From patchwork Fri Nov 11 10:22:22 2016 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Viresh Kumar X-Patchwork-Id: 81811 Delivered-To: patch@linaro.org Received: by 10.182.1.168 with SMTP id 8csp1382990obn; Fri, 11 Nov 2016 02:22:58 -0800 (PST) X-Received: by 10.99.218.85 with SMTP id l21mr3631850pgj.102.1478859778535; Fri, 11 Nov 2016 02:22:58 -0800 (PST) Return-Path: Received: from vger.kernel.org (vger.kernel.org. [209.132.180.67]) by mx.google.com with ESMTP id fc7si2712159pab.324.2016.11.11.02.22.58; Fri, 11 Nov 2016 02:22:58 -0800 (PST) Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67; Authentication-Results: mx.google.com; dkim=pass header.i=@linaro.org; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE dis=NONE) header.from=linaro.org Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932331AbcKKKWo (ORCPT + 27 others); Fri, 11 Nov 2016 05:22:44 -0500 Received: from mail-pf0-f176.google.com ([209.85.192.176]:35015 "EHLO mail-pf0-f176.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755623AbcKKKWk (ORCPT ); Fri, 11 Nov 2016 05:22:40 -0500 Received: by mail-pf0-f176.google.com with SMTP id i88so9998339pfk.2 for ; Fri, 11 Nov 2016 02:22:39 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linaro.org; s=google; h=from:to:cc:subject:date:message-id:in-reply-to:references :in-reply-to:references; bh=/WEaDxDJ1TcDow2WXmGAS/Iyj0j+/CT2WWFH7ppSOIQ=; b=ixDz7k3ijNfAKGsIofHr73ZTG2zQNiMlvoM4VfXlnDhRAjdcuNQ99OaOm6OWaJ/qaG Awr3LesqNr/HfsQQoqjPmcrkEaZxZB2INbSFpNu8oIQ/CCyCKkdJAeEtw6GjmkaX3iRp pSxyBttHqM7ifeefYmJg5g/Ukw00IkelOR/Xo= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:in-reply-to:references; bh=/WEaDxDJ1TcDow2WXmGAS/Iyj0j+/CT2WWFH7ppSOIQ=; b=IDmlMUk8+Lya1DakbBXt1eV0VnsP1qVTq6ugGiuvSwvqHUBBuBH2tDh2rWhX6t8yP+ CxB607489bJPBa62oeThBoickwKWDDLtSd0zT3NLrA4aS/YSY3GiboAYdL246s83gM3K GTexwLRdcUYnX22LxzRyYrpOanr7xOUryqA+JhNCgVZvweKt9/+8UpLhdhLbjfiAZ7jl Puor29R0VOOBTpHm8VnrOQ5PuTfwYK16CW1Zp/yHEc69cyJdwLIiFG9Z87sPXeFNdcex gihaF6046o9E4YLWYbhaB4iuSsvEbSaZfv0Tm0vi5zjoULiKhUovYYgda8+V8d5ZmtK3 x0Lw== X-Gm-Message-State: ABUngvelhs0kCLBkJfghxd4Dgh2qpbiLnw5bBqAnM3ge1ZmWENy9Csgh0KtBvvDQ5ERqeJSx X-Received: by 10.98.57.144 with SMTP id u16mr5510210pfj.142.1478859759406; Fri, 11 Nov 2016 02:22:39 -0800 (PST) Received: from localhost ([122.172.89.192]) by smtp.gmail.com with ESMTPSA id dc3sm13879913pad.32.2016.11.11.02.22.38 (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Fri, 11 Nov 2016 02:22:38 -0800 (PST) From: Viresh Kumar To: Rafael Wysocki , Ingo Molnar , Peter Zijlstra Cc: linaro-kernel@lists.linaro.org, linux-pm@vger.kernel.org, linux-kernel@vger.kernel.org, Vincent Guittot , Juri Lelli , Robin Randhawa , Viresh Kumar , Steve Muckle Subject: [PATCH 2/3] cpufreq: schedutil: move slow path from workqueue to SCHED_FIFO task Date: Fri, 11 Nov 2016 15:52:22 +0530 Message-Id: <85bf45982709e06f7f42e1b8f8315945e9d9b6d0.1478858983.git.viresh.kumar@linaro.org> X-Mailer: git-send-email 2.7.1.410.g6faf27b In-Reply-To: References: In-Reply-To: References: Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org If slow path frequency changes are conducted in a SCHED_OTHER context then they may be delayed for some amount of time, including indefinitely, when real time or deadline activity is taking place. Move the slow path to a real time kernel thread. In the future the thread should be made SCHED_DEADLINE. The RT priority is arbitrarily set to 50 for now. Hackbench results on ARM Exynos, dual core A15 platform for 10 iterations: $ hackbench -s 100 -l 100 -g 10 -f 20 Before After --------------------------------- 1.808 1.603 1.847 1.251 2.229 1.590 1.952 1.600 1.947 1.257 1.925 1.627 2.694 1.620 1.258 1.621 1.919 1.632 1.250 1.240 Average: 1.8829 1.5041 Based on initial work by Steve Muckle. Signed-off-by: Steve Muckle Signed-off-by: Viresh Kumar --- kernel/sched/cpufreq_schedutil.c | 62 ++++++++++++++++++++++++++++++++-------- 1 file changed, 50 insertions(+), 12 deletions(-) -- 2.7.1.410.g6faf27b diff --git a/kernel/sched/cpufreq_schedutil.c b/kernel/sched/cpufreq_schedutil.c index ccb2ab89affb..045ce0a4e6d1 100644 --- a/kernel/sched/cpufreq_schedutil.c +++ b/kernel/sched/cpufreq_schedutil.c @@ -12,6 +12,7 @@ #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt #include +#include #include #include @@ -35,8 +36,10 @@ struct sugov_policy { /* The next fields are only needed if fast switch cannot be used. */ struct irq_work irq_work; - struct work_struct work; + struct kthread_work work; struct mutex work_lock; + struct kthread_worker worker; + struct task_struct *thread; bool work_in_progress; bool need_freq_update; @@ -291,9 +294,10 @@ static void sugov_update_shared(struct update_util_data *hook, u64 time, raw_spin_unlock(&sg_policy->update_lock); } -static void sugov_work(struct work_struct *work) +static void sugov_work(struct kthread_work *work) { - struct sugov_policy *sg_policy = container_of(work, struct sugov_policy, work); + struct sugov_policy *sg_policy = + container_of(work, struct sugov_policy, work); mutex_lock(&sg_policy->work_lock); __cpufreq_driver_target(sg_policy->policy, sg_policy->next_freq, @@ -308,7 +312,7 @@ static void sugov_irq_work(struct irq_work *irq_work) struct sugov_policy *sg_policy; sg_policy = container_of(irq_work, struct sugov_policy, irq_work); - schedule_work_on(smp_processor_id(), &sg_policy->work); + kthread_queue_work(&sg_policy->worker, &sg_policy->work); } /************************** sysfs interface ************************/ @@ -362,9 +366,23 @@ static struct kobj_type sugov_tunables_ktype = { static struct cpufreq_governor schedutil_gov; +static void sugov_policy_free(struct sugov_policy *sg_policy) +{ + if (!sg_policy->policy->fast_switch_enabled) { + kthread_flush_worker(&sg_policy->worker); + kthread_stop(sg_policy->thread); + } + + mutex_destroy(&sg_policy->work_lock); + kfree(sg_policy); +} + static struct sugov_policy *sugov_policy_alloc(struct cpufreq_policy *policy) { struct sugov_policy *sg_policy; + struct task_struct *thread; + struct sched_param param = { .sched_priority = 50 }; + int ret; sg_policy = kzalloc(sizeof(*sg_policy), GFP_KERNEL); if (!sg_policy) @@ -372,16 +390,36 @@ static struct sugov_policy *sugov_policy_alloc(struct cpufreq_policy *policy) sg_policy->policy = policy; init_irq_work(&sg_policy->irq_work, sugov_irq_work); - INIT_WORK(&sg_policy->work, sugov_work); mutex_init(&sg_policy->work_lock); raw_spin_lock_init(&sg_policy->update_lock); - return sg_policy; -} -static void sugov_policy_free(struct sugov_policy *sg_policy) -{ - mutex_destroy(&sg_policy->work_lock); - kfree(sg_policy); + /* kthread only required for slow path */ + if (policy->fast_switch_enabled) + return sg_policy; + + kthread_init_work(&sg_policy->work, sugov_work); + kthread_init_worker(&sg_policy->worker); + thread = kthread_create(kthread_worker_fn, &sg_policy->worker, + "sugov:%d", cpumask_first(policy->related_cpus)); + if (IS_ERR(thread)) { + mutex_destroy(&sg_policy->work_lock); + kfree(sg_policy); + pr_err("failed to create sugov thread: %ld\n", PTR_ERR(thread)); + return NULL; + } + sg_policy->thread = thread; + + ret = sched_setscheduler_nocheck(thread, SCHED_FIFO, ¶m); + if (ret) { + sugov_policy_free(sg_policy); + pr_warn("%s: failed to set SCHED_FIFO\n", __func__); + return NULL; + } + + kthread_bind_mask(thread, policy->related_cpus); + wake_up_process(thread); + + return sg_policy; } static struct sugov_tunables *sugov_tunables_alloc(struct sugov_policy *sg_policy) @@ -541,7 +579,7 @@ static void sugov_stop(struct cpufreq_policy *policy) synchronize_sched(); irq_work_sync(&sg_policy->irq_work); - cancel_work_sync(&sg_policy->work); + kthread_cancel_work_sync(&sg_policy->work); } static void sugov_limits(struct cpufreq_policy *policy)