From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261442AbULIDvQ (ORCPT ); Wed, 8 Dec 2004 22:51:16 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261450AbULIDvQ (ORCPT ); Wed, 8 Dec 2004 22:51:16 -0500 Received: from pop5-1.us4.outblaze.com ([205.158.62.125]:55238 "HELO pop5-1.us4.outblaze.com") by vger.kernel.org with SMTP id S261442AbULIDvN (ORCPT ); Wed, 8 Dec 2004 22:51:13 -0500 Subject: Re: oops with dual xeon 2.8ghz 4gb ram +smp, software raid, lvm, and xfs From: Nigel Cunningham Reply-To: ncunningham@linuxmail.org To: Andrew Morton Cc: David Greaves , phil@dier.us, Linux Kernel Mailing List In-Reply-To: <20041208011546.2d509d1b.akpm@osdl.org> References: <20041122130622.27edf3e6.phil@dier.us> <20041122161725.21adb932.akpm@osdl.org> <20041124094549.4c51d6d5.phil@dier.us> <20041124151234.714f30d4.akpm@osdl.org> <41A9B693.30905@dgreaves.com> <20041128102751.2dac71f7.akpm@osdl.org> <41B6C34B.7030700@dgreaves.com> <20041208011546.2d509d1b.akpm@osdl.org> Content-Type: text/plain Message-Id: <1102540749.4209.8.camel@desktop.cunninghams> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.6-1mdk Date: Thu, 09 Dec 2004 14:50:46 +1100 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi Andrew. On Wed, 2004-12-08 at 20:15, Andrew Morton wrote: > David Greaves wrote: > > > > I did as you suggested and it's been fine until I got this last night. > > > > Dec 8 06:50:04 cu kernel: slab: Internal list corruption detected in > > cache 'vm_area_struct'(41), slabp cfedd000(13). > > That's totally different from the previous oops (it was in dcache). > > I'd be suspecting either a random memory scribble or flakey hardware. It > could well be the latter if you're not using any unusual > drivers/filesystems/etc. I'm seeing similar things occasionally with 2.6.9+kgdb+suspend2+Win4Lin on ht/preempt/regparm/4gb highmem. I've come to the conclusion it's probably not directly suspend (I can do 100 cycles on the trot), but haven't been able to reliably reproduce it. The corruption is always in fs related data, but apart from that seems random. Seeing the mention of being unable to allocate a page at the bottom of David's email makes me wonder if the difficulties with memory freeing are triggering some code that's not properly handling failed page allocations. Regards, Nigel -- Nigel Cunningham Pastoral Worker Christian Reformed Church of Tuggeranong PO Box 1004, Tuggeranong, ACT 2901 You see, at just the right time, when we were still powerless, Christ died for the ungodly. -- Romans 5:6