From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753311AbcHZSly (ORCPT ); Fri, 26 Aug 2016 14:41:54 -0400 Received: from mail-pa0-f50.google.com ([209.85.220.50]:33658 "EHLO mail-pa0-f50.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751449AbcHZSlt (ORCPT ); Fri, 26 Aug 2016 14:41:49 -0400 From: Steve Muckle X-Google-Original-From: Steve Muckle To: Peter Zijlstra , Ingo Molnar , "Rafael J . Wysocki" Cc: linux-kernel@vger.kernel.org, linux-pm@vger.kernel.org, Vincent Guittot , Morten Rasmussen , Dietmar Eggemann , Juri Lelli , Patrick Bellasi , Steve Muckle Subject: [PATCH 0/2] utilization changes for schedutil Date: Fri, 26 Aug 2016 11:40:46 -0700 Message-Id: <1472236848-17038-1-git-send-email-smuckle@linaro.org> X-Mailer: git-send-email 2.7.3 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org The first patch (which is a re-send) corrects an issue with utilization on SMT platforms. Currently CFS utilization is being roughly doubled on these platforms due to differences in the way PELT and overall max CPU capacity are calculated. The second patch moves away from the policy of going to fmax during RT task activity, instead using rt_avg as an estimate of RT utilization. Steve Muckle (2): sched: cpufreq: ignore SMT when determining max cpu capacity sched: cpufreq: use rt_avg as estimate of required RT CPU capacity kernel/sched/cpufreq_schedutil.c | 26 ++++++++++++++++++-------- 1 file changed, 18 insertions(+), 8 deletions(-) -- 2.7.3