From mboxrd@z Thu Jan 1 00:00:00 1970 From: Chris Jones Subject: e1000e corruption timeline and logs - ubuntu Date: Fri, 26 Sep 2008 02:21:07 +0100 Message-ID: <48DC3903.5060207@canonical.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit To: jkosina@suse.cz, netdev@vger.kernel.org, linux-kernel@vger.kernel.org, tim.gardner@canonical.com, "Brandeburg, Jesse" Return-path: Received: from adelie.canonical.com ([91.189.90.139]:60392 "EHLO adelie.canonical.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752522AbYIZBVK (ORCPT ); Thu, 25 Sep 2008 21:21:10 -0400 Sender: netdev-owner@vger.kernel.org List-ID: Hi (Apologies for replying in a new thread and crossposting, I'm not subscribed to these lists and maybe this will make for useful reading. Also please CC me in replies) wrt Jiri Kosina's post: alpine.LNX.1.10.0809260042380.3389 () pegasus ! suse ! cz - I hit this bug at the end of August after upgrading to Ubuntu Intrepid (which was alpha at the time). Perhaps my timeline[0] would help provide some information to rule potential culprits out or in by way of Ubuntu's code at the time. Also I have some logs[2-5] that cover the period surrounding the corruption happening. The timeline shows a <24 hour window from my upgrading to Ubuntu Intrepid (and thus getting a 2.6.27 kernel), and the damage actually happening. In conjunction with pages like https://launchpad.net/ubuntu/+source/linux and https://launchpad.net/ubuntu/+source/xserver-xorg-video-intel it should be possible to figure out exactly what source I had in use. My reading of those pages suggests I had this .config[1] (which makes sense, it's the oldest 2.6.27 one I have installed and there doesn't seem to have been a subsequent publishing of an Ubuntu kernel within the failure window). It also suggests I had xserver-xorg-video-intel version2:2.4.1-1ubuntu1 (I would not necessarily trust the version number cf the vanilla intel driver - tracking down the source package in Launchpad or via #ubuntu-x on freenode would show exactly which patches were applied). Similarly, libgl1-mesa-dri 7.1-1ubuntu1. Apart from (obviously) any occurances where they failed to be written to disk, I have kern.log[2] and messages[3] for the period (although not syslog itself, that gets rotated out too quickly, unfortunately, but I expect kern.log will have any interesting stuff there is to be found). The mail in [0] references a kernel WARNING bug on Launchpad which ought to be in kern.log, but just in case it isn't, that bug has it attached. I also have logs[4][5] which show all package activity before and after the point of failure, should you be curious about any other versions. Finally I have some lspci info[6] for the chip from before the error. Please let me know if there is anything else I can try and pull off my (now post-RMA) laptop (Thinkpad X300) which may help. [0] http://mairukipa.tenshu.net/e1000e/timeline.txt [1] http://mairukipa.tenshu.net/e1000e/config-2.6.27-1-generic [2] http://mairukipa.tenshu.net/e1000e/kern.log.2.gz [3] http://mairukipa.tenshu.net/e1000e/messages.2.gz [4] http://mairukipa.tenshu.net/e1000e/dpkg.log-2008-08-28 [5] http://mairukipa.tenshu.net/e1000e/dpkg.log-2008-08-29 [6] http://mairukipa.tenshu.net/e1000e/lspci.txt Cheers, -- Chris Jones cmsj@canonical.com www.canonical.com