From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from ipmail04.adl6.internode.on.net ([150.101.137.141]:51509 "EHLO ipmail04.adl6.internode.on.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750779AbbGVVKx (ORCPT ); Wed, 22 Jul 2015 17:10:53 -0400 Date: Thu, 23 Jul 2015 07:10:50 +1000 From: Dave Chinner To: Josef Bacik Cc: Chandan Rajendra , Brian Foster , fstests@vger.kernel.org, linux-btrfs@vger.kernel.org Subject: Re: [PATCH] generic/311: Disable dmesg check Message-ID: <20150722211050.GZ3902@dastard> References: <1437109003-2357-1-git-send-email-chandan@linux.vnet.ibm.com> <55ACEFC4.9040502@fb.com> <20150720221220.GU7943@dastard> <8410729.YTcdQ5sv7i@localhost.localdomain> <20150722000105.GX3902@dastard> <55AEE5D3.6020509@fb.com> <20150722052700.GY3902@dastard> <55AF7E77.5020803@fb.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <55AF7E77.5020803@fb.com> Sender: linux-btrfs-owner@vger.kernel.org List-ID: On Wed, Jul 22, 2015 at 07:28:55AM -0400, Josef Bacik wrote: > On 07/22/2015 01:27 AM, Dave Chinner wrote: > >>>Josef, Chris, is this really how btrfs handles metadata write > >>>failures? This handling of errors seems like a design flaw rather > >>>than a desireable behaviour to me.... > >>> > >> > >>Jeeze you are cranky Dave, lets think about what this test is doing. > > > >Do you really see anything above that indicates I'm cranky? I'm just > >following the logic as i read it, reaching a conclusion and the > >saying "this doesn't sound right - can an expert please confirm I've > >got this right?". Last thing I expected was to be shouted at and > >treated like an idiot. > > > >BTW, you'll know when I'm cranky - swearing is a real good > >indication that I'm pissed off at something or someone. > > Sorry, I didn't mean to come across angry or condescending. > Obviously I appreciate your expertise and questions, I wasn't pissed > or anything, this is just how I normally talk, you know I use > swearing as punctuation. Yes, I know that, and that doesn't worry me in the least. You don't normally shout, though. > >>This is part of the reason for dm-log-writes, to solve this kind of > >>shitty situation that dm-flakey introduces. > > > >Which you've also never explained in any of the dm-log-write patches > >you've posted in the past. IOWs, get the dm-log-writes support code > >reviewed and fix the dm-flakey tests to use it so this problem goes > >away. > > The dm-log-writes target is already upstream, I sent the xfstests > stuff months ago, not sure if they ever made it in. I'll check and > get them posted again if not. Thanks, They didn't, because it wasn't clear when the kernel code was going to be committed. I did notice yesterday that it was upstream. Cheers, Dave. -- Dave Chinner david@fromorbit.com