drm/amdgpu: remove ctx->lock

KMD reports a warning on holding a lock from drm_syncobj_find_fence,
when running amdgpu_test case “syncobj timeline test”.

ctx->lock was designed to prevent concurrent "amdgpu_ctx_wait_prev_fence"
calls and avoid dead reservation lock from GPU reset. since no reservation
lock is held in latest GPU reset any more, ctx->lock can be simply removed
and concurrent "amdgpu_ctx_wait_prev_fence" call also can be prevented by
PD root bo reservation lock.

call stacks:
=================
//hold lock
amdgpu_cs_ioctl->amdgpu_cs_parser_init->mutex_lock(&parser->ctx->lock);
…
//report warning
amdgpu_cs_dependencies->amdgpu_cs_process_syncobj_timeline_in_dep \
->amdgpu_syncobj_lookup_and_add_to_sync -> drm_syncobj_find_fence \
-> lockdep_assert_none_held_once
…
amdgpu_cs_ioctl->amdgpu_cs_parser_fini->mutex_unlock(&parser->ctx->lock);

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

drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c

@@ -127,8 +127,6 @@ static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, union drm_amdgpu_cs
 		goto free_chunk;
 	}
 
-	mutex_lock(&p->ctx->lock);
-
 	/* skip guilty context job */
 	if (atomic_read(&p->ctx->guilty) == 1) {
 		ret = -ECANCELED;
@@ -585,6 +583,16 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
 		}
 	}
 
+	/* Move fence waiting after getting reservation lock of
+	 * PD root. Then there is no need on a ctx mutex lock.
+	 */
+	r = amdgpu_ctx_wait_prev_fence(p->ctx, p->entity);
+	if (unlikely(r != 0)) {
+		if (r != -ERESTARTSYS)
+			DRM_ERROR("amdgpu_ctx_wait_prev_fence failed.\n");
+		goto error_validate;
+	}
+
 	amdgpu_cs_get_threshold_for_moves(p->adev, &p->bytes_moved_threshold,
 					  &p->bytes_moved_vis_threshold);
 	p->bytes_moved = 0;
@@ -700,7 +708,6 @@ static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser, int error,
 	dma_fence_put(parser->fence);
 
 	if (parser->ctx) {
-		mutex_unlock(&parser->ctx->lock);
 		amdgpu_ctx_put(parser->ctx);
 	}
 	if (parser->bo_list)
@@ -944,7 +951,7 @@ static int amdgpu_cs_ib_fill(struct amdgpu_device *adev,
 	if (parser->job->uf_addr && ring->funcs->no_user_fence)
 		return -EINVAL;
 
-	return amdgpu_ctx_wait_prev_fence(parser->ctx, parser->entity);
+	return 0;
 }
 
 static int amdgpu_cs_process_fence_dep(struct amdgpu_cs_parser *p,
@@ -1363,7 +1370,6 @@ int amdgpu_cs_ioctl(struct drm_device *dev, void *data, struct drm_file *filp)
 		goto out;
 
 	r = amdgpu_cs_submit(&parser, cs);
-
 out:
 	amdgpu_cs_parser_fini(&parser, r, reserved_buffers);
 

drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c

@@ -230,7 +230,6 @@ static int amdgpu_ctx_init(struct amdgpu_device *adev,
 
 	kref_init(&ctx->refcount);
 	spin_lock_init(&ctx->ring_lock);
-	mutex_init(&ctx->lock);
 
 	ctx->reset_counter = atomic_read(&adev->gpu_reset_counter);
 	ctx->reset_counter_query = ctx->reset_counter;
@@ -352,7 +351,6 @@ static void amdgpu_ctx_fini(struct kref *ref)
 		}
 	}
 	amdgpu_ctx_set_stable_pstate(ctx, AMDGPU_CTX_STABLE_PSTATE_NONE);
-	mutex_destroy(&ctx->lock);
 	kfree(ctx);
 }
 

drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h

@@ -49,7 +49,6 @@ struct amdgpu_ctx {
 	bool				preamble_presented;
 	int32_t				init_priority;
 	int32_t				override_priority;
-	struct mutex			lock;
 	atomic_t			guilty;
 	unsigned long			ras_counter_ce;
 	unsigned long			ras_counter_ue;