From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Rafael J. Wysocki" Subject: Re: [patch 3/3] clockevents: Fix resume logic - updated version Date: Wed, 9 May 2007 20:55:44 +0200 Message-ID: <200705092055.45688.rjw@sisk.pl> References: <20070430102837.748238000@linutronix.de> <200705091452.08092.rjw@sisk.pl> <20070509101445.fa6b60cb.akpm@linux-foundation.org> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: Received: from ogre.sisk.pl ([217.79.144.158]:50685 "EHLO ogre.sisk.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757118AbXEISvD (ORCPT ); Wed, 9 May 2007 14:51:03 -0400 In-Reply-To: <20070509101445.fa6b60cb.akpm@linux-foundation.org> Content-Disposition: inline Sender: linux-acpi-owner@vger.kernel.org List-Id: linux-acpi@vger.kernel.org To: Andrew Morton Cc: tglx@linutronix.de, Ingo Molnar , LKML , John Stultz , linux-acpi@vger.kernel.org On Wednesday, 9 May 2007 19:14, Andrew Morton wrote: > On Wed, 9 May 2007 14:52:07 +0200 "Rafael J. Wysocki" wrote: > > > > Andrew, could you please apply the appended patch and see if that > > > helps (should apply to -mm2)? > > > > Argh, sorry. This needs yet another patch (sent for review to linux-pm) to > > be applied. The following one is against -mm2: > > > > --- > > NOTE: This is not a complete solution, because it removes the enabling of GPEs > > from the resume-during-hibernation code path entirely, which probbably is not a > > good idea in general. > > --- > > kernel/power/disk.c | 1 - > > 1 file changed, 1 deletion(-) > > > > Index: linux-2.6.21-mm2/kernel/power/disk.c > > =================================================================== > > --- linux-2.6.21-mm2.orig/kernel/power/disk.c > > +++ linux-2.6.21-mm2/kernel/power/disk.c > > @@ -205,7 +205,6 @@ int hibernate(void) > > > > if (in_suspend) { > > enable_nonboot_cpus(); > > - platform_finish(); > > device_resume(); > > resume_console(); > > pr_debug("PM: writing image.\n"); > > It now hangs in a similar fashion in the device_resume() call. > > If I back off Thomas's clockevents-fix-resume-logic-updated-version.patch > and include just the above patch the machine does suspend and resume > correctly. > > I can delve further into the device_resume() hang if we think it would be > useful. Frankly, I'm not sure. On the one hand, I'd like to know wth is going on in there, but on the other hand, I don't know if that knowledge will help us in general. It looks like we'll need to revamp the ACPI stuff in the hibernation/suspend code paths anyway. > Maybe Linus's clock-stomping tracer can help here, if I can remember how to > use it. It's described a bit in Documentation/power/s2ram.txt .