xfs
[Top] [All Lists]

Re: [PATCH 2/2] xfs: fix splice/direct-IO deadlock

To: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject: Re: [PATCH 2/2] xfs: fix splice/direct-IO deadlock
From: Dave Chinner <david@xxxxxxxxxxxxx>
Date: Thu, 29 Nov 2012 08:33:14 +1100
Cc: linux-fsdevel@xxxxxxxxxxxxxxx, xfs@xxxxxxxxxxx
In-reply-to: <20121128160736.GA14888@xxxxxxxxxxxxx>
References: <1354068768-4316-1-git-send-email-david@xxxxxxxxxxxxx> <1354068768-4316-3-git-send-email-david@xxxxxxxxxxxxx> <20121128160736.GA14888@xxxxxxxxxxxxx>
User-agent: Mutt/1.5.21 (2010-09-15)
On Wed, Nov 28, 2012 at 11:07:36AM -0500, Christoph Hellwig wrote:
> On Wed, Nov 28, 2012 at 01:12:48PM +1100, Dave Chinner wrote:
> > From: Dave Chinner <dchinner@xxxxxxxxxx>
> > 
> > lockdep reports splice vs direct-io write lock inversions due to
> > generic_file_splice_write() taking the inode->i_mutex inside
> > XFS_IOLOCK_EXCL context. These lock contexts are inverted, hence can
> > deadlock.  Remove the XFS_IOLOCK_EXCL locking context from the outer
> > function and drive it inwards to the actor function that only locks
> > the inode when the lock is really needed,
> 
> punctuation?
> 
> Otherwise the patch looks fine, but I'd love to understand why the
> generic code thes te I_MUTEX_CHILD annotation and we can get away
> without it.
> 
> Also can you add a testcase for this to xfstests?

I've had trouble reproducing it reliably. The only cases I've seen
it occur are some custom FIO workloads. I'll try again to see if I
can come up with something that reliably hits it...

Cheers,

Dave.
-- 
Dave Chinner
david@xxxxxxxxxxxxx

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