
- * vmw_sou_dmabuf_clip - Callback to encode a readback cliprect.
+ * vmw_sou_bo_clip - Callback to encode a readback cliprect.
  *
  * @dirty: The closure structure
  *
  * Encodes a BLIT_GMRFB_TO_SCREEN cliprect.
  */
-static void vmw_sou_dmabuf_clip(struct vmw_kms_dirty *dirty)
+static void vmw_sou_bo_clip(struct vmw_kms_dirty *dirty)
 {
-	struct vmw_kms_sou_dmabuf_blit *blit = dirty->cmd;
+	struct vmw_kms_sou_bo_blit *blit = dirty->cmd;
 
 	blit += dirty->num_hits;
 	blit->header = SVGA_CMD_BLIT_GMRFB_TO_SCREEN;
@@ -1012,10 +1020,10 @@ static void vmw_sou_dmabuf_clip(struct vmw_kms_dirty *dirty)
 }
 
 /**
- * vmw_kms_do_dmabuf_dirty - Dirty part of a dma-buffer backed framebuffer
+ * vmw_kms_do_bo_dirty - Dirty part of a buffer-object backed framebuffer
  *
  * @dev_priv: Pointer to the device private structure.
- * @framebuffer: Pointer to the dma-buffer backed framebuffer.
+ * @framebuffer: Pointer to the buffer-object backed framebuffer.
  * @clips: Array of clip rects.
  * @vclips: Alternate array of clip rects. Either @clips or @vclips must
  * be NULL.
@@ -1025,12 +1033,12 @@ static void vmw_sou_dmabuf_clip(struct vmw_kms_dirty *dirty)
  * @out_fence: If non-NULL, will return a ref-counted pointer to a
  * struct vmw_fence_obj. The returned fence pointer may be NULL in which
  * case the device has already synchronized.
- * @crtc: If crtc is passed, perform dmabuf dirty on that crtc only.
+ * @crtc: If crtc is passed, perform bo dirty on that crtc only.
  *
  * Returns 0 on success, negative error code on failure. -ERESTARTSYS if
  * interrupted.
  */
-int vmw_kms_sou_do_dmabuf_dirty(struct vmw_private *dev_priv,
+int vmw_kms_sou_do_bo_dirty(struct vmw_private *dev_priv,
 				struct vmw_framebuffer *framebuffer,
 				struct drm_clip_rect *clips,
 				struct drm_vmw_rect *vclips,
@@ -1039,8 +1047,8 @@ int vmw_kms_sou_do_dmabuf_dirty(struct vmw_private *dev_priv,
 				struct vmw_fence_obj **out_fence,
 				struct drm_crtc *crtc)
 {
-	struct vmw_dma_buffer *buf =
-		container_of(framebuffer, struct vmw_framebuffer_dmabuf,
+	struct vmw_buffer_object *buf =
+		container_of(framebuffer, struct vmw_framebuffer_bo,
 			     base)->buffer;
 	struct vmw_kms_dirty dirty;
 	int ret;
@@ -1050,14 +1058,14 @@ int vmw_kms_sou_do_dmabuf_dirty(struct vmw_private *dev_priv,
 	if (ret)
 		return ret;
 
-	ret = do_dmabuf_define_gmrfb(dev_priv, framebuffer);
+	ret = do_bo_define_gmrfb(dev_priv, framebuffer);
 	if (unlikely(ret != 0))
 		goto out_revert;
 
 	dirty.crtc = crtc;
-	dirty.fifo_commit = vmw_sou_dmabuf_fifo_commit;
-	dirty.clip = vmw_sou_dmabuf_clip;
-	dirty.fifo_reserve_size = sizeof(struct vmw_kms_sou_dmabuf_blit) *
+	dirty.fifo_commit = vmw_sou_bo_fifo_commit;
+	dirty.clip = vmw_sou_bo_clip;
+	dirty.fifo_reserve_size = sizeof(struct vmw_kms_sou_bo_blit) *
 		num_clips;
 	ret = vmw_kms_helper_dirty(dev_priv, framebuffer, clips, vclips,
 				   0, 0, num_clips, increment, &dirty);
@@ -1116,12 +1124,12 @@ static void vmw_sou_readback_clip(struct vmw_kms_dirty *dirty)
 
 /**
  * vmw_kms_sou_readback - Perform a readback from the screen object system to
- * a dma-buffer backed framebuffer.
+ * a buffer-object backed framebuffer.
  *
  * @dev_priv: Pointer to the device private structure.
  * @file_priv: Pointer to a struct drm_file identifying the caller.
  * Must be set to NULL if @user_fence_rep is NULL.
- * @vfb: Pointer to the dma-buffer backed framebuffer.
+ * @vfb: Pointer to the buffer-object backed framebuffer.
  * @user_fence_rep: User-space provided structure for fence information.
  * Must be set to non-NULL if @file_priv is non-NULL.
  * @vclips: Array of clip rects.
@@ -1139,8 +1147,8 @@ int vmw_kms_sou_readback(struct vmw_private *dev_priv,
 			 uint32_t num_clips,
 			 struct drm_crtc *crtc)
 {
-	struct vmw_dma_buffer *buf =
-		container_of(vfb, struct vmw_framebuffer_dmabuf, base)->buffer;
+	struct vmw_buffer_object *buf =
+		container_of(vfb, struct vmw_framebuffer_bo, base)->buffer;
 	struct vmw_kms_dirty dirty;
 	int ret;
 
@@ -1149,7 +1157,7 @@ int vmw_kms_sou_readback(struct vmw_private *dev_priv,
 	if (ret)
 		return ret;
 
-	ret = do_dmabuf_define_gmrfb(dev_priv, vfb);
+	ret = do_bo_define_gmrfb(dev_priv, vfb);
 	if (unlikely(ret != 0))
 		goto out_revert;
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c b/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c
index 73b8e9a16368..fe4842ca3b6e 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_shader.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -159,7 +159,7 @@ static int vmw_gb_shader_init(struct vmw_private *dev_priv,
 			      SVGA3dShaderType type,
 			      uint8_t num_input_sig,
 			      uint8_t num_output_sig,
-			      struct vmw_dma_buffer *byte_code,
+			      struct vmw_buffer_object *byte_code,
 			      void (*res_free) (struct vmw_resource *res))
 {
 	struct vmw_shader *shader = vmw_res_to_shader(res);
@@ -178,7 +178,7 @@ static int vmw_gb_shader_init(struct vmw_private *dev_priv,
 
 	res->backup_size = size;
 	if (byte_code) {
-		res->backup = vmw_dmabuf_reference(byte_code);
+		res->backup = vmw_bo_reference(byte_code);
 		res->backup_offset = offset;
 	}
 	shader->size = size;
@@ -306,7 +306,7 @@ static int vmw_gb_shader_unbind(struct vmw_resource *res,
 	(void) vmw_execbuf_fence_commands(NULL, dev_priv,
 					  &fence, NULL);
 
-	vmw_fence_single_bo(val_buf->bo, fence);
+	vmw_bo_fence_single(val_buf->bo, fence);
 
 	if (likely(fence != NULL))
 		vmw_fence_obj_unreference(&fence);
@@ -537,7 +537,7 @@ static int vmw_dx_shader_unbind(struct vmw_resource *res,
 
 	(void) vmw_execbuf_fence_commands(NULL, dev_priv,
 					  &fence, NULL);
-	vmw_fence_single_bo(val_buf->bo, fence);
+	vmw_bo_fence_single(val_buf->bo, fence);
 
 	if (likely(fence != NULL))
 		vmw_fence_obj_unreference(&fence);
@@ -723,7 +723,7 @@ int vmw_shader_destroy_ioctl(struct drm_device *dev, void *data,
 }
 
 static int vmw_user_shader_alloc(struct vmw_private *dev_priv,
-				 struct vmw_dma_buffer *buffer,
+				 struct vmw_buffer_object *buffer,
 				 size_t shader_size,
 				 size_t offset,
 				 SVGA3dShaderType shader_type,
@@ -801,7 +801,7 @@ out:
 
 
 static struct vmw_resource *vmw_shader_alloc(struct vmw_private *dev_priv,
-					     struct vmw_dma_buffer *buffer,
+					     struct vmw_buffer_object *buffer,
 					     size_t shader_size,
 					     size_t offset,
 					     SVGA3dShaderType shader_type)
@@ -862,12 +862,12 @@ static int vmw_shader_define(struct drm_device *dev, struct drm_file *file_priv,
 {
 	struct vmw_private *dev_priv = vmw_priv(dev);
 	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
-	struct vmw_dma_buffer *buffer = NULL;
+	struct vmw_buffer_object *buffer = NULL;
 	SVGA3dShaderType shader_type;
 	int ret;
 
 	if (buffer_handle != SVGA3D_INVALID_ID) {
-		ret = vmw_user_dmabuf_lookup(tfile, buffer_handle,
+		ret = vmw_user_bo_lookup(tfile, buffer_handle,
 					     &buffer, NULL);
 		if (unlikely(ret != 0)) {
 			DRM_ERROR("Could not find buffer for shader "
@@ -906,7 +906,7 @@ static int vmw_shader_define(struct drm_device *dev, struct drm_file *file_priv,
 
 	ttm_read_unlock(&dev_priv->reservation_sem);
 out_bad_arg:
-	vmw_dmabuf_unreference(&buffer);
+	vmw_bo_unreference(&buffer);
 	return ret;
 }
 
@@ -983,7 +983,7 @@ int vmw_compat_shader_add(struct vmw_private *dev_priv,
 			  struct list_head *list)
 {
 	struct ttm_operation_ctx ctx = { false, true };
-	struct vmw_dma_buffer *buf;
+	struct vmw_buffer_object *buf;
 	struct ttm_bo_kmap_obj map;
 	bool is_iomem;
 	int ret;
@@ -997,8 +997,8 @@ int vmw_compat_shader_add(struct vmw_private *dev_priv,
 	if (unlikely(!buf))
 		return -ENOMEM;
 
-	ret = vmw_dmabuf_init(dev_priv, buf, size, &vmw_sys_ne_placement,
-			      true, vmw_dmabuf_bo_free);
+	ret = vmw_bo_init(dev_priv, buf, size, &vmw_sys_ne_placement,
+			      true, vmw_bo_bo_free);
 	if (unlikely(ret != 0))
 		goto out;
 
@@ -1031,7 +1031,7 @@ int vmw_compat_shader_add(struct vmw_private *dev_priv,
 				 res, list);
 	vmw_resource_unreference(&res);
 no_reserve:
-	vmw_dmabuf_unreference(&buf);
+	vmw_bo_unreference(&buf);
 out:
 	return ret;
 }
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_simple_resource.c b/drivers/gpu/drm/vmwgfx/vmwgfx_simple_resource.c
index a0cb310665cc..6ebc5affde14 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_simple_resource.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_simple_resource.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2016 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2016 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_so.c b/drivers/gpu/drm/vmwgfx/vmwgfx_so.c
index d3573c37c436..e9b6b7baa009 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_so.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_so.c
@@ -1,6 +1,6 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
- * Copyright © 2014-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2014-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_so.h b/drivers/gpu/drm/vmwgfx/vmwgfx_so.h
index 268738387b5e..b80c7252f2fd 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_so.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_so.h
@@ -1,6 +1,6 @@
+/* SPDX-License-Identifier: GPL-2.0 OR MIT */
 /**************************************************************************
- * Copyright © 2014-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2014-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c b/drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c
index 67331f01ef32..93f6b96ca7bb 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_stdu.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /******************************************************************************
  *
- * COPYRIGHT © 2014-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * COPYRIGHT (C) 2014-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -44,7 +44,7 @@
 enum stdu_content_type {
 	SAME_AS_DISPLAY = 0,
 	SEPARATE_SURFACE,
-	SEPARATE_DMA
+	SEPARATE_BO
 };
 
