From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: with ECARTIS (v1.0.0; list xfs); Tue, 28 Oct 2008 22:27:25 -0700 (PDT) Received: from cuda.sgi.com (cuda2.sgi.com [192.48.168.29]) by oss.sgi.com (8.12.11.20060308/8.12.11/SuSE Linux 0.7) with ESMTP id m9T5RBbq007633 for ; Tue, 28 Oct 2008 22:27:11 -0700 Received: from ipmail01.adl6.internode.on.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id BF132556CFC for ; Tue, 28 Oct 2008 22:27:11 -0700 (PDT) Received: from ipmail01.adl6.internode.on.net (ipmail01.adl6.internode.on.net [203.16.214.146]) by cuda.sgi.com with ESMTP id 5rnpDDNClKvpkk2i for ; Tue, 28 Oct 2008 22:27:11 -0700 (PDT) Date: Wed, 29 Oct 2008 16:27:07 +1100 From: Dave Chinner Subject: Re: [patch 0/9] writeback data integrity and other fixes (take 3) Message-ID: <20081029052707.GH4985@disturbed> References: <20081028144715.683011000@suse.de> <20081028153953.GB3082@wotan.suse.de> <20081028222746.GB4985@disturbed> <20081029001653.GF15599@wotan.suse.de> <20081029031645.GE4985@disturbed> <20081029040014.GB17624@wotan.suse.de> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20081029040014.GB17624@wotan.suse.de> Sender: xfs-bounce@oss.sgi.com Errors-to: xfs-bounce@oss.sgi.com List-Id: xfs To: Nick Piggin Cc: akpm@linux-foundation.org, xfs@oss.sgi.com, linux-fsdevel@vger.kernel.org, Chris Mason , linux-ext4@vger.kernel.org On Wed, Oct 29, 2008 at 05:00:14AM +0100, Nick Piggin wrote: > On Wed, Oct 29, 2008 at 02:16:45PM +1100, Dave Chinner wrote: > > FWIW, the core issue here is that we've got to do the > > filemap_fdatawait() call in the ->fsync method because ->fsync > > gets called before we've waited for the data I/O to complete. > > XFS updates inode state on I/O completion, so we *must* wait > > for data I/O to complete before logging the inode changes. I > > think btrfs has the same problem.... > > Interesting. Does that mean you can do without the final filemap_fdatawait? We could, yes. > Can you do the first fdatawait without i_mutex held? I don't see why not - I/O completion is only touching the XFS inode. XFS has it's own inode locks for I/O and inode exclusion, and the mapping tree lock protects the tree walk that the fdatawait is doing... > There was some talk IIRC about moving all this logic into the filesystem > to provide more flexibility here. If there is still enough interest, > we should get that moving... It would simplify some of the code in XFS, that's for sure ;) Cheers, Dave. -- Dave Chinner david@fromorbit.com