If an attempt is made to mount a bad ufs file system, the following backtrace
may be seen.
Call Trace:
[<ffffffff81443d50>] mutex_lock+0x18/0x70
[<ffffffff8127d6cc>] ufs_sync_fs+0x2c/0x278
[<ffffffff8120d814>] __sync_filesystem+0x74/0xb8
[<ffffffff8120d934>] sync_filesystem+0x4c/0x80
[<ffffffff811e23dc>] generic_shutdown_super+0x34/0x110
[<ffffffff811e24dc>] kill_block_super+0x24/0x80
[<ffffffff811e2644>] deactivate_locked_super+0x6c/0x98
[<ffffffff811e32e0>] mount_bdev+0x218/0x220
[<ffffffff811e295c>] mount_fs+0xe4/0x220
[<ffffffff81201064>] vfs_kern_mount+0x64/0x110
[<ffffffff812014fc>] do_kern_mount+0x4c/0x138
[<ffffffff81202d58>] do_mount+0x440/0x898
[<ffffffff8122c870>] compat_sys_mount+0xd8/0x248
The backtrace occurs if ufs_read_cylinder_structures() reports an error.
The calling function ufs_fill_super() tries to clean up, but does not return
the error to the calling code. This ultimately causes the backtrace.
Fix problem by returning an error from ufs_fill_super() if
ufs_read_cylinder_structures() fails.
Signed-off-by: Guenter Roeck <guenter.roeck@xxxxxxxxxxxx>
---
Resend, this time copying the linux-fsdevel list.
If I should send it to some other list, please let me know.
fs/ufs/super.c | 6 +++++-
1 files changed, 5 insertions(+), 1 deletions(-)
diff --git a/fs/ufs/super.c b/fs/ufs/super.c
index 3915ade..c398221 100644
--- a/fs/ufs/super.c
+++ b/fs/ufs/super.c
@@ -1174,8 +1174,12 @@ magic_found:
* Read cylinder group structures
*/
if (!(sb->s_flags & MS_RDONLY))
- if (!ufs_read_cylinder_structures(sb))
+ if (!ufs_read_cylinder_structures(sb)) {
+ dput(sb->s_root);
+ sb->s_root = NULL;
+ ret = -ENODEV;
goto failed;
+ }
UFSD("EXIT\n");
return 0;
--
1.7.3.1
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
