From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756387Ab3A3R2H (ORCPT ); Wed, 30 Jan 2013 12:28:07 -0500 Received: from mail-ve0-f169.google.com ([209.85.128.169]:54981 "EHLO mail-ve0-f169.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756316Ab3A3R2E (ORCPT ); Wed, 30 Jan 2013 12:28:04 -0500 MIME-Version: 1.0 In-Reply-To: <20130130034204.GB30965@pd.tnic> References: <20130129202848.GE25415@pd.tnic> <20130130034204.GB30965@pd.tnic> From: Bjorn Helgaas Date: Wed, 30 Jan 2013 10:27:42 -0700 Message-ID: Subject: Re: Uhhuh. NMI received for unknown reason 2c on CPU 0. To: Borislav Petkov , Bjorn Helgaas , x86@kernel.org, lkml , Konstantin Khlebnikov , "Rafael J. Wysocki" Content-Type: text/plain; charset=ISO-8859-1 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, Jan 29, 2013 at 8:42 PM, Borislav Petkov wrote: > On Tue, Jan 29, 2013 at 02:32:56PM -0700, Bjorn Helgaas wrote: >> Konstantin has some fixes for an e1000e power management issue related >> to suspend/resume that he observed on an x220. He didn't see an NMI, >> and apparently his problem has been around for a long time, > > Yeah, this is one of those issues you don't see *every* s/r cycle and > besides, I just got this box and haven't run 3.{6,7} on it yet (maybe > never will :-)). > >> so no idea whether it could be related. I just noticed the conjunction >> of thinkpad/e1000e/resume/power saving in both reports. >> >> https://lkml.org/lkml/2013/1/18/147 > > Yes, thanks Bjorn, that was a good suggestion. Btw, from reading the > thread, those patches still need cooking a bit more, AFAICR people's > objections/comments. Or should I go ahead and test them? You're right, I don't think we're quite ready to merge those patches. But if your NMI is easy to reproduce, it might be worth removing e1000e altogether to see if it still happens. I noticed in your original log that the NMI occurred 5 seconds after the e1000e message, and I could imagine some CPU or PCI response timeout being 5 seconds. > It's just that I'm overly cautious every time I hear e1000e is involved: > > www.linux-magazine.com/content/download/62169/484085/file/Security_Lessons_Ftrace.pdf Thanks for the pointer, that was an interesting read :) Bjorn