From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1764858AbZAQQh3 (ORCPT ); Sat, 17 Jan 2009 11:37:29 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1756801AbZAQQhU (ORCPT ); Sat, 17 Jan 2009 11:37:20 -0500 Received: from mail.gmx.net ([213.165.64.20]:46816 "HELO mail.gmx.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1756228AbZAQQhT (ORCPT ); Sat, 17 Jan 2009 11:37:19 -0500 X-Authenticated: #14349625 X-Provags-ID: V01U2FsdGVkX1+4nWRMS7+41TDlAcMWNpMhws20qU4TAC7XARdgbu pjEpDR7JzzdGg7 Subject: Re: [git pull] scheduler fixes From: Mike Galbraith To: Ingo Molnar Cc: Avi Kivity , Kevin Shanahan , Andrew Morton , a.p.zijlstra@chello.nl, torvalds@linux-foundation.org, linux-kernel@vger.kernel.org In-Reply-To: <20090117162519.GD10825@elte.hu> References: <1232173776.7073.21.camel@marge.simson.net> <1232186054.6813.48.camel@marge.simson.net> <1232186877.14073.59.camel@laptop> <1232188484.6813.85.camel@marge.simson.net> <1232193617.14073.67.camel@laptop> <1232194752.6273.5.camel@marge.simson.net> <20090117044316.bda7d0bd.akpm@linux-foundation.org> <1232198574.16303.8.camel@marge.simson.net> <20090117160115.GA31601@elte.hu> <1232209281.5987.4.camel@marge.simson.net> <20090117162519.GD10825@elte.hu> Content-Type: text/plain Date: Sat, 17 Jan 2009 17:37:12 +0100 Message-Id: <1232210232.5987.20.camel@marge.simson.net> Mime-Version: 1.0 X-Mailer: Evolution 2.22.1.1 Content-Transfer-Encoding: 7bit X-Y-GMX-Trusted: 0 X-FuHaFi: 0.58 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Sat, 2009-01-17 at 17:25 +0100, Ingo Molnar wrote: > * Mike Galbraith wrote: > > > On Sat, 2009-01-17 at 17:01 +0100, Ingo Molnar wrote: > > > * Mike Galbraith wrote: > > > > > > > On Sat, 2009-01-17 at 04:43 -0800, Andrew Morton wrote: > > > > > http://bugzilla.kernel.org/show_bug.cgi?id=12465 just popped up - another > > > > > scheduler regression. It has been bisected. > > > > > > > > Seems pretty clear. I'd suggest reverting it. > > > > > > We can revert it (and will revert it if no solution is found), but i'd > > > also like to understand why it happens, because that kind of > > > regression from this change is unexpected - we might be hiding some > > > bug that could pop up under less debuggable circumstances, so we need > > > to understand it while we have a chance. > > > > Agree. However, with the sched_mc stuff, mysql+oltp now does better > > with NEWIDLE on than off as well, as does an nfs kbuild. > > Didnt you come up with the verdict that sched_mc=2 is not a win - or has > that changed? If we should change the defaults then please send a > re-tuning patch against the latest code. sched_mc=2 was better than sched_mc=1. The other balancing changes put a dent in mysql+oltp peak and immediately after peak. Setting sched_mc=2 brought back the loss that was otherwise there all the way through the curve back to 28 level, so with sched_mc=2, there was only the slight loss of peak, and larger loss of post-peak. > Still, the ping delays of multiple seconds are completely unacceptable and > need to be understood, or these will come back and might bite us in a lot > less fortunate place. NEWIDLE and WAKE_BALANCE has micro-effects on > latencies - anything in the user-visible range is highly anomalous at > these low load levels. > > Ingo