Re: [PATCH] xfs: avoid direct I/O write vs buffered I/O race

To:	Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject:	Re: [PATCH] xfs: avoid direct I/O write vs buffered I/O race
From:	Dave Chinner <david@xxxxxxxxxxxxx>
Date:	Tue, 30 Aug 2011 16:29:24 +1000
Cc:	xfs@xxxxxxxxxxx
In-reply-to:	<20110827144253.GA16381@xxxxxxxxxxxxx>
References:	<20110827144253.GA16381@xxxxxxxxxxxxx>
User-agent:	Mutt/1.5.21 (2010-09-15)

On Sat, Aug 27, 2011 at 10:42:53AM -0400, Christoph Hellwig wrote:
> Currently a buffered reader or writer can add pages to the pagecache
> while we are waiting for the iolock in xfs_file_dio_aio_write.  Prevent
> this by re-checking mapping->nrpages after we got the iolock, and if
> nessecary upgrade the lock to exclusive mode.  To simplify this a bit
> only take the ilock inside of xfs_file_aio_write_checks.
> 
> Signed-off-by: Christoph Hellwig <hch@xxxxxx>

Looks sane. Pushing the ILOCK completely into
xfs_file_aio_write_checks() is a nice segregation of locking
responsibilities.

Reviewed-by: Dave Chinner <dchinner@xxxxxxxxxx>

-- 
Dave Chinner
david@xxxxxxxxxxxxx

<Prev in Thread]	Current Thread	[Next in Thread>
[PATCH] xfs: avoid direct I/O write vs buffered I/O race, Christoph Hellwig Re: [PATCH] xfs: avoid direct I/O write vs buffered I/O race, Dave Chinner <=

Previous by Date:	Re: [PATCH 2/2] xfs: fix ->write_inode return values, Dave Chinner
Next by Date:	Re: [PATCH] xfs: avoid synchronous transactions when deleting attr blocks, Dave Chinner
Previous by Thread:	[PATCH] xfs: avoid direct I/O write vs buffered I/O race, Christoph Hellwig
Next by Thread:	[PATCH] xfs: avoid synchronous transactions when deleting attr blocks, Christoph Hellwig
Indexes:	[Date] [Thread] [Top] [All Lists]