summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/amd/amdgpu
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/gpu/drm/amd/amdgpu')
-rw-r--r--drivers/gpu/drm/amd/amdgpu/amdgpu.h1
-rw-r--r--drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c5
-rw-r--r--drivers/gpu/drm/amd/amdgpu/amdgpu_device.c29
-rw-r--r--drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c5
-rw-r--r--drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c3
-rw-r--r--drivers/gpu/drm/amd/amdgpu/cik.c2
-rw-r--r--drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c24
-rw-r--r--drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c16
-rw-r--r--drivers/gpu/drm/amd/amdgpu/navi10_sdma_pkt_open.h16
-rw-r--r--drivers/gpu/drm/amd/amdgpu/nv.c4
-rw-r--r--drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c14
-rw-r--r--drivers/gpu/drm/amd/amdgpu/soc15.c4
-rw-r--r--drivers/gpu/drm/amd/amdgpu/vi.c2
13 files changed, 84 insertions, 41 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 2992a49ad4a5..8ac1581a6b53 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -945,6 +945,7 @@ struct amdgpu_device {
/* s3/s4 mask */
bool in_suspend;
+ bool in_hibernate;
/* record last mm index being written through WREG32*/
unsigned long last_mm_index;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
index 9dff792c9290..6a5b91d23fd9 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
@@ -1343,7 +1343,7 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu(
}
/* Free the BO*/
- amdgpu_bo_unref(&mem->bo);
+ drm_gem_object_put_unlocked(&mem->bo->tbo.base);
mutex_destroy(&mem->lock);
kfree(mem);
@@ -1688,7 +1688,8 @@ int amdgpu_amdkfd_gpuvm_import_dmabuf(struct kgd_dev *kgd,
| KFD_IOC_ALLOC_MEM_FLAGS_WRITABLE
| KFD_IOC_ALLOC_MEM_FLAGS_EXECUTABLE;
- (*mem)->bo = amdgpu_bo_ref(bo);
+ drm_gem_object_get(&bo->tbo.base);
+ (*mem)->bo = bo;
(*mem)->va = va;
(*mem)->domain = (bo->preferred_domains & AMDGPU_GEM_DOMAIN_VRAM) ?
AMDGPU_GEM_DOMAIN_VRAM : AMDGPU_GEM_DOMAIN_GTT;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 559dc24ef436..affde2de2a0d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -2008,8 +2008,24 @@ static void amdgpu_device_fill_reset_magic(struct amdgpu_device *adev)
*/
static bool amdgpu_device_check_vram_lost(struct amdgpu_device *adev)
{
- return !!memcmp(adev->gart.ptr, adev->reset_magic,
- AMDGPU_RESET_MAGIC_NUM);
+ if (memcmp(adev->gart.ptr, adev->reset_magic,
+ AMDGPU_RESET_MAGIC_NUM))
+ return true;
+
+ if (!adev->in_gpu_reset)
+ return false;
+
+ /*
+ * For all ASICs with baco/mode1 reset, the VRAM is
+ * always assumed to be lost.
+ */
+ switch (amdgpu_asic_reset_method(adev)) {
+ case AMD_RESET_METHOD_BACO:
+ case AMD_RESET_METHOD_MODE1:
+ return true;
+ default:
+ return false;
+ }
}
/**
@@ -2340,6 +2356,8 @@ static int amdgpu_device_ip_suspend_phase1(struct amdgpu_device *adev)
{
int i, r;
+ amdgpu_device_set_pg_state(adev, AMD_PG_STATE_UNGATE);
+ amdgpu_device_set_cg_state(adev, AMD_CG_STATE_UNGATE);
for (i = adev->num_ip_blocks - 1; i >= 0; i--) {
if (!adev->ip_blocks[i].status.valid)
@@ -3354,15 +3372,12 @@ int amdgpu_device_suspend(struct drm_device *dev, bool fbcon)
}
}
- amdgpu_device_set_pg_state(adev, AMD_PG_STATE_UNGATE);