 /**
@@ -58,7 +58,7 @@ enum stdu_content_type {
  * @bottom: Bottom side of bounding box.
  * @fb_left: Left side of the framebuffer/content bounding box
  * @fb_top: Top of the framebuffer/content bounding box
- * @buf: DMA buffer when DMA-ing between buffer and screen targets.
+ * @buf: buffer object when DMA-ing between buffer and screen targets.
  * @sid: Surface ID when copying between surface and screen targets.
  */
 struct vmw_stdu_dirty {
@@ -68,7 +68,7 @@ struct vmw_stdu_dirty {
 	s32 fb_left, fb_top;
 	u32 pitch;
 	union {
-		struct vmw_dma_buffer *buf;
+		struct vmw_buffer_object *buf;
 		u32 sid;
 	};
 };
@@ -178,13 +178,9 @@ static int vmw_stdu_define_st(struct vmw_private *dev_priv,
 	cmd->body.height = mode->vdisplay;
 	cmd->body.flags  = (0 == cmd->body.stid) ? SVGA_STFLAG_PRIMARY : 0;
 	cmd->body.dpi    = 0;
-	if (stdu->base.is_implicit) {
-		cmd->body.xRoot  = crtc_x;
-		cmd->body.yRoot  = crtc_y;
-	} else {
-		cmd->body.xRoot  = stdu->base.gui_x;
-		cmd->body.yRoot  = stdu->base.gui_y;
-	}
+	cmd->body.xRoot  = crtc_x;
+	cmd->body.yRoot  = crtc_y;
+
 	stdu->base.set_gui_x = cmd->body.xRoot;
 	stdu->base.set_gui_y = cmd->body.yRoot;
 
@@ -374,11 +370,14 @@ static void vmw_stdu_crtc_mode_set_nofb(struct drm_crtc *crtc)
 {
 	struct vmw_private *dev_priv;
 	struct vmw_screen_target_display_unit *stdu;
-	int ret;
-
+	struct drm_connector_state *conn_state;
+	struct vmw_connector_state *vmw_conn_state;
+	int x, y, ret;
 
-	stdu     = vmw_crtc_to_stdu(crtc);
+	stdu = vmw_crtc_to_stdu(crtc);
 	dev_priv = vmw_priv(crtc->dev);
+	conn_state = stdu->base.connector.state;
+	vmw_conn_state = vmw_connector_state_to_vcs(conn_state);
 
 	if (stdu->defined) {
 		ret = vmw_stdu_bind_st(dev_priv, stdu, NULL);
@@ -397,8 +396,16 @@ static void vmw_stdu_crtc_mode_set_nofb(struct drm_crtc *crtc)
 	if (!crtc->state->enable)
 		return;
 
+	if (stdu->base.is_implicit) {
+		x = crtc->x;
+		y = crtc->y;
+	} else {
+		x = vmw_conn_state->gui_x;
+		y = vmw_conn_state->gui_y;
+	}
+
 	vmw_svga_enable(dev_priv);
-	ret = vmw_stdu_define_st(dev_priv, stdu, &crtc->mode, crtc->x, crtc->y);
+	ret = vmw_stdu_define_st(dev_priv, stdu, &crtc->mode, x, y);
 
 	if (ret)
 		DRM_ERROR("Failed to define Screen Target of size %dx%d\n",
@@ -414,6 +421,7 @@ static void vmw_stdu_crtc_helper_prepare(struct drm_crtc *crtc)
 static void vmw_stdu_crtc_atomic_enable(struct drm_crtc *crtc,
 					struct drm_crtc_state *old_state)
 {
+	struct drm_plane_state *plane_state = crtc->primary->state;
 	struct vmw_private *dev_priv;
 	struct vmw_screen_target_display_unit *stdu;
 	struct vmw_framebuffer *vfb;
@@ -422,7 +430,7 @@ static void vmw_stdu_crtc_atomic_enable(struct drm_crtc *crtc,
 
 	stdu     = vmw_crtc_to_stdu(crtc);
 	dev_priv = vmw_priv(crtc->dev);
-	fb       = crtc->primary->fb;
+	fb       = plane_state->fb;
 
 	vfb = (fb) ? vmw_framebuffer_to_vfb(fb) : NULL;
 
@@ -507,14 +515,14 @@ static int vmw_stdu_crtc_page_flip(struct drm_crtc *crtc,
 
 
 /**
- * vmw_stdu_dmabuf_clip - Callback to encode a suface DMA command cliprect
+ * vmw_stdu_bo_clip - Callback to encode a suface DMA command cliprect
  *
  * @dirty: The closure structure.
  *
  * Encodes a surface DMA command cliprect and updates the bounding box
  * for the DMA.
  */
-static void vmw_stdu_dmabuf_clip(struct vmw_kms_dirty *dirty)
+static void vmw_stdu_bo_clip(struct vmw_kms_dirty *dirty)
 {
 	struct vmw_stdu_dirty *ddirty =
 		container_of(dirty, struct vmw_stdu_dirty, base);
@@ -542,14 +550,14 @@ static void vmw_stdu_dmabuf_clip(struct vmw_kms_dirty *dirty)
 }
 
 /**
- * vmw_stdu_dmabuf_fifo_commit - Callback to fill in and submit a DMA command.
+ * vmw_stdu_bo_fifo_commit - Callback to fill in and submit a DMA command.
  *
  * @dirty: The closure structure.
  *
  * Fills in the missing fields in a DMA command, and optionally encodes
  * a screen target update command, depending on transfer direction.
  */
-static void vmw_stdu_dmabuf_fifo_commit(struct vmw_kms_dirty *dirty)
+static void vmw_stdu_bo_fifo_commit(struct vmw_kms_dirty *dirty)
 {
 	struct vmw_stdu_dirty *ddirty =
 		container_of(dirty, struct vmw_stdu_dirty, base);
@@ -593,13 +601,13 @@ static void vmw_stdu_dmabuf_fifo_commit(struct vmw_kms_dirty *dirty)
 
 
 /**
- * vmw_stdu_dmabuf_cpu_clip - Callback to encode a CPU blit
+ * vmw_stdu_bo_cpu_clip - Callback to encode a CPU blit
  *
  * @dirty: The closure structure.
  *
  * This function calculates the bounding box for all the incoming clips.
  */
-static void vmw_stdu_dmabuf_cpu_clip(struct vmw_kms_dirty *dirty)
+static void vmw_stdu_bo_cpu_clip(struct vmw_kms_dirty *dirty)
 {
 	struct vmw_stdu_dirty *ddirty =
 		container_of(dirty, struct vmw_stdu_dirty, base);
@@ -623,14 +631,14 @@ static void vmw_stdu_dmabuf_cpu_clip(struct vmw_kms_dirty *dirty)
 
 
 /**
- * vmw_stdu_dmabuf_cpu_commit - Callback to do a CPU blit from DMAbuf
+ * vmw_stdu_bo_cpu_commit - Callback to do a CPU blit from buffer object
  *
  * @dirty: The closure structure.
  *
  * For the special case when we cannot create a proxy surface in a
  * 2D VM, we have to do a CPU blit ourselves.
  */
-static void vmw_stdu_dmabuf_cpu_commit(struct vmw_kms_dirty *dirty)
+static void vmw_stdu_bo_cpu_commit(struct vmw_kms_dirty *dirty)
 {
 	struct vmw_stdu_dirty *ddirty =
 		container_of(dirty, struct vmw_stdu_dirty, base);
@@ -651,7 +659,7 @@ static void vmw_stdu_dmabuf_cpu_commit(struct vmw_kms_dirty *dirty)
 	if (width == 0 || height == 0)
 		return;
 
-	/* Assume we are blitting from Guest (dmabuf) to Host (display_srf) */
+	/* Assume we are blitting from Guest (bo) to Host (display_srf) */
 	dst_pitch = stdu->display_srf->base_size.width * stdu->cpp;
 	dst_bo = &stdu->display_srf->res.backup->base;
 	dst_offset = ddirty->top * dst_pitch + ddirty->left * stdu->cpp;
@@ -711,13 +719,13 @@ out_cleanup:
 }
 
 /**
- * vmw_kms_stdu_dma - Perform a DMA transfer between a dma-buffer backed
+ * vmw_kms_stdu_dma - Perform a DMA transfer between a buffer-object backed
  * framebuffer and the screen target system.
  *
  * @dev_priv: Pointer to the device private structure.
  * @file_priv: Pointer to a struct drm-file identifying the caller. May be
  * set to NULL, but then @user_fence_rep must also be set to NULL.
- * @vfb: Pointer to the dma-buffer backed framebuffer.
+ * @vfb: Pointer to the buffer-object backed framebuffer.
  * @clips: Array of clip rects. Either @clips or @vclips must be NULL.
  * @vclips: Alternate array of clip rects. Either @clips or @vclips must
  * be NULL.
@@ -746,8 +754,8 @@ int vmw_kms_stdu_dma(struct vmw_private *dev_priv,
 		     bool interruptible,
 		     struct drm_crtc *crtc)
 {
-	struct vmw_dma_buffer *buf =
-		container_of(vfb, struct vmw_framebuffer_dmabuf, base)->buffer;
+	struct vmw_buffer_object *buf =
+		container_of(vfb, struct vmw_framebuffer_bo, base)->buffer;
 	struct vmw_stdu_dirty ddirty;
 	int ret;
 	bool cpu_blit = !(dev_priv->capabilities & SVGA_CAP_3D);
@@ -769,8 +777,8 @@ int vmw_kms_stdu_dma(struct vmw_private *dev_priv,
 	ddirty.fb_left = ddirty.fb_top = S32_MAX;
 	ddirty.pitch = vfb->base.pitches[0];
 	ddirty.buf = buf;
-	ddirty.base.fifo_commit = vmw_stdu_dmabuf_fifo_commit;
-	ddirty.base.clip = vmw_stdu_dmabuf_clip;
+	ddirty.base.fifo_commit = vmw_stdu_bo_fifo_commit;
+	ddirty.base.clip = vmw_stdu_bo_clip;
 	ddirty.base.fifo_reserve_size = sizeof(struct vmw_stdu_dma) +
 		num_clips * sizeof(SVGA3dCopyBox) +
 		sizeof(SVGA3dCmdSurfaceDMASuffix);
@@ -779,8 +787,8 @@ int vmw_kms_stdu_dma(struct vmw_private *dev_priv,
 
 
 	if (cpu_blit) {
-		ddirty.base.fifo_commit = vmw_stdu_dmabuf_cpu_commit;
-		ddirty.base.clip = vmw_stdu_dmabuf_cpu_clip;
+		ddirty.base.fifo_commit = vmw_stdu_bo_cpu_commit;
+		ddirty.base.clip = vmw_stdu_bo_cpu_clip;
 		ddirty.base.fifo_reserve_size = 0;
 	}
 
@@ -926,7 +934,7 @@ int vmw_kms_stdu_surface_dirty(struct vmw_private *dev_priv,
 	if (ret)
 		return ret;
 
-	if (vfbs->is_dmabuf_proxy) {
+	if (vfbs->is_bo_proxy) {
 		ret = vmw_kms_update_proxy(srf, clips, num_clips, inc);
 		if (ret)
 			goto out_finish;
@@ -1074,7 +1082,7 @@ vmw_stdu_primary_plane_cleanup_fb(struct drm_plane *plane,
  * @new_state: info on the new plane state, including the FB
  *
  * This function allocates a new display surface if the content is
- * backed by a DMA.  The display surface is pinned here, and it'll
+ * backed by a buffer object.  The display surface is pinned here, and it'll
  * be unpinned in .cleanup_fb()
  *
  * Returns 0 on success
@@ -1104,13 +1112,13 @@ vmw_stdu_primary_plane_prepare_fb(struct drm_plane *plane,
 	}
 
 	vfb = vmw_framebuffer_to_vfb(new_fb);
-	new_vfbs = (vfb->dmabuf) ? NULL : vmw_framebuffer_to_vfbs(new_fb);
+	new_vfbs = (vfb->bo) ? NULL : vmw_framebuffer_to_vfbs(new_fb);
 
 	if (new_vfbs && new_vfbs->surface->base_size.width == hdisplay &&
 	    new_vfbs->surface->base_size.height == vdisplay)
 		new_content_type = SAME_AS_DISPLAY;
-	else if (vfb->dmabuf)
-		new_content_type = SEPARATE_DMA;
+	else if (vfb->bo)
+		new_content_type = SEPARATE_BO;
 	else
 		new_content_type = SEPARATE_SURFACE;
 
@@ -1123,10 +1131,10 @@ vmw_stdu_primary_plane_prepare_fb(struct drm_plane *plane,
 		display_base_size.depth  = 1;
 
