xfs
[Top] [All Lists]

Re: [PATCH 2/5] xfs: reduce ilock hold times in xfs_file_aio_write_check

To: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject: Re: [PATCH 2/5] xfs: reduce ilock hold times in xfs_file_aio_write_checks
From: Dave Chinner <david@xxxxxxxxxxxxx>
Date: Tue, 27 Mar 2012 09:15:59 +1100
Cc: xfs@xxxxxxxxxxx
In-reply-to: <20120326211603.268269812@xxxxxxxxxxxxxxxxxxxxxx>
References: <20120326211421.518374058@xxxxxxxxxxxxxxxxxxxxxx> <20120326211603.268269812@xxxxxxxxxxxxxxxxxxxxxx>
User-agent: Mutt/1.5.21 (2010-09-15)
On Mon, Mar 26, 2012 at 05:14:23PM -0400, Christoph Hellwig wrote:
> We do not need the ilock for generic_write_checks and the i_size read,
> which are protected by i_mutex and/or iolock, so reduce the ilock
> critical section to just the call to xfs_zero_eof.
> 
> Signed-off-by: Christoph Hellwig <hch@xxxxxx>

That answers one of the open questions I had about my patch -
whether we needed to hold the ilock at all over
generic_write_checks(). I wasn't 100% sure, but this patch confirms
my thinking on it.

Reviewed-by: Dave Chinner <dchinner@xxxxxxxxxx>

-- 
Dave Chinner
david@xxxxxxxxxxxxx

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