- amdgpu_device_set_cg_state(adev, AMD_CG_STATE_UNGATE);
-
- amdgpu_amdkfd_suspend(adev, !fbcon);
-
amdgpu_ras_suspend(adev);
r = amdgpu_device_ip_suspend_phase1(adev);
+ amdgpu_amdkfd_suspend(adev, !fbcon);
+
/* evict vram memory */
amdgpu_bo_evict_vram(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
index 8ea86ffdea0d..a735d79a717b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c
@@ -85,9 +85,10 @@
* - 3.34.0 - Non-DC can flip correctly between buffers with different pitches
* - 3.35.0 - Add drm_amdgpu_info_device::tcc_disabled_mask
* - 3.36.0 - Allow reading more status registers on si/cik
+ * - 3.37.0 - L2 is invalidated before SDMA IBs, needed for correctness
*/
#define KMS_DRIVER_MAJOR 3
-#define KMS_DRIVER_MINOR 36
+#define KMS_DRIVER_MINOR 37
#define KMS_DRIVER_PATCHLEVEL 0
int amdgpu_vram_limit = 0;
@@ -1180,7 +1181,9 @@ static int amdgpu_pmops_freeze(struct device *dev)
struct amdgpu_device *adev = drm_dev->dev_private;
int r;
+ adev->in_hibernate = true;
r = amdgpu_device_suspend(drm_dev, true);
+ adev->in_hibernate = false;
if (r)
return r;
return amdgpu_asic_reset(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
index 9ae7b61f696a..25ddb482466a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fb.c
@@ -133,8 +133,7 @@ static int amdgpufb_create_pinned_object(struct amdgpu_fbdev *rfbdev,
u32 cpp;
u64 flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS |
- AMDGPU_GEM_CREATE_VRAM_CLEARED |
- AMDGPU_GEM_CREATE_CPU_GTT_USWC;
+ AMDGPU_GEM_CREATE_VRAM_CLEARED;
info = drm_get_format_info(adev->ddev, mode_cmd);
cpp = info->cpp[0];
diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c
index 006f21ef7ddf..62635e58e45e 100644
--- a/drivers/gpu/drm/amd/amdgpu/cik.c
+++ b/drivers/gpu/drm/amd/amdgpu/cik.c
@@ -1358,8 +1358,6 @@ static int cik_asic_reset(struct amdgpu_device *adev)
int r;
if (cik_asic_reset_method(adev) == AMD_RESET_METHOD_BACO) {
- if (!adev->in_suspend)
- amdgpu_inc_vram_lost(adev);
r = amdgpu_dpm_baco_reset(adev);
} else {
r = cik_asic_pci_config_reset(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
index d78059fd2c72..0e0daf0021b6 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
@@ -279,7 +279,7 @@ static const struct soc15_reg_golden golden_settings_gc_10_1_2_nv12[] =
#define DEFAULT_SH_MEM_CONFIG \
((SH_MEM_ADDRESS_MODE_64 << SH_MEM_CONFIG__ADDRESS_MODE__SHIFT) | \
- (SH_MEM_ALIGNMENT_MODE_DWORD << SH_MEM_CONFIG__ALIGNMENT_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))
@@ -4273,7 +4273,7 @@ static int gfx_v10_0_update_gfx_clock_gating(struct amdgpu_device *adev,
/* === CGCG /CGLS for GFX 3D Only === */
gfx_v10_0_update_3d_clock_gating(adev, enable);
/* === MGCG + MGLS === */
- /* gfx_v10_0_update_medium_grain_clock_gating(adev, enable); */
+ gfx_v10_0_update_medium_grain_clock_gating(adev, enable);
}
if (adev->cg_flags &
@@ -4353,11 +4353,7 @@ static int gfx_v10_0_set_powergating_state(void *handle,
switch (adev->asic_type) {
case CHIP_NAVI10:
case CHIP_NAVI14:
- if (!enable) {
- amdgpu_gfx_off_ctrl(adev, false);
- cancel_delayed_work_sync(&adev->gfx.gfx_off_delay_work);
- } else
- amdgpu_gfx_off_ctrl(adev, true);
+ amdgpu_gfx_off_ctrl(adev, enable);
break;
default:
break;
