From patchwork Tue Jun 3 10:13:47 2014 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Srivatsa S. Bhat" X-Patchwork-Id: 4286371 Return-Path: X-Original-To: patchwork-linux-pm@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork2.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.19.201]) by patchwork2.web.kernel.org (Postfix) with ESMTP id 1DA68BEEA7 for ; Tue, 3 Jun 2014 10:15:14 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id 28B3620253 for ; Tue, 3 Jun 2014 10:15:13 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 04B2F201C8 for ; Tue, 3 Jun 2014 10:15:12 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751748AbaFCKPK (ORCPT ); Tue, 3 Jun 2014 06:15:10 -0400 Received: from e23smtp09.au.ibm.com ([202.81.31.142]:44409 "EHLO e23smtp09.au.ibm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751511AbaFCKPJ (ORCPT ); Tue, 3 Jun 2014 06:15:09 -0400 Received: from /spool/local by e23smtp09.au.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Tue, 3 Jun 2014 20:15:07 +1000 Received: from d23dlp03.au.ibm.com (202.81.31.214) by e23smtp09.au.ibm.com (202.81.31.206) with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted; Tue, 3 Jun 2014 20:15:05 +1000 Received: from d23relay05.au.ibm.com (d23relay05.au.ibm.com [9.190.235.152]) by d23dlp03.au.ibm.com (Postfix) with ESMTP id 910AA3578047; Tue, 3 Jun 2014 20:15:04 +1000 (EST) Received: from d23av03.au.ibm.com (d23av03.au.ibm.com [9.190.234.97]) by d23relay05.au.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id s539rDQo45744334; Tue, 3 Jun 2014 19:53:13 +1000 Received: from d23av03.au.ibm.com (localhost [127.0.0.1]) by d23av03.au.ibm.com (8.14.4/8.14.4/NCO v10.0 AVout) with ESMTP id s53AF3JP023070; Tue, 3 Jun 2014 20:15:04 +1000 Received: from srivatsabhat.in.ibm.com (srivatsabhat.in.ibm.com [9.124.35.63] (may be forged)) by d23av03.au.ibm.com (8.14.4/8.14.4/NCO v10.0 AVin) with ESMTP id s53AF1V3022849; Tue, 3 Jun 2014 20:15:02 +1000 Message-ID: <538D9FDB.6070607@linux.vnet.ibm.com> Date: Tue, 03 Jun 2014 15:43:47 +0530 From: "Srivatsa S. Bhat" User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:15.0) Gecko/20120828 Thunderbird/15.0 MIME-Version: 1.0 To: Viresh Kumar CC: "Rafael J. Wysocki" , Vaidyanathan Srinivasan , "ego@linux.vnet.ibm.com" , "linux-pm@vger.kernel.org" , Linux Kernel Mailing List Subject: Re: [PATCH] cpufreq: governor: Be friendly towards latency-sensitive bursty workloads References: <20140526205337.1100.55275.stgit@srivatsabhat.in.ibm.com> <538D9631.9090500@linux.vnet.ibm.com> <538D9DBD.2030605@linux.vnet.ibm.com> In-Reply-To: X-TM-AS-MML: disable X-Content-Scanned: Fidelis XPS MAILER x-cbid: 14060310-3568-0000-0000-0000059E7670 Sender: linux-pm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-pm@vger.kernel.org X-Spam-Status: No, score=-7.5 required=5.0 tests=BAYES_00, RCVD_IN_DNSWL_HI, RP_MATCHES_RCVD, UNPARSEABLE_RELAY autolearn=unavailable version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP On 06/03/2014 03:38 PM, Viresh Kumar wrote: > On 3 June 2014 15:34, Srivatsa S. Bhat wrote: >> Well, the method I used keeps the organization such that the code following >> the comment does precisely what the comment says (i.e, get the sampling_rate, >> fetch the multiplier, and then multiply). So I feel it makes it easier to >> understand. > > It looked like the comment is there only for this special statement: > >>>> + sampling_rate *= od_dbs_info->rate_mult; > > And so suggested that :) > > Anyway move this up as it doesn't belong to comment for sure. >>> + od_dbs_info = dbs_data->cdata->get_cpu_dbs_info_s(cpu); > Fair enough :) Here it is: Acked-by: Viresh Kumar --- To unsubscribe from this list: send the line "unsubscribe linux-pm" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c index e1c6433..2597bbe 100644 --- a/drivers/cpufreq/cpufreq_governor.c +++ b/drivers/cpufreq/cpufreq_governor.c @@ -36,14 +36,29 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu) struct od_dbs_tuners *od_tuners = dbs_data->tuners; struct cs_dbs_tuners *cs_tuners = dbs_data->tuners; struct cpufreq_policy *policy; + unsigned int sampling_rate; unsigned int max_load = 0; unsigned int ignore_nice; unsigned int j; - if (dbs_data->cdata->governor == GOV_ONDEMAND) + if (dbs_data->cdata->governor == GOV_ONDEMAND) { + struct od_cpu_dbs_info_s *od_dbs_info = + dbs_data->cdata->get_cpu_dbs_info_s(cpu); + + /* + * Sometimes, the ondemand governor uses an additional + * multiplier to give long delays. So apply this multiplier to + * the 'sampling_rate', so as to keep the wake-up-from-idle + * detection logic a bit conservative. + */ + sampling_rate = od_tuners->sampling_rate; + sampling_rate *= od_dbs_info->rate_mult; + ignore_nice = od_tuners->ignore_nice_load; - else + } else { + sampling_rate = cs_tuners->sampling_rate; ignore_nice = cs_tuners->ignore_nice_load; + } policy = cdbs->cur_policy; @@ -96,7 +111,29 @@ void dbs_check_cpu(struct dbs_data *dbs_data, int cpu) if (unlikely(!wall_time || wall_time < idle_time)) continue; - load = 100 * (wall_time - idle_time) / wall_time; + /* + * If the CPU had gone completely idle, and a task just woke up + * on this CPU now, it would be unfair to calculate 'load' the + * usual way for this elapsed time-window, because it will show + * near-zero load, irrespective of how CPU intensive the new + * task is. This is undesirable for latency-sensitive bursty + * workloads. + * + * To avoid this, we reuse the 'load' from the previous + * time-window and give this task a chance to start with a + * reasonably high CPU frequency. + * + * Detecting this situation is easy: the governor's deferrable + * timer would not have fired during CPU-idle periods. Hence + * an unusually large 'wall_time' (as compared to the sampling + * rate) indicates this scenario. + */ + if (unlikely(wall_time > (2 * sampling_rate))) { + load = j_cdbs->prev_load; + } else { + load = 100 * (wall_time - idle_time) / wall_time; + j_cdbs->prev_load = load; + } if (load > max_load) max_load = load; @@ -323,6 +360,10 @@ int cpufreq_governor_dbs(struct cpufreq_policy *policy, j_cdbs->cur_policy = policy; j_cdbs->prev_cpu_idle = get_cpu_idle_time(j, &j_cdbs->prev_cpu_wall, io_busy); + j_cdbs->prev_load = 100 * (j_cdbs->prev_cpu_wall - + j_cdbs->prev_cpu_idle) / + j_cdbs->prev_cpu_wall; + if (ignore_nice) j_cdbs->prev_cpu_nice = kcpustat_cpu(j).cpustat[CPUTIME_NICE]; diff --git a/drivers/cpufreq/cpufreq_governor.h b/drivers/cpufreq/cpufreq_governor.h index bfb9ae1..b56552b 100644 --- a/drivers/cpufreq/cpufreq_governor.h +++ b/drivers/cpufreq/cpufreq_governor.h @@ -134,6 +134,7 @@ struct cpu_dbs_common_info { u64 prev_cpu_idle; u64 prev_cpu_wall; u64 prev_cpu_nice; + unsigned int prev_load; struct cpufreq_policy *cur_policy; struct delayed_work work; /*