Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S938046AbdDSQbZ (ORCPT ); Wed, 19 Apr 2017 12:31:25 -0400 Received: from mail-qt0-f169.google.com ([209.85.216.169]:33339 "EHLO mail-qt0-f169.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S937930AbdDSQbW (ORCPT ); Wed, 19 Apr 2017 12:31:22 -0400 MIME-Version: 1.0 In-Reply-To: References: <1491815909-13345-1-git-send-email-vincent.guittot@linaro.org> <20170410173802.orygigjbcpefqtdv@hirez.programming.kicks-ass.net> <20170411075221.GA30421@linaro.org> <20170413133928.jmykwcq4qq5grktk@hirez.programming.kicks-ass.net> <20170413161307.q7g3wtlvh3c76zvp@hirez.programming.kicks-ass.net> From: Vincent Guittot Date: Wed, 19 Apr 2017 18:31:00 +0200 Message-ID: Subject: Re: [PATCH v2] sched/fair: update scale invariance of PELT To: Peter Zijlstra Cc: Ingo Molnar , linux-kernel , Dietmar Eggemann , Morten Rasmussen , Yuyang Du , Paul Turner , Ben Segall Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1917 Lines: 47 On 14 April 2017 at 10:49, Vincent Guittot wrote: > On 13 April 2017 at 18:13, Peter Zijlstra wrote: >> On Thu, Apr 13, 2017 at 05:16:20PM +0200, Vincent Guittot wrote: >>> On 13 April 2017 at 15:39, Peter Zijlstra wrote: >> >>> > OK, so the reason util_avg varies is because we compute it wrong. And I >>> > think we can easily fix that once we pull out all the factors (which >>> > would mean your patch and the pulling out of weight patch which still >>> > needs to be finished). >>> >>> That would be great to remove this unwanted variation. >> >> So the problem with the _avg stuff is that we include the d3 segment, >> that is the unfinished current window. Since we only re-compute the _avg >> whenever we roll over, the intent already seems to be to only compute it >> on completed windows. > > yes make sense > >> >> But because 'complicated/expensive', its hard to not include d3 and thus >> we get the wobble. >> >> Once we compute pure running/runnable sums, without extra contrib >> factors, we can simply subtract our d3 term from sum when doing the >> division and change the divider to LOAD_AVG_MAX*y, getting the stable >> _avg over all completed windows. > > I'm going to make it a try to check that it removes the variation i'm seeing I have sent a patchset based on your proposal that fix this variation issue > >> >> (you could do the same with factors, but then we get to do a bunch of >> extra multiplications which aren't free). >> >>> > >>> > But you're comparing against util_sum here, that behaves slightly >>> > different. I think you want 'util_sum >= 1024 * (LOAD_AVG_MAX - 1024)' >>> > instead. >>> >>> yes, the variation happens on the util_sum >> >> Well, for util_sum its simple to ignore the current window, which is >> what the suggested equation does (note that LOAD_AVG_MAX*y == >> LOAD_AVG_MAX-1024).