From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758536Ab0LCIhy (ORCPT ); Fri, 3 Dec 2010 03:37:54 -0500 Received: from smtp-out.google.com ([216.239.44.51]:24687 "EHLO smtp-out.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754816Ab0LCIhw convert rfc822-to-8bit (ORCPT ); Fri, 3 Dec 2010 03:37:52 -0500 DomainKey-Signature: a=rsa-sha1; c=nofws; d=google.com; s=beta; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-type:content-transfer-encoding; b=X9gR3BsTuFaSAK54Bo/HlxflT86h8UOtffSAwMyR5yUAXM/vJPg8sHet7GL3eltcXR JKr88PdFuAQoq8D/1V5A== MIME-Version: 1.0 In-Reply-To: <1291358909.4378.17.camel@marge.simson.net> References: <20101121133744.GA10765@elte.hu> <1290700829.4759.16.camel@maggy.simson.net> <1290954299.30515.15.camel@marge.simson.net> <4CF5C379.8030204@google.com> <1291184173.7466.147.camel@marge.simson.net> <4CF87C14.8000708@google.com> <1291358909.4378.17.camel@marge.simson.net> From: Paul Turner Date: Fri, 3 Dec 2010 00:37:19 -0800 Message-ID: Subject: Re: [PATCH v4] sched: automated per session task groups To: Mike Galbraith Cc: Ingo Molnar , Oleg Nesterov , Peter Zijlstra , Linus Torvalds , LKML Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8BIT X-System-Of-Record: true Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Dec 2, 2010 at 10:48 PM, Mike Galbraith wrote: > On Thu, 2010-12-02 at 21:11 -0800, Paul Turner wrote: >> On 11/30/10 22:16, Mike Galbraith wrote: >> > On Tue, 2010-11-30 at 19:39 -0800, Paul Turner wrote: >> >> On 11/28/10 06:24, Mike Galbraith wrote: >> >>> >> >>> Something else is seriously wrong though.  36.1 with attached (plus >> >>> sched, cgroup: Fixup broken cgroup movement) works a treat, whereas >> >>> 37.git and tip with fixlet below both suck rocks.  With a make -j40 >> >>> running, wakeup-latency is showing latencies of>100ms, amarok skips, >> >>> mouse lurches badly.. generally horrid.  Something went south. >> >> >> >> I'm looking at this. >> >> >> >> The share:share ratios looked good in static testing, but perhaps we >> >> need a little more wake-up boost to improve interactivity. >> > >> > Yeah, feels like a wakeup issue.  I too did a (brief) static test, and >> > that looked ok. >> > >> >     -Mike >> > >> >> Hey Mike, >> >> Does something like the below help? > > Unfortunately not.  For example, Xorg+mplayer needs (30 sec refresh).. > >  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  P COMMAND >  6487 root      20   0  366m  30m 5100 S   31  0.4   2:04.83 2 Xorg >  4454 root      20   0  318m  28m  15m S   29  0.4   0:38.06 3 mplayer > > ..but gets this when a heavy kbuild is running along with them. > >  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  P COMMAND >  6487 root      20   0  366m  30m 5136 S   12  0.4   2:25.98 1 Xorg >  5595 root      20   0  318m  28m  15m R    8  0.4   0:09.31 3 mplayer > > There are 4 task groups active at this time, Xorg, mplayer, Amarok and > konsole where the kbuild is running make -j40. > Hmm.. unfortunate. Ok -- based on the traces of synthetic loads and the traces of their share on wake-up I think this is the right track at least, will refine it tomorrow. Thanks for trying it. >        -Mike > > >