drm/amdgpu: fix shadow BO restoring

Don't grab the reservation lock any more and simplify the handling quite
a bit.

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

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

@@ -2950,54 +2950,6 @@ static int amdgpu_device_ip_post_soft_reset(struct amdgpu_device *adev)
 	return 0;
 }
 
-/**
- * amdgpu_device_recover_vram_from_shadow - restore shadowed VRAM buffers
- *
- * @adev: amdgpu_device pointer
- * @ring: amdgpu_ring for the engine handling the buffer operations
- * @bo: amdgpu_bo buffer whose shadow is being restored
- * @fence: dma_fence associated with the operation
- *
- * Restores the VRAM buffer contents from the shadow in GTT.  Used to
- * restore things like GPUVM page tables after a GPU reset where
- * the contents of VRAM might be lost.
- * Returns 0 on success, negative error code on failure.
- */
-static int amdgpu_device_recover_vram_from_shadow(struct amdgpu_device *adev,
-						  struct amdgpu_ring *ring,
-						  struct amdgpu_bo *bo,
-						  struct dma_fence **fence)
-{
-	uint32_t domain;
-	int r;
-
-	if (!bo->shadow)
-		return 0;
-
-	r = amdgpu_bo_reserve(bo, true);
-	if (r)
-		return r;
-	domain = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type);
-	/* if bo has been evicted, then no need to recover */
-	if (domain == AMDGPU_GEM_DOMAIN_VRAM) {
-		r = amdgpu_bo_validate(bo->shadow);
-		if (r) {
-			DRM_ERROR("bo validate failed!\n");
-			goto err;
-		}
-
-		r = amdgpu_bo_restore_from_shadow(adev, ring, bo,
-						 NULL, fence, true);
-		if (r) {
-			DRM_ERROR("recover page table failed!\n");
-			goto err;
-		}
-	}
-err:
-	amdgpu_bo_unreserve(bo);
-	return r;
-}
-
 /**
  * amdgpu_device_recover_vram - Recover some VRAM contents
  *
@@ -3006,16 +2958,15 @@ static int amdgpu_device_recover_vram_from_shadow(struct amdgpu_device *adev,
  * Restores the contents of VRAM buffers from the shadows in GTT.  Used to
  * restore things like GPUVM page tables after a GPU reset where
  * the contents of VRAM might be lost.
- * Returns 0 on success, 1 on failure.
+ *
+ * Returns:
+ * 0 on success, negative error code on failure.
  */
 static int amdgpu_device_recover_vram(struct amdgpu_device *adev)
 {
-	struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring;
-	struct amdgpu_bo *bo, *tmp;
 	struct dma_fence *fence = NULL, *next = NULL;
-	long r = 1;
+	struct amdgpu_bo *shadow;
-	int i = 0;
+	long r = 1, tmo;
-	long tmo;
 
 	if (amdgpu_sriov_runtime(adev))
 		tmo = msecs_to_jiffies(8000);
@@ -3024,44 +2975,40 @@ static int amdgpu_device_recover_vram(struct amdgpu_device *adev)
 
 	DRM_INFO("recover vram bo from shadow start\n");
 	mutex_lock(&adev->shadow_list_lock);
-	list_for_each_entry_safe(bo, tmp, &adev->shadow_list, shadow_list) {
+	list_for_each_entry(shadow, &adev->shadow_list, shadow_list) {
-		next = NULL;
+
-		amdgpu_device_recover_vram_from_shadow(adev, ring, bo, &next);
+		/* No need to recover an evicted BO */
+		if (shadow->tbo.mem.mem_type != TTM_PL_TT ||
+		    shadow->parent->tbo.mem.mem_type != TTM_PL_VRAM)
+			continue;
+
+		r = amdgpu_bo_restore_shadow(shadow, &next);
+		if (r)
+			break;
+
 		if (fence) {
 			r = dma_fence_wait_timeout(fence, false, tmo);
-			if (r == 0)
-				pr_err("wait fence %p[%d] timeout\n", fence, i);
-			else if (r < 0)
-				pr_err("wait fence %p[%d] interrupted\n", fence, i);
-			if (r < 1) {
 			dma_fence_put(fence);
 			fence = next;
+			if (r <= 0)
 				break;
-			}
+		} else {
-			i++;
-		}
-
-		dma_fence_put(fence);
 			fence = next;
 		}
+	}
 	mutex_unlock(&adev->shadow_list_lock);
 
-	if (fence) {
+	if (fence)
-		r = dma_fence_wait_timeout(fence, false, tmo);
+		tmo = dma_fence_wait_timeout(fence, false, tmo);
-		if (r == 0)
-			pr_err("wait fence %p[%d] timeout\n", fence, i);
-		else if (r < 0)
-			pr_err("wait fence %p[%d] interrupted\n", fence, i);
-
-	}
 	dma_fence_put(fence);
 
