From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753513AbaJJJOq (ORCPT ); Fri, 10 Oct 2014 05:14:46 -0400 Received: from casper.infradead.org ([85.118.1.10]:45163 "EHLO casper.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752586AbaJJJOo (ORCPT ); Fri, 10 Oct 2014 05:14:44 -0400 Date: Fri, 10 Oct 2014 11:14:41 +0200 From: Peter Zijlstra To: Yuyang Du Cc: Morten Rasmussen , mingo@redhat.com, dietmar.eggemann@arm.com, pjt@google.com, bsegall@google.com, vincent.guittot@linaro.org, nicolas.pitre@linaro.org, mturquette@linaro.org, rjw@rjwysocki.net, linux-kernel@vger.kernel.org Subject: Re: [PATCH 1/7] sched: Introduce scale-invariant load tracking Message-ID: <20141010091441.GF10832@worktop.programming.kicks-ass.net> References: <1411403047-32010-1-git-send-email-morten.rasmussen@arm.com> <1411403047-32010-2-git-send-email-morten.rasmussen@arm.com> <20141008005007.GA7017@intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20141008005007.GA7017@intel.com> User-Agent: Mutt/1.5.22.1 (2013-10-16) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, Oct 08, 2014 at 08:50:07AM +0800, Yuyang Du wrote: > Hi Morten, > > Sorry for late jumping in. > > The problem seems to be self-evident. But for the implementation to be > equally attractive it needs to account for every freq change for every task, > or anything less than that makes it less attractive. I'm not entirely sure that is indeed required. > But this should be very hard. Intel Architecture has limitation to capture all > the freq changes in software and also the intel_pstate should have no > notification. So current Intel arch takes P-states as hints and then can mostly lower their actual frequency, right? In this case still accounting at the higher frequency is not a problem, the hardware conserves more power than we know, but that's the right kind of error to have :-) For the full automatic hardware, that's basically hardware without DVFS capability so we should not bother at all and simply disable this scaling. > For every task, this makes the updating of the entire queue in load tracking > more needed, so once again, ping maintainers for the rewrite patches, :) Could you remind me what your latest version is; please reply to those patches with a new email so that I can more conveniently locate it.