@@ -4918,6 +4914,19 @@ static void gfx_v10_0_ring_emit_reg_write_reg_wait(struct amdgpu_ring *ring,
ref, mask);
}
+static void gfx_v10_0_ring_soft_recovery(struct amdgpu_ring *ring,
+ unsigned vmid)
+{
+ struct amdgpu_device *adev = ring->adev;
+ uint32_t value = 0;
+
+ value = REG_SET_FIELD(value, SQ_CMD, CMD, 0x03);
+ value = REG_SET_FIELD(value, SQ_CMD, MODE, 0x01);
+ value = REG_SET_FIELD(value, SQ_CMD, CHECK_VMID, 1);
+ value = REG_SET_FIELD(value, SQ_CMD, VM_ID, vmid);
+ WREG32_SOC15(GC, 0, mmSQ_CMD, value);
+}
+
static void
gfx_v10_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev,
uint32_t me, uint32_t pipe,
@@ -5309,6 +5318,7 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_gfx = {
.emit_wreg = gfx_v10_0_ring_emit_wreg,
.emit_reg_wait = gfx_v10_0_ring_emit_reg_wait,
.emit_reg_write_reg_wait = gfx_v10_0_ring_emit_reg_write_reg_wait,
+ .soft_recovery = gfx_v10_0_ring_soft_recovery,
};
static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_compute = {
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
index e6b113ed2f40..d2d9dce68c2f 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c
@@ -1234,6 +1234,10 @@ struct amdgpu_gfxoff_quirk {
static const struct amdgpu_gfxoff_quirk amdgpu_gfxoff_quirk_list[] = {
/* https://bugzilla.kernel.org/show_bug.cgi?id=204689 */
{ 0x1002, 0x15dd, 0x1002, 0x15dd, 0xc8 },
+ /* https://bugzilla.kernel.org/show_bug.cgi?id=207171 */
+ { 0x1002, 0x15dd, 0x103c, 0x83e7, 0xd3 },
+ /* GFXOFF is unstable on C6 parts with a VBIOS 113-RAVEN-114 */
+ { 0x1002, 0x15dd, 0x1002, 0x15dd, 0xc6 },
{ 0, 0, 0, 0, 0 },
};
@@ -5023,10 +5027,9 @@ static int gfx_v9_0_set_powergating_state(void *handle,
switch (adev->asic_type) {
case CHIP_RAVEN:
case CHIP_RENOIR:
- if (!enable) {
+ if (!enable)
amdgpu_gfx_off_ctrl(adev, false);
- cancel_delayed_work_sync(&adev->gfx.gfx_off_delay_work);
- }
+
if (adev->pg_flags & AMD_PG_SUPPORT_RLC_SMU_HS) {
gfx_v9_0_enable_sck_slow_down_on_power_up(adev, true);
gfx_v9_0_enable_sck_slow_down_on_power_down(adev, true);
@@ -5050,12 +5053,7 @@ static int gfx_v9_0_set_powergating_state(void *handle,
amdgpu_gfx_off_ctrl(adev, true);
break;
case CHIP_VEGA12:
- if (!enable) {
- amdgpu_gfx_off_ctrl(adev, false);
- cancel_delayed_work_sync(&adev->gfx.gfx_off_delay_work);
- } else {
- amdgpu_gfx_off_ctrl(adev, true);
- }
+ amdgpu_gfx_off_ctrl(adev, enable);
break;
default:
break;
diff --git a/drivers/gpu/drm/amd/amdgpu/navi10_sdma_pkt_open.h b/drivers/gpu/drm/amd/amdgpu/navi10_sdma_pkt_open.h
index 074a9a09c0a7..a5b60c9a2418 100644
--- a/drivers/gpu/drm/amd/amdgpu/navi10_sdma_pkt_open.h
+++ b/drivers/gpu/drm/amd/amdgpu/navi10_sdma_pkt_open.h
@@ -73,6 +73,22 @@
#define SDMA_OP_AQL_COPY 0
#define SDMA_OP_AQL_BARRIER_OR 0
+#define SDMA_GCR_RANGE_IS_PA (1 << 18)
+#define SDMA_GCR_SEQ(x) (((x) & 0x3) << 16)
+#define SDMA_GCR_GL2_WB (1 << 15)
+#define SDMA_GCR_GL2_INV (1 << 14)
+#define SDMA_GCR_GL2_DISCARD (1 << 13)
+#define SDMA_GCR_GL2_RANGE(x) (((x) & 0x3) << 11)
+#define SDMA_GCR_GL2_US (1 << 10)
+#define SDMA_GCR_GL1_INV (1 << 9)
+#define SDMA_GCR_GLV_INV (1 << 8)
+#define SDMA_GCR_GLK_INV (1 << 7)
+#define SDMA_GCR_GLK_WB (1 << 6)
+#define SDMA_GCR_GLM_INV (1 << 5)
+#define SDMA_GCR_GLM_WB (1 << 4)
+#define SDMA_GCR_GL1_RANGE(x) (((x) & 0x3) << 2)
