xfs
[Top] [All Lists]

Re: [PATCH] xfs: call xfs_idestroy_fork() in xfs_ilock() critical sectio

To: Waiman Long <Waiman.Long@xxxxxx>
Subject: Re: [PATCH] xfs: call xfs_idestroy_fork() in xfs_ilock() critical section
From: Brian Foster <bfoster@xxxxxxxxxx>
Date: Wed, 22 Apr 2015 15:11:37 -0400
Cc: Dave Chinner <david@xxxxxxxxxxxxx>, xfs@xxxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <1429724021-7675-1-git-send-email-Waiman.Long@xxxxxx>
References: <1429724021-7675-1-git-send-email-Waiman.Long@xxxxxx>
User-agent: Mutt/1.5.23 (2014-03-12)
On Wed, Apr 22, 2015 at 01:33:41PM -0400, Waiman Long wrote:
> The commit f7be2d7f594cbc ("xfs: push down inactive transaction
> mgmt for truncate") refactored the xfs_inactive() function
> in fs/xfs/xfs_inode.c.  However, it also moved the call to
> xfs_idestroy_fork() from inside the xfs_ilock() critical section to
> outside. That was causing memory corruption and strange failures like
> deferencing NULL pointers in some circumstances.
> 
> This patch moves the xfs_idestroy_fork() call back into an xfs_ilock()
> critical section to avoid memory corruption problem.
> 
> Signed-off-by: Waiman Long <Waiman.Long@xxxxxx>
> ---

Interesting... so from your previous mail we have an inactive/reclaim
racing with an xfs_iflush_fork() of the attr fork, or something of that
nature? Is there a specific reproducer or is it some kind of stress
test?

Good catch in any case, it looks like a deviation from the previous
code...

>  fs/xfs/xfs_inode.c |    5 ++++-
>  1 files changed, 4 insertions(+), 1 deletions(-)
> 
> diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
> index 6163767..31850fb 100644
> --- a/fs/xfs/xfs_inode.c
> +++ b/fs/xfs/xfs_inode.c
> @@ -1900,8 +1900,11 @@ xfs_inactive(
>                       return;
>       }
>  
> -     if (ip->i_afp)
> +     if (ip->i_afp) {
> +             xfs_ilock(ip, XFS_ILOCK_EXCL);
>               xfs_idestroy_fork(ip, XFS_ATTR_FORK);
> +             xfs_iunlock(ip, XFS_ILOCK_EXCL);
> +     }

It probably doesn't matter, but I wonder if it would be better to just
place the lock outside of the ip->i_afp check to preserve the original
behavior if nothing else...

Brian

>  
>       ASSERT(ip->i_d.di_anextents == 0);
>  
> -- 
> 1.7.1
> 
> _______________________________________________
> xfs mailing list
> xfs@xxxxxxxxxxx
> http://oss.sgi.com/mailman/listinfo/xfs

<Prev in Thread] Current Thread [Next in Thread>