From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1757923AbYELT3O (ORCPT ); Mon, 12 May 2008 15:29:14 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751635AbYELT27 (ORCPT ); Mon, 12 May 2008 15:28:59 -0400 Received: from ogre.sisk.pl ([217.79.144.158]:40894 "EHLO ogre.sisk.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751046AbYELT26 (ORCPT ); Mon, 12 May 2008 15:28:58 -0400 From: "Rafael J. Wysocki" To: Ingo Molnar Subject: Re: [Bug #10640] audio skipping on 2.6.26-rc1 Date: Mon, 12 May 2008 21:29:14 +0200 User-Agent: KMail/1.9.6 (enterprise 20070904.708012) Cc: Parag Warudkar , linux-kernel@vger.kernel.org, Miklos Szeredi References: <20080512075252.GB26074@elte.hu> In-Reply-To: <20080512075252.GB26074@elte.hu> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200805122129.15149.rjw@sisk.pl> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Monday, 12 of May 2008, Ingo Molnar wrote: > > * Parag Warudkar wrote: > > > Rafael J. Wysocki sisk.pl> writes: > > > > > Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=10640 > > > Subject : audio skipping on 2.6.26-rc1 > > > Submitter : Miklos Szeredi szeredi.hu> > > > Date : 2008-05-07 7:56 (5 days old) > > > References : http://marc.info/?l=linux-kernel&m=121014707707781&w=2 > > > Handled-By : Ingo Molnar elte.hu> > > > > This is a duplicate of [Bug #10616] Horrendous Audio Stutter - current > > git . > > that should be resolved via your patch which i pushed upstream (commit > v2.6.26-rc1-83-gaac6abc), which marks group scheduling experimental and > non-default. > > i think the audio skipping from Miklos, if it does not go away with > disabling group scheduling, might also be solved by: > > - commit v2.6.26-rc1-181-g4615112 ("sched: fix weight calculations") > from Mike. > > - or by commit v2.6.26-rc1-82-g3e51f33 from Peter ("sched: add optional > support for CONFIG_HAVE_UNSTABLE_SCHED_CLOCK") which makes the > scheduler clock more reliable [and hence produce no artificial > latencies] on NOHZ > > - or by commit v2.6.26-rc1-66-ga992241 from Peter ("sched: fix > normalized sleeper"). Miklos, can you please verify if the problem is present in 2.6.26-rc2? Rafael