sysfs: Use kill_anon_super
authorEric W. Biederman <ebiederm@xmission.com>
Mon, 20 Aug 2007 12:36:29 +0000 (21:36 +0900)
committerGreg Kroah-Hartman <gregkh@suse.de>
Fri, 12 Oct 2007 21:51:07 +0000 (14:51 -0700)
Since sysfs no longer stores fs directory information in the dcache
on a permanent basis kill_litter_super it is inappropriate and actively
wrong.  It will decrement the count on all dentries left in the
dcache before trying to free them.

At the moment this is not biting us only because we never unmount sysfs.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Tejun Heo <htejun@gmail.com>
Cc: Cornelia Huck <cornelia.huck@de.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
fs/sysfs/mount.c

index 92f407fb1263de1dd0c4d867e6e6cc82b1b70c48..ac7625631fcfc0f5ae369a998b49804d1c961fe6 100644 (file)
@@ -71,7 +71,7 @@ static int sysfs_get_sb(struct file_system_type *fs_type,
 static struct file_system_type sysfs_fs_type = {
        .name           = "sysfs",
        .get_sb         = sysfs_get_sb,
-       .kill_sb        = kill_litter_super,
+       .kill_sb        = kill_anon_super,
 };
 
 int __init sysfs_init(void)