drm/ttm: remove nonsense wait in ttm_bo_cleanup_refs_and_unlock
authorChristian König <christian.koenig@amd.com>
Wed, 26 Jul 2017 12:59:10 +0000 (14:59 +0200)
committerAlex Deucher <alexander.deucher@amd.com>
Thu, 17 Aug 2017 19:45:55 +0000 (15:45 -0400)
With shared reservation objects the assumption that no fence
could have been added isn't true any more.

Additional to that the BO is about to be destroyed, so removing the
fences now has no advantage whatsoever.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drivers/gpu/drm/ttm/ttm_bo.c

index 22b57020790da749c4494becbabc981164191ded..0f5c59f2e92c8ceb466f765db53a3a1f1a6bc19f 100644 (file)
@@ -505,13 +505,6 @@ static int ttm_bo_cleanup_refs_and_unlock(struct ttm_buffer_object *bo,
                        spin_unlock(&glob->lru_lock);
                        return 0;
                }
-
-               /*
-                * remove sync_obj with ttm_bo_wait, the wait should be
-                * finished, and no new wait object should have been added.
-                */
-               ret = ttm_bo_wait(bo, false, true);
-               WARN_ON(ret);
        }
 
        if (ret || unlikely(list_empty(&bo->ddestroy))) {