From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933053AbXCEK2r (ORCPT ); Mon, 5 Mar 2007 05:28:47 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S933046AbXCEK2r (ORCPT ); Mon, 5 Mar 2007 05:28:47 -0500 Received: from www.osadl.org ([213.239.205.134]:39179 "EHLO mail.tglx.de" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S933044AbXCEK2b (ORCPT ); Mon, 5 Mar 2007 05:28:31 -0500 Subject: Re: 2.6.21-rc2-git3 soft lockup detected on CPU#0 (crash dump kernel) From: Thomas Gleixner Reply-To: tglx@linutronix.de To: Ingo Molnar Cc: Michal Piotrowski , LKML In-Reply-To: <20070305080824.GA20583@elte.hu> References: <6bffcb0e0703031531h3f0a86c6sda2522c0ca504765@mail.gmail.com> <1173018209.24738.135.camel@localhost.localdomain> <6bffcb0e0703040636q5a4ed3ekf5a0461052e8e018@mail.gmail.com> <1173022668.24738.141.camel@localhost.localdomain> <6bffcb0e0703040828j46a339c2ofabf7c0e20f6ae09@mail.gmail.com> <1173026442.24738.143.camel@localhost.localdomain> <6bffcb0e0703040847g1c04940bm4c22895ec2540bd1@mail.gmail.com> <1173027823.24738.156.camel@localhost.localdomain> <6bffcb0e0703040915h442b4d6cq8bb47cd961165916@mail.gmail.com> <20070305080824.GA20583@elte.hu> Content-Type: text/plain Date: Mon, 05 Mar 2007 11:34:29 +0100 Message-Id: <1173090869.24738.198.camel@localhost.localdomain> Mime-Version: 1.0 X-Mailer: Evolution 2.6.1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Mon, 2007-03-05 at 09:08 +0100, Ingo Molnar wrote: > * Michal Piotrowski wrote: > > > http://www.stardust.webpages.pl/files/tbf/bitis-gabonica/2.6.21-rc2-git3-kdump/log2.txt > > weird. Could you do a dump with SysRq-T, SysRq-Q done /twice/ after each > other, to see how the system progresses while in this lockup? Plus > please also try to get a register dump via SysRq-P. > > Btw., the task state printout seems busted: > > udev_run_devd ? C15B0030 2544 979 853 (L-TLB) > > shouldnt that be 'R' instead of the question mark? ? are dead ones, [] do_exit+0x6f9/0x6fd [] sys_exit_group+0x0/0x11 [] sys_exit_group+0xf/0x11 [] sysenter_past_esp+0x5f/0x99 The interesting part is here: 1 lock held by udevd/918: #0: (&mm->mmap_sem){----}, at: [] do_page_fault+0x150/0x50d But udevd/918 is nowhere in the sysrq-t output. tglx