 		/*
-		 * If content buffer is a DMA buf, then we have to construct
-		 * surface info
+		 * If content buffer is a buffer object, then we have to
+		 * construct surface info
 		 */
-		if (new_content_type == SEPARATE_DMA) {
+		if (new_content_type == SEPARATE_BO) {
 
 			switch (new_fb->format->cpp[0]*8) {
 			case 32:
@@ -1149,6 +1157,9 @@ vmw_stdu_primary_plane_prepare_fb(struct drm_plane *plane,
 			content_srf.flags             = 0;
 			content_srf.mip_levels[0]     = 1;
 			content_srf.multisample_count = 0;
+			content_srf.multisample_pattern =
+				SVGA3D_MS_PATTERN_NONE;
+			content_srf.quality_level = SVGA3D_MS_QUALITY_NONE;
 		} else {
 			content_srf = *new_vfbs->surface;
 		}
@@ -1177,6 +1188,8 @@ vmw_stdu_primary_plane_prepare_fb(struct drm_plane *plane,
 				 content_srf.multisample_count,
 				 0,
 				 display_base_size,
+				 content_srf.multisample_pattern,
+				 content_srf.quality_level,
 				 &vps->surf);
 			if (ret != 0) {
 				DRM_ERROR("Couldn't allocate STDU surface.\n");
@@ -1211,12 +1224,12 @@ vmw_stdu_primary_plane_prepare_fb(struct drm_plane *plane,
 	vps->content_fb_type = new_content_type;
 
 	/*
-	 * This should only happen if the DMA buf is too large to create a
+	 * This should only happen if the buffer object is too large to create a
 	 * proxy surface for.
-	 * If we are a 2D VM with a DMA buffer then we have to use CPU blit
+	 * If we are a 2D VM with a buffer object then we have to use CPU blit
 	 * so cache these mappings
 	 */
-	if (vps->content_fb_type == SEPARATE_DMA &&
+	if (vps->content_fb_type == SEPARATE_BO &&
 	    !(dev_priv->capabilities & SVGA_CAP_3D))
 		vps->cpp = new_fb->pitches[0] / new_fb->width;
 
@@ -1275,7 +1288,7 @@ vmw_stdu_primary_plane_atomic_update(struct drm_plane *plane,
 		if (ret)
 			DRM_ERROR("Failed to bind surface to STDU.\n");
 
-		if (vfb->dmabuf)
+		if (vfb->bo)
 			ret = vmw_kms_stdu_dma(dev_priv, NULL, vfb, NULL, NULL,
 					       &vclips, 1, 1, true, false,
 					       crtc);
@@ -1285,8 +1298,6 @@ vmw_stdu_primary_plane_atomic_update(struct drm_plane *plane,
 							 1, 1, NULL, crtc);
 		if (ret)
 			DRM_ERROR("Failed to update STDU.\n");
-
-		crtc->primary->fb = plane->state->fb;
 	} else {
 		crtc = old_state->crtc;
 		stdu = vmw_crtc_to_stdu(crtc);
@@ -1487,7 +1498,7 @@ static int vmw_stdu_init(struct vmw_private *dev_priv, unsigned unit)
 		goto err_free_connector;
 	}
 
-	(void) drm_mode_connector_attach_encoder(connector, encoder);
+	(void) drm_connector_attach_encoder(connector, encoder);
 	encoder->possible_crtcs = (1 << unit);
 	encoder->possible_clones = 0;
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
index b236c48bf265..e125233e074b 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_surface.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -33,6 +33,10 @@
 #include "vmwgfx_binding.h"
 #include "device_include/svga3d_surfacedefs.h"
 
+#define SVGA3D_FLAGS_64(upper32, lower32) (((uint64_t)upper32 << 32) | lower32)
+#define SVGA3D_FLAGS_UPPER_32(svga3d_flags) (svga3d_flags >> 32)
+#define SVGA3D_FLAGS_LOWER_32(svga3d_flags) \
+	(svga3d_flags & ((uint64_t)U32_MAX))
 
 /**
  * struct vmw_user_surface - User-space visible surface resource
@@ -81,7 +85,16 @@ static int vmw_gb_surface_unbind(struct vmw_resource *res,
 				 bool readback,
 				 struct ttm_validate_buffer *val_buf);
 static int vmw_gb_surface_destroy(struct vmw_resource *res);
-
+static int
+vmw_gb_surface_define_internal(struct drm_device *dev,
+			       struct drm_vmw_gb_surface_create_ext_req *req,
+			       struct drm_vmw_gb_surface_create_rep *rep,
+			       struct drm_file *file_priv);
+static int
+vmw_gb_surface_reference_internal(struct drm_device *dev,
+				  struct drm_vmw_surface_arg *req,
+				  struct drm_vmw_gb_surface_ref_ext_rep *rep,
+				  struct drm_file *file_priv);
 
 static const struct vmw_user_resource_conv user_surface_conv = {
 	.object_type = VMW_RES_SURFACE,
@@ -224,7 +237,12 @@ static void vmw_surface_define_encode(const struct vmw_surface *srf,
 	cmd->header.id = SVGA_3D_CMD_SURFACE_DEFINE;
 	cmd->header.size = cmd_len;
 	cmd->body.sid = srf->res.id;
-	cmd->body.surfaceFlags = srf->flags;
+	/*
+	 * Downcast of surfaceFlags, was upcasted when received from user-space,
+	 * since driver internally stores as 64 bit.
+	 * For legacy surface define only 32 bit flag is supported.
+	 */
+	cmd->body.surfaceFlags = (SVGA3dSurface1Flags)srf->flags;
 	cmd->body.format = srf->format;
 	for (i = 0; i < DRM_VMW_MAX_SURFACE_FACES; ++i)
 		cmd->body.face[i].numMipLevels = srf->mip_levels[i];
@@ -468,7 +486,7 @@ static int vmw_legacy_srf_dma(struct vmw_resource *res,
 	(void) vmw_execbuf_fence_commands(NULL, dev_priv,
 					  &fence, NULL);
 
-	vmw_fence_single_bo(val_buf->bo, fence);
+	vmw_bo_fence_single(val_buf->bo, fence);
 
 	if (likely(fence != NULL))
 		vmw_fence_obj_unreference(&fence);
@@ -760,7 +778,8 @@ int vmw_surface_define_ioctl(struct drm_device *dev, void *data,
 	srf = &user_srf->srf;
 	res = &srf->res;
 
-	srf->flags = req->flags;
+	/* Driver internally stores as 64-bit flags */
+	srf->flags = (SVGA3dSurfaceAllFlags)req->flags;
 	srf->format = req->format;
 	srf->scanout = req->scanout;
 
@@ -785,6 +804,8 @@ int vmw_surface_define_ioctl(struct drm_device *dev, void *data,
 	srf->base_size = *srf->sizes;
 	srf->autogen_filter = SVGA3D_TEX_FILTER_NONE;
 	srf->multisample_count = 0;
+	srf->multisample_pattern = SVGA3D_MS_PATTERN_NONE;
+	srf->quality_level = SVGA3D_MS_QUALITY_NONE;
 
 	cur_bo_offset = 0;
 	cur_offset = srf->offsets;
@@ -842,12 +863,12 @@ int vmw_surface_define_ioctl(struct drm_device *dev, void *data,
 	if (dev_priv->has_mob && req->shareable) {
 		uint32_t backup_handle;
 
-		ret = vmw_user_dmabuf_alloc(dev_priv, tfile,
-					    res->backup_size,
-					    true,
-					    &backup_handle,
-					    &res->backup,
-					    &user_srf->backup_base);
+		ret = vmw_user_bo_alloc(dev_priv, tfile,
+					res->backup_size,
+					true,
+					&backup_handle,
+					&res->backup,
+					&user_srf->backup_base);
 		if (unlikely(ret != 0)) {
 			vmw_resource_unreference(&res);
 			goto out_unlock;
@@ -990,7 +1011,8 @@ int vmw_surface_reference_ioctl(struct drm_device *dev, void *data,
 	user_srf = container_of(base, struct vmw_user_surface, prime.base);
 	srf = &user_srf->srf;
 
-	rep->flags = srf->flags;
+	/* Downcast of flags when sending back to user space */
+	rep->flags = (uint32_t)srf->flags;
 	rep->format = srf->format;
 	memcpy(rep->mip_levels, srf->mip_levels, sizeof(srf->mip_levels));
 	user_sizes = (struct drm_vmw_size __user *)(unsigned long)
@@ -1031,6 +1053,10 @@ static int vmw_gb_surface_create(struct vmw_resource *res)
 		SVGA3dCmdHeader header;
 		SVGA3dCmdDefineGBSurface_v2 body;
 	} *cmd2;
+	struct {
+		SVGA3dCmdHeader header;
+		SVGA3dCmdDefineGBSurface_v3 body;
+	} *cmd3;
 
 	if (likely(res->id != -1))
 		return 0;
@@ -1047,7 +1073,11 @@ static int vmw_gb_surface_create(struct vmw_resource *res)
 		goto out_no_fifo;
 	}
 
-	if (srf->array_size > 0) {
+	if (dev_priv->has_sm4_1 && srf->array_size > 0) {
+		cmd_id = SVGA_3D_CMD_DEFINE_GB_SURFACE_V3;
+		cmd_len = sizeof(cmd3->body);
+		submit_len = sizeof(*cmd3);
+	} else if (srf->array_size > 0) {
 		/* has_dx checked on creation time. */
 		cmd_id = SVGA_3D_CMD_DEFINE_GB_SURFACE_V2;
 		cmd_len = sizeof(cmd2->body);
@@ -1060,6 +1090,7 @@ static int vmw_gb_surface_create(struct vmw_resource *res)
 
 	cmd = vmw_fifo_reserve(dev_priv, submit_len);
 	cmd2 = (typeof(cmd2))cmd;
+	cmd3 = (typeof(cmd3))cmd;
 	if (unlikely(!cmd)) {
 		DRM_ERROR("Failed reserving FIFO space for surface "
 			  "creation.\n");
@@ -1067,12 +1098,27 @@ static int vmw_gb_surface_create(struct vmw_resource *res)
 		goto out_no_fifo;
 	}
 
-	if (srf->array_size > 0) {
+	if (dev_priv->has_sm4_1 && srf->array_size > 0) {
+		cmd3->header.id = cmd_id;
+		cmd3->header.size = cmd_len;
+		cmd3->body.sid = srf->res.id;
+		cmd3->body.surfaceFlags = srf->flags;
+		cmd3->body.format = srf->format;
+		cmd3->body.numMipLevels = srf->mip_levels[0];
+		cmd3->body.multisampleCount = srf->multisample_count;
+		cmd3->body.multisamplePattern = srf->multisample_pattern;
+		cmd3->body.qualityLevel = srf->quality_level;
+		cmd3->body.autogenFilter = srf->autogen_filter;
+		cmd3->body.size.width = srf->base_size.width;
+		cmd3->body.size.height = srf->base_size.height;
+		cmd3->body.size.depth = srf->base_size.depth;
+		cmd3->body.arraySize = srf->array_size;
+	} else if (srf->array_size > 0) {
 		cmd2->header.id = cmd_id;
 		cmd2->header.size = cmd_len;
 		cmd2->body.sid = srf->res.id;
 		cmd2->body.surfaceFlags = srf->flags;
-		cmd2->body.format = cpu_to_le32(srf->format);
+		cmd2->body.format = srf->format;
 		cmd2->body.numMipLevels = srf->mip_levels[0];
 		cmd2->body.multisampleCount = srf->multisample_count;
 		cmd2->body.autogenFilter = srf->autogen_filter;
@@ -1085,7 +1131,7 @@ static int vmw_gb_surface_create(struct vmw_resource *res)
 		cmd->header.size = cmd_len;
 		cmd->body.sid = srf->res.id;
 		cmd->body.surfaceFlags = srf->flags;
-		cmd->body.format = cpu_to_le32(srf->format);
+		cmd->body.format = srf->format;
 		cmd->body.numMipLevels = srf->mip_levels[0];
 		cmd->body.multisampleCount = srf->multisample_count;
 		cmd->body.autogenFilter = srf->autogen_filter;
@@ -1210,7 +1256,7 @@ static int vmw_gb_surface_unbind(struct vmw_resource *res,
 	(void) vmw_execbuf_fence_commands(NULL, dev_priv,
 					  &fence, NULL);
 
