Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756074AbYGHAA2 (ORCPT ); Mon, 7 Jul 2008 20:00:28 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1754664AbYGHAAS (ORCPT ); Mon, 7 Jul 2008 20:00:18 -0400 Received: from e5.ny.us.ibm.com ([32.97.182.145]:37602 "EHLO e5.ny.us.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754650AbYGHAAR (ORCPT ); Mon, 7 Jul 2008 20:00:17 -0400 Subject: Re: [PATCH] [0/9] Use 64bit x86 machine check code for 32bit too From: Max Asbock To: "H. Peter Anvin" Cc: Andi Kleen , x86@kernel.org, linux-kernel@vger.kernel.org In-Reply-To: <48728C81.7080708@kernel.org> References: <200807041120.678642023@firstfloor.org> <48728A0E.7010802@kernel.org> <48728C81.7080708@kernel.org> Content-Type: text/plain; charset=UTF-8 Date: Mon, 07 Jul 2008 17:00:15 -0700 Message-Id: <1215475215.6036.10.camel@w-amax.beaverton.ibm.com> Mime-Version: 1.0 X-Mailer: Evolution 2.22.2 Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1459 Lines: 40 On Mon, 2008-07-07 at 14:37 -0700, H. Peter Anvin wrote: > H. Peter Anvin wrote: > > Andi Kleen wrote: > >> This patchkit uses the 64bit machine check code which is better in many > >> ways on 32bit x86 too. This is also the basis for some future machine > >> check work. > > > > Hi Andi, > > > > I tried applying your patch kit to -tip, but I'm getting build failures > > on i386; gcc-4.1.1-51.fc5.x86_64 with attached configuration. > > > > FWIW, this is rooting the branch on top of current linus. Applied on > top of tip/auto-x86-next or tip/master, the patch does not apply. > > -hpa The patches won't apply to -tip because Venki's patch has to be backed out of mce_64.c and it looks like thread_info.h has been merged in the mean time. I was able to apply the patches to 2.6.26-rc8 though, build it and boot it. However it turns out the build warnings about "mce_64.c:822: warning: passing argument 2 of ‘strict_strtoull’ makes integer from pointer without a cast" are justified. It looks like strict_strtoull has the last two arguments switched with respect to simple_strtoull. You get a NULL pointer dereference as soon as you write to one of the machinecheck sysfs control files. Max -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/