diff options
Diffstat (limited to 'drivers/gpu/drm/amd')
81 files changed, 1141 insertions, 420 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/Kconfig b/drivers/gpu/drm/amd/amdgpu/Kconfig index a04f2fc7bf37..f6e5c0282fc1 100644 --- a/drivers/gpu/drm/amd/amdgpu/Kconfig +++ b/drivers/gpu/drm/amd/amdgpu/Kconfig @@ -27,8 +27,7 @@ config DRM_AMDGPU_CIK config DRM_AMDGPU_USERPTR bool "Always enable userptr write support" depends on DRM_AMDGPU - depends on ARCH_HAS_HMM - select HMM_MIRROR + depends on HMM_MIRROR help This option selects CONFIG_HMM and CONFIG_HMM_MIRROR if it isn't already selected to enabled full userptr support. diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile index 7bbcf1b20cfd..56e084367b93 100644 --- a/drivers/gpu/drm/amd/amdgpu/Makefile +++ b/drivers/gpu/drm/amd/amdgpu/Makefile @@ -54,7 +54,9 @@ amdgpu-y += amdgpu_device.o amdgpu_kms.o \ amdgpu_gtt_mgr.o amdgpu_vram_mgr.o amdgpu_virt.o amdgpu_atomfirmware.o \ amdgpu_vf_error.o amdgpu_sched.o amdgpu_debugfs.o amdgpu_ids.o \ amdgpu_gmc.o amdgpu_xgmi.o amdgpu_csa.o amdgpu_ras.o amdgpu_vm_cpu.o \ - amdgpu_vm_sdma.o amdgpu_pmu.o amdgpu_discovery.o + amdgpu_vm_sdma.o amdgpu_discovery.o + +amdgpu-$(CONFIG_PERF_EVENTS) += amdgpu_pmu.o # add asic specific block amdgpu-$(CONFIG_DRM_AMDGPU_CIK)+= cik.o cik_ih.o kv_smc.o kv_dpm.o \ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 596f7e07b5a8..8199d201b43a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -164,6 +164,7 @@ extern int amdgpu_async_gfx_ring; extern int amdgpu_mcbp; extern int amdgpu_discovery; extern int amdgpu_mes; +extern int amdgpu_noretry; #ifdef CONFIG_DRM_AMDGPU_SI extern int amdgpu_si_support; @@ -775,6 +776,7 @@ struct amdgpu_device { struct mutex grbm_idx_mutex; struct dev_pm_domain vga_pm_domain; bool have_disp_power_ref; + bool have_atomics_support; /* BIOS */ bool is_atom_fw; @@ -1216,6 +1218,10 @@ int amdgpu_dm_display_resume(struct amdgpu_device *adev ); static inline int amdgpu_dm_display_resume(struct amdgpu_device *adev) { return 0; } #endif + +void amdgpu_register_gpu_instance(struct amdgpu_device *adev); +void amdgpu_unregister_gpu_instance(struct amdgpu_device *adev); + #include "amdgpu_object.h" /* used by df_v3_6.c and amdgpu_pmu.c */ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index fab3eb173b05..9fa4f25a3745 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -668,6 +668,13 @@ bool amdgpu_amdkfd_is_kfd_vmid(struct amdgpu_device *adev, u32 vmid) return false; } +bool amdgpu_amdkfd_have_atomics_support(struct kgd_dev *kgd) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)kgd; + + return adev->have_atomics_support; +} + #ifndef CONFIG_HSA_AMD bool amdkfd_fence_check_mm(struct dma_fence *f, struct mm_struct *mm) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index 93a25c799d75..b6076d19e442 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -135,6 +135,7 @@ 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); void amdgpu_amdkfd_set_compute_idle(struct kgd_dev *kgd, bool idle); +bool amdgpu_amdkfd_have_atomics_support(struct kgd_dev *kgd); struct kfd2kgd_calls *amdgpu_amdkfd_gfx_7_get_functions(void); struct kfd2kgd_calls *amdgpu_amdkfd_gfx_8_0_get_functions(void); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index 0aa81456ec32..1d3ee9c42f7e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -504,7 +504,7 @@ static int init_user_pages(struct kgd_mem *mem, struct mm_struct *mm, goto out; } - ret = amdgpu_ttm_tt_get_user_pages(bo->tbo.ttm, bo->tbo.ttm->pages); + ret = amdgpu_ttm_tt_get_user_pages(bo, bo->tbo.ttm->pages); if (ret) { pr_err("%s: Failed to get user pages: %d\n", __func__, ret); goto unregister_out; @@ -813,7 +813,7 @@ static int process_sync_pds_resv(struct amdkfd_process_info *process_info, ret = amdgpu_sync_resv(NULL, sync, pd->tbo.resv, - AMDGPU_FENCE_OWNER_UNDEFINED, false); + AMDGPU_FENCE_OWNER_KFD, false); if (ret) return ret; } @@ -1729,8 +1729,7 @@ static int update_invalid_user_pages(struct amdkfd_process_info *process_info, bo = mem->bo; /* Get updated user pages */ - ret = amdgpu_ttm_tt_get_user_pages(bo->tbo.ttm, - bo->tbo.ttm->pages); + ret = amdgpu_ttm_tt_get_user_pages(bo, bo->tbo.ttm->pages); if (ret) { pr_debug("%s: Failed to get user pages: %d\n", __func__, ret); @@ -1740,6 +1739,12 @@ static int update_invalid_user_pages(struct amdkfd_process_info *process_info, } amdgpu_ttm_tt_get_user_pages_done(bo->tbo.ttm); + + /* Mark the BO as valid unless it was invalidated + * again concurrently. + */ + if (atomic_cmpxchg(&mem->invalid, invalid, 0) != invalid) + return -EAGAIN; } return 0; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 37adce981fa3..e069de8b54e6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -633,7 +633,7 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, return -ENOMEM; } - r = amdgpu_ttm_tt_get_user_pages(bo->tbo.ttm, e->user_pages); + r = amdgpu_ttm_tt_get_user_pages(bo, e->user_pages); if (r) { kvfree(e->user_pages); e->user_pages = NULL; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c index 20ce158490db..6d54decef7f8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c @@ -106,10 +106,10 @@ static int amdgpu_debugfs_process_reg_op(bool read, struct file *f, ssize_t result = 0; int r; bool pm_pg_lock, use_bank, use_ring; - unsigned instance_bank, sh_bank, se_bank, me, pipe, queue; + unsigned instance_bank, sh_bank, se_bank, me, pipe, queue, vmid; pm_pg_lock = use_bank = use_ring = false; - instance_bank = sh_bank = se_bank = me = pipe = queue = 0; + instance_bank = sh_bank = se_bank = me = pipe = queue = vmid = 0; if (size & 0x3 || *pos & 0x3 || ((*pos & (1ULL << 62)) && (*pos & (1ULL << 61)))) @@ -135,6 +135,7 @@ static int amdgpu_debugfs_process_reg_op(bool read, struct file *f, me = (*pos & GENMASK_ULL(33, 24)) >> 24; pipe = (*pos & GENMASK_ULL(43, 34)) >> 34; queue = (*pos & GENMASK_ULL(53, 44)) >> 44; + vmid = (*pos & GENMASK_ULL(58, 54)) >> 54; use_ring = 1; } else { @@ -152,7 +153,7 @@ static int amdgpu_debugfs_process_reg_op(bool read, struct file *f, sh_bank, instance_bank); } else if (use_ring) { mutex_lock(&adev->srbm_mutex); - amdgpu_gfx_select_me_pipe_q(adev, me, pipe, queue); + amdgpu_gfx_select_me_pipe_q(adev, me, pipe, queue, vmid); } if (pm_pg_lock) @@ -185,7 +186,7 @@ end: amdgpu_gfx_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); mutex_unlock(&adev->grbm_idx_mutex); } else if (use_ring) { - amdgpu_gfx_select_me_pipe_q(adev, 0, 0, 0); + amdgpu_gfx_select_me_pipe_q(adev, 0, 0, 0, 0); mutex_unlock(&adev->srbm_mutex); } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index e886be292f86..5a7f893cf724 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -2537,6 +2537,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, hash_init(adev->mn_hash); mutex_init(&adev->lock_reset); mutex_init(&adev->virt.dpm_mutex); + mutex_init(&adev->psp.mutex); r = amdgpu_device_check_arguments(adev); if (r) @@ -2596,6 +2597,17 @@ int amdgpu_device_init(struct amdgpu_device *adev, if (adev->rio_mem == NULL) DRM_INFO("PCI I/O BAR is not found.\n"); + /* enable PCIE atomic ops */ + r = pci_enable_atomic_ops_to_root(adev->pdev, + PCI_EXP_DEVCAP2_ATOMIC_COMP32 | + PCI_EXP_DEVCAP2_ATOMIC_COMP64); + if (r) { + adev->have_atomics_support = false; + DRM_INFO("PCIE atomic ops is not supported\n"); + } else { + adev->have_atomics_support = true; + } + amdgpu_device_get_pcie_info(adev); if (amdgpu_mcbp) @@ -2604,7 +2616,7 @@ int amdgpu_device_init(struct amdgpu_device *adev, if (amdgpu_mes && adev->asic_type >= CHIP_NAVI10) adev->enable_mes = true; - if (amdgpu_discovery) { + if (amdgpu_discovery && adev->asic_type >= CHIP_NAVI10) { r = amdgpu_discovery_init(adev); if (r) { dev_err(adev->dev, "amdgpu_discovery_init failed\n"); @@ -2798,7 +2810,8 @@ fence_driver_init: return r; } - r = amdgpu_pmu_init(adev); + if (IS_ENABLED(CONFIG_PERF_EVENTS)) + r = amdgpu_pmu_init(adev); if (r) dev_err(adev->dev, "amdgpu_pmu_init failed\n"); @@ -2870,9 +2883,10 @@ void amdgpu_device_fini(struct amdgpu_device *adev) amdgpu_debugfs_regs_cleanup(adev); device_remove_file(adev->dev, &dev_attr_pcie_replay_count); amdgpu_ucode_sysfs_fini(adev); - amdgpu_pmu_fini(adev); + if (IS_ENABLED(CONFIG_PERF_EVENTS)) + amdgpu_pmu_fini(adev); amdgpu_debugfs_preempt_cleanup(adev); - if (amdgpu_discovery) + if (amdgpu_discovery && adev->asic_type >= CHIP_NAVI10) amdgpu_discovery_fini(adev); } @@ -3559,6 +3573,12 @@ static int amdgpu_do_asic_reset(struct amdgpu_hive_info *hive, if (vram_lost) amdgpu_device_fill_reset_magic(tmp_adev); + /* + * Add this ASIC as tracked as reset was already + * complete successfully. + */ + amdgpu_register_gpu_instance(tmp_adev); + r = amdgpu_device_ip_late_init(tmp_adev); if (r) goto out; @@ -3693,8 +3713,19 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, device_list_handle = &device_list; } + /* + * Mark these ASICs to be reseted as untracked first + * And add them back after reset completed + */ + list_for_each_entry(tmp_adev, device_list_handle, gmc.xgmi.head) + amdgpu_unregister_gpu_instance(tmp_adev); + /* block all schedulers and reset given job's ring */ list_for_each_entry(tmp_adev, device_list_handle, gmc.xgmi.head) { + /* disable ras on ALL IPs */ + if (amdgpu_device_ip_need_full_reset(tmp_adev)) + amdgpu_ras_suspend(tmp_adev); + for (i = 0; i < AMDGPU_MAX_RINGS; ++i) { struct amdgpu_ring *ring = tmp_adev->rings[i]; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c index e049ae6a76fb..1481899f86c1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c @@ -123,7 +123,7 @@ static int hw_id_map[MAX_HWIP] = { [UVD_HWIP] = UVD_HWID, [VCE_HWIP] = VCE_HWID, [DF_HWIP] = DF_HWID, - [DCE_HWIP] = DCEAZ_HWID, + [DCE_HWIP] = DMU_HWID, [OSSSYS_HWIP] = OSSSYS_HWID, [SMUIO_HWIP] = SMUIO_HWID, [PWR_HWIP] = PWR_HWID, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 3913a75924c6..f2e8b4238efd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -140,8 +140,9 @@ uint amdgpu_smu_memory_pool_size = 0; uint amdgpu_dc_feature_mask = 0; int amdgpu_async_gfx_ring = 1; int amdgpu_mcbp = 0; -int amdgpu_discovery = 0; +int amdgpu_discovery = -1; int amdgpu_mes = 0; +int amdgpu_noretry; struct amdgpu_mgpu_info mgpu_info = { .mutex = __MUTEX_INITIALIZER(mgpu_info.mutex), @@ -247,7 +248,8 @@ module_param_named(msi, amdgpu_msi, int, 0444); * By default(with no lockup_timeout settings), the timeout for all non-compute(GFX, SDMA and Video) * jobs is 10000. And there is no timeout enforced on compute jobs. */ -MODULE_PARM_DESC(lockup_timeout, "GPU lockup timeout in ms (default: 10000 for non-compute jobs and no timeout for compute jobs), " +MODULE_PARM_DESC(lockup_timeout, "GPU lockup timeout in ms (default: 10000 for non-compute jobs and infinity timeout for compute jobs." + " 0: keep default value. negative: infinity timeout), " "format is [Non-Compute] or [GFX,Compute,SDMA,Video]"); module_param_string(lockup_timeout, amdgpu_lockup_timeout, sizeof(amdgpu_lockup_timeout), 0444); @@ -581,18 +583,36 @@ MODULE_PARM_DESC(async_gfx_ring, "Asynchronous GFX rings that could be configured with either different priorities (HP3D ring and LP3D ring), or equal priorities (0 = disabled, 1 = enabled (default))"); module_param_named(async_gfx_ring, amdgpu_async_gfx_ring, int, 0444); +/** + * DOC: mcbp (int) + * It is used to enable mid command buffer preemption. (0 = disabled (default), 1 = enabled) + */ MODULE_PARM_DESC(mcbp, "Enable Mid-command buffer preemption (0 = disabled (default), 1 = enabled)"); module_param_named(mcbp, amdgpu_mcbp, int, 0444); +/** + * DOC: discovery (int) + * Allow driver to discover hardware IP information from IP Discovery table at the top of VRAM. + * (-1 = auto (default), 0 = disabled, 1 = enabled) + */ MODULE_PARM_DESC(discovery, "Allow driver to discover hardware IPs from IP Discovery table at the top of VRAM"); module_param_named(discovery, amdgpu_discovery, int, 0444); +/** + * DOC: mes (int) + * Enable Micro Engine Scheduler. This is a new hw scheduling engine for gfx, sdma, and compute. + * (0 = disabled (default), 1 = enabled) + */ MODULE_PARM_DESC(mes, "Enable Micro Engine Scheduler (0 = disabled (default), 1 = enabled)"); module_param_named(mes, amdgpu_mes, int, 0444); +MODULE_PARM_DESC(noretry, + "Disable retry faults (0 = retry enabled (default), 1 = retry disabled)"); +module_param_named(noretry, amdgpu_noretry, int, 0644); + #ifdef CONFIG_HSA_AMD /** * DOC: sched_policy (int) @@ -669,17 +689,6 @@ MODULE_PARM_DESC(ignore_crat, "Ignore CRAT table during KFD initialization (0 = use CRAT (default), 1 = ignore CRAT)"); /** - * DOC: noretry (int) - * This parameter sets sh_mem_config.retry_disable. Default value, 0, enables retry. - * Setting 1 disables retry. - * Retry is needed for recoverable page faults. - */ -int noretry; -module_param(noretry, int, 0644); -MODULE_PARM_DESC(noretry, - "Set sh_mem_config.retry_disable on Vega10 (0 = retry enabled (default), 1 = retry disabled)"); - -/** * DOC: halt_if_hws_hang (int) * Halt if HWS hang is detected. Default value, 0, disables the halt on hang. * Setting 1 enables halt on hang. @@ -1302,7 +1311,8 @@ int amdgpu_device_get_job_timeout_settings(struct amdgpu_device *adev) * By default timeout for non compute jobs is 10000. * And there is no timeout enforced on compute jobs. */ - adev->gfx_timeout = adev->sdma_timeout = adev->video_timeout = 10000; + adev->gfx_timeout = msecs_to_jiffies(10000); + adev->sdma_timeout = adev->video_timeout = adev->gfx_timeout; adev->compute_timeout = MAX_SCHEDULE_TIMEOUT; if (strnlen(input, AMDGPU_MAX_TIMEOUT_PARAM_LENTH)) { @@ -1312,10 +1322,13 @@ int amdgpu_device_get_job_timeout_settings(struct amdgpu_device *adev) if (ret) return ret; - /* Invalidate 0 and negative values */ - if (timeout <= 0) { + if (timeout == 0) { index++; continue; + } else if (timeout < 0) { + timeout = MAX_SCHEDULE_TIMEOUT; + } else { + timeout = msecs_to_jiffies(timeout); } switch (index++) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h index dad2186f4ed5..df8a23554831 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h @@ -31,7 +31,8 @@ struct amdgpu_gds { uint32_t gds_size; uint32_t gws_size; uint32_t oa_size; - uint32_t gds_compute_max_wave_id; + uint32_t gds_compute_max_wave_id; + uint32_t vgt_gs_max_wave_id; }; struct amdgpu_gds_reg_offset { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c index 1f9f27061e2f..939f8305511b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c @@ -327,8 +327,7 @@ int amdgpu_gem_userptr_ioctl(struct drm_device *dev, void *data, } if (args->flags & AMDGPU_GEM_USERPTR_VALIDATE) { - r = amdgpu_ttm_tt_get_user_pages(bo->tbo.ttm, - bo->tbo.ttm->pages); + r = amdgpu_ttm_tt_get_user_pages(bo, bo->tbo.ttm->pages); if (r) goto release_object; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h index f96407ba9770..1199b5828b90 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h @@ -195,7 +195,7 @@ struct amdgpu_gfx_funcs { uint32_t wave, uint32_t start, uint32_t size, uint32_t *dst); void (*select_me_pipe_q)(struct amdgpu_device *adev, u32 me, u32 pipe, - u32 queue); + u32 queue, u32 vmid); }; struct amdgpu_ngg_buf { @@ -327,7 +327,7 @@ struct amdgpu_gfx { #define amdgpu_gfx_get_gpu_clock_counter(adev) (adev)->gfx.funcs->get_gpu_clock_counter((adev)) #define amdgpu_gfx_select_se_sh(adev, se, sh, instance) (adev)->gfx.funcs->select_se_sh((adev), (se), (sh), (instance)) -#define amdgpu_gfx_select_me_pipe_q(adev, me, pipe, q) (adev)->gfx.funcs->select_me_pipe_q((adev), (me), (pipe), (q)) +#define