From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Doug Smythies" Subject: RE: Ask for help on governor Date: Wed, 13 Dec 2017 08:49:46 -0800 Message-ID: <000801d37432$637cdab0$2a769010$@net> References: <000801d37364$d48f6ed0$7dae4c70$@net> <000f01d373bf$deacca10$9c065e30$@net> P0Mmegy2kTqmMP0MoeccOV P9fLe8wpe7WleP9fMeeBW3 Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from cmta19.telus.net ([209.171.16.92]:42741 "EHLO cmta19.telus.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753164AbdLMQtu (ORCPT ); Wed, 13 Dec 2017 11:49:50 -0500 In-Reply-To: P9fLe8wpe7WleP9fMeeBW3 Content-Language: en-ca Sender: linux-pm-owner@vger.kernel.org List-Id: linux-pm@vger.kernel.org To: 'Viresh Kumar' Cc: "'Rafael J. Wysocki'" , 'Andy Tang' , "'Rafael J. Wysocki'" , 'Linux PM' , 'Doug Smythies' Hi Viresh, In addition to what I wrote earlier: I can also make the conservative governor work under the intel_cpufreq scaling driver, with the default 500 uSec sampling time, by increasing the /sys/devices/system/cpu/cpufreq/conservative/freq_step value from the default of 5 to 100. In this case there is not a specific break point freq_step, but rather a very gradual increase in CPU freq as a function of freq_step for a 100% load on a CPU. Now, I wonder if the two issues here might be related after all. ... Doug