-	if (r > 0)
+	if (r <= 0 || tmo <= 0) {
-		DRM_INFO("recover vram bo from shadow done\n");
-	else
 		DRM_ERROR("recover vram bo from shadow failed\n");
+		return -EIO;
+	}
 
-	return (r > 0) ? 0 : 1;
+	DRM_INFO("recover vram bo from shadow done\n");
+	return 0;
 }
 
 /**

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

@@ -553,7 +553,7 @@ static int amdgpu_bo_create_shadow(struct amdgpu_device *adev,
 	if (!r) {
 		bo->shadow->parent = amdgpu_bo_ref(bo);
 		mutex_lock(&adev->shadow_list_lock);
-		list_add_tail(&bo->shadow_list, &adev->shadow_list);
+		list_add_tail(&bo->shadow->shadow_list, &adev->shadow_list);
 		mutex_unlock(&adev->shadow_list_lock);
 	}
 
@@ -685,13 +685,10 @@ int amdgpu_bo_validate(struct amdgpu_bo *bo)
 }
 
 /**
- * amdgpu_bo_restore_from_shadow - restore an &amdgpu_bo buffer object
+ * amdgpu_bo_restore_shadow - restore an &amdgpu_bo shadow
- * @adev: amdgpu device object
+ *
- * @ring: amdgpu_ring for the engine handling the buffer operations
+ * @shadow: &amdgpu_bo shadow to be restored
- * @bo: &amdgpu_bo buffer to be restored
- * @resv: reservation object with embedded fence
  * @fence: dma_fence associated with the operation
- * @direct: whether to submit the job directly
  *
  * Copies a buffer object's shadow content back to the object.
  * This is used for recovering a buffer from its shadow in case of a gpu
@@ -700,36 +697,19 @@ int amdgpu_bo_validate(struct amdgpu_bo *bo)
  * Returns:
  * 0 for success or a negative error code on failure.
  */
-int amdgpu_bo_restore_from_shadow(struct amdgpu_device *adev,
+int amdgpu_bo_restore_shadow(struct amdgpu_bo *shadow, struct dma_fence **fence)
-				  struct amdgpu_ring *ring,
-				  struct amdgpu_bo *bo,
-				  struct reservation_object *resv,
-				  struct dma_fence **fence,
-				  bool direct)
 
 {
-	struct amdgpu_bo *shadow = bo->shadow;
+	struct amdgpu_device *adev = amdgpu_ttm_adev(shadow->tbo.bdev);
-	uint64_t bo_addr, shadow_addr;
+	struct amdgpu_ring *ring = adev->mman.buffer_funcs_ring;
-	int r;
+	uint64_t shadow_addr, parent_addr;
 
-	if (!shadow)
+	shadow_addr = amdgpu_bo_gpu_offset(shadow);
-		return -EINVAL;
+	parent_addr = amdgpu_bo_gpu_offset(shadow->parent);
 
-	bo_addr = amdgpu_bo_gpu_offset(bo);
+	return amdgpu_copy_buffer(ring, shadow_addr, parent_addr,
-	shadow_addr = amdgpu_bo_gpu_offset(bo->shadow);
+				  amdgpu_bo_size(shadow), NULL, fence,
-
+				  true, false);
-	r = reservation_object_reserve_shared(bo->tbo.resv);
-	if (r)
-		goto err;
-
-	r = amdgpu_copy_buffer(ring, shadow_addr, bo_addr,
-			       amdgpu_bo_size(bo), resv, fence,
-			       direct, false);
-	if (!r)
-		amdgpu_bo_fence(bo, *fence, true);
-
-err:
-	return r;
 }
 
 /**

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

@@ -273,12 +273,8 @@ int amdgpu_bo_backup_to_shadow(struct amdgpu_device *adev,
 			       struct reservation_object *resv,
 			       struct dma_fence **fence, bool direct);
 int amdgpu_bo_validate(struct amdgpu_bo *bo);
-int amdgpu_bo_restore_from_shadow(struct amdgpu_device *adev,
+int amdgpu_bo_restore_shadow(struct amdgpu_bo *shadow,
-				  struct amdgpu_ring *ring,
+			     struct dma_fence **fence);
-				  struct amdgpu_bo *bo,
-				  struct reservation_object *resv,
-				  struct dma_fence **fence,
-				  bool direct);
 uint32_t amdgpu_bo_get_preferred_pin_domain(struct amdgpu_device *adev,
 					    uint32_t domain);