amdgpu_gfx_select_me_pipe_q(adev, me, pipe, q, vmid) (adev)->gfx.funcs->select_me_pipe_q((adev), (me), (pipe), (q), (vmid)) /** * amdgpu_gfx_create_bitmask - create a bitmask diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index 5832cd8f4ff1..0cf7e8606fd3 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -44,7 +44,7 @@ #include "amdgpu_display.h" #include "amdgpu_ras.h" -static void amdgpu_unregister_gpu_instance(struct amdgpu_device *adev) +void amdgpu_unregister_gpu_instance(struct amdgpu_device *adev) { struct amdgpu_gpu_instance *gpu_instance; int i; @@ -105,7 +105,7 @@ done_free: dev->dev_private = NULL; } -static void amdgpu_register_gpu_instance(struct amdgpu_device *adev) +void amdgpu_register_gpu_instance(struct amdgpu_device *adev) { struct amdgpu_gpu_instance *gpu_instance; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c index 623f56a1485f..3971c201f320 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c @@ -45,49 +45,12 @@ #include <linux/firmware.h> #include <linux/module.h> -#include <linux/hmm.h> -#include <linux/interval_tree.h> - #include <drm/drm.h> #include "amdgpu.h" #include "amdgpu_amdkfd.h" /** - * struct amdgpu_mn - * - * @adev: amdgpu device pointer - * @mm: process address space - * @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 - * @mirror: HMM mirror function support - * - * Data for each amdgpu device and process address space. - */ -struct amdgpu_mn { - /* constant after initialisation */ - struct amdgpu_device *adev; - struct mm_struct *mm; - enum amdgpu_mn_type type; - - /* only used on destruction */ - struct work_struct work; - - /* protected by adev->mn_lock */ - struct hlist_node node; - - /* objects protected by lock */ - struct rw_semaphore lock; - struct rb_root_cached objects; - - /* HMM mirror */ - struct hmm_mirror mirror; -}; - -/** * struct amdgpu_mn_node * * @it: interval node defining start-last of the affected address range diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h index f5b67c63ed6b..b8ed68943625 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.h @@ -24,17 +24,53 @@ #ifndef __AMDGPU_MN_H__ #define __AMDGPU_MN_H__ -/* - * HMM mirror - */ -struct amdgpu_mn; -struct hmm_range; +#include <linux/types.h> +#include <linux/hmm.h> +#include <linux/rwsem.h> +#include <linux/workqueue.h> +#include <linux/interval_tree.h> enum amdgpu_mn_type { AMDGPU_MN_TYPE_GFX, AMDGPU_MN_TYPE_HSA, }; +/** + * struct amdgpu_mn + * + * @adev: amdgpu device pointer + * @mm: process address space + * @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 + * @mirror: HMM mirror function support + * + * Data for each amdgpu device and process address space. + */ +struct amdgpu_mn { + /* constant after initialisation */ + struct amdgpu_device *adev; + struct mm_struct *mm; + enum amdgpu_mn_type type; + + /* only used on destruction */ + struct work_struct work; + + /* protected by adev->mn_lock */ + struct hlist_node node; + + /* objects protected by lock */ + struct rw_semaphore lock; + struct rb_root_cached objects; + +#ifdef CONFIG_HMM_MIRROR + /* HMM mirror */ + struct hmm_mirror mirror; +#endif +}; + #if defined(CONFIG_HMM_MIRROR) void amdgpu_mn_lock(struct amdgpu_mn *mn); void amdgpu_mn_unlock(struct amdgpu_mn *mn); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c index 193d53720d9b..8b7efd0a7028 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c @@ -2077,11 +2077,6 @@ static ssize_t amdgpu_hwmon_show_sclk(struct device *dev, (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return -EINVAL; - /* sanity check PP is enabled */ - if (!(adev->powerplay.pp_funcs && - adev->powerplay.pp_funcs->read_sensor)) - return -EINVAL; - /* get the sclk */ r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GFX_SCLK, (void *)&sclk, &size); @@ -2112,11 +2107,6 @@ static ssize_t amdgpu_hwmon_show_mclk(struct device *dev, (ddev->switch_power_state != DRM_SWITCH_POWER_ON)) return -EINVAL; - /* sanity check PP is enabled */ - if (!(adev->powerplay.pp_funcs && - adev->powerplay.pp_funcs->read_sensor)) - return -EINVAL; - /* get the sclk */ r = amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_GFX_MCLK, (void *)&mclk, &size); @@ -2996,13 +2986,10 @@ void amdgpu_pm_compute_clocks(struct amdgpu_device *adev) } if (is_support_sw_smu(adev)) { - struct smu_context *smu = &adev->smu; struct smu_dpm_context *smu_dpm = &adev->smu.smu_dpm; - mutex_lock(&(smu->mutex)); smu_handle_task(&adev->smu, smu_dpm->dpm_level, AMD_PP_TASK_DISPLAY_CONFIG_CHANGE); - mutex_unlock(&(smu->mutex)); } else { if (adev->powerplay.pp_funcs->dispatch_tasks) { if (!amdgpu_device_has_dc_support(adev)) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index e69ad6e089c5..c027e5e7713e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -130,6 +130,8 @@ psp_cmd_submit_buf(struct psp_context *psp, int index; int timeout = 2000; + mutex_lock(&psp->mutex); + memset(psp->cmd_buf_mem, 0, PSP_CMD_BUFFER_SIZE); memcpy(psp->cmd_buf_mem, cmd, sizeof(struct psp_gfx_cmd_resp)); @@ -139,6 +141,7 @@ psp_cmd_submit_buf(struct psp_context *psp, fence_mc_addr, index); if (ret) { atomic_dec(&psp->fence_value); + mutex_unlock(&psp->mutex); return ret; } @@ -161,8 +164,10 @@ psp_cmd_submit_buf(struct psp_context *psp, ucode->ucode_id); DRM_WARN("psp command failed and response status is (%d)\n", psp->cmd_buf_mem->resp.status); - if (!timeout) + if (!timeout) { + mutex_unlock(&psp->mutex); return -EINVAL; + } } /* get xGMI session id from response buffer */ @@ -172,6 +177,7 @@ psp_cmd_submit_buf(struct psp_context *psp, ucode->tmr_mc_addr_lo = psp->cmd_buf_mem->resp.fw_addr_lo; ucode->tmr_mc_addr_hi = psp->cmd_buf_mem->resp.fw_addr_hi; } + mutex_unlock(&psp->mutex); return ret; } @@ -763,6 +769,15 @@ static int psp_hw_start(struct psp_context *psp) int ret; if (!amdgpu_sriov_vf(adev) || !adev->in_gpu_reset) { + if (psp->kdb_bin_size && + (psp->funcs->bootloader_load_kdb != NULL)) { + ret = psp_bootloader_load_kdb(psp); + if (ret) { + DRM_ERROR("PSP load kdb failed!\n"); + return ret; + } + } + ret = psp_bootloader_load_sysdrv(psp); if (ret) { DRM_ERROR("PSP load sysdrv failed!\n"); @@ -1188,10 +1203,16 @@ failed: int psp_gpu_reset(struct amdgpu_device *adev) { + int ret; + if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) return 0; - return psp_mode1_reset(&adev->psp); + mutex_lock(&adev->psp.mutex); + ret = psp_mode1_reset(&adev->psp); + mutex_unlock(&adev->psp.mutex); + + return ret; } int psp_rlc_autoload_start(struct psp_context *psp) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h index 6039acc84346..e0fc2a790e53 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h @@ -42,6 +42,12 @@ struct psp_context; struct psp_xgmi_node_info; struct psp_xgmi_topology_info; +enum psp_bootloader_cmd { + PSP_BL__LOAD_SYSDRV = 0x10000, + PSP_BL__LOAD_SOSDRV = 0x20000, + PSP_BL__LOAD_KEY_DATABASE = 0x80000, +}; + enum psp_ring_type { PSP_RING_TYPE__INVALID = 0, @@ -73,6 +79,7 @@ enum psp_reg_prog_id { struct psp_funcs { int (*init_microcode)(struct psp_context *psp); + int (*bootloader_load_kdb)(struct psp_context *psp); int (*bootloader_load_sysdrv)(struct psp_context *psp); int (*bootloader_load_sos)(struct psp_context *psp); int (*ring_init)(struct psp_context *psp, enum psp_ring_type ring_type); @@ -156,9 +163,11 @@ struct psp_context uint32_t sys_bin_size; uint32_t sos_bin_size; uint32_t toc_bin_size; + uint32_t kdb_bin_size; uint8_t *sys_start_addr; uint8_t *sos_start_addr; uint8_t *toc_start_addr; + uint8_t *kdb_start_addr; /* tmr buffer */ struct amdgpu_bo *tmr_bo; @@ -201,6 +210,7 @@ struct psp_context uint8_t *ta_ras_start_addr; struct psp_xgmi_context xgmi_context; struct psp_ras_context ras; + struct mutex mutex; }; struct amdgpu_psp_funcs { @@ -219,6 +229,8 @@ struct amdgpu_psp_funcs { (psp)->funcs->compare_sram_data((psp), (ucode), (type)) #define psp_init_microcode(psp) \ ((psp)->funcs->init_microcode ? (psp)->funcs->init_microcode((psp)) : 0) +#define psp_bootloader_load_kdb(psp) \ + ((psp)->funcs->bootloader_load_kdb ? (psp)->funcs->bootloader_load_kdb((psp)) : 0) #define psp_bootloader_load_sysdrv(psp) \ ((psp)->funcs->bootloader_load_sysdrv ? (psp)->funcs->bootloader_load_sysdrv((psp)) : 0) #define psp_bootloader_load_sos(psp) \ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index c9faa69cd677..e51b48ac48eb 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -391,6 +391,7 @@ int amdgpu_ttm_copy_mem_to_mem(struct amdgpu_device *adev, src_node_start = amdgpu_mm_node_addr(src->bo, ++src_mm, src->mem); src_node_size = (src_mm->size << PAGE_SHIFT); + src_page_offset = 0; } else { src_node_start += cur_size; src_page_offset = src_node_start & (PAGE_SIZE - 1); @@ -400,6 +401,7 @@ int amdgpu_ttm_copy_mem_to_mem(struct amdgpu_device *adev, dst_node_start = amdgpu_mm_node_addr(dst->bo, ++dst_mm, dst->mem); dst_node_size = (dst_mm->size << PAGE_SHIFT); + dst_page_offset = 0; } else { dst_node_start += cur_size; dst_page_offset = dst_node_start & (PAGE_SIZE - 1); @@ -487,6 +489,7 @@ static int amdgpu_move_vram_ram(struct ttm_buffer_object *bo, bool evict, placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT; r = ttm_bo_mem_space(bo, &placement, &tmp_mem, ctx); if (unlikely(r)) { + pr_err("Failed to find GTT space for blit from VRAM\n"); return r; } @@ -545,6 +548,7 @@ static int amdgpu_move_ram_vram(struct ttm_buffer_object *bo, bool evict, placements.flags = TTM_PL_MASK_CACHING | TTM_PL_FLAG_TT; r = ttm_bo_mem_space(bo, &placement, &tmp_mem, ctx); if (unlikely(r)) { + pr_err("Failed to find GTT space for blit to VRAM\n"); return r; } @@ -565,6 +569,30 @@ out_cleanup: } /** + * amdgpu_mem_visible - Check that memory can be accessed by ttm_bo_move_memcpy + * + * Called by amdgpu_bo_move() + */ +static bool amdgpu_mem_visible(struct amdgpu_device *adev, + struct ttm_mem_reg *mem) +{ + struct drm_mm_node *nodes = mem->mm_node; + + if (mem->mem_type == TTM_PL_SYSTEM || + mem->mem_type == TTM_PL_TT) + return true; + if (mem->mem_type != TTM_PL_VRAM) + return false; + + /* ttm_mem_reg_ioremap only supports contiguous memory */ + if (nodes->size != mem->num_pages) + return false; + + return ((nodes->start + nodes->size) << PAGE_SHIFT) + <= adev->gmc.visible_vram_size; +} + +/** * amdgpu_bo_move - Move a buffer object to a new memory location * * Called by ttm_bo_handle_move_mem() @@ -608,8 +636,10 @@ static int amdgpu_bo_move(struct ttm_buffer_object *bo, bool evict, return 0; } - if (!adev->mman.buffer_funcs_enabled) + if (!adev->mman.buffer_funcs_enabled) { + r = -ENODEV; goto memcpy; + } if (old_mem->mem_type == TTM_PL_VRAM && new_mem->mem_type == TTM_PL_SYSTEM) { @@ -624,10 +654,16 @@ static int amdgpu_bo_move(struct ttm_buffer_object *bo, bool evict, if (r) { memcpy: - r = ttm_bo_move_memcpy(bo, ctx, new_mem); - if (r) { + /* Check that all memory is CPU accessible */ + if (!amdgpu_mem_visible(adev, old_mem) || + !amdgpu_mem_visible(adev, new_mem)) { + pr_err("Move buffer fallback to memcpy unavailable\n"); return r; } + + r = ttm_bo_move_memcpy(bo, ctx, new_mem); + if (r) + return r; } if (bo->type == ttm_bo_type_device && @@ -731,8 +767,10 @@ struct amdgpu_ttm_tt { #define MAX_RETRY_HMM_RANGE_FAULT 16 -int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages) +int amdgpu_ttm_tt_get_user_pages(struct amdgpu_bo *bo, struct page **pages) { + struct hmm_mirror *mirror = bo->mn ? &bo->mn->mirror : NULL; + struct ttm_tt *ttm = bo->tbo.ttm; struct amdgpu_ttm_tt *gtt = (void *)ttm; struct mm_struct *mm = gtt->usertask->mm; unsigned long start = gtt->userptr; @@ -746,6 +784,12 @@ int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages) if (!mm) /* Happens during process shutdown */ return -ESRCH; + if (unlikely(!mirror)) { + DRM_DEBUG_DRIVER("Failed to get hmm_mirror\n"); + r = -EFAULT; + goto out; + } + vma = find_vma(mm, start); if (unlikely(!vma || start < vma->vm_start)) { r = -EFAULT; @@ -775,7 +819,7 @@ int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages) 0 : range->flags[HMM_PFN_WRITE]; range->pfn_flags_mask = 0; range->pfns = pfns; - hmm_range_register(range, mm, start, + hmm_range_register(range, mirror, start, start + ttm->num_pages * PAGE_SIZE, PAGE_SHIFT); retry: @@ -2051,9 +2095,9 @@ int amdgpu_fill_buffer(struct amdgpu_bo *bo, mm_node = bo->tbo.mem.mm_node; num_loops = 0; while (num_pages) { - uint32_t byte_count = mm_node->size << PAGE_SHIFT; + uint64_t byte_count = mm_node->size << PAGE_SHIFT; - num_loops += DIV_ROUND_UP(byte_count, max_bytes); + num_loops += DIV_ROUND_UP_ULL(byte_count, max_bytes); num_pages -= mm_node->size; ++mm_node; } @@ -2079,12 +2123,13 @@ int amdgpu_fill_buffer(struct amdgpu_bo *bo, mm_node = bo->tbo.mem.mm_node; while (num_pages) { - uint32_t byte_count = mm_node->size << PAGE_SHIFT; + uint64_t byte_count = mm_node->size << PAGE_SHIFT; uint64_t dst_addr; dst_addr = amdgpu_mm_node_addr(&bo->tbo, mm_node, &bo->tbo.mem); while (byte_count) { - uint32_t cur_size_in_bytes = min(byte_count, max_bytes); + uint32_t cur_size_in_bytes = min_t(uint64_t, byte_count, + max_bytes); amdgpu_emit_fill_buffer(adev, &job->ibs[0], src_data, dst_addr, cur_size_in_bytes); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h index c2b7669004ba..caa76c693700 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h @@ -102,10 +102,11 @@ int amdgpu_ttm_alloc_gart(struct ttm_buffer_object *bo); int amdgpu_ttm_recover_gart(struct ttm_buffer_object *tbo); #if IS_ENABLED(CONFIG_DRM_AMDGPU_USERPTR) -int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages); +int amdgpu_ttm_tt_get_user_pages(struct amdgpu_bo *bo, struct page **pages); bool amdgpu_ttm_tt_get_user_pages_done(struct ttm_tt *ttm); #else -static inline int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages) +static inline int amdgpu_ttm_tt_get_user_pages(struct amdgpu_bo *bo, + struct page **pages) { return -EPERM; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c index c352a519ddd4..bfaa0eac3213 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c @@ -262,6 +262,12 @@ void amdgpu_ucode_print_psp_hdr(const struct common_firmware_header *hdr) le32_to_cpu(psp_hdr_v1_1->toc_offset_bytes)); DRM_DEBUG("toc_size_bytes: %u\n", le32_to_cpu(psp_hdr_v1_1->toc_size_bytes)); + DRM_DEBUG("kdb_header_version: %u\n", + le32_to_cpu(psp_hdr_v1_1->kdb_header_version)); + DRM_DEBUG("kdb_offset_bytes: %u\n", + le32_to_cpu(psp_hdr_v1_1->kdb_offset_bytes)); + DRM_DEBUG("kdb_size_bytes: %u\n", + le32_to_cpu(psp_hdr_v1_1->kdb_size_bytes)); } } else { DRM_ERROR("Unknown PSP ucode version: %u.