From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933617AbXDGRo5 (ORCPT ); Sat, 7 Apr 2007 13:44:57 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S966194AbXDGRo5 (ORCPT ); Sat, 7 Apr 2007 13:44:57 -0400 Received: from gateway-1237.mvista.com ([63.81.120.158]:49431 "EHLO gateway-1237.mvista.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933617AbXDGRo5 (ORCPT ); Sat, 7 Apr 2007 13:44:57 -0400 Subject: Re: [PATCH] timekeeping: drop irq-context clocksource polling From: Daniel Walker To: Andrew Morton Cc: linux-kernel@vger.kernel.org, johnstul@us.ibm.com, mingo@elte.hu In-Reply-To: <20070407031922.ee49478f.akpm@linux-foundation.org> References: <20070405210316.785839431@mvista.com> <20070407031922.ee49478f.akpm@linux-foundation.org> Content-Type: text/plain Date: Sat, 07 Apr 2007 10:43:59 -0700 Message-Id: <1175967839.15973.156.camel@imap.mvista.com> Mime-Version: 1.0 X-Mailer: Evolution 2.8.2.1 (2.8.2.1-3.fc6) Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Sat, 2007-04-07 at 03:19 -0700, Andrew Morton wrote: > On Thu, 05 Apr 2007 14:03:16 -0700 Daniel Walker wrote: > > > Before this change the timekeeping code would poll the clocksource > > list every interrupt. This changes that so the clocksource list is > > only checked when there has been and update, and no longer checks > > in interrupt context. > > I get a complete lockup on i386 SMP - before the kernel has printed anything. > > I'm suspecting a recursive taking of xtime_lock. Looks like this path , arch/i386/kernel/tsc.c: time_cpufreq_notifier(); <-- takes xtime_lock mark_tsc_unstable(); clocksource_change_rating(&clocksource_tsc, 0); timekeeping_change_clocksource(); <-- takes xtime_lock I'm not sure why the time_cpufreq_notifier is taking the xtime_lock tho . Daniel