-	vmw_fence_single_bo(val_buf->bo, fence);
+	vmw_bo_fence_single(val_buf->bo, fence);
 
 	if (likely(fence != NULL))
 		vmw_fence_obj_unreference(&fence);
@@ -1256,194 +1302,55 @@ static int vmw_gb_surface_destroy(struct vmw_resource *res)
 
 /**
  * vmw_gb_surface_define_ioctl - Ioctl function implementing
- *                               the user surface define functionality.
+ * the user surface define functionality.
  *
- * @dev:            Pointer to a struct drm_device.
- * @data:           Pointer to data copied from / to user-space.
- * @file_priv:      Pointer to a drm file private structure.
+ * @dev: Pointer to a struct drm_device.
+ * @data: Pointer to data copied from / to user-space.
+ * @file_priv: Pointer to a drm file private structure.
  */
 int vmw_gb_surface_define_ioctl(struct drm_device *dev, void *data,
 				struct drm_file *file_priv)
 {
-	struct vmw_private *dev_priv = vmw_priv(dev);
-	struct vmw_user_surface *user_srf;
-	struct vmw_surface *srf;
-	struct vmw_resource *res;
-	struct vmw_resource *tmp;
 	union drm_vmw_gb_surface_create_arg *arg =
 	    (union drm_vmw_gb_surface_create_arg *)data;
-	struct drm_vmw_gb_surface_create_req *req = &arg->req;
 	struct drm_vmw_gb_surface_create_rep *rep = &arg->rep;
-	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
-	int ret;
-	uint32_t size;
-	uint32_t backup_handle = 0;
-
-	if (req->multisample_count != 0)
-		return -EINVAL;
-
-	if (req->mip_levels > DRM_VMW_MAX_MIP_LEVELS)
-		return -EINVAL;
+	struct drm_vmw_gb_surface_create_ext_req req_ext;
 
-	if (unlikely(vmw_user_surface_size == 0))
-		vmw_user_surface_size = ttm_round_pot(sizeof(*user_srf)) +
-			128;
-
-	size = vmw_user_surface_size + 128;
-
-	/* Define a surface based on the parameters. */
-	ret = vmw_surface_gb_priv_define(dev,
-			size,
-			req->svga3d_flags,
-			req->format,
-			req->drm_surface_flags & drm_vmw_surface_flag_scanout,
-			req->mip_levels,
-			req->multisample_count,
-			req->array_size,
-			req->base_size,
-			&srf);
-	if (unlikely(ret != 0))
-		return ret;
-
-	user_srf = container_of(srf, struct vmw_user_surface, srf);
-	if (drm_is_primary_client(file_priv))
-		user_srf->master = drm_master_get(file_priv->master);
+	req_ext.base = arg->req;
+	req_ext.version = drm_vmw_gb_surface_v1;
+	req_ext.svga3d_flags_upper_32_bits = 0;
+	req_ext.multisample_pattern = SVGA3D_MS_PATTERN_NONE;
+	req_ext.quality_level = SVGA3D_MS_QUALITY_NONE;
+	req_ext.must_be_zero = 0;
 
-	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
-	if (unlikely(ret != 0))
-		return ret;
-
-	res = &user_srf->srf.res;
-
-
-	if (req->buffer_handle != SVGA3D_INVALID_ID) {
-		ret = vmw_user_dmabuf_lookup(tfile, req->buffer_handle,
-					     &res->backup,
-					     &user_srf->backup_base);
-		if (ret == 0) {
-			if (res->backup->base.num_pages * PAGE_SIZE <
-			    res->backup_size) {
-				DRM_ERROR("Surface backup buffer is too small.\n");
-				vmw_dmabuf_unreference(&res->backup);
-				ret = -EINVAL;
-				goto out_unlock;
-			} else {
-				backup_handle = req->buffer_handle;
-			}
-		}
-	} else if (req->drm_surface_flags & drm_vmw_surface_flag_create_buffer)
-		ret = vmw_user_dmabuf_alloc(dev_priv, tfile,
-					    res->backup_size,
-					    req->drm_surface_flags &
-					    drm_vmw_surface_flag_shareable,
-					    &backup_handle,
-					    &res->backup,
-					    &user_srf->backup_base);
-
-	if (unlikely(ret != 0)) {
-		vmw_resource_unreference(&res);
-		goto out_unlock;
-	}
-
-	tmp = vmw_resource_reference(res);
-	ret = ttm_prime_object_init(tfile, res->backup_size, &user_srf->prime,
-				    req->drm_surface_flags &
-				    drm_vmw_surface_flag_shareable,
-				    VMW_RES_SURFACE,
-				    &vmw_user_surface_base_release, NULL);
-
-	if (unlikely(ret != 0)) {
-		vmw_resource_unreference(&tmp);
-		vmw_resource_unreference(&res);
-		goto out_unlock;
-	}
-
-	rep->handle      = user_srf->prime.base.hash.key;
-	rep->backup_size = res->backup_size;
-	if (res->backup) {
-		rep->buffer_map_handle =
-			drm_vma_node_offset_addr(&res->backup->base.vma_node);
-		rep->buffer_size = res->backup->base.num_pages * PAGE_SIZE;
-		rep->buffer_handle = backup_handle;
-	} else {
-		rep->buffer_map_handle = 0;
-		rep->buffer_size = 0;
-		rep->buffer_handle = SVGA3D_INVALID_ID;
-	}
-
-	vmw_resource_unreference(&res);
-
-out_unlock:
-	ttm_read_unlock(&dev_priv->reservation_sem);
-	return ret;
+	return vmw_gb_surface_define_internal(dev, &req_ext, rep, file_priv);
 }
 
 /**
  * vmw_gb_surface_reference_ioctl - Ioctl function implementing
- *                                  the user surface reference functionality.
+ * the user surface reference functionality.
  *
- * @dev:            Pointer to a struct drm_device.
- * @data:           Pointer to data copied from / to user-space.
- * @file_priv:      Pointer to a drm file private structure.
+ * @dev: Pointer to a struct drm_device.
+ * @data: Pointer to data copied from / to user-space.
+ * @file_priv: Pointer to a drm file private structure.
  */
 int vmw_gb_surface_reference_ioctl(struct drm_device *dev, void *data,
 				   struct drm_file *file_priv)
 {
-	struct vmw_private *dev_priv = vmw_priv(dev);
 	union drm_vmw_gb_surface_reference_arg *arg =
 	    (union drm_vmw_gb_surface_reference_arg *)data;
 	struct drm_vmw_surface_arg *req = &arg->req;
 	struct drm_vmw_gb_surface_ref_rep *rep = &arg->rep;
-	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
-	struct vmw_surface *srf;
-	struct vmw_user_surface *user_srf;
-	struct ttm_base_object *base;
-	uint32_t backup_handle;
-	int ret = -EINVAL;
+	struct drm_vmw_gb_surface_ref_ext_rep rep_ext;
+	int ret;
+
+	ret = vmw_gb_surface_reference_internal(dev, req, &rep_ext, file_priv);
 
-	ret = vmw_surface_handle_reference(dev_priv, file_priv, req->sid,
-					   req->handle_type, &base);
 	if (unlikely(ret != 0))
 		return ret;
 
-	user_srf = container_of(base, struct vmw_user_surface, prime.base);
-	srf = &user_srf->srf;
-	if (!srf->res.backup) {
-		DRM_ERROR("Shared GB surface is missing a backup buffer.\n");
-		goto out_bad_resource;
-	}
-
-	mutex_lock(&dev_priv->cmdbuf_mutex); /* Protect res->backup */
-	ret = vmw_user_dmabuf_reference(tfile, srf->res.backup,
-					&backup_handle);
-	mutex_unlock(&dev_priv->cmdbuf_mutex);
-
-	if (unlikely(ret != 0)) {
-		DRM_ERROR("Could not add a reference to a GB surface "
-			  "backup buffer.\n");
-		(void) ttm_ref_object_base_unref(tfile, base->hash.key,
-						 TTM_REF_USAGE);
-		goto out_bad_resource;
-	}
-
-	rep->creq.svga3d_flags = srf->flags;
-	rep->creq.format = srf->format;
-	rep->creq.mip_levels = srf->mip_levels[0];
-	rep->creq.drm_surface_flags = 0;
-	rep->creq.multisample_count = srf->multisample_count;
-	rep->creq.autogen_filter = srf->autogen_filter;
-	rep->creq.array_size = srf->array_size;
-	rep->creq.buffer_handle = backup_handle;
-	rep->creq.base_size = srf->base_size;
-	rep->crep.handle = user_srf->prime.base.hash.key;
-	rep->crep.backup_size = srf->res.backup_size;
-	rep->crep.buffer_handle = backup_handle;
-	rep->crep.buffer_map_handle =
-		drm_vma_node_offset_addr(&srf->res.backup->base.vma_node);
-	rep->crep.buffer_size = srf->res.backup->base.num_pages * PAGE_SIZE;
-
-out_bad_resource:
-	ttm_base_object_unref(&base);
+	rep->creq = rep_ext.creq.base;
+	rep->crep = rep_ext.crep;
 
 	return ret;
 }
@@ -1461,6 +1368,8 @@ out_bad_resource:
  * @multisample_count:
  * @array_size: Surface array size.
  * @size: width, heigh, depth of the surface requested
+ * @multisample_pattern: Multisampling pattern when msaa is supported
+ * @quality_level: Precision settings
  * @user_srf_out: allocated user_srf.  Set to NULL on failure.
  *
  * GB surfaces allocated by this function will not have a user mode handle, and
@@ -1470,13 +1379,15 @@ out_bad_resource:
  */
 int vmw_surface_gb_priv_define(struct drm_device *dev,
 			       uint32_t user_accounting_size,
-			       uint32_t svga3d_flags,
+			       SVGA3dSurfaceAllFlags svga3d_flags,
 			       SVGA3dSurfaceFormat format,
 			       bool for_scanout,
 			       uint32_t num_mip_levels,
 			       uint32_t multisample_count,
 			       uint32_t array_size,
 			       struct drm_vmw_size size,
+			       SVGA3dMSPattern multisample_pattern,
+			       SVGA3dMSQualityLevel quality_level,
 			       struct vmw_surface **srf_out)
 {
 	struct vmw_private *dev_priv = vmw_priv(dev);
@@ -1487,7 +1398,8 @@ int vmw_surface_gb_priv_define(struct drm_device *dev,
 	};
 	struct vmw_surface *srf;
 	int ret;
-	u32 num_layers;
+	u32 num_layers = 1;
+	u32 sample_count = 1;
 
 	*srf_out = NULL;
 