%u\n", diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h index f46944453c6e..c1fb6dc86440 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h @@ -85,6 +85,9 @@ struct psp_firmware_header_v1_1 { uint32_t toc_header_version; uint32_t toc_offset_bytes; uint32_t toc_size_bytes; + uint32_t kdb_header_version; + uint32_t kdb_offset_bytes; + uint32_t kdb_size_bytes; }; /* version_major=1, version_minor=0 */ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index 07a7e3820b7b..59dd204498c5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -390,7 +390,8 @@ static uint32_t parse_clk(char *buf, bool min) if (!ptr) break; ptr+=2; - clk = simple_strtoul(ptr, NULL, 10); + if (kstrtou32(ptr, 10, &clk)) + return 0; } while (!min); return clk * 100; diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c index 2932ade7dbd0..1675d5837c3c 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c @@ -109,6 +109,13 @@ static const struct soc15_reg_golden golden_settings_gc_10_0_nv10[] = /* Pending on emulation bring up */ }; +#define DEFAULT_SH_MEM_CONFIG \ + ((SH_MEM_ADDRESS_MODE_64 << SH_MEM_CONFIG__ADDRESS_MODE__SHIFT) | \ + (SH_MEM_ALIGNMENT_MODE_UNALIGNED << SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT) | \ + (SH_MEM_RETRY_MODE_ALL << SH_MEM_CONFIG__RETRY_MODE__SHIFT) | \ + (3 << SH_MEM_CONFIG__INITIAL_INST_PREFETCH__SHIFT)) + + static void gfx_v10_0_set_ring_funcs(struct amdgpu_device *adev); static void gfx_v10_0_set_irq_funcs(struct amdgpu_device *adev); static void gfx_v10_0_set_gds_init(struct amdgpu_device *adev); @@ -995,6 +1002,12 @@ static void gfx_v10_0_read_wave_vgprs(struct amdgpu_device *adev, uint32_t simd, start + SQIND_WAVE_VGPRS_OFFSET, size, dst); } +static void gfx_v10_0_select_me_pipe_q(struct amdgpu_device *adev, + u32 me, u32 pipe, u32 q, u32 vm) + { + nv_grbm_select(adev, me, pipe, q, vm); + } + static const struct amdgpu_gfx_funcs gfx_v10_0_gfx_funcs = { .get_gpu_clock_counter = &gfx_v10_0_get_gpu_clock_counter, @@ -1002,6 +1015,7 @@ static const struct amdgpu_gfx_funcs gfx_v10_0_gfx_funcs = { .read_wave_data = &gfx_v10_0_read_wave_data, .read_wave_sgprs = &gfx_v10_0_read_wave_sgprs, .read_wave_vgprs = &gfx_v10_0_read_wave_vgprs, + .select_me_pipe_q = &gfx_v10_0_select_me_pipe_q, }; static void gfx_v10_0_gpu_early_init(struct amdgpu_device *adev) @@ -1408,7 +1422,6 @@ static u32 gfx_v10_0_init_pa_sc_tile_steering_override(struct amdgpu_device *ade static void gfx_v10_0_init_compute_vmid(struct amdgpu_device *adev) { int i; - uint32_t sh_mem_config; uint32_t sh_mem_bases; /* @@ -1419,15 +1432,11 @@ static void gfx_v10_0_init_compute_vmid(struct amdgpu_device *adev) */ sh_mem_bases = DEFAULT_SH_MEM_BASES | (DEFAULT_SH_MEM_BASES << 16); - sh_mem_config = SH_MEM_ADDRESS_MODE_64 | - SH_MEM_ALIGNMENT_MODE_UNALIGNED << - SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT; - mutex_lock(&adev->srbm_mutex); for (i = FIRST_COMPUTE_VMID; i < LAST_COMPUTE_VMID; i++) { nv_grbm_select(adev, 0, 0, 0, i); /* CP and shaders */ - WREG32_SOC15(GC, 0, mmSH_MEM_CONFIG, sh_mem_config); + WREG32_SOC15(GC, 0, mmSH_MEM_CONFIG, DEFAULT_SH_MEM_CONFIG); WREG32_SOC15(GC, 0, mmSH_MEM_BASES, sh_mem_bases); } nv_grbm_select(adev, 0, 0, 0, 0); @@ -1520,17 +1529,8 @@ static void gfx_v10_0_constants_init(struct amdgpu_device *adev) for (i = 0; i < adev->vm_manager.id_mgr[AMDGPU_GFXHUB].num_ids; i++) { nv_grbm_select(adev, 0, 0, 0, i); /* CP and shaders */ - if (i == 0) { - tmp = REG_SET_FIELD(0, SH_MEM_CONFIG, ALIGNMENT_MODE, - SH_MEM_ALIGNMENT_MODE_UNALIGNED); - tmp = REG_SET_FIELD(tmp, SH_MEM_CONFIG, RETRY_MODE, 0); - WREG32_SOC15(GC, 0, mmSH_MEM_CONFIG, tmp); - WREG32_SOC15(GC, 0, mmSH_MEM_BASES, 0); - } else { - tmp = REG_SET_FIELD(0, SH_MEM_CONFIG, ALIGNMENT_MODE, - SH_MEM_ALIGNMENT_MODE_UNALIGNED); - tmp = REG_SET_FIELD(tmp, SH_MEM_CONFIG, RETRY_MODE, 0); - WREG32_SOC15(GC, 0, mmSH_MEM_CONFIG, tmp); + WREG32_SOC15(GC, 0, mmSH_MEM_CONFIG, DEFAULT_SH_MEM_CONFIG); + if (i != 0) { 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, @@ -1544,24 +1544,6 @@ static void gfx_v10_0_constants_init(struct amdgpu_device *adev) gfx_v10_0_init_compute_vmid(adev); - mutex_lock(&adev->grbm_idx_mutex); - /* - * making sure that the following register writes will be broadcasted - * to all the shaders - */ - gfx_v10_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); - - tmp = REG_SET_FIELD(0, PA_SC_FIFO_SIZE, SC_FRONTEND_PRIM_FIFO_SIZE, - adev->gfx.config.sc_prim_fifo_size_frontend); - tmp = REG_SET_FIELD(tmp, PA_SC_FIFO_SIZE, SC_BACKEND_PRIM_FIFO_SIZE, - adev->gfx.config.sc_prim_fifo_size_backend); - tmp = REG_SET_FIELD(tmp, PA_SC_FIFO_SIZE, SC_HIZ_TILE_FIFO_SIZE, - adev->gfx.config.sc_hiz_tile_fifo_size); - tmp = REG_SET_FIELD(tmp, PA_SC_FIFO_SIZE, SC_EARLYZ_TILE_FIFO_SIZE, - adev->gfx.config.sc_earlyz_tile_fifo_size); - WREG32_SOC15(GC, 0, mmPA_SC_FIFO_SIZE, tmp); - - mutex_unlock(&adev->grbm_idx_mutex); } static void gfx_v10_0_enable_gui_idle_interrupt(struct amdgpu_device *adev, @@ -4215,6 +4197,15 @@ static void gfx_v10_0_ring_emit_ib_gfx(struct amdgpu_ring *ring, unsigned vmid = AMDGPU_JOB_GET_VMID(job); u32 header, control = 0; + /* Prevent a hw deadlock due to a wave ID mismatch between ME and GDS. + * This resets the wave ID counters. (needed by transform feedback) + * TODO: This might only be needed on a VMID switch when we change + * the GDS OA mapping, not sure. + */ + amdgpu_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1)); + amdgpu_ring_write(ring, mmVGT_GS_MAX_WAVE_ID); + amdgpu_ring_write(ring, ring->adev->gds.vgt_gs_max_wave_id); + if (ib->flags & AMDGPU_IB_FLAG_CE) header = PACKET3(PACKET3_INDIRECT_BUFFER_CNST, 2); else @@ -4252,6 +4243,22 @@ static void gfx_v10_0_ring_emit_ib_compute(struct amdgpu_ring *ring, unsigned vmid = AMDGPU_JOB_GET_VMID(job); u32 control = INDIRECT_BUFFER_VALID | ib->length_dw | (vmid << 24); + /* Currently, there is a high possibility to get wave ID mismatch + * between ME and GDS, leading to a hw deadlock, because ME generates + * different wave IDs than the GDS expects. This situation happens + * randomly when at least 5 compute pipes use GDS ordered append. + * The wave IDs generated by ME are also wrong after suspend/resume. + * Those are probably bugs somewhere else in the kernel driver. + * + * Writing GDS_COMPUTE_MAX_WAVE_ID resets wave ID counters in ME and + * GDS to 0 for this ring (me/pipe). + */ + if (ib->flags & AMDGPU_IB_FLAG_RESET_GDS_MAX_WAVE_ID) { + amdgpu_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1)); + amdgpu_ring_write(ring, mmGDS_COMPUTE_MAX_WAVE_ID); + amdgpu_ring_write(ring, ring->adev->gds.gds_compute_max_wave_id); + } + amdgpu_ring_write(ring, PACKET3(PACKET3_INDIRECT_BUFFER, 2)); BUG_ON(ib->gpu_addr & 0x3); /* Dword align */ amdgpu_ring_write(ring, @@ -4278,11 +4285,7 @@ static void gfx_v10_0_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, amdgpu_ring_write(ring, PACKET3(PACKET3_RELEASE_MEM, 6)); amdgpu_ring_write(ring, (PACKET3_RELEASE_MEM_GCR_SEQ | PACKET3_RELEASE_MEM_GCR_GL2_WB | - PACKET3_RELEASE_MEM_GCR_GL2_INV | - PACKET3_RELEASE_MEM_GCR_GL2_US | - PACKET3_RELEASE_MEM_GCR_GL1_INV | - PACKET3_RELEASE_MEM_GCR_GLV_INV | - PACKET3_RELEASE_MEM_GCR_GLM_INV | + PACKET3_RELEASE_MEM_GCR_GLM_INV | /* must be set with GLM_WB */ PACKET3_RELEASE_MEM_GCR_GLM_WB | PACKET3_RELEASE_MEM_CACHE_POLICY(3) | PACKET3_RELEASE_MEM_EVENT_TYPE(CACHE_FLUSH_AND_INV_TS_EVENT) | @@ -4948,7 +4951,7 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_gfx = { 5 + /* HDP_INVL */ 8 + 8 + /* FENCE x2 */ 2, /* SWITCH_BUFFER */ - .emit_ib_size = 4, /* gfx_v10_0_ring_emit_ib_gfx */ + .emit_ib_size = 7, /* gfx_v10_0_ring_emit_ib_gfx */ .emit_ib = gfx_v10_0_ring_emit_ib_gfx, .emit_fence = gfx_v10_0_ring_emit_fence, .emit_pipeline_sync = gfx_v10_0_ring_emit_pipeline_sync, @@ -4987,7 +4990,7 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_compute = { SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 + 2 + /* gfx_v10_0_ring_emit_vm_flush */ 8 + 8 + 8, /* gfx_v10_0_ring_emit_fence x3 for user fence, vm fence */ - .emit_ib_size = 4, /* gfx_v10_0_ring_emit_ib_compute */ + .emit_ib_size = 7, /* gfx_v10_0_ring_emit_ib_compute */ .emit_ib = gfx_v10_0_ring_emit_ib_compute, .emit_fence = gfx_v10_0_ring_emit_fence, .emit_pipeline_sync = gfx_v10_0_ring_emit_pipeline_sync, @@ -5020,7 +5023,7 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_kiq = { SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 7 + 2 + /* gfx_v10_0_ring_emit_vm_flush */ 8 + 8 + 8, /* gfx_v10_0_ring_emit_fence_kiq x3 for user fence, vm fence */ - .emit_ib_size = 4, /* gfx_v10_0_ring_emit_ib_compute */ + .emit_ib_size = 7, /* gfx_v10_0_ring_emit_ib_compute */ .emit_ib = gfx_v10_0_ring_emit_ib_compute, .emit_fence = gfx_v10_0_ring_emit_fence_kiq, .test_ring = gfx_v10_0_ring_test_ring, @@ -5096,10 +5099,10 @@ static void gfx_v10_0_set_gds_init(struct amdgpu_device *adev) /* init asic gds info */ switch (adev->asic_type) { case CHIP_NAVI10: - adev->gds.gds_size = 0x10000; - break; default: adev->gds.gds_size = 0x10000; + adev->gds.gds_compute_max_wave_id = 0x4ff; + adev->gds.vgt_gs_max_wave_id = 0x3ff; break; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c index 789e900905e9..7f0a63628c43 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c @@ -3043,7 +3043,7 @@ static void gfx_v6_0_read_wave_sgprs(struct amdgpu_device *adev, uint32_t simd, } static void gfx_v6_0_select_me_pipe_q(struct amdgpu_device *adev, - u32 me, u32 pipe, u32 q) + u32 me, u32 pipe, u32 q, u32 vm) { DRM_INFO("Not implemented\n"); } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c index 341b5024e598..0db9f488da7e 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c @@ -4169,9 +4169,9 @@ static void gfx_v7_0_read_wave_sgprs(struct amdgpu_device *adev, uint32_t simd, } static void gfx_v7_0_select_me_pipe_q(struct amdgpu_device *adev, - u32 me, u32 pipe, u32 q) + u32 me, u32 pipe, u32 q, u32 vm) { - cik_srbm_select(adev, me, pipe, q, 0); + cik_srbm_select(adev, me, pipe, q, vm); } static const struct amdgpu_gfx_funcs gfx_v7_0_gfx_funcs = { diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index a57b5abf96a0..5f401b41ef7c 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -3436,9 +3436,9 @@ static void gfx_v8_0_select_se_sh(struct amdgpu_device *adev, } static void gfx_v8_0_select_me_pipe_q(struct amdgpu_device *adev, - u32 me, u32 pipe, u32 q) + u32 me, u32 pipe, u32 q, u32 vm) { - vi_srbm_select(adev, me, pipe, q, 0); + vi_srbm_select(adev, me, pipe, q, vm); } static u32 gfx_v8_0_get_rb_active_bitmap(struct amdgpu_device *adev) @@ -3925,11 +3925,10 @@ static int gfx_v8_0_init_save_restore_list(struct amdgpu_device *adev) int list_size; unsigned int *register_list_format = - kmalloc(adev->gfx.rlc.reg_list_format_size_bytes, GFP_KERNEL); + kmemdup(adev->gfx.rlc.register_list_format, + adev->gfx.rlc.reg_list_format_size_bytes, GFP_KERNEL); if (!register_list_format) return -ENOMEM; - memcpy(register_list_format, adev->gfx.rlc.register_list_format, - adev->gfx.rlc.reg_list_format_size_bytes); gfx_v8_0_parse_ind_reg_list(register_list_format, RLC_FormatDirectRegListLength, diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 9b413f6fa588..f4c4eea62526 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -1313,9 +1313,9 @@ static void gfx_v9_0_read_wave_vgprs(struct amdgpu_device *adev, uint32_t simd, } static void gfx_v9_0_select_me_pipe_q(struct amdgpu_device *adev, - u32 me, u32 pipe, u32 q) + u32 me, u32 pipe, u32 q, u32 vm) { - soc15_grbm_select(adev, me, pipe, q, 0); + soc15_grbm_select(adev, me, pipe, q, vm); } static const struct amdgpu_gfx_funcs gfx_v9_0_gfx_funcs = { @@ -1942,11 +1942,15 @@ static void gfx_v9_0_constants_init(struct amdgpu_device *adev) if (i == 0) { tmp = REG_SET_FIELD(0, SH_MEM_CONFIG, ALIGNMENT_MODE, SH_MEM_ALIGNMENT_MODE_UNALIGNED); + tmp = REG_SET_FIELD(tmp, SH_MEM_CONFIG, RETRY_DISABLE, + !!amdgpu_noretry); WREG32_SOC15_RLC(GC, 0, mmSH_MEM_CONFIG, tmp); WREG32_SOC15_RLC(GC, 0, mmSH_MEM_BASES, 0); } else { tmp = REG_SET_FIELD(0, SH_MEM_CONFIG, ALIGNMENT_MODE, SH_MEM_ALIGNMENT_MODE_UNALIGNED); + tmp = REG_SET_FIELD(tmp, SH_MEM_CONFIG, RETRY_DISABLE, + !!amdgpu_noretry); WREG32_SOC15_RLC(GC, 0, mmSH_MEM_CONFIG, tmp); tmp = REG_SET_FIELD(0, SH_MEM_BASES, PRIVATE_BASE, (adev->gmc.private_aperture_start >> 48)); @@ -1960,25 +1964,6 @@ static void gfx_v9_0_constants_init(struct amdgpu_device *adev) mutex_unlock(&adev->srbm_mutex); gfx_v9_0_init_compute_vmid(adev); - - mutex_lock(&adev->grbm_idx_mutex); - /* - * making sure that the following register writes will be broadcasted - * to all the shaders - */ - gfx_v9_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); - - WREG32_SOC15_RLC(GC, 0, mmPA_SC_FIFO_SIZE, - (adev->gfx.config.sc_prim_fifo_size_frontend << - PA_SC_FIFO_SIZE__SC_FRONTEND_PRIM_FIFO_SIZE__SHIFT) | - (adev->gfx.config.sc_prim_fifo_size_backend << - PA_SC_FIFO_SIZE__SC_BACKEND_PRIM_FIFO_SIZE__SHIFT) | - (adev->gfx.config.sc_hiz_tile_fifo_size << - PA_SC_FIFO_SIZE__SC_HIZ_TILE_FIFO_SIZE__SHIFT) | - (adev->gfx.config.sc_earlyz_tile_fifo_size << - PA_SC_FIFO_SIZE__SC_EARLYZ_TILE_FIFO_SIZE__SHIFT)); - mutex_unlock(&adev->grbm_idx_mutex); - } static void gfx_v9_0_wait_for_rlc_serdes(struct amdgpu_device *adev) @@ -2093,11 +2078,10 @@ static int gfx_v9_1_init_rlc_save_restore_list(struct amdgpu_device *adev) u32 tmp = 0; u32 *register_list_format = - kmalloc(adev->gfx.rlc.reg_list_format_size_bytes, GFP_KERNEL); + kmemdup(adev->gfx.rlc.register_list_format, + adev->gfx.rlc.reg_list_format_size_bytes, GFP_KERNEL); if (!register_list_format) return -ENOMEM; - memcpy(register_list_format, adev->gfx.rlc.register_list_format, - adev->gfx.rlc.reg_list_format_size_bytes); /* setup unique_indirect_regs array and indirect_start_offsets array */ unique_indirect_reg_count = ARRAY_SIZE(unique_indirect_regs); diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c index 9f0f189fc111..15986748f59f 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v1_0.c @@ -236,7 +236,8 @@ static void gfxhub_v1_0_setup_vmid_config(struct amdgpu_device *adev) block_size); /* Send no-retry XNACK on fault to suppress VM fault storm. */ tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, - RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 1); + RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, + !amdgpu_noretry); WREG32_SOC15_OFFSET(GC, 0, mmVM_CONTEXT1_CNTL, i, tmp); WREG32_SOC15_OFFSET(GC, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32, i*2, 0); WREG32_SOC15_OFFSET(GC, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32, i*2, 0); diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_0.