f2fs: record quota during dot{,dot} recovery
authorChao Yu <yuchao0@huawei.com>
Mon, 24 Jul 2017 09:12:06 +0000 (17:12 +0800)
committerJaegeuk Kim <jaegeuk@kernel.org>
Mon, 31 Jul 2017 23:48:31 +0000 (16:48 -0700)
In ->lookup(), we will have a try to recover dot or dotdot for
corrupted directory, once disk quota is on, if it allocates new
block during dotdot recovery, we need to record disk quota info
for the allocation, so this patch fixes this issue by adding
missing dquot_initialize() in __recover_dot_dentries.

Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
fs/f2fs/namei.c

index 26b4d2b548128c14952ff449d83fd921914a7201..98e13f123d8737c8724b007ca10ee8abd1461c84 100644 (file)
@@ -266,6 +266,10 @@ static int __recover_dot_dentries(struct inode *dir, nid_t pino)
                return 0;
        }
 
+       err = dquot_initialize(dir);
+       if (err)
+               return err;
+
        f2fs_balance_fs(sbi, true);
 
        f2fs_lock_op(sbi);