From mboxrd@z Thu Jan 1 00:00:00 1970 Content-Type: multipart/mixed; boundary="===============6170374644836844812==" MIME-Version: 1.0 From: Dave Jones To: lkp@lists.01.org Subject: Re: More information please. Re: [fs] 54cc07a761: BUG: kernel test crashed Date: Tue, 26 Jul 2016 17:18:59 -0400 Message-ID: <20160726211859.GB9577@codemonkey.org.uk> In-Reply-To: <87wpk8p49p.fsf@x220.int.ebiederm.org> List-Id: --===============6170374644836844812== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable On Tue, Jul 26, 2016 at 02:50:58PM -0500, Eric W. Biederman wrote: > > Yes that's the right directions to follow. We'll make the reports more > > understandable and the test/bisects more reliable. > = > Thanks. Does trinity have a random seed it can export/import to rerun > the same tests? I ask because there was a failure these tests caught > that if the right kernel options was enabled was 100% reliable and it > blamed a commit 10 patches down from the indicated commit. It does log the random seed on startup, and can replay it with the -s option There is some variance if you were to do two runs with the same seed back to back, because on startup, it walks /proc & /sys, and they'd see differen= t things so for eg "read() from fd 100" might mean something entirely different. But as the testbot runs a pretty minimal userspace, only runs trinity once, and really early in boot, this point might not matter. Dave --===============6170374644836844812==-- From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1161161AbcGZVTI (ORCPT ); Tue, 26 Jul 2016 17:19:08 -0400 Received: from arcturus.aphlor.org ([188.246.204.175]:42974 "EHLO arcturus.aphlor.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1161029AbcGZVTE (ORCPT ); Tue, 26 Jul 2016 17:19:04 -0400 Date: Tue, 26 Jul 2016 17:18:59 -0400 From: Dave Jones To: "Eric W. Biederman" Cc: Fengguang Wu , containers@lists.linux-foundation.org, lkp@01.org, LKML Subject: Re: [LKP] More information please. Re: [fs] 54cc07a761: BUG: kernel test crashed Message-ID: <20160726211859.GB9577@codemonkey.org.uk> Mail-Followup-To: Dave Jones , "Eric W. Biederman" , Fengguang Wu , containers@lists.linux-foundation.org, lkp@01.org, LKML References: <5793dfa5.mvXPj4LJOS0dlz3i%xiaolong.ye@intel.com> <877fc9a6mb.fsf@x220.int.ebiederm.org> <20160726143545.GA23656@wfg-t540p.sh.intel.com> <87vazsv4cn.fsf@x220.int.ebiederm.org> <20160726151344.GB8811@wfg-t540p.sh.intel.com> <87wpk8p49p.fsf@x220.int.ebiederm.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <87wpk8p49p.fsf@x220.int.ebiederm.org> User-Agent: Mutt/1.6.0 (2016-04-01) X-Spam-Flag: skipped (authorised relay user) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Jul 26, 2016 at 02:50:58PM -0500, Eric W. Biederman wrote: > > Yes that's the right directions to follow. We'll make the reports more > > understandable and the test/bisects more reliable. > > Thanks. Does trinity have a random seed it can export/import to rerun > the same tests? I ask because there was a failure these tests caught > that if the right kernel options was enabled was 100% reliable and it > blamed a commit 10 patches down from the indicated commit. It does log the random seed on startup, and can replay it with the -s option There is some variance if you were to do two runs with the same seed back to back, because on startup, it walks /proc & /sys, and they'd see different things so for eg "read() from fd 100" might mean something entirely different. But as the testbot runs a pretty minimal userspace, only runs trinity once, and really early in boot, this point might not matter. Dave