From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S934082AbXC2Ty2 (ORCPT ); Thu, 29 Mar 2007 15:54:28 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S934143AbXC2Ty2 (ORCPT ); Thu, 29 Mar 2007 15:54:28 -0400 Received: from waste.org ([66.93.16.53]:34733 "EHLO waste.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S934082AbXC2Ty1 (ORCPT ); Thu, 29 Mar 2007 15:54:27 -0400 Date: Thu, 29 Mar 2007 14:40:49 -0500 From: Matt Mackall To: Mariusz =?utf-8?Q?Koz=C5=82owski?= Cc: Ingo Molnar , Andrew Morton , Adrian Bunk , linux-kernel@vger.kernel.org, Thomas Gleixner , john stultz Subject: Re: 2.6.21-rc5-mm1 Message-ID: <20070329194048.GG4892@waste.org> References: <20070326205706.a750bb35.akpm@linux-foundation.org> <200703292045.13140.m.kozlowski@tuxland.pl> <20070329183857.GA13362@elte.hu> <200703292102.56503.m.kozlowski@tuxland.pl> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <200703292102.56503.m.kozlowski@tuxland.pl> User-Agent: Mutt/1.5.9i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Mar 29, 2007 at 09:02:56PM +0200, Mariusz Kozłowski wrote: > > > > > This is verified and repeatable _every_ single time I tried. > > > > > Unfortunately the last thing seen on the screen before system is > > > > > frozen is 'eth0: link down'. So my guess is that when hunting for > > > > > hangs I found something else that can hang my laptop (netconsole > > > > > that is). > > > > > > > > fix-sysfs-reclaim-crash.patch is known-bad. Did you try reverting > > > > that? > > > > > > Yes, just tested it. Sorry. It doesn't change anything. > > > > could you try a !CONFIG_SMP kernel to make sure it's not some > > self-locking of netconsole that is interfering? (Is it perhaps the > > 'eth0: link down' message that is trigger from within netconsole?) > > This is from .config: > > # CONFIG_SMP is not set > > Can this be that 'link down' is emitted by netconsole when the link is > already down and kernel gets locked? Does it make any sense? Yes, that's probably exactly the problem. But it probably involves locks. I think a preemptible kernel can also cause the problem? -- Mathematics is the supreme nostalgia of our time.