+#define SDMA_GCR_GLI_INV(x) (((x) & 0x3) << 0)
+
/*define for op field*/
#define SDMA_PKT_HEADER_op_offset 0
#define SDMA_PKT_HEADER_op_mask 0x000000FF
diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c
index 033cbbca2072..52318b03c424 100644
--- a/drivers/gpu/drm/amd/amdgpu/nv.c
+++ b/drivers/gpu/drm/amd/amdgpu/nv.c
@@ -351,8 +351,6 @@ static int nv_asic_reset(struct amdgpu_device *adev)
struct smu_context *smu = &adev->smu;
if (nv_asic_reset_method(adev) == AMD_RESET_METHOD_BACO) {
- if (!adev->in_suspend)
- amdgpu_inc_vram_lost(adev);
ret = smu_baco_enter(smu);
if (ret)
return ret;
@@ -360,8 +358,6 @@ static int nv_asic_reset(struct amdgpu_device *adev)
if (ret)
return ret;
} else {
- if (!adev->in_suspend)
- amdgpu_inc_vram_lost(adev);
ret = nv_asic_mode1_reset(adev);
}
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c
index ebfd2cdf4e65..d2840c2f6286 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c
@@ -382,6 +382,18 @@ static void sdma_v5_0_ring_emit_ib(struct amdgpu_ring *ring,
unsigned vmid = AMDGPU_JOB_GET_VMID(job);
uint64_t csa_mc_addr = amdgpu_sdma_get_csa_mc_addr(ring, vmid);
+ /* Invalidate L2, because if we don't do it, we might get stale cache
+ * lines from previous IBs.
+ */
+ amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_GCR_REQ));
+ amdgpu_ring_write(ring, 0);
+ amdgpu_ring_write(ring, (SDMA_GCR_GL2_INV |
+ SDMA_GCR_GL2_WB |
+ SDMA_GCR_GLM_INV |
+ SDMA_GCR_GLM_WB) << 16);
+ amdgpu_ring_write(ring, 0xffffff80);
+ amdgpu_ring_write(ring, 0xffff);
+
/* An IB packet must end on a 8 DW boundary--the next dword
* must be on a 8-dword boundary. Our IB packet below is 6
* dwords long, thus add x number of NOPs, such that, in
@@ -1595,7 +1607,7 @@ static const struct amdgpu_ring_funcs sdma_v5_0_ring_funcs = {
SOC15_FLUSH_GPU_TLB_NUM_WREG * 3 +
SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 6 * 2 +
10 + 10 + 10, /* sdma_v5_0_ring_emit_fence x3 for user fence, vm fence */
- .emit_ib_size = 7 + 6, /* sdma_v5_0_ring_emit_ib */
+ .emit_ib_size = 5 + 7 + 6, /* sdma_v5_0_ring_emit_ib */
.emit_ib = sdma_v5_0_ring_emit_ib,
.emit_fence = sdma_v5_0_ring_emit_fence,
.emit_pipeline_sync = sdma_v5_0_ring_emit_pipeline_sync,
diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index a40499d51c93..d42a8d8a0dea 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -569,14 +569,10 @@ static int soc15_asic_reset(struct amdgpu_device *adev)
switch (soc15_asic_reset_method(adev)) {
case AMD_RESET_METHOD_BACO:
- if (!adev->in_suspend)
- amdgpu_inc_vram_lost(adev);
return soc15_asic_baco_reset(adev);
case AMD_RESET_METHOD_MODE2:
return amdgpu_dpm_mode2_reset(adev);
default:
- if (!adev->in_suspend)
- amdgpu_inc_vram_lost(adev);
return soc15_asic_mode1_reset(adev);
}
}
diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
index 78b35901643b..3ce10e05d0d6 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -765,8 +765,6 @@ static int vi_asic_reset(struct amdgpu_device *adev)
int r;
if (vi_asic_reset_method(adev) == AMD_RESET_METHOD_BACO) {
- if (!adev->in_suspend)
- amdgpu_inc_vram_lost(adev);
r = amdgpu_dpm_baco_reset(adev);
} else {
r = vi_asic_pci_config_reset(adev);
generated by cgit (git 2.34.1) at 2024-06-09 17:54:15 +0000