Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933547AbcK2Ou5 (ORCPT ); Tue, 29 Nov 2016 09:50:57 -0500 Received: from foss.arm.com ([217.140.101.70]:48722 "EHLO foss.arm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S933539AbcK2Ouq (ORCPT ); Tue, 29 Nov 2016 09:50:46 -0500 Date: Tue, 29 Nov 2016 14:50:42 +0000 From: Morten Rasmussen To: Vincent Guittot Cc: Peter Zijlstra , Ingo Molnar , linux-kernel , Matt Fleming , Dietmar Eggemann , Wanpeng Li , Yuyang Du , Mike Galbraith Subject: Re: [PATCH 1/2 v2] sched: fix find_idlest_group for fork Message-ID: <20161129145041.GC1716@e105550-lin.cambridge.arm.com> References: <1480088073-11642-1-git-send-email-vincent.guittot@linaro.org> <1480088073-11642-2-git-send-email-vincent.guittot@linaro.org> <20161129105758.GA1716@e105550-lin.cambridge.arm.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.24 (2015-08-30) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1539 Lines: 42 On Tue, Nov 29, 2016 at 02:04:27PM +0100, Vincent Guittot wrote: > On 29 November 2016 at 11:57, Morten Rasmussen wrote: > > On Fri, Nov 25, 2016 at 04:34:32PM +0100, Vincent Guittot wrote: > >> @@ -5708,13 +5708,6 @@ static int select_idle_cpu(struct task_struct *p, struct sched_domain *sd, int t > >> > >> avg_cost = this_sd->avg_scan_cost; > >> > >> - /* > >> - * Due to large variance we need a large fuzz factor; hackbench in > >> - * particularly is sensitive here. > >> - */ > >> - if ((avg_idle / 512) < avg_cost) > >> - return -1; > >> - > >> time = local_clock(); > >> > >> for_each_cpu_wrap(cpu, sched_domain_span(sd), target, wrap) { > > > > I don't quite get this fix, but it is very likely because I haven't paid > > enough attention. > > > > Are you saying that removing the avg_cost check is improving hackbench > > performance? I thought it was supposed to help hackbench? I'm confused > > :-( > > Yes, avg_cost check prevents some tasks migration at the end of the > tests when some threads have already finished their loop letting some > CPUs idle whereas others threads are still competing on the same CPUS Okay, thanks. > > Should we do the same for SD_BALANCE_EXEC? > > I asked myself if i should add SD_BALANCE_EXEC but decided to only > keep SD_BALANCE_FORK for now as no regression has been raised for now. Fair enough. FWIW, with the label renaming suggested by mfleming, you can add my reviewed/acked-by if you like. Morten