@@ -1562,19 +1474,23 @@ int vmw_surface_gb_priv_define(struct drm_device *dev,
 	srf->autogen_filter    = SVGA3D_TEX_FILTER_NONE;
 	srf->array_size        = array_size;
 	srf->multisample_count = multisample_count;
+	srf->multisample_pattern = multisample_pattern;
+	srf->quality_level = quality_level;
 
 	if (array_size)
 		num_layers = array_size;
 	else if (svga3d_flags & SVGA3D_SURFACE_CUBEMAP)
 		num_layers = SVGA3D_MAX_SURFACE_FACES;
-	else
-		num_layers = 1;
+
+	if (srf->flags & SVGA3D_SURFACE_MULTISAMPLE)
+		sample_count = srf->multisample_count;
 
 	srf->res.backup_size   =
-		svga3dsurface_get_serialized_size(srf->format,
-						  srf->base_size,
-						  srf->mip_levels[0],
-						  num_layers);
+		svga3dsurface_get_serialized_size_extended(srf->format,
+							   srf->base_size,
+							   srf->mip_levels[0],
+							   num_layers,
+							   sample_count);
 
 	if (srf->flags & SVGA3D_SURFACE_BIND_STREAM_OUTPUT)
 		srf->res.backup_size += sizeof(SVGA3dDXSOState);
@@ -1599,3 +1515,266 @@ out_unlock:
 	ttm_read_unlock(&dev_priv->reservation_sem);
 	return ret;
 }
+
+/**
+ * vmw_gb_surface_define_ext_ioctl - Ioctl function implementing
+ * the user surface define functionality.
+ *
+ * @dev: Pointer to a struct drm_device.
+ * @data: Pointer to data copied from / to user-space.
+ * @file_priv: Pointer to a drm file private structure.
+ */
+int vmw_gb_surface_define_ext_ioctl(struct drm_device *dev, void *data,
+				struct drm_file *file_priv)
+{
+	union drm_vmw_gb_surface_create_ext_arg *arg =
+	    (union drm_vmw_gb_surface_create_ext_arg *)data;
+	struct drm_vmw_gb_surface_create_ext_req *req = &arg->req;
+	struct drm_vmw_gb_surface_create_rep *rep = &arg->rep;
+
+	return vmw_gb_surface_define_internal(dev, req, rep, file_priv);
+}
+
+/**
+ * vmw_gb_surface_reference_ext_ioctl - Ioctl function implementing
+ * the user surface reference functionality.
+ *
+ * @dev: Pointer to a struct drm_device.
+ * @data: Pointer to data copied from / to user-space.
+ * @file_priv: Pointer to a drm file private structure.
+ */
+int vmw_gb_surface_reference_ext_ioctl(struct drm_device *dev, void *data,
+				   struct drm_file *file_priv)
+{
+	union drm_vmw_gb_surface_reference_ext_arg *arg =
+	    (union drm_vmw_gb_surface_reference_ext_arg *)data;
+	struct drm_vmw_surface_arg *req = &arg->req;
+	struct drm_vmw_gb_surface_ref_ext_rep *rep = &arg->rep;
+
+	return vmw_gb_surface_reference_internal(dev, req, rep, file_priv);
+}
+
+/**
+ * vmw_gb_surface_define_internal - Ioctl function implementing
+ * the user surface define functionality.
+ *
+ * @dev: Pointer to a struct drm_device.
+ * @req: Request argument from user-space.
+ * @rep: Response argument to user-space.
+ * @file_priv: Pointer to a drm file private structure.
+ */
+static int
+vmw_gb_surface_define_internal(struct drm_device *dev,
+			       struct drm_vmw_gb_surface_create_ext_req *req,
+			       struct drm_vmw_gb_surface_create_rep *rep,
+			       struct drm_file *file_priv)
+{
+	struct vmw_private *dev_priv = vmw_priv(dev);
+	struct vmw_user_surface *user_srf;
+	struct vmw_surface *srf;
+	struct vmw_resource *res;
+	struct vmw_resource *tmp;
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+	int ret;
+	uint32_t size;
+	uint32_t backup_handle = 0;
+	SVGA3dSurfaceAllFlags svga3d_flags_64 =
+		SVGA3D_FLAGS_64(req->svga3d_flags_upper_32_bits,
+				req->base.svga3d_flags);
+
+	if (!dev_priv->has_sm4_1) {
+		/*
+		 * If SM4_1 is not support then cannot send 64-bit flag to
+		 * device.
+		 */
+		if (req->svga3d_flags_upper_32_bits != 0)
+			return -EINVAL;
+
+		if (req->base.multisample_count != 0)
+			return -EINVAL;
+
+		if (req->multisample_pattern != SVGA3D_MS_PATTERN_NONE)
+			return -EINVAL;
+
+		if (req->quality_level != SVGA3D_MS_QUALITY_NONE)
+			return -EINVAL;
+	}
+
+	if ((svga3d_flags_64 & SVGA3D_SURFACE_MULTISAMPLE) &&
+	    req->base.multisample_count == 0)
+		return -EINVAL;
+
+	if (req->base.mip_levels > DRM_VMW_MAX_MIP_LEVELS)
+		return -EINVAL;
+
+	if (unlikely(vmw_user_surface_size == 0))
+		vmw_user_surface_size = ttm_round_pot(sizeof(*user_srf)) +
+			128;
+
+	size = vmw_user_surface_size + 128;
+
+	/* Define a surface based on the parameters. */
+	ret = vmw_surface_gb_priv_define(dev,
+					 size,
+					 svga3d_flags_64,
+					 req->base.format,
+					 req->base.drm_surface_flags &
+					 drm_vmw_surface_flag_scanout,
+					 req->base.mip_levels,
+					 req->base.multisample_count,
+					 req->base.array_size,
+					 req->base.base_size,
+					 req->multisample_pattern,
+					 req->quality_level,
+					 &srf);
+	if (unlikely(ret != 0))
+		return ret;
+
+	user_srf = container_of(srf, struct vmw_user_surface, srf);
+	if (drm_is_primary_client(file_priv))
+		user_srf->master = drm_master_get(file_priv->master);
+
+	ret = ttm_read_lock(&dev_priv->reservation_sem, true);
+	if (unlikely(ret != 0))
+		return ret;
+
+	res = &user_srf->srf.res;
+
+	if (req->base.buffer_handle != SVGA3D_INVALID_ID) {
+		ret = vmw_user_bo_lookup(tfile, req->base.buffer_handle,
+					 &res->backup,
+					 &user_srf->backup_base);
+		if (ret == 0) {
+			if (res->backup->base.num_pages * PAGE_SIZE <
+			    res->backup_size) {
+				DRM_ERROR("Surface backup buffer too small.\n");
+				vmw_bo_unreference(&res->backup);
+				ret = -EINVAL;
+				goto out_unlock;
+			} else {
+				backup_handle = req->base.buffer_handle;
+			}
+		}
+	} else if (req->base.drm_surface_flags &
+		   drm_vmw_surface_flag_create_buffer)
+		ret = vmw_user_bo_alloc(dev_priv, tfile,
+					res->backup_size,
+					req->base.drm_surface_flags &
+					drm_vmw_surface_flag_shareable,
+					&backup_handle,
+					&res->backup,
+					&user_srf->backup_base);
+
+	if (unlikely(ret != 0)) {
+		vmw_resource_unreference(&res);
+		goto out_unlock;
+	}
+
+	tmp = vmw_resource_reference(res);
+	ret = ttm_prime_object_init(tfile, res->backup_size, &user_srf->prime,
+				    req->base.drm_surface_flags &
+				    drm_vmw_surface_flag_shareable,
+				    VMW_RES_SURFACE,
+				    &vmw_user_surface_base_release, NULL);
+
+	if (unlikely(ret != 0)) {
+		vmw_resource_unreference(&tmp);
+		vmw_resource_unreference(&res);
+		goto out_unlock;
+	}
+
+	rep->handle      = user_srf->prime.base.hash.key;
+	rep->backup_size = res->backup_size;
+	if (res->backup) {
+		rep->buffer_map_handle =
+			drm_vma_node_offset_addr(&res->backup->base.vma_node);
+		rep->buffer_size = res->backup->base.num_pages * PAGE_SIZE;
+		rep->buffer_handle = backup_handle;
+	} else {
+		rep->buffer_map_handle = 0;
+		rep->buffer_size = 0;
+		rep->buffer_handle = SVGA3D_INVALID_ID;
+	}
+
+	vmw_resource_unreference(&res);
+
+out_unlock:
+	ttm_read_unlock(&dev_priv->reservation_sem);
+	return ret;
+}
+
+/**
+ * vmw_gb_surface_reference_internal - Ioctl function implementing
+ * the user surface reference functionality.
+ *
+ * @dev: Pointer to a struct drm_device.
+ * @req: Pointer to user-space request surface arg.
+ * @rep: Pointer to response to user-space.
+ * @file_priv: Pointer to a drm file private structure.
+ */
+static int
+vmw_gb_surface_reference_internal(struct drm_device *dev,
+				  struct drm_vmw_surface_arg *req,
+				  struct drm_vmw_gb_surface_ref_ext_rep *rep,
+				  struct drm_file *file_priv)
+{
+	struct vmw_private *dev_priv = vmw_priv(dev);
+	struct ttm_object_file *tfile = vmw_fpriv(file_priv)->tfile;
+	struct vmw_surface *srf;
+	struct vmw_user_surface *user_srf;
+	struct ttm_base_object *base;
+	uint32_t backup_handle;
+	int ret = -EINVAL;
+
+	ret = vmw_surface_handle_reference(dev_priv, file_priv, req->sid,
+					   req->handle_type, &base);
+	if (unlikely(ret != 0))
+		return ret;
+
+	user_srf = container_of(base, struct vmw_user_surface, prime.base);
+	srf = &user_srf->srf;
+	if (!srf->res.backup) {
+		DRM_ERROR("Shared GB surface is missing a backup buffer.\n");
+		goto out_bad_resource;
+	}
+
+	mutex_lock(&dev_priv->cmdbuf_mutex); /* Protect res->backup */
+	ret = vmw_user_bo_reference(tfile, srf->res.backup, &backup_handle);
+	mutex_unlock(&dev_priv->cmdbuf_mutex);
+
+	if (unlikely(ret != 0)) {
+		DRM_ERROR("Could not add a reference to a GB surface "
+			  "backup buffer.\n");
+		(void) ttm_ref_object_base_unref(tfile, base->hash.key,
+						 TTM_REF_USAGE);
+		goto out_bad_resource;
+	}
+
+	rep->creq.base.svga3d_flags = SVGA3D_FLAGS_LOWER_32(srf->flags);
+	rep->creq.base.format = srf->format;
+	rep->creq.base.mip_levels = srf->mip_levels[0];
+	rep->creq.base.drm_surface_flags = 0;
+	rep->creq.base.multisample_count = srf->multisample_count;
+	rep->creq.base.autogen_filter = srf->autogen_filter;
+	rep->creq.base.array_size = srf->array_size;
+	rep->creq.base.buffer_handle = backup_handle;
+	rep->creq.base.base_size = srf->base_size;
+	rep->crep.handle = user_srf->prime.base.hash.key;
+	rep->crep.backup_size = srf->res.backup_size;
+	rep->crep.buffer_handle = backup_handle;
+	rep->crep.buffer_map_handle =
+		drm_vma_node_offset_addr(&srf->res.backup->base.vma_node);
+	rep->crep.buffer_size = srf->res.backup->base.num_pages * PAGE_SIZE;
+
+	rep->creq.version = drm_vmw_gb_surface_v1;
+	rep->creq.svga3d_flags_upper_32_bits =
+		SVGA3D_FLAGS_UPPER_32(srf->flags);
+	rep->creq.multisample_pattern = srf->multisample_pattern;
+	rep->creq.quality_level = srf->quality_level;
+	rep->creq.must_be_zero = 0;
+
+out_bad_resource:
+	ttm_base_object_unref(&base);
+
+	return ret;
+}
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_buffer.c
index 21111fd091f9..31786b200afc 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_buffer.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_buffer.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009-2015 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2015 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
@@ -798,7 +798,7 @@ static int vmw_verify_access(struct ttm_buffer_object *bo, struct file *filp)
 	struct ttm_object_file *tfile =
 		vmw_fpriv((struct drm_file *)filp->private_data)->tfile;
 
-	return vmw_user_dmabuf_verify_access(bo, tfile);
+	return vmw_user_bo_verify_access(bo, tfile);
 }
 
 static int vmw_ttm_io_mem_reserve(struct ttm_bo_device *bdev, struct ttm_mem_reg *mem)
@@ -852,7 +852,7 @@ static void vmw_move_notify(struct ttm_buffer_object *bo,
 			    bool evict,
 			    struct ttm_mem_reg *mem)
 {
-	vmw_resource_move_notify(bo, mem);
+	vmw_bo_move_notify(bo, mem);
 	vmw_query_move_notify(bo, mem);
 }
 
@@ -864,7 +864,7 @@ static void vmw_move_notify(struct ttm_buffer_object *bo,
  */
 static void vmw_swap_notify(struct ttm_buffer_object *bo)
 {
-	vmw_resource_swap_notify(bo);
+	vmw_bo_swap_notify(bo);
 	(void) ttm_bo_wait(bo, false, false);
 }
 
