[PATCH] xfs: test for shut down fs in xfs_dir_fsync()

[Dave Chinner]

On Mon, Apr 28, 2014 at 11:35:16AM -0500, Eric Sandeen wrote:
> Similar to xfs_file_fsync(), I think xfs_dir_fsync() needs
> to test for a shut down fs, lest we go down paths we'll
> never be able to complete; Boris reported that during some
> stress tests he had threads stuck in xlog_cil_force_lsn
> via xfs_dir_fsync().
> 
> [ 3663.361709] sfsuspend-par   D ffff88042f0b4540     0  3981   3947 0x00000080
> 
> [ 3663.394472] Call Trace:
> [ 3663.397199]  [<ffffffff815f1889>] schedule+0x29/0x70
> [ 3663.402743]  [<ffffffffa01feda5>] xlog_cil_force_lsn+0x185/0x1a0 [xfs]
> [ 3663.416249]  [<ffffffffa01fd3af>] _xfs_log_force_lsn+0x6f/0x2f0 [xfs]
> [ 3663.429271]  [<ffffffffa01a339d>] xfs_dir_fsync+0x7d/0xe0 [xfs]
> [ 3663.435873]  [<ffffffff811df8c5>] do_fsync+0x65/0xa0
> [ 3663.441408]  [<ffffffff811dfbc0>] SyS_fsync+0x10/0x20
> [ 3663.447043]  [<ffffffff815fc7d9>] system_call_fastpath+0x16/0x1b

Wow, I believe it's taken this long for us to notice that we can't
break out of xlog_cil_force_lsn() if we fail on xlog_write()
from a CIL push.

I'd say that xlog_cil_force_lsn() needs log shutdown checks before
it goes to sleep in xlog_wait()....

> Reported-by: Boris Ranto <branto at redhat.com>
> Signed-off-by: Eric Sandeen <sandeen at redhat.com>
> ---
> 
> NB: While I've not asked Boris to test this yet, it seems
> clear (?) that dir_fsync should behave the same as
> file_fsync() in the face of a shut-down fs.
> 
> diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
> index 4c749ab..2b94362 100644
> --- a/fs/xfs/xfs_file.c
> +++ b/fs/xfs/xfs_file.c
> @@ -146,6 +146,9 @@ xfs_dir_fsync(
>  
>  	trace_xfs_dir_fsync(ip);
>  
> +	if (XFS_FORCED_SHUTDOWN(mp))
> +		return -XFS_ERROR(EIO);
> +

That won't hurt, but it won't fix the problem.

Cheers,

Dave.
-- 
Dave Chinner
david at fromorbit.com