From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261677AbVFKKlr (ORCPT ); Sat, 11 Jun 2005 06:41:47 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261680AbVFKKlr (ORCPT ); Sat, 11 Jun 2005 06:41:47 -0400 Received: from one.firstfloor.org ([213.235.205.2]:31648 "EHLO one.firstfloor.org") by vger.kernel.org with ESMTP id S261677AbVFKKlp (ORCPT ); Sat, 11 Jun 2005 06:41:45 -0400 To: martin@cs.uga.edu Cc: linux-kernel@vger.kernel.org Subject: Re: PROBLEM: OOPSes in PREEMPT SMP for AMD Opteron Dual-Core with Memhole Mapping (non tainted kernel) References: <200506071836.12076.martin@cs.uga.edu> From: Andi Kleen Date: Sat, 11 Jun 2005 12:41:44 +0200 In-Reply-To: <200506071836.12076.martin@cs.uga.edu> (Jacob Martin's message of "Tue, 7 Jun 2005 18:36:12 -0400") Message-ID: User-Agent: Gnus/5.110002 (No Gnus v0.2) Emacs/21.3 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Jacob Martin writes: > On second thought, the MTRR stuff may not have anything to do with this. I think it has or the strange software memhole (what is that?) > > > I think this is a bug in the preemptable kernel (I was warned, but didn't listen :). You could verify that by disabling PREEMPT. Butactually it should not be that bad. >>From your oopses i more suspect it is hardware actually: > Jun 7 14:11:27 optimator Unable to handle kernel paging request at 00000000000025b0 RIP: > Jun 7 14:11:27 optimator {pte_alloc_map+170} It crashes on first accessing newly allocated memory for page tables. Most likely something is wrong with your memory or memory map. Maybe it is related to your "discrete mtrr mapping" or your "software memhole" whatever they are? I would suggest to try without these. -Andi