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_0.c index b7de60a15623..d605b4963f8a 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_0.c @@ -215,7 +215,8 @@ static void gfxhub_v2_0_setup_vmid_config(struct amdgpu_device *adev) adev->vm_manager.block_size - 9); /* Send no-retry XNACK on fault to suppress VM fault storm. */ tmp = REG_SET_FIELD(tmp, GCVM_CONTEXT1_CNTL, - RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0); + RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, + !amdgpu_noretry); WREG32_SOC15_OFFSET(GC, 0, mmGCVM_CONTEXT1_CNTL, i, tmp); WREG32_SOC15_OFFSET(GC, 0, mmGCVM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32, i*2, 0); WREG32_SOC15_OFFSET(GC, 0, mmGCVM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32, i*2, 0); diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c index cec7c1fb14bf..5eeb72fcc123 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c @@ -245,8 +245,9 @@ static void gmc_v10_0_flush_gpu_tlb(struct amdgpu_device *adev, mutex_lock(&adev->mman.gtt_window_lock); gmc_v10_0_flush_vm_hub(adev, vmid, AMDGPU_MMHUB, 0); - if (!adev->mman.buffer_funcs_enabled || !adev->ib_pool_ready || - adev->asic_type != CHIP_NAVI10) { + if (!adev->mman.buffer_funcs_enabled || + !adev->ib_pool_ready || + adev->in_gpu_reset) { gmc_v10_0_flush_vm_hub(adev, vmid, AMDGPU_GFXHUB, 0); mutex_unlock(&adev->mman.gtt_window_lock); return; diff --git a/drivers/gpu/drm/amd/amdgpu/mes_v10_1.h b/drivers/gpu/drm/amd/amdgpu/mes_v10_1.h index 17b9b53fa892..9afd6ddb01e9 100644 --- a/drivers/gpu/drm/amd/amdgpu/mes_v10_1.h +++ b/drivers/gpu/drm/amd/amdgpu/mes_v10_1.h @@ -22,7 +22,7 @@ */ #ifndef __MES_V10_1_H__ -#define __MES_v10_1_H__ +#define __MES_V10_1_H__ extern const struct amdgpu_ip_block_version mes_v10_1_ip_block; diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c index 05d1d448c8f5..dc5ce03034d3 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c @@ -265,7 +265,8 @@ static void mmhub_v1_0_setup_vmid_config(struct amdgpu_device *adev) block_size); /* Send no-retry XNACK on fault to suppress VM fault storm. */ tmp = REG_SET_FIELD(tmp, VM_CONTEXT1_CNTL, - RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 1); + RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, + !amdgpu_noretry); WREG32_SOC15_OFFSET(MMHUB, 0, mmVM_CONTEXT1_CNTL, i, tmp); WREG32_SOC15_OFFSET(MMHUB, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32, i*2, 0); WREG32_SOC15_OFFSET(MMHUB, 0, mmVM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32, i*2, 0); diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c index 37a1a318ae63..0f9549f19ade 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c @@ -205,7 +205,8 @@ static void mmhub_v2_0_setup_vmid_config(struct amdgpu_device *adev) adev->vm_manager.block_size - 9); /* Send no-retry XNACK on fault to suppress VM fault storm. */ tmp = REG_SET_FIELD(tmp, MMVM_CONTEXT1_CNTL, - RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, 0); + RETRY_PERMISSION_OR_INVALID_PAGE_FAULT, + !amdgpu_noretry); WREG32_SOC15_OFFSET(MMHUB, 0, mmMMVM_CONTEXT1_CNTL, i, tmp); WREG32_SOC15_OFFSET(MMHUB, 0, mmMMVM_CONTEXT1_PAGE_TABLE_START_ADDR_LO32, i*2, 0); WREG32_SOC15_OFFSET(MMHUB, 0, mmMMVM_CONTEXT1_PAGE_TABLE_START_ADDR_HI32, i*2, 0); diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c index af20ffb55c54..662612f89c70 100644 --- a/drivers/gpu/drm/amd/amdgpu/nv.c +++ b/drivers/gpu/drm/amd/amdgpu/nv.c @@ -31,6 +31,7 @@ #include "amdgpu_vce.h" #include "amdgpu_ucode.h" #include "amdgpu_psp.h" +#include "amdgpu_smu.h" #include "atom.h" #include "amd_pcie.h" @@ -255,6 +256,39 @@ static void nv_gpu_pci_config_reset(struct amdgpu_device *adev) } #endif +static int nv_asic_mode1_reset(struct amdgpu_device *adev) +{ + u32 i; + int ret = 0; + + amdgpu_atombios_scratch_regs_engine_hung(adev, true); + + dev_info(adev->dev, "GPU mode1 reset\n"); + + /* disable BM */ + pci_clear_master(adev->pdev); + + pci_save_state(adev->pdev); + + ret = psp_gpu_reset(adev); + if (ret) + dev_err(adev->dev, "GPU mode1 reset failed\n"); + + pci_restore_state(adev->pdev); + + /* wait for asic to come out of reset */ + for (i = 0; i < adev->usec_timeout; i++) { + u32 memsize = adev->nbio_funcs->get_memsize(adev); + + if (memsize != 0xffffffff) + break; + udelay(1); + } + + amdgpu_atombios_scratch_regs_engine_hung(adev, false); + + return ret; +} static int nv_asic_reset(struct amdgpu_device *adev) { @@ -266,8 +300,15 @@ static int nv_asic_reset(struct amdgpu_device *adev) amdgpu_atombios_scratch_regs_engine_hung(adev, false); #endif + int ret = 0; + struct smu_context *smu = &adev->smu; - return 0; + if (smu_baco_is_support(smu)) + ret = smu_baco_reset(smu); + else + ret = nv_asic_mode1_reset(adev); + + return ret; } static int nv_set_uvd_clocks(struct amdgpu_device *adev, u32 vclk, u32 dclk) @@ -348,8 +389,10 @@ int nv_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &smu_v11_0_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) else if (amdgpu_device_has_dc_support(adev)) amdgpu_device_ip_block_add(adev, &dm_ip_block); +#endif amdgpu_device_ip_block_add(adev, &gfx_v10_0_ip_block); amdgpu_device_ip_block_add(adev, &sdma_v5_0_ip_block); if (adev->firmware.load_type == AMDGPU_FW_LOAD_DIRECT && diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c index 759b2768c51c..41b72588adcf 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c @@ -103,6 +103,9 @@ static int psp_v11_0_init_microcode(struct psp_context *psp) adev->psp.toc_bin_size = le32_to_cpu(sos_hdr_v1_1->toc_size_bytes); adev->psp.toc_start_addr = (uint8_t *)adev->psp.sys_start_addr + le32_to_cpu(sos_hdr_v1_1->toc_offset_bytes); + adev->psp.kdb_bin_size = le32_to_cpu(sos_hdr_v1_1->kdb_size_bytes); + adev->psp.kdb_start_addr = (uint8_t *)adev->psp.sys_start_addr + + le32_to_cpu(sos_hdr_v1_1->kdb_offset_bytes); } break; default: @@ -177,6 +180,48 @@ out: return err; } +static int psp_v11_0_bootloader_load_kdb(struct psp_context *psp) +{ + int ret; + uint32_t psp_gfxdrv_command_reg = 0; + struct amdgpu_device *adev = psp->adev; + uint32_t sol_reg; + + /* Check tOS sign of life register to confirm sys driver and sOS + * are already been loaded. + */ + sol_reg = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_81); + if (sol_reg) { + psp->sos_fw_version = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_58); + dev_info(adev->dev, "sos fw version = 0x%x.\n", psp->sos_fw_version); + return 0; + } + + /* Wait for bootloader to signify that is ready having bit 31 of C2PMSG_35 set to 1 */ + ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_35), + 0x80000000, 0x80000000, false); + if (ret) + return ret; + + memset(psp->fw_pri_buf, 0, PSP_1_MEG); + + /* Copy PSP KDB binary to memory */ + memcpy(psp->fw_pri_buf, psp->kdb_start_addr, psp->kdb_bin_size); + + /* Provide the sys driver to bootloader */ + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, + (uint32_t)(psp->fw_pri_mc_addr >> 20)); + psp_gfxdrv_command_reg = PSP_BL__LOAD_KEY_DATABASE; + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_35, + psp_gfxdrv_command_reg); + + /* Wait for bootloader to signify that is ready having bit 31 of C2PMSG_35 set to 1*/ + ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_35), + 0x80000000, 0x80000000, false); + + return ret; +} + static int psp_v11_0_bootloader_load_sysdrv(struct psp_context *psp) { int ret; @@ -190,7 +235,7 @@ static int psp_v11_0_bootloader_load_sysdrv(struct psp_context *psp) sol_reg = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_81); if (sol_reg) { psp->sos_fw_version = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_58); - printk("sos fw version = 0x%x.\n", psp->sos_fw_version); + dev_info(adev->dev, "sos fw version = 0x%x.\n", psp->sos_fw_version); return 0; } @@ -208,7 +253,7 @@ static int psp_v11_0_bootloader_load_sysdrv(struct psp_context *psp) /* Provide the sys driver to bootloader */ WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, (uint32_t)(psp->fw_pri_mc_addr >> 20)); - psp_gfxdrv_command_reg = 1 << 16; + psp_gfxdrv_command_reg = PSP_BL__LOAD_SYSDRV; WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_35, psp_gfxdrv_command_reg); @@ -249,7 +294,7 @@ static int psp_v11_0_bootloader_load_sos(struct psp_context *psp) /* Provide the PSP secure OS to bootloader */ WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, (uint32_t)(psp->fw_pri_mc_addr >> 20)); - psp_gfxdrv_command_reg = 2 << 16; + psp_gfxdrv_command_reg = PSP_BL__LOAD_SOSDRV; WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_35, psp_gfxdrv_command_reg); @@ -534,7 +579,7 @@ psp_v11_0_sram_map(struct amdgpu_device *adev, case AMDGPU_UCODE_ID_RLC_G: *sram_offset = 0x2000; - if (adev->asic_type != CHIP_NAVI10) { + if (adev->asic_type < CHIP_NAVI10) { *sram_addr_reg_offset = SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_UCODE_ADDR); *sram_data_reg_offset = SOC15_REG_OFFSET(GC, 0, mmRLC_GPM_UCODE_DATA); } else { @@ -545,7 +590,7 @@ psp_v11_0_sram_map(struct amdgpu_device *adev, case AMDGPU_UCODE_ID_SDMA0: *sram_offset = 0x0; - if (adev->asic_type != CHIP_NAVI10) { + if (adev->asic_type < CHIP_NAVI10) { *sram_addr_reg_offset = SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_UCODE_ADDR); *sram_data_reg_offset = SOC15_REG_OFFSET(SDMA0, 0, mmSDMA0_UCODE_DATA); } else { @@ -822,6 +867,7 @@ static int psp_v11_0_rlc_autoload_start(struct psp_context *psp) static const struct psp_funcs psp_v11_0_funcs = { .init_microcode = psp_v11_0_init_microcode, + .bootloader_load_kdb = psp_v11_0_bootloader_load_kdb, .bootloader_load_sysdrv = psp_v11_0_bootloader_load_sysdrv, .bootloader_load_sos = psp_v11_0_bootloader_load_sos, .ring_init = psp_v11_0_ring_init, diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c index 2ea772692037..019c47feee42 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c @@ -155,7 +155,7 @@ static int psp_v3_1_bootloader_load_sysdrv(struct psp_context *psp) /* Provide the sys driver to bootloader */ WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, (uint32_t)(psp->fw_pri_mc_addr >> 20)); - psp_gfxdrv_command_reg = 1 << 16; + psp_gfxdrv_command_reg = PSP_BL__LOAD_SYSDRV; WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_35, psp_gfxdrv_command_reg); @@ -218,7 +218,7 @@ static int psp_v3_1_bootloader_load_sos(struct psp_context *psp) /* Provide the PSP secure OS to bootloader */ WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, (uint32_t)(psp->fw_pri_mc_addr >> 20)); - psp_gfxdrv_command_reg = 2 << 16; + psp_gfxdrv_command_reg = PSP_BL__LOAD_SOSDRV; WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_35, psp_gfxdrv_command_reg); diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index 87152d8ef0df..23265414d448 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -649,8 +649,6 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev) #if defined(CONFIG_DRM_AMD_DC) else if (amdgpu_device_has_dc_support(adev)) amdgpu_device_ip_block_add(adev, &dm_ip_block); -#else -# warning "Enable CONFIG_DRM_AMD_DC for display support on SOC15." #endif if (!(adev->asic_type == CHIP_VEGA20 && amdgpu_sriov_vf(adev))) { amdgpu_device_ip_block_add(adev, &uvd_v7_0_ip_block); @@ -671,8 +669,6 @@ int soc15_set_ip_blocks(struct amdgpu_device *adev) #if defined(CONFIG_DRM_AMD_DC) else if (amdgpu_device_has_dc_support(adev)) amdgpu_device_ip_block_add(adev, &dm_ip_block); -#else -# warning "Enable CONFIG_DRM_AMD_DC for display support on SOC15." #endif amdgpu_device_ip_block_add(adev, &vcn_v1_0_ip_block); break; @@ -717,9 +713,15 @@ static void soc15_get_pcie_usage(struct amdgpu_device *adev, uint64_t *count0, return; /* Set the 2 events that we wish to watch, defined above */ - /* Reg 40 is # received msgs, Reg 104 is # of posted requests sent */ + /* Reg 40 is # received msgs */ perfctr = REG_SET_FIELD(perfctr, PCIE_PERF_CNTL_TXCLK, EVENT0_SEL, 40); - perfctr = REG_SET_FIELD(perfctr, PCIE_PERF_CNTL_TXCLK, EVENT1_SEL, 104); + /* Pre-VG20, Reg 104 is # of posted requests sent. On VG20 it's 108 */ + if (adev->asic_type == CHIP_VEGA20) + perfctr = REG_SET_FIELD(perfctr, PCIE_PERF_CNTL_TXCLK, + EVENT1_SEL, 108); + else + perfctr = REG_SET_FIELD(perfctr, PCIE_PERF_CNTL_TXCLK, + EVENT1_SEL, 104); /* Write to enable desired perf counters */ WREG32_PCIE(smnPCIE_PERF_CNTL_TXCLK, perfctr); diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c index d40ed1a828dd..6575ddcfcf00 100644 --- a/drivers/gpu/drm/amd/amdgpu/vi.c +++ b/drivers/gpu/drm/amd/amdgpu/vi.c @@ -59,7 +59,6 @@ #include "vid.h" #include "vi.h" -#include "vi_dpm.h" #include "gmc_v8_0.h" #include "gmc_v7_0.h" #include "gfx_v8_0.h" diff --git a/drivers/gpu/drm/amd/amdgpu/vi_dpm.h b/drivers/gpu/drm/amd/amdgpu/vi_dpm.h deleted file mode 100644 index c43e03fddfba..000000000000 --- a/drivers/gpu/drm/amd/amdgpu/vi_dpm.h +++ /dev/null @@ -1,32 +0,0 @@ -/* - * Copyright 2014 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 __VI_DPM_H__ -#define __VI_DPM_H__ - -extern const struct amd_ip_funcs cz_dpm_ip_funcs; -int cz_smu_init(struct amdgpu_device *adev); -int cz_smu_start(struct amdgpu_device *adev); -int cz_smu_fini(struct amdgpu_device *adev); - -#endif diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c index f91126f5f1be..26b15cc56c31 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c @@ -839,7 +839,7 @@ static int kfd_ioctl_get_clock_counters(struct file *filep, /* No access to rdtsc. Using raw monotonic time */ args->cpu_clock_counter = ktime_get_raw_ns(); - args->system_clock_counter = ktime_get_boot_ns(); + args->system_clock_counter = ktime_get_boottime_ns(); /* Since the counter is in nano-seconds we use 1GHz frequency */ args->system_clock_freq = 1000000000; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c index 2e2f7241f381..3322a443dfb2 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c @@ -487,7 +487,6 @@ 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); @@ -504,17 +503,15 @@ struct kfd_dev *kgd2kfd_probe(struct kgd_dev *kgd, * 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) { + kfd->pci_atomic_requested = amdgpu_amdkfd_have_atomics_support(kgd); + if (device_info->needs_pci_atomics && + !kfd->pci_atomic_requested) { dev_info(kfd_device, "skipped device %x:%x, PCI rejects atomics\n", pdev->vendor, pdev->device); kfree(kfd); return NULL; - } else if (!ret) - kfd->pci_atomic_requested = true; + } kfd->kgd = kgd; kfd->device_info = device_info; 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 584748c23f14..e6a4288bfaa6 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c @@ -1157,12 +1157,7 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q, mqd_mgr = dqm->mqd_mgrs[get_mqd_type_from_queue_type( q->properties.type)]; - /* - * Eviction state logic: mark all queues as evicted, even ones - * not currently active. Restoring inactive queues later only - * updates the is_evicted flag but is a no-op otherwise. - */ - q->properties.is_evicted = !!qpd->evicted; + if (q->properties.type == KFD_QUEUE_TYPE_SDMA || q->properties.type == KFD_QUEUE_TYPE_SDMA_XGMI) dqm->asic_ops.init_sdma_vm(dqm, q, qpd); @@ -1173,9 +1168,16 @@ static int create_queue_cpsch(struct device_queue_manager *dqm, struct queue *q, retval = -ENOMEM; goto out_deallocate_doorbell; } + + dqm_lock(dqm); + /* + * Eviction state logic: mark all queues as evicted, even ones + * not currently active. Restoring inactive queues later only + * updates the is_evicted flag but is a no-op otherwise. + */ + q->properties.is_evicted = !!qpd->evicted; mqd_mgr->init_mqd(mqd_mgr, &q->mqd, q->mqd_mem_obj, &q->gart_mqd_addr, &q->properties); - dqm_lock(dqm); list_add(&q->list, &qpd->queues_list); qpd->queue_count++; 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 index e9fe39382371..95a82ac455f2 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager_v9.c @@ -61,7 +61,7 @@ static int update_qpd_v9(struct device_queue_manager *dqm, qpd->sh_mem_config = SH_MEM_ALIGNMENT_MODE_UNALIGNED << SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT; - if (noretry && + if (amdgpu_noretry && !dqm->dev->device_info->needs_iommu_device) qpd->sh_mem_config |= 1 << SH_MEM_CONFIG__RETRY_DISABLE__SHIFT; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c index 07f02f8e4fe4..2d5ddf199bd0 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue_v9.c @@ -134,6 +134,7 @@ static int pm_runlist_v9(struct packet_manager *pm, uint32_t *buffer, packet->bitfields4.ib_size = ib_size_in_dwords; packet->bitfields4.chain = chain ? 1 : 0; packet->bitfields4.offload_polling = 0; + packet->bitfields4.chained_runlist_idle_disable = chain ? 1 : 0; packet->bitfields4.valid = 1; packet->bitfields4.process_cnt = concurrent_proc_cnt; packet->ordinal2 = lower_32_bits(ib); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c index c72c8f5fd54c..ccf6b2310316 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c @@ -203,11 +203,15 @@ static int pm_create_runlist_ib(struct packet_manager *pm, pr_debug("Finished map process and queues to runlist\n"); - if (is_over_subscription) + if (is_over_subscription) { + if (!pm->is_over_subscription) + pr_warn("Runlist is getting oversubscribed. Expect reduced ROCm performance.