From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from cuda.sgi.com (cuda3.sgi.com [192.48.176.15]) by oss.sgi.com (8.14.3/8.14.3/SuSE Linux 0.8) with ESMTP id qAD339oV015027 for ; Mon, 12 Nov 2012 21:03:09 -0600 Received: from ipmail07.adl2.internode.on.net (ipmail07.adl2.internode.on.net [150.101.137.131]) by cuda.sgi.com with ESMTP id S58qH93DNomaUQu6 for ; Mon, 12 Nov 2012 19:05:10 -0800 (PST) Date: Tue, 13 Nov 2012 14:05:07 +1100 From: Dave Chinner Subject: Re: [PATCH 0/3] xfs: fixes for 3.7-rc6 Message-ID: <20121113030507.GE25039@dastard> References: <1352718586-3538-1-git-send-email-david@fromorbit.com> <20121113020150.GY9783@sgi.com> MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: <20121113020150.GY9783@sgi.com> List-Id: XFS Filesystem from SGI List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: xfs-bounces@oss.sgi.com Errors-To: xfs-bounces@oss.sgi.com To: Ben Myers Cc: xfs@oss.sgi.com On Mon, Nov 12, 2012 at 08:01:50PM -0600, Ben Myers wrote: > Hi Dave, > > On Mon, Nov 12, 2012 at 10:09:43PM +1100, Dave Chinner wrote: > > These are the remaining fixes I have in my tree for the 3.7 cycle. > > One is a corruption fix, the other two are problems I'm seeing > > during shutdown and recovery testing. That testing lasts 400-500 > > cycles (3-4 hours) now before it fails with another, as yet unknown > > buffer/perag reference assert failure. Without these fixes, it lasts > > 3-4 cycles before assert failing or triggering a BUG. > > > > I'd like these fixes to be in the 3.7 cycle if at all possible, with > > both patch 1 and 2 being stable kernel candidates. > > I think we can make that happen. I noticed that you posted patches 2 and 3 > earlier today and had to chuckle, but I understand. Fixing a nasty bug or regression bug and asking it be merged into the current -rc cycle after initial testing is not unusual, so I'm not sure why you'd chuckle at that.... > We'll put this 3.7 work at > the top of the priority queue, followed by your 3.8 queue, and then the few > items left for the userspace release. Just FYI, the first thing that needs to be done after the userspace release is to resync the libxfs/libxlog code with the kernel code. That needs to be done before any of the CRC changes can be merged, so be prepared for a pretty major update that'll need review soon after the release. IIRC, the last kernel/user resync I did was about 15,000 lines of change I don't expect this to be quite as big (the previous one contained the btree consolidation changes), but I'd expect it to be in the order of 2-3000 lines of change to be needed. > Thanks for reposting your 3.8 queue, BTW. Sometimes I think that frequent > reposts are inconvenient, but in this case I think it will be helpful. When I'm having trouble tracking of several of my own patch sets for the same release, it's time for re-organisation and a repost. Besides, it's a habit Christoph and I got into a while back when we had lots of work going on and we needed to make sure nothing slipped through the cracks..... :) Cheers, Dave. -- Dave Chinner david@fromorbit.com _______________________________________________ xfs mailing list xfs@oss.sgi.com http://oss.sgi.com/mailman/listinfo/xfs