From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757047AbXKWS3X (ORCPT ); Fri, 23 Nov 2007 13:29:23 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752861AbXKWS3O (ORCPT ); Fri, 23 Nov 2007 13:29:14 -0500 Received: from dizzy.i200.net ([74.86.142.210]:55094 "EHLO dizzy.i200.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752700AbXKWS3N (ORCPT ); Fri, 23 Nov 2007 13:29:13 -0500 Message-ID: <47471BF9.8020100@quo.to> Date: Fri, 23 Nov 2007 12:29:13 -0600 From: Jordan Russell User-Agent: Thunderbird 2.0.0.9 (Windows/20071031) MIME-Version: 1.0 To: Chuck Ebbert CC: Andrew Morton , linux-kernel@vger.kernel.org, Thomas Gleixner , Ingo Molnar , john stultz Subject: Re: 2.6.23.1: Random hangs during boot with "tsc" clocksource References: <472B59E8.3080709@quo.to> <20071106221548.c8d7b41b.akpm@linux-foundation.org> <47436BA0.5050809@quo.to> <47446690.5090808@redhat.com> In-Reply-To: <47446690.5090808@redhat.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Chuck Ebbert wrote: > On 11/20/2007 06:20 PM, Jordan Russell wrote: >> Same problem with 2.6.23.8. >> >> Are there any specific (TSC related?) patches I should try reverting? >> >> Would it help if I captured the dmesg/SysRq output from one of the >> hanging boots? >> >> Any other information that might be useful in getting to the bottom of this? >> > > Did you try this one? You are seeing problems with preemption disabled, > but it's at least worth trying. > > > > From: Marin Mitov > To: linux-kernel@vger.kernel.org > Subject: [PATCH]new_TSC_based_delay_tsc() > Cc: Ingo Molnar > Date: Tue, 20 Nov 2007 21:32:27 +0200 Thanks for the response. I backported that patch to 2.6.23.8, but it didn't make a difference. I also went ahead and tested 2.6.24-rc3 (with no changes to the existing config settings): With the new CPU_IDLE option set to "n", it still hangs, but much less frequently than on 2.6.23.x. In 25 tries, there were 3 hangs, all after "input: AT Translated Set 2 keyboard as /class/input/input0". With CPU_IDLE set to "y", it didn't hang at all in 25 tries. However, CPU_IDLE=y produces these additional messages, which may explain why: Marking TSC unstable due to: TSC halts in idle. ... Time: acpi_pm clocksource has been installed. Not sure what else to try at this point... -- Jordan Russell