From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261959AbVAHJxX (ORCPT ); Sat, 8 Jan 2005 04:53:23 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261825AbVAHJvw (ORCPT ); Sat, 8 Jan 2005 04:51:52 -0500 Received: from grendel.digitalservice.pl ([217.67.200.140]:65215 "HELO mail.digitalservice.pl") by vger.kernel.org with SMTP id S261858AbVAHJse (ORCPT ); Sat, 8 Jan 2005 04:48:34 -0500 From: "Rafael J. Wysocki" To: ncunningham@linuxmail.org, Pavel Machek Subject: Re: 2.6.10-mm2: swsusp regression [update] Date: Sat, 8 Jan 2005 10:49:02 +0100 User-Agent: KMail/1.7.1 Cc: Linux Kernel Mailing List References: <20050106002240.00ac4611.akpm@osdl.org> <1105135940.2488.39.camel@desktop.cunninghams> <200501080156.06145.rjw@sisk.pl> In-Reply-To: <200501080156.06145.rjw@sisk.pl> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-2" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200501081049.02862.rjw@sisk.pl> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Saturday, 8 of January 2005 01:56, Rafael J. Wysocki wrote: > On Friday, 7 of January 2005 23:12, Nigel Cunningham wrote: > > Hi. > > > > On Fri, 2005-01-07 at 23:45, Rafael J. Wysocki wrote: > > > > ..so... could you go through sysdev_register()s, one by one, > > > > commenting them to see which one causes the regression? That driver > > > > then needs to be fixed. > > > > > > > > Go after mtrr and time in first places. > > > > > > OK, but it'll take some time. > > > > There's an MTRR fix in the -overloaded ck patches. Perhaps it is what > > you're after. (Or perhaps it's already included :>) > > > > http://kem.p.lodz.pl/~peter/cko/fixes/2.6.10-cko1-swsusp_fix.patch > > Thanks for pointing it out. I have adapted this patch to -mm2, but > unfortunately it does not fix the issue. Still searching. ;-) The regression is caused by the timer driver. Obviously, turning timer_resume() in arch/x86_64/kernel/time.c into a NOOP makes it go away. It looks like a locking problem to me. I'll try to find a fix, although someone who knows more about these things would probably do it faster. :-) Greets, RJW -- - Would you tell me, please, which way I ought to go from here? - That depends a good deal on where you want to get to. -- Lewis Carroll "Alice's Adventures in Wonderland"