Btrfs: return as soon as possible when edquot happens
authorWang Shilong <wangsl-fnst@cn.fujitsu.com>
Wed, 6 Mar 2013 11:51:47 +0000 (11:51 +0000)
committerChris Mason <chris.mason@fusionio.com>
Thu, 14 Mar 2013 18:57:29 +0000 (14:57 -0400)
If one of qgroup fails to reserve firstly, we should return immediately,
it is unnecessary to continue check.

Signed-off-by: Wang Shilong <wangsl-fnst@cn.fujitsu.com>
Signed-off-by: Josef Bacik <jbacik@fusionio.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>
fs/btrfs/qgroup.c

index aee4b1cc3d988ccc0dbde615000d07bcf5d35051..5471e47d6559eafeb36e55035cb427710c5ee4ba 100644 (file)
@@ -1525,21 +1525,23 @@ int btrfs_qgroup_reserve(struct btrfs_root *root, u64 num_bytes)
 
                if ((qg->lim_flags & BTRFS_QGROUP_LIMIT_MAX_RFER) &&
                    qg->reserved + qg->rfer + num_bytes >
-                   qg->max_rfer)
+                   qg->max_rfer) {
                        ret = -EDQUOT;
+                       goto out;
+               }
 
                if ((qg->lim_flags & BTRFS_QGROUP_LIMIT_MAX_EXCL) &&
                    qg->reserved + qg->excl + num_bytes >
-                   qg->max_excl)
+                   qg->max_excl) {
                        ret = -EDQUOT;
+                       goto out;
+               }
 
                list_for_each_entry(glist, &qg->groups, next_group) {
                        ulist_add(ulist, glist->group->qgroupid,
                                  (uintptr_t)glist->group, GFP_ATOMIC);
                }
        }
-       if (ret)
-               goto out;
 
        /*
         * no limits exceeded, now record the reservation into all qgroups