From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S934151Ab2GLRCv (ORCPT ); Thu, 12 Jul 2012 13:02:51 -0400 Received: from mx1.redhat.com ([209.132.183.28]:41845 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S934096Ab2GLRCu (ORCPT ); Thu, 12 Jul 2012 13:02:50 -0400 Date: Thu, 12 Jul 2012 13:02:42 -0400 From: Dave Jones To: Kay Sievers Cc: Linux Kernel , Linus Torvalds , Greg Kroah-Hartmann Subject: Re: 3.5-rc6 printk formatting problem during oom-kill. Message-ID: <20120712170242.GA11890@redhat.com> Mail-Followup-To: Dave Jones , Kay Sievers , Linux Kernel , Linus Torvalds , Greg Kroah-Hartmann References: <20120709180328.GA4105@redhat.com> <1341859731.801.0.camel@mop> <20120712005405.GA1613@redhat.com> <20120712140552.GA13745@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Jul 12, 2012 at 06:47:57PM +0200, Kay Sievers wrote: > > > > Also, I have no idea how the hell the 'Modules linked in:' line (9th line) ended up being printed /after/ the > > > > module listing began (2nd line). > > They do not belong together. The second line is just another call to > the same print_modules() done from: > arch/x86/kernel/dumpstack_64.c :: show_regs() > > While we already called print_modules() a few cycles earlier from: > kernel/watchdog :: watchdog_timer_fn() ah, derp. Good spotting. > Hmm, it does not trigger your pattern. I tried adding an rmmod in that > loop, but that crashes after a few seconds. Some modules are just not > meant to be removed. :) yeah, that's bothersome. TTBOMK, we've never survived a 'load all modules/unload all modules' ever. > I forced the watchdog to trigger by setting the timeout to 1s, but all > looks still fine: > > Could it possibly be that we get some sort of corruption somewhere > else while running trinity and load modules? I suppose anything is possible, but it's surprisingly consistent. Dave