On 13/11/17 10:54 AM, Christian König wrote:
> Needed for the next patch.
>
> v2: actually predicate all unlocks
>
> Signed-off-by: Christian König <christian.koenig at amd.com>
[...]
> @@ -531,24 +533,28 @@ static int ttm_bo_cleanup_refs_and_unlock(struct ttm_buffer_object *bo,
> return -EBUSY;
>
> spin_lock(&glob->lru_lock);
> - ret = reservation_object_trylock(bo->resv) ? 0 : -EBUSY;
> -
> - /*
> - * We raced, and lost, someone else holds the reservation now,
> - * and is probably busy in ttm_bo_cleanup_memtype_use.
> - *
> - * Even if it's not the case, because we finished waiting any
> - * delayed destruction would succeed, so just return success
> - * here.
> - */
> - if (ret) {
> - spin_unlock(&glob->lru_lock);
> - return 0;
> + if (unlock_resv) {
> + ret = reservation_object_trylock(bo->resv) ? 0 : -EBUSY;
> + /*
> + * We raced, and lost, someone else holds the reservation now,
> + * and is probably busy in ttm_bo_cleanup_memtype_use.
> + *
> + * Even if it's not the case, because we finished waiting any
> + * delayed destruction would succeed, so just return success
> + * here.
> + */
> + if (ret) {
> + spin_unlock(&glob->lru_lock);
> + return 0;
> + }
> + } else {
> + ret = 0;
> }
> }
This could be simplified slightly:
if (unlock_resv && !reservation_object_trylock(bo->resv)) {
/*
* We raced, and lost, someone else holds the reservation now,
* and is probably busy in ttm_bo_cleanup_memtype_use.
*
* Even if it's not the case, because we finished waiting any
* delayed destruction would succeed, so just return success
* here.
*/
spin_unlock(&glob->lru_lock);
return 0;
}
ret = 0;
}
Either way, this patch is
Reviewed-and-Tested-by: Michel Dänzer <michel.daenzer at amd.com>
--
Earthling Michel Dänzer | http://www.amd.com
Libre software enthusiast | Mesa and X developer
