]> git.proxmox.com Git - mirror_ubuntu-kernels.git/commitdiff
drm/vmwgfx: Detach backing store from its resources when it is evicted
authorThomas Hellstrom <thellstrom@vmware.com>
Wed, 21 Nov 2012 10:29:13 +0000 (11:29 +0100)
committerThomas Hellstrom <thellstrom@vmware.com>
Fri, 17 Jan 2014 06:52:24 +0000 (07:52 +0100)
When the backing store buffer is evicted, Issue a readback from the
resources and notify the resources that they are no longer bound to
a valid backing store.

Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Zack Rusin <zackr@vmware.com>
drivers/gpu/drm/vmwgfx/vmwgfx_resource.c

index 1a62eedb0fea8bb26d96346cbd181c2a50d2c56b..6cd1560c15470bdc1f7182a3f64a1169eb7ed85c 100644 (file)
@@ -1293,11 +1293,54 @@ void vmw_fence_single_bo(struct ttm_buffer_object *bo,
  * @mem:            The truct ttm_mem_reg indicating to what memory
  *                  region the move is taking place.
  *
- * For now does nothing.
+ * Evicts the Guest Backed hardware resource if the backup
+ * buffer is being moved out of MOB memory.
+ * Note that this function should not race with the resource
+ * validation code as long as it accesses only members of struct
+ * resource that remain static while bo::res is !NULL and
+ * while we have @bo reserved. struct resource::backup is *not* a
+ * static member. The resource validation code will take care
+ * to set @bo::res to NULL, while having @bo reserved when the
+ * buffer is no longer bound to the resource, so @bo:res can be
+ * used to determine whether there is a need to unbind and whether
+ * it is safe to unbind.
  */
 void vmw_resource_move_notify(struct ttm_buffer_object *bo,
                              struct ttm_mem_reg *mem)
 {
+       struct vmw_dma_buffer *dma_buf;
+
+       if (mem == NULL)
+               return;
+
+       if (bo->destroy != vmw_dmabuf_bo_free &&
+           bo->destroy != vmw_user_dmabuf_destroy)
+               return;
+
+       dma_buf = container_of(bo, struct vmw_dma_buffer, base);
+
+       if (mem->mem_type != VMW_PL_MOB) {
+               struct vmw_resource *res, *n;
+               struct ttm_bo_device *bdev = bo->bdev;
+               struct ttm_validate_buffer val_buf;
+
+               val_buf.bo = bo;
+
+               list_for_each_entry_safe(res, n, &dma_buf->res_list, mob_head) {
+
+                       if (unlikely(res->func->unbind == NULL))
+                               continue;
+
+                       (void) res->func->unbind(res, true, &val_buf);
+                       res->backup_dirty = true;
+                       res->res_dirty = false;
+                       list_del_init(&res->mob_head);
+               }
+
+               spin_lock(&bdev->fence_lock);
+               (void) ttm_bo_wait(bo, false, false, false);
+               spin_unlock(&bdev->fence_lock);
+       }
 }
 
 /**