Merge tag 'amd-drm-next-6.5-2023-06-16' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-6.5-2023-06-16:

amdgpu:
- Misc display fixes
- W=1 fixes
- Improve scheduler naming
- DCN 3.1.4 fixes
- kdoc fixes
- Enable W=1
- VCN 4.0 fix
- xgmi fixes
- TOPDOWN fix for large BAR systems
- eDP fix
- PSR fixes
- SubVP fixes
- Freesync fix
- DPIA fix
- SMU 13.0.5 fixes
- vblflash fix
- RAS fixes
- SDMA 4 fix
- BO locking fix
- BO backing store fix
- NBIO 7.9 fixes
- GC 9.4.3 fixes
- GPU reset recovery fixes
- HMM fix

amdkfd:
- Fix NULL check
- Trap fixes
- Queue count fix
- Add event age tracking

radeon:
- fbdev client fix

scheduler:
- Avoid an infinite loop

UAPI:
- Add KFD event age tracking:
  Proposed ROCT-Thunk-Interface:
  https://github.com/RadeonOpenCompute/ROCT-Thunk-Interface/commit/efdbf6cfbc026bd68ac3c35d00dacf84370eb81e
  https://github.com/RadeonOpenCompute/ROCT-Thunk-Interface/commit/1820ae0a2db85b6f584611dc0cde1a00e7c22915
  Proposed ROCR-Runtime:
  https://github.com/RadeonOpenCompute/ROCR-Runtime/compare/master...zhums:ROCR-Runtime:new_event_wait_review
  https://github.com/RadeonOpenCompute/ROCR-Runtime/commit/e1f5bdb88eb882ac798aeca2c00ea3fbb2dba459
  https://github.com/RadeonOpenCompute/ROCR-Runtime/commit/7d26afd14107b5c2a754c1a3f415d89f3aabb503

drm:
- DP MST fix

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230616163548.7706-1-alexander.deucher@amd.com

1 files changed, 11 insertions, 3 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
index 410acdd4554c..0dc9c655c4fb 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
@@ -333,7 +333,7 @@ static int amdgpu_ctx_init(struct amdgpu_ctx_mgr *mgr, int32_t priority,
 
 	ctx->reset_counter = atomic_read(&mgr->adev->gpu_reset_counter);
 	ctx->reset_counter_query = ctx->reset_counter;
-	ctx->vram_lost_counter = atomic_read(&mgr->adev->vram_lost_counter);
+	ctx->generation = amdgpu_vm_generation(mgr->adev, &fpriv->vm);
 	ctx->init_priority = priority;
 	ctx->override_priority = AMDGPU_CTX_PRIORITY_UNSET;
 
@@ -432,6 +432,7 @@ int amdgpu_ctx_get_entity(struct amdgpu_ctx *ctx, u32 hw_ip, u32 instance,
 			  u32 ring, struct drm_sched_entity **entity)
 {
 	int r;
+	struct drm_sched_entity *ctx_entity;
 
 	if (hw_ip >= AMDGPU_HW_IP_NUM) {
 		DRM_ERROR("unknown HW IP type: %d\n", hw_ip);
@@ -455,7 +456,14 @@ int amdgpu_ctx_get_entity(struct amdgpu_ctx *ctx, u32 hw_ip, u32 instance,
 			return r;
 	}
 
-	*entity = &ctx->entities[hw_ip][ring]->entity;
+	ctx_entity = &ctx->entities[hw_ip][ring]->entity;
+	r = drm_sched_entity_error(ctx_entity);
+	if (r) {
+		DRM_DEBUG("error entity %p\n", ctx_entity);
+		return r;
+	}
+
+	*entity = ctx_entity;
 	return 0;
 }
 
@@ -586,7 +594,7 @@ static int amdgpu_ctx_query2(struct amdgpu_device *adev,
 	if (ctx->reset_counter != atomic_read(&adev->gpu_reset_counter))
 		out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_RESET;
 
-	if (ctx->vram_lost_counter != atomic_read(&adev->vram_lost_counter))
+	if (ctx->generation != amdgpu_vm_generation(adev, &fpriv->vm))
 		out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_VRAMLOST;
 
 	if (atomic_read(&ctx->guilty))