\n"); retval = pm->pmf->runlist(pm, &rl_buffer[rl_wptr], *rl_gpu_addr, alloc_size_bytes / sizeof(uint32_t), true); + } + pm->is_over_subscription = is_over_subscription; for (i = 0; i < alloc_size_bytes / sizeof(uint32_t); i++) pr_debug("0x%2X ", rl_buffer[i]); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_ai.h b/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_ai.h index 49ab66b703fa..e3e21404cfa0 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_ai.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_ai.h @@ -120,7 +120,7 @@ struct pm4_mes_runlist { uint32_t ib_size:20; uint32_t chain:1; uint32_t offload_polling:1; - uint32_t reserved2:1; + uint32_t chained_runlist_idle_disable:1; uint32_t valid:1; uint32_t process_cnt:4; uint32_t reserved3:4; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h index f88c0d706b1d..3933fb6a371e 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h @@ -157,7 +157,7 @@ extern int ignore_crat; /* * Set sh_mem_config.retry_disable on Vega10 */ -extern int noretry; +extern int amdgpu_noretry; /* * Halt if HWS hang is detected @@ -937,6 +937,7 @@ struct packet_manager { bool allocated; struct kfd_mem_obj *ib_buffer_obj; unsigned int ib_size_bytes; + bool is_over_subscription; const struct packet_manager_funcs *pmf; }; 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 da0958625861..7e6c3ee82f5b 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process_queue_manager.c @@ -150,6 +150,9 @@ void pqm_uninit(struct process_queue_manager *pqm) struct process_queue_node *pqn, *next; list_for_each_entry_safe(pqn, next, &pqm->queues, process_queue_list) { + if (pqn->q && pqn->q->gws) + amdgpu_amdkfd_remove_gws_from_process(pqm->process->kgd_process_info, + pqn->q->gws); uninit_queue(pqn->q); list_del(&pqn->process_queue_list); kfree(pqn); diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig index 7073cfcf04e8..f954bf61af28 100644 --- a/drivers/gpu/drm/amd/display/Kconfig +++ b/drivers/gpu/drm/amd/display/Kconfig @@ -5,6 +5,7 @@ menu "Display Engine Configuration" config DRM_AMD_DC bool "AMD DC - Enable new display engine" default y + select SND_HDA_COMPONENT if SND_HDA_CORE select DRM_AMD_DC_DCN1_0 if X86 && !(KCOV_INSTRUMENT_ALL && KCOV_ENABLE_COMPARISONS) help Choose this option if you want to use the new display engine 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 df08ea172595..4a29f72334d0 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -56,6 +56,7 @@ #include <linux/pm_runtime.h> #include <linux/pci.h> #include <linux/firmware.h> +#include <linux/component.h> #include <drm/drm_atomic.h> #include <drm/drm_atomic_uapi.h> @@ -65,6 +66,7 @@ #include <drm/drm_fourcc.h> #include <drm/drm_edid.h> #include <drm/drm_vblank.h> +#include <drm/drm_audio_component.h> #if defined(CONFIG_DRM_AMD_DC_DCN1_0) #include "ivsrcid/dcn/irqsrcs_dcn_1_0.h" @@ -508,6 +510,139 @@ static void amdgpu_dm_fbc_init(struct drm_connector *connector) } +static int amdgpu_dm_audio_component_get_eld(struct device *kdev, int port, + int pipe, bool *enabled, + unsigned char *buf, int max_bytes) +{ + struct drm_device *dev = dev_get_drvdata(kdev); + struct amdgpu_device *adev = dev->dev_private; + struct drm_connector *connector; + struct drm_connector_list_iter conn_iter; + struct amdgpu_dm_connector *aconnector; + int ret = 0; + + *enabled = false; + + mutex_lock(&adev->dm.audio_lock); + + drm_connector_list_iter_begin(dev, &conn_iter); + drm_for_each_connector_iter(connector, &conn_iter) { + aconnector = to_amdgpu_dm_connector(connector); + if (aconnector->audio_inst != port) + continue; + + *enabled = true; + ret = drm_eld_size(connector->eld); + memcpy(buf, connector->eld, min(max_bytes, ret)); + + break; + } + drm_connector_list_iter_end(&conn_iter); + + mutex_unlock(&adev->dm.audio_lock); + + DRM_DEBUG_KMS("Get ELD : idx=%d ret=%d en=%d\n", port, ret, *enabled); + + return ret; +} + +static const struct drm_audio_component_ops amdgpu_dm_audio_component_ops = { + .get_eld = amdgpu_dm_audio_component_get_eld, +}; + +static int amdgpu_dm_audio_component_bind(struct device *kdev, + struct device *hda_kdev, void *data) +{ + struct drm_device *dev = dev_get_drvdata(kdev); + struct amdgpu_device *adev = dev->dev_private; + struct drm_audio_component *acomp = data; + + acomp->ops = &amdgpu_dm_audio_component_ops; + acomp->dev = kdev; + adev->dm.audio_component = acomp; + + return 0; +} + +static void amdgpu_dm_audio_component_unbind(struct device *kdev, + struct device *hda_kdev, void *data) +{ + struct drm_device *dev = dev_get_drvdata(kdev); + struct amdgpu_device *adev = dev->dev_private; + struct drm_audio_component *acomp = data; + + acomp->ops = NULL; + acomp->dev = NULL; + adev->dm.audio_component = NULL; +} + +static const struct component_ops amdgpu_dm_audio_component_bind_ops = { + .bind = amdgpu_dm_audio_component_bind, + .unbind = amdgpu_dm_audio_component_unbind, +}; + +static int amdgpu_dm_audio_init(struct amdgpu_device *adev) +{ + int i, ret; + + if (!amdgpu_audio) + return 0; + + adev->mode_info.audio.enabled = true; + + adev->mode_info.audio.num_pins = adev->dm.dc->res_pool->audio_count; + + for (i = 0; i < adev->mode_info.audio.num_pins; i++) { + adev->mode_info.audio.pin[i].channels = -1; + adev->mode_info.audio.pin[i].rate = -1; + adev->mode_info.audio.pin[i].bits_per_sample = -1; + adev->mode_info.audio.pin[i].status_bits = 0; + adev->mode_info.audio.pin[i].category_code = 0; + adev->mode_info.audio.pin[i].connected = false; + adev->mode_info.audio.pin[i].id = + adev->dm.dc->res_pool->audios[i]->inst; + adev->mode_info.audio.pin[i].offset = 0; + } + + ret = component_add(adev->dev, &amdgpu_dm_audio_component_bind_ops); + if (ret < 0) + return ret; + + adev->dm.audio_registered = true; + + return 0; +} + +static void amdgpu_dm_audio_fini(struct amdgpu_device *adev) +{ + if (!amdgpu_audio) + return; + + if (!adev->mode_info.audio.enabled) + return; + + if (adev->dm.audio_registered) { + component_del(adev->dev, &amdgpu_dm_audio_component_bind_ops); + adev->dm.audio_registered = false; + } + + /* TODO: Disable audio? */ + + adev->mode_info.audio.enabled = false; +} + +void amdgpu_dm_audio_eld_notify(struct amdgpu_device *adev, int pin) +{ + struct drm_audio_component *acomp = adev->dm.audio_component; + + if (acomp && acomp->audio_ops && acomp->audio_ops->pin_eld_notify) { + DRM_DEBUG_KMS("Notify ELD: %d\n", pin); + + acomp->audio_ops->pin_eld_notify(acomp->audio_ops->audio_ptr, + pin, -1); + } +} + static int amdgpu_dm_init(struct amdgpu_device *adev) { struct dc_init_data init_data; @@ -518,6 +653,7 @@ static int amdgpu_dm_init(struct amdgpu_device *adev) memset(&init_data, 0, sizeof(init_data)); mutex_init(&adev->dm.dc_lock); + mutex_init(&adev->dm.audio_lock); if(amdgpu_dm_irq_init(adev)) { DRM_ERROR("amdgpu: failed to initialize DM IRQ support.\n"); @@ -621,6 +757,8 @@ error: static void amdgpu_dm_fini(struct amdgpu_device *adev) { + amdgpu_dm_audio_fini(adev); + amdgpu_dm_destroy_drm_device(&adev->dm); /* DC Destroy TODO: Replace destroy DAL */ @@ -641,6 +779,7 @@ static void amdgpu_dm_fini(struct amdgpu_device *adev) adev->dm.freesync_module = NULL; } + mutex_destroy(&adev->dm.audio_lock); mutex_destroy(&adev->dm.dc_lock); return; @@ -1535,10 +1674,7 @@ static int dce110_register_irq_handlers(struct amdgpu_device *adev) int i; unsigned client_id = AMDGPU_IRQ_CLIENTID_LEGACY; - if (adev->asic_type == CHIP_VEGA10 || - adev->asic_type == CHIP_VEGA12 || - adev->asic_type == CHIP_VEGA20 || - adev->asic_type == CHIP_RAVEN) + if (adev->asic_type >= CHIP_VEGA10) client_id = SOC15_IH_CLIENTID_DCE; int_params.requested_polarity = INTERRUPT_POLARITY_DEFAULT; @@ -1891,6 +2027,10 @@ static int amdgpu_dm_mode_config_init(struct amdgpu_device *adev) if (r) return r; + r = amdgpu_dm_audio_init(adev); + if (r) + return r; + return 0; } @@ -4837,6 +4977,7 @@ 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 */ + aconnector->audio_inst = -1; mutex_init(&aconnector->hpd_lock); /* @@ -5731,6 +5872,81 @@ cleanup: kfree(bundle); } +static void amdgpu_dm_commit_audio(struct drm_device *dev, + struct drm_atomic_state *state) +{ + struct amdgpu_device *adev = dev->dev_private; + struct amdgpu_dm_connector *aconnector; + struct drm_connector *connector; + struct drm_connector_state *old_con_state, *new_con_state; + struct drm_crtc_state *new_crtc_state; + struct dm_crtc_state *new_dm_crtc_state; + const struct dc_stream_status *status; + int i, inst; + + /* Notify device removals. */ + for_each_oldnew_connector_in_state(state, connector, old_con_state, new_con_state, i) { + if (old_con_state->crtc != new_con_state->crtc) { + /* CRTC changes require notification. */ + goto notify; + } + + if (!new_con_state->crtc) + continue; + + new_crtc_state = drm_atomic_get_new_crtc_state( + state, new_con_state->crtc); + + if (!new_crtc_state) + continue; + + if (!drm_atomic_crtc_needs_modeset(new_crtc_state)) + continue; + + notify: + aconnector = to_amdgpu_dm_connector(connector); + + mutex_lock(&adev->dm.audio_lock); + inst = aconnector->audio_inst; + aconnector->audio_inst = -1; + mutex_unlock(&adev->dm.audio_lock); + + amdgpu_dm_audio_eld_notify(adev, inst); + } + + /* Notify audio device additions. */ + for_each_new_connector_in_state(state, connector, new_con_state, i) { + if (!new_con_state->crtc) + continue; + + new_crtc_state = drm_atomic_get_new_crtc_state( + state, new_con_state->crtc); + + if (!new_crtc_state) + continue; + + if (!drm_atomic_crtc_needs_modeset(new_crtc_state)) + continue; + + new_dm_crtc_state = to_dm_crtc_state(new_crtc_state); + if (!new_dm_crtc_state->stream) + continue; + + status = dc_stream_get_status(new_dm_crtc_state->stream); + if (!status) + continue; + + aconnector = to_amdgpu_dm_connector(connector); + + mutex_lock(&adev->dm.audio_lock); + inst = status->audio_inst; + aconnector->audio_inst = inst; + mutex_unlock(&adev->dm.audio_lock); + + amdgpu_dm_audio_eld_notify(adev, inst); + } +} + /* * Enable interrupts on CRTCs that are newly active, undergone * a modeset, or have active planes again. @@ -6109,6 +6325,9 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state) /* Enable interrupts for CRTCs going from 0 to n active planes. */ amdgpu_dm_enable_crtc_interrupts(dev, state, false); + /* Update audio instances for each connector. */ + amdgpu_dm_commit_audio(dev, state); + /* * send vblank event on all events not handled in flip and * mark consumed event for drm_atomic_helper_commit_hw_done 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 baca5dc22b92..b89cbbfcc0e9 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -144,6 +144,28 @@ struct amdgpu_display_manager { struct mutex dc_lock; /** + * @audio_lock: + * + * Guards access to audio instance changes. + */ + struct mutex audio_lock; + + /** + * @audio_component: + * + * Used to notify ELD changes to sound driver. + */ + struct drm_audio_component *audio_component; + + /** + * @audio_registered: + * + * True if the audio component has been registered + * successfully, false otherwise. + */ + bool audio_registered; + + /** * @irq_handler_list_low_tab: * * Low priority IRQ handler table. @@ -254,6 +276,9 @@ struct amdgpu_dm_connector { int max_vfreq ; int pixel_clock_mhz; + /* Audio instance - protected by audio_lock. */ + int audio_inst; + struct mutex hpd_lock; bool fake_enable; 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 index eac09bfe3be2..592fa499c9f8 100644 --- 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 @@ -308,7 +308,8 @@ static void pp_to_dc_clock_levels_with_voltage( 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); + DRM_INFO("DM_PPLIB:\t %d in kHz, %d in mV\n", pp_clks->data[i].clocks_in_khz, + pp_clks->data[i].voltage_in_mv); 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; } @@ -910,11 +911,11 @@ void dm_pp_get_funcs( /* todo set_pme_wa_enable cause 4k@6ohz display not light up */ funcs->nv_funcs.set_pme_wa_enable = NULL; /* todo debug waring message */ - funcs->nv_funcs.set_hard_min_uclk_by_freq = NULL; + funcs->nv_funcs.set_hard_min_uclk_by_freq = pp_nv_set_hard_min_uclk_by_freq; /* todo compare data with window driver*/ - funcs->nv_funcs.get_maximum_sustainable_clocks = NULL; + funcs->nv_funcs.get_maximum_sustainable_clocks = pp_nv_get_maximum_sustainable_clocks; /*todo compare data with window driver */ - funcs->nv_funcs.get_uclk_dpm_states = NULL; + funcs->nv_funcs.get_uclk_dpm_states = pp_nv_get_uclk_dpm_states; break; #endif default: diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c index 14c4c60c59db..4ef4dc63e221 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc.c @@ -1190,14 +1190,12 @@ struct dc_state *dc_create_state(struct dc *dc) struct dc_state *dc_copy_state(struct dc_state *src_ctx) { int i, j; - struct dc_state *new_ctx = kzalloc(sizeof(struct dc_state), - GFP_KERNEL); + struct dc_state *new_ctx = kmemdup(src_ctx, + sizeof(struct dc_state), GFP_KERNEL); if (!new_ctx) return NULL; - memcpy(new_ctx, src_ctx, sizeof(struct dc_state)); - for (i = 0; i < MAX_PIPES; i++) { struct pipe_ctx *cur_pipe = &new_ctx->res_ctx.pipe_ctx[i]; 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 c17db5c144aa..8dbf759eba45 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c @@ -3072,8 +3072,8 @@ uint32_t dc_link_bandwidth_kbps( * but the difference is minimal and is in a safe direction, * which all works well around potential ambiguity of DP 1.4a spec. */ - long long fec_link_bw_kbps = link_bw_kbps * 970LL; - link_bw_kbps = (uint32_t)(fec_link_bw_kbps / 1000LL); + link_bw_kbps = mul_u64_u32_shr(BIT_ULL(32) * 970LL / 1000, + link_bw_kbps, 32); } #endif 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 173fcfb5abe6..51a78283a86d 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c @@ -175,32 +175,22 @@ struct resource_pool *dc_create_resource_pool(struct dc *dc, if (res_pool != NULL) { struct dc_firmware_info fw_info = { { 0 } }; - if (dc->ctx->dc_bios->funcs->get_firmware_info( - dc->ctx->dc_bios, &fw_info) == BP_RESULT_OK) { - res_pool->ref_clocks.xtalin_clock_inKhz = fw_info.pll_info.crystal_frequency; - - if (IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) { - // On FPGA these dividers are currently not configured by GDB - res_pool->ref_clocks.dccg_ref_clock_inKhz = res_pool->ref_clocks.xtalin_clock_inKhz; - res_pool->ref_clocks.dchub_ref_clock_inKhz = res_pool->ref_clocks.xtalin_clock_inKhz; - } else if (res_pool->dccg && res_pool->hubbub) { - // If DCCG reference frequency cannot be determined (usually means not set to xtalin) then this is a critical error - // as this value must be known for DCHUB programming - (res_pool->dccg->funcs->get_dccg_ref_freq)(res_pool->dccg, - fw_info.pll_info.crystal_frequency, - &res_pool->ref_clocks.dccg_ref_clock_inKhz); - - // Similarly, if DCHUB reference frequency cannot be determined, then it is also a critical error - (res_pool->hubbub->funcs->get_dchub_ref_freq)(res_pool->hubbub, - res_pool->ref_clocks.dccg_ref_clock_inKhz, - &res_pool->ref_clocks.dchub_ref_clock_inKhz); - } else { - // Not all ASICs have DCCG sw component - res_pool->ref_clocks.dccg_ref_clock_inKhz = res_pool->ref_clocks.xtalin_clock_inKhz; - res_pool->ref_clocks.dchub_ref_clock_inKhz = res_pool->ref_clocks.xtalin_clock_inKhz; - } - } else - ASSERT_CRITICAL(false); + if (dc->ctx->dc_bios->funcs->get_firmware_info(dc->ctx->dc_bios, + &fw_info) == BP_RESULT_OK) { + res_pool->ref_clocks.xtalin_clock_inKhz = + fw_info.pll_info.crystal_frequency; + /* initialize with firmware data first, no all + * ASIC have DCCG SW component. FPGA or + * simulation need initialization of + * dccg_ref_clock_inKhz, dchub_ref_clock_inKhz + * with xtalin_clock_inKhz + */ + res_pool->ref_clocks.dccg_ref_clock_inKhz = + res_pool->ref_clocks.xtalin_clock_inKhz; + res_pool->ref_clocks.dchub_ref_clock_inKhz = + res_pool->ref_clocks.xtalin_clock_inKhz; + } else + ASSERT_CRITICAL(false); } return res_pool; @@ -2011,6 +2001,9 @@ enum dc_status resource_map_pool_resources( if (context->streams[i] == stream) { context->stream_status[i].primary_otg_inst = pipe_ctx->stream_res.tg->inst; context->stream_status[i].stream_enc_inst = pipe_ctx->stream_res.stream_enc->id; + context->stream_status[i].audio_inst = + pipe_ctx->stream_res.audio ? pipe_ctx->stream_res.audio->inst : -1; + return DC_OK; } 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 a7c769a1b40b..af7f8be230f7 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c @@ -181,12 +181,10 @@ struct dc_stream_state *dc_copy_stream(const struct dc_stream_state *stream) { struct dc_stream_state *new_stream; - new_stream = kzalloc(sizeof(struct dc_stream_state), GFP_KERNEL); + new_stream = kmemdup(stream, sizeof(struct dc_stream_state), GFP_KERNEL); if (!new_stream) return NULL; - memcpy(new_stream, stream, sizeof(struct dc_stream_state)); - if (new_stream->sink) dc_sink_retain(new_stream->sink); diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h index e253a5c591f6..0fa1c26bc20d 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_stream.h +++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h @@ -42,6 +42,7 @@ struct dc_stream_status { int primary_otg_inst; int stream_enc_inst; int plane_count; + int audio_inst; struct timing_sync_info timing_sync_info; struct dc_plane_state *plane_states[MAX_SURFACE_NUM]; }; diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/Makefile b/drivers/gpu/drm/amd/display/dc/dcn20/Makefile index 1b68de27ba74..e9721a906592 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dcn20/Makefile @@ -10,7 +10,13 @@ ifdef CONFIG_DRM_AMD_DC_DSC_SUPPORT DCN20 += dcn20_dsc.o endif -CFLAGS_dcn20_resource.o := -mhard-float -msse -mpreferred-stack-boundary=4 +ifneq ($(call cc-option, -mpreferred-stack-boundary=4),) + cc_stack_align := -mpreferred-stack-boundary=4 +else ifneq ($(call cc-option, -mstack-alignment=16),) + cc_stack_align := -mstack-alignment=16 +endif + +CFLAGS_dcn20_resource.o := -mhard-float -msse $(cc_stack_align) AMD_DAL_DCN20 = $(addprefix $(AMDDALPATH)/dc/dcn20/,$(DCN20)) diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c index 4b0d8b9f61da..0b84a322b8a2 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c @@ -22,6 +22,7 @@ * Authors: AMD * */ +#include <linux/delay.h> #include "dm_services.h" #include "dm_helpers.h" @@ -522,6 +523,7 @@ static void dcn20_init_hw(struct dc *dc) struct dc_bios *dcb = dc->ctx->dc_bios; struct resource_pool *res_pool = dc->res_pool; struct dc_state *context = dc->current_state; + struct dc_firmware_info fw_info = { { 0 } }; if (dc->clk_mgr && dc->clk_mgr->funcs->init_clocks) dc->clk_mgr->funcs->init_clocks(dc->clk_mgr); @@ -545,6 +547,30 @@ static void dcn20_init_hw(struct dc *dc) } else { if (!dcb->funcs->is_accelerated_mode(dcb)) { bios_golden_init(dc); + if (dc->ctx->dc_bios->funcs->get_firmware_info( + dc->ctx->dc_bios, &fw_info) == BP_RESULT_OK) { + res_pool->ref_clocks.xtalin_clock_inKhz = fw_info.pll_info.crystal_frequency; + + if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) { + if (res_pool->dccg && res_pool->hubbub) { + + (res_pool->dccg->funcs->get_dccg_ref_freq)(res_pool->dccg, + fw_info.pll_info.crystal_frequency, + &res_pool->ref_clocks.dccg_ref_clock_inKhz); + + (res_pool->hubbub->funcs->get_dchub_ref_freq)(res_pool->hubbub, + res_pool->ref_clocks.dccg_ref_clock_inKhz, + &res_pool->ref_clocks.dchub_ref_clock_inKhz); + } else { + // Not all ASICs have DCCG sw component + res_pool->ref_clocks.dccg_ref_clock_inKhz = + res_pool->ref_clocks.xtalin_clock_inKhz; + res_pool->ref_clocks.dchub_ref_clock_inKhz = + res_pool->ref_clocks.xtalin_clock_inKhz; + } + } + } else + ASSERT_CRITICAL(false); disable_vga(dc->hwseq); } diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c index fb8aff7360ec..d200bc3cec71 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c @@ -2415,7 +2415,7 @@ struct pipe_ctx *dcn20_acquire_idle_pipe_for_layer( ASSERT(0); if (!idle_pipe) - return false; + return NULL; idle_pipe->stream = head_pipe->stream; idle_pipe->stream_res.tg = head_pipe->stream_res.tg; @@ -2576,6 +2576,9 @@ static void cap_soc_clocks( && max_clocks.uClockInKhz != 0) bb->clock_limits[i].dram_speed_mts = (max_clocks.uClockInKhz / 1000) * 16; + // HACK: Force every uclk to max for now to "disable" uclk switching. + bb->clock_limits[i].dram_speed_mts = (max_clocks.uClockInKhz / 1000) * 16; + if ((bb->clock_limits[i].fabricclk_mhz > (max_clocks.fabricClockInKhz / 1000)) && max_clocks.fabricClockInKhz != 0) bb->clock_limits[i].fabricclk_mhz = (max_clocks.fabricClockInKhz / 1000); @@ -2647,7 +2650,7 @@ static void update_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_ calculated_states[i].dram_speed_mts = uclk_states[i] * 16 / 1000; // FCLK:UCLK ratio is 1.08 - min_fclk_required_by_uclk = ((unsigned long long)uclk_states[i]) * 1080 / 1000000; + min_fclk_required_by_uclk = mul_u64_u32_shr(BIT_ULL(32) * 1080 / 1000000, uclk_states[i], 32); calculated_states[i].fabricclk_mhz = (min_fclk_required_by_uclk < min_dcfclk) ? min_dcfclk : min_fclk_required_by_uclk; @@ -2783,6 +2786,8 @@ static bool init_soc_bounding_box(struct dc *dc, le32_to_cpu(bb->vmm_page_size_bytes); dcn2_0_soc.dram_clock_change_latency_us = fixed16_to_double_to_cpu(bb->dram_clock_change_latency_us); + // HACK!! Lower uclock latency switch time so we don't switch + dcn2_0_soc.dram_clock_change_latency_us = 10; dcn2_0_soc.writeback_dram_clock_change_latency_us = fixed16_to_double_to_cpu(bb->writeback_dram_clock_change_latency_us); dcn2_0_soc.return_bus_width_bytes = @@ -2824,6 +2829,7 @@ static bool init_soc_bounding_box(struct dc *dc, struct pp_smu_nv_clock_table max_clocks = {0}; unsigned int uclk_states[8] = {0}; unsigned int num_states = 0; + int i; enum pp_smu_status status; bool clock_limits_available = false; bool uclk_states_available = false; @@ -2845,6 +2851,10 @@ static bool init_soc_bounding_box(struct dc *dc, clock_limits_available = (status == PP_SMU_RESULT_OK); } + // HACK: Use the max uclk_states value for all elements. + for (i = 0; i < num_states; i++) + uclk_states[i] = uclk_states[num_states - 1]; + if (clock_limits_available && uclk_states_available && num_states) update_bounding_box(dc, &dcn2_0_soc, &max_clocks, uclk_states, num_states); else if (clock_limits_available) diff --git a/drivers/gpu/drm/amd/display/dc/dsc/Makefile b/drivers/gpu/drm/amd/display/dc/dsc/Makefile index c5d5b94e2604..e019cd9447e8 100644 --- a/drivers/gpu/drm/amd/display/dc/dsc/Makefile +++ b/drivers/gpu/drm/amd/display/dc/dsc/Makefile @@ -1,10 +1,18 @@ # # Makefile for the 'dsc' sub-component of DAL. -CFLAGS_rc_calc.o := -mhard-float -msse -mpreferred-stack-boundary=4 -CFLAGS_rc_calc_dpi.o := -mhard-float -msse -mpreferred-stack-boundary=4 -CFLAGS_codec_main_amd.o := -mhard-float -msse -mpreferred-stack-boundary=4 -CFLAGS_dc_dsc.o := -mhard-float -msse -mpreferred-stack-boundary=4 +ifneq ($(call cc-option, -mpreferred-stack-boundary=4),) + cc_stack_align := -mpreferred-stack-boundary=4 +else ifneq ($(call cc-option, -mstack-alignment=16),) + cc_stack_align := -mstack-alignment=16 +endif + +dsc_ccflags := -mhard-float -msse $(cc_stack_align) + +CFLAGS_rc_calc.o := $(dsc_ccflags) +CFLAGS_rc_calc_dpi.o := $(dsc_ccflags) +CFLAGS_codec_main_amd.o := $(dsc_ccflags) +CFLAGS_dc_dsc.o := $(dsc_ccflags) DSC = dc_dsc.o rc_calc.o rc_calc_dpi.o diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h index d352b8d76365..a0a7211438f2 100644 --- a/drivers/gpu/drm/amd/include/amd_shared.h +++ b/drivers/gpu/drm/amd/include/amd_shared.h @@ -26,7 +26,7 @@ #include <drm/amd_asic_type.h> -#define AMD_MAX_USEC_TIMEOUT 200000 /* 200 ms */ +#define AMD_MAX_USEC_TIMEOUT 1000000 /* 1000 ms */ /* * Chip flags diff --git a/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c b/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c index 31152d495f69..f1565c448de5 100644 --- a/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c +++ b/drivers/gpu/drm/amd/powerplay/amdgpu_smu.c @@ -69,6 +69,9 @@ int smu_set_soft_freq_range(struct smu_context *smu, enum smu_clk_type clk_type, if (min <= 0 && max <= 0) return -EINVAL; + if (!smu_clk_dpm_is_enabled(smu, clk_type)) + return 0; + clk_id = smu_clk_get_index(smu, clk_type); if (clk_id < 0) return clk_id; @@ -102,6 +105,9 @@ int smu_set_hard_freq_range(struct smu_context *smu, enum smu_clk_type clk_type, if (min <= 0 && max <= 0) return -EINVAL; + if (!smu_clk_dpm_is_enabled(smu, clk_type)) + return 0; + clk_id = smu_clk_get_index(smu, clk_type); if (clk_id < 0) return clk_id; @@ -135,23 +141,8 @@ int smu_get_dpm_freq_range(struct smu_context *smu, enum smu_clk_type clk_type, if (!min && !max) return -EINVAL; - switch (clk_type) { - case SMU_UCLK: - if (!smu_feature_is_enabled(smu, SMU_FEATURE_DPM_UCLK_BIT)) { - pr_warn("uclk dpm is not enabled\n"); - return 0; - } - break; - case SMU_GFXCLK: - case SMU_SCLK: - if (!smu_feature_is_enabled(smu, SMU_FEATURE_DPM_GFXCLK_BIT)) { - pr_warn("gfxclk dpm is not enabled\n"); - return 0; - } - break; - default: - break; - } + if (!smu_clk_dpm_is_enabled(smu, clk_type)) + return 0; mutex_lock(&smu->mutex); clk_id = smu_clk_get_index(smu, clk_type); @@ -194,6 +185,9 @@ int smu_get_dpm_freq_by_index(struct smu_context *smu, enum smu_clk_type clk_typ if (!value) return -EINVAL; + if (!smu_clk_dpm_is_enabled(smu, clk_type)) + return 0; + clk_id = smu_clk_get_index(smu, clk_type); if (clk_id < 0) return clk_id; @@ -222,6 +216,35 @@ int smu_get_dpm_level_count(struct smu_context *smu, enum smu_clk_type clk_type, return smu_get_dpm_freq_by_index(smu, clk_type, 0xff, value); } +bool smu_clk_dpm_is_enabled(struct smu_context *smu, enum smu_clk_type clk_type) +{ + enum smu_feature_mask feature_id = 0; + + switch (clk_type) { + case SMU_MCLK: + case SMU_UCLK: + feature_id = SMU_FEATURE_DPM_UCLK_BIT; + break; + case SMU_GFXCLK: + case SMU_SCLK: + feature_id = SMU_FEATURE_DPM_GFXCLK_BIT; + break; + case SMU_SOCCLK: + feature_id = SMU_FEATURE_DPM_SOCCLK_BIT; + break; + default: + return true; + } + + if(!smu_feature_is_enabled(smu, feature_id)) { + pr_warn("smu %d clk dpm feature %d is not enabled\n", clk_type, feature_id); + return false; + } + + return true; +} + + int smu_dpm_set_power_gate(struct smu_context *smu, uint32_t block_type, bool gate) { @@ -300,7 +323,7 @@ int smu_common_read_sensor(struct smu_context *smu, enum amd_pp_sensors sensor, return ret; } -int smu_update_table(struct smu_context *smu, enum smu_table_id table_index, +int smu_update_table(struct smu_context *smu, enum smu_table_id table_index, int argument, void *table_data, bool drv2smu) { struct smu_table_context *smu_table = &smu->smu_table; @@ -327,7 +350,7 @@ int smu_update_table(struct smu_context *smu, enum smu_table_id table_index, ret = smu_send_smc_msg_with_param(smu, drv2smu ? SMU_MSG_TransferTableDram2Smu : SMU_MSG_TransferTableSmu2Dram, - table_id); + table_id | ((argument & 0xFFFF) << 16)); if (ret) return ret; @@ -634,6 +657,11 @@ static int smu_sw_init(void *handle) bitmap_zero(smu->smu_feature.supported, SMU_FEATURE_MAX); bitmap_zero(smu->smu_feature.enabled, SMU_FEATURE_MAX); bitmap_zero(smu->smu_feature.allowed, SMU_FEATURE_MAX); + + mutex_init(&smu->smu_baco.mutex); + smu->smu_baco.state = SMU_BACO_STATE_EXIT; + smu->smu_baco.platform_support = false; + smu->watermarks_bitmap = 0; smu->power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT; smu->default_power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT; @@ -1103,11 +1131,20 @@ static int smu_suspend(void *handle) int ret; struct amdgpu_device *adev = (struct amdgpu_device *)handle; struct smu_context *smu = &adev->smu; + bool baco_feature_is_enabled = smu_feature_is_enabled(smu, SMU_FEATURE_BACO_BIT); ret = smu_system_features_control(smu, false); if (ret) return ret; + if (adev->in_gpu_reset && baco_feature_is_enabled) { + ret = smu_feature_set_enabled(smu, SMU_FEATURE_BACO_BIT, true); + if (ret) { + pr_warn("set BACO feature enabled failed, return %d\n", ret); + return ret; + } + } + smu->watermarks_bitmap &= ~(WATERMARKS_LOADED); if (adev->asic_type >= CHIP_NAVI10 && @@ -1358,10 +1395,10 @@ int smu_adjust_power_state_dynamic(struct smu_context *smu, break; case AMD_DPM_FORCED_LEVEL_AUTO: + case AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD: ret = smu_unforce_dpm_levels(smu); break; - case AMD_DPM_FORCED_LEVEL_PROFILE_STANDARD: case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_SCLK: case AMD_DPM_FORCED_LEVEL_PROFILE_MIN_MCLK: case AMD_DPM_FORCED_LEVEL_PROFILE_PEAK: @@ -1371,8 +1408,9 @@ int smu_adjust_power_state_dynamic(struct smu_context *smu, &soc_mask); if (ret) return ret; - smu_force_clk_levels(smu, PP_SCLK, 1 << sclk_mask); - smu_force_clk_levels(smu, PP_MCLK, 1 << mclk_mask); + smu_force_clk_levels(smu, SMU_SCLK, 1 << sclk_mask); + smu_force_clk_levels(smu, SMU_MCLK, 1 << mclk_mask); + smu_force_clk_levels(smu, SMU_SOCCLK, 1 << soc_mask); break; case AMD_DPM_FORCED_LEVEL_MANUAL: @@ -1427,17 +1465,16 @@ int smu_handle_task(struct smu_context *smu, enum amd_dpm_forced_level smu_get_performance_level(struct smu_context *smu) { struct smu_dpm_context *smu_dpm_ctx = &(smu->smu_dpm); + enum amd_dpm_forced_level level; if (!smu_dpm_ctx->dpm_context) return -EINVAL; mutex_lock(&(smu->mutex)); - if (smu_dpm_ctx->dpm_level != smu_dpm_ctx->saved_dpm_level) { - smu_dpm_ctx->saved_dpm_level = smu_dpm_ctx->dpm_level; - } + level = smu_dpm_ctx->dpm_level; mutex_unlock(&(smu->mutex)); - return smu_dpm_ctx->dpm_level; + return level; } int smu_force_performance_level(struct smu_context *smu, enum amd_dpm_forced_level level) diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c index 0b0d83c2a678..a24beaa4fb01 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/hwmgr.c @@ -327,7 +327,7 @@ int hwmgr_resume(struct pp_hwmgr *hwmgr) if (ret) return ret; - ret = psm_adjust_power_state_dynamic(hwmgr, true, NULL); + ret = psm_adjust_power_state_dynamic(hwmgr, false, NULL); return ret; } 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 ae64ff7153d6..b760f95e7fa7 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 @@ -916,8 +916,10 @@ static int init_thermal_controller( PHM_PlatformCaps_ThermalController ); - if (0 == powerplay_table->usFanTableOffset) + if (0 == powerplay_table->usFanTableOffset) { + hwmgr->thermal_controller.use_hw_fan_control = 1; return 0; + } fan_table = (const PPTable_Generic_SubTable_Header *) (((unsigned long)powerplay_table) + @@ -1065,8 +1067,6 @@ static int pp_tables_v1_0_initialize(struct pp_hwmgr *hwmgr) PP_ASSERT_WITH_CODE((NULL != hwmgr->pptable), "Failed to allocate hwmgr->pptable!", return -ENOMEM); - memset(hwmgr->pptable, 0x00, sizeof(struct phm_ppt_v1_information)); - powerplay_table = get_powerplay_table(hwmgr); PP_ASSERT_WITH_CODE((NULL != powerplay_table), diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c index c5986d28fbf1..487aeee1cf8a 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/smu7_hwmgr.c @@ -3495,7 +3495,7 @@ static int smu7_get_gpu_power(struct pp_hwmgr *hwmgr, u32 *query) ixSMU_PM_STATUS_95, 0); for (i = 0; i < 10; i++) { - mdelay(500); + msleep(500); smum_send_msg_to_smc(hwmgr, PPSMC_MSG_PmStatusLogSample); tmp = cgs_read_ind_register(hwmgr->device, CGS_IND_REG__SMC, diff --git a/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h index 2818df46481c..1af992fb0bde 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h +++ b/drivers/gpu/drm/amd/powerplay/inc/amdgpu_smu.h @@ -241,6 +241,7 @@ enum smu_message_type SMU_MSG_PowerUpJpeg, SMU_MSG_PowerDownJpeg, SMU_MSG_BacoAudioD3PME, + SMU_MSG_ArmD3, SMU_MSG_MAX_COUNT, }; @@ -489,6 +490,19 @@ struct mclock_latency_table { struct mclk_latency_entries entries[MAX_REGULAR_DPM_NUM]; }; +enum smu_baco_state +{ + SMU_BACO_STATE_ENTER = 0, + SMU_BACO_STATE_EXIT, +}; + +struct smu_baco_context +{ + struct mutex mutex; + uint32_t state; + bool platform_support; +}; + #define WORKLOAD_POLICY_MAX 7 struct smu_context { @@ -505,6 +519,7 @@ struct smu_context struct smu_power_context smu_power; struct smu_feature smu_feature; struct amd_pp_display_configuration *display_config; + struct smu_baco_context smu_baco; void *od_settings; uint32_t pstate_sclk; @@ -680,6 +695,11 @@ struct smu_funcs int (*register_irq_handler)(struct smu_context *smu); int (*set_azalia_d3_pme)(struct smu_context *smu); int (*get_max_sustainable_clocks_by_dc)(struct smu_context *smu, struct pp_smu_nv_clock_table *max_clocks); + bool (*baco_is_support)(struct smu_context *smu); + enum smu_baco_state (*baco_get_state)(struct smu_context *smu); + int (*baco_set_state)(struct smu_context *smu, enum smu_baco_state state); + int (*baco_reset)(struct smu_context *smu); + }; #define smu_init_microcode(smu) \ @@ -892,6 +912,12 @@ struct smu_funcs ((smu)->funcs->get_max_sustainable_clocks_by_dc ? (smu)->funcs->get_max_sustainable_clocks_by_dc((smu), (max_clocks)) : 0) #define smu_get_uclk_dpm_states(smu, clocks_in_khz, num_states) \ ((smu)->ppt_funcs->get_uclk_dpm_states ? (smu)->ppt_funcs->get_uclk_dpm_states((smu), (clocks_in_khz), (num_states)) : 0) +#define smu_baco_is_support(smu) \ + ((smu)->funcs->baco_is_support? (smu)->funcs->baco_is_support((smu)) : false) +#define smu_baco_get_state(smu, state) \ + ((smu)->funcs->baco_get_state? (smu)->funcs->baco_get_state((smu), (state)) : 0) +#define smu_baco_reset(smu) \ + ((smu)->funcs->baco_reset? (smu)->funcs->baco_reset((smu)) : 0) extern int smu_get_atom_data_table(struct smu_context *smu, uint32_t table, uint16_t *size, uint8_t *frev, uint8_t *crev, @@ -911,7 +937,7 @@ extern int smu_feature_is_supported(struct smu_context *smu, extern int smu_feature_set_supported(struct smu_context *smu, enum smu_feature_mask mask, bool enable); -int smu_update_table(struct smu_context *smu, uint32_t table_index, +int smu_update_table(struct smu_context *smu, enum smu_table_id table_index, int argument, void *table_data, bool drv2smu); bool is_support_sw_smu(struct amdgpu_device *adev); @@ -947,5 +973,6 @@ int smu_set_hard_freq_range(struct smu_context *smu, enum smu_clk_type clk_type, enum amd_dpm_forced_level smu_get_performance_level(struct smu_context *smu); int smu_force_performance_level(struct smu_context *smu, enum amd_dpm_forced_level level); int smu_set_display_count(struct smu_context *smu, uint32_t count); +bool smu_clk_dpm_is_enabled(struct smu_context *smu, enum smu_clk_type clk_type); #endif diff --git a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h index 47dbecca5adb..c5989cb38b1b 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h +++ b/drivers/gpu/drm/amd/powerplay/inc/hwmgr.h @@ -695,6 +695,7 @@ struct pp_thermal_controller_info { uint8_t ucType; uint8_t ucI2cLine; uint8_t ucI2cAddress; + uint8_t use_hw_fan_control; struct pp_fan_info fanInfo; struct pp_advance_fan_control_parameters advanceFanControlParameters; }; diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if.h b/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if.h index 195c4ae67058..755d51f9c6a9 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if.h +++ b/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if.h @@ -27,7 +27,7 @@ // *** IMPORTANT *** // SMU TEAM: Always increment the interface version if // any structure is changed in this file -#define SMU11_DRIVER_IF_VERSION 0x12 +#define SMU11_DRIVER_IF_VERSION 0x13 #define PPTABLE_V20_SMU_VERSION 3 @@ -615,6 +615,7 @@ typedef struct { uint16_t UclkAverageLpfTau; uint16_t GfxActivityLpfTau; uint16_t UclkActivityLpfTau; + uint16_t SocketPowerLpfTau; uint32_t MmHubPadding[8]; @@ -665,7 +666,8 @@ typedef struct { uint32_t ThrottlerStatus ; uint8_t LinkDpmLevel; - uint8_t Padding[3]; + uint16_t AverageSocketPower; + uint8_t Padding; uint32_t MmHubPadding[7]; diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if_navi10.h b/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if_navi10.h index a8b31bc50054..adbbfebbb1e5 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if_navi10.h +++ b/drivers/gpu/drm/amd/powerplay/inc/smu11_driver_if_navi10.h @@ -26,7 +26,7 @@ // *** IMPORTANT *** // SMU TEAM: Always increment the interface version if // any structure is changed in this file -#define SMU11_DRIVER_IF_VERSION 0x32 +#define SMU11_DRIVER_IF_VERSION 0x33 #define PPTABLE_NV10_SMU_VERSION 8 @@ -813,8 +813,8 @@ typedef struct { uint16_t UclkAverageLpfTau; uint16_t GfxActivityLpfTau; uint16_t UclkActivityLpfTau; + uint16_t SocketPowerLpfTau; - uint16_t Padding; // Padding - ignore uint32_t MmHubPadding[8]; // SMU internal use } DriverSmuConfig_t; @@ -853,7 +853,7 @@ typedef struct { uint8_t CurrGfxVoltageOffset ; uint8_t CurrMemVidOffset ; uint8_t Padding8 ; - uint16_t CurrSocketPower ; + uint16_t AverageSocketPower ; uint16_t TemperatureEdge ; uint16_t TemperatureHotspot ; uint16_t TemperatureMem ; diff --git a/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h index d93cd76269b4..2fff4b16cb4e 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h +++ b/drivers/gpu/drm/amd/powerplay/inc/smu_v11_0.h @@ -105,6 +105,14 @@ struct smu_11_0_power_context { enum smu_11_0_power_state power_state; }; +enum smu_v11_0_baco_seq { + BACO_SEQ_BACO = 0, + BACO_SEQ_MSR, + BACO_SEQ_BAMACO, + BACO_SEQ_ULPS, + BACO_SEQ_COUNT, +}; + void smu_v11_0_set_smu_funcs(struct smu_context *smu); #endif diff --git a/drivers/gpu/drm/amd/powerplay/navi10_ppt.c b/drivers/gpu/drm/amd/powerplay/navi10_ppt.c index 27e5c8088f1b..2dae0ae0829e 100644 --- a/drivers/gpu/drm/amd/powerplay/navi10_ppt.c +++ b/drivers/gpu/drm/amd/powerplay/navi10_ppt.c @@ -115,6 +115,7 @@ static int navi10_message_map[SMU_MSG_MAX_COUNT] = { MSG_MAP(PowerUpJpeg, PPSMC_MSG_PowerUpJpeg), MSG_MAP(PowerDownJpeg, PPSMC_MSG_PowerDownJpeg), MSG_MAP(BacoAudioD3PME, PPSMC_MSG_BacoAudioD3PME), + MSG_MAP(ArmD3, PPSMC_MSG_ArmD3), }; static int navi10_clk_map[SMU_CLK_COUNT] = { @@ -330,7 +331,10 @@ navi10_get_allowed_feature_mask(struct smu_context *smu, | FEATURE_MASK(FEATURE_DS_DCEFCLK_BIT) | FEATURE_MASK(FEATURE_FW_DSTATE_BIT) | FEATURE_MASK(FEATURE_BACO_BIT) - | FEATURE_MASK(FEATURE_ACDC_BIT); + | FEATURE_MASK(FEATURE_ACDC_BIT) + | FEATURE_MASK(FEATURE_GFX_SS_BIT) + | FEATURE_MASK(FEATURE_APCC_DFLL_BIT) + | FEATURE_MASK(FEATURE_FW_CTF_BIT); if (adev->pm.pp_feature & PP_MCLK_DPM_MASK) *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_DPM_UCLK_BIT) @@ -338,8 +342,7 @@ navi10_get_allowed_feature_mask(struct smu_context *smu, | FEATURE_MASK(FEATURE_MEM_MVDD_SCALING_BIT); if (adev->pm.pp_feature & PP_GFXOFF_MASK) { - *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_GFX_SS_BIT) - | FEATURE_MASK(FEATURE_GFXOFF_BIT); + *(uint64_t *)feature_mask |= FEATURE_MASK(FEATURE_GFXOFF_BIT); /* TODO: remove it once fw fix the bug */ *(uint64_t *)feature_mask &= ~FEATURE_MASK(FEATURE_FW_DSTATE_BIT); } @@ -464,9 +467,6 @@ static int navi10_append_powerplay_table(struct smu_context *smu) smc_pptable->MvddRatio = smc_dpm_table->MvddRatio; if (adev->pm.pp_feature & PP_GFXOFF_MASK) { - *(uint64_t *)smc_pptable->FeaturesToRun |= FEATURE_MASK(FEATURE_GFX_SS_BIT) - | FEATURE_MASK(FEATURE_GFXOFF_BIT); - /* TODO: remove it once SMU fw fix it */ smc_pptable->DebugOverrides |= DPM_OVERRIDE_DISABLE_DFLL_PLL_SHUTDOWN; } @@ -478,6 +478,7 @@ static int navi10_store_powerplay_table(struct smu_context *smu) { struct smu_11_0_powerplay_table *powerplay_table = NULL; struct smu_table_context *table_context = &smu->smu_table; + struct smu_baco_context *smu_baco = &smu->smu_baco; if (!table_context->power_play_table) return -EINVAL; @@ -489,6 +490,12 @@ static int navi10_store_powerplay_table(struct smu_context *smu) table_context->thermal_controller_type = powerplay_table->thermal_controller_type; + mutex_lock(&smu_baco->mutex); + if (powerplay_table->platform_caps & SMU_11_0_PP_PLATFORM_CAP_BACO || + powerplay_table->platform_caps & SMU_11_0_PP_PLATFORM_CAP_MACO) + smu_baco->platform_support = true; + mutex_unlock(&smu_baco->mutex); + return 0; } @@ -598,13 +605,15 @@ static int navi10_get_current_clk_freq_by_table(struct smu_context *smu, enum smu_clk_type clk_type, uint32_t *value) { - static SmuMetrics_t metrics = {0}; + static SmuMetrics_t metrics; int ret = 0, clk_id = 0; if (!value) return -EINVAL; - ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, (void *)&metrics, false); + memset(&metrics, 0, sizeof(metrics)); + + ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, 0, (void *)&metrics, false); if (ret) return ret; @@ -699,7 +708,7 @@ static int navi10_force_clk_levels(struct smu_context *smu, static int navi10_populate_umd_state_clk(struct smu_context *smu) { int ret = 0; - uint32_t min_sclk_freq = 0; + uint32_t min_sclk_freq = 0, min_mclk_freq = 0; ret = smu_get_dpm_freq_range(smu, SMU_SCLK, &min_sclk_freq, NULL); if (ret) @@ -707,6 +716,12 @@ static int navi10_populate_umd_state_clk(struct smu_context *smu) smu->pstate_sclk = min_sclk_freq * 100; + ret = smu_get_dpm_freq_range(smu, SMU_MCLK, &min_mclk_freq, NULL); + if (ret) + return ret; + + smu->pstate_mclk = min_mclk_freq * 100; + return ret; } @@ -817,27 +832,20 @@ static int navi10_force_dpm_limit_value(struct smu_context *smu, bool highest) return ret; } -static int navi10_unforce_dpm_levels(struct smu_context *smu) { - +static int navi10_unforce_dpm_levels(struct smu_context *smu) +{ int ret = 0, i = 0; uint32_t min_freq, max_freq; enum smu_clk_type clk_type; - struct clk_feature_map { - enum smu_clk_type clk_type; - uint32_t feature; - } clk_feature_map[] = { - {SMU_GFXCLK, SMU_FEATURE_DPM_GFXCLK_BIT}, - {SMU_MCLK, SMU_FEATURE_DPM_UCLK_BIT}, - {SMU_SOCCLK, SMU_FEATURE_DPM_SOCCLK_BIT}, + enum smu_clk_type clks[] = { + SMU_GFXCLK, + SMU_MCLK, + SMU_SOCCLK, }; - for (i = 0; i < ARRAY_SIZE(clk_feature_map); i++) { - if (!smu_feature_is_enabled(smu, clk_feature_map[i].feature)) - continue; - - clk_type = clk_feature_map[i].clk_type; - + for (i = 0; i < ARRAY_SIZE(clks); i++) { + clk_type = clks[i]; ret = smu_get_dpm_freq_range(smu, clk_type, &min_freq, &max_freq); if (ret) return ret; @@ -858,17 +866,18 @@ static int navi10_get_gpu_power(struct smu_context *smu, uint32_t *value) if (!value) return -EINVAL; - ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, (void *)&metrics, + ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, 0, (void *)&metrics, false); if (ret) return ret; - *value = metrics.CurrSocketPower << 8; + *value = metrics.AverageSocketPower << 8; return 0; } static int navi10_get_current_activity_percent(struct smu_context *smu, + enum amd_pp_sensors sensor, uint32_t *value) { int ret = 0; @@ -879,12 +888,22 @@ static int navi10_get_current_activity_percent(struct smu_context *smu, msleep(1); - ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, + ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, 0, (void *)&metrics, false); if (ret) return ret; - *value = metrics.AverageGfxActivity; + switch (sensor) { + case AMDGPU_PP_SENSOR_GPU_LOAD: + *value = metrics.AverageGfxActivity; + break; + case AMDGPU_PP_SENSOR_MEM_LOAD: + *value = metrics.AverageUclkActivity; + break; + default: + pr_err("Invalid sensor for retrieving clock activity\n"); + return -EINVAL; + } return 0; } @@ -902,13 +921,15 @@ static bool navi10_is_dpm_running(struct smu_context *smu) static int navi10_get_fan_speed(struct smu_context *smu, uint16_t *value) { - SmuMetrics_t metrics = {0}; + SmuMetrics_t metrics; int ret = 0; if (!value) return -EINVAL; - ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, + memset(&metrics, 0, sizeof(metrics)); + + ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, 0, (void *)&metrics, false); if (ret) return ret; @@ -974,7 +995,7 @@ static int navi10_get_power_profile_mode(struct smu_context *smu, char *buf) /* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */ workload_type = smu_workload_get_type(smu, i); result = smu_update_table(smu, - SMU_TABLE_ACTIVITY_MONITOR_COEFF | workload_type << 16, + SMU_TABLE_ACTIVITY_MONITOR_COEFF, workload_type, (void *)(&activity_monitor), false); if (result) { pr_err("[%s] Failed to get activity monitor!", __func__); @@ -1047,7 +1068,7 @@ static int navi10_set_power_profile_mode(struct smu_context *smu, long *input, u return -EINVAL; ret = smu_update_table(smu, - SMU_TABLE_ACTIVITY_MONITOR_COEFF | WORKLOAD_PPLIB_CUSTOM_BIT << 16, + SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT, (void *)(&activity_monitor), false); if (ret) { pr_err("[%s] Failed to get activity monitor!", __func__); @@ -1091,7 +1112,7 @@ static int navi10_set_power_profile_mode(struct smu_context *smu, long *input, u } ret = smu_update_table(smu, - SMU_TABLE_ACTIVITY_MONITOR_COEFF | WORKLOAD_PPLIB_CUSTOM_BIT << 16, + SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT, (void *)(&activity_monitor), true); if (ret) { pr_err("[%s] Failed to set activity monitor!", __func__); @@ -1134,14 +1155,14 @@ static int navi10_get_profiling_clk_mask(struct smu_context *smu, ret = smu_get_dpm_level_count(smu, SMU_MCLK, &level_count); if (ret) return ret; - *sclk_mask = level_count - 1; + *mclk_mask = level_count - 1; } if(soc_mask) { ret = smu_get_dpm_level_count(smu, SMU_SOCCLK, &level_count); if (ret) return ret; - *sclk_mask = level_count - 1; + *soc_mask = level_count - 1; } } @@ -1247,6 +1268,41 @@ static int navi10_set_watermarks_table(struct smu_context *smu, return 0; } +static int navi10_thermal_get_temperature(struct smu_context *smu, + enum amd_pp_sensors sensor, + uint32_t *value) +{ + SmuMetrics_t metrics; + int ret = 0; + + if (!value) + return -EINVAL; + + ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, 0, (void *)&metrics, false); + if (ret) + return ret; + + switch (sensor) { + case AMDGPU_PP_SENSOR_HOTSPOT_TEMP: + *value = metrics.TemperatureHotspot * + SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; + break; + case AMDGPU_PP_SENSOR_EDGE_TEMP: + *value = metrics.TemperatureEdge * + SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; + break; + case AMDGPU_PP_SENSOR_MEM_TEMP: + *value = metrics.TemperatureMem * + SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; + break; + default: + pr_err("Invalid sensor for retrieving temp\n"); + return -EINVAL; + } + + return 0; +} + static int navi10_read_sensor(struct smu_context *smu, enum amd_pp_sensors sensor, void *data, uint32_t *size) @@ -1260,14 +1316,21 @@ static int navi10_read_sensor(struct smu_context *smu, *(uint32_t *)data = pptable->FanMaximumRpm; *size = 4; break; + case AMDGPU_PP_SENSOR_MEM_LOAD: case AMDGPU_PP_SENSOR_GPU_LOAD: - ret = navi10_get_current_activity_percent(smu, (uint32_t *)data); + ret = navi10_get_current_activity_percent(smu, sensor, (uint32_t *)data); *size = 4; break; case AMDGPU_PP_SENSOR_GPU_POWER: ret = navi10_get_gpu_power(smu, (uint32_t *)data); *size = 4; break; + case AMDGPU_PP_SENSOR_HOTSPOT_TEMP: + case AMDGPU_PP_SENSOR_EDGE_TEMP: + case AMDGPU_PP_SENSOR_MEM_TEMP: + ret = navi10_thermal_get_temperature(smu, sensor, (uint32_t *)data); + *size = 4; + break; default: return -EINVAL; } diff --git a/drivers/gpu/drm/amd/powerplay/smu_v11_0.c b/drivers/gpu/drm/amd/powerplay/smu_v11_0.c index 632a20587c8b..95c7c4dae523 100644 --- a/drivers/gpu/drm/amd/powerplay/smu_v11_0.c +++ b/drivers/gpu/drm/amd/powerplay/smu_v11_0.c @@ -40,6 +40,7 @@ #include "asic_reg/mp/mp_11_0_offset.h" #include "asic_reg/mp/mp_11_0_sh_mask.h" #include "asic_reg/nbio/nbio_7_4_offset.h" +#include "asic_reg/nbio/nbio_7_4_sh_mask.h" #include "asic_reg/smuio/smuio_11_0_0_offset.h" #include "asic_reg/smuio/smuio_11_0_0_sh_mask.h" @@ -67,9 +68,9 @@ static int smu_v11_0_read_arg(struct smu_context *smu, uint32_t *arg) static int smu_v11_0_wait_for_response(struct smu_context *smu) { struct amdgpu_device *adev = smu->adev; - uint32_t cur_value, i; + uint32_t cur_value, i, timeout = adev->usec_timeout * 10; - for (i = 0; i < adev->usec_timeout; i++) { + for (i = 0; i < timeout; i++) { cur_value = RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90); if ((cur_value & MP1_C2PMSG_90__CONTENT_MASK) != 0) break; @@ -77,7 +78,7 @@ static int smu_v11_0_wait_for_response(struct smu_context *smu) } /* timeout means wrong logic */ - if (i == adev->usec_timeout) + if (i == timeout) return -ETIME; return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90) == 0x1 ? 