lkml.org 
[lkml]   [2012]   [Feb]   [5]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 5/8] xfs: Protect xfs_file_aio_write() & xfs_setattr_size() with sb_start_write - sb_end_write
On Fri, Feb 03, 2012 at 10:30:05PM -0600, Eric Sandeen wrote:
> On 1/20/12 2:34 PM, Jan Kara wrote:
> > Replace racy xfs_wait_for_freeze() check in xfs_file_aio_write() with
> > a reliable sb_start_write() - sb_end_write() locking.
>
> Here's what I'm running with now. With this and my modified
> patch6, I can pass xfstests 068 on xfs.

Just a quick question this raises - is .splice_write() protected?

Cheers,

Dave.
>
> -Eric
>
>
>
> diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
> index 7e5bc87..f1cacdc 100644
> --- a/fs/xfs/xfs_file.c
> +++ b/fs/xfs/xfs_file.c
> @@ -874,11 +874,11 @@ xfs_file_aio_write(
> if (ocount == 0)
> return 0;
>
> - xfs_wait_for_freeze(ip->i_mount, SB_FREEZE_WRITE);
> -
> if (XFS_FORCED_SHUTDOWN(ip->i_mount))
> return -EIO;
>
> + sb_start_write(inode->i_sb, SB_FREEZE_WRITE);
> +

This check really should go before the shutdown check - if the
filesystem shuts down while we are freezing or attempting to freeze,
we want to abort the write after we are woken....

Cheers,

Dave.
--
Dave Chinner
david@fromorbit.com


\
 
 \ /
  Last update: 2012-02-06 00:13    [W:0.066 / U:1.908 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site