From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754855Ab1AaKOZ (ORCPT ); Mon, 31 Jan 2011 05:14:25 -0500 Received: from ogre.sisk.pl ([217.79.144.158]:51449 "EHLO ogre.sisk.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752683Ab1AaKOW convert rfc822-to-8bit (ORCPT ); Mon, 31 Jan 2011 05:14:22 -0500 From: "Rafael J. Wysocki" To: Ozan =?utf-8?q?=C3=87a=C4=9Flayan?= Subject: Re: Common boot/shutdown issues with the latest 2.6.37 (Mostly Asus laptops) Date: Mon, 31 Jan 2011 11:14:11 +0100 User-Agent: KMail/1.13.5 (Linux/2.6.38-rc2+; KDE/4.4.4; x86_64; ; ) Cc: Lionel Debroux , linux-acpi@vger.kernel.org, linux-kernel@vger.kernel.org, platform-driver-x86@vger.kernel.org, colin.king@canonical.com References: <556624.21214.qm@web28409.mail.ukl.yahoo.com> <201101311050.49722.rjw@sisk.pl> <4D4688DD.6030502@pardus.org.tr> In-Reply-To: <4D4688DD.6030502@pardus.org.tr> MIME-Version: 1.0 Content-Type: Text/Plain; charset="utf-8" Content-Transfer-Encoding: 8BIT Message-Id: <201101311114.11935.rjw@sisk.pl> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Monday, January 31, 2011, Ozan Çağlayan wrote: > On 31.01.2011 11:50, Rafael J. Wysocki wrote: > > On Monday, January 31, 2011, Ozan Çağlayan wrote: > >> On 31.01.2011 00:48, Rafael J. Wysocki wrote: > >> > >>> Does booting with "nolapic" help on all of the affected systems? > >>> > >>> Also, does disabling the CPUidle during boot help? > >>> > >>> Finally, is the problem reproducible with CONFIG_NO_HZ unset? > >> > >> Well on two machines, nolapic cures the boot hang and on one machine it > >> cures the random shutdown hangs. > >> > >> I'll compile a kernel with CONFIG_NO_HZ to see what happens but how do I > >> disable cpuidle during boot? I've checked the kernel-parameters.txt but > >> there's no cpuidle related boot parameter there. > > > > Booting with intel_idle.max_cstate=0 should do the trick. > > Okay but on Asus X61S idle fallbacks to cpuidle: > > intel_idle: MWAIT substates: 0x3122220 > intel_idle: does not run on family 6 model 23 > > so forcing max_cstate to 0 will turn off the cpuidle on those machines too? In those cases you can use processor.max_cstate=0 I think. Rafael