0 : -EIO; @@ -260,14 +261,20 @@ static int smu_v11_0_check_fw_version(struct smu_context *smu) smu_minor = (smu_version >> 8) & 0xff; smu_debug = (smu_version >> 0) & 0xff; - + /* + * 1. if_version mismatch is not critical as our fw is designed + * to be backward compatible. + * 2. New fw usually brings some optimizations. But that's visible + * only on the paired driver. + * Considering above, we just leave user a warning message instead + * of halt driver loading. + */ if (if_version != smu->smc_if_version) { pr_info("smu driver if version = 0x%08x, smu fw if version = 0x%08x, " "smu fw version = 0x%08x (%d.%d.%d)\n", smu->smc_if_version, if_version, smu_version, smu_major, smu_minor, smu_debug); - pr_err("SMU driver if version not matched\n"); - ret = -EINVAL; + pr_warn("SMU driver if version not matched\n"); } return ret; @@ -702,7 +709,7 @@ static int smu_v11_0_write_pptable(struct smu_context *smu) struct smu_table_context *table_context = &smu->smu_table; int ret = 0; - ret = smu_update_table(smu, SMU_TABLE_PPTABLE, + ret = smu_update_table(smu, SMU_TABLE_PPTABLE, 0, table_context->driver_pptable, true); return ret; @@ -721,7 +728,7 @@ static int smu_v11_0_write_watermarks_table(struct smu_context *smu) if (!table->cpu_addr) return -EINVAL; - ret = smu_update_table(smu, SMU_TABLE_WATERMARKS, table->cpu_addr, + ret = smu_update_table(smu, SMU_TABLE_WATERMARKS, 0, table->cpu_addr, true); return ret; @@ -1088,7 +1095,7 @@ static int smu_v11_0_get_current_clk_freq(struct smu_context *smu, uint32_t *value) { int ret = 0; - uint32_t freq; + uint32_t freq = 0; if (clk_id >= SMU_CLK_COUNT || !value) return -EINVAL; @@ -1642,6 +1649,92 @@ static int smu_v11_0_set_azalia_d3_pme(struct smu_context *smu) return ret; } +static int smu_v11_0_baco_set_armd3_sequence(struct smu_context *smu, enum smu_v11_0_baco_seq baco_seq) +{ + return smu_send_smc_msg_with_param(smu, SMU_MSG_ArmD3, baco_seq); +} + +static bool smu_v11_0_baco_is_support(struct smu_context *smu) +{ + struct amdgpu_device *adev = smu->adev; + struct smu_baco_context *smu_baco = &smu->smu_baco; + uint32_t val; + bool baco_support; + + mutex_lock(&smu_baco->mutex); + baco_support = smu_baco->platform_support; + mutex_unlock(&smu_baco->mutex); + + if (!baco_support) + return false; + + if (!smu_feature_is_enabled(smu, SMU_FEATURE_BACO_BIT)) + return false; + + val = RREG32_SOC15(NBIO, 0, mmRCC_BIF_STRAP0); + if (val & RCC_BIF_STRAP0__STRAP_PX_CAPABLE_MASK) + return true; + + return false; +} + +static enum smu_baco_state smu_v11_0_baco_get_state(struct smu_context *smu) +{ + struct smu_baco_context *smu_baco = &smu->smu_baco; + enum smu_baco_state baco_state = SMU_BACO_STATE_EXIT; + + mutex_lock(&smu_baco->mutex); + baco_state = smu_baco->state; + mutex_unlock(&smu_baco->mutex); + + return baco_state; +} + +static int smu_v11_0_baco_set_state(struct smu_context *smu, enum smu_baco_state state) +{ + + struct smu_baco_context *smu_baco = &smu->smu_baco; + int ret = 0; + + if (smu_v11_0_baco_get_state(smu) == state) + return 0; + + mutex_lock(&smu_baco->mutex); + + if (state == SMU_BACO_STATE_ENTER) + ret = smu_send_smc_msg_with_param(smu, SMU_MSG_EnterBaco, BACO_SEQ_BACO); + else + ret = smu_send_smc_msg(smu, SMU_MSG_ExitBaco); + if (ret) + goto out; + + smu_baco->state = state; +out: + mutex_unlock(&smu_baco->mutex); + return ret; +} + +static int smu_v11_0_baco_reset(struct smu_context *smu) +{ + int ret = 0; + + ret = smu_v11_0_baco_set_armd3_sequence(smu, BACO_SEQ_BACO); + if (ret) + return ret; + + ret = smu_v11_0_baco_set_state(smu, SMU_BACO_STATE_ENTER); + if (ret) + return ret; + + msleep(10); + + ret = smu_v11_0_baco_set_state(smu, SMU_BACO_STATE_EXIT); + if (ret) + return ret; + + return ret; +} + static const struct smu_funcs smu_v11_0_funcs = { .init_microcode = smu_v11_0_init_microcode, .load_microcode = smu_v11_0_load_microcode, @@ -1690,6 +1783,10 @@ static const struct smu_funcs smu_v11_0_funcs = { .register_irq_handler = smu_v11_0_register_irq_handler, .set_azalia_d3_pme = smu_v11_0_set_azalia_d3_pme, .get_max_sustainable_clocks_by_dc = smu_v11_0_get_max_sustainable_clocks_by_dc, + .baco_is_support = smu_v11_0_baco_is_support, + .baco_get_state = smu_v11_0_baco_get_state, + .baco_set_state = smu_v11_0_baco_set_state, + .baco_reset = smu_v11_0_baco_reset, }; void smu_v11_0_set_smu_funcs(struct smu_context *smu) diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c index 6c81cb91ebae..15590fd86ef4 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/ci_smumgr.c @@ -2705,8 +2705,6 @@ static int ci_initialize_mc_reg_table(struct pp_hwmgr *hwmgr) cgs_write_register(hwmgr->device, mmMC_SEQ_PMG_CMD_MRS2_LP, cgs_read_register(hwmgr->device, mmMC_PMG_CMD_MRS2)); cgs_write_register(hwmgr->device, mmMC_SEQ_WR_CTL_2_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_WR_CTL_2)); - memset(table, 0x00, sizeof(pp_atomctrl_mc_reg_table)); - result = atomctrl_initialize_mc_reg_table(hwmgr, module_index, table); if (0 == result) diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c index 9e0dd56fe7c5..732005c03a82 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/iceland_smumgr.c @@ -2634,8 +2634,6 @@ static int iceland_initialize_mc_reg_table(struct pp_hwmgr *hwmgr) cgs_write_register(hwmgr->device, mmMC_SEQ_PMG_CMD_MRS2_LP, cgs_read_register(hwmgr->device, mmMC_PMG_CMD_MRS2)); cgs_write_register(hwmgr->device, mmMC_SEQ_WR_CTL_2_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_WR_CTL_2)); - memset(table, 0x00, sizeof(pp_atomctrl_mc_reg_table)); - result = atomctrl_initialize_mc_reg_table(hwmgr, module_index, table); if (0 == result) diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c index 196cd3a429e5..dc754447f0dd 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/polaris10_smumgr.c @@ -2094,6 +2094,10 @@ static int polaris10_thermal_setup_fan_table(struct pp_hwmgr *hwmgr) return 0; } + /* use hardware fan control */ + if (hwmgr->thermal_controller.use_hw_fan_control) + return 0; + tmp64 = hwmgr->thermal_controller.advanceFanControlParameters. usPWMMin * duty100; do_div(tmp64, 10000); diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c index ba3394303b9c..f19bac7ef7ba 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/tonga_smumgr.c @@ -3117,8 +3117,6 @@ static int tonga_initialize_mc_reg_table(struct pp_hwmgr *hwmgr) cgs_write_register(hwmgr->device, mmMC_SEQ_WR_CTL_2_LP, cgs_read_register(hwmgr->device, mmMC_SEQ_WR_CTL_2)); - memset(table, 0x00, sizeof(pp_atomctrl_mc_reg_table)); - result = atomctrl_initialize_mc_reg_table(hwmgr, module_index, table); if (!result) diff --git a/drivers/gpu/drm/amd/powerplay/vega20_ppt.c b/drivers/gpu/drm/amd/powerplay/vega20_ppt.c index 0f14fe14ecd8..bb9bb09cfc7a 100644 --- a/drivers/gpu/drm/amd/powerplay/vega20_ppt.c +++ b/drivers/gpu/drm/amd/powerplay/vega20_ppt.c @@ -319,7 +319,7 @@ static int vega20_tables_init(struct smu_context *smu, struct smu_table *tables) AMDGPU_GEM_DOMAIN_VRAM); smu_table->metrics_table = kzalloc(sizeof(SmuMetrics_t), GFP_KERNEL); - if (smu_table->metrics_table) + if (!smu_table->metrics_table) return -ENOMEM; smu_table->metrics_time = 0; @@ -441,7 +441,6 @@ static int vega20_store_powerplay_table(struct smu_context *smu) { ATOM_Vega20_POWERPLAYTABLE *powerplay_table = NULL; struct smu_table_context *table_context = &smu->smu_table; - int ret; if (!table_context->power_play_table) return -EINVAL; @@ -455,9 +454,7 @@ static int vega20_store_powerplay_table(struct smu_context *smu) table_context->thermal_controller_type = powerplay_table->ucThermalControllerType; table_context->TDPODLimit = le32_to_cpu(powerplay_table->OverDrive8Table.ODSettingsMax[ATOM_VEGA20_ODSETTING_POWERPERCENTAGE]); - ret = vega20_setup_od8_information(smu); - - return ret; + return 0; } static int vega20_append_powerplay_table(struct smu_context *smu) @@ -992,7 +989,7 @@ static int vega20_print_clk_levels(struct smu_context *smu, break; case SMU_SOCCLK: - ret = smu_get_current_clk_freq(smu, PPCLK_SOCCLK, &now); + ret = smu_get_current_clk_freq(smu, SMU_SOCCLK, &now); if (ret) { pr_err("Attempt to get current socclk Failed!"); return ret; @@ -1013,7 +1010,7 @@ static int vega20_print_clk_levels(struct smu_context *smu, break; case SMU_FCLK: - ret = smu_get_current_clk_freq(smu, PPCLK_FCLK, &now); + ret = smu_get_current_clk_freq(smu, SMU_FCLK, &now); if (ret) { pr_err("Attempt to get current fclk Failed!"); return ret; @@ -1028,7 +1025,7 @@ static int vega20_print_clk_levels(struct smu_context *smu, break; case SMU_DCEFCLK: - ret = smu_get_current_clk_freq(smu, PPCLK_DCEFCLK, &now); + ret = smu_get_current_clk_freq(smu, SMU_DCEFCLK, &now); if (ret) { pr_err("Attempt to get current dcefclk Failed!"); return ret; @@ -1502,11 +1499,17 @@ static int vega20_set_default_od8_setttings(struct smu_context *smu) od8_settings = kzalloc(sizeof(struct vega20_od8_settings), GFP_KERNEL); - if (od8_settings) + if (!od8_settings) return -ENOMEM; smu->od_settings = (void *)od8_settings; + ret = vega20_setup_od8_information(smu); + if (ret) { + pr_err("Retrieve board OD limits failed!\n"); + return ret; + } + if (smu_feature_is_enabled(smu, SMU_FEATURE_DPM_SOCCLK_BIT)) { if (od8_settings->od_feature_capabilities[ATOM_VEGA20_ODFEATURE_GFXCLK_LIMITS] && od8_settings->od_settings_max[OD8_SETTING_GFXCLK_FMAX] > 0 && @@ -1677,7 +1680,7 @@ static int vega20_get_metrics_table(struct smu_context *smu, int ret = 0; if (!smu_table->metrics_time || time_after(jiffies, smu_table->metrics_time + HZ / 1000)) { - ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, + ret = smu_update_table(smu, SMU_TABLE_SMU_METRICS, 0, (void *)smu_table->metrics_table, false); if (ret) { pr_info("Failed to export SMU metrics table!\n"); @@ -1706,7 +1709,7 @@ static int vega20_set_default_od_settings(struct smu_context *smu, if (!table_context->overdrive_table) return -ENOMEM; - ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, + ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, 0, table_context->overdrive_table, false); if (ret) { pr_err("Failed to export over drive table!\n"); @@ -1718,7 +1721,7 @@ static int vega20_set_default_od_settings(struct smu_context *smu, return ret; } - ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, + ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, 0, table_context->overdrive_table, true); if (ret) { pr_err("Failed to import over drive table!\n"); @@ -1802,7 +1805,7 @@ static int vega20_get_power_profile_mode(struct smu_context *smu, char *buf) /* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */ workload_type = smu_workload_get_type(smu, i); result = smu_update_table(smu, - SMU_TABLE_ACTIVITY_MONITOR_COEFF | workload_type << 16, + SMU_TABLE_ACTIVITY_MONITOR_COEFF, workload_type, (void *)(&activity_monitor), false); if (result) { pr_err("[%s] Failed to get activity monitor!", __func__); @@ -1888,7 +1891,7 @@ static int vega20_set_power_profile_mode(struct smu_context *smu, long *input, u if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) { ret = smu_update_table(smu, - SMU_TABLE_ACTIVITY_MONITOR_COEFF | WORKLOAD_PPLIB_CUSTOM_BIT << 16, + SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT, (void *)(&activity_monitor), false); if (ret) { pr_err("[%s] Failed to get activity monitor!", __func__); @@ -1943,7 +1946,7 @@ static int vega20_set_power_profile_mode(struct smu_context *smu, long *input, u } ret = smu_update_table(smu, - SMU_TABLE_ACTIVITY_MONITOR_COEFF | WORKLOAD_PPLIB_CUSTOM_BIT << 16, + SMU_TABLE_ACTIVITY_MONITOR_COEFF, WORKLOAD_PPLIB_CUSTOM_BIT, (void *)(&activity_monitor), true); if (ret) { pr_err("[%s] Failed to set activity monitor!", __func__); @@ -2492,7 +2495,7 @@ static int vega20_update_od8_settings(struct smu_context *smu, struct smu_table_context *table_context = &smu->smu_table; int ret; - ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, + ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, 0, table_context->overdrive_table, false); if (ret) { pr_err("Failed to export over drive table!\n"); @@ -2503,7 +2506,7 @@ static int vega20_update_od8_settings(struct smu_context *smu, if (ret) return ret; - ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, + ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, 0, table_context->overdrive_table, true); if (ret) { pr_err("Failed to import over drive table!\n"); @@ -2767,7 +2770,7 @@ static int vega20_odn_edit_dpm_table(struct smu_context *smu, break; case PP_OD_RESTORE_DEFAULT_TABLE: - ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, table_context->overdrive_table, false); + ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, 0, table_context->overdrive_table, false); if (ret) { pr_err("Failed to export over drive table!\n"); return ret; @@ -2776,7 +2779,7 @@ static int vega20_odn_edit_dpm_table(struct smu_context *smu, break; case PP_OD_COMMIT_DPM_TABLE: - ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, table_context->overdrive_table, true); + ret = smu_update_table(smu, SMU_TABLE_OVERDRIVE, 0, table_context->overdrive_table, true); if (ret) { pr_err("Failed to import over drive table!\n"); return ret; @@ -3016,15 +3019,17 @@ static int vega20_get_fan_speed_percent(struct smu_context *smu, uint32_t *speed) { int ret = 0; - uint32_t percent = 0; - uint32_t current_rpm; + uint32_t current_rpm = 0, percent = 0; PPTable_t *pptable = smu->smu_table.driver_pptable; ret = smu_get_current_rpm(smu, ¤t_rpm); + if (ret) + return ret; + percent = current_rpm * 100 / pptable->FanMaximumRpm; *speed = percent > 100 ? 100 : percent; - return ret; + return 0; } static int vega20_get_gpu_power(struct smu_context *smu, uint32_t *value) |