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c
index e771091d2cd3..7b1e5a5cbd2c 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_ttm_glue.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2009-2011 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2009-2011 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_va.c b/drivers/gpu/drm/vmwgfx/vmwgfx_va.c
index b4162fd78600..ebc1d83c34b4 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_va.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_va.c
@@ -1,7 +1,7 @@
+// SPDX-License-Identifier: GPL-2.0 OR MIT
 /**************************************************************************
  *
- * Copyright © 2012-2016 VMware, Inc., Palo Alto, CA., USA
- * All Rights Reserved.
+ * Copyright 2012-2016 VMware, Inc., Palo Alto, CA., USA
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the
diff --git a/drivers/gpu/drm/xen/xen_drm_front.c b/drivers/gpu/drm/xen/xen_drm_front.c
index 1b0ea9ac330e..6b6d5ab82ec3 100644
--- a/drivers/gpu/drm/xen/xen_drm_front.c
+++ b/drivers/gpu/drm/xen/xen_drm_front.c
@@ -188,8 +188,8 @@ int xen_drm_front_dbuf_create(struct xen_drm_front_info *front_info,
 	buf_cfg.be_alloc = front_info->cfg.be_alloc;
 
 	shbuf = xen_drm_front_shbuf_alloc(&buf_cfg);
-	if (!shbuf)
-		return -ENOMEM;
+	if (IS_ERR(shbuf))
+		return PTR_ERR(shbuf);
 
 	ret = dbuf_add_to_list(front_info, shbuf, dbuf_cookie);
 	if (ret < 0) {
@@ -543,8 +543,8 @@ static int xen_drm_drv_init(struct xen_drm_front_info *front_info)
 	front_info->drm_info = drm_info;
 
 	drm_dev = drm_dev_alloc(&xen_drm_driver, dev);
-	if (!drm_dev) {
-		ret = -ENOMEM;
+	if (IS_ERR(drm_dev)) {
+		ret = PTR_ERR(drm_dev);
 		goto fail;
 	}
 
@@ -623,7 +623,7 @@ static int displback_initwait(struct xen_drm_front_info *front_info)
 	if (ret < 0)
 		return ret;
 
-	DRM_INFO("Have %d conector(s)\n", cfg->num_connectors);
+	DRM_INFO("Have %d connector(s)\n", cfg->num_connectors);
 	/* Create event channels for all connectors and publish */
 	ret = xen_drm_front_evtchnl_create_all(front_info);
 	if (ret < 0)
@@ -737,9 +737,8 @@ static int xen_drv_probe(struct xenbus_device *xb_dev,
 	 * is not correct: to fix this call of_dma_configure() with a NULL
 	 * node to set default DMA ops.
 	 */
-	dev->bus->force_dma = true;
 	dev->coherent_dma_mask = DMA_BIT_MASK(32);
-	ret = of_dma_configure(dev, NULL);
+	ret = of_dma_configure(dev, NULL, true);
 	if (ret < 0) {
 		DRM_ERROR("Cannot setup DMA ops, ret %d", ret);
 		return ret;
@@ -778,7 +777,7 @@ static int xen_drv_remove(struct xenbus_device *dev)
 	 */
 	while ((xenbus_read_unsigned(front_info->xb_dev->otherend, "state",
 				     XenbusStateUnknown) != XenbusStateInitWait) &&
-				     to--)
+				     --to)
 		msleep(10);
 
 	if (!to) {
diff --git a/drivers/gpu/drm/xen/xen_drm_front.h b/drivers/gpu/drm/xen/xen_drm_front.h
index 2c2479b571ae..5693b4a4b02b 100644
--- a/drivers/gpu/drm/xen/xen_drm_front.h
+++ b/drivers/gpu/drm/xen/xen_drm_front.h
@@ -126,12 +126,12 @@ struct xen_drm_front_drm_info {
 
 static inline u64 xen_drm_front_fb_to_cookie(struct drm_framebuffer *fb)
 {
-	return (u64)fb;
+	return (uintptr_t)fb;
 }
 
 static inline u64 xen_drm_front_dbuf_to_cookie(struct drm_gem_object *gem_obj)
 {
-	return (u64)gem_obj;
+	return (uintptr_t)gem_obj;
 }
 
 int xen_drm_front_mode_set(struct xen_drm_front_drm_pipeline *pipeline,
diff --git a/drivers/gpu/drm/xen/xen_drm_front_shbuf.c b/drivers/gpu/drm/xen/xen_drm_front_shbuf.c
index d5705251a0d6..d333b67cc1a0 100644
--- a/drivers/gpu/drm/xen/xen_drm_front_shbuf.c
+++ b/drivers/gpu/drm/xen/xen_drm_front_shbuf.c
@@ -122,7 +122,7 @@ static void guest_calc_num_grefs(struct xen_drm_front_shbuf *buf)
 }
 
 #define xen_page_to_vaddr(page) \
-		((phys_addr_t)pfn_to_kaddr(page_to_xen_pfn(page)))
+		((uintptr_t)pfn_to_kaddr(page_to_xen_pfn(page)))
 
 static int backend_unmap(struct xen_drm_front_shbuf *buf)
 {
@@ -383,7 +383,7 @@ xen_drm_front_shbuf_alloc(struct xen_drm_front_shbuf_cfg *cfg)
 
 	buf = kzalloc(sizeof(*buf), GFP_KERNEL);
 	if (!buf)
-		return NULL;
+		return ERR_PTR(-ENOMEM);
 
 	if (cfg->be_alloc)
 		buf->ops = &backend_ops;
diff --git a/drivers/gpu/drm/zte/zx_hdmi.c b/drivers/gpu/drm/zte/zx_hdmi.c
index 13ea90f7a185..78655269d843 100644
--- a/drivers/gpu/drm/zte/zx_hdmi.c
+++ b/drivers/gpu/drm/zte/zx_hdmi.c
@@ -272,7 +272,7 @@ static int zx_hdmi_connector_get_modes(struct drm_connector *connector)
 
 	hdmi->sink_is_hdmi = drm_detect_hdmi_monitor(edid);
 	hdmi->sink_has_audio = drm_detect_monitor_audio(edid);
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 	ret = drm_add_edid_modes(connector, edid);
 	kfree(edid);
 
@@ -326,7 +326,7 @@ static int zx_hdmi_register(struct drm_device *drm, struct zx_hdmi *hdmi)
 	drm_connector_helper_add(&hdmi->connector,
 				 &zx_hdmi_connector_helper_funcs);
 
-	drm_mode_connector_attach_encoder(&hdmi->connector, encoder);
+	drm_connector_attach_encoder(&hdmi->connector, encoder);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/zte/zx_plane.c b/drivers/gpu/drm/zte/zx_plane.c
index d1931f5ea0b2..ae8c53b4b261 100644
--- a/drivers/gpu/drm/zte/zx_plane.c
+++ b/drivers/gpu/drm/zte/zx_plane.c
@@ -446,7 +446,7 @@ static const struct drm_plane_helper_funcs zx_gl_plane_helper_funcs = {
 
 static void zx_plane_destroy(struct drm_plane *plane)
 {
-	drm_plane_helper_disable(plane);
+	drm_plane_helper_disable(plane, NULL);
 	drm_plane_cleanup(plane);
 }
 
diff --git a/drivers/gpu/drm/zte/zx_tvenc.c b/drivers/gpu/drm/zte/zx_tvenc.c
index 0de1a71ca4e0..b73afb212fb2 100644
--- a/drivers/gpu/drm/zte/zx_tvenc.c
+++ b/drivers/gpu/drm/zte/zx_tvenc.c
@@ -297,7 +297,7 @@ static int zx_tvenc_register(struct drm_device *drm, struct zx_tvenc *tvenc)
 			   DRM_MODE_CONNECTOR_Composite);
 	drm_connector_helper_add(connector, &zx_tvenc_connector_helper_funcs);
 
-	drm_mode_connector_attach_encoder(connector, encoder);
+	drm_connector_attach_encoder(connector, encoder);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/zte/zx_vga.c b/drivers/gpu/drm/zte/zx_vga.c
index 3e7e33cd3dfa..23d1ff4355a0 100644
--- a/drivers/gpu/drm/zte/zx_vga.c
+++ b/drivers/gpu/drm/zte/zx_vga.c
@@ -109,7 +109,7 @@ static int zx_vga_connector_get_modes(struct drm_connector *connector)
 	 */
 	zx_writel(vga->mmio + VGA_AUTO_DETECT_SEL, VGA_DETECT_SEL_HAS_DEVICE);
 
-	drm_mode_connector_update_edid_property(connector, edid);
+	drm_connector_update_edid_property(connector, edid);
 	ret = drm_add_edid_modes(connector, edid);
 	kfree(edid);
 
@@ -175,7 +175,7 @@ static int zx_vga_register(struct drm_device *drm, struct zx_vga *vga)
 
 	drm_connector_helper_add(connector, &zx_vga_connector_helper_funcs);
 
-	ret = drm_mode_connector_attach_encoder(connector, encoder);
+	ret = drm_connector_attach_encoder(connector, encoder);
 	if (ret) {
 		DRM_DEV_ERROR(dev, "failed to attach encoder: %d\n", ret);
 		goto clean_connector;
diff --git a/drivers/gpu/host1x/bus.c b/drivers/gpu/host1x/bus.c
index 88a3558b7916..815bdb42e3f0 100644
--- a/drivers/gpu/host1x/bus.c
+++ b/drivers/gpu/host1x/bus.c
@@ -314,6 +314,11 @@ static int host1x_device_match(struct device *dev, struct device_driver *drv)
 	return strcmp(dev_name(dev), drv->name) == 0;
 }
 
+static int host1x_dma_configure(struct device *dev)
+{
+	return of_dma_configure(dev, dev->of_node, true);
+}
+
 static const struct dev_pm_ops host1x_device_pm_ops = {
 	.suspend = pm_generic_suspend,
 	.resume = pm_generic_resume,
@@ -326,8 +331,8 @@ static const struct dev_pm_ops host1x_device_pm_ops = {
 struct bus_type host1x_bus_type = {
 	.name = "host1x",
 	.match = host1x_device_match,
+	.dma_configure	= host1x_dma_configure,
 	.pm = &host1x_device_pm_ops,
-	.force_dma = true,
 };
 
 static void __host1x_device_del(struct host1x_device *device)
@@ -416,7 +421,7 @@ static int host1x_device_add(struct host1x *host1x,
 	device->dev.bus = &host1x_bus_type;
 	device->dev.parent = host1x->dev;
 
-	of_dma_configure(&device->dev, host1x->dev->of_node);
+	of_dma_configure(&device->dev, host1x->dev->of_node, true);
 
 	err = host1x_device_parse_dt(device, driver);
 	if (err < 0) {
diff --git a/drivers/gpu/host1x/cdma.c b/drivers/gpu/host1x/cdma.c
index 28541b280739..91df51e631b2 100644
--- a/drivers/gpu/host1x/cdma.c
+++ b/drivers/gpu/host1x/cdma.c
@@ -51,7 +51,7 @@ static void host1x_pushbuffer_destroy(struct push_buffer *pb)
 	struct host1x_cdma *cdma = pb_to_cdma(pb);
 	struct host1x *host1x = cdma_to_host1x(cdma);
 
-	if (!pb->phys)
+	if (!pb->mapped)
 		return;
 
 	if (host1x->domain) {
@@ -127,7 +127,7 @@ static int host1x_pushbuffer_init(struct push_buffer *pb)
 iommu_free_iova:
 	__free_iova(&host1x->iova, alloc);
 iommu_free_mem:
-	dma_free_wc(host1x->dev, pb->alloc_size, pb->mapped, pb->phys);
+	dma_free_wc(host1x->dev, size, pb->mapped, pb->phys);
 
 	return err;
 }
@@ -247,7 +247,7 @@ static void cdma_start_timer_locked(struct host1x_cdma *cdma,
 static void stop_cdma_timer_locked(struct host1x_cdma *cdma)
 {
 	cancel_delayed_work(&cdma->timeout.wq);
-	cdma->timeout.client = 0;
+	cdma->timeout.client = NULL;
 }
 
