drm/amdgpu: deprecate guilty handling

The guilty handling tried to establish a second way of signaling problems with
the GPU back to userspace. This caused quite a bunch of issue we had to work
around, especially lifetime issues with the drm_sched_entity.

Just drop the handling altogether and use the dma_fence based approach instead.

v2: fix reversed condition in entity check (Alex)

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

4 files changed, 24 insertions, 12 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index 32af8cce3df8..c42ae3e6fdd1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -60,11 +60,6 @@ static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p,
 	if (!p->ctx)
 		return -EINVAL;
 
-	if (atomic_read(&p->ctx->guilty)) {
-		amdgpu_ctx_put(p->ctx);
-		return -ECANCELED;
-	}
-
 	amdgpu_sync_create(&p->sync);
 	drm_exec_init(&p->exec, DRM_EXEC_INTERRUPTIBLE_WAIT |
 		      DRM_EXEC_IGNORE_DUPLICATES, 0);
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
index 7af86a32c0c5..0d7f6cd74f79 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
@@ -255,7 +255,7 @@ static int amdgpu_ctx_init_entity(struct amdgpu_ctx *ctx, u32 hw_ip,
 	}
 
 	r = drm_sched_entity_init(&entity->entity, drm_prio, scheds, num_scheds,
-				  &ctx->guilty);
+				  NULL);
 	if (r)
 		goto error_free_entity;
 
@@ -579,6 +579,27 @@ static int amdgpu_ctx_query(struct amdgpu_device *adev,
 
 #define AMDGPU_RAS_COUNTE_DELAY_MS 3000
 
+static bool amdgpu_ctx_guilty(struct amdgpu_ctx *ctx)
+{
+	int i, j, r;
+
+	for (i = 0; i < AMDGPU_HW_IP_NUM; ++i) {
+		for (j = 0; j < amdgpu_ctx_num_entities[i]; ++j) {
+			struct amdgpu_ctx_entity *ctx_entity;
+
+			ctx_entity = ctx->entities[i][j];
+			if (!ctx_entity)
+				continue;
+
+			r = drm_sched_entity_error(&ctx_entity->entity);
+			if (r == -ETIME)
+				return true;
+		}
+	}
+
+	return false;
+}
+
 static int amdgpu_ctx_query2(struct amdgpu_device *adev,
 			     struct amdgpu_fpriv *fpriv, uint32_t id,
 			     union drm_amdgpu_ctx_out *out)
@@ -607,7 +628,7 @@ static int amdgpu_ctx_query2(struct amdgpu_device *adev,
 	if (ctx->generation != amdgpu_vm_generation(adev, &fpriv->vm))
 		out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_VRAMLOST;
 
-	if (atomic_read(&ctx->guilty))
+	if (amdgpu_ctx_guilty(ctx))
 		out->state.flags |= AMDGPU_CTX_QUERY2_FLAGS_GUILTY;
 
 	if (amdgpu_in_reset(adev))
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h
index cf8d700a22fe..e444b2088d40 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h
@@ -50,7 +50,6 @@ struct amdgpu_ctx {
 	int32_t				init_priority;
 	int32_t				override_priority;
 	uint32_t			stable_pstate;
-	atomic_t			guilty;
 	bool				preamble_presented;
 	uint64_t			generation;
 	unsigned long			ras_counter_ce;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index f18e46502829..942f0251c748 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -5112,12 +5112,12 @@ link_reset_failed:
 int amdgpu_device_pre_asic_reset(struct amdgpu_device *adev,
 				 struct amdgpu_reset_context *reset_context)
 {
-	int i, r = 0;
 	struct amdgpu_job *job = NULL;
 	struct dma_fence *fence = NULL;
 	struct amdgpu_device *tmp_adev = reset_context->reset_req_dev;
 	bool need_full_reset =
 		test_bit(AMDGPU_NEED_FULL_RESET, &reset_context->flags);
+	int i, r;
 
 	if (reset_context->reset_req_dev == adev)
 		job = reset_context->job;
@@ -5143,9 +5143,6 @@ int amdgpu_device_pre_asic_reset(struct amdgpu_device *adev,
 
 	amdgpu_fence_driver_isr_toggle(adev, false);
 
-	if (job && job->vm)
-		drm_sched_increase_karma(&job->base);
-
 	r = amdgpu_reset_prepare_hwcontext(adev, reset_context);
 	/* If reset handler not implemented, continue; otherwise return */
 	if (r == -EOPNOTSUPP)