Re: [PATCH 07/12] repair: fix another ABBA deadlock in inode prefetching

xfs

Re: [PATCH 07/12] repair: fix another ABBA deadlock in inode prefetching

To:	Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject:	Re: [PATCH 07/12] repair: fix another ABBA deadlock in inode prefetching
From:	Dave Chinner <david@xxxxxxxxxxxxx>
Date:	Tue, 13 Dec 2011 13:35:14 +1100
Cc:	xfs@xxxxxxxxxxx
In-reply-to:	<20111202174742.307774458@xxxxxxxxxxxxxxxxxxxxxx>
References:	<20111202174619.179530033@xxxxxxxxxxxxxxxxxxxxxx> <20111202174742.307774458@xxxxxxxxxxxxxxxxxxxxxx>
User-agent:	Mutt/1.5.21 (2010-09-15)

On Fri, Dec 02, 2011 at 12:46:26PM -0500, Christoph Hellwig wrote:
> The inode prefetching code has a fixed limit of inodes that might are
> submitted at a time.  Unfortunately the buffers for them get locked
> once the prefetching starts.  That way the threads processing the inode
> might get stuck on buffer locked, but not submitted for reading yet.
> 
> Fix this by kicking the queue as soon as we would have to wait on the
> ra_count semaphore.
> 
> Reported-by: Arkadiusz Mi??kiewicz <arekm@xxxxxxxx>
> Tested-by: Arkadiusz Mi??kiewicz <arekm@xxxxxxxx>
> Signed-off-by: Christoph Hellwig <hch@xxxxxx>

Looks safe - I can't think up a scenario that would deadlock on
the second sem_wait call...

Reviewed-by: Dave Chinner <dchinner@xxxxxxxxxx>

-- 
Dave Chinner
david@xxxxxxxxxxxxx

[More with this subject...]

<Prev in Thread]	Current Thread	[Next in Thread>
[PATCH 00/12] xfs_repair queue, Christoph Hellwig [PATCH 04/12] xfsprogs: allow linking against libtcmalloc, Christoph Hellwig Re: [PATCH 04/12] xfsprogs: allow linking against libtcmalloc, Dave Chinner Re: [PATCH 04/12] xfsprogs: allow linking against libtcmalloc, Christoph Hellwig [PATCH 07/12] repair: fix another ABBA deadlock in inode prefetching, Christoph Hellwig Re: [PATCH 07/12] repair: fix another ABBA deadlock in inode prefetching, Dave Chinner <= [PATCH 05/12] repair: update extent count after zapping duplicate blocks, Christoph Hellwig Re: [PATCH 05/12] repair: update extent count after zapping duplicate blocks, Dave Chinner [PATCH 03/12] repair: allocate and free extent records individually, Christoph Hellwig Re: [PATCH 03/12] repair: allocate and free extent records individually, Dave Chinner [PATCH 09/12] repair: kill check_inode_block, Christoph Hellwig [PATCH 02/12] repair: allocate and free inode records individually, Christoph Hellwig Re: [PATCH 02/12] repair: allocate and free inode records individually, Dave Chinner [PATCH 12/12] repair: cleanup inode record macros, Christoph Hellwig [PATCH 10/12] repair: mark local functions static, Christoph Hellwig [PATCH 11/12] repair: move extern declarations to headers, Christoph Hellwig

Previous by Date:	Re: [PATCH 06/12] repair: use recursive buffer locking, Dave Chinner
Next by Date:	Re: [PATCH 08/12] repair: handle filesystems with the log in allocation group 0, Dave Chinner
Previous by Thread:	[PATCH 07/12] repair: fix another ABBA deadlock in inode prefetching, Christoph Hellwig
Next by Thread:	[PATCH 05/12] repair: update extent count after zapping duplicate blocks, Christoph Hellwig
Indexes:	[Date] [Thread] [Top] [All Lists]