 /*
diff --git a/drivers/gpu/host1x/cdma.h b/drivers/gpu/host1x/cdma.h
index 286d49386be9..e97e17b82370 100644
--- a/drivers/gpu/host1x/cdma.h
+++ b/drivers/gpu/host1x/cdma.h
@@ -44,7 +44,7 @@ struct host1x_job;
 struct push_buffer {
 	void *mapped;			/* mapped pushbuffer memory */
 	dma_addr_t dma;			/* device address of pushbuffer */
-	phys_addr_t phys;		/* physical address of pushbuffer */
+	dma_addr_t phys;		/* physical address of pushbuffer */
 	u32 fence;			/* index we've written */
 	u32 pos;			/* index to write to */
 	u32 size;
@@ -58,7 +58,7 @@ struct buffer_timeout {
 	u32 syncpt_val;			/* syncpt value when completed */
 	ktime_t start_ktime;		/* starting time */
 	/* context timeout information */
-	int client;
+	struct host1x_client *client;
 };
 
 enum cdma_event {
diff --git a/drivers/gpu/host1x/debug.c b/drivers/gpu/host1x/debug.c
index dc77ec452ffc..329e4a3d8ae7 100644
--- a/drivers/gpu/host1x/debug.c
+++ b/drivers/gpu/host1x/debug.c
@@ -103,7 +103,7 @@ static void show_syncpts(struct host1x *m, struct output *o)
 
 static void show_all(struct host1x *m, struct output *o, bool show_fifo)
 {
-	int i;
+	unsigned int i;
 
 	host1x_hw_show_mlocks(m, o);
 	show_syncpts(m, o);
diff --git a/drivers/gpu/host1x/dev.c b/drivers/gpu/host1x/dev.c
index 03db71173f5d..f1d5f76e9c33 100644
--- a/drivers/gpu/host1x/dev.c
+++ b/drivers/gpu/host1x/dev.c
@@ -223,10 +223,14 @@ static int host1x_probe(struct platform_device *pdev)
 		struct iommu_domain_geometry *geometry;
 		unsigned long order;
 
+		err = iova_cache_get();
+		if (err < 0)
+			goto put_group;
+
 		host->domain = iommu_domain_alloc(&platform_bus_type);
 		if (!host->domain) {
 			err = -ENOMEM;
-			goto put_group;
+			goto put_cache;
 		}
 
 		err = iommu_attach_group(host->domain, host->group);
@@ -234,6 +238,7 @@ static int host1x_probe(struct platform_device *pdev)
 			if (err == -ENODEV) {
 				iommu_domain_free(host->domain);
 				host->domain = NULL;
+				iova_cache_put();
 				iommu_group_put(host->group);
 				host->group = NULL;
 				goto skip_iommu;
@@ -308,6 +313,9 @@ fail_detach_device:
 fail_free_domain:
 	if (host->domain)
 		iommu_domain_free(host->domain);
+put_cache:
+	if (host->group)
+		iova_cache_put();
 put_group:
 	iommu_group_put(host->group);
 
@@ -328,6 +336,7 @@ static int host1x_remove(struct platform_device *pdev)
 		put_iova_domain(&host->iova);
 		iommu_detach_group(host->domain, host->group);
 		iommu_domain_free(host->domain);
+		iova_cache_put();
 		iommu_group_put(host->group);
 	}
 
diff --git a/drivers/gpu/host1x/dev.h b/drivers/gpu/host1x/dev.h
index 43e9fabb43a1..36f44ffebe73 100644
--- a/drivers/gpu/host1x/dev.h
+++ b/drivers/gpu/host1x/dev.h
@@ -78,7 +78,6 @@ struct host1x_syncpt_ops {
 	void (*load_wait_base)(struct host1x_syncpt *syncpt);
 	u32 (*load)(struct host1x_syncpt *syncpt);
 	int (*cpu_incr)(struct host1x_syncpt *syncpt);
-	int (*patch_wait)(struct host1x_syncpt *syncpt, void *patch_addr);
 	void (*assign_to_channel)(struct host1x_syncpt *syncpt,
 	                          struct host1x_channel *channel);
 	void (*enable_protection)(struct host1x *host);
@@ -183,13 +182,6 @@ static inline int host1x_hw_syncpt_cpu_incr(struct host1x *host,
 	return host->syncpt_op->cpu_incr(sp);
 }
 
-static inline int host1x_hw_syncpt_patch_wait(struct host1x *host,
-					      struct host1x_syncpt *sp,
-					      void *patch_addr)
-{
-	return host->syncpt_op->patch_wait(sp, patch_addr);
-}
-
 static inline void host1x_hw_syncpt_assign_to_channel(
 	struct host1x *host, struct host1x_syncpt *sp,
 	struct host1x_channel *ch)
diff --git a/drivers/gpu/host1x/hw/channel_hw.c b/drivers/gpu/host1x/hw/channel_hw.c
index 9af758785a11..d188f9068b91 100644
--- a/drivers/gpu/host1x/hw/channel_hw.c
+++ b/drivers/gpu/host1x/hw/channel_hw.c
@@ -104,8 +104,7 @@ static int channel_submit(struct host1x_job *job)
 	sp = host->syncpt + job->syncpt_id;
 	trace_host1x_channel_submit(dev_name(ch->dev),
 				    job->num_gathers, job->num_relocs,
-				    job->num_waitchk, job->syncpt_id,
-				    job->syncpt_incrs);
+				    job->syncpt_id, job->syncpt_incrs);
 
 	/* before error checks, return current max */
 	prev_max = job->syncpt_end = host1x_syncpt_read_max(sp);
@@ -165,7 +164,7 @@ static int channel_submit(struct host1x_job *job)
 	trace_host1x_channel_submitted(dev_name(ch->dev), prev_max, syncval);
 
 	/* schedule a submit complete interrupt */
-	err = host1x_intr_add_action(host, job->syncpt_id, syncval,
+	err = host1x_intr_add_action(host, sp, syncval,
 				     HOST1X_INTR_ACTION_SUBMIT_COMPLETE, ch,
 				     completed_waiter, NULL);
 	completed_waiter = NULL;
diff --git a/drivers/gpu/host1x/hw/syncpt_hw.c b/drivers/gpu/host1x/hw/syncpt_hw.c
index 7dfd47d74f89..a23bb3352d02 100644
--- a/drivers/gpu/host1x/hw/syncpt_hw.c
+++ b/drivers/gpu/host1x/hw/syncpt_hw.c
@@ -96,16 +96,6 @@ static int syncpt_cpu_incr(struct host1x_syncpt *sp)
 	return 0;
 }
 
-/* remove a wait pointed to by patch_addr */
-static int syncpt_patch_wait(struct host1x_syncpt *sp, void *patch_addr)
-{
-	u32 override = host1x_class_host_wait_syncpt(HOST1X_SYNCPT_RESERVED, 0);
-
-	*((u32 *)patch_addr) = override;
-
-	return 0;
-}
-
 /**
  * syncpt_assign_to_channel() - Assign syncpoint to channel
  * @sp: syncpoint
@@ -156,7 +146,6 @@ static const struct host1x_syncpt_ops host1x_syncpt_ops = {
 	.load_wait_base = syncpt_read_wait_base,
 	.load = syncpt_load,
 	.cpu_incr = syncpt_cpu_incr,
-	.patch_wait = syncpt_patch_wait,
 	.assign_to_channel = syncpt_assign_to_channel,
 	.enable_protection = syncpt_enable_protection,
 };
diff --git a/drivers/gpu/host1x/intr.c b/drivers/gpu/host1x/intr.c
index 8b4fad0ab35d..9629c009d10f 100644
--- a/drivers/gpu/host1x/intr.c
+++ b/drivers/gpu/host1x/intr.c
@@ -144,7 +144,7 @@ static const action_handler action_handlers[HOST1X_INTR_ACTION_COUNT] = {
 static void run_handlers(struct list_head completed[HOST1X_INTR_ACTION_COUNT])
 {
 	struct list_head *head = completed;
-	int i;
+	unsigned int i;
 
 	for (i = 0; i < HOST1X_INTR_ACTION_COUNT; ++i, ++head) {
 		action_handler handler = action_handlers[i];
@@ -211,11 +211,11 @@ static void syncpt_thresh_work(struct work_struct *work)
 				host1x_syncpt_load(host->syncpt + id));
 }
 
-int host1x_intr_add_action(struct host1x *host, unsigned int id, u32 thresh,
-			   enum host1x_intr_action action, void *data,
-			   struct host1x_waitlist *waiter, void **ref)
+int host1x_intr_add_action(struct host1x *host, struct host1x_syncpt *syncpt,
+			   u32 thresh, enum host1x_intr_action action,
+			   void *data, struct host1x_waitlist *waiter,
+			   void **ref)
 {
-	struct host1x_syncpt *syncpt;
 	int queue_was_empty;
 
 	if (waiter == NULL) {
@@ -234,19 +234,17 @@ int host1x_intr_add_action(struct host1x *host, unsigned int id, u32 thresh,
 	waiter->data = data;
 	waiter->count = 1;
 
-	syncpt = host->syncpt + id;
-
 	spin_lock(&syncpt->intr.lock);
 
 	queue_was_empty = list_empty(&syncpt->intr.wait_head);
 
 	if (add_waiter_to_queue(waiter, &syncpt->intr.wait_head)) {
 		/* added at head of list - new threshold value */
-		host1x_hw_intr_set_syncpt_threshold(host, id, thresh);
+		host1x_hw_intr_set_syncpt_threshold(host, syncpt->id, thresh);
 
 		/* added as first waiter - enable interrupt */
 		if (queue_was_empty)
-			host1x_hw_intr_enable_syncpt_intr(host, id);
+			host1x_hw_intr_enable_syncpt_intr(host, syncpt->id);
 	}
 
 	spin_unlock(&syncpt->intr.lock);
diff --git a/drivers/gpu/host1x/intr.h b/drivers/gpu/host1x/intr.h
index 1370c2bb75b8..6db96af484fe 100644
--- a/drivers/gpu/host1x/intr.h
+++ b/drivers/gpu/host1x/intr.h
@@ -22,6 +22,7 @@
 #include <linux/interrupt.h>
 #include <linux/workqueue.h>
 
+struct host1x_syncpt;
 struct host1x;
 
 enum host1x_intr_action {
@@ -75,9 +76,10 @@ struct host1x_waitlist {
  *
  * This is a non-blocking api.
  */
-int host1x_intr_add_action(struct host1x *host, unsigned int id, u32 thresh,
-	enum host1x_intr_action action, void *data,
-	struct host1x_waitlist *waiter, void **ref);
+int host1x_intr_add_action(struct host1x *host, struct host1x_syncpt *syncpt,
+			   u32 thresh, enum host1x_intr_action action,
+			   void *data, struct host1x_waitlist *waiter,
+			   void **ref);
 
 /*
  * Unreference an action submitted to host1x_intr_add_action().
diff --git a/drivers/gpu/host1x/job.c b/drivers/gpu/host1x/job.c
index db509ab8874e..e2f4a4d93d20 100644
--- a/drivers/gpu/host1x/job.c
+++ b/drivers/gpu/host1x/job.c
@@ -34,8 +34,7 @@
 #define HOST1X_WAIT_SYNCPT_OFFSET 0x8
 
 struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
-				    u32 num_cmdbufs, u32 num_relocs,
-				    u32 num_waitchks)
+				    u32 num_cmdbufs, u32 num_relocs)
 {
 	struct host1x_job *job = NULL;
 	unsigned int num_unpins = num_cmdbufs + num_relocs;
@@ -46,7 +45,6 @@ struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
 	total = sizeof(struct host1x_job) +
 		(u64)num_relocs * sizeof(struct host1x_reloc) +
 		(u64)num_unpins * sizeof(struct host1x_job_unpin_data) +
-		(u64)num_waitchks * sizeof(struct host1x_waitchk) +
 		(u64)num_cmdbufs * sizeof(struct host1x_job_gather) +
 		(u64)num_unpins * sizeof(dma_addr_t) +
 		(u64)num_unpins * sizeof(u32 *);
@@ -62,12 +60,10 @@ struct host1x_job *host1x_job_alloc(struct host1x_channel *ch,
 
 	/* Redistribute memory to the structs  */
 	mem += sizeof(struct host1x_job);
-	job->relocarray = num_relocs ? mem : NULL;
+	job->relocs = num_relocs ? mem : NULL;
 	mem += num_relocs * sizeof(struct host1x_reloc);
 	job->unpins = num_unpins ? mem : NULL;
 	mem += num_unpins * sizeof(struct host1x_job_unpin_data);
-	job->waitchk = num_waitchks ? mem : NULL;
-	mem += num_waitchks * sizeof(struct host1x_waitchk);
 	job->gathers = num_cmdbufs ? mem : NULL;
 	mem += num_cmdbufs * sizeof(struct host1x_job_gather);
 	job->addr_phys = num_unpins ? mem : NULL;
@@ -100,84 +96,18 @@ void host1x_job_put(struct host1x_job *job)
 EXPORT_SYMBOL(host1x_job_put);
 
 void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *bo,
-			   u32 words, u32 offset)
+			   unsigned int words, unsigned int offset)
 {
-	struct host1x_job_gather *cur_gather = &job->gathers[job->num_gathers];
+	struct host1x_job_gather *gather = &job->gathers[job->num_gathers];
+
+	gather->words = words;
+	gather->bo = bo;
+	gather->offset = offset;
 
