From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754105AbbLDAsu (ORCPT ); Thu, 3 Dec 2015 19:48:50 -0500 Received: from v094114.home.net.pl ([79.96.170.134]:61828 "HELO v094114.home.net.pl" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1751944AbbLDAst (ORCPT ); Thu, 3 Dec 2015 19:48:49 -0500 From: "Rafael J. Wysocki" To: Viresh Kumar Cc: linaro-kernel@lists.linaro.org, linux-pm@vger.kernel.org, ashwin.chaugule@linaro.org, "Rafael J. Wysocki" , open list Subject: Re: [PATCH V2 5/6] cpufreq: governor: replace per-cpu delayed work with timers Date: Fri, 04 Dec 2015 02:18:41 +0100 Message-ID: <10439879.00aCyM9quW@vostro.rjw.lan> User-Agent: KMail/4.11.5 (Linux/4.1.0-rc5+; KDE/4.11.5; x86_64; ; ) In-Reply-To: <691a3524afaa8d08d61987b1c2913948b4e59f01.1449115453.git.viresh.kumar@linaro.org> References: <691a3524afaa8d08d61987b1c2913948b4e59f01.1449115453.git.viresh.kumar@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: 7Bit Content-Type: text/plain; charset="utf-8" Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thursday, December 03, 2015 09:37:53 AM Viresh Kumar wrote: > cpufreq governors evaluate load at sampling rate and based on that they > update frequency for a group of CPUs belonging to the same cpufreq > policy. > > This is required to be done in a single thread for all policy->cpus, but > because we don't want to wakeup idle CPUs to do just that, we use > deferrable work for this. If we would have used a single delayed > deferrable work for the entire policy, there were chances that the CPU > required to run the handler can be in idle and we might end up not > changing the frequency for the entire group with load variations. > > And so we were forced to keep per-cpu works, and only the one that > expires first need to do the real work and others are rescheduled for > next sampling time. > > We have been using the more complex solution until now, where we used a > delayed deferrable work for this, which is a combination of a timer and > a work. > > This could be made lightweight by keeping per-cpu deferred timers with a > single work item, which is scheduled by the first timer that expires. > > This patch does just that and here are important changes: > - The timer handler will run in irq context and so we need to use a > spin_lock instead of the timer_mutex. And so a separate timer_lock is > created. This also makes the use of the mutex and lock quite clear, as > we know what exactly they are protecting. > - A new field 'skip_work' is added to track when the timer handlers can > queue a work. More comments present in code. > > Suggested-by: Rafael J. Wysocki > Signed-off-by: Viresh Kumar > Reviewed-by: Ashwin Chaugule I've tentatively queued this one up, but I still have a couple of questions. > --- > drivers/cpufreq/cpufreq_governor.c | 139 +++++++++++++++++++++---------------- > drivers/cpufreq/cpufreq_governor.h | 20 ++++-- > drivers/cpufreq/cpufreq_ondemand.c | 8 +-- > 3 files changed, 98 insertions(+), 69 deletions(-) [cut] > @@ -250,14 +247,44 @@ static void dbs_timer(struct work_struct *work) > sampling_rate = od_tuners->sampling_rate; > } > > - if (!need_load_eval(cdbs->shared, sampling_rate)) > - modify_all = false; > + eval_load = need_load_eval(shared, sampling_rate); > > - delay = dbs_data->cdata->gov_dbs_timer(policy, modify_all); > - gov_queue_work(dbs_data, policy, delay, modify_all); > + /* > + * Make sure cpufreq_governor_limits() isn't evaluating load in > + * parallel. > + */ > + mutex_lock(&shared->timer_mutex); > + delay = dbs_data->cdata->gov_dbs_timer(policy, eval_load); > + mutex_unlock(&shared->timer_mutex); > + > + shared->skip_work--; Is there any reason for incrementing and decrementing this instead of setting it to either 0 or 1 (or maybe either 'true' or 'false' for that matter)? If my reading of the patch is correct, it can only be either 0 or 1 anyway, right? > + gov_add_timers(policy, delay); > +} > + > +static void dbs_timer_handler(unsigned long data) > +{ > + struct cpu_dbs_info *cdbs = (struct cpu_dbs_info *)data; > + struct cpu_common_dbs_info *shared = cdbs->shared; > + struct cpufreq_policy *policy; > + unsigned long flags; > + > + spin_lock_irqsave(&shared->timer_lock, flags); > + policy = shared->policy; Why do we need policy here? > + > + /* > + * Timer handler isn't allowed to queue work at the moment, because: > + * - Another timer handler has done that > + * - We are stopping the governor > + * - Or we are updating the sampling rate of ondemand governor > + */ > + if (shared->skip_work) > + goto unlock; > + > + shared->skip_work++; > + queue_work(system_wq, &shared->work); > > unlock: What about writing the above as if (!shared->work_in_progress) { shared->work_in_progress = true; queue_work(system_wq, &shared->work); } and then you won't need the unlock label. > - mutex_unlock(&shared->timer_mutex); > + spin_unlock_irqrestore(&shared->timer_lock, flags); > } > > static void set_sampling_rate(struct dbs_data *dbs_data, Thanks, Rafael