-	cur_gather->words = words;
-	cur_gather->bo = bo;
-	cur_gather->offset = offset;
 	job->num_gathers++;
 }
 EXPORT_SYMBOL(host1x_job_add_gather);
 
-/*
- * NULL an already satisfied WAIT_SYNCPT host method, by patching its
- * args in the command stream. The method data is changed to reference
- * a reserved (never given out or incr) HOST1X_SYNCPT_RESERVED syncpt
- * with a matching threshold value of 0, so is guaranteed to be popped
- * by the host HW.
- */
-static void host1x_syncpt_patch_offset(struct host1x_syncpt *sp,
-				       struct host1x_bo *h, u32 offset)
-{
-	void *patch_addr = NULL;
-
-	/* patch the wait */
-	patch_addr = host1x_bo_kmap(h, offset >> PAGE_SHIFT);
-	if (patch_addr) {
-		host1x_syncpt_patch_wait(sp,
-					 patch_addr + (offset & ~PAGE_MASK));
-		host1x_bo_kunmap(h, offset >> PAGE_SHIFT, patch_addr);
-	} else
-		pr_err("Could not map cmdbuf for wait check\n");
-}
-
-/*
- * Check driver supplied waitchk structs for syncpt thresholds
- * that have already been satisfied and NULL the comparison (to
- * avoid a wrap condition in the HW).
- */
-static int do_waitchks(struct host1x_job *job, struct host1x *host,
-		       struct host1x_job_gather *g)
-{
-	struct host1x_bo *patch = g->bo;
-	int i;
-
-	/* compare syncpt vs wait threshold */
-	for (i = 0; i < job->num_waitchk; i++) {
-		struct host1x_waitchk *wait = &job->waitchk[i];
-		struct host1x_syncpt *sp =
-			host1x_syncpt_get(host, wait->syncpt_id);
-
-		/* validate syncpt id */
-		if (wait->syncpt_id > host1x_syncpt_nb_pts(host))
-			continue;
-
-		/* skip all other gathers */
-		if (patch != wait->bo)
-			continue;
-
-		trace_host1x_syncpt_wait_check(wait->bo, wait->offset,
-					       wait->syncpt_id, wait->thresh,
-					       host1x_syncpt_read_min(sp));
-
-		if (host1x_syncpt_is_expired(sp, wait->thresh)) {
-			dev_dbg(host->dev,
-				"drop WAIT id %u (%s) thresh 0x%x, min 0x%x\n",
-				wait->syncpt_id, sp->name, wait->thresh,
-				host1x_syncpt_read_min(sp));
-
-			host1x_syncpt_patch_offset(sp, patch,
-						   g->offset + wait->offset);
-		}
-
-		wait->bo = NULL;
-	}
-
-	return 0;
-}
-
 static unsigned int pin_job(struct host1x *host, struct host1x_job *job)
 {
 	unsigned int i;
@@ -186,7 +116,7 @@ static unsigned int pin_job(struct host1x *host, struct host1x_job *job)
 	job->num_unpins = 0;
 
 	for (i = 0; i < job->num_relocs; i++) {
-		struct host1x_reloc *reloc = &job->relocarray[i];
+		struct host1x_reloc *reloc = &job->relocs[i];
 		struct sg_table *sgt;
 		dma_addr_t phys_addr;
 
@@ -267,14 +197,14 @@ unpin:
 
 static int do_relocs(struct host1x_job *job, struct host1x_job_gather *g)
 {
-	int i = 0;
 	u32 last_page = ~0;
 	void *cmdbuf_page_addr = NULL;
 	struct host1x_bo *cmdbuf = g->bo;
+	unsigned int i;
 
 	/* pin & patch the relocs for one gather */
 	for (i = 0; i < job->num_relocs; i++) {
-		struct host1x_reloc *reloc = &job->relocarray[i];
+		struct host1x_reloc *reloc = &job->relocs[i];
 		u32 reloc_addr = (job->reloc_addr_phys[i] +
 				  reloc->target.offset) >> reloc->shift;
 		u32 *target;
@@ -331,17 +261,6 @@ static bool check_reloc(struct host1x_reloc *reloc, struct host1x_bo *cmdbuf,
 	return true;
 }
 
-static bool check_wait(struct host1x_waitchk *wait, struct host1x_bo *cmdbuf,
-		       unsigned int offset)
-{
-	offset *= sizeof(u32);
-
-	if (wait->bo != cmdbuf || wait->offset != offset)
-		return false;
-
-	return true;
-}
-
 struct host1x_firewall {
 	struct host1x_job *job;
 	struct device *dev;
@@ -349,9 +268,6 @@ struct host1x_firewall {
 	unsigned int num_relocs;
 	struct host1x_reloc *reloc;
 
-	unsigned int num_waitchks;
-	struct host1x_waitchk *waitchk;
-
 	struct host1x_bo *cmdbuf;
 	unsigned int offset;
 
@@ -378,20 +294,6 @@ static int check_register(struct host1x_firewall *fw, unsigned long offset)
 		fw->reloc++;
 	}
 
-	if (offset == HOST1X_WAIT_SYNCPT_OFFSET) {
-		if (fw->class != HOST1X_CLASS_HOST1X)
-			return -EINVAL;
-
-		if (!fw->num_waitchks)
-			return -EINVAL;
-
-		if (!check_wait(fw->waitchk, fw->cmdbuf, fw->offset))
-			return -EINVAL;
-
-		fw->num_waitchks--;
-		fw->waitchk++;
-	}
-
 	return 0;
 }
 
@@ -550,14 +452,12 @@ static inline int copy_gathers(struct host1x_job *job, struct device *dev)
 	struct host1x_firewall fw;
 	size_t size = 0;
 	size_t offset = 0;
-	int i;
+	unsigned int i;
 
 	fw.job = job;
 	fw.dev = dev;
-	fw.reloc = job->relocarray;
+	fw.reloc = job->relocs;
 	fw.num_relocs = job->num_relocs;
-	fw.waitchk = job->waitchk;
-	fw.num_waitchks = job->num_waitchk;
 	fw.class = job->class;
 
 	for (i = 0; i < job->num_gathers; i++) {
@@ -604,8 +504,8 @@ static inline int copy_gathers(struct host1x_job *job, struct device *dev)
 		offset += g->words * sizeof(u32);
 	}
 
-	/* No relocs and waitchks should remain at this point */
-	if (fw.num_relocs || fw.num_waitchks)
+	/* No relocs should remain at this point */
+	if (fw.num_relocs)
 		return -EINVAL;
 
 	return 0;
@@ -616,19 +516,6 @@ int host1x_job_pin(struct host1x_job *job, struct device *dev)
 	int err;
 	unsigned int i, j;
 	struct host1x *host = dev_get_drvdata(dev->parent);
-	DECLARE_BITMAP(waitchk_mask, host1x_syncpt_nb_pts(host));
-
-	bitmap_zero(waitchk_mask, host1x_syncpt_nb_pts(host));
-	for (i = 0; i < job->num_waitchk; i++) {
-		u32 syncpt_id = job->waitchk[i].syncpt_id;
-
-		if (syncpt_id < host1x_syncpt_nb_pts(host))
-			set_bit(syncpt_id, waitchk_mask);
-	}
-
-	/* get current syncpt values for waitchk */
-	for_each_set_bit(i, waitchk_mask, host1x_syncpt_nb_pts(host))
-		host1x_syncpt_load(host->syncpt + i);
 
 	/* pin memory */
 	err = pin_job(host, job);
@@ -663,10 +550,6 @@ int host1x_job_pin(struct host1x_job *job, struct device *dev)
 		err = do_relocs(job, g);
 		if (err)
 			break;
-
-		err = do_waitchks(job, host, g);
-		if (err)
-			break;
 	}
 
 out:
diff --git a/drivers/gpu/host1x/job.h b/drivers/gpu/host1x/job.h
index 4bda51d503ec..188400e00192 100644
--- a/drivers/gpu/host1x/job.h
+++ b/drivers/gpu/host1x/job.h
@@ -20,10 +20,10 @@
 #define __HOST1X_JOB_H
 
 struct host1x_job_gather {
-	u32 words;
+	unsigned int words;
 	dma_addr_t base;
 	struct host1x_bo *bo;
-	u32 offset;
+	unsigned int offset;
 	bool handled;
 };
 
diff --git a/drivers/gpu/host1x/syncpt.c b/drivers/gpu/host1x/syncpt.c
index a2a952adc136..a5dbf1ba4645 100644
--- a/drivers/gpu/host1x/syncpt.c
+++ b/drivers/gpu/host1x/syncpt.c
@@ -57,8 +57,8 @@ static struct host1x_syncpt *host1x_syncpt_alloc(struct host1x *host,
 						 struct host1x_client *client,
 						 unsigned long flags)
 {
-	int i;
 	struct host1x_syncpt *sp = host->syncpt;
+	unsigned int i;
 	char *name;
 
 	mutex_lock(&host->syncpt_mutex);
@@ -255,7 +255,7 @@ int host1x_syncpt_wait(struct host1x_syncpt *sp, u32 thresh, long timeout,
 	}
 
 	/* schedule a wakeup when the syncpoint value is reached */
-	err = host1x_intr_add_action(sp->host, sp->id, thresh,
+	err = host1x_intr_add_action(sp->host, sp, thresh,
 				     HOST1X_INTR_ACTION_WAKEUP_INTERRUPTIBLE,
 				     &wq, waiter, &ref);
 	if (err)
@@ -373,12 +373,6 @@ bool host1x_syncpt_is_expired(struct host1x_syncpt *sp, u32 thresh)
 		return (s32)(current_val - thresh) >= 0;
 }
 
-/* remove a wait pointed to by patch_addr */
-int host1x_syncpt_patch_wait(struct host1x_syncpt *sp, void *patch_addr)
-{
-	return host1x_hw_syncpt_patch_wait(sp->host, sp, patch_addr);
-}
-
 int host1x_syncpt_init(struct host1x *host)
 {
 	struct host1x_syncpt_base *bases;
diff --git a/drivers/gpu/host1x/syncpt.h b/drivers/gpu/host1x/syncpt.h
index 9d88d37c2397..d98e22325e9d 100644
--- a/drivers/gpu/host1x/syncpt.h
+++ b/drivers/gpu/host1x/syncpt.h
@@ -124,7 +124,4 @@ static inline int host1x_syncpt_is_valid(struct host1x_syncpt *sp)
 	return sp->id < host1x_syncpt_nb_pts(sp->host);
 }
 
-/* Patch a wait by replacing it with a wait for syncpt 0 value 0 */
-int host1x_syncpt_patch_wait(struct host1x_syncpt *sp, void *patch_addr);
-
 #endif
diff --git a/drivers/gpu/vga/Kconfig b/drivers/gpu/vga/Kconfig
index 29437eabe095..b677e5d524e6 100644
--- a/drivers/gpu/vga/Kconfig
+++ b/drivers/gpu/vga/Kconfig
@@ -6,7 +6,7 @@ config VGA_ARB
 	  Some "legacy" VGA devices implemented on PCI typically have the same
 	  hard-decoded addresses as they did on ISA. When multiple PCI devices
 	  are accessed at same time they need some kind of coordination. Please
-	  see Documentation/vgaarbiter.txt for more details. Select this to
+	  see Documentation/gpu/vgaarbiter.rst for more details. Select this to
 	  enable VGA arbiter.
 
 config VGA_ARB_MAX_GPUS
diff --git a/drivers/gpu/vga/vgaarb.c b/drivers/gpu/vga/vgaarb.c
index 1c5e74cb9279..c61b04555779 100644
--- a/drivers/gpu/vga/vgaarb.c
+++ b/drivers/gpu/vga/vgaarb.c
@@ -1,6 +1,6 @@
 /*
  * vgaarb.c: Implements the VGA arbitration. For details refer to
- * Documentation/vgaarbiter.txt
+ * Documentation/gpu/vgaarbiter.rst
  *
  *
  * (C) Copyright 2005 Benjamin Herrenschmidt